Citation
Genetic and evolutionary characterization of plant pathogenic xanthomonads based on DNA sequences related to the hrp genes

Material Information

Title:
Genetic and evolutionary characterization of plant pathogenic xanthomonads based on DNA sequences related to the hrp genes
Creator:
Leite, Rui Pereira, 1957-
Publication Date:
Language:
English
Physical Description:
ix, 245 leaves : ill., photos ; 29 cm.

Subjects

Subjects / Keywords:
Bacteria ( jstor )
Diseases ( jstor )
DNA ( jstor )
Fatty acids ( jstor )
Nucleotide sequences ( jstor )
Peppers ( jstor )
Stall ( jstor )
Tomatoes ( jstor )
Xanthomonas ( jstor )
Xanthomonas campestris ( jstor )
Dissertations, Academic -- Plant Pathology -- UF
Plant Pathology thesis Ph.D
City of Gainesville ( local )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1994.
Bibliography:
Includes bibliographical references (leaves 226-244).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Rui Pereira Leite Jr.

Record Information

Source Institution:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
021117626 ( ALEPH )
33038024 ( OCLC )

Downloads

This item has the following downloads:


Full Text
157
Materials and Methods
Bacterial strains and culture conditions
The bacterial strains isolated from pepper and tomato materials which were
used in this study and their sources are listed in Table 7-1. All strains had been
identified previously as members of the Xanthomonas by fatty acid analysis (N. C.
Hodge, personal communication). All strains were streaked on nutrient agar (Becton
Dickinson) and single colonies were selected. Nutrient broth cultures were grown 24
hr on a rotatory shaker (150 rpm) at 28 C. Strains were stored in sterile tap-water at
room temperature.
Seed sources and artificial infestation
Pepper and tomato seeds were obtained from fruits or were provided by
commercial seed companies. Naturally infested seeds of pepper and tomato were
obtained by extracting seeds from diseased fruits or were provided by the Georgia
Department of Agriculture at Tilton, GA. In some experiments, bacterial suspensions
ofX. campestris pv. vesicatoria 75-3 was added to the seed wash mixture at the
appropriate concentrations before ultrasonication.
Plant material and plant inoculations
All plants were maintained in a growth chamber at 28-30C during inoculation
and incubation. The pepper cultivar. Early Calwonder and the tomato cultivar, Bonny
Best were used in the pathogenicity tests. In pathogenicity tests, fully expanded leaves
of plants were inoculated with bacterial suspensions by infiltrating the bacteria into the
intercellular spaces by using a 1 ml plastic syringe with a 27 gauge needle. The


25
The xanthomonads that cause diseases on citrus also comprise a genetically
diverse group of bacteria. Strains of the citrus pathogens X. campes tris pv. citri and X.
campestris pv. citrumelo which belong to the DNA homology group VIII of Vauterin
et al. (1993) are only about 60% similar to one another based on DNA-DNA homology
(Egel et al., 1991; Vauterin et al., 1991b). X campestris pv. citri canker A strains are,
however, highly related to other pathovars of X. campestris. The canker A strains of
this bacterium are related at over 90% by DNA-DNA homology to X. campestris pv.
glycines and X. campestris pv. malvacearum (Egel et al., 1991; Vauterin et al., 1993).
On the other hand, strains of the citrus bacterial pathogen, X. campestris pv. citrumelo,
are closely related to strains of other pathovars of X campestris which includes X
campestris pv. alfalfae and A! campestris pv. fici (Egel et al., 1991; Vauterin et al.,
1993). A strong relationship exists between the citrus pathogens based on
corroborating data obtained from DNA-DNA homology studies and analyses of fatty
acid composition (Graham et al., 1990; Vauterin et al., 1991b), SDS-PAGE of proteins
(Vauterin et al., 1991b), and RFLP of genomic DNA (Gabriel et al., 1988, 1989;
Graham et al., 1990; Vauterin et al., 1991b).
Different techniques of genomic fingerprinting have been used to explore the
genetic diversity of the plant pathogenic xanthomonads for identification and
comparison purposes. Genomic fingerprintings produced by using frequent-cutting
endonucleases were examined for establishing the genetic relationships of the
xanthomonads that cause diseases on citrus (Hartung and Civerolo, 1987; Pruvost et al.,
1992). The citrus canker A and B groups of X. campestris pv. citri each produced very
conserved genomic fingerprintings which distinguished strains of both groups (Hartung
and Civerolo, 1987). Although strains of the citrus bacterial spot pathogen^
campestris pv. citrumelo could be distinguished from both canker groups of X.


175
M1 23 45678
1 375
564
Fig. 7-5. Amplification of the 355 bp DNA fragment related to the complementation
group B of the hrp gene cluster of Xanthomonas campestris pv. vesicatoria from DNA
extracted from different pepper seed lots. Lanes: M, phage X restricted with EcoRl and
HindilI; 1, SP2.92; 2, SP66.92; 3, SP124.92; 4, SP133.92; 5, SP135.92; 6, SP306.92;
7, Greenhouse grown; 8, Jupiter. Molecular sizes are given in bases.


196
Table A-1 Continued
Pathovar
Strain
Source8
pv. dieffenbachiae
X260, X422, X736, X738, X739, X745,
X757, X763, X790, X1238, X1241, X1246,
X1272, XI292, X1298, X1426
ARC
pv. fici
X125, X151, X208, X212, X224, X548,
X687, X702
ARC
pv. gardneri
XG101
DCH
XV6, 1066
RES
pv. glycines
G-56, 86-16, 86-17, 86-18, 86-20, 87-2
RES
pv. holcicola
G-23
JEL
pv. incanae
9561-1, UK152, UK153
RES
pv.
macul i fol i igardeniae
X22j
DPI
pv. malvacearum
RIATC
RES
XcmA, XcmB, XcmC, XcmD, XcmE, XcmF,
XcmH, XcmN
DWG
pv. manihotis
Xml25D
RES
pv. papavericola
XP5
DCH
pv. pelargonii
XCP2, XCP10, XCP17, XCP36, XCP44,
XCP54, XCP58, XCP60
JBJ
pv. phaseoli
81-19, 82-1, 82-17, 85-6, EK11, XCPH4,
XP20
RES
pv. phaseoli "ftiscans"
XP163A
DCH
pv. physalidicola
XP172
DCH
pv. poinsettiicola
071-424, X87, X202, X259, X351, X352,
X354
RES
pv. pruni
X1219L, X1219S, X1220L, X1220S
ARC
pv. raphani
16B, 69-2, 69-4B, 69-8R
RES
pv. secalis
XC129C
RES
pv. translucens
80-1
RES
Continued on following page


102
as probes revealed the presence of sequence similarity in the genomic DNA of several
pathovars of X. campestris and related species of Xanthomonas (Chapter 3; Bonas et
al., 1991; Stall and Minsavage, 1990). Further, the hrp genes of the xanthomonads
seem to hybridize to genomic DNA of P. solanacearum when low stringency
conditions are used (Arlat et al., 1991; Boucher et al., 1987). The hrp gene clusters of
the xanthomonads are also functionally interchangeable. Heterologous
complementation of the hrp genes has been achieved for several pathovars ofX.
campestris (Chapter 3; Arlat et al., 1991; Bonas et al., 1991). Despite all the
information accumulated regarding the hrp genes, little is known about the
evolutionary relationship among the hrp genes of the xanthomonads.
Inferences of the relationships of hrp genes of the plant pathogenic
xanthomonads based on molecular phylogeny would certainly improve the
understanding of the genetics of pathogenicity of these plant pathogens. Dissimilar hrp
gene clusters in strains with relatively divergent genetic background might indicate that
the hrp region has coevolved with the rest of the genome from a common bacterial
ancestor. On the contrary, similar hrp gene clusters in strains with relatively divergent
genetic background might indicate a more recent horizontal genetic movement between
bacterial strains. Alternatively, variability in the genetic relatedness between bacterial
strains for different regions of the hrp genes may support the hypothesis of different
origins of the hrp genes. In this study, I examined the evolutionary relationship of two
DNA sequences of the hrp genes from different plant pathogenic xanthomonads. DNA
fragments of the bacterial genome related to the hrpB and hrpC/D complementation
groups of the hrp gene cluster of X campestris pv. vesicatoria (Chapter 3; Bonas et al.,
1991) were enzymatically amplified from different plant pathogenic xanthomonads and
then digested with frequent-cutting restriction endonucleases. The restriction fragment


primers
RST RST
9 10
RST
3
RST
2
8.0 kb
2.7 kb
RST
21
RST
22
UJ
"J


234
Hodge, N. C., Chase, A. R., and Stall, R. E. 1992. Diversity of four species of
Xanthomonas as determined by cellular fatty acid anlyses. (Abstr.) Phytopathology
82:1153.
Holben, W. E., Jansson, J. K., Chelm, B. K., and Tiedje, J. M. 1988. DNA probe
method for the detection of specific microorganisms in the soil bacterial community.
Appl. Environ. Microbiol. 54: 703-711.
Holt, J. G., Krieg, N. R., Sneath, P. H. A., Staley, J. T., and Williams, S. T., eds. 1994.
Bergey's Manual of Determinative Bacteriology, 9th ed. Williams & Wilkins,
Baltimore, MD. 787 pp.
Huang, Y., Xu, P., and Sequeira, L. 1990. A second cluster of genes that specify
pathogenicity and host response in Pseudomonas solanacearum. Mol. Plant-Microbe
Interact. 3:48-53.
Hwang, I. Y., Lim, S. M., and Shaw, P. D. 1992. Cloning and characterization of
pathogenicity genes from Xanthomonas campestris pv. glycines. J. Bacteriol. 174:
1923-1931.
Jones, J. B., Chase, A. R., and Harris, G. K. 1993a. Evaluation of the Biolog GN
MicroPlate system for identification of some plant-pathogenic bacteria. Plant Dis. 77:
553-558.
Jones, J. B., Jones, J. P., Stall, R. E., and Zitter, T. A. 1991. Compendium of Tomato
Diseases. American Phytopathological Society Press, St. Paul, MN. 73 pp.
Jones, J. B., Minsavage, G. V., Stall, R. E., Kelly, R. O., and Bouzar, H. 1993b.
Genetic analysis of a DNA region involved in expression of two epitopes associated
with lipopolysaccharide in Xanthomonas campestris pv. vesicatoria. Phytopathology
83: 551-556.
Jones, J. B., Pohronezny, K. L., Stall, R. E., and Jones, J. P. 1986. Survival of
Xanthomonas campestris pv. vesicatoria in Florida on tomato crop residue.
Phytopathology 76: 430-434.
Kamoun, S., and Kado, C. I. 1990. A plant-inducible gene of Xanthomonas
campestris pv. campestris encodes an exocellular component required for growth in the
host and hypersensitivity on nonhosts. J. Bacteriol. 172: 5165-5172.


167
Results
Development of a DNA amplification approach for detection of plant pathogenic
xanthomonads in seed extracts
Cells of X campestris pv. vesicatoria were detected in extracts of tomato seeds
by amplification of an 840-bp DNA fragment of the hrp gene cluster. The fragment
was amplified from bulked DNA preparations obtained from the extracts to which cells
of strain 75-3 were added (Fig. 7-1 and 7-2). The 355-bp hrpB fragment was also
amplified from the same samples (data not shown). The hrp fragments were amplified
from the samples processed with and without the presence of PVPP and sodium
ascorbate (Fig. 7-1 and 7-2). Similar results were also obtained with other pepper and
tomato seed lots (data not shown). The two reagents have been added to environmental
and plant samples in order to assure the extraction of nucleic acids with the quality
required for consistent enzymatic amplification of DNA fragments (Holben et al.,
1988; Minsavage et al., 1994). The addition of PVPP, or sodium ascorbate, or both at
different concentrations to the seed washings apparently did not interfere with the
recovery of DNA, nor the amplification of the hrp fragments by polymerase chain
reaction (Fig. 7-1 and 7-2). Other methods of sample preparation, such as the quick
approach of DNA isolation (Kawasaki, 1990), did not produced consistent results in the
amplification of hrp fragments for the detection of X. campestris pv. vesicatoria present
in pepper and tomato seed washings (data not shown).
Investigation was carried out to further examine the quality of the DNA
extracted from tomato seed washings with sodium ascorbate and PVPP added to the
final concentration of 0.2 M and 0.1% respectively, or without these reagents for
amplification of the hrp fragments. Aliquots of 5 pi, 10 pi, and 15 pi of DNA
extracted from tomato seed washings containing cells of X. campestris pv. vesicatoria


Table D-lContinued
Species/Pathovar Species/Pathovar
12
13
14
15
1. pv. alfalfae
.5000
.5000
.8163
.9388
2. pv. alfalfae
.5000
.5000
.7755
.8980
3. pv. armoraceae
.3182
.3182
.3111
.3111
4. pv. begoniae A
.1818
.1818
.3556
.4889
5. pv. begoniae A
.2174
.2174
.3830
.5532
6. pv. begonia B
.2609
.2609
.4255
.5532
7. pv. bilvae
.7083
.7083
.6531
.7347
8. pv. campestris
.3182
.3182
.3111
.3111
9. pv. carotae A
.4091
.4091
.3556
.3556
10. pv. carotae B
.3182
.3182
.3111
.3111
11. pv. citri A
.7500
.7500
.4898
.6122
12. pv. citri B
1.000
.5306
.5306
13. pv. citri C
.5306
.5306
14. pv. citrumelo
15. pv. citrumelo
16. pv. citrumelo
17. pv. citrumelo
18. pv. citrumelo
19. pv. citrumelo
20. pv. citrumelo
21. pv. citrumelo
.7600
16
17
18
19
20
21
.8333
.8333
.9167
.8511
.7083
.8333
.7917
.7917
.8750
.8936
.6667
.7917
.2727
.2727
.2727
.2791
.2727
.3182
.5455
.5455
.5909
.6047
.5000
.5455
.5217
.5217
.5652
.5778
.4783
.6087
.5217
.5217
.5652
.5778
.5217
.6522
.6667
.6667
.7083
.6383
.6250
.6667
.2727
.2727
.2727
.2791
.2727
.3182
.3182
.3182
.3182
.3256
.3182
.3636
.2727
.2727
.2727
.2791
.2727
.3182
.5833
.5833
.5833
.5957
.4583
.5833
.4583
.4583
.4583
.4255
.5833
.5000
.4583
.4583
.4583
.4255
.5833
.5000
.7755
.7755
.7347
.6667
.6939
.6939
.8163
.8163
.8980
.8333
.6939
.8980
1.000
.9167
.8936
.7917
.7500
.9167
.8936
.7917
.7500
.
.9362
.7917
.7917
.7234
.7660
.7500
Continued on following page


68
Fig. 4-1. Restriction profiles established for the 840-bp /irp-related fragment amplified
from different strains of plant pathogenic Xanthomonas spp. and restricted with the
endonuclease Cfol. Lane M, phage X restricted with ft/I. Molecular sizes are given in
bases. See Table 4-2 for identification of strains of Xanthomonas spp. for each
restriction pattern.


11
Hodge et al., 1992; Vauterin et al., 1991b, 1992; Yang et al., 1993). Extensive studies
carried out to study the fatty acid composition of strains of the genus Xanthomonas
included bacteria from more than 100 pathovars oX. campestris (Hodge et al., 1992;
Yang et al., 1993). The members of the X. campestris have been divided in 24 different
groups on the basis of fatty acid composition (Yang et al., 1993). Some pathovars ofX.
campestris, such as X campestris pv. maculifoliigardeniae, X. campestris pv.
malvacearum, and X campestris pv. pelargonii, were homogeneous and had a distinct
fatty acid profiles (Hodge et al., 1992). In some cases, pathovars pathogenic for
members of the same family of plants could be grouped together based on fatty acid
composition. This is the case for the pathovars X. campestris pv. cerealis, X.
campestris pv. hordei, X campestris pv. secalis, X campestris pv. translucens, and X.
campestris pv. undulosa which cause diseases on cereal crops (Vauterin et al., 1992;
Yang et al., 1993), and also for the pathovars from crucifers X campestris pv. aberrans,
X campestris pv. armoraciae, X campestris pv. barbareae, X. campestris pv.
campestris, X. campestris pv. incanae, and X. campestris pv. raphani (Yang et al.,
1993). Many pathovars from leguminous plants, such as X campestris pv. alfalfae, X
campestris pv. glycines, X campestris pv. phaseoli, X campestris pv. phaseoli
"fuscans", X. campestris pv. rhynchosiae, X campestris pv. sesbaniae, and X.
campestris pv. vignicola, also clustered in the same group based on fatty acid analysis
(Yang et al., 1993). This grouping delineated on the basis of fatty acid composition
agreed with the relationships established based on DNA-DNA hybridization and
protein profile (Hildebrand et al., 1990; Vauterin et al., 1991a, 1992; Yang et al.,
1993).
In contrast to some pathovars, the strains of X. campestris causing diseases of
citrus are composed of a heterogeneous group of strains. Besides the heterogeneity in


192
lack of host speciation (Arlat et al. 1992; Bonas et al., 1991). Although the hrp genes
are necessary in the plant-pathogen interaction, other factors in the bacterial pathogen
are likely to be involved in host specialization (Fenselau et al., 1992; Gough et al.,
1992).
On the contrary, there is also indication of horizontal movement of the hrp
genes among plant pathogenic xanthomonads. This hypothesis is supported by two
lines of evidences, the presence of similar hrp sequences in strains with divergent
genetic background and variability in the relatedness between strains depending on the
hrp region examined. In the first case, the hrp genes of the strains ofX. campestris pv.
vesicatoria A were very closely related to the homologous sequences of the strains of
X. campestris pv. citrumelo. The estimated genetic relatedness between strains of these
two pathovars for both hrp-related regions examined ranged in all from 0.71 to 0.96.
Furthermore, the phylogenetic analysis strongly supports a monophyletic relationship
for the hrp genes of the strains of both pathovars. In contrast, DNA homology studies
indicated that the strains of these two pathovars are not closely related, with genomic
similarity usually below 0.58 (Egel, 1991).
Stronger evidence to support the contention of horizontal movement of the hrp
genes among the plant pathogenic xanthomonads is the variability in the relatedness
between strains for the two distinct hrp regions examined. For instance, the hrpC/D
region ofX. campestris pv. vignicola is phylogenetically monophyletic and closely
related to the homologous hrp region of a group of pathovars of X. campestris that
includes X. campestris pv. bilvae, X campestris pv. citri, X campestris pv. glycines A,
X. campestris pv. malvacearum, X. campestris pv. phaseoli B, X. campestris pv.
phaseoli "fuscans", and X campestris pv. vitians B. On the contrary, the hrpB region
of A! campestris pv. vignicola though monophyletic is closely related to the


LITERATURE CITED
Alvarez, A. M., Benedict, A. A., and Gottwald, T. R. 1990. Serological relationships
among Xanthomonas campestris strains associated with citrus bacterial spot. (Abstr.)
Phytopathology 80: 964
Alvarez, A. M., Benedict, A. A., and Mizumoto, C. Y. 1985. Identification of
xanthomonads and grouping of strains of Xanthomonas campestris pv. campestris with
monoclonal antibodies. Phytopathology 75:722-728.
Alvarez, A. M., Benedict, A. A., Mizumoto, C. Y., Pollard, L. W., and Civerolo, E. L.
1991. Analysis of Xanthomonas campestris pv. citri and X. c. citrumelo with
monoclonal antibodies. Phytopathology 81:857-865.
Alvarez, A. M. and Lou, K. 1982. Rapid field identification of a bacterial pathogen by
an enzyme-linked immunosorbent assay (ELISA). (Abstr.) Phytopathology 72: 947.
Anderson, M. J., and Nameth, S. T. 1990. Development of a polyclonal antibody-
based serodiagnostic assay for the detection of Xanthomonas campestris pv. pelargonii
in geranium plants. Phytopathology 80: 357-360.
Angeles-Ramos, R., Vidaver, A. K., and Flynn, P. 1991. Characterization of epiphytic
Xanthomonas campestris pv. phaseoli and pectolytic xanthomonads recovered from
symptomless weeds in the Dominican Republic. Phytopathology 81: 677-681.
Arlat, M., Gough, C. L., Barber, C. E., Boucher, C., and Daniels, M. J. 1991.
Xanthomonas campestris contains a cluster of hrp genes related to the larger hrp cluster
of Pseudomonas solanacearum. Mol. Plant-Microbe Interact. 4:593-601.
Arlat, M., Gough, C. L., Zischek, C., Barberis, P. A., Trigalet, A., and Boucher, C. A.
1992. Transcriptional organization and expression of the large hrp gene cluster of
Pseudomonas solanacearum. Mol. Plant-Microbe Interact. 5:187-193.
Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., and Seidman, K., eds.. 1987.
Current Protocols in Molecular Biology. J. Wiley & Sons, New York.
226


APPENDIX B
SAS PROGRAM TO ESTIMATE THE SIMILARITY VALUES
FROM RESTRICTION FRAGMENT DATA
options ls=80 ps=54 nodate nonumber;
data dnatype;
infile bands lrecl=835;
input bacteria $ bandl -band 100; /* Read data from file Bacteria and set number of
*/
/* fragments compared */
proc sort data=dna_type;
by bacteria;
data dna type;
set dna type;
bactl=_n_;
bact2=_n_;
proc transpose data=dna_type out=transout; /* Create a new set with rows of the */
var band 1 -band 100; /* original data matrix becoming
*/
/* columns, and columns becoming */
/* rows */
data coefmtx;
coef=;
bactl=;
bact2=;
%macro calccoef;
%do i=l %to 3;
%do j=%eval (&i+l) %to 4;
/* Invoke the macro processor and execute */
/* the calculations until j = # of strains to be */
/* compared; i =j -1 */
201


237
McGuire, R. G., and Jones, J.B. 1989. Detection of Xanthomonas campestris pv.
vesicatoria in tomato. Pages 59-62 in: Detection of Bacteria in Seed and Other
Planting Material. A.W. Saettler, N.W. Schaad, and D.A. Roth, eds. American
Phytopathological Society Press, St. Paul, MN.
McGuire, R. G., Jones, J. B., and Sasser, M. 1986. Tween media for semiselective
isolation of Xanthomonas campestris pv. vesicatoria from soil and plant material.
Plant Dis. 70:887-891.
Miller, J. H. 1972. Experiments in Molecular Genetics. Cold Spring Harbor
Laboratory, Cold Spring Harbor, NY. 466 pp.
Miller, L. T. 1982. Single derivatization method for routine analyses of bacterial
esters, including hydroxy acids. J. Clin. Microbiol. 16: 584-586.
Miller, S. A., and Martin, R. R. 1988. Molecular diagnosis of plant disease. Annu.
Rev. Phytopathol. 26: 409-432.
Ming, D., Ye, H., Schaad, N. W., and Roth, D. A. 1991. Selective recovery of
Xanthomonas spp. from rice seed. Phytopathology 81: 1358-1363.
Minsavage, G. V., Dahlbeck, D., Whalen, M. C., Kearney, B., Bonas, U., Staskawicz,
B. J., and Stall, R. E. 1990. Gene-for-gene relationships specifying disease resistance
in Xanthomonas campestris pv. vesicatoria pepper interactions. Mol. Plant-Microbe
Interact. 3: 41-47.
Minsavage, G. V., and Schaad, N. W. 1983. Characterization of membrane proteins of
Xanthomonas campestris pv. campestris. Phytopathology 73:747-755.
Minsavage, G. V., Thompson, C. M., Hopkins, D. L., Leite, R. M. V. B. C., and Stall,
R. E. 1994. Development of a polymerase chain reaction protocol for detection of
Xylella fastidiosa in plant tissue. Phytopathology 84:456-461.
Morris, C. 1985. Diversity of Epiphytic Bacteria on Snap Bean Leaflets Based on
Nutrient Utilization Abilities: Biological and Statistical Considerations. Ph.D.
dissertation. University of Wisconsin, Madison. 210 pp.
Nei, M. 1987. Molecular Evolutionary Genetics. Columbia University Press, New
York, NY.


Table D-2Continued
Species/Pathovar Species/Pathovar
24
25
26
27
28
29
31
34
35
36
1. pv. alfalfae
.6190
.7111
.8696
.9333
.9333
.3902
.4545
.4545
.5854
.4651
2. pv. alfalfae
.6190
.7111
.8696
.9333
.9333
.3902
.4545
.4545
.5854
.4651
3. pv. armoraciae
.5238
.5778
.5217
.4444
.4889
.6341
.5455
.9545
.5366
.5116
4. pv. begoniae A
.9231
.5714
.7442
.6667
.7143
.5789
.5366
.5366
.8421
.5500
5. pv. begoniae A
.9231
.5714
.7442
.6667
.7143
.5789
.5366
.5366
.8421
.5500
6. pv. begoniae B
.9231
.5714
.7442
.6667
.7143
.5789
.5366
.5366
.8421
.5500
7. pv. bilvae
.6047
.5217
.5957
.5217
.5652
.4286
.8444
.5778
.6190
.8182
8. pv. campestris
.5714
.5333
.5652
.4889
.5333
.6829
.5455
1.000
.5854
.5116
11. pv. citri A
.5714
.4889
.5217
.4444
.4889
.4390
.9091
.5455
.5854
.8837
12. pv. citri B
.5714
.4889
.5217
.4444
.4889
.3902
.8182
.5455
.5854
.7907
13. pv. citri C
.5714
.4889
.5217
.4444
.4889
.3902
.8182
.5455
.5854
.7907
14. pv. citrumelo
.6512
.7826
.9362
1.000
.8696
.4286
.4889
.4889
.6190
.5000
15. pv. citrumelo
.7273
.8511
1.000
.9362
.9362
.5116
.5652
.5652
.6977
.5778
16. pv. citrumelo
.7273
.8511
1.000
.9362
.9362
.5116
.5652
.5652
.6977
.5778
17. pv. ctirumelo
.6977
.7826
.9362
.8696
1.000
.4762
.5333
.5333
.6667
.5455
18. pv. citrumelo
.6512
.7826
.9362
1.000
.8696
.4286
.4889
.4889
.6190
.5000
19. pv. citrumelo
.6190
.6667
.8261
.8889
.8889
.4390
.4545
.4545
.5854
.4186
20. pv. citrumelo
.7273
.8511
1.000
.9362
.9362
.5116
.5652
.5652
.6977
.5778
21. pv. citrumelo
.6190
.7111
.8696
.9333
.9333
.3902
.4545
.4545
.5854
.4651
22. pv. citrumelo
.7273
.8085
.9583
.8936
.8936
.5581
.5652
.5652
.6977
.5333
Continued on following page


Table 7-1Continued
a DAM, D. A. Maddox, Seed Testing of America, Longmont, CO; GOK, G. O'Keefe, Department of Agriculture, State of
Georgia, Tifton, GA; JD, J. Dodson, Woodland Research Station, Petoseed, Woodland, CA; JW, J. Watterson, Woodland
Reserach Station, Petoseed, Woodland, CA; MD, M. Derie, Washington State University, Pullman, WA; WW, W. Wiebe,
Rogers NK, Woodland CA.
b Identification of the strain was based on comparison of the hrp profile established to the different groups of plant
pathogenic xanthomonads (see Chapter 4).
c Identification based on the MIDI TSBA Library (version 3.80; Microbial Identification System software, Newark, DE).
Number in parenthesis is the similarity index.
d -, no /irp-related fragment amplification; +, /zr/?-related fragment amplification; (+) weak /irp-related fragment
amplification.
e -, no reaction; +, pathogenic reaction; HR, hypersensitive reaction.
f Strains of X. campestris pv. campestris could not be distinguished from strains of X. campestris pv. armoraciae by the
restriction analysis of both hrp fragments (see Chapter 4).
8 na, not applicable; nd, not determined.
On
On


169
M 1 2 3 4 5
1
Fig. 7-2. Amplification of the 840-bp fragment of the complementation group B of the
hrp gene cluster from seeds of tomato containing 106 CFU/ml of Xanthomonas
campestris pv. vesicatoria 75-3 and different amounts of sodium ascorbate. Lanes: M,
phage X restricted with EcoRl and HindiII; 1, no sodium ascorbate nor PVPP; 2, no
sodium ascorbate; 3, 0.02 M of sodium ascorbate; 4, 0.2 M of sodium ascorbate; 5,2 M
sodium ascorbate. Molecular sizes are given in bases.


110
100
pv. vesicatoria B (11)
84
42
68
94
34
74
84
88
26r
54rl
pv. malvacearum (9)
I pv. vitians B (1)
i pv. phaseoli B (1)
1 pv. glycines A (6)
pv. citri A (o)
pv. citri C (6)
pv. citri B (7)
X. campestris XCF (1)
pv. bilvae (1)
64
14
42
84
64
60 36~TT_
44f~l
pv. phaseoli Tuscans" (1)
pv. citrumelo (1)
58| pv. alfalfae (1)
pv. alfalfae (1)
_ pv. citrumelo (2)
pv. citrumelo (5)
pv. citrumelo (2)
6_6r
20
pv. fici A (1)
pv. citrumelo (1)
pv. fici A (1)
4
K
18
5
24
r
14
4l
20
44r
20r
54
XOl
34~E?p~~~
78
78
72
80
pv. poinsettiicola A (2)
pv. pruni (4)
pv. fici A (1)
pv. physalidicola (1)
pv. vesicatoria A (11
pv. vesicatoria A (24)
pv. vesicatoria A (2)
pv. citrumelo (1)
pv. citrumelo (1)
pv. citrumelo (1)
pv. citrumelo (2)
_ pv. dieffenbachiae B (2)
_ pv. vignicola (3)
X. campestris X198 (1)
pv. maculifoliigardeniae (1)
pv. manihotis (1)
pv. vitians A (3)
pv. phaseoli A (4)
pv. dieffenbachiae A (4)
pv. dieffenbachiae A (4)
pv. begoniae (1)
pv. begoniae (7)
pv. begoniae (1)
68
40
80
pv. poinsettiicola B (3)
pv. fici B (2)
- X. fragariae (9)
96 | pv. armoraciae (3)
I pv. raphani B (1)
32
pv. campestris (9)
pv. incanae (1)
pv. raphani A (3)


20
The identification of strains of Xanthomonas to species and to pathovar was correct in
97% and 6% of the cases, respectively, when the original data base was used.
However, the accuracy of the Biolog system for identification of pathovars of X.
campestris was substantially improved when the database was amended by the addition
of data obtained for different strains of X campestris. Nevertheless, the Biolog results
confirmed the close relatedness between the various pathovars determined in previous
work, and substantiated the difficulty of separation of pathovars on the basis of
phenotypic features alone.
The metabolic variation of strains of X campestris that cause diseases on citrus
was investigated by using the Biolog system (Hartung and Civerolo, 1991; Vemiere et
al., 1993). The Biolog metabolic profile was useful for differentiating strains of the
citrus canker pathogen X campestris pv. citri from strains of the citrus bacterial spot
agent X. campestris pv. citrumelo. Furthermore, the different citrus canker groups of X.
campestris pv. citri could be differentiated based on the utilization of three carbon
sources (Vemiere et al., 1993). Although the correct identification of theJf campestris
pv. citri strains was very poor by using the original data base of the system, the
identification was significantly improved when the database was supplemented with
additional data (Vemiere et al., 1993). The citrus bacterial spot pathogen X. campestris
pv. citrumelo was more variable in the metabolic profile than X. campestris pv. citri
(Hartung and Civerolo, 1991). All the profiles examined placed the strains within the
species X campestris. However, some strains were identified as X. campestris pv.
dieffenbachiae (Hartung and Civerolo, 1991). The relatedness of strains ofX.
campestris pv. citrumelo to other pathovars of X. campestris has also been determined
on the basis of DNA-DNA hybridization (Egel et al., 1991), RFLP analysis (Gabriel et
al., 1988, 1989), and fatty acid composition (Graham et al., 1990).


APPENDIX D
ESTIMATES OF SIMILARITY VALUES FOR PLANT PATHOGENIC XANTHOMONADS BASED ON
COMPARISON OF RESTRICTION FRAGMENT DATA OF THE hrp RELATED DNA SEQUENCES
Table D-l. Similarity values for plant pathogenic strains of different pathovars of Xanthomonas campestris and X. fragariae
based on the genetic analysis of the hrpC/D related DNA sequences.
Species/Pathovar
Species/Pathovar
2
3
4
5
6
7
8
9
10
11
1. pv. alfalfae
.9583
.3182
.5000
.5217
.5217
.7083
.3182
.3636
.3182
.5833
2. pv. alfalfae
.3182
.5000
.5217
.5217
.6667
.3182
.3636
.3182
.5833
3. pv. armoraciae
.2500
.3333
.3810
.3182
1.000
.8500
1.000
.3636
4. pv. begonia A
.
.9048
.8571
.3636
.2500
.3000
.2500
.3636
5. pv. begonia B
.
.9545
.3913
.3333
.3333
.3333
.3913
6. pv. begoniae C
.3913
.3810
.3810
.3810
.3913
7. pv. bilvae
.
.3182
.4091
.3182
.7917
8. pv. campestris
.
.
.8500
1.000
.3636
9. pv. carotae A
.
.
.
.8500
.4545
10. pv. carotae B
.
.
.
.
.3636
11. pv. citri A




Continued on following page


their lives. I am also indebted to my own parents who have offered whatever support
was needed.
in


24
The most striking feature of the plant pathogenic xanthomonads is the genetic
diversity among strains of the same pathovar or different pathovars of X. campestris
that cause diseases on highly related hosts. The diversity strongly illustrates the
divergence between pathogenicity characteristics and genetic backgrounds. The
pathovars X. campestris pv. gardneri, X. campestris pv. physalidicola, and X
campestris pv. vesicatoria, which cause diseases on solanaceous plants, are genetically
diverse on the basis of DNA-DNA homology (Hildebrand et al., 1990; Palleroni et ah,
1993; Stall et ah, 1994; Vauterin et ah, 1993). X. campestris pv. vesicatoria is
composed of the two genetically unrelated groups A and B (Stall et ah, 1994). X.
campestris pv. vesicatoria group A belongs to homology group VIII of Vauterin et ah
(1993), as well as X. campestris pv. physalidicola. Further, X. campestris pv.
vesicatoria group A and X. campestris pv. physalidicola are also genetically closely
related to several other pathovars of X. campestris that cause diseases on different
plants (Hildebrand et ah, 1990; Palleroni et ah, 1993; Vauterin et ah, 1993). Strains of
X campestris pv. vesicatoria group A are only about 33% related to strains ofX.
campestris pv. vesicatoria group B, on the basis of DNA-DNA homology (Stall et ah,
1994). Strains of X. campestris pv. vesicatoria group B for instance is distinct from
other xanthomonads and they are the only members of the DNA homology group XVI
(Vauterin et ah, 1993). The grouping of X. campestris pv. vesicatoria are also well
supported by fatty acid analysis, serology, and phenotypic features (Stall et ah, 1994).
X. campestris pv. gardneri is also not related genetically to the other solanaceous
pathogens. Instead, X. campestris pv. gardneri constitutes a group which contains the
pathovars X. campestris pv. carotae, X. campestris pv. pelargonii, and X. campestris pv.
taraxaci (Hildebrand et ah, 1990; Palleroni et ah, 1993).


232
Gilbertson, R. L., Maxwell, D. P., Hagedom, D. J., and Leong, S. A. 1989.
Development and application of a plasmid DNA probe for detection of bacteria causing
common bacterial blight of bean. Phytopathology 79:518-525.
Gilbertson, R. L., Rand, R. E., and Hagedom, D. J. 1990. Survival of Xanthomonas
campestris pv. phaseoli and pectolytic strains o Xanthomonas campestris in bean
debris. Plant Dis. 74: 322-327.
Gitaitis, R. D., Chang, C. J., Sijam, K., and Dowler, C. C. 1991. A differential
medium for semiselective isolation of Xanthomonas campestris pv. vesicatoria and
other cellulolytic xanthomonads from various natural sources. Plant Dis. 75: 1274-
1278.
Gitaitis, R., McCarter, S., and Jones, J. 1992. Disease control in tomato transplants
produced in Georgia and Florida. Plant Dis. 76: 651-656.
Gitaitis, R. D., Sasser, M. J., Beaver, R. W., Mclnnes, T. B., and Stall, R. E. 1987.
Pectolytic xanthomonads in mixed infections with Pseudomonas syringae pv. syringae,
P. syringae pv. tomato, and Xanthomonas campestris pv. vesicatoria in tomato and
pepper transplants. Phytopathology 77:611-615.
Gottwald, T. R., Alvarez, A. M., Hartung, J. S., and Benedict, A. A. 1991. Diversity
of Xanthomonas campestris pv. citrumelo strains associated with epidemics of citrus
bacterial spot in Florida citrus nurseries: Correlation of detached leaf, monoclonal
antibody, and restriction fragment length polymorphism assays. Phytopathology 81:
749-753.
Gough, C. L., Genin, S., Zischek, C., and Boucher, C. A. 1992. hrp genes of
Pseudomonas solanacearum are homologous to pathogenicity determinants of animal
pathogenic bacteria and are conserved among plant pathogenic. Mol. Plant-Microbe
Interact. 5: 384-389.
Graham, J. H., and Gottwald, T. R. 1988. Citrus canker and citrus bacterial spot in
Florida: Research fmdings-future considerations. The Citrus Industry 69: 42-51.
Graham, J. H., and Gottwald, T. R. 1990. Variation in aggressiveness of Xanthomonas
campestris pv. citrumelo associated with citrus bacterial spot in Florida citrus nurseries.
Phytopathology 80: 190-196.
Graham, J.H., and Gottwald, T.R. 1991. Research perspectives on eradication of citrus
bacterial diseases in Florida. Plant Dis. 75:1193-1200.


123
with divergent genetic background and the variability in the relatedness between strains
depending on the hrp region examined. Strains of X. campestris pv. vesicatoria A
belong to the same clade as the strains of X. campestris pv. citrumelo for both hrp
genes examined (Fig. 5-1 and 5-2). Furthermore, the estimated genetic relatedness
between strains of these two pathovars for the hrp genes regions ranged from about
0.71 to 0.96 (Appendix D), although these strains cause diseases on different hosts. On
the contrary, DNA-DNA hybridization studies indicated that these two pathovars are
not genetically closely related based on comparisons of the entire bacterial genome
which was compared (Egel, 1991). In fact, the relatedness between X. campestris pv.
citrumelo and X campestris pv. vesicatoria was about 0.58 which is close to the
similarity determined between strains of X. campestris pv. citrumelo and X. campestris
pv. citri discussed before (Egel, 1991; Egel et al., 1991). X. campestris pv. citrumelo
has been reported occurring only in Florida (Schoulties et al., 1987). Although the
origin of these strains is not very clear yet (Schoulties et al., 1987), the strains of this
pathovar may have evolved from an endemic flora of xanthomonads that is present in
Florida (Hartung and Civerolo, 1989). The analysis of the hrp genes of the strains of X.
campestris pv. citrumelo revealed some degree of variability, but there is strong
support for a monophyletic relationship among them. Furthermore, they are also
closely related and monophyletic in regard to both regions of the hrp genes to strains of
other pathovars of X campestris, i.e. X. campestris pv. alfalfae, X. campestris pv. fici,
X campestris pv. poinsettiicola, X. campestris pv. pruni, and X. campestris pv.
vesicatoria A, and the strain XI98 of campestris, which are also endemic to the same
geographic area. A more extensive study of the genetic relationships among these
pathogens also including nonpathogenic xanthomonads endemic to Florida may


19
diseases on grasses. The analysis of 257 phenotypic characteristics of strains from
grasses revealed three distinct groups corresponding to (1).V campestris pv. graminis,
and A"! campestris pv. phleipratensis, (2) X campestris pv. arrhenatheri, and (3) X
campestris pv. poae (Van den Mooter et al., 1987b). These groups are also in
agreement with the ones established on the basis of pathogenicity features. However,
most of the strains of pathovars of X. campestris have heterogeneous phenotypic
characteristics and only have the pathogenicity to the same host plant as the common
feature (Burkholder and Starr, 1948; Dye, 1962; Van den Mooter and Swings, 1990).
This can be illustrated by the case of X campestris pv. dieffenbachiae. Chase et al.
(1992) analyzed the phenotypic features of 149 strains of X. campestris pv.
dieffenbachiae obtained from aroid plants and found large variation in their reactions.
Minimum pH for growth in nutrient broth was the only useful feature to differentiate
typical strains obtained from Syngonium spp. from all other aroid strains. Although
some trends were observed, physiological tests, such as hydrolysis of both pectin and
starch, and growth on different media, failed to differentiate the strains to their host of
origin. The strains of X campestris pv. dieffenbachiae from different hosts overlapped
in their reactions for these tests, which made it impossible to relate the phenotypic
features to the host of origin.
Commercially available systems for identification of bacteria on the basis of
phenotypic characteristics have also been examined for differentiation of plant
pathogenic xanthomonads (Chase et al., 1992; Jones et al., 1993a; Vauterin et al.,
1990b; Vemiere et al., 1993). The Biolog GN MicroPlate System, which is based on
differential utilization of 95 carbon sources with consequent changes in redox potential
of the bacterial suspension (Bochner, 1989), was evaluated for accuracy in identifying
some plant pathogenic bacteria, including strains of Xanthomonas (Jones et al., 1993a).


Table 4-2Continued
Species/Pathovar
Strain
840 bp
1,075 bp
Taql
Saw3AI HaeIII
Cfo\
Taql
Sau3AI Hae III
Cfol
pv. raphani
69-2, 69-4B, 69-8R
1
3
3
3
4
6
7
4
16B
1
3
3
4
4
6
7
4
pv. taraxaci
XT11A
nd
nd
nd
nd
3
3
6
2
pv. vesicatoria
group A
75-3, 82-4, 85-10, 87-21, 87-
35T, 87-44T, 87-48T, 89-8, 89-
10, 90-20, 90-21, 90-27, 90-40,
91-66, 91-72, 92-11, 92-15, 92-
16, 92-17, 92-119, 1712,6107,
XV14, XV17
3
6
16
12
8
11
16
17
92-118,92-120
3
6
16
13
8
12
17
17
90-60
3
6
16
12
8
12
17
17
group B
141.0226A, 0350A, 695,853,
1062, B-3,B-20, BA28-1, BV5-
5, XV56
1
1
1
1
1
1
1
1
pv. vignicola
81-30, 82-38, G-55
6
7
16
12
7
9
11
11
Continued on following page
OO
O


105
suggested by Nei and Li (1979). A program written for the SAS system was used for
the calculations (Appendix C).
The evolutionary relationship among strains was examined by using programs
from the computer package PHYLIP (Felsenstein, 1991, 1993). Phylogenetic trees
were inferred for each of the two hrp regions individually by using a parsimony
criterion and a distance matrix method. The restriction fragment data encoded 0 or 1
were used as input for reconstruction of an unrooted phylogenetic tree by the Wagner
parsimony criterion of the BOOT program (Felsenstein, 1991). The strains G-23 ofY
campestris pv. holcicola and XV56 of X campestris pv. vesicatoria group B were taken
as the outgroups to infer the topology of the phylogenetic trees for the hrpC/D and
hrpB regions, respectively. No assumptions were made for the ancestral character state
and the confidence intervals of the estimates of the inferred phylogenetic trees were
determined by analyzing a total of 100 bootstrap samples (Felsenstein, 1985, 1991).
The distance matrix method unweight pair-group method with arithmetic mean
(UPGMA) of phylogenetic reconstruction (Nei, 1987) was also applied to the data for
reconstruction of a rooted phylogenetic tree. The option UPGMA of the program
NEIGHBOR (Felsenstein, 1993) was performed for this analysis. The input data
consisted of the estimates of the number of nucleotide substitutions per site (5) between
strains determined by the method developed by Nei and Li (1979). For both methods,
the data of each strain were entered in a random order.
Results
DNA fragments related to the hrpB and hrpC/D complementation groups of the
hrp genes ofX. campestris pv. vesicatoria were amplified from strains representing the


Fig. 5-4. Rooted phylogenetic tree inferred from restriction analysis data of the 840-bp
DNA fragment related to the hrpB complementation group of the hrp genes of
Xanthomonas campestris pv. vesicatoria generated by the method unweight pair-group
method with arithmetic mean (UPGMA). The shaded boxes delineate the major clades
identified in the analysis of the hrpC/D-related DNA fragment. Also included are the
number of strains examined for each taxa (numbers in parenthesis).


53
1990) on the conservation of the hrp region among the plant pathogenic xanthomonads.
In contrast, the DNA of A campestris pv. secalis and X campestris pv. translucens did
not hybridize to the hrp fragments. X campestris pv. secalis is genetically only weakly
related to a few pathovars of X campestris based on DNA-DNA hybridization
experiments (Hildebrand et al., 1990). Previously, X campestris pv. translucens has
shown only weak hybridization to DNA probes representing the entire hrp gene cluster
region of X campestris pv. vesicatoria (Bonas et al., 1991; Stall and Minsavage, 1990).
Although hybridization to DNA of all strains of X. campestris included in this study
was not observed, the hrp region seems useful for detection and identification of a large
number of plant pathogenic xanthomonads. A major advantage in using hrp sequence
is the lack of homology to DNA of non plant pathogenic xanthomonads, as observed
for X maltophilia and the opportunistic X. campestris strains T-55 and INA, as well as
other plant pathogens of the genera Acidovorax, Agrobacterium, Clavibacter, Erwinia,
Pseudomonas, and Xylella.
The three pairs of oligonucleotide primers described in this study are specific
for the hrpB, hrpC, and hrpD regions of A campestris pv. vesicatoria 75-3 and were
used to amplify homologous DNA fragments from X. fragariae and from 31 of 33
plant pathogenic taxa of A campestris tested, which comprise at least 28 different
pathovars of this species. In all cases, each set of primers amplified DNA fragments
identical in size, suggesting a high degree of structural conservation between operons,
as seen with primers RST21 and RST22. Cloned regions of DNA of A campestris pv.
vesicatoria group B and A! campestris pv. pelargonii, from which the hrp related
fragments were also amplified, fully restored pathogenicity to several nonpathogenic
Tn3-gws mutants of A campestris pv. vesicatoria 85-10. This supports the contention


CHAPTER 5
PHYLOGENETIC ANALYSIS PLANT PATHOGENIC Xanthomonas
BASED ON DNA SEQUENCES RELATED TO THE hrp GENES
The genus Xanthomonas Dowson 1939 includes bacteria that occur worldwide
and cause economically important diseases on many plants. The host range of the
xanthomonads spans over 392 mono and dicotyledonous plant species (Hayward, 1993;
Leyns et al., 1984). X. campestris is certainly the most complex species among the
xanthomonads. This species is highly diverse and comprises at least 125 different
pathovars (Bradbury, 1984; Dye et al., 1980; Hayward, 1993). The classification of the
plant pathogenic xanthomonads has been based largely on the capability of the bacterial
strain to cause a characteristic disease (Bradbury, 1984; Dye et al., 1980; Vauterin et
al., 1990a; Young et al., 1992). Consequently, bacteria with features different from
pathogenicity may be classified under the same taxonomic unit. On the other hand,
genetically similar bacteria may be placed in different taxa because they cause diseases
on different plants.
The establishment of relationships among the different pathovars and species of
Xanthomonas has been attempted by using different approaches, such as metabolic,
fatty acid, and protein profiling (Chase et al., 1992; Hildebrand et al., 1993; Hodge et
al., 1992; Van den Mooter and Swings, 1990; Vauterin et al., 1991b), and nucleic acid
analyses (Hildebrand et al., 1990; Palleroni et al., 1993; Vauterin et al., 1993).
Although the pathovars of X. campestris have almost identical biochemical and
physiological characteristics (Van den Mooter and Swings, 1990), diversity seems to
100


236
oryzae pv. oryzae with a repetitive DNA element. Appl. Environ. Microbiol. 58:
2188-2195.
Leach, J. E., White, F. F., Rhoads, M. L., and Leung, H. 1990. A repetitive DNA
sequence differentiates Xanthomonas campes tris pv. oryzae from other pathovars of X.
campestris. Mol. Plant-Microbe Interact. 3:238-246.
Leite, R. P., Jr. 1990. Cancro Ctrico: Preven9ao e Controle no Paran. Funda^ao
Instituto Agronmico do Paran, Londrina, Brazil. 51 pp.
Liao, C. H., and Wells, J. M. 1987. Association of pectolytic strains of Xanthomonas
campestris with soft rots of fruits and vegetables at retail markets. Phytopathology 77:
418-422.
Lindgren, P. B., Peet, R. C., and Panopoulos, N. J. 1986. Gene cluster of
Pseudomonas syringae pv. "phaseolicola" controls hypersensitivity on nonhost plants.
J. Bacteriol. 168:512-522.
Link, G. K. K., and Sharp, C. G. 1927. Correlation of host and serological specificity
of Bact. campestre, Bact. flaccumfaciens, Bad. phaseoli, and Bad. phaseoli sojense.
Bot.Gaz. 83: 145-160.
Lipp, R. L., Alvarez, A. M., Benedict, A. A., and Berestecky, J. 1992. Use of
monoclonal antibodies and pathogenicity tests to characterize strains of Xanthomonas
campestris pv. dieffenbachiae from aroids. Phytopathology 82: 677-682.
Leyns, F., DeCleene, M., Swings, J.-G., and DeLey, J. 1984. The host range of the
gems Xanthomonas. Bot. Rev. 50:308-356.
Maas, J. L., Finney, M. M., Civerolo, E. L., and Sasser, M. 1985. Association of an
unusual strain of Xanthomonas campestris with apple. Phytopathology 75: 438-445.
Mahanta, I. C., and Addy, S. K. 1977. Serological specificity o Xanthomonas oryzae
incitant of bacterial blight of rice. Int. J. Syst. Bacteriol. 27: 383-385.
Manulis, S., Gafni, Y., Clark, E., Zutra, D., Ophir, Y., and Barash, I. 1991.
Identification of a plasmid DNA probe for detection of strains of Erwinia herbicola
pathogenic on Gypsophila paniculata. Phytopathology 81:54-57.
Marco, G. M., and Stall, R. E. 1983. Control of bacterial spot of pepper initiated by
strains of Xanthomonas campestris pv. vesicatoria that differ in sensitivity to copper.
Plant Dis. 67:779-781.


10
between the serological heterogeneity and the host range of the xanthomonads.
Pathovar specific MABs have been developed that react specifically with all strains of
a pathovar of X. campestris that usually cause disease in a single host, as for example,
X. campestris pv. begoniae and X. campestris pv. pelargonii. On the contrary, no
pathovar specific MABs were produced that react with all strains of a pathovar that
have a large host range, such as X campestris pv. campestris, X. campestris pv. citri,
and X. campestris pv. dieffenbachiae.
Fattv Acids Composition
The analysis of fatty acids by gas chromatography has become a common
technique in bacterial classification and identification. Comprehensive studies have
been carried out on the relationship of fatty acid profiles within several groups of plant
pathogenic xanthomonads (Chase et al., 1992; Graham et al., 1990; Hodge et al., 1992;
Vauterin et al., 1991b, 1992; Yang et al., 1993). At least 65 different fatty acids were
found within the members of the genus Xanthomonas. Three fatty acids, 11:0 iso, 11:0
iso 30H, and 13:0 iso OH, are characteristic for all xanthomonads and they were useful
for differentiating Xanthomonas from other plant pathogenic bacteria (Vauterin et al.,
1992; Yang et al., 1993). Furthermore, the species X. albilineans, X. fragariae, and Y
populi were homogeneous based on fatty acid composition, and they could be clearly
differentiated from each other (Hodge et al., 1992; Vauterin et al., 1992; Yang et al.,
1993). Strains ofY oryzae have unique fatty acid composition within the genus
Xanthomonas and they form two major groups which correspond to the pathovars X.
oryzae pv. oryzae and X. oryzae pv. oryzicola (Vauterin et al., 1992; Yang et al., 1993).
As observed on other characteristics, X. campestris also comprises a
heterogeneous group in relation to the fatty acid composition (Chase et al., 1992;


56
In conclusion, the results presented here indicate that plant pathogenic strains of
X. campestris and related xanthomonad species can be detected and may be identified
by analysis of DNA fragments amplified with hrp gene-specific primers. The
conservation of the hrp DNA sequence among a large number of pathovars ofX.
campestris, as well as in related Xanthomonas spp., but lack of the hrp DNA sequence
among non-plant pathogenic bacteria, makes this method a useful tool for detection and
identification of many plant pathogens. Consequently, hrp oligonucleotide primers
may be also useful to determine the pathogenic nature of unknown xanthomonads.
This is particularly significant for assessing the complex population of plant pathogenic
and nonpathogenic xanthomonads associated with plants and plant parts. The presence
of plant pathogenic strains in such samples may be determined by amplification of the
hrp fragments without the need of the troublesome methods of isolation of the
organism and inoculation into potential host plants. Moreover, RFLPs detected in the
genome of different strains seem valuable for the study of the relatedness of plant
pathogenic xanthomonads, particularly among Xanthomonas spp. and pathovars of X.
campestris. Of course, this has to be extended by testing larger numbers of strains for
each species, subspecies, or pathovar. The genetic methods of analyzing populations of
bacteria will provide valuable additional information for taxonomic, ecological, and
epidemiological studies of plant pathogenic xanthomonads.


Table D-lContinued
Species/Pathovar Species/Pathovar
46. pv. pruni
47. pv. raphani A
48. pv. raphani B
49. pv. taraxaci
50. pv. vesicatoria A
51. pv. vesicatoria A
52. pv. vesicatoria A
53. pv. vesicatoria B
54. pv. vignicola
55. pv. vitians A
56. pv. vitians B
57. pv. vitians C
58. X. campe sir is XI98
59. X. campestris XCF
60. X. campestris X52
61.X. fragariae
53 54 55
.4490
.5306
.5200
.4091
.3636
.3556
.4091
.3636
.3556
.5000
.3182
.3111
.3333
.4583
.6122
.3830
.4681
.6667
.3830
.4681
.6667
.4167
.3673
.2857
56 57 58
.5217
.3556
.8980
.3902
.4500
.3636
.3902
.4500
.3636
.3415
.8500
.3636
.4000
.3182
.8750
.4091
.2791
.7660
.4091
.2791
.7660
.4444
.5909
.4167
.6667
.3636
.4583
.3478
.3111
.5306
.3902
.4889
.3636
59
60
61
.7347
.3182
.2041
.4545
.5128
.3182
.4545
.5128
.3182
.3182
.4615
.3636
.6250
.2791
.1667
.6383
.3333
.2128
.6383
.3333
.2128
.4167
.4186
.2500
.7083
.3256
.3333
.4082
.2727
.1224
.7556
.2500
.2667
.3636
.4615
.3636
.7083
.2791
.1667
.
.3256
.2500
.3256
216


page
D ESTIMATES OF SIMILARITY VALUES FOR
PLANT PATHOGENIC XANTHOMONADS BASED
ON COMPARISON OF RESTRICTION FRAGMENT
DATA OF THE hrp RELATED DNA SEQUENCES 205
LITERATURE CITED 226
BIOGRAPHICAL SKETCH 245
vi


22
Nucleic acid based techniques have also been applied to the identification and
detection of the plant pathogenic xanthomonads. Genomic and plasmid DNA
fingerprinting has been used in several instances for the differentiation of a number of
groups of xanthomonads (Berthier et al., 1993; Cooksey and Graham, 1989; Hartung
and Civerolo, 1987; Pruvost et al., 1992). In the same way, genomic and plasmid DNA
probes were also developed and examined for the purpose of identification and
detection of plant pathogenic xanthomonads (Garde and Bender, 1991; Gilbertson et
al., 1989; Hartung, 1992; Lazo and Gabriel, 1987; Leach et al., 1990, 1992).
The genomic relationship among plant pathogenic xanthomonads determined
by DNA-DNA hybridization has revealed a group of bacteria with diverse genetic
background, with DNA homologies between pathovars o X. campestris ranging from
values as low as 15% to 39% (Vauterin et al., 1990b). These results strongly reinforce
the contention that strains of X. campestris do not constitute a single species.
Furthermore, the genetic relationship established on the basis of DNA-DNA homology
does not always correlate with the pathogenicity of the bacteria toward a given range of
host plants. The genetic relationship of strains of plant pathogenic xanthomonads
representing 44 X. campestris pathovars and Xanthomonas spp. was examined by SI
DNA-DNA hybridization (Hildebrand et al., 1990; Palleroni et al., 1993). Several
groups of xanthomonads were defined on the basis of DNA-DNA homology. The
largest group consisted of strains belonging to 25 pathovars of X. campestris
(Hildebrand et al., 1990; Palleroni et al., 1993). This group corresponds to DNA
homology group VIII established by Vauterin et al. (1993), which also includes 22
additional pathovars of V. campestris and X axonopodis. Another cluster comprises a
group ofX. campestris pv. carotae, X. campestris pv. gardneri, X campestris pv.
pelargonii, and A", campestris pv. taraxaci (Hildebrand et al., 1990; Palleroni et al.,


Table 4-2Continued
Species/Pathovar
Strain
840 bp
1,075 bp
Taql
Sa3AI HaeIII
Cfol
Taql
Sau3A\ Hae III
Cfol
pv. malvacearum
RIATC, XcmA, XcmB, XcmC,
XcmD, XcmE, XcmF, XcmH,
XcmN
2
5
6
6
6
9
12
10
pv. manihotis
Xml25D
2
7
11
10
9
14
22
21
pv. papavericola
XP5
nd
nd
nd
nd
3
2
3
3
pv. pelargonii
XCP2, XCP10, XCP17, XCP36,
XCP54, XCP58, XCP60
nd
nd
nd
nd
3
2
3
2
pv. phaseoli A
81-19, 82-1, 85-6, XCPH4
2
7
12
10
9
14
22
19
B
82-17
2
5
6
7
7
9
10
7
pv. phaseoli 'Tuscans"
XP163A
4
5
15
9
7
9
11
12
pv. physalidicola
XP172
3
6
16
12
nd
nd
nd
nd
pv. poinsettiicola
B
X87, X202, X259
10
4
5
5
5
8
9
6
A
X352, 071-424
3
6
16
12
8
11
15
8
pv. pruni
X1219L, X1219S, X1220L,
X1220S
3
6
16
12
8
11
19
8
Continued on following page


90
distinct restriction patterns that corresponded to the grouping established previously on
the basis of pathogenic features (Fig. 4-11; Table 4-2). For instance, the strains of X.
campestris pv. citri of the citrus canker groups A, B, and C each produced
characteristic restriction patterns for all the different combinations of fragment-
restriction endonuclease tested (Chapter 6; Table 4-2). Similarly, strains ofX.
campestris pv. vesicatoria groups A and B also produced characteristic restriction
banding patterns (Table 4-2). The two groups of X. campestris pv. vesicatoria could be
differentiated by any combination hrp fragment/restriction endonuclease (Table 4-2).
Strains of X. campestris pv. dieffenbachiae that cause diseases in plants of the family
Araceae also produced characteristic restriction patterns (Fig. 4-11; Table 4-2). Among
the strains of X. campestris pv. dieffenbachiae, three RFLP groups were established on
the basis of restriction analysis of the hrp related fragments that corresponded to the
host of origin. The strains X422, X790, XI51, and XI272 were isolated from
Anthurium sp., strains X260 and X763 were isolated from Syngonium sp., and strains
X736, X738, X739, and X745 were isolated from Xanthosoma sigittifolium (Table 4-
2). Further, the strains from Anthurium sp. could be distinguished from the X.
sigittifolium strains only by comparison of the 1,075-bp fragment restricted with the
endonuclease Sau3A\ (Table 4-2).
The strains of X campestris pv. fici and X. campestris pv. poinsettiicola each
were divided in two different groups based on the distinct restriction banding patterns
for both /irp-related fragments (Fig. 4-12; Table 4-2). Further, sequence variability was
also observed within the groups, particularly for the X. campestris pv. fici group A
(Table 4-2). However, the banding patterns obtained for the three strains of this group
were very similar and they have several bands in common (Fig. 4-12; Table 4-2). The
most striking feature of these two pathovars is the similarity of the restriction banding


137
Restriction analysis
The 840- and 1,075-bp hrp gene cluster fragments amplified from strains of
different pathovars of X. campestris were each restricted with either Cfol, HaeIII,
Sau3Al, or Taql. The banding patterns for each set of fragments from the pathovars of
X. campestris included in this study were variable. The banding patterns of the 1,075
bp hrp gene cluster fragment amplified from strains of different pathovars ofX.
campestris restricted with the endonucleases Cfol and Haelll are presented in Fig. 6-3.
The banding patterns of the strains of X. campestris pv. citrumelo were very similar to
the patterns obtained forX. campestris pv. vesicatoria strain 75-3 (Fig. 6-3). Also, X
campestris pv. alfalfa, X. campestris pv. fici, and the strain XI98 of X. campestris
produced patterns similar to the strains of X. campestris pv. citrumelo with certain
combinations of fragment-restriction endonucleases (Fig. 6-3). X campestris pv.
bilvae, X campestris XCF, and the strains of X. campestris pv. citri also made up a
group with very similar banding pattern (Fig. 6-3). On the other hand, X. campestris
pv. maculifoliigardeniae had a more distinct restriction fragment profile (Fig. 6-3).
Although variability was also observed in the banding patterns obtained with the
endonucleases Sau3A\ and Taql, a characteristic pattern for each group or pathovar of
X. campestris was less evident for these two endonucleases (data not shown).
Similarly, restriction analysis of the 840-bp hrp fragment also produced a pattern of
variation for the different strains of X. campestris (data not shown).
Sixteen strains of A. campestris pv. citrumelo, representing all three
aggressiveness groups, were analyzed by restriction analysis of the amplified hrp
fragments. The banding patterns of five strains of the highly aggressive group of X.
campestris pv. citrumelo were identical to each other when restricted with either Cfol
(Fig. 6-4A), Haelll (Fig. 6-4B), Sau3Al or Taql (data not shown). For certain


194
or uniformity of the different taxa of xanthomonads assessed on the basis of restriction
analysis of hrp-related sequences apparently agrees very closely with the structure
established by using other methods. However, the restriction banding profiles
generated for the hrp-related fragments may be an easier and a more discriminating
approach for identification of plant pathogenic xanthomonads, compared to other
nucleic acid-based approaches, such as the genomic fingerprinting or restriction
fragment length polymorphism analysis by using random or specific DNA probes.
However, more extensive work is necessary to further characterize the hrp-related
sequence variation in other groups of plant pathogenic xanthomonads. Furthermore,
DNA-DNA hybridization data is also necessary to determine the consistency of the
groupings established on such a small region of the bacterial genome as the hrp genes.


193
homologous hrp region of X. campestris pv. begoniae, X. campestris pv.
dieffenbachiae, X. campestris pv. maculifoliigardeniae, X. campestris pv. manihotis,
and X. campestris pv. phaseoli A. Since there is no genetic and functional indication of
selection pressure regarding the hrp genes, the most likely hypothesis to explain the
source of the variability in the relatedness for these two regions of the genome of X.
campestris pv. vignicola remains in the origin of the hrpB and hrpC/D genes from
distinct ancestors. The coexistence in the same biological niche may have provided
opportunities for the lateral movement of this region of bacterial genome between
xanthomonads. The horizontal movement of genetic material between bacteria is a
common and important mechanism in bacterial evolution (Krawiec and Riley, 1990).
Furthermore, the resemblance at the protein level of the hrp genes of the xanthomonads
with genes involved in the secretion of pathogenicity factors of genetically distant
organisms, such as the animal pathogens of the genus Yersinia (Fenselau et al., 1992;
Gough et al., 1992), is also intriguing. If this means a convergent functional evolution
or a common bacterial ancestor remains to be clarified. Nevertheless, the phylogeny of
the hrp genes of plant pathogenic xanthomonads may provide a framework and a
rational basis through which origins and differentiation of the xanthomonads may be
assessed.
The classification of plant pathogenic bacteria at pathovar level was not based
initially on the genetic or other intrinsic characteristic of the organism but rather on the
host from which the bacteria were isolated (Bradbury, 1984; Dye et al., 1980).
Furthermore, comprehensive studies on the genetic and phenotypic characteristics have
supported the existence of a high degree of diversity among strains of a given pathovar
ofX. campestris. The genetic analysis of the /2/77-related sequences also indicates a
large diversity among the plant pathogenic xanthomonads. Furthermore, the diversity


Table 7-3Continued
Amplification of fragments related to the hrpB of Xanthomonas campestris pv. vesicatoria by 30 cycles of polymerase chain
reaction.
bGOK, G. O'Keefe, Department of Agriculture, State of Georgia, Tifton, GA; JFW, J. F. Wang, The Asian Vegetable
Research and Development Center Tainan, Taiwan; MM, M. Meadows, Rogers NK, Naples, FL; RST, R. E. Stall, University
of Florida, Gainesville.
cPresence of plant pathogenic xanthomonads was determined by isolation of bacteria or by detection by amplification of hrp-
related fragment from diseased tissue.
d- negative result; + positive result; (+), weak positive result; nd, not determined.


239
Qhobela, M, and Claflin, L. E. 1988. Characterization of Xanthomonas campestris
pv. pennamericanum pv. nov., causal agent of bacterial leaf streak of pearl millet. Int.
J. Syst. Bacteriol. 38: 362-366.
Qhobela, M, and Claflin, L. E. 1992. Eastern and southern african strains of
Xanthomonas campestris pv. vasculorum are distinguishable by restriction fragment
length polymorphism of DNA and polyacrylamide gel electrophoresis of membrane
proteins. Plant Pathol. 41:113-121.
Qhobela, M, Leach, J. E., Claflin, L. E., and Pearson, D. L. 1991. Characterization of
strains of Xanthomonas campestris pv. holcicola by PAGE of membrane proteins and
by REA and RFLP analysis of genomic DNA. Plant Dis. 75: 32-36.
Roth, D. A. 1989. Review of extraction and isolation methods. Pages 3-8 in:
Detection of Bacteria in Seeds and Other Planting Material. A.W. Saettler, N.W.
Schaad, and D.A. Roth, eds. American Phytopathological Society Press, St. Paul, MN.
Saettler, A. W., Schaad, N. W., and Roth, D. A., eds. 1989. Detection of Bacteria in
Seeds and Other Planting Material. American Phytopathological Society Press, St.
Paul, MN. 122pp
Sambrook, J., Fritsch, E. F., and Maniatis, T. 1989. Molecular Cloning: A Laboratory
Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 1546 pp.
SAS Institute Inc. 1988. SAS/STAT User's Guide. SAS Institute Inc., Cary, NC.
1028 pp.
Sasser, M. 1990. Identification of bacteria through fatty acid analysis. Pages 199-204
in: Methods in Phytobacteriology. Z. Klement, K. Rudolph, and D. Sands, eds.
Akadmiai Kiad, Budapest, Hungary.
Schaad, N.W. 1979. Serological identification of plant pathogenic bacteria. Annu.
Rev. Phytopathol. 17:123-147.
Schaad, N.W. 1982. Detection of seedbome bacterial plant pathogens. Plant Dis. 66:
885-890.
Schaad, N. W., ed. 1988. Laboratory Guide for Identification of Plant Pathogenic
Bacteria. American Phytopathological Society Press, St. Paul, MN. 164 pp.


108
pv. holcicola (I)
-pv. vesicatoria Bill)
96
X. campestris X52 (1)
pv. pomsettiicola B (3)
pv. glycines B (I)
pv. fici B (2)
pv. raphani B in
pv. raphani A (3)
pv. papavericola (1)
pv. pelargonii (7)
pv. carotae B (I)
pv. campestris (9)
pv. armoraciae (3)
pv. incanae (1)
pv. gardneri (3)
pv. taraxaci (1)
. vitinns C (2)
carotae A (6)
pruni (4)
citrurnelo
pv. citrurnelo iI)
pv. citrurnelo (1)
pv. alfalfae (I)
pv. alfalfae (I)
pv. poinsettiicola A (2)
X. campestris X198
pv. fici A (I)
pv. citrurnelo (2)
v. citrurnelo (2),
pv. vesicatoria A
pv. vesicatoria A
pv. vesicatoria A (24)
pv. citrurnelo ill
pv. citrurnelo (I)
pv. citrurnelo (I)
pv. fici A
pv. fici A .
pv. citrurnelo (5)
99
(2)
81
ill
pv. vitians B (1)
pv. malvacearum (9)
pv. phaseoli B (1)
pv. glycines A (6)
pv. citri A (6)
phaseoli Tuscans" (I)
pv. vignicola (3)
pv. citri C (6)
pv. citri B (7)
X. campestris XCF (1)
bilvae (1^
91
. pv. dieffenbachiae A (4)
pv. djefTenbachiae A (4)
pv. phaseoli A (4)
pv. vitians A (3)
pv. manihotis (1)
pv. begoniae (I)
pv. maculifoliigardeniae (I)
pv. begoniae (1)
pv. begoniae (7)
pv. dieffenbachiae B (2)


40
M 1 23 456 78 9 10 11 12
Fig. 3-2. Amplification of fragments of the hrp gene cluster from Xanthomonas
campestris pv. vesicatoria 75-3. The following DNAs were used: 75-3 (lanes 1, 5, and
9), plasmid pXV9 (lanes 2, 6, and 10), plasmid pXV5.5 (lanes 3, 7, and 11), and
plasmid pXV5.1 (lanes 4, 8, and 12). Lanes: M, phage X restricted with EcoRI and
Hindlll; 1 to 4, amplification with primers RST9 and RST10; 5 to 8, amplification with
primers RST2 and RST3; 9 to 12, amplification with primers RST21 and RST22.
Molecular sizes are given in base pairs.


231
Felsenstein, J. 1985. Confidence limits on phylogenies: An approach using the
bootstrap. Evolution 39:783-791.
Felsenstein, J. 1991. Phylogeny Inference Package, Version 3.4. Department of
Genetics, University of Washington, Pullman.
Felsenstein, J. 1993. Phylogeny Inference Package, Version 3.5. Department of
Genetics, University of Washington, Pullman.
' Fenselau, S., Balbo, I., and Bonas, U. 1992. Determinants of pathogenicity in
Xanthomonas campestris pv. vesicatoria are related to proteins involved in secretion in
bacterial pathogens of animals. Mol. Plant-Microbe Interact. 5: 390-396.
Figurski, D., and Helinski, D. R. 1979. Replication of an origin containing derivative
of plasmid RK2 dependent on a plasmid function provided in trans. Proc. Natl. Acad.
Sci. USA 76: 1648-1979.
Fitch, W. M., and Margoliash, E. 1967. Construction of phylogenetic trees. Science
155:279-284.
Franken, A. A. J. M., Zilverentant, J. F., Boonekamp, P. M., and Schots, A. 1992.
Specificity of polyclonal and monoclonal antibodies for the identification of
Xanthomonas campestris pv. campestris. Neth. J. PI. Path. 98: 81-94.
Gabriel, D. W., Kingsley, M. T., Hunter, J. E., and Gottwald, T. 1989. Reinstatement
of Xanthomonas citri (ex Hasse) and X. phaseoli (ex Smith) to species and
reclassification of all A! campestris pv. citri strains. Int. J. Syst. Bacteriol. 39: 14-22.
Gabriel, D. W., Hunter, J. E., Kingsley, M. T., Miller, J. W., and Lazo, G. R. 1988.
Clonal population structure of Xanthomonas campestris and genetic diversity among
citrus canker strains. Mol. Plant-Microbe Interact. 1: 59-65.
Garde, S., and Bender, C. L. 1991. DNA probes for detection of copper resistance
genes in Xanthomonas campestris pv. vesicatoria. Appl. Environ. Microbiol. 57:
2435-2439.
Gardner, M. W., and Kendrick, J. B. 1921. Bacterial spot of tomato. J. Agr. Res. 21:
123-156.
Gardner, M. W., and Kendrick, J. B. 1923. Bacterial spot of tomato and pepper.
Phytopathology 13:307-315.


63
s of annealing at 61C, and 45 s of extension at 72C for the primers RST21 and
RST22. The last extension step for both profiles was extended to 5 min.
The amplified DNA sequences were detected by electrophoresis in 0.9%
agarose gels in TAE buffer (40 mM Trisacetate, 1 mM EDTA, pH 8.2) according to
standard procedures (Ausubel et al., 1987; Sambrook et al., 1989).
Restriction endonuclease analysis
The amplified DNA fragments were restricted with either endonucleases Cfol,
HaeIII, Saw3AI, or Taql, under conditions specified by the manufacturer (Promega).
The restricted fragments were separated by electrophoresis in 4% agarose gels (3%
NuSieve and 1% Seakem GTG [FMC BioProducts, Rockland, ME]) in TAE buffer at 8
V/cm. The gel was stained with 0.5 pg of ethidium bromide per ml for 40 min and
then destained in 1 mM MgS04 for 1 hour and photographed over a UV
transilluminator (Fotodyne Inc., New Berlin, WI).
Establishment of the restriction fragment patterns
The DNA restriction banding patterns were determined by visual examination
of the /irp-related fragments digested with each of the four endonucleases. The
banding patterns obtained for each fragment were used for comparison with the
patterns obtained for homologous fragment amplified from strains of a given pathovar
and also from strains of different pathovars of X. campestris and of the species X.
fragariae. A restriction profile number was assigned to each unique DNA restriction
banding pattern. This restriction profile number was then used to determine the RFLP
group of each strain or group of strains of the plant pathogenic xanthomonads included
in this study.


168
M 1 2 3 4 5 6 7
1 375
564-
Fig. 7-1. Amplification of the 840-bp fragment of the complementation group B of the
hrp gene cluster from seeds of tomato containing 106 CFU/ml of Xanthomonas
campestris pv. vesicatoria 75-3 and different amounts of PVPP. Lanes: M, phage X
restricted with £coRI and HindiII; 1, no PVPP nor sodium ascorbate; 2, no PVPP; 3,
1% of PVPP; 4, 2% of PVPP; 5, 3% of PVPP; 6, 4% of PVPP; 7, 5% of PVPP.
Molecular sizes are given in bases.


179
Fig. 7-6. Restriction analysis of the 1,075-bp DNA fragments related to the hrpC/D
complementation group and amplified from strains of Xanthomonas campestris isolated
from pepper and tomato seed and seedling, and restricted with the endonucleases Taql
(Lanes 2 to 10) and HaeIII (Lanes 12 to 20). Lanes: 1 and 11, phage k restricted with
Pstl; 2 and 12, 157; 3 and 13, 524A-1; 4 and 14, 75-0-3; 5 and 15, 7B-0-1; 6 and 16,
T1087; 7 and 17, 524A-2; 8 and 18, l-A-0-1; 9 and 19, DM-1; 10 and 20, X. campestris
pv. vesicatoria 75-3. Molecular sizes are given in number of bases.


183
lots out of the 11 which produced positive results in the DNA amplification procedure.
This is not surprising because X. campestris pv. vesicatoria has the capability to survive
for long periods of time in seeds and to initiate disease (Bashan et al., 1982; Gardner and
Kendrick, 1921, 1923).
A diverse group of xanthomonads was found associated with pepper and tomato
seeds. The majority of the 56 strains examined have the basic characteristics of plant
pathogenic xanthomonads. They produced pathogenic or hypersensitive reaction on
plants in pathogenicity tests, and DNA sequences related to the hrp genes were amplified
from their genome. On the contrary, seven strains did not produce reaction on plants, and
they failed to produce DNA sequences similar to the hrp genes in DNA amplification
assay. These strains resemble the opportunistic xanthomonads that have been described
in association with pepper and tomato transplants (Gitaitis et ah, 1987, 1992). Among
the plant pathogenic xanthomonads, only nine strains were positively identified as X.
campestris pv. vesicatoria, the pepper and tomato pathogen (Table 7-1). Another group
included potential plant pathogenic strains that were not pathogenic to pepper and tomato
but instead produced a hypersensitive reaction on these plants. Based on the analyses of
DNA sequences related to hrp genes and fatty acid composition, they were highly related
to different pathovars of X. campestris, but not to X campestris pv. vesicatoria.
Furthermore, the identification of the strains based on restriction profiles of the DNA
sequences related to the hrp genes agreed very closely with the identification by fatty acid
analysis. Some of these strains were identified as X. campestris pv. campestris though
their pathogenicity to cabbage was not determined. Strains similar to X. campestris pv.
campestris and pathogenic to cabbage have been reported previously on pepper and
tomato transplants (Gitaitis et al., 1987). Other strains were similar to X. campestris pv.
raphani which has also been found commonly in association with pepper and tomato


233
Graham, J. H., Hartung, J. S., Stall, R. E., and Chase, A.R. 1990. Pathological,
restriction-fragment length polymorphism, and fatty acid profile relationships between
Xanthomonas campestris from citrus and noncitrus hosts. Phytopathology 80: 829-
836.
Haefele, D., and Webb, R. 1982. The use of ultrasound to facilitate the harvesting and
quantification of epiphytic and phytopathogenic microorganisms. (Abstr.)
Phytopathology 72: 947.
Hartung, J. S. 1992. Plasmid-based hybridization probes for detection and
identification of Xanthomonas campestris pv. citri. Plant Dis. 76: 889-893.
Hartung, J. S., and Civerolo, E. L. 1987. Genomic fingerprints of Xanthomonas
campestris pv. citri strains from Asia, South America, and Florida. Phytopathology
77: 282-285.
Hartung, J. S., and Civerolo, E. L. 1989. Restriction fragment length polymorphisms
distinguish Xanthomonas campestris strains isolated from Florida citrus nurseries from
X. c. pv. citri. Phytopathology 79: 793-799.
Hartung, J. S., and Civerolo, E. L. 1991. Variation among strains of Xanthomonas
campestris causing citrus bacterial spot. Plant Dis. 75:622-626.
Hartung, J. S., Daniel, J. F., and Pruvost, O. P. 1993. Detection of Xanthomonas
campestris pv. citri by the polymerase chain reaction method. Appl. Environ.
Microbiol. 59: 1143-1148.
Hayward, A. C. 1993. The hosts of Xanthomonas. Pages 1-119 in: Xanthomonas.
J.G. Swings, and E.L. Civerolo, eds. Chapman & Hall, London, United Kingdom.
Higgis, B. B. 1922. The bacterial spot of pepper. Phytopathology 12:501-516.
Hildebrand, D. C., Hendson, M., and Schroth, M. N. 1993. Usefulness of nutritional
screening for the identification of Xanthomonas campestris DNA homology groups and
pathovars. J. Appl. Bacteriol. 75: 447-455.
Hildebrand, D. C., Palleroni, N. J., and Schroth, M. N. 1990. Deoxyribonucleic acid
relatedness of 24 xanthomonad strains representing 23 Xanthomonas campestris
pathovars and Xanthomonas fragariae. J. Appl. Bacteriol. 68: 263-269.


18
Mooter et al., 1987b), DNA-DNA hybridization (Kersters et al., 1989; Vauterin et al.,
1992), SDS-PAGE (Van den Mooter et al., 1987b; Vauterin et al., 1992), and fatty acid
profiles (Vauterin et al., 1992; Yang et al., 1993) have failed to consistently
differentiate strains of X. campestris pv. graminis from other pathovars o X.
campestris which cause diseases on cereals and grasses, such as X. campestris pv.
arrhenatheri, X campestris pv. cerealis, X campestris pv. hordei, X. campestris pv.
phlei, X campestris pv. phleipratensis, X campestris pv. poae, X. campestris pv.
secalis, X. campestris pv. translucens, and X. campestris pv. undulosa. Furthermore,
the misnamed species Pseudomonas gardneri, P. vitiswoodrowii, and P.
mangiferaeindicae were indistinguishable from X. campestris strains on the basis of
phenotypic features (Dye, 1966; Hildebrand et al., 1993; Van den Mooter and Swings,
1990). The latter also agrees with studies of DNA-rRNA hybridization (De Vos et al.,
1985), DNA-DNA hybridization (Hildebrand et al., 1990; Palleroni et al., 1993), and
fatty acid composition (Yang et al., 1993).
X campestris contains by far the largest group of strains within the genus
Xanthomonas, consisting of at least 125 different pathovars (Bradbury, 1984). Despite
certain variations in phenotypic characteristics (Hildebrand et al., 1993), strains ofY
campestris form a rather indistinguishable group (Burkholder and Starr, 1948; Dye,
1962; Van den Mooter and Swings, 1990). In a limited number of cases, however,
distinctive phenotypic traits have been linked to the grouping based on pathogenicity to
a given host species. Two of these cases are the X campestris pv. manihotis and X
campestris pv. cassavae, which cause diseases on cassava. All strains of each pathovar
formed a single group based on phenotypic features (Van den Mooter and Swings,
1990; Van den Mooter et al., 1987a). Correlation between phenotypic features and
pathogenic specialization was also established for the xanthomonads that cause wilt


ACKNOWLEDGMENTS
I would like to express my sincere gratitude to Dr. Robert Stall for his constant
support and encouragement during the course of this doctoral program. His vast
expertise and willing to offer all necessary support were invaluable. I would also like
to thank the other members who served on my supervisory committee, Dr. Jeffry Jones,
Dr. James Preston, and Dr. Daryl Pring, for their support and suggestions. I am
indebted to Dr. Jerry Bartz for reviewing this dissertation.
Appreciation is extended to the Instituto Agronmico do Paran (IAPAR), which
gave permission for this study leave and provided financial support, and to the
Conselho Nacional de Pesquisa (CNPq), which has granted a scholarship for this
graduate program.
I am indebted to Dr. Ulla Bonas of the Institut des Sciences Vgtales, Centre
National de la Recherche Scientifique, Gif-sur-Yvette, France. She provided the DNA
sequences of the hrp genes that were the keystone of this work. Appreciation is
extended to Gail Somodi, Agricultural Research and Education Center in Bradenton,
FL, who performed the ELISA tests. Jay Harrison, IFAS consulting Division of the
Department of Statistics, was very helpful with the SAS programming.
Special thanks are given to Jerry Minsavage for his invaluable technical assistance
and friendship, and to Cheri Hodge for the fatty acid analysis and her friendship.
Finally, I do not have enough words to thank my family, Claudia, my wife, and our
two sons, Ruy and Rafael. They have provided unlimited understanding and support
during these years of graduate studies. They have shared their love, their dreams, and
11


188
840- and 1,075-bp DNA fragments with frequent-cutting endonucleases indicated the
presence of characteristic sequence variation. A study carried out to determine the
extension of the diversity of /irp-related sequences of 192 strains of plant pathogenic
xanthomonads revealed that only a few taxa of the xanthomonads are homogeneous.
This includes X. fragariae, X. campestris pv. begoniae, X campestris pv. campestris,
X. campestris pv. malvacearum, and X campestris pv. pelargonii. The homogeneous
population structure of these taxa has also been supported by using different methods,
such as DNA-DNA hybridization, fatty acid composition, and SDS-PAGE of proteins
(Vauterin et al., 1990a, 1991a; Yang et al., 1993). On the contrary, the majority of the
pathovars of X campestris seem to comprise an heterogeneous group of strains as
determined by the analysis of the /?rp-related sequences. Whereas strains of some
pathovars formed distinct and uniform subgroups, e.g. X. campestris pv. citri, X
campestris pv. dieffenbachiae, and X campestris pv. vesicatoria, other pathovars were
composed of a highly diverse group of strains. The latter may be well characterized by
the population structure determined for X campestris pv. citrumelo. The 16 strains of
X. campestris pv. citrumelo included in the study were separated in nine different
groups on the basis of the restriction banding profiles generated for the two hrp-related
fragments digested with four different endonucleases. The diverse nature of the strains
of X. campestris pv. citrumelo has been reported previously based on DNA homology,
restriction fragment length polymorphism analysis, fatty acid composition, and SDS-
PAGE of proteins (Egel et al., 1991; Gabriel et al., 1988, 1989; Graham et al., 1990;
Hartung and Civerolo, 1987, 1989; Vauterin et al., 1991a).
The most striking feature of the restriction analysis of the amplified /zrp-related
fragments is the possibility for differentiation of almost all the pathovars and groups of
X. campestris included in the study on the basis of the restriction banding profile of the


148
pv. citruinelo F94
pv. citruinelo F306
pv. citrumelo F100
pv. vesicatoria 75-3
pv. citruinelo 534
pv. citruinelo F348
pv. fici X151
pv. citruinelo FI
pv. citrumelo F86
pv. alfalfae 82-1
pv. citrumelo F311
pv. citrumelo F6
X. campes tris X198
X. campestris XCF
pv. bilvae XCB
pv. citri 339
pv. citri B84
pv. citri 9771
pv. maculifoliigardeniae X22j
j
.030 .025 .020 .015 .010 .005 0
Genetic distance


89
M123456789 10
Fig. 4-12. Restriction analysis of the 1,075 bp DNA fragments of the hrp gene cluster
amplified from strains of Xanthomonas campestris pv. fici (Lanes 1 to 5) and
Xanthomonas campestris pv. poinsettiicola (Lanes 6 to 10) and restricted with Taql.
Lane M, phage X restricted with Pstl. Lanes 1, X125; 2, XI51; 3, X208; 4, X212; 5,
X702; 6, X87; 7, X202; 8, X352; 9, 071-424; \0,X. campestris pv. vesicatoria 75-3.
Molecular sizes are given in bases.


243
Vauterin, L., and Vauterin, P. 1992. Computer-aided objective comparison of
electrophoresis patterns for grouping and identification of microorganisms. Eur.
Microbiol. 1:37-41.
Vauterin, L., Yang, P., Hoste, B., Pot, B., Swings, J., and Kersters, K. 1992.
Taxonomy of xanthomonads from cereals and grasses based on SDS-PAGE of
proteins, fatty acid analysis and DNA hybridization. J. Gen. Microbiol. 138: 1467-
1477.
Vauterin, L., Yang, P., Hoste, B., Vancanneyt, M., Civerolo, E.L., Swings, J., and
Kersters, K. 1991b. Differentiation of Xanthomonas campestris pv. citri strains by
sodium dodecyl sulfate-polyacrylamide gel electrophoresis of proteins, fatty acid
analysis, and DNA-DNA hybridization. Int. J. Syst. Bacteriol. 41: 535-542.
Verdier, V., Dongo, P., and Boher, B. 1993. Assessment of genetic diversity among
strains o Xanthomonas campestris pv. manihotis. J. Gen. Microbiol. 139: 2591-2601.
Vemiere, C., Pruvost, O., Civerolo, E. L., Gambin, O., Jacquemoud-Collet, J. P., and
Luisetti, J. 1993. Evaluation of the Biolog substrate utilization system to identify and
assess metabolic variation among strains of Xanthomonas campestris pv. citri. Appl.
Environ. Microbiol. 59: 243-249.
Volcani, Z. 1969. The effect of mode of irrigation and wind direction on disease
severity caused by Xanthomonas vesicatoria on tomato in Israel. Plant Dis. Reptr. 53:
459-461.
Voloudakis, A. E., Bender, C. L., and Cooksey, D. A. 1993. Similarity between
copper resistance genes from Xanthomonas campestris and Pseudomonas syringae.
Appl. Environ. Microbiol. 59: 1627-1634.
Wayne, L. G., Brenner, D. J., Colwell, R. R., Grimount, P. A. D., Kandler, O.,
Krichevsky, M. I., Moore, L. H., Moore, W. E. C., Murray, R. G. E., Stackebrandt, E.,
Starr, M. P., and Truper, H. G. 1987. Report of the ad hoc committee on
reconciliation of approaches to bacterial systematics. Int. J. Syst. Bacteriol. 37: 463-
464.
Weber, K., and Osborn, M. 1960. The reliability of molecular weight determinations
by dodecyl sulfate-polyacrylamide gel electrophoresis. The Journal of Biological
Chemistry 244: 4406-4412.
White, H. E. 1930. Bacterial spot of radish and turnip. Phytopathology 20: 653-662.


Table D-2Continued
Species/Pathovar
Species/Pathovar
50
51
52
53
54
55
56
58
59
61
50. pv. vesicatoria A
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
51. pv. vesicatoria A
.
.9796
.5000
.9167
.7556
.5652
.7917
.6250
.4000
52. pv. vesicatoria A
.5128
.9362
.7273
.5778
.8085
.5957
.3636
53. pv. vesicatoria B
.5263
.5714
.5556
.4211
.5263
.4000
54. pv. vignicola
.7907
.5909
.8696
.6087
.3721
55. pv. vitians A
.5854
.6977
.6047
.5000
56. pv. vitians B
.5000
.8182
.4390
58. X campestrisX 198
.5217
.2791
59. X. campestris XCF
61. A! fragariae
*


.4186
to
to
L/t


153
epidemiological studies. Furthermore, the use of oligonucleotide primers for the hrp
gene cluster region provide certainty for identification of strains X. campestris that may
not be possible with other methods.


146
56
100
100
59
16
15
25
17
54
35
33
pv. citri 9771
98 | pv. citri 339
* pV. citri B84
79 | X. campes iris XCF
pv. bilvae XCB
pv. citrumelo F100
pv. vesicatoria 75-3
pv. citrumelo F306
pv. citrumelo F94
L pv. citrumelo F86
X. campes tris XI98
pv. alfalfae 82-1
pv. citrumelo F6
pv. citrumelo F311
pv. citrumelo F348
87 pv. fici X151
pv. citrumelo FI
pv. citrumleo 534
pv. maculifoliigardeniae X22j
13
441 '
t


61
Materials and Methods
Bacterial strains and culture conditions
The bacterial strains used in this study, their taxonomic designation, and their
sources are listed in the Appendix A. All the bacterial strains had previously been
identified as members of Xanthomonas spp. by fatty acid analysis (Nancy C. Hodge,
personal communication). The strains ofX. campestris were grown on nutrient agar
(Becton Dickson, Cockeysville, MD). Strains of X. albilineans and X fragariae were
grown on Wilbrink's medium (Koike, 1965). Broth cultures were grown 24 hours on a
rotatory shaker (150 rpm) at 28C. All strains were streaked on appropriate media and
cultures were obtained from single colonies. The strains were stored and maintained in
tap water at room temperature, or in 30% glycerol at -70C, or both.
DNA extraction
Total bacterial genomic DNA was prepared by using the phenol-chloroform
extraction procedure described by Ausubel et al. (1987) with minor modifications.
Bacterial cells were harvested by centrifuging for 2 min at 16,000 g. The pellet was
washed once in 1 ml of distilled water, and then resuspended in 567 pi of TE buffer (10
mM Tris-Cl, pH 8.0; 1 mM EDTA, pH 8.0). Lysis solution containing Proteinase K
(Boehringer Mannheim, Indianapolis, IN) and sodium dodecyl sulfate (SDS) (Sigma,
St. Louis, MO) were added to final concentrations of 100 pg/ml and 0.5%,
respectively. The samples were incubated for 1 h at 37C. Sodium chloride and
hexadecyltrimethyl ammonium bromide (Sigma) were added for a final concentration
of 0.7 M and 1%, respectively, and incubated for 10 min at 65C. DNA extraction was


131
vacuum for 20 min and then the pellet was redissolved in 100 pi of TE and stored at
4C.
DNA amplification
The two sets of oligonucleotide primers used in this study were designed based
on nucleotide sequences of the hrp gene cluster ofX. campestris pv. vesicatoria (U.
Bonas, personal communication). The two primers RST2 and RST3 delineate an 840-
bp region and primers RST21 and RST22 delineate a 1,075-bp region of the hrp
complementation groups B and C/D of X. campestris pv. vesicatoria, respectively
(Chapter 3). Oligonucleotide primers were synthesized with a model 394 DNA
Synthesizer (Applied Biosystems, Foster City, CA) by the DNA Synthesis Laboratory,
University of Florida, Gainesville.
DNA was amplified in a total volume of 50 pi. The reaction mixture contained
5 pi of 10X buffer (500 mM KC1, 100 mM TrisCl [pH 9.0 at 25C], 1% Triton X-
100), 1.5 mM MgCl2, 200 pM of each deoxynucleotide triphosphate (Boehringer
Mannheim), 25 pmol of each primer, and 2.5 units of Taq polymerase (Promega,
Madison, WI). The amount of template DNA added was 100 ng of purified total
bacterial DNA The reaction mixture was covered with 50 pi of light mineral oil. A
total of 30 amplification cycles were performed in an automated thermocycler (MJ
Research, Watertown, MA). Each cycle consisted of 30 s of denaturation at 95C, 30 s
of annealing at 62C, and 45 s of extension at 72C for the primers RST2 and RST3
and 30 s at 95C, 45 s at 61 C and 1.5 min at 72C, respectively, for the primers
RST21 and RST22. The last extension step was extended to 5 min.
Amplified DNAs were detected by electrophoresis in 0.9% agarose gels in TAE
buffer (40 mM Tris-acetate, 1 mM EDTA, pH 8.2) at 5 V/cm of gel (Sambrook et al.,


CHAPTER 8
SUMMARY AND CONCLUSIONS
The plant pathogenic xanthomonads form a diverse group of usually yellow-
pigmented bacteria that occur worldwide and cause disease on many plants. The
differentiation of these bacteria at the subgeneric level is by the capability of the
bacterial strain to cause characteristic disease or by reference to their host range (Dye et
al., 1980; Vauterin et al., 1990a; Young et al., 1992). In fact, it is almost impossible to
differentiate the xanthomonads by biochemical and physiological features without
knowing their plant hosts (Bradbury, 1984; Dye, 1962; Holt et al., 1994; Schaad, 1988;
Van den Mooter and Swings, 1990). On the contrary, the genomic relationship among
plant pathogenic xanthomonads has revealed a group of bacteria with diverse genetic
background (Hildebrand et al., 1990; Palleroni et al., 1993; Vauterin et al., 1993). In
this work, I investigated a specific region of the bacterial genome, the DNA sequences
related to the hypersensitive reaction and pathogenicity (hrp) genes, for differentiation
of the xanthomonads and also for the establishment of the evolutionary relationships
among them.
Three sets of oligonucleotide primers selected from the DNA sequence of the
hrp genes of X campestris pv. vesicatoria were tested for amplification of DNA
fragments from different plant pathogenic xanthomonads. Two sets of primers, RST2
plus RST3 and RST9 plus RST10, are specific for the hrpB, and the set RST21 plus
RST22 is specific for a region of the hrpC/D of X. campestris pv. vesicatoria 75-3.
DNA fragments were amplified from X fragariae and from 31 of 33 plant pathogenic
taxa oiX. campestris tested, which comprise at least 28 different pathovars of this
186


75
Fig. 4-8. Restriction profiles established for the 1,075-bp /irp-related fragment
amplified from different strains of plant pathogenic Xanthomonas spp. and restricted
with the endonuclease Taql. Lane M, phage X restricted with Pstl. Molecular sizes are
given in bases. See Table 4-2 for identification of strains of Xanthomonas spp. for each
restriction pattern.


Table D-1 Continued
Species/Pathovar Species/Pathovar
42
44
45
46
47
48
49
50
51
52
1. pv. alfalfae
.5217
.3256
.9167
.9388
.4091
.4091
.3636
.7917
.7660
.7660
2. pv. alfalfae
.5217
.3256
.8750
.8980
.4091
.4091
.3636
.7500
.7234
.7234
3. pv. armoraceae
.2381
.4615
.2727
.3111
.4500
.4500
.8500
.2273
.2326
.2326
4. pv. begoniae A
.2381
.2564
.5909
.4889
.3500
.3500
.3500
.5000
.5116
.5116
5. pv. begoniae A
.2727
.2927
.5652
.5532
.3810
.3810
.3810
.4783
.5333
.5333
6. pv. begonia B
.2727
.3415
.5652
.5532
.3810
.3810
.4286
.4783
.5333
.5333
7. pv. bilvae
.6957
.3721
.7083
.7347
.4091
.4091
.3182
.6250
.6383
.6383
8. pv. campestris
.2381
.4615
.2727
.3111
.4500
.4500
.8500
.2273
.2326
.2326
9. pv. carotae A
.3333
.5128
.3182
.3556
.5000
.5000
.8000
.2727
.2791
.2791
10. pv. carotae B
.2381
.4615
.2727
.3111
.4500
.4500
.8500
.2273
.2326
.2326
11. pv. citri A
.8261
.2791
.5833
.6122
.4091
.4091
.3636
.5417
.5532
.5532
12. pv. citri B
.8261
.3256
.4583
.5306
.3636
.3636
.3182
.4583
.4681
.4681
13. pv. citri C
.8261
.3256
.4583
.5306
.3636
.3636
.3182
.4583
.4681
.4681
14. pv. citrumelo
.5106
.3182
.7347
.7600
.3556
.3556
.3556
.7755
.7500
.7500
15. pv. citrumelo
.5532
.3182
.8980
1.000
.4000
.4000
.3556
.7755
.7500
.7500
16. pv. citrumelo
.5652
.2791
.9167
.8163
.3182
.3182
.3636
.9583
.8511
.8511
17. pv. citrumelo
.5652
.2791
.9167
.8163
.3182
.3182
.3636
.9583
.8511
.8511
18. pv. citrumelo
.4783
.2791
1.000
.8980
.3636
.3636
.3636
.8750
.7660
.7660
19. pv. citrumelo
.4889
.2857
.9362
.8333
.3256
.3256
.3721
.8511
.7391
.7391
20. pv. citrumelo
.3913
.3256
.7917
.6939
.3636
.3636
.3636
.7500
.7660
.7660
21. pv. citrumelo
.5217
.3721
.7917
.8980
.3636
.3636
.3636
.7083
.8085
.8085
22. pv. citrumelo
.4783
.3256
.7917
.8163
.4091
.4091
.3182
.7083
.8085
.8085
Continued on following page


Fig. 6-6. Unrooted tree for 19 strains of Xanthomonas campestris inferred from
restriction analysis data of DNA fragments related to the complementation groups B
and C/D of the hrp genes of X. campestris pv. vesicatoria generated by the BOOT
procedure from the PHYLIP computer package by using the Wagner parsimony
criterion. Numbers at each node indicate the bootstrap percentages from 100 samples.
Bootstrap values less than 50 indicate that the assemblage is not well supported by the
data.


55
The identification of strains of X. campestris at the pathovar level is difficult
even by using different techniques, such as fatty acid analysis (Chase et al., 1992;
Graham et al., 1990), serology (Alvarez et al., 1991), metabolic profile (Chase et al.,
1992; Van den Mooter and Swings, 1990), and SDS-PAGE of proteins (Vauterin et al.,
1991a); thus, the restriction analysis of amplified hrp related fragments may be a
valuable tool for identification of subgroups of plant pathogenic strains and pathovars
of campestris. For example, restriction analysis with frequent-cutting endonucleases
produced a characteristic restriction pattern for the 840- or 1,075-bp DNA fragments
amplified from the Xanthomonas spp. included in this work; this appears to be highly
conserved within each group of certain plant pathogenic xanthomonads (Chapter 4). In
this way, RFLP profiles of a particular hrp region could be established for each plant
pathogenic group of xanthomonads thus facilitating the identification and classification
of these bacterial strains.
The use of oligonucleotide primers provides a sensitive and specific tool for
detection of DNA by amplification. Theoretically, the limit of detection of an
amplifiable DNA sequence is estimated to be as low as one single target cell in the
reaction mixture (Steffan and Atlas, 1991). In our studies, we were able to find a
detectable signal for as low as 0.25 pg of total bacterial DNA. This level of sensitivity
is comparable to those obtained by others without the use of any technique to enhance
the signal (Pickup, 1991; Steffan and Atlas, 1988). Furthermore, X. campestris pv.
vesicatoria could be detected in plant samples containing less than 100 CFU/ml,
without prior enrichment or cultivation of the organism (Chapter 7). In addition to the
sensitivity of the technique, the specificity of the oligonucleotide primers to plant
pathogenic xanthomonads certainly assures selectivity against background nontarget
microorganisms which are always present in the samples.


185
reactions on plants under artificial inoculations. Therefore, additional studies are
necessary to determine the role of these xanthomonads found in the seeds of pepper and
tomato in the subsequent development of bacterial diseases on seedlings and plants.
In conclusion, a procedure based on amplification of hrp-related fragments was
examined for sensitive and specific detection of plant pathogenic xanthomonads
associated with pepper and tomato seeds. The complexity of the xanthomonads
community found on seeds was assessed, and confirmed previous results obtained by
other researchers using different methods (Gitaitis et al., 1987, 1992). Furthermore, the
combined sensitivity and specificity of the DNA amplification approach certainly makes
the procedure an indispensable tool for the detection and identification of plant
pathogenic xanthomonads in seeds and other propagative materials. Although the
different xanthomonads could not be distinguished by the size of the amplified DNA
fragment, the restriction fragment length polymorphism present in the DNA sequences
related to the hrp genes obtained from different strains of Xanthomonas was valuable for
differentiation of these plant pathogenic xanthomonads. Furthermore, the comparison of
the restriction profile to predetermined profiles of several pathovars of X. campestris
(Chapter 4) allowed the tentative and precise identification of these plant pathogenic
xanthomonads.


174
xanthomonads in naturally contaminated seeds. Plant pathogenic xanthomonads in
pepper and tomato seed lots were detected consistently by amplification of the hrpB
fragments in DNA extracted from washes of 11 seed lots out of the 15 tested ( Table 7-
3). The technique specificaly amplified the expected DNA fragments despite the
background bacterial microflora of more than 105 CFU/g of seed (Fig. 7-5; Table 7-3).
The primers RST9 and RST10 failed to produce amplification of the 355-bp hrp-
related fragment from the pepper seed lots SP 135.92 and PK-l-PF though
amplification of the 840-bp /irp-related fragment indicated the presence of plant
pathogenic xanthomonads in these seeds (Table 7-3). This is not surprising because the
set of primers RST9 and RST10 seem to allow DNA amplification only from strains of
a limited number of pathovars of A. campestris (Chapter 3).
Attempts to recover plant pathogenic xanthomonads from different seed lots by
plating on general and semiselective media was unsuccessful. Whereas no colonies
typical of xanthomonads on both media used, YNA and Tween B, were found with the
seed lots containing low bacterial population, a large background bacterial microflora
in extracts from other seed lots prevented the identification of any possible
xanthomonads that may have grown. However, the presence of viable plant pathogenic
xanthomonads in 5 seed lots was confirmed by detection of the bacteria in seedlings
from seeds (Table 7-3). The detection of the plant pathogenic xanthomonads in the
seedlings was accomplished by isolating the bacteria from lesions in the leaflets, or by
amplification of the hrp-related fragments from bulked DNA extracted from washings
of the aerial part of the seedlings grown under greenhouse conditions. Furthermore, the
presence of viable plant pathogenic xanthomonads in the pepper seed lots SP2.92,
SP66.92, SP124.92, SP133.92, SP 135.92, and SP306.92 had been determined
previously (G. O'Keefe, personal communication).


134
parsimony criterion. No assumptions were made regarding the ancestral character
state, and the pathovar X. campestris pv. maculifoliigardeniae strain X22j was taken as
the outgroup to infer the topology of the phylogenetic tree. A total of 100 bootstrap
samples were analyzed to determine the confidence intervals of the estimates of the
inferred phylogenetic tree (Felsenstein, 1985, 1991). The KITSCH program was used
to infer a rooted phylogenetic tree by the Fitch-Margoliash method (Felsenstein, 1991;
Fitch, and Margoliash, 1967). The input data consisted of a distance matrix of pairwise
estimates of the number of nucleotide substitutions per site, (5), between strains, and
negative branch was not allowed.
Results
DNA amplification
The 840- and 1,075-bp fragments of the hrp gene cluster in X. campestris pv.
vesicatoria were successfully amplified with primers RST2 plus RST3 (Fig. 6-1) and
RST21 plus RST22 (Fig. 6-2), respectively. The same size fragments were also
successfully amplified from DNA of all strains of the other pathovars of X campestris
(Fig. 6-1 and 6-2). The DNA fragments were also amplified from 16 strains of A!
campestris pv. citrumelo representing the three aggressiveness groups and from 19
strains of A! campestris pv. citri groups A, B, and C without variation in size (data not
shown). The sequence similarity of the two DNA fragments amplified from strains of
different pathovars of X. campestris to the hrp gene cluster of X. campestris pv.
vesicatoria was further confirmed by Southern hybridization analysis. The amplified
DNA fragments of the different strains of X. campestris hybridized to the respective
internal probes specific for both hrp gene cluster fragments (data not shown).


94
DNA sequences (Chapter 5). However, the restriction analysis of the two hrp
fragments with four restriction endonucleases consistently allowed the differentiation
of almost all the pathovars and groups of X. campestris, even for those genetically
closely related (Table 4-2). In a few cases strains of different groups of plant
pathogenic xanthomonads could not be distinguished by restriction analysis of hrp-
related sequence. These include a strain of X. campestris pv. incanae and one of X.
campestris pv. carotae which were identical to strains of X. campestris pv. campestris;
a strain of X campestris pv. phaseoli which was identical to strains ofX. campestris pv
glycines group A; strains F59 and F86 of X. campestris pv. citrumelo which were
indistinguishable from strains ofY campestris pv. pruni; strain X125 ofX. campestris
pv. fici group A which was identical to strains of X. campestris pv. poinsettiicola group
A; strain XV2 ofX. campestris pv. vitians which was identical to strains ofX.
campestris pv. malvacearum; and strain X52 isolated from Hibiscus sp. was
indistinguishable from strains ofX. campestris pv. poinsettiicola group B.
Strains of a few taxa of the plant pathogenic xanthomonads were homogeneous
on the basis of restriction analysis of the hrp-related sequences. Strains of X. fragariae
produced identical and almost unique restriction patterns for all combinations of hrp-
related fragment and restriction endonuclease. The genetic analysis of the hrp-related
sequences suggests that strains of X. fragariae may be formed by a clonal population
even though the strains were isolated from plant materials from different geographic
locations. The uniformity of the X. fragariae population has also been supported by
analysis of fatty acid composition (Yang et al., 1993) and by SDS-PAGE of proteins
(Vauterin et al., 1991a). Similarly, strains of some pathovars ofX. campestris were
also highly homogeneous with regard to the restriction banding patterns of the Arp-
related fragments, i.e. X. campestris pv. begoniae, X campestris pv. campestris, X.


4
fragariae and of 28 pathovars of X. campestris. DNA sequence variation in the hrp
related fragments was investigated by restriction endonuclease analysis, and the
evolutionary relationship of these hrp gene sequences was assessed based on
phylogenetic analysis of the restriction fragment data. The DNA amplification
technique was also examined for specific detection and identification of plant
pathogenic xanthomonads and to detect plant pathogens associated with plant samples.


142
hrp genes of the different strains of X. campestris. The genetic divergence between
strains was estimated based on the data for 106 restriction fragments obtained from the
combination of two hrp gene cluster fragments and four endonucleases. A pairwise
matrix of the genetic distances, 5, was calculated for the 18 distinct banding patterns
(Table 6-1). X. campestris pv. citri 339 was included in the genetic analysis as a
representative of group C, although the restriction banding patterns of the hrp
fragments were identical to those of the strains of group B of X. campestris pv. citri.
The largest genetic divergence value was 0.082 nucleotide substitution per site between
X. campestris pv. maculifoliigardeniae X22j and A! campestris pv. citri 9771 of group
A (Table 6-1). However, most of the estimates of nucleotide substitutions per site are
smaller than 0.05, which is considered the upper limit to give accurate estimates of
genetic distance based on restriction fragment data (Nei, 1987).
Strains of X. campestris pv. citrumelo that represent all three aggressiveness
groups exhibited nine different restriction patterns of the hrp fragments that were
divergent from 0.002 to 0.022 nucleotide substitution per site (Table 6-1). Similarly,
strains of A! campestris pv. citri groups A, B, and C showed a low genetic divergence
for the hrp genes, ranging from 0.000 to 0.014 nucleotide substitution per site (Table 6-
1). As mentioned above, the banding patterns of the strains of X. campestris pv. citri
groups B and C were identical to each other for all combinations of hrp gene cluster
fragments and restriction endonucleases. On the other hand, the hrp genes of strains of
X campestris pv. citrumelo were very poorly related to the ones of X campestris pv.
citri, with divergence ranging from 0.050 to 0.064 nucleotide substitution per site
(Table 6-1).


CHAPTER 6
GENETIC ANALYSIS OF hrp RELATED DNA SEQUENCES OF
Xanthomonas campestris STRAINS CAUSING DISEASES OF
CITRUS
Citrus canker, caused by strains of Xanthomonas campestris pv. citri group A,
represents an important problem for production of citrus worldwide (Civerolo, 1984).
This disease is characterized by raised lesions on leaves, stems, and fruits. Strains of X.
campestris pv. citri group A have a relatively wide host range and cause symptoms of
various degrees in all commercial citrus varieties (Stall and Seymour, 1983). In severe
cases, abscission of fruits and leaves may result (Civerolo, 1984; Stall and Seymour,
1983). Other xanthomonads that cause similar symptoms on citrus are strains of
groups B and C of X. campestris pv. citri. They are of less importance than strains of
X. campestris pv. citri group A and have comparatively limited host ranges. Citrus
bacterial spot is another bacterial disease of citrus caused by a xanthomonad, and
symptoms are similar to citrus canker with a few important differences (Schoulties et
al., 1987). The pathogen, referred to as X campestris pv. citrumelo (Gabriel et al.,
1989), causes flat, watersoaked lesions in young leaves. Strains ofX campestris pv.
citrumelo cause symptoms primarily on trifoliate orange (Poncirus trifoliata) and its
hybrids, such as Swingle citrumelo (Citrus paradisi X P. trifoliata) (Graham and
Gottwald, 1988).
Although X. campestris pv. citri and X campestris pv. citrumelo cause similar
diseases of citrus, there is evidence for differences between these pathovars. In
addition to the pathogenicity differences listed above, X campestris pv. citrumelo
127


Fig. 3-1. Structural organization of the hrp region in Xanthomonas campestris pv. vesicatoria. (A) FcoRI fragments of the
hrp region. (B) Position and orientation of the hrp loci, designated hrpA to hrpF (Bonas et al., 1991; Schulte and Bonas,
1992). (C) Position and orientation of the open reading frames (ORFs). The sizes of the loci are based on a combination of
genetic and sequence analysis from which possible open reading frames (ORF) are predicted (Ulla Bonas, personal
communication). Only the open reading frames relevant for this study, hrpB5 to hrpB8, hrpC3, and hrpD, are shown here.
For each RST oligonucleotide primer used for DNA amplification, its position in the DNA sequence is indicated by an
asterisk (*).


35
Nonradioactive DNA labeling and Detection kit (Boehringer Mannheim, Indianapolis,
IN) as specified by the manufacturer. Clones containing the desired insert of the hrp
gene cluster of X. campestris pv. vesicatoria, or in vitro amplified hrp fragments used
as probes, were labeled by the random primer (Feinberg and Vogelstein, 1983)
incorporation of digoxigenin-labeled deoxyuridine-triphosphate (DIG-UTP). Before
use, the probes were denatured by boiling for 10 min followed by chilling in an ice
ethanol slurry. Hybridization was carried out at 68C with 0.5X SSC and 0.1%
(wt/vol) SDS. The membranes were prewashed twice at room temperature for 5 min in
IX SSC containing 0.1% (wt/vol) SDS. Two final washes were completed at 65 C for
15 min in 0.5X SSC containing 0.1% (wt/vol) SDS.
DNA amplification
Three sets of oligonucleotide primers were selected from the nucleotide
sequence of the hrp region of X campestris pv. vesicatoria (Ulla Bonas, personal
communication). Primers RST2 (5'AGGCCCTGGAAGGTGCCCTGGA3') and RST3
(5'ATCGCACTGCGTACCGCGCGCGA3') delineated a 840-bp fragment, RST9
(5'GGCACTATGCAATGACTG3') and RST10 (5'AATACGCTGGAACTGCTG3')
delineated a 355-bp fragment, and RST21
(5'GCACGCTCCAGATCAGCATCGAGG3') and RST22
(5'GGCATCTGCATGCGTGCTCTCCGA3') delineated a 1,075-bp fragment. The
primers map to the complementation groups hrpB, hrpC, and hrpD of X campestris pv.
vesicatoria (Fig. 3-1). Furthermore, the sequences of the oligonucleotide primers RST3
and RST9 originate from the hrpB6, a gene for a putative ATPase that seems to be
highly conserved among different bacteria at the protein sequence level (Fenselau et al.,
1992). Oligonucleotide primers were synthesized with a model 394 DNA Synthesizer


149
was taken as the outgroup to infer the topology of the tree (Fig. 6-6). The strains ofX.
campestris can be divided into three major clades based on the hrp genes, with A"!
campestris pv. maculifoliigardeniae X22j as the sole member of one clade (Fig. 6-6).
The second clade is the largest one and comprises all strains oiX. campestris pv.
citrumelo, X. campestris pv. alfalfae, X. campestris pv. fici, X. campestris pv.
vesicatoria, andX. campestris from S. reginae (Fig. 6-6). The third clade is formed by
all the strains of X. campestris pv. citri, X. campestris pv. bilvae, and a strain of X.
campestris from Feronia sp. (Fig. 6-6). The assemblage of these phylogenetic clades is
highly supported by the bootstrap values (Fig. 6-6). Furthermore, X. campestris pv.
maculifoliigardeniae X22j was indeed chosen as an appropriate outgroup to evaluate
the relationships among X. campestris strains causing disease on citrus on the basis of
the hrp genes. The rooted tree obtained by the KITSCH program, selected among 1908
trees examined, also indicates that X. campestris pv. maculifoliigardeniae X22j is basal
to the remainder^ campestris strains included in this study (Fig. 6-7).
The inferred phylogenetic trees seem to support the hypothesis that the hrp gene
cluster of X. campestris pv. citrumelo strains from the three aggressiveness groups are
monophyletic and closely related to other pathovars of X. campestris, including X.
campestris pv. alfalfae, X. campestris pv. fici, and X. campestris pv. vesicatoria.
Similarly, the monophyly of the hrp genes of X. campestris pv. citri, X. campestris pv.
bilvae, andX. campestris XCF is also supported (Fig. 6-6 and 6-7).
Discussion
Strains of all pathovars of X. campestris included in this study have an hrp gene
cluster on the basis of hybridization of genomic DNA with the hrp gene cluster from X.


Fig. 5-2. Unrooted phylogenetic tree inferred from restriction analysis data of the 840-
bp DNA fragment related to the hrpB complementation group of the hrp genes of
Xanthomonas campestris pv. vesicatoria generated by the Wagner parsimony criterion.
The values on each node indicate the levels of support derived from 100 bootstrapped
trees. The shaded boxes delineate the major clades identified in the analysis of the
hrpC/D-rdated DNA fragment. Also included are the number of strains examined for
each taxa (numbers in parenthesis).


152
gene cluster, were compared. Nevertheless, the results of the genetic analysis of the
hrp related regions are consistent with those obtained when the entire genome was
randomly examined. From this and the previously cited research, the two groups of
strains within X. campestris pv. citri should probably be differentiated at some
taxonomic level.
Two groups of X campestris pv. citrumelo were also distinguished after
restriction enzyme digestion of the amplified hrp fragments. Strains of the highly
aggressive group were very uniform for all fragment-endonuclease combinations and
had a characteristic restriction banding pattern. On the other hand, the moderately and
weakly aggressive groups of the bacterium had diverse restriction banding patterns for
both amplified hrp fragments. This concurs with previous studies of the moderately
and weakly aggressive groups of X. campestris pv. citrumelo (Egel, 1991; Egel et al.,
1991; Gabriel et ah, 1988, 1989; Hartung and Civerolo, 1987, 1989; Kubicek et ah,
1989). Even with genetic diversity among the moderately and weakly aggressive
strains, all strains included under X. campestris pv. citrumelo are 80% similar using
DNA-DNA hybridization (Egel, 1991; Egel et ah, 1991). As with strains ofX.
campestris pv. citri, the two groups of strains of X. campestris pv. citrumelo should
also be distinguished at the taxonomic level.
The results presented in this work demonstrate that strains of X. campestris
causing disease in citrus can be reliably differentiated and identified by restriction
analysis of amplified fragments related to the hrp gene cluster. The number of strains
examined were small, but the genetic diversity within different pathovars of X.
campestris could be assessed. The reliable identification of the citrus pathogen using
DNA amplification will greatly facilitate disease diagnosis, as well as ecological and


87
Fig. 4-10. Restriction analysis of the (A) 840 bp and (B) 1,075 bp DNA fragments of
the hrp gene cluster amplified from strains of Xanthomonas campestris pv. begoniae
and restricted with HaeIII (Lanes 2 to 10) and Cfo\ (Lanes 12 to 20). Lanes 1 and 11,
phage X restricted with Pstl. Lanes 2 and 12, X274; 3 and 13, X281; 4 and 14, X329; 5
and 15, X610; 6 and 16, X627; 7 and 17, X1490; 8 and 18, X1492; 9 and 19, X1496;
10 and 20, XCB9. Molecular sizes are given in bases.


Fig. 5-3. Rooted phylogenetic tree inferred from restriction analysis data of the 1,075-
bp DNA fragment related to the hrpC/D complementation group of the hrp genes of
Xanthomonas campestris pv. vesicatoria generated by the method unweight pair-group
method with arithmetic mean (UPGMA). The shaded boxes delineate the major clades
identified in the analysis of the hrpC/D-related DNA fragment. Also included are the
number of strains examined for each taxa (numbers in parenthesis).


190
the 56 strains isolated from pepper and tomato seeds or seedlings, only nine strains
were positively identified as X. campestris pv. vesicatoria, the pepper and tomato
pathogen. The other strains formed two distinct groups. Seven strains did not produce
a reaction on pepper and tomato plants, and they failed to produce DNA sequences
similar to the hrp genes. These strains resemble the opportunistic xanthomonads that
have been described previously in association with pepper and tomato transplants
(Gitaitis et al., 1987, 1992). The other group of xanthomonads included potential plant
pathogenic strains that were not pathogenic to pepper and tomato, but instead caused a
hypersensitive reaction or no reaction on these plants under artificial inoculation.
These strains were tentatively identified as belonging to different pathovars of X.
campestris, but not to X campestris pv. vesicatoria, on the basis of the hrp analysis and
fatty acid composition. The presence of a diverse group of xanthomonads associated
with pepper and tomato seeds and transplants is of major significance for the health
inspection in certification programs, particularly with the procedures currently
available for the detection and identification of plant pathogenic xanthomonads (Holt,
1994; Saettler et al., 1989; Schaad, 1988). Although more investigation may be
necessary, the combined specificity and sensitivity of the DNA amplification approach
of /zrp-related fragments seems highly promising for the detection and identification of
plant pathogenic xanthomonads associated with propagative plant materials.
Differences observed in the number of common restriction fragments of the
amplified DNA sequences related to the hrp genes indicated that there is variation in
the relatedness of the hrp genes of the different strains of plant pathogenic
xanthomonads. The relatedness of the hrp genes of the different plant pathogenic
xanthomonads was further investigated by phylogenetic analysis. This analysis
revealed a diverse evolutionary relationship for the hrp genes of the plant pathogenic


85
Fig. 4-9. Restriction analysis of the 840-bp hrp-related fragments amplified from
strains of Xanthomonas fragariae and restricted with HaeIII (Lanes 2 to 10) and Cfol
(Lanes 12 to 20). Lanes 1 and 11, phage X restricted with Pstl. Lanes 2 and 12,
X1238; 3 and 13, X1241; 4 and 14, X1244; 5 and 15, X1246; 6 and 16, X1292; 7 and
17, X1298; 8 and 18, X1426; 9 and 19, GC6265; 10 and 20, X. campestris pv.
vesicatoria 75-3. Molecular sizes are given in bases.


27
campestris pv. manihotis correlated with pathogenicity on the host plant (Berthier et
al., 1993).
Extensive RELP analyses were conducted in some pathovars ofX. campestris,
such &sX. campestris pv. citri and X. campestris pv. citrumelo (Gabriel et al., 1988,
1989; Gottwald et al., 1991; Graham et al., 1990; Hartung, 1992; Hartung and
Civerolo, 1989, 1991; Lazo and Gabriel, 1987; Lazo et al., 1987),^ campestris pv.
manihotis (Verdier et al., 1993), X. campestris pv. vasculorum (Qhobela and Claflin,
1992), X. campestris pv. pennamericanum (Qhobela and Claflin, 1988), and X oryzae
pv. oryzae (Leach et al., 1990, 1992). The most comprehensive RFLP analyses were
carried out to determine the genetic diversity of the xanthomonads causing diseases of
citrus (Gabriel et al., 1988, 1989; Gottwald et al., 1991; Graham et al., 1990; Hartung,
1992; Hartung and Civerolo, 1989, 1991; Lazo and Gabriel, 1987; Lazo et al., 1987).
The RFLP analyses characterized the strains of the citrus bacterial spot pathogen, X.
campestris pv. citrumelo, as distinct from all forms of the citrus canker pathogen
(Gabriel et al., 1988, 1989; Hartung and Civerolo, 1989). In contrast to the
characteristic restriction patterns of the X campestris pv. citri canker groups A and B,
X. campestris pv. citrumelo gave a wide range of variation in the RFLP analyses
(Gabriel et al., 1988, 1989; Graham et al., 1990; Hartung and Civerolo, 1989, 1991).
The RFLP analyses further supported that strains of the citrus bacterial spot pathogen
are not genetically closely related to any recognized group of X. campestris pv. citri
(Hartung and Civerolo, 1987, 1989; Gabriel et al., 1988, 1989). Instead, the analysis
has revealed a significant genetic relationship between strains of X. campestris pv.
citrumelo and strains of non-citrus pathogens, such as X. campestris pv. alfalfae and X
campestris pv. fici (Gabriel et al., 1988,1989; Graham et al., 1990). The relatedness of
strains of X. campestris pv. citrumelo to strains of the other pathovars oiX. campestris


106
species A! fragariae and 29 pathovars of X. campestris included in the study (data not
shown). However, the amplification of the hrpB-re\ated fragment from X. campestris
pv. carotae, X. campestris pv. gardneri, X. campestris pv. glycines B, X. campestris pv.
papavericola, X. campestris pv. pelargonii, X. campestris pv. vitians C, and a strain of
X. campestris from Hibiscus sp. usually produced low yield of DNA whereas no
amplification of this fragment was obtained from X campestris pv. holcicola. These
strains were not included in the phylogenetic analysis of the hrpB region. The hrp-
related fragments amplified with each set of oligonucleotide primers from different
plant pathogenic xanthomonads were of identical size.
Unrooted parsimony analyses were carried out on 100 and 63 restriction
fragment data sets produced by the digestion of the hrpC/D- and hrpB-related
fragments, respectively, with four different frequent-cutting restriction endonucleases.
The phylogenetic trees inferred for each amplified fragment are very similar in
topology to each other (Fig. 5-1 and 5-2). Ten major clades were identified in the
analysis of the hrpC/D-related fragment amplified from the plant pathogenic
xanthomonads (Fig. 5-1). Clades 1, 2, 4, 9, and 10 include only a single taxon whereas
the remaining clades are comprised by taxa representing different pathovars of X.
campestris (Fig. 5-1 and 5-2). Further, the members of each clade can be easily
identified based on the restriction banding profile of the hrpC/D-related fragment
produced with the endonuclease Taql (Table 5-1). The result of the phylogenetic
analysis of the hrpB region is consistent with the phylogeny determined for the hrpC/D
region with a few important differences. The clades 3 and 4 which contain pathovars
of2f. campestris that cause diseases on brassica plants merged in a larger single clade
whereas the single member of clade 9, X. campestris pv. dieffenbachiae B, were placed
in clade 7 (Fig. 5-2). Another difference is the shift of the strains of X. campestris pv.


Table D-lContinued
Species/Pathovar Species/Pathovar
42
44
45
46
47
48
49
50
51
52
23. pv. dieffenbachiae A
.2609
.1395
.4583
.4490
.2727
.2727
.2273
.4167
.4255
.4255
24. pv. dieffenbachiae A
.2174
.1860
.4583
.4490
.3182
.3182
.2273
.4167
.4681
.4681
25. pv. dieffenbachiae B
.3810
.2564
.7273
.6667
.4000
.4000
.3500
.6364
.6977
.6977
26. pv. fici A
.4783
.2791
1.000
.8980
.3636
.3636
.3636
.8750
.7660
.7660
27. pv. fici A
.5106
.3636
.7347
.8000
.4000
.4000
.4000
.7347
.7083
.7083
28. pv. fici A
.5106
.3636
.7347
.8000
.4000
.4000
.4000
.7347
.7083
.7083
29. pv. fici B
.2381
.7692
.2273
.2667
.4000
.4000
.4000
.2273
.2791
.2791
30. pv. gardneri
.2381
.4615
.3636
.3556
.4500
.4500
.9000
.3182
.2791
.2791
31. pv. glycines A
.8261
.2791
.5833
.6122
.4091
.4091
.3636
.5417
.5532
.5532
32. pv. glycines B
.2381
.7692
.2273
.2667
.4000
.4000
.4000
.2273
.2791
.2791
33. pv. holcicola
.1739
.2326
.1667
.1633
.2273
.2273
.2727
.1250
.1277
.1277
34. pv. incanae
.2381
.4615
.2727
.3111
.4500
.4500
.8500
.2273
.2326
.2326
35. pv. maculifoliigardeniae
.2273
.2927
.5652
.5532
.4286
.4286
.3810
.4348
.4889
.4889
36. pv. malvacearum
.6512
.2500
.4889
.5217
.3902
.3902
.3415
.4000
.4091
.4091
37. pv. manihotis
.2979
.2727
.5714
.5600
.3556
.3556
.3111
.5306
.5833
.5833
38. pv. papavericola
.3256
.5000
.3556
.3913
.4878
.4878
.7805
.2667
.2727
.2727
39. pv. pelargonii
.3256
.5000
.3111
.3478
.4878
.4878
.8293
.2667
.2727
.2727
40. pv. phaeoli A
.2667
.1905
.5106
.5000
.3256
.3256
.2326
.4681
.5217
.5217
41. pv. phaseoli B
.8261
.2791
.5833
.6122
.4091
.4091
.3636
.5417
.5532
.5532
42. pv. phaseoli "fuscans"
,
.2439
.4783
.5532
.2857
.2857
.2381
.5652
.5778
.5778
44. pv. poinsettiicola B


.2791
.3182
.5128
.5128
.4615
.2791
.3333
.3333
Continued on following page


Table 4-2Continued
Species/Pathovar
Strain
840 bp
1,075 bp
Taql
Sau3Al HaeIII
CM
Taql
Sau3A\ Haelll
CM
pv. vitians A
ICPB101, XVIT, X1215
2
7
12
10
9
14
22
17
B
XV2
2
5
6
6
6
9
12
10
C
ICPB164, XV3
nd
nd
nd
nd
3
2
4
2
undetermined and
isolated from Feronia
sp.
XCF
4
5
9
9
7
10
10
10
undetermined and
isolated from Hibiscus
sp.
X52
nd
nd
nd
nd
5
8
9
6
undetermined and
isolated from Strelitzia
reginae
X198
9
7
16
12
8
11
15
8
X. fragariae
GC-6259, GC-6265, X1238,
X1241, X1244, XI246, X1292,
X1298, XI426
1
2
2
2
2
5
2
13
Restriction pattern group; Number of the restriction pattern groups refer to lanes in Fig. 4-1 to 4-8.
bnd, not determined.


173
Table 7-2. Sensitivity of ELISA and DNA amplification procedures for detection of
Xanthomonas campestris pv. vesicatoria in tomato seeds.
Sample Total population of X. campestris recovered ELISA Amplification
bacteria on YNA on Tween B medium Anm492a of hrp
(CFU/ml) (CFU/ml) fragment*5
355 bp 840 bp
1
2.9 X 108
2.0 X 108
2.79 0.01
c

2
7.3 X 107
6.7 X 107
2.09 0.24


3
5.4 X 106
4.6 X 106
0.51 0.01


4
5.6 X 105
5.5 X 105
0.21 0.01


5
6.8 X 104
6.8 X 104
0.16 0.02


6
5.8 X 103
5.2 X 103
0.16 0.01


7
1.7 X 102
1.4 X 102
0.17 0.01

-
8
2.6 X 102
7.0 X 101
0.17 0.01
-
-
9
1.5 X 102
not detected
0.16 0.01
-
-
aValues are the average of the spectrophotometer readings of three wells.
bAliquots of 300 pi and 600 pi of the seed washings were used in the ELISA assay and
for extraction of DNA, respectively. In the DNA amplification assay was used 10 pi of
DNA template.
c no hrp fragment amplification; +, hrp fragment amplification.


CHAPTER 7
EVALUATION OF A DNA AMPLIFICATION APPROACH FOR
DETECTION AND IDENTIFICATION OF PLANT PATHOGENIC
XANTHOMONADS ASSOCIATED WITH PEPPER AND TOMATO
SEEDS
Bacterial spot caused by X. campestris pv. vesicatoria is one of the most
important diseases for pepper and tomato productions worldwide (Jones et al., 1991;
Stall, 1993). The disease is characterized by necrotic lesions on leaves, stems, and
fruits. In warm and rainy weather, the disease is usually severe and may cause heavy
defoliation of the plants that results in reduced yield (Pohronezny et al., 1986).
Furthermore, diseased fruits are usually commercially depreciated and may not be
suitable for fresh-market (Cox, 1966; Pohronezny et al., 1983). Bacterial spot is also of
major concern in the certification program in the transplant industry in southern USA
where producers are expected to provide disease-free transplants (Gitaitis et al., 1987,
1992).
Control of bacterial spot has been hampered by many factors. No bactericide is
completely effective to control the disease and tolerance to bactericides, such as
streptomycin and copper compounds, is widespread among strains of X campestris pv.
vesicatoria (Marco and Stall, 1983; Thayer and Stall, 1961). This has lead to reduced
efficacy of bactericide sprays for the control of the disease in the field. Sources of
resistance to bacterial spot were identified in pepper (Sowell and Dempsey, 1977) and
tomato (Scott and Jones, 1989) and have been used in breeding programs to develop
horticulturally desirable pepper and tomato cultivars. However, sudden shifts in races
ofX campestris pv. vesicatoria can overcome the resistanceand poses a major concern
154


3
strains ofX. campestris were divided into 15 DNA homology groups (Vauterin et al.,
1993). Comparisons by genomic fingerprinting and RFLP analysis have also shown
significant genetic diversity among and within different pathovars of X. campestris
(Egel et al., 1991; Lazo et al., 1987; Stall et al., 1994). Furthermore, physiologically
and biochemically similar xanthomonads that are pathogenic on the same host plant
may have a diverse genetic background (Egel et al., 1991; Hildebrand et al., 1990;
Palleroni et al., 1993; Stall et al., 1994). On the other hand, the application of
conventional genetic analysis for routine identification, as well as for taxonomic
purposes, may be hampered by some limiting features. The determination of DNA-
DNA homologies are limited by technical difficulties and by experimental error in
obtaining the data (Sneath, 1989). Further, rearrangements of the bacterial
chromosome may result in changes in the phenotype and RFLP patterns without
appreciably affecting overall sequence similarity (Egel et al., 1991; Young et al., 1992).
Alternatively, analysis of specific DNA sequences of the bacterial genome should
provide accurate information regarding the relationships of the strains and pathovars of
X. campestris without the shortcomings mentioned above. The procedure may be a
useful tool for the specific detection and identification of plant pathogenic bacteria.
The primary objective of this work is to investigate specific sequences of the
bacterial genome of plant pathogenic xanthomonads for the specific identification of
these plant pathogens. Oligonucleotide primers were designed based on the sequence
of the hrp gene (hypersensitive reaction and pathogenicity) cluster of X. campestris pv.
vesicatoria (Bonas et al., 1991; Ulla Bonas, personal communication), and tested for
the specific amplification of DNA fragments related to the hrp gene cluster from the
bacterial genome of different plant pathogenic xanthomonads by polymerase chain
reaction. DNA fragments related to the hrp genes were amplified from strains of X.


CHAPTER 1
INTRODUCTION
The genus Xanthomonas Dowson 1939 includes Gram-negative and usually
yellow-pigmented bacteria that occur worldwide and cause disease on many plants.
The host range spans over 124 monocotyledonous and 268 dicotyledonous plant
species (Hayward, 1993; Leyns et al., 1984). Among the species of Xanthomonas, X.
campestris is the most diverse and includes at least 125 different pathovars that are
distinguished by causing different diseases of plants (Bradbury, 1984; Dye et al.,
1980). Although the determination of the genus Xanthomonas and its species presents
no great problem, the identification and characterization at the subgeneric level is still
difficult (Van den Mooter and Swings, 1990; Vauterin et al., 1990a, 1991a; Young et
al., 1992). The basis for differentiation of plant pathogenic bacteria at the subgeneric
level is by the capability of the bacterial strain to cause characteristic disease, or by
reference to their host range (Dye et al., 1980; Vauterin et al., 1990a; Young et al.,
1992). However, pathogenic reactions are variable. Strains may differ in
aggressiveness, causing a broad range of symptoms when artificially inoculated into
host plant species. The strains may also differ in the pathogenic reaction when
artificially inoculated onto nonhost plants. Furthermore, these pathovars may not be
readily differentiated by other phenotypic features (Van den Mooter and Swings,
1990). In addition, yellow-pigmented bacteria apparently belonging to the genus
Xanthomonas and nonpathogenic to the plants from which they were isolated are
usually associated with plant material (Angeles-Ramos et al., 1991; Gitaitis et al.,
1


48
Pseudomonas, and Xylella, and from the non plant pathogens E. herbicola, X.
maltophilia, and the opportunistic strains of X. campestris T-55 and INA, when either
RST2 plus RST3 or RST21 plus RST22 were used (Table 3-1). The failure to amplify
the DNA fragments from all those bacterial strains was expected because of the lack of
hybridization to the three hrp fragments from X. campestris pv. vesicatoria 75-3 (Table
3-1).
DNA fragments delineated by the primers RST9 and RST10 were consistently
amplified only from strains of X. campestris pv. fci, X campestris pv. physalidicola,
X. campestris pv. vesicatoria, and A! campestris X198 (Fig. 3-5; Table 3-1). However,
some pathovars of X. campestris, including alfalfae, citrumelo, maculifoliigardinae,
and manihotis, as well as the strain XCF, sometimes produced low yield in the
amplification of the 355 bp fragment (Fig. 3-5; Table 3-1).
The identity of these hrp related fragments amplified from different strains of
plant pathogenic Xanthomonas spp. was further confirmed by Southern hybridization
analysis. Internal portions of the 840- and 1,075-bp DNA fragments, as well as the
entire 355 bp fragment amplified from X. campestris pv. vesicatoria 75-3, were used as
probes. The internal probe for the 840 bp fragment consisted of a 271-bp insert of the
plasmid pXV840, and the internal probe for the 1,075-bp fragment consisted of a 335-
bp insert of the plasmid pXV1075. The inserts were obtained from fragments
amplified from DNA of A campestris pv. vesicatoria strain 75-3 by cloning Sau3Al
digests into the BamHl site of the vector pBluescript II KS +/- (Stratagene, La Jolla,
CA). In all cases, the DNA fragments amplified from different strains of Xanthomonas
spp. with each set of primers hybridized with the respective probe (data not shown).


Table 7-1 Continued
Strain
Origin
Reaction on
plants
Amplification
of hrp fragment
Tentative identification
Source
pepper tomato
hrpB
hrpC/D
hrp analysis
FAA
BSA9
pepper seed
HR
HR
(+)
+
pv. carotae
pv. carotae (0.824)
JW
BSA10
pepper seed
HR
HR
(+)
+
pv. carotae
pv. carotae (0.903)
JW
BSA11
pepper seed
HR
HR
(+)
+
pv. carotae
pv. carotae (0.852)
JW
BSA12
pepper seed
HR
HR
(+)
+
pv. carotae
pv. carotae (0.767)
JW
BSA13
pepper seed
-
HR
(+)
+
pv. carotae
pv. carotae (0.753)
JW
BSA14
pepper seed
HR
HR
(+)
+
pv. carotae
pv. carotae (0.577)
JW
BSA15
pepper seed
HR
HR
(+)
+
pv. carotae
pv. carotae (0.748)
JW
BSA16
pepper seed
-
-
(+)
+
pv. carotae
pv. carotae (0.841)
JW
BSA17
pepper seed
-
-
(+)
+
pv. carotae
pv. carotae (0.868)
JW
BSA18
pepper seed
HR
HR
-
-
na
pv. malvacearum (0.736)
JW
BSA19
pepper seed
HR
HR
-
-
na
pv. malvacearum (0.720)
JW
BSA20
pepper seed
-
HR
(+)
+
pv. carotae
pv. carotae (0.789)
JW
BSA21
pepper seed
-
-
(+)
+
pv. carotae
pv. carotae (0.905)
JW
BSA22
pepper seed
-
-
(+)
+
pv. carotae
pv. carotae (0.809)
JW
BSA23
pepper seed
HR
HR
(+)
+
pv. carotae
pv. carotae (0.782)
JW
BSA24
pepper seed
-
HR
(+)
+
pv. carotae
pv. carotae (0.779)
JW
BSA25
pepper seed
-
-
(+)
+
pv. carotae
pv. carotae (0.752)
JW
Continued on the following page


122
analysis of the hrp genes revealed the same level of relatedness found when the entire
genome were compared. Whereas the estimates of the similarity for the two regions of
the hrp genes ranged from 0.41 to 0.61, the DNA-DNA hybridization relatedness
ranged from 0.55 to 0.63 (Egel et al., 1991). In contrast, strains of A campestris pv.
citri canker A are genetically highly related to strains of X. campestris pv.
malvacearum based on DNA-DNA hybridization studies (Egel et al., 1991). The
analysis of the hrp genes also revealed a high similarity between these two groups of
plant pathogenic xanthomonads with values ranging from 0.80 to 0.88 for both regions
of the /^-related sequences examined (Appendix D). Further, these two groups o X.
campestris are monophyletic in regard to the evolution of the hrp genes (Fig. 5-1 and
5-2).
Additional evidence to support the coevolution of the hrp genes and the rest of
the bacterial genome comes from the pathovars X. campestris pv. carotae, X.
campestris pv. gardneri, and X. campestris pv. pelargonii. Although these
xanthomonads have very distinct host ranges, they are genetically closely related based
on DNA-DNA hybridization (Hildebrand et al., 1990; Palleroni et al., 1993). The
phylogenetic analysis of the hrp-related sequences also supports the contention that
these pathovars of X. campestris are monophyletic and genetically closely related in
regard to the hrp genes. The examples discussed above not only point to a coevolution
of the hrp genes and the rest of the genome from a common bacterial ancestor for
certain plant pathogenic xanthomonads, but also substantiate the divergence between
hrp genes and host speciation.
Another major finding of this study is the indication of lateral movement of the
hrp genes between plant pathogenic xanthomonads. This hypothesis is supported by
two lines of evidences, the presence of similar hrp gene cluster sequences in strains


93
pv. vesicatoria group B is genetically and phenotypically very distinct from X.
campestris pv. vesicatoria group A, although both groups are presently lumped into the
same pathovar and cause similar disease symptoms on solanaceous plants (Stall et al.,
1994). These two groups of strains have different genetic backgrounds and they are
only about 33% similar on the basis of DNA homology (Stall et al., 1994). On the
other hand, the majority of the plant pathogenic xanthomonads produced restriction
banding patterns that were shared by strains of different groups or pathovars ofX.
campestris. Furthermore, different restriction profiles may have DNA bands in
common. The presence of similarities in the restriction banding patterns of strains of
different taxa strongly supports the presence of some degree of genetic relatedness
among them.
Several pathovars of X. campestris were determined to be genetically very
closely related to each other on the basis of DNA homology (Hildebrand et al., 1990;
Palleroni et al., 1993; Vauterin et al., 1993) and RFLP analysis (Graham et al., 1990;
Gottwald et al., 1991; Lazo et al., 1987; Verdier et al., 1993). For example, the
pathovars X campestris pv. alfalfae, X campestris pv. begoniae, X. campestris pv.
cassavae, X. campestris pv. citri, X. campestris pv. citrumelo, X campestris pv.
dieffenbachiae, X campestris pv. glycines, X. campestris pv. malvacearum, X
campestris pv. manihotis, X. campestris pv. phaseoli, X campestris pv. phaseoli
"fuscans", X. campestris pv. poinsettiicola group A, X. campestris pv. vesicatoria group
A, and X campestris pv. vignicola belong to the DNA homology group VIII
established by Vauterin et al. (1993) whereas the pathovars X. campestris pv.
armoraciae, X campestris pv. campestris, X. campestris pv. incanae, and X. campestris
pv. raphani belong to the group XII (Vauterin et al., 1993). Similar grouping was
obtained in the phylogenetic analysis of the restriction fragment data of the hrp-related


Table D-2Continued
Species/Pathovar Species/Pathovar
50
51
52
53
54
55
56
58
59
61
1. pv. alfalfae
.8696
.8511
.8696
.3784
.8000
.6190
.4651
.8444
.4889
.2381
2. pv. alfalfae
.8696
.8511
.8696
.3784
.8000
.6190
.4651
.8444
.4889
.2381
3. pv. armoraciae
.5217
.5106
.5217
.5405
.5333
.5238
.5116
.4444
.5778
.5238
4. pv. begoniae A
.7442
.7727
.7442
.5294
.7619
.9231
.5500
.6667
.6190
.5128
5. pv. begoniae A
.7442
.7727
.7442
.5294
.7619
.9231
.5500
.6667
.6190
.5128
6. pv. begoniae B
.7442
.7727
.7442
.5294
.7619
.9231
.5500
.6667
.6190
.5128
7. pv. bilvae
.5957
.6250
.5957
.5263
.6087
.6047
.8182
.5217
1.000
.4186
8. pv. campestris
.5652
.5532
.5652
.5946
.5778
.5714
.5116
.4889
.5778
.5714
11. pv. citri A
.5217
.5532
.5217
.4865
.5333
.5714
.8837
.4444
.8444
.4762
12. pv. citri B
.5217
.5532
.5217
.4865
.5333
.5714
.7907
.4444
.9333
.4286
13. pv. citri C
.5217
.5532
.5217
.4865
.5333
.5714
.7907
.4444
.9333
.4286
14. pv. citrumelo
.9362
.9167
.9362
.4211
.8696
.6512
.5000
.7826
.5217
.2791
15. pv. citrumelo
1.000
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
16. pv. citrumelo
1.000
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
17. pv. ctirumelo
.9362
.9167
.9362
.4737
.8696
.6977
.5455
.8696
.5652
.3256
18. pv. citrumelo
.9362
.9167
.9362
.4211
.8696
.6512
.5000
.7826
.5217
.2791
19. pv. citrumelo
.8261
.8085
.8261
.3784
.7556
.6190
.4186
.8000
.4889
.2857
20. pv. citrumelo
1.000
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
21. pv. citrumelo
.8696
.8511
.8696
.3784
.8000
.6190
.4651
.8444
.4889
.2381
22. pv. citrumelo
.9583
.9388
.9583
.5128
.8936
.7273
.5333
.7660
.5957
.4091
23. pv. dieffenbachiae A
.7273
.7556
.7273
.5714
.7907
1.000
.5854
.6977
.6047
.5000
24. pv. dieffenbachiae A
.7273
.7556
.7273
.5714
.7907
1.000
.5854
.6977
.6047
.5000
Continued on following page


Ill
vignicola from clade 6 to clade 7 in the phylogenetic analyses of the hrpC/D and hrpB
regions, respectively (Fig. 5-1 and 5-2).
The results of the bootstrapped resampling analyses presented in the branches
of the trees largely support the monophyletic nature of the major clades in almost all
the cases (Fig. 5-1 and 5-2). Further, the topology of these trees reconstructed by the
parsimony procedure is consistent with the phylogenetic analyses based on the distance
matrix approaches UPGMA (Fig. 5-3 and 5-4) and neighbor-joining (not shown).
Further, the UPGMA analyses also indicate that the strains G-23 ofX. campes tris pv.
holcicola and XV56 of X. campestris pv. vesicatoria group B taken to infer the
topology of the unrooted parsimony trees for the hrpC/D and hrpB regions,
respectively, were appropriate outgroups because they were basal to the remainder of
Xanthomonas spp. included in this study (Fig. 5-3 and 5-4).
The present phylogenetic framework supports the monophyletic nature of the
hrp gene sequences of plant pathogenic xanthomonads that belongs to different
pathovars of X campestris and cause diseases in different hosts. For instance, clade 6
comprises the pathovars X. campestris pv. citri, X campestris pv. bilvae, X campestris
pv. glycines A, X. campestris pv. malvacearum, X campestris pv. phaseoli B, X.
campestris pv. phaseoli "fuscans", and A", campestris pv. vitians B (Fig. 5-1 and 5-2).
Although these pathogens cause diseases on different plants, there is strong agreement
in the parsimony analyses to indicate a common bacterial ancestor for both hrp genes
regions of these pathovars. The bootstrap resampling analyses support the branching of
this clade at 78% and 94% for the hrpC/D and hrpB regions, respectively (Fig. 5-1 and
5-2). The estimates of the similarity between hrp genes for each region also indicates a
close genetic relatedness for the members within this clade. The estimates of the
similarity of the hrp genes region within clade 6 ranged from 0.65 to 1.00 and from


Table D-lContinued
Species/Pathovar
Species/Pathovar
22
23
24
25
26
27
28
29
30
31
23. pv. dieffenbachiae A
.9583
.5000
.4583
.4082
.4082
.1364
.2273
.2917
24. pv. dieffenbachiae A
.5455
.4583
.4082
.4082
.1818
.2273
.2500
25. pv. dieffenbachiae B
.7273
.4889
.4889
.2000
.3500
.5000
26. pv. fici A
.7347
.7347
.2273
.3636
.5833
27. pv. fici A
1.000
.3111
.4000
.5306
28. pv. fici A
.3111
.4000
.5306
29. pv. fici B
.
.
.4000
.2273
30. pv. gardneri
.
.
.3636
31. pv. glycines A
.
.
Continued on following page
to
o
oo


117
Table 5-1. Within clade similarity values generated by comparison of the restriction
profiles obtained by digestion of the /zrp-related fragments amplified from plant
pathogenic Xanthomonas spp. with restriction endonuclease enzymes.
Clade3
Taql
groupb
Number of taxa in the
cladec
Similarity within clade
hrpB
hrpC/D
hrpB
hrpC/D
1
1
1
1


2
2
1
1


3
3
}5
10
0.96-1.00de
0.78-1.0
4
5
2
0.96
1.00
5
5
2
4
0.87f
0.77-1.00
6
6 and 7
10
11
053-1.00
0.65-1.00
7
8
22
20
0.77-0.78
0.67-1.00
8
9
9
9
0.84-1.00
0.64-0.96
9
10
1
1


10
11
na§
1


aClade established on the basis of the phylogenetic analysis of the restriction fragment
data of the 1,075-bp fragment related to the hrpC/D region of the hrp genes of X.
campestris pv. vesicatoria.
bGroup established based on the restriction profile of the 1,075-bp fragment related to
the hrpC/D region of the hrp genes of X. campestris pv. vesicatoria restricted with the
endonuclease Taq\ (see Chapter 4).
cNumber of taxa determined based on the phylogenetic analysis of each hrp region
individually.
dValues are the similarities estimated by using the equation proposed by Nei and Li
(1979) for each /zrp-related fragment restricted with either endonuclease Cfol, HaelW,
Sau3Al, and Taql.
eOnly three taxa were compared.
fOnly two taxa were compared.
8 na, not applicable.


199
Table A-2. List of bacterial strains and plasmids used in molecular transformation
and conjugation in
this study.
Strain
Relevant characteristics
Source or reference3
Bacteria
Xanthomonas campestris
pv. vesicatoria
22
nonpathogenic Tn3-gus insertion mutant
of strain 85-10
Bonas et al., 1991
44
nonpathogenic Tn3-gs insertion mutant
of strain 85-10
Bonas et al., 1991
75
nonpathogenic Tn3-gus insertion mutant
of strain 85-10
Bonas et al., 1991
85
nonpathogenic Tn3-gi/.v insertion mutant
of strain 85-10
Bonas et al., 1991
137
nonpathogenic Tn3-gw.v insertion mutant
of strain 85-10
Bonas et al., 1991
318
nonpathogenic Tn3-gns insertion mutant
of strain 85-10
Bonas et al., 1991
Escherichia coli
DH5a
FTecA 80dlacZM15
BRL
HB101
F'recA
BRL
Plasmids
pLAFR3
Tetrr/x+ RK2replicon
Staskawicz et al., 1987
pBluescript-KS+
Ampr, Bluescript
Stratagene
pRK2013
KmrTraRK2+Mob+ColEl replicn
Figurski and Helinski,
1979
pXV9
pLAFR3 cosmid clone from X. c. pv.
vesicatoria 75-3
Bonas et al., 1991
pXV847
pBluescript clone containing internal Sau
3 A hrp fragment amplified from X. c. pv.
vesicatoria 75-3
This study
pXVllll
pBluescript clone containing internal Sau
3 A hrp fragment amplified from X. c. pv.
vesicatoria 75-3
This study
pXV5.1
pLAFR3 Eco RI subclone from pXV9
This study
pXV5.5
pLAFR3 Eco Rl subclone from pXV9
This study
Continued on the following page


82
Variability of homologous hrp-related fragments at species and pathovar levels
Strains of 26 campestris pv. holcicola, X. campestris pv. vesicatoria group B,
and X. fragariae each produced characteristic restriction patterns for both /^-related
fragments digested with four different endonucleases. In the case of the 1,075-bp hrp-
related fragment amplified from strains of the three taxa and restricted with either Cfol,
HaeIII, SaulAl, or Taql, the restriction banding patterns produced were unique (Table
4-2). Furthermore, the restriction pattern obtained for strains of 26 campestris pv.
vesicatoria group B and X. fragariae were identical within each group for all four
endonucleases (Table 4-2). In regard to the 840-bp hrp-related fragment, the restriction
patterns were also unique for the strains of X. campestris pv. vesicatoria group B and X.
fragariae with the exception of the banding patterns obtained by using the restriction
endonuclease Taql (Table 4-2). In this case, the restriction pattern of strains ofX.
campestris pv. vesicatoria group B and X. fragariae were identical to each other and
also to strains of another group of plant pathogenic xanthomonads that included X.
campestris pv. armoraciae, X. campestris pv. begoniae, X campestris pv. campestris,
X campestris pv. fici group B, X. campestris pv. incanae, and V campestris pv.
raphani (Table 4-2). The 840-bp /^-related fragment was not amplified from the
strain ofX. campestris pv. holcicola (Table 4-1).
On the contrary, strains of the other pathovars of X. campestris usually
produced restriction banding patterns that were common to other groups of
xanthomonads in at least one combination of Arp-related fragment and restriction
endonuclease (Table 4-2). However, the combined profiles of four endonucleases and
two Arp-related fragments provided differentiation for almost all groups of plant
pathogenic xanthomonads included in this study (Table 4-2). Among the exceptions
are the strain 9561-1 ofX. campestris pv. incanae and strain #16 of26 campestris pv.


Table D-2Continued
Species/Pathovar
Species/Pathovar
14
15
16
17
18
19
20
21
22
23
1. pv. alfalfae
.9333
.8696
.8696
.9333
.9333
.9545
.8696
1.000
.8261
.6190
2. pv. alfalfae
.9333
.8696
.8696
.9333
.9333
.9545
.8696
1.000
.8261
.6190
3. pv. armoraciae
.4444
.5217
.5217
.4889
.4444
.4091
.5217
.4091
.5217
.5238
4. pv. begoniae A
.6667
.7442
.7442
.7143
.6667
.6341
.7442
.6341
.7442
.9231
5. pv. begoniae A
.6667
.7442
.7442
.7143
.6667
.6341
.7442
.6341
.7442
.9231
6. pv. begoniae B
.6667
.7442
.7442
.7143
.6667
.6341
.7442
.6341
.7442
.9231
7. pv. bilvae
.5217
.5957
.5957
.5652
.5217
.4889
.5957
.4889
.5957
.6047
8. pv. campestris
.4889
.5652
.5652
.5333
.4889
.4545
.5652
.4545
.5652
.5714
11. pv. citri A
.4444
.5217
.5217
.4889
.4444
.4091
.5217
.4091
.5217
.5714
12. pv. citri B
.4444
.5217
.5217
.4889
.4444
.4091
.5217
.4091
.5217
.5714
13. pv. citri C
.4444
.5217
.5217
.4889
.4444
.4091
.5217
.4091
.5217
.5714
14. pv. citrumelo
.9362
.9362
.8696
1.000
.8889
.9362
.9333
.8936
.6512
15. pv. citrumelo
1.000
.9362
.9362
.8261
1.000
.8696
.9583
.7273
16. pv. citrumelo
.9362
.9362
.8261
1.000
.8696
.9583
.7273
17. pv. ctirumelo
.8696
.8889
.9362
.9333
.8936
.6977
18. pv. citrumelo
.8889
.9362
.9333
.8936
.6512
19. pv. citrumelo
.
.8261
.9545
.8696
.6190
20. pv. citrumelo
.
.8696
.9583
.7273
21. pv. citrumelo
#
.
.8261
.6190
22. pv. citrumelo
.
#
.7273
23. pv. dieffenbachiae A



Continued on following page


BIOGRAPHICAL SKETCH
Rui Pereira Leite, Jr., was bom in Sao Pedro, Sao Paulo, Brazil on May 26,
1957, to Ruy Pereira Leite and Daysy Fortes Pereria Leite. He attended Rafael de
Moura Campos Elementary and Instituto de Educado Cardoso de Almeida High
School, Botucatu, Sao Paulo, where he graduated in 1974. In 1975, Rui enrolled at the
Faculdade de Cincias Agronmicas of the Universidade Paulista Jlio de Mesquita
Filho UNESP, Botucatu, Sao Paulo, where he majored in agriculture and took courses
in plant pathology from Dr. Chukichi Kurozawa. Rui graduated with a B.S. from
Universidade Paulista Jlio de Mesquita Filho UNESP in 1978. After graduating, Rui
took the position of plant pathologist in the Department of Plant Pathology at the
Funda9o Instituto Agronmico do Paran IAPAR, Londrina, Paran. At IAPAR,
Rui carried out research on characterization and control of bacterial diseases of fruit
crops. In August 1984, Rui got a study leave and enrolled in the graduate program of
the Department of Plant Pathology, at the University of Wisconsin, Madison. In
Madison, Rui worked with Dr. Doug Rouse on the ecology of Pseudomonas syringae
pv. syringae that causes the bacterial brown spot on snap beans. Rui graduated with a
M.S. degree in plant pathology from the University of Wisconsin in 1986 and returned
to his position in the Department of Plant Pathology at IAPAR. In August 1990, Rui
enrolled in the graduate program at the University of Florida, Department of Plant
Pathology, with Dr. Robert Stall as advisor. Rui was married in 1980 to Claudia
Stulzer, and they have two sons, Ruy and Rafael.
245


21
The enzymatic API ZYM galleries (API Systems S.A., La-Balme-les-Grottes,
Montalieu-Vercieu, France), has also been tested for differentiation and identification
of plant pathogenic xanthomonads (Vauterin et al., 1990b). The enzymatic profiles of
X. campestris pv. begoniae and X. campestris pv. pelargonii were investigated with the
API ZYM system. The profiles were highly uniform within each pathovar, and they
could be differentiated by three enzymatic activities, chymotrypsin, a-D-galactosidase,
and a-D-glucosidase. However, the API ZYM system did not seem to be
discriminative at the pathovar level when larger numbers of pathovars were included in
the study (Vauterin et al., 1993).
Nucleic Acids Analysis
A number of techniques have been used for characterization of nucleic acids of
the xanthomonads. DNA-rRNA hybridization has been of major importance in
differentiation of the xanthomonads at the intrageneric level (De Vos and De Ley,
1983; Palleroni et al., 1973), whereas DNA-DNA hybridization has become the basis
for determining genetic relationships among bacterial strains and for establishing
classification schemes at the subgeneric level (Holt et al., 1994; Krieg and Holt, 1984;
Wayne et al., 1987). Furthermore, DNA-DNA reassociation is considered to be the
reference standard to determine the phylogeny, and consequently the taxonomy of
bacteria (Wayne et al., 1987). RFLP analysis of genomic DNA has been useful to
distinguish several groups of xanthomonads, to study genetic diversity, and to establish
phylogenetic relationships among different pathovars of X. campestris (Gabriel et al.,
1988, 1989; Gottwald et al., 1991; Graham et al., 1990; Hartung and Civerolo, 1989,
1991; Lazo and Gabriel, 1987; Lazo et al., 1987; Leach et al., 1990, 1992; Qhobela and
Claflin, 1992; Qhobela et al., 1991; Verdier et al., 1993).


26
campestris pv. citri, a wide diversity of genomic fingerprints (Hartung and Civerolo,
1987) was found. The genetic diversity ofX. campestris pv. citrumelo was also
determined by analysis of restriction patterns derived from pulsed-field gel
electrophoresis of genomic DNA fragments generated by rare-cutting endonucleases
(Egel et al., 1991). In contrast to the diversity of the strains of2f campestris pv.
citrumelo, strains of X. campestris pv. citri canker A and B produced characteristic
restriction patterns by pulsed-field gel electrophoresis (Egel et al., 1991).
Attempts have also been made to differentiate X. campestris pathovars by RFLP
analysis of plasmid and genomic DNAs based on hybridization with different DNA
probes. In most of the cases, random cloned fragments of the bacterial genome were
used as DNA probes in RFLP analysis of the xanthomonads. Specific regions of the
genome have been selected for DNA probes, such as rRNA sequences (Berthier et al.,
1993; DeParasis and Roth, 1990), repetitive DNA sequences (Leach et al., 1990, 1992),
copper resistance genes (Garde and Bender, 1991), and plasmid DNA sequences
(Hartung, 1992; Gilbertson et al., 1989; Lazo and Gabriel, 1987). Lazo et al. (1987)
used a cloned DNA fragment derived from X campestris pv. citrumelo for RFLP
analysis to differentiate pathovars of X. campestris. The RFLP analysis involving
strains representing 26 pathovars of X. campestris revealed profiles highly conserved
and characteristic for each pathovar tested. By using more than one DNA probe, or by
digesting the genomic DNA with different restriction endonucleases, it was possible to
differentiate all the pathovars ofX. campestris included in the study (Lazo et al., 1987).
An rRNA probe was also used to distinguish pathovars of X. campestris (Berthier et al.,
1993). Further, the RFLP patterns established for the pathovars X. campestris pv.
begoniae, X campestris pv. dieffenbachiae, X campestris pv. malvacearum, and X.


65
Table 4-1. Amplification of the hrp-related fragments from strains of Xanthomonas
campestris and related Xanthomonas spp.
Species/Pathovar
No. of strains
tested
No. of strains with positive
amplification of /irp-related fragment
840 bp 1,075 bp
X campestris
pv. alfalfae
2
2
2
pv. armoraciae
3
3
3
pv. begoniae
9
9
9
pv. bilvae
1
1
1
pv. campestris
9
9
9
pv. carotae
7
7(7)a
7
pv. celebensis
1
0
0
pv. citri
canker A
6
6
6
canker B
7
7
7
canker C
6
6
6
pv. citrumelo
16
16
16
pv. dieffenbachiae
10
10
10
pv. fici
8
5
5
pv. gardneri
3
3(3)
3
pv. glycines
7
7
7
pv. holcicola
1
0
1
pv. incanae
1
1
1
pv. maculifoliigardeniae
1
1
1
pv. malvacearum
9
9
9
pv. manihotis
1
1
1
pv. papavericola
1
1(1)
1
pv. pelargonii
8
7(7)
7
pv. phaseoli
7
5
5
pv. phaseoli "fuscans"
1
1
1
pv. physalidicola
1
1
1
pv. poinsettiicola
7
4
6
pv. pruni
5
4
4
pv. raphani
4
4
4
pv. secalis
1
0
0
pv. taraxaci
1
1(D
1
Continued on following page


189
hrp-related sequences. Fifty different restriction fragment length polymorphism groups
of plant pathogenic xanthomonads were established based on the banding patterns
generated by the analysis of two hrp-related fragments digested with four
endonucleases. Although the banding pattern for a combination /zrp-related fragment
and endonuclease may be shared by more than one group, the profile of all
combinations hrp fragments and endonucleases allowed the specific differentiation and
identification of the different plant pathogenic xanthomonads included in the study.
However, some strains of different groups of plant pathogenic xanthomonads could not
be distinguished by restriction analysis of the /^-related sequences. These include a
strain of X. campestris pv. incanae and one of X. campestris pv. carotae which were
identical to strains ofX. campestris pv. campestris; a strain of A! campestris pv.
phaseoli identical to strains of X. campestris pv. glycines group A; strains F59 and F86
of A! campestris pv. citrumelo were indistinguishable from strains ofX. campestris pv.
pruni; strain XI25 ofX. campestris pv. fici group A was identical to strains ofX.
campestris pv. poinsettiicola group A; strain XV2 of X. campestris pv. vitians was
identical to strains of X. campestris pv. malvacearum; and strain X52 isolated from
Hibiscus sp. was indistinguishable from strains of X. campestris pv. poinsettiicola
group B. No further investigation was carried out with those strains, but it may be
necessary a better assessment of the identity of these strains by using different
approaches.
To illustrate the usefulness of the amplification and analysis of /np-related
fragments in the study of plant pathogenic xanthomonads, I examined the
xanthomonads associated with pepper and tomato seeds. A complex xanthomonad
community was found in association with these seeds as determined by restriction
analysis of /zr/j-related DNA sequences and confirmed by fatty acid analysis. Among


141
M 1 2 3 4 5 6 7 89 10 11 12 13 141516171819
Fig. 6-5. Restriction analysis of the 1,075-bp DNA fragment of the complementation
groups C/D of the hrp gene cluster amplified from strains of Xanthomonas campestris
pv. citri restricted with the endonuclease Haelll. Lanes: M, phage X restricted with the
Pstl; 1-6, group A strains 9771, 3340, 9760-2, 3213, Tl, and 115-A; 7 to 13, group B
strains B64, B69, B80, B84, B93, B94 and B148; 14 to 18, group C strains 338, 339,
340, 341, and 342; 19, X. c. pv. vesicatoria strain 75-3. Molecular sizes are given in
bases.


Table 4-2. Restriction pattern of the 840- and 1.075-bp /zrp-related DNA fragments amplified from different strains of plant
pathogenic Xanthomonas spp. and digested with four restriction endonucleases
Species/Pathovar
Strain
840 bp
1,075 bp
Taq\
5au3AI HaelU
Cfo\
Taql
Saw3AI HaelU
Cfol
X. campestris
pv. alfalfae
G-22 (KS)
8a
6
16
12
8
11
13
8
82-1
8
6
16
12
8
13
13
8
pv. armoraciae
63-27,417, 756
1
3
3
4
3
4
3
2
pv. begoniae
XCB9, X274, X329, X627,
XI490, XI492, XI496
1
6
14
10
9
14
20
15
X610
1
6
14
10
9
14
21
15
X281
1
6
14
10
9
14
24
22
pv. bilvae
XCB
4
5
9
9
7
10
10
9
pv. campestris
33913, 62-1, 62-9a, 65-6b, 70-3,
70-5, 71-2, 83-1. 83-2
1
3
3
3
3
4
3
2
pv. carotae
#3, #5, #7, #9, #12, #13
ndb
nd
nd
nd
3
2
5
2
#16
nd
nd
nd
nd
3
4
3
2
Continued on following page


23
1993). Two minor clusters were also established. One group comprises the pathovars
X. campestris pv. campestris, X. campestris pv. plantaginis, and X campestris pv.
raphani, and the other group comprises X. campestris pv. celebensis and X. campestris
pv. juglandis (Palleroni et al., 1993).
Although the genetic relatedness among plant pathogenic xanthomonads which
cause disease in different plants does not support a correlation between genomic
groupings and pathogenicity features, there are some exceptions. The pathovars ofX.
campestris which cause diseases on closely related leguminous plants, X. campestris
pv. glycines, X campestris pv. lespedezae, X. campestris pv. phaseoli, X. campestris
pv. phaseoli "fuscans", and X campestris pv. vignicola, cluster in DNA homology
group VIII of Vauterin et al. (1993); X. campestris pv. pisi was the only leguminous
pathogen not included in this group (Hildebrand et al., 1990; Palleroni et al., 1993;
Vauterin et al., 1993). However, the leguminous pathogens are also genetically highly
related to strains of several pathovars of X. campestris that cause diseases on different
plants (Hildebrand et al., 1990; Palleroni et al., 1993; Vauterin et al., 1993). The
strains ofX. campestris pv. arrhenatheri, X. campestris pv. cerealis, X. campestris pv.
graminis, X. campestris pv. hordei, X. campestris pv. phlei, X campestris pv.
phleipratensis, X campestris pv. poae, X. campestris pv. secalis, and X. campestris pv.
translucens, that cause diseases on gramineous hosts cluster together in the DNA
homology group IX of Vauterin et al. (1993). This grouping also correlates strongly
with groupings based on SDS-PAGE and fatty acid analysis (Vauterin et al., 1992).
The DNA homology group XII contains the six pathovars from crucifers, X campestris
pv. aberrans, X campestris pv. armoraciae, X. campestris pv. barbareae, X campestris
pv. campestris, X. campestris pv. incanae, and X campestris pv. raphani (Vauterin et
al., 1993).


180
the group B ofX. campestris pv. vesicatoria, and two to group C (Table 7-1). The fatty
acid analysis did not identify the groups of strains and strains 75-0-3 and 9310 were
similar to the fatty acid library ofX. campestris pv. manihotis, and the strains T-93-23
and 7B-0-1 that were very similar to X campestris pv. alfalfae and X campestris pv. fici,
respectively (Table 7-1). Most of the strains that produced hypersensitive reaction on
pepper and tomato were identified as X. campestris pv. armoraciae, X. campestris pv.
campestris, X. campestris pv. carotae, or X campestris pv. raphani, with a high degree of
agreement in the identification by analyses of DNA sequences related to the hrp genes
and by fatty acid composition (Table 7-1). Fifteen strains isolated from pepper seeds
were identified as X. campestris pv. carotae by analyses of DNA sequences related to the
hrp genes and fatty acid composition, though some variability was observed on the
reaction on plants (Table 7-1). The restriction analysis of the /^-related fragments
amplified from the strains SP290.92, 157, T1083, DM-1, LM-1. 9311, 724-4, BSA1,
BSA2, and BSA3 produced pattern different from those determined for 50 different
groups of plant pathogenic xanthomonads (Chapter 4), or the yield obtained in the DNA
amplification was too low for restriction fragment length polymorphism analysis.
Therefore, the identification of these strains on the basis of the restriction profile of the
hrp-related fragments to pathovar or group of X. campestris was not possible.
Discussion
There is an increased need for specific and sensitive methods for detection of
plant pathogenic bacteria associated with seeds and other plant propagative materials
(Miller and Martin, 1988; Schaad, 1982; Saettler et al., 1989). Based on the results
presented here, the amplification and analysis of DNA sequences related to the hrp genes


69
Fig. 4-2. Restriction profiles established for the 840-bp hrp-related fragment amplified
from different strains of plant pathogenic Xanthomonas spp. and restricted with the
endonuclease HaeIII. Lane M, phage X restricted with Pstl. Molecular sizes are given
in bases. See Table 4-2 for identification of strains of Xanthomonas spp. for each
restriction pattern.


7
nature of antigenic determinants in immunogens from whole cell preparations is likely
to play a major role in the low specificity of polyclonal antibodies. Consequently,
cross-reactions with different plant pathogenic xanthomonads pose a limitation in the
identification of specific pathovars of X. campestris when these antibodies are used.
Attempts have been made to increase specificity of polyclonal antisera by
different preparations of immunogens or by partially purified antigen preparations.
Thaveechai and Schaad (1984) compared four different immunogen preparations of
whole bacterial cells. However, none of the immunogens was specific enough in
immunofluorescence to differentiate X. campestris pv. campestris from other pathovars
of X. campestris (Thaveechai and Schaad, 1984). Partially purified antigens including
enzymes, membrane proteins, and ribosomes have also been used as immunogens.
However, polyclonal antisera produced against partially purified antigens have not
provided enough specificity to distinguish among different pathovars of X. campestris.
Thaveechai and Schaad (1984) typed 25 strains of2f campestris pv. campestris in four
serotypes by using antisera produced against ribosomes. However, the antiserum
cross-reacted with strains of X campestris pv. malvacearum, X. campestris pv.
translucens, and X. campestris pv. vesicatoria (Thaveechai and Schaad, 1984). Similar
results were obtained with polyclonal antisera produced against membrane proteins of
X. campestris pv. translucens (Azad and Schaad, 1988). Although strains of A!
campestris pv. translucens from different gramineous hosts were grouped in two
serotypes by using antisera to membrane proteins, X. campestris pv. begoniae was
serologically indistinguishable from X. campestris pv. translucens in Ouchterlony and
immunofluorescence tests (Azad and Schaad, 1988). Further, no correlation was
observed between host of origin and serotype (Azad and Schaad, 1988).


187
species. In all cases, each set of primers amplified DNA fragments identical in size,
suggesting a high degree of structural conservation between operon, as seen with the
primers RST21 and RST22. Furthermore, the pathogenicity of several nonpathogenic
Tn3-gus mutants of X campestris pv. vesicatoria 85-10 was fully restored with cloned
regions of DNA o2l campestris pv. vesicatoria group B and X. campestris pv.
pelargonii, from which the hrp-rdated fragments were amplified. This supports the
contention that the fragments were amplified from DNA sequences which also control
the pathogenicity in other xanthomonads.
In contrast to the narrow spectrum of oligonucleotide primers previously
described for detection and identification of only certain strains of X. campestris
(Garde and Bender, 1991; Hartung, 1992), the hrp-specific primer sets RST2 plus
RST3 and RST21 plus RST22 were useful for identification of a large number of plant
pathogenic xanthomonads. This is not surprising, because the hrp region seems to be
very conserved among different plant pathogenic xanthomonads as determined in the
present and previous studies (Bonas et al., 1991; Stall and Minsavage, 1990). On the
other hand, the primers RST9 and RST10 allowed amplification of DNA fragment only
from a limited number of pathovars of X. campestris, although the hrp fragment
amplified with these primers from X. campestris pv. vesicatoria 75-3 hybridized to the
majority of the plant pathogenic xanthomonads included in this study. These results
indicate differences in the DNA sequences of the xanthomonads corresponding to one
or both primers used. However, the set of primers RST9 and RST10 seems useful for
specific detection of strains ofX. campestris pv. vesicatoria group A, X. campestris pv.
ci,X. campestris pv. physalidicola, and X campestris XI98.
Although no size variation was observed for the /irp-related fragments
amplified from different plant pathogenic xanthomonads, the restriction analysis of the


Sowell, G., Jr., and Dempsey, A. H. 1977. Additional sources of resistance to
bacterial spot of pepper. Plant Dis. Reptr. 61: 684-686.
241
Stall, R. E. 1993. Xanthomonas campestris pv. vesicatoria: Cause of bacterial spot of
tomato and pepper. Pages 57-60 in: Xanthomonas. J. G. Swings and E. L. Civerolo,
eds. Chapman & Hall, London, United Kingdom.
Stall, R. E., Beaulieu, C., Egel, D., Hodge, N. C., Leite, R. P., Minsavage, G. V.,
Bouzar, H., Jones, J. B., Alvarez, A. M., and Benedict, A. A. 1994. Two genetically
diverse groups of strains are included in Xanthomonas campestris pv. vesicatoria. Int.
J. Syst. Bacteriol. 44: 47-53.
Stall, R. E., and Civerolo, E. L. 1991. Research relating to the recent outbreak of
citrus canker in Florida. Annu. Rev. Phytopathol. 29: 399-420.
Stall, R. E., and Hodge, N. C. 1989. Use of fatty acid profiles to identify strains of
Xanthomonas campestris pv. citri from the citrus canker epidemic in Florida. (Abstr.)
Phytopathology 79: 376
Stall, R. E., and Minsavage, G. V. 1990. The use of hrp genes to identify
opportunistic xanthomonads. Pages 369-374 in: Proceedings of the 7th International
Conference on Plant Pathogenic Bacteria. Z. Klement, eds. Akadmiai Kiad,
Budapest, Hungary.
Stall, R. E., and Seymour, C. P. 1983. Canker, a threat to citrus in the gulf-coast
states. Plant Dis. 67:581-585.
Steffan, R. J., and Atlas, R. M. 1988. DNA amplification to enhance the detection of
genetically engineered bacteria in environmental samples. Appl. Environ. Microbiol.
54:2185-2191.
Steffan, R. J., and Atlas, R. M. 1991. Polymerase chain reaction: Applications in
environmental microbiology. Annu. Rev. Microbiol. 45:137-161.
Steffan, R. J., Goksoyr, J., Bej, A. K., and Atlas, R. M. 1988. Recovery of DNA from
soils and sediments. Appl. Environ. Microbiol. 54:2908-2915.
Taylor, J. 1962. The estimation of numbers of bacteria by ten-fold dilution series. J.
Appl. Bacteriol. 25:54-61.


74
Fig. 4-7. Restriction profiles established for the 1,075-bp hrp-related fragment
amplified from different strains of plant pathogenic Xanthomonas spp. and restricted
with the endonuclease Sau3Ad. Lane M, phage X restricted with Pstl. Molecular sizes
are given in bases. See Table 4-2 for identification of strains of Xanthomonas spp. for
each restriction pattern.


GENETIC AND EVOLUTIONARY CHARACTERIZATION OF
PLANT PATHOGENIC XANTHOMONADS BASED ON DNA
SEQUENCES RELATED TO THE hrp GENES
By
RUI PEREIRA LEITE, Jr.
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1994

ACKNOWLEDGMENTS
I would like to express my sincere gratitude to Dr. Robert Stall for his constant
support and encouragement during the course of this doctoral program. His vast
expertise and willing to offer all necessary support were invaluable. I would also like
to thank the other members who served on my supervisory committee, Dr. Jeffry Jones,
Dr. James Preston, and Dr. Daryl Pring, for their support and suggestions. I am
indebted to Dr. Jerry Bartz for reviewing this dissertation.
Appreciation is extended to the Instituto Agronmico do Paran (IAPAR), which
gave permission for this study leave and provided financial support, and to the
Conselho Nacional de Pesquisa (CNPq), which has granted a scholarship for this
graduate program.
I am indebted to Dr. Ulla Bonas of the Institut des Sciences Vgtales, Centre
National de la Recherche Scientifique, Gif-sur-Yvette, France. She provided the DNA
sequences of the hrp genes that were the keystone of this work. Appreciation is
extended to Gail Somodi, Agricultural Research and Education Center in Bradenton,
FL, who performed the ELISA tests. Jay Harrison, IFAS consulting Division of the
Department of Statistics, was very helpful with the SAS programming.
Special thanks are given to Jerry Minsavage for his invaluable technical assistance
and friendship, and to Cheri Hodge for the fatty acid analysis and her friendship.
Finally, I do not have enough words to thank my family, Claudia, my wife, and our
two sons, Ruy and Rafael. They have provided unlimited understanding and support
during these years of graduate studies. They have shared their love, their dreams, and
11

their lives. I am also indebted to my own parents who have offered whatever support
was needed.
in

TABLE OF CONTENTS
page
ACKNOWLEDGMENTS ii
ABSTRACT vii
CHAPTERS
1 INTRODUCTION 1
2 REVIEW OF LITERATURE 5
Serology 5
Fatty Acids Composition 10
Protein Profiles 12
Phenotypic Characteristics 17
Nucleic Acids Analysis 21
3 OLIGONUCLEOTIDE PRIMERS FOR DETECTION AND
IDENTIFICATION OF PLANT PATHOGENIC STRAINS OF
Xanthomonas BY AMPLIFICATION OF DNA SEQUENCES
RELATED TO THE hrp GENES OF Xanthomonas campes tris
PV. VESICATORIA 30
Material and Methods 33
Results 39
Discussion 51
4 CHARACTERIZATION OF PLANT PATHOGENIC
Xanthomonas BASED ON RESTRICTION ANALYSIS OF
AMPLIFIED DNA SEQUENCES RELATED TO THE hrp
GENES 57
Material and Methods 61
Results 64
Discussion 91
iv

page
5 PHYLOGENETIC ANALYSIS OF PLANT
PATHOGENIC Xanthomonas BASED ON DNA
SEQUENCES RELATED TO THE hrp GENES 100
Material and Methods 103
Results 105
Discussion 121
6 GENETIC ANALYSIS OF hrp RELATED DNA
SEQUENCES OF Xanthomonas campestris STRAINS
CAUSING DISEASES OF CITRUS 127
Material and Methods 129
Results 134
Discussion 149
7 EVALUATION OF A DNA AMPLIFICATION
APPROACH FOR DETECTION AND IDENTIFICATION
OF PLANT PATHOGENIC XANTHOMONADS
ASSOCIATED WITH PEPPER AND TOMATO SEEDS 154
Material and Methods 157
Results 167
Discussion 180
8 SUMMARY AND CONCLUSIONS 186
APPENDICES
A BACTERIAL STRAINS AND PLASMIDS USED IN
THIS STUDY 195
B SAS PROGRAM TO ESTIMATE THE SIMILARITY
VALUES FROM RESTRICTION FRAGMENT DATA 201
C SAS PROGRAM TO ESTIMATE THE NUCLEOTIDE
SUBSTITUTION BASED ON RESTRICTION
FRAGMENT DATA 204
v

page
D ESTIMATES OF SIMILARITY VALUES FOR
PLANT PATHOGENIC XANTHOMONADS BASED
ON COMPARISON OF RESTRICTION FRAGMENT
DATA OF THE hrp RELATED DNA SEQUENCES 205
LITERATURE CITED 226
BIOGRAPHICAL SKETCH 245
vi

Abstract of Dissertation Presented to the Graduate School of the University of Florida
in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
GENETIC AND EVOLUTIONARY CHARACTERIZATION OF
PLANT PATHOGENIC XANTHOMONADS BASED ON DNA
SEQUENCES RELATED TO THE hrp GENES
By
Rui Pereira Leite, Jr.
August 1994
Chairman: Robert E. Stall
Major Department: Plant Pathology
Three pairs of oligonucleotide primers specific for different regions of the
hypersensitive reaction and pathogenicity (hrp) genes of Xanthomonas campestris pv.
vesicatoria were designed and used for amplification of homologous DNA sequences
from X fragariae and from 28 pathovars of X campestris. No amplification occurred
with genomic DNA from plant pathogenic strains of X. campestris pv. secalis, X.
campestris pv. translucens, and X albilineans, from nonpathogenic xanthomonads, or
from plant pathogenic strains of the genera Acidovorax, Agrobacterium, Clavibacter,
Erwinia, Pseudomonas, and Xylella. DNA fragments amplified with a particular
primer pair were of identical size from each of the different plant pathogenic
xanthomonads. However, the restriction analysis of these fragments by using frequent-
cutting endonucleases revealed variation in the banding pattern of these hrp-related
fragments. The banding patterns allowed distinction of the strains representing a
pathovar or species of plant pathogenic xanthomonads.
vii

The population structure of some taxa, i.e. X. jragariae, X. campestris pv.
begoniae, X campestris pv. campestris, X. campestris pv. malvacearum, and X.
campestris pv. pelargonii, was homogeneous. However, the majority of the pathovars
of X campestris presented a high degree of sequence variability in the /^-related
fragments for the strains within pathovars. For instance, the citrus pathogens X.
campestris pv. citri strains in groups A, B, and C, which cause citrus canker A, B, and
C, respectively, and X. campestris pv. citrumelo strains in the highly aggressive group,
which cause the citrus bacterial spot, had a characteristic and homogeneous restriction
banding pattern within each group. In contrast, restriction fragment polymorphism was
evident among strains of the moderately and weakly aggressive groups of X. campestris
pv. citrumelo.
The phylogenetic analysis of the hrp genes revealed a diverse evolutionary
relationship for this region of the bacterial genome of the xanthomonads. Whereas the
hypothesis of coevolution of the hrp region with the rest of the genome from a common
ancestor is supported, there is also evidence of horizontal movement of the hrp genes
between the different plant pathogenic xanthomonads. Nevertheless, the results
obtained do not substantiate the existence of genetic and functional selection pressure
regarding the hrp genes within the xanthomonads. One of the major findings of this
study is the divergence between hrp gene sequence and host specificity, which
indicates that factors other than the hrp genes are likely to be involved in host
speciation.
Strains of plant pathogenic xanthomonads could be detected by amplification of
/irp-related DNA sequences from washings of pepper and tomato seeds. The diversity
of the xanthomonads found on seeds assessed on the basis of restriction analysis of
/irp-related sequences agreed very closely with the population structure determined by
viii

using other methods. The restriction banding profiles generated for the hrp fragments
may be an easy and a discriminating approach for identification of plant pathogenic
xanthomonads.
IX

CHAPTER 1
INTRODUCTION
The genus Xanthomonas Dowson 1939 includes Gram-negative and usually
yellow-pigmented bacteria that occur worldwide and cause disease on many plants.
The host range spans over 124 monocotyledonous and 268 dicotyledonous plant
species (Hayward, 1993; Leyns et al., 1984). Among the species of Xanthomonas, X.
campestris is the most diverse and includes at least 125 different pathovars that are
distinguished by causing different diseases of plants (Bradbury, 1984; Dye et al.,
1980). Although the determination of the genus Xanthomonas and its species presents
no great problem, the identification and characterization at the subgeneric level is still
difficult (Van den Mooter and Swings, 1990; Vauterin et al., 1990a, 1991a; Young et
al., 1992). The basis for differentiation of plant pathogenic bacteria at the subgeneric
level is by the capability of the bacterial strain to cause characteristic disease, or by
reference to their host range (Dye et al., 1980; Vauterin et al., 1990a; Young et al.,
1992). However, pathogenic reactions are variable. Strains may differ in
aggressiveness, causing a broad range of symptoms when artificially inoculated into
host plant species. The strains may also differ in the pathogenic reaction when
artificially inoculated onto nonhost plants. Furthermore, these pathovars may not be
readily differentiated by other phenotypic features (Van den Mooter and Swings,
1990). In addition, yellow-pigmented bacteria apparently belonging to the genus
Xanthomonas and nonpathogenic to the plants from which they were isolated are
usually associated with plant material (Angeles-Ramos et al., 1991; Gitaitis et al.,
1

2
1987; Maas et al., 1985; Liao and Wells, 1987; Stall and Minsavage, 1990). These
opportunistic bacteria can be identified as xanthomonads by the presence of
xanthomonadins and by unique fatty acid profiles.
Attempts have been made to differentiate the pathovars and strains of A
campestris with several techniques. Traditional methods for the identification and
characterization of plant pathogenic xanthomonads rely upon performing
predetermined biochemical, serological, and pathological tests on pure cultures of the
bacteria (Bradbury, 1984; Holt et al., 1994; Schaad, 1988). More recently, methods
based on metabolic and protein profiles (Chase et al., 1992; Hildebrand et al., 1993;
Van den Mooter and Swings, 1990; Vauterin et al., 1991ab), and fatty acid analysis
(Chase et al., 1992; Yang et al., 1993) have been used for taxonomic and identification
purposes. Polyclonal and monoclonal antisera produced against strains ofX.
campestris have been used for detection and identification, but they have provided
variable results (Alvarez and Lou, 1982; Alvarez et al., 1991; Benedict et al., 1990;
Jones et al., 1993b). Nucleic acid based techniques have become a premier approach
for detection and identification of plant pathogenic bacteria (Bereswill et al., 1992;
Manulis et al., 1991; Schaad et al., 1989; Seal et al., 1992), including members of the
xanthomonads (Berthier et al., 1993; Garde and Bender, 1991; Gilbertson et al., 1989;
Hartung, 1992; Hartung et al., 1993; Hildebrand et al., 1990; Lazo and Gabriel, 1987;
Lazo et al., 1987; Leach et al., 1990,1992).
Genetic relationships among plant pathogenic xanthomonads have been
established based on DNA-DNA homology, total genomic fingerprinting, and analysis
of restriction fragment length polymorphism (RFLP) in their DNAs (Vauterin et al.,
1990a). DNA-DNA hybridization studies have shown great variation in the relatedness
ofX. campestris pathovars (Hildebrand et al., 1990; Palleroni et al., 1993), and the

3
strains ofX. campestris were divided into 15 DNA homology groups (Vauterin et al.,
1993). Comparisons by genomic fingerprinting and RFLP analysis have also shown
significant genetic diversity among and within different pathovars of X. campestris
(Egel et al., 1991; Lazo et al., 1987; Stall et al., 1994). Furthermore, physiologically
and biochemically similar xanthomonads that are pathogenic on the same host plant
may have a diverse genetic background (Egel et al., 1991; Hildebrand et al., 1990;
Palleroni et al., 1993; Stall et al., 1994). On the other hand, the application of
conventional genetic analysis for routine identification, as well as for taxonomic
purposes, may be hampered by some limiting features. The determination of DNA-
DNA homologies are limited by technical difficulties and by experimental error in
obtaining the data (Sneath, 1989). Further, rearrangements of the bacterial
chromosome may result in changes in the phenotype and RFLP patterns without
appreciably affecting overall sequence similarity (Egel et al., 1991; Young et al., 1992).
Alternatively, analysis of specific DNA sequences of the bacterial genome should
provide accurate information regarding the relationships of the strains and pathovars of
X. campestris without the shortcomings mentioned above. The procedure may be a
useful tool for the specific detection and identification of plant pathogenic bacteria.
The primary objective of this work is to investigate specific sequences of the
bacterial genome of plant pathogenic xanthomonads for the specific identification of
these plant pathogens. Oligonucleotide primers were designed based on the sequence
of the hrp gene (hypersensitive reaction and pathogenicity) cluster of X. campestris pv.
vesicatoria (Bonas et al., 1991; Ulla Bonas, personal communication), and tested for
the specific amplification of DNA fragments related to the hrp gene cluster from the
bacterial genome of different plant pathogenic xanthomonads by polymerase chain
reaction. DNA fragments related to the hrp genes were amplified from strains of X.

4
fragariae and of 28 pathovars of X. campestris. DNA sequence variation in the hrp
related fragments was investigated by restriction endonuclease analysis, and the
evolutionary relationship of these hrp gene sequences was assessed based on
phylogenetic analysis of the restriction fragment data. The DNA amplification
technique was also examined for specific detection and identification of plant
pathogenic xanthomonads and to detect plant pathogens associated with plant samples.

CHAPTER 2
REVIEW OF LITERATURE
The genus Xanthomonas includes six species containing plant pathogenic
bacteria (Bradbury, 1984, 1986) that occur worldwide and cause economically
important diseases on several plant crops. The species X. campestris is the largest and
consists of at least 125 different pathovars (Bradbury, 1984, 1986). The basis for
differentiation and identification of these plant pathogenic bacteria at the pathovar level
is by means of the capability of the bacterial strain to cause characteristic disease or by
their host range (Dye et al., 1980; Vauterin et al., 1990a; Young et al., 1992). The
identification of the xanthomonads at the genus and species level does not present a
significant problem, and can be achieved by using different techniques, such as
biochemical and physiological tests, serology, protein profiling, and fatty acid and
nucleic acid analysis (Bradbury, 1984; Holt et al., 1994; Schaad, 1988; Vauterin et al.,
1990ab). Despite a range of different techniques available, the differentiation at
infrasubspecific level is still difficult, and it is not uncommon that strains belonging to
different pathovars may be genetically and phenotypically closely related.
Serology
Several attempts have been made to distinguish between pathovars of X.
campestris based on serological reactions. Early work on the use of serology for
differentiation of plant pathogenic xanthomonads by using agglutination and
precipitation tests suggested that this technique would be specific for differentiation
5

6
and identification of this group of bacteria (Link and Sharp, 1927; Williams and Glass,
1931). Elrod and Braun (1947) carried out an extensive study which included different
species of Xanthomonas and several pathovars of X. campestris, thus identifying five
immunological groups which have a high degree of serological homogeneity within
them. However, the serogroups did not show good or significant correlation with
nomenspecies and pathotypes. Subsequent work has confirmed the complex nature of
the serological relationships among different taxonomic groups of the plant pathogenic
xanthomonads.
Polyclonal antibodies produced against whole cells ofX. campestris pv.
campestris cross-reacted with strains of X. campestris pv. armoraciae, X. campestris pv.
phaseoli "fuscans" and X campestris pv. vesicatoria (Franken et al., 1992). In
immunofluorescence microscopy, these polyclonal antibodies also cross-reacted with
non-xanthomonad bacteria. In a similar study, polyclonal antibodies produced to
whole cells of X campestris pv. pelargonii also reacted to strains of seven different
pathovars of campestris (Anderson and Nameth, 1990). Conflicting data have also
emerged regarding the relationship between serotypes and pathotypes. O'Brien et al.
(1967) distinguished pepper strains from tomato strains ofX. campestris pv. vesicatoria
by serological reaction, but no correlation between serotype and pathotype of X.
campestris pv. vesicatoria was found by Charudattan et al. (1973). They found that
tomato and pepper strains of the bacterium included both serotypes. Further, no
correlation was observed for hydrolysis of starch and resistance to streptomycin with
serotypes. These contrasting results are not exclusive to X. campestris pv. vesicatoria,
but have also been observed for other pathovars of X. campestris (Bach et ah, 1978;
Civerolo and Fan, 1982) and even to other plant pathogenic species of Xanthomonas
(Obata and Tsuboi, 1972; Mahanta and Addy, 1977). Furthermore, the heterogeneous

7
nature of antigenic determinants in immunogens from whole cell preparations is likely
to play a major role in the low specificity of polyclonal antibodies. Consequently,
cross-reactions with different plant pathogenic xanthomonads pose a limitation in the
identification of specific pathovars of X. campestris when these antibodies are used.
Attempts have been made to increase specificity of polyclonal antisera by
different preparations of immunogens or by partially purified antigen preparations.
Thaveechai and Schaad (1984) compared four different immunogen preparations of
whole bacterial cells. However, none of the immunogens was specific enough in
immunofluorescence to differentiate X. campestris pv. campestris from other pathovars
of X. campestris (Thaveechai and Schaad, 1984). Partially purified antigens including
enzymes, membrane proteins, and ribosomes have also been used as immunogens.
However, polyclonal antisera produced against partially purified antigens have not
provided enough specificity to distinguish among different pathovars of X. campestris.
Thaveechai and Schaad (1984) typed 25 strains of2f campestris pv. campestris in four
serotypes by using antisera produced against ribosomes. However, the antiserum
cross-reacted with strains of X campestris pv. malvacearum, X. campestris pv.
translucens, and X. campestris pv. vesicatoria (Thaveechai and Schaad, 1984). Similar
results were obtained with polyclonal antisera produced against membrane proteins of
X. campestris pv. translucens (Azad and Schaad, 1988). Although strains of A!
campestris pv. translucens from different gramineous hosts were grouped in two
serotypes by using antisera to membrane proteins, X. campestris pv. begoniae was
serologically indistinguishable from X. campestris pv. translucens in Ouchterlony and
immunofluorescence tests (Azad and Schaad, 1988). Further, no correlation was
observed between host of origin and serotype (Azad and Schaad, 1988).

8
Monoclonal antibodies (MABs) have been examined quite extensively for
specific identification of plant pathogenic xanthomonads. MABs have been produced
for various groups of these plant pathogens, including X albilineans, X. oryzae, and
several pathovars of X. campestris (Vauterin et al., 1993). Further, MABs have been
developed for identification of xanthomonads at different taxonomic levels (Alvarez et
al., 1985; Benedict et al., 1989, 1990). Alvarez et al. (1985) obtained a MAB that
identifies the genus Xanthomonas. This MAB was specific for all 436 xanthomonads
tested, but did not react with bacteria of 12 other genera. MABs were produced that
were specific for X oryzae pv. oryzae and for X. oryzae pv. oryzicola at pathovar level
(Benedict et al., 1989). These pathovar-specific antibodies reacted with all tested
strains of each pathovar, but neither of the MABs reacted with strains of other
pathovars and species, or with strains of other bacterial genera. Also, for MABs
specific for epitopes on the lipopolysaccharide of X. campestris pv. begoniae and X.
campestris pv. pelargonii were produced that were pathovar-specific and reacted only
with their respective strains (Benedict et al., 1990). These monoclonal antibodies did
not react with strains of other xanthomonads or non-xanthomonads tested.
Contrary to the pathovar-specific MABs mentioned above, no pathovar specific
MABs have been found that could be used to specifically detect all strains within
certain pathovars of X campestris, such as X. campestris pv. campestris (Alvarez et al.,
1985; Franken et al., 1992), X. campestris pv. citri (Alvarez et al., 1991), X. campestris
pv. citrumelo (Alvarez et al., 1991; Gottwald et al., 1991; Permar and Gottwald, 1989),
and X campestris pv. dieffenbachiae (Lipp et al., 1992). Based on the reactions of six
antibodies, 200 strains of A! campestris pv. campestris were grouped in six serogroups
(Alvarez et al., 1991). However, this grouping did not imply closer genetic
relationships among members of a group than with members of other groups or

9
geographical distribution of the strains. In a similar study, six MABs were used to
group 323 strains of A campestris pv. dieffenbachiae isolated from different plants of
the Araceae family (Lipp et al., 1992). The strains of A! campestris pv. dieffenbachiae
were serologically heterogeneous and 12 major serogroups were identified. The
serotype did not correspond strongly to the host of origin. Strains of A campestris that
cause diseases on citrus are also serologically diverse (Alvarez et ah, 1990, 1991;
Benedict et ah, 1985; Gottwald et ah, 1991; Permar and Gottwald, 1989). An MAB
developed to A campestris pv. citri of the group A of citrus canker did not react with
strains of B, C, and D forms of citrus canker (Alvarez et ah, 1991). Although this
antibody also did not react with 130 other Aanthomonas pathovars and species, it
reacted with some strains of A campestris pv. citrumelo, the citrus bacterial spot agent,
and with a strain of A campestris pv. manihotis. However, in serological studies with
monoclonal and polyclonal antibodies a close antigenic relationship existed between
the strains of the groups B, C, and D forms of citrus canker (Alvarez et ah, 1991;
Civerolo and Fan, 1982). The citrus bacterial spot pathogen, A campestris pv.
citrumelo, is also serologically heterogeneous, and no single MAB reacted with all
strains of this pathogen (Alvarez et ah, 1991; Gottwald et ah, 1991). Although A
campestris pv. citrumelo is serologically distinct from strains of A campestris pv. citri,
some strains of the citrus bacterial spot pathogen share a common epitope with
members of other pathovars of A campestris (Alvarez et ah, 1991; Gottwald et ah,
1991). The diversity of A campestris pv. citrumelo is also reflected genetically and in
differential host reactions (Alvarez et ah, 1991; Egel et ah, 1991; Gottwald et ah, 1991;
Graham and Gottwald, 1990).
An interesting point emerging from these studies on serological relationships
among different pathovars of A campestris by using MABs is the inverse correlation

10
between the serological heterogeneity and the host range of the xanthomonads.
Pathovar specific MABs have been developed that react specifically with all strains of
a pathovar of X. campestris that usually cause disease in a single host, as for example,
X. campestris pv. begoniae and X. campestris pv. pelargonii. On the contrary, no
pathovar specific MABs were produced that react with all strains of a pathovar that
have a large host range, such as X campestris pv. campestris, X. campestris pv. citri,
and X. campestris pv. dieffenbachiae.
Fattv Acids Composition
The analysis of fatty acids by gas chromatography has become a common
technique in bacterial classification and identification. Comprehensive studies have
been carried out on the relationship of fatty acid profiles within several groups of plant
pathogenic xanthomonads (Chase et al., 1992; Graham et al., 1990; Hodge et al., 1992;
Vauterin et al., 1991b, 1992; Yang et al., 1993). At least 65 different fatty acids were
found within the members of the genus Xanthomonas. Three fatty acids, 11:0 iso, 11:0
iso 30H, and 13:0 iso OH, are characteristic for all xanthomonads and they were useful
for differentiating Xanthomonas from other plant pathogenic bacteria (Vauterin et al.,
1992; Yang et al., 1993). Furthermore, the species X. albilineans, X. fragariae, and Y
populi were homogeneous based on fatty acid composition, and they could be clearly
differentiated from each other (Hodge et al., 1992; Vauterin et al., 1992; Yang et al.,
1993). Strains ofY oryzae have unique fatty acid composition within the genus
Xanthomonas and they form two major groups which correspond to the pathovars X.
oryzae pv. oryzae and X. oryzae pv. oryzicola (Vauterin et al., 1992; Yang et al., 1993).
As observed on other characteristics, X. campestris also comprises a
heterogeneous group in relation to the fatty acid composition (Chase et al., 1992;

11
Hodge et al., 1992; Vauterin et al., 1991b, 1992; Yang et al., 1993). Extensive studies
carried out to study the fatty acid composition of strains of the genus Xanthomonas
included bacteria from more than 100 pathovars oX. campestris (Hodge et al., 1992;
Yang et al., 1993). The members of the X. campestris have been divided in 24 different
groups on the basis of fatty acid composition (Yang et al., 1993). Some pathovars ofX.
campestris, such as X campestris pv. maculifoliigardeniae, X. campestris pv.
malvacearum, and X campestris pv. pelargonii, were homogeneous and had a distinct
fatty acid profiles (Hodge et al., 1992). In some cases, pathovars pathogenic for
members of the same family of plants could be grouped together based on fatty acid
composition. This is the case for the pathovars X. campestris pv. cerealis, X.
campestris pv. hordei, X campestris pv. secalis, X campestris pv. translucens, and X.
campestris pv. undulosa which cause diseases on cereal crops (Vauterin et al., 1992;
Yang et al., 1993), and also for the pathovars from crucifers X campestris pv. aberrans,
X campestris pv. armoraciae, X campestris pv. barbareae, X. campestris pv.
campestris, X. campestris pv. incanae, and X. campestris pv. raphani (Yang et al.,
1993). Many pathovars from leguminous plants, such as X campestris pv. alfalfae, X
campestris pv. glycines, X campestris pv. phaseoli, X campestris pv. phaseoli
"fuscans", X. campestris pv. rhynchosiae, X campestris pv. sesbaniae, and X.
campestris pv. vignicola, also clustered in the same group based on fatty acid analysis
(Yang et al., 1993). This grouping delineated on the basis of fatty acid composition
agreed with the relationships established based on DNA-DNA hybridization and
protein profile (Hildebrand et al., 1990; Vauterin et al., 1991a, 1992; Yang et al.,
1993).
In contrast to some pathovars, the strains of X. campestris causing diseases of
citrus are composed of a heterogeneous group of strains. Besides the heterogeneity in

12
the fatty acid profile, the grouping obtained by fatty acid analysis does not agree
entirely with the grouping obtained by using other methods, such as by genetic analysis
and phenotypic features. Strains of X. campestris pv. citri that cause the citrus canker
disease can be clearly differentiated from strains ofX. campestris pv. citrumelo, the
citrus bacterial spot pathogen (Vauterin et al., 1991b; Yang et al., 1993). However,
there are some discrepancies between the delineation of the citrus canker groups by
fatty acid profile compared to genetic analysis and protein profile (Vauterin et al.,
1991b). Based on fatty acid analysis, the strains of citrus canker B were
indistinguishable from strains of the citrus canker A, but they were distinct from the
strains of groups C and D of citrus canker (Vauterin et al., 1991b). However, the citrus
canker A strains are considered a homogeneous and distinct group from the strains of
citrus canker B, C, and D based on DNA-DNA hybridization, RFLP analysis, and
protein profiles (Egel et al., 1991; Gabriel et al., 1989; Hartung and Civerolo, 1987;
Vauterin et al., 1991b). Contrasting results on fatty acid analysis were also obtained
for the citrus bacterial pathogen, X. campestris pv. citrumelo. Vauterin et al. (1991b)
found that strains within X. campestris pv. citrumelo constituted a distinct and
homogeneous group based on fatty acid composition. On the other hand, Graham et al.
(1990) observed considerable diversity among the X. campestris pv. citrumelo strains,
and some of these strains were closely related to non citrus strains of X. campestris on
the basis of fatty acid composition.
Protein Profiles
El-Sharkawy and Huisingh (197lab) were the first to demonstrate the
usefulness of electrophoresis of proteins for differentiation and identification of
Xanthomonas. The patterns of native negatively charged proteins extracted from whole

13
bacterial cells varied considerably and indicated the occurrence of taxon-specific
protein profiles that could be used for differentiation and identification of plant
pathogenic xanthomonads. The presence of distinct and specific patterns of esterases at
the pathovar level was particularly valuable in differentiating and identifying the
different groups of plant pathogenic Xanthomonas spp. (El-Sharkawy and Huisingh,
1971a). However, a more reliable and reproducible techniques were required to resolve
the protein profile at subspecific level. In this respect, sodium dodecyl sulphate-
polyacrylamide gel electrophoresis (SDS-PAGE) of total cellular proteins has become a
key technique for the separation and characterization of a wide variety of proteins
(Laemmli, 1970; Weber and Osborn, 1960). Furthermore, this technique has been
useful for differentiating the plant pathogenic xanthomonads on the basis of whole-cell
protein (Bouzar et al., 1994; Kersters and De Ley, 1975; Van den Mooter et al.,
1987ab; Vauterin et al., 1990a, 199lab, 1992). Variation in the electrophoretic patterns
of specific proteins of the bacterial cell, such as total cell envelope and total membrane
proteins, has also been investigated by using SDS-PAGE for differentiation of some
pathovars ofX. campestris (Dos Santos and Dianese, 1985; Minsavage and Schaad,
1983; Qhobela and Claflin, 1992; Qhobela et al., 1991).
Vauterin et al. (1991a) analyzed whole-cell proteins of an extensive collection
of plant pathogenic bacteria of the genus Xanthomonas, comprising all species and 27
pathovars of X campestris, by using SDS-PAGE analysis. On the basis of SDS-PAGE
protein profiles, the plant pathogenic species X. albilineans, X. axonopodis, X.
fragariae, and X. oryzae could be distinguished from each other and also from X.
campestris pathovars, whereas X populi was rather similar to X. campestris. Several
distinct protein profiles were observed within X. campestris, and some of the groups
established corresponded to described pathovars of this species. Certain pathovars

14
constituted distinct and homogeneous groups based on protein profiles, such as X.
campestris pv. campestris, X campestris pv. graminis, X. campestris pv. hyacinthi, X.
campestris pv. pelargonii, X. campestris pv. pruni, and X. campestris pv. theicola. On
the other hand, the pathovars X. campestris pv. alfalfae, X. campestris pv. cajani, X.
campestris pv. glycines, X campestris pv. phaseoli, X campestris pv. phaseoli
"fuscans", and X. campestris pv. vignicola which cause diseases on leguminous plants
could not be differentiated from one another. The relatedness of this group of
pathogens of leguminous plants has also been confirmed by DNA-DNA hybridization
(Hildebrand et al., 1990; Vauterin et al., 1990ab) and by fatty acid composition (Yang
et al., 1993). Moreover, some pathovars, including X. campestris pv. alfalfae, X.
campestris pv. citri, X campestris pv. dieffenbachiae, X campestris pv. poinsettiicola,
X campestris pv. vesicatoria, X campestris pv. vignicola, and X. campestris pv.
vitians, were heterogeneous on the basis of protein profiles, consisting of two or more
distinct protein groups (Vauterin et al., 1991b). The diversity on the basis of protein
profile for some of these pathovars, i.e. X campestris pv. citri, X. campestris pv.
dieffenbachiae, and X. campestris pv. vesicatoria, correlates with the heterogeneity
based on genetic analysis and fatty acid composition (Chase at al., 1992; Egel et al.,
1991; Graham et al., 1990; Stall et al., 1994; Vauterin et al., 1991ab).
The complexity of the grouping of strains of X campestris that cause diseases
on citrus on the basis of whole-cell protein profiles (Vauterin et al., 1991ab) is in
agreement with the diversity determined by genetic analysis (Egel et al., 1991; Gabriel
et al. 1989; Hartung and Civerolo, 1987) and fatty acid composition (Graham et al.,
1990; Vauterin et al., 1991b). Based on protein profile, the strains of the citrus canker
A of campestris pv. citri form a well defined and homogeneous group (Vauterin et
al., 1991b). The uniformity of this group was also revealed by DNA-DNA

15
hybridization, DNA fingerprinting, RFLP analyses, and fatty acid composition (Egel et
al., 1991; Gabriel et al., 1987; Graham et al., 1990; Hartung and Civerolo, 1987;
Vauterin et ah, 1991b). The strains of the citrus canker B and D groups ofX.
campestris pv. citri constituted one group, and strains of citrus canker C formed
another distinct group on the basis of protein profile (Vauterin et ah, 1991b). However,
the groupings based on DNA-DNA hybridization, fatty acid composition, and RFLP
analysis do not entirely agree with the groups determined by protein analysis (Vauterin
et ah, 1991b). Although distinct from the citrus canker A strains ofX. campestris pv.
citri, strains of X. campestris pv. citrumelo also showed a quite uniform protein
patterns on the SDS-PAGE (Vauterin et ah, 1991b). This contrasts with the diversity
determined for the strains of this pathovar based on fatty acid composition (Graham et
ah, 1990), RFLP analyses (Gabriel et ah, 1989; Hartung and Civerolo, 1987), and
DNA-DNA hybridization (Egel et ah, 1991).
SDS-PAGE of whole-cell proteins was also used in comparative studies of the
xanthomonads that cause diseases on cassava, X. campestris pv. manihotis and A!
campestris pv. cassavae (Van den Mooter et ah, 1987a). Strains of X. campestris pv.
manihotis constituted a distinct and homogeneous group on the basis of protein
profiles, whereas the strains of X. campestris pv. cassavae formed two groups. These
results also agree with the groupings obtained by numerical analyses of phenotypic
features (Van den Mooter et ah, 1987a), DNA-DNA hybridization (Van den Mooter et
ah, 1987a), and electrophoresis of total membrane proteins (Dos Santos and Dianese,
1985).
X. campestris pv. vesicatoria, the cause of the bacterial spot disease of tomato
and pepper, comprises heterogeneous groups in relation to protein profiles (Bouzar et
ah, 1994; Vauterin et ah, 1991a). This diversity has also been confirmed by genetic

16
analyses and phenotypic characteristics (Stall et al., 1994; Vauterin et al., 1990a). Two
major distinct groups have been delineated by analysis of whole-cell protein profiles
(Bouzar et al., 1994; Vauterin et al., 1991a), utilization of carbon sources (Bouzar et
al., 1992), fatty acid composition (Stall et al., 1994; Vauterin et al., 1990b), and DNA-
DNA hybridization (Stall et al., 1994). Furthermore, these two groups can be easily
distinguished by the presence of unique protein bands on the basis of silver staining of
SDS-lysed cell proteins (Bouzar et al., 1994).
A comparative study was carried out of the Xanthomonas species and X.
campestris pathovars that cause diseases on cereals and grasses by using SDS-PAGE of
whole-cell proteins (Vauterin et al., 1992). The pathovars of the "translucens" group,
which includes X. campestris pv. cerealis, X campestris pv. hordei, X. campestris pv.
secalis, X. campestris pv. translucens, and X. campestris pv. undulosa, could not be
distinguished from one another. The pathogens of grasses which formed another
homogeneous group were X. campestris pv. arrhenatheri, X. campestris pv. phlei, and
X. campestris pv. poae, and had protein profiles very similar to those of the
"translucens" group. DNA-DNA hybridization data, phenotypic features, and fatty acid
composition also support the close relationship between the two groups (Van den
Mooter et al., 1987b; Vauterin et al., 1992). On the other hand, X. axonopodis and X.
campestris pv. vasculorum were heterogeneous, with two distinct groups being found
in each taxon. The type A group of X. axonopodis was very closely related to one
group ofX. campestris pv. vasculorum, whereas the type B group of A! axonopodis had
atypical protein and fatty acid profiles which do not resemble xanthomonads. X
campestris pv. vasculorum type B group was indistinguishable from X. campestris pv.
holcicola based on protein profile, as well as on fatty acid composition and DNA-DNA
hybridization (Vauterin et al., 1992).

17
Phenotypic Characteristics
The differentiation and identification of Xanthomonas spp. have largely
depended on phenotypic characteristics (Bradbury, 1984; Holt et al., 1994; Schaad,
1988). Although members of the xanthomonads can be distinguished from the bacteria
of other genera on the basis of phenotypic features (Bradbury, 1984; Holt et al., 1994;
Schaad, 1988), it has been almost impossible to differentiate within the Xanthomonas
by biochemical and physiological features alone without knowing their plant hosts
(Bradbury, 1984; Dye, 1962; Holt et al., 1994; Schaad, 1988; Van den Mooter and
Swings, 1990). In fact, this was the major reason for the inclusion of the majority of
the former species of Xanthomonas into a single species, X. campestris (Bradbury,
1984; Dye et al., 1980).
Extensive studies have been carried out to compare the phenotypic
characteristics of strains representing the different species of the genus Xanthomonas,
pathovars of X campestris, and other bacteria genetically related to the xanthomonads
(Dye, 1962; Hildebrand et al., 1993; Van den Mooter and Swings, 1990). Van den
Mooter and Swings (1990) examined 295 morphological, biochemical, and
physiological features of 266 strains of Xanthomonas spp. and related bacteria. The
plant pathogenic xanthomonads were placed in seven different groups by cluster
analysis. The groups were delineated by strains of the taxaX. albilineans, X.
axonopodis, X campestris, X. campestris pv. graminis, X fragariae, X. oryzae, and X
populi, which largely correlate with the existing species of the genus Xanthomonas.
Strains ofX. campestris pv. graminis did not cluster with other pathovars ofV
campestris, and they were considered a homogeneous and different group on the basis
of phenotypic characteristics. However, other studies of phenotypic features (Van den

18
Mooter et al., 1987b), DNA-DNA hybridization (Kersters et al., 1989; Vauterin et al.,
1992), SDS-PAGE (Van den Mooter et al., 1987b; Vauterin et al., 1992), and fatty acid
profiles (Vauterin et al., 1992; Yang et al., 1993) have failed to consistently
differentiate strains of X. campestris pv. graminis from other pathovars o X.
campestris which cause diseases on cereals and grasses, such as X. campestris pv.
arrhenatheri, X campestris pv. cerealis, X campestris pv. hordei, X. campestris pv.
phlei, X campestris pv. phleipratensis, X campestris pv. poae, X. campestris pv.
secalis, X. campestris pv. translucens, and X. campestris pv. undulosa. Furthermore,
the misnamed species Pseudomonas gardneri, P. vitiswoodrowii, and P.
mangiferaeindicae were indistinguishable from X. campestris strains on the basis of
phenotypic features (Dye, 1966; Hildebrand et al., 1993; Van den Mooter and Swings,
1990). The latter also agrees with studies of DNA-rRNA hybridization (De Vos et al.,
1985), DNA-DNA hybridization (Hildebrand et al., 1990; Palleroni et al., 1993), and
fatty acid composition (Yang et al., 1993).
X campestris contains by far the largest group of strains within the genus
Xanthomonas, consisting of at least 125 different pathovars (Bradbury, 1984). Despite
certain variations in phenotypic characteristics (Hildebrand et al., 1993), strains ofY
campestris form a rather indistinguishable group (Burkholder and Starr, 1948; Dye,
1962; Van den Mooter and Swings, 1990). In a limited number of cases, however,
distinctive phenotypic traits have been linked to the grouping based on pathogenicity to
a given host species. Two of these cases are the X campestris pv. manihotis and X
campestris pv. cassavae, which cause diseases on cassava. All strains of each pathovar
formed a single group based on phenotypic features (Van den Mooter and Swings,
1990; Van den Mooter et al., 1987a). Correlation between phenotypic features and
pathogenic specialization was also established for the xanthomonads that cause wilt

19
diseases on grasses. The analysis of 257 phenotypic characteristics of strains from
grasses revealed three distinct groups corresponding to (1).V campestris pv. graminis,
and A"! campestris pv. phleipratensis, (2) X campestris pv. arrhenatheri, and (3) X
campestris pv. poae (Van den Mooter et al., 1987b). These groups are also in
agreement with the ones established on the basis of pathogenicity features. However,
most of the strains of pathovars of X. campestris have heterogeneous phenotypic
characteristics and only have the pathogenicity to the same host plant as the common
feature (Burkholder and Starr, 1948; Dye, 1962; Van den Mooter and Swings, 1990).
This can be illustrated by the case of X campestris pv. dieffenbachiae. Chase et al.
(1992) analyzed the phenotypic features of 149 strains of X. campestris pv.
dieffenbachiae obtained from aroid plants and found large variation in their reactions.
Minimum pH for growth in nutrient broth was the only useful feature to differentiate
typical strains obtained from Syngonium spp. from all other aroid strains. Although
some trends were observed, physiological tests, such as hydrolysis of both pectin and
starch, and growth on different media, failed to differentiate the strains to their host of
origin. The strains of X campestris pv. dieffenbachiae from different hosts overlapped
in their reactions for these tests, which made it impossible to relate the phenotypic
features to the host of origin.
Commercially available systems for identification of bacteria on the basis of
phenotypic characteristics have also been examined for differentiation of plant
pathogenic xanthomonads (Chase et al., 1992; Jones et al., 1993a; Vauterin et al.,
1990b; Vemiere et al., 1993). The Biolog GN MicroPlate System, which is based on
differential utilization of 95 carbon sources with consequent changes in redox potential
of the bacterial suspension (Bochner, 1989), was evaluated for accuracy in identifying
some plant pathogenic bacteria, including strains of Xanthomonas (Jones et al., 1993a).

20
The identification of strains of Xanthomonas to species and to pathovar was correct in
97% and 6% of the cases, respectively, when the original data base was used.
However, the accuracy of the Biolog system for identification of pathovars of X.
campestris was substantially improved when the database was amended by the addition
of data obtained for different strains of X campestris. Nevertheless, the Biolog results
confirmed the close relatedness between the various pathovars determined in previous
work, and substantiated the difficulty of separation of pathovars on the basis of
phenotypic features alone.
The metabolic variation of strains of X campestris that cause diseases on citrus
was investigated by using the Biolog system (Hartung and Civerolo, 1991; Vemiere et
al., 1993). The Biolog metabolic profile was useful for differentiating strains of the
citrus canker pathogen X campestris pv. citri from strains of the citrus bacterial spot
agent X. campestris pv. citrumelo. Furthermore, the different citrus canker groups of X.
campestris pv. citri could be differentiated based on the utilization of three carbon
sources (Vemiere et al., 1993). Although the correct identification of theJf campestris
pv. citri strains was very poor by using the original data base of the system, the
identification was significantly improved when the database was supplemented with
additional data (Vemiere et al., 1993). The citrus bacterial spot pathogen X. campestris
pv. citrumelo was more variable in the metabolic profile than X. campestris pv. citri
(Hartung and Civerolo, 1991). All the profiles examined placed the strains within the
species X campestris. However, some strains were identified as X. campestris pv.
dieffenbachiae (Hartung and Civerolo, 1991). The relatedness of strains ofX.
campestris pv. citrumelo to other pathovars of X. campestris has also been determined
on the basis of DNA-DNA hybridization (Egel et al., 1991), RFLP analysis (Gabriel et
al., 1988, 1989), and fatty acid composition (Graham et al., 1990).

21
The enzymatic API ZYM galleries (API Systems S.A., La-Balme-les-Grottes,
Montalieu-Vercieu, France), has also been tested for differentiation and identification
of plant pathogenic xanthomonads (Vauterin et al., 1990b). The enzymatic profiles of
X. campestris pv. begoniae and X. campestris pv. pelargonii were investigated with the
API ZYM system. The profiles were highly uniform within each pathovar, and they
could be differentiated by three enzymatic activities, chymotrypsin, a-D-galactosidase,
and a-D-glucosidase. However, the API ZYM system did not seem to be
discriminative at the pathovar level when larger numbers of pathovars were included in
the study (Vauterin et al., 1993).
Nucleic Acids Analysis
A number of techniques have been used for characterization of nucleic acids of
the xanthomonads. DNA-rRNA hybridization has been of major importance in
differentiation of the xanthomonads at the intrageneric level (De Vos and De Ley,
1983; Palleroni et al., 1973), whereas DNA-DNA hybridization has become the basis
for determining genetic relationships among bacterial strains and for establishing
classification schemes at the subgeneric level (Holt et al., 1994; Krieg and Holt, 1984;
Wayne et al., 1987). Furthermore, DNA-DNA reassociation is considered to be the
reference standard to determine the phylogeny, and consequently the taxonomy of
bacteria (Wayne et al., 1987). RFLP analysis of genomic DNA has been useful to
distinguish several groups of xanthomonads, to study genetic diversity, and to establish
phylogenetic relationships among different pathovars of X. campestris (Gabriel et al.,
1988, 1989; Gottwald et al., 1991; Graham et al., 1990; Hartung and Civerolo, 1989,
1991; Lazo and Gabriel, 1987; Lazo et al., 1987; Leach et al., 1990, 1992; Qhobela and
Claflin, 1992; Qhobela et al., 1991; Verdier et al., 1993).

22
Nucleic acid based techniques have also been applied to the identification and
detection of the plant pathogenic xanthomonads. Genomic and plasmid DNA
fingerprinting has been used in several instances for the differentiation of a number of
groups of xanthomonads (Berthier et al., 1993; Cooksey and Graham, 1989; Hartung
and Civerolo, 1987; Pruvost et al., 1992). In the same way, genomic and plasmid DNA
probes were also developed and examined for the purpose of identification and
detection of plant pathogenic xanthomonads (Garde and Bender, 1991; Gilbertson et
al., 1989; Hartung, 1992; Lazo and Gabriel, 1987; Leach et al., 1990, 1992).
The genomic relationship among plant pathogenic xanthomonads determined
by DNA-DNA hybridization has revealed a group of bacteria with diverse genetic
background, with DNA homologies between pathovars o X. campestris ranging from
values as low as 15% to 39% (Vauterin et al., 1990b). These results strongly reinforce
the contention that strains of X. campestris do not constitute a single species.
Furthermore, the genetic relationship established on the basis of DNA-DNA homology
does not always correlate with the pathogenicity of the bacteria toward a given range of
host plants. The genetic relationship of strains of plant pathogenic xanthomonads
representing 44 X. campestris pathovars and Xanthomonas spp. was examined by SI
DNA-DNA hybridization (Hildebrand et al., 1990; Palleroni et al., 1993). Several
groups of xanthomonads were defined on the basis of DNA-DNA homology. The
largest group consisted of strains belonging to 25 pathovars of X. campestris
(Hildebrand et al., 1990; Palleroni et al., 1993). This group corresponds to DNA
homology group VIII established by Vauterin et al. (1993), which also includes 22
additional pathovars of V. campestris and X axonopodis. Another cluster comprises a
group ofX. campestris pv. carotae, X. campestris pv. gardneri, X campestris pv.
pelargonii, and A", campestris pv. taraxaci (Hildebrand et al., 1990; Palleroni et al.,

23
1993). Two minor clusters were also established. One group comprises the pathovars
X. campestris pv. campestris, X. campestris pv. plantaginis, and X campestris pv.
raphani, and the other group comprises X. campestris pv. celebensis and X. campestris
pv. juglandis (Palleroni et al., 1993).
Although the genetic relatedness among plant pathogenic xanthomonads which
cause disease in different plants does not support a correlation between genomic
groupings and pathogenicity features, there are some exceptions. The pathovars ofX.
campestris which cause diseases on closely related leguminous plants, X. campestris
pv. glycines, X campestris pv. lespedezae, X. campestris pv. phaseoli, X. campestris
pv. phaseoli "fuscans", and X campestris pv. vignicola, cluster in DNA homology
group VIII of Vauterin et al. (1993); X. campestris pv. pisi was the only leguminous
pathogen not included in this group (Hildebrand et al., 1990; Palleroni et al., 1993;
Vauterin et al., 1993). However, the leguminous pathogens are also genetically highly
related to strains of several pathovars of X. campestris that cause diseases on different
plants (Hildebrand et al., 1990; Palleroni et al., 1993; Vauterin et al., 1993). The
strains ofX. campestris pv. arrhenatheri, X. campestris pv. cerealis, X. campestris pv.
graminis, X. campestris pv. hordei, X. campestris pv. phlei, X campestris pv.
phleipratensis, X campestris pv. poae, X. campestris pv. secalis, and X. campestris pv.
translucens, that cause diseases on gramineous hosts cluster together in the DNA
homology group IX of Vauterin et al. (1993). This grouping also correlates strongly
with groupings based on SDS-PAGE and fatty acid analysis (Vauterin et al., 1992).
The DNA homology group XII contains the six pathovars from crucifers, X campestris
pv. aberrans, X campestris pv. armoraciae, X. campestris pv. barbareae, X campestris
pv. campestris, X. campestris pv. incanae, and X campestris pv. raphani (Vauterin et
al., 1993).

24
The most striking feature of the plant pathogenic xanthomonads is the genetic
diversity among strains of the same pathovar or different pathovars of X. campestris
that cause diseases on highly related hosts. The diversity strongly illustrates the
divergence between pathogenicity characteristics and genetic backgrounds. The
pathovars X. campestris pv. gardneri, X. campestris pv. physalidicola, and X
campestris pv. vesicatoria, which cause diseases on solanaceous plants, are genetically
diverse on the basis of DNA-DNA homology (Hildebrand et al., 1990; Palleroni et ah,
1993; Stall et ah, 1994; Vauterin et ah, 1993). X. campestris pv. vesicatoria is
composed of the two genetically unrelated groups A and B (Stall et ah, 1994). X.
campestris pv. vesicatoria group A belongs to homology group VIII of Vauterin et ah
(1993), as well as X. campestris pv. physalidicola. Further, X. campestris pv.
vesicatoria group A and X. campestris pv. physalidicola are also genetically closely
related to several other pathovars of X. campestris that cause diseases on different
plants (Hildebrand et ah, 1990; Palleroni et ah, 1993; Vauterin et ah, 1993). Strains of
X campestris pv. vesicatoria group A are only about 33% related to strains ofX.
campestris pv. vesicatoria group B, on the basis of DNA-DNA homology (Stall et ah,
1994). Strains of X. campestris pv. vesicatoria group B for instance is distinct from
other xanthomonads and they are the only members of the DNA homology group XVI
(Vauterin et ah, 1993). The grouping of X. campestris pv. vesicatoria are also well
supported by fatty acid analysis, serology, and phenotypic features (Stall et ah, 1994).
X. campestris pv. gardneri is also not related genetically to the other solanaceous
pathogens. Instead, X. campestris pv. gardneri constitutes a group which contains the
pathovars X. campestris pv. carotae, X. campestris pv. pelargonii, and X. campestris pv.
taraxaci (Hildebrand et ah, 1990; Palleroni et ah, 1993).

25
The xanthomonads that cause diseases on citrus also comprise a genetically
diverse group of bacteria. Strains of the citrus pathogens X. campes tris pv. citri and X.
campestris pv. citrumelo which belong to the DNA homology group VIII of Vauterin
et al. (1993) are only about 60% similar to one another based on DNA-DNA homology
(Egel et al., 1991; Vauterin et al., 1991b). X campestris pv. citri canker A strains are,
however, highly related to other pathovars of X. campestris. The canker A strains of
this bacterium are related at over 90% by DNA-DNA homology to X. campestris pv.
glycines and X. campestris pv. malvacearum (Egel et al., 1991; Vauterin et al., 1993).
On the other hand, strains of the citrus bacterial pathogen, X. campestris pv. citrumelo,
are closely related to strains of other pathovars of X campestris which includes X
campestris pv. alfalfae and A! campestris pv. fici (Egel et al., 1991; Vauterin et al.,
1993). A strong relationship exists between the citrus pathogens based on
corroborating data obtained from DNA-DNA homology studies and analyses of fatty
acid composition (Graham et al., 1990; Vauterin et al., 1991b), SDS-PAGE of proteins
(Vauterin et al., 1991b), and RFLP of genomic DNA (Gabriel et al., 1988, 1989;
Graham et al., 1990; Vauterin et al., 1991b).
Different techniques of genomic fingerprinting have been used to explore the
genetic diversity of the plant pathogenic xanthomonads for identification and
comparison purposes. Genomic fingerprintings produced by using frequent-cutting
endonucleases were examined for establishing the genetic relationships of the
xanthomonads that cause diseases on citrus (Hartung and Civerolo, 1987; Pruvost et al.,
1992). The citrus canker A and B groups of X. campestris pv. citri each produced very
conserved genomic fingerprintings which distinguished strains of both groups (Hartung
and Civerolo, 1987). Although strains of the citrus bacterial spot pathogen^
campestris pv. citrumelo could be distinguished from both canker groups of X.

26
campestris pv. citri, a wide diversity of genomic fingerprints (Hartung and Civerolo,
1987) was found. The genetic diversity ofX. campestris pv. citrumelo was also
determined by analysis of restriction patterns derived from pulsed-field gel
electrophoresis of genomic DNA fragments generated by rare-cutting endonucleases
(Egel et al., 1991). In contrast to the diversity of the strains of2f campestris pv.
citrumelo, strains of X. campestris pv. citri canker A and B produced characteristic
restriction patterns by pulsed-field gel electrophoresis (Egel et al., 1991).
Attempts have also been made to differentiate X. campestris pathovars by RFLP
analysis of plasmid and genomic DNAs based on hybridization with different DNA
probes. In most of the cases, random cloned fragments of the bacterial genome were
used as DNA probes in RFLP analysis of the xanthomonads. Specific regions of the
genome have been selected for DNA probes, such as rRNA sequences (Berthier et al.,
1993; DeParasis and Roth, 1990), repetitive DNA sequences (Leach et al., 1990, 1992),
copper resistance genes (Garde and Bender, 1991), and plasmid DNA sequences
(Hartung, 1992; Gilbertson et al., 1989; Lazo and Gabriel, 1987). Lazo et al. (1987)
used a cloned DNA fragment derived from X campestris pv. citrumelo for RFLP
analysis to differentiate pathovars of X. campestris. The RFLP analysis involving
strains representing 26 pathovars of X. campestris revealed profiles highly conserved
and characteristic for each pathovar tested. By using more than one DNA probe, or by
digesting the genomic DNA with different restriction endonucleases, it was possible to
differentiate all the pathovars ofX. campestris included in the study (Lazo et al., 1987).
An rRNA probe was also used to distinguish pathovars of X. campestris (Berthier et al.,
1993). Further, the RFLP patterns established for the pathovars X. campestris pv.
begoniae, X campestris pv. dieffenbachiae, X campestris pv. malvacearum, and X.

27
campestris pv. manihotis correlated with pathogenicity on the host plant (Berthier et
al., 1993).
Extensive RELP analyses were conducted in some pathovars ofX. campestris,
such &sX. campestris pv. citri and X. campestris pv. citrumelo (Gabriel et al., 1988,
1989; Gottwald et al., 1991; Graham et al., 1990; Hartung, 1992; Hartung and
Civerolo, 1989, 1991; Lazo and Gabriel, 1987; Lazo et al., 1987),^ campestris pv.
manihotis (Verdier et al., 1993), X. campestris pv. vasculorum (Qhobela and Claflin,
1992), X. campestris pv. pennamericanum (Qhobela and Claflin, 1988), and X oryzae
pv. oryzae (Leach et al., 1990, 1992). The most comprehensive RFLP analyses were
carried out to determine the genetic diversity of the xanthomonads causing diseases of
citrus (Gabriel et al., 1988, 1989; Gottwald et al., 1991; Graham et al., 1990; Hartung,
1992; Hartung and Civerolo, 1989, 1991; Lazo and Gabriel, 1987; Lazo et al., 1987).
The RFLP analyses characterized the strains of the citrus bacterial spot pathogen, X.
campestris pv. citrumelo, as distinct from all forms of the citrus canker pathogen
(Gabriel et al., 1988, 1989; Hartung and Civerolo, 1989). In contrast to the
characteristic restriction patterns of the X campestris pv. citri canker groups A and B,
X. campestris pv. citrumelo gave a wide range of variation in the RFLP analyses
(Gabriel et al., 1988, 1989; Graham et al., 1990; Hartung and Civerolo, 1989, 1991).
The RFLP analyses further supported that strains of the citrus bacterial spot pathogen
are not genetically closely related to any recognized group of X. campestris pv. citri
(Hartung and Civerolo, 1987, 1989; Gabriel et al., 1988, 1989). Instead, the analysis
has revealed a significant genetic relationship between strains of X. campestris pv.
citrumelo and strains of non-citrus pathogens, such as X. campestris pv. alfalfae and X
campestris pv. fici (Gabriel et al., 1988,1989; Graham et al., 1990). The relatedness of
strains of X. campestris pv. citrumelo to strains of the other pathovars oiX. campestris

28
was further supported on the basis of DNA homology and phenotypic characteristics
(Egel et al., 1991; Graham et ah, 1990).
Nucleic acid based techniques have also been examined for the specific
detection of plant pathogenic xanthomonads (Garde and Bender, 1991; Gilbertson et
ah, 1989; Hartung, 1992; Hartung et ah, 1993). Garde and Bender (1991) developed
DNA probes for detection of copper resistance in X. campestris pv. vesicatoria.
Although the DNA probes seem specific to the genes conferring resistance to copper in
X. campestris pv. vesicatoria and apparently did not hybridize to DNA of copper
resistant P. syringae pv. tomato (Garde and Bender, 1991), the cop operon of P.
syringae pv. tomato not only hybridized to copper sensitive bacteria (Cooksey et ah,
1990), but also to strains of ^ campestris pv. vesicatoria (Voloudakis et ah, 1993). In
another study, Gilbertson et ah (1989) used DNA probes obtained from a cryptic
plasmid for specific detection of the bean pathogens X. campestris pv. phaseoli and X
campestris pv. phaseoli "fiiscans". Although the DNA probes did not hybridize to
nonpathogenic xanthomonads isolated from bean plants, they did hybridize to total
genomic DNA of bacterial strains of other pathovars ofX. campestris (Gilbertson et ah,
1989). However, strains of these other pathovars of A! campestris were thought
unlikely to be associated with bean plants. Plasmid based DNA probes have also been
used for detection of the citrus canker pathogen, X. campestris pv. citri (Hartung,
1992). The DNA probes were developed for specific detection of the citrus canker
pathogen, X. campestris pv. citri which include four different groups. The DNA probes
were highly specific for the citrus canker pathogens and did not hybridize to strains of
the citrus bacterial spot pathogen, X. campestris pv. citrumelo. However, cross
hybridization was observed in strains of X campestris pv. bilvae and X campestris pv.
vignicola (Hartung, 1992). Furthermore, a DNA amplification-based procedure was

29
developed for specific and sensitive detection of the citrus canker pathogen based on
the DNA sequence of these plasmid probes (Hartung et al., 1993).

CHAPTER 3
OLIGONUCLEOTIDE PRIMERS FOR DETECTION AND
IDENTIFICATION OF PLANT PATHOGENIC STRAINS OF
Xanthomonas BY AMPLIFICATION OF DNA SEQUENCES
RELATED TO THE hrp GENES OF Xanthomonas campestris PV.
VESICATORIA
The genus Xanthomonas Dowson 1939 includes Gram-negative, usually
yellow-pigmented bacteria that occur worldwide and cause plant diseases. Over 124
monocotyledonous and 268 dicotyledonous plant species are hosts of Xanthomonas
(Leyns et al., 1984). Among the species o Xanthomonas, X. campestris comprises at
least 125 different pathovars that are distinguished by the diseases they cause
(Bradbury, 1984). The genus Xanthomonas also includes strains which may be
associated with plant material, but are not pathogenic to the plants from which they
were isolated (Angeles-Ramos et al., 1991; Gitaitis et al., 1987; Liao and Wells, 1987;
Maas et al., 1985). These opportunistic bacteria can be identified as xanthomonads by
the presence of xanthomonadins and by unique fatty acid profiles. Although the
identification of bacteria in the genus Xanthomonas presents no great problem, sub
generic identification of xanthomonads is still difficult.
Traditional methods for the detection and identification of plant pathogenic
xanthomonads rely on isolating the organism of interest in pure culture and performing
predetermined biochemical, serological, and pathological tests (Saettler et al., 1989;
Schaad, 1988). Sometimes, non-selective or selective enrichments are required to
increase the sensitivity of the isolation, which may be complicated by the presence of
fast-growing contaminating bacteria associated with plant tissue (Saettler et al., 1989).
30

31
More recently, methods based on metabolic and protein profiling (Chase et al., 1992;
Van den Mooter and Swings, 1990; Vauterin et al., 199lab), and fatty acid analysis
(Chase et al., 1992; Yang et al., 1993) have been used for identification, but isolation
and purification of the bacterial strain is still required. Polyclonal or monoclonal
antisera produced against strains of X. campestris have been used for detection and
identification, but they have provided variable results. Polyclonal antisera may cross-
react with other bacteria and may be unable to differentiate specific strains or pathovars
ofX. campestris (Alvarez and Lou, 1982). Several monoclonal antisera were produced
which reacted specifically with all strains of some pathovars ofX. campestris that
infect relatively few genera of hosts (Alvarez et al., 1991; Benedict et al., 1990).
However, for certain X. campestris pathovars, mainly those that infect several hosts
from different genera, no pathovar-specific monoclonal antisera that react with all
strains of the respective pathovar have been found (Alvarez et al., 1991; Jones et al.,
1993b).
Nucleic acid-based techniques have also been applied for detection and
identification of plant pathogenic bacteria (Bereswill et al., 1992; Manulis et al., 1991;
Schaad et al., 1989; Seal et al., 1992), including some members of the xanthomonads
(Garde and Bender, 1991; Gilbertson et al., 1989; Hartung, 1992; Hartung et al., 1993;
Lazo and Gabriel, 1987; Lazo et al., 1987). The techniques developed for detection
and identification of xanthomonads were based on random probes (Lazo et al., 1987),
or on plasmid DNA fragments specific for a few pathovars of X campestris (Gilbertson
et al., 1989; Hartung, 1992; Hartung et al., 1993; Lazo and Gabriel, 1987), or even for a
group of strains (Garde and Bender, 1991). Highly conserved regions in the bacterial
genome of plant pathogenic bacteria could be more useful for the selection of specific

32
DNA probes for detection and identification of a larger number of strains, pathovars, or
species of Xanthomonas.
The hrp gene clusters that determine hypersensitivity and pathogenicity may be
appropriate for selection of probes for detection and identification of plant pathogenic
bacteria. The hrp gene cluster is required by bacterial plant pathogens to produce
symptoms on susceptible hosts and a hypersensitive reaction (HR) on resistant hosts, or
on nonhosts (Willis et al., 1991), and has been found in several plant pathogenic
bacteria, such as Erwinia amylovora (Beer et al., 1991), Pseudomonas solanacearum
(Boucher et al., 1987), P. syringae pv. phaseolicola (Lindgren et al., 1986), and A
campestris pv. vesicatoria (Bonas et al., 1991). Furthermore, hrp functions seem to be
highly conserved among a number of plant pathogenic bacteria (Ulla Bonas, personal
communication; Fenselau et al., 1992; Gough et al., 1992; Hwang et al., 1992). The
hrp genes of plant pathogenic bacteria are also very similar at the protein level to genes
that are involved in the secretion of pathogenicity factors by bacterial pathogens of
mammals (Fenselau et al., 1992; Gough et al., 1992). By contrast, nonpathogenic
bacteria are unable to produce symptoms on susceptible hosts and HR on nonhosts,
apparently because they do not possess DNA sequences similar to hrp genes (Lindgren
et al., 1986; Stall and Minsavage, 1990). Physically and functionally similar hrp
sequences occur among several pathovars of X. campestris, but not in opportunistic
xanthomonads (Bonas et al., 1991; Stall and Minsavage, 1990).
The objective of this study was to examine sequences of the hrp genes of A
campestris pv. vesicatoria for detection and identification of plant pathogenic
xanthomonads. Oligonucleotide primers specific for hrp genes were tested for their
suitability for identification of these xanthomonads by the polymerase chain reaction

33
(PCR). Furthermore, the reliability of identification of the xanthomonads to subgeneric
classification was assessed by restriction analysis of amplified DNA fragments.
Materials and Methods
Bacterial strains, plasmids, and culture conditions
The bacterial strains and plasmids used in this study and their sources are listed
in Appendix A. The identity of bacterial strains used here was confirmed by fatty acid
analysis (Nancy C. Hodge, personal communication). All strains of Acidovorax
avenae, Agrobacterium tumefaciens, Erwinia spp., Clavibacter michiganense,
Pseudomonas spp., X. campes tris, and X. maltophilia were grown on nutrient agar
(Becton Dickinson, Cockeysville, MD). Nutrient broth cultures were grown 24 hours
on a rotatory shaker (150 rpm) at 28C. Strains of A albilineans andX. fragariae were
cultivated on Wilbrink's medium (Koike, 1965). Strains of Xylella fastidiosa were
grown on PW medium (Davis et al., 1981). Strains of Escherichia coli were cultivated
on Luria-Bertani medium at 37C (Miller, 1972). All strains were stored in sterile tap
water at room temperature or in 30% glycerol at -70C, or both. Antibiotics were used
to maintain selection for resistance markers at the following final concentrations:
ampicillin, 100 pg/ml; tetracycline, 10 pg/ml; rifampicin, 100 pg/ml; and
spectinomycin, 50 pg/ml.
Plant material and plant inoculations
All plants were maintained in a growth chamber at 28-30C during inoculation
and incubation. The pepper cultivar Early Calwonder (ECW) and the near-isogenic
lines ECW-1 OR, ECW-20R, and ECW-30R have been described elsewhere (Minsavage

34
et al., 1990). These lines provided a susceptible reaction or a hypersensitive reaction,
depending on the strain used.
Fully expanded leaves of plants were inoculated with bacterial suspensions by
infiltrating the bacteria into the intercellular spaces with a 1 ml plastic syringe with a
27 gauge needle. The concentration of the inoculum was approximately 5 X 108
colony forming units (CFU) per milliliter in sterile tap water, determined by measuring
the optical density in a Spectronic 20 spectrophotometer (Bausch and Lomb, Inc.,
Rochester, NY). Plant reactions were scored over a period of several days.
DNA manipulations
Total genomic DNA was isolated by phenol extraction and ethanol precipitation
essentially as described by Ausubel et al. (1987). Plasmid mini-prep, preparation of
competent cells, ligation, and transformation of E. coli cells were performed by
standard procedures (Ausubel et al., 1987; Sambrook et al., 1989). For
complementation analysis, the helper plasmid pRK2013 was used in triparental mating
to mobilize pLAFR3 clones from E. coli into Xanthomonas cells (Ditta et al., 1980;
Figurski and Helinski, 1979).
Hybridization analysis
Total genomic DNA and amplified DNA fragments were electrophoresed in
0.7% agarose according to standard procedures (Sambrook et al., 1989). The DNA was
then denatured in 0.4 N NaOH and 0.6 M NaCl for 30 min, neutralized for 30 min in
0.5 M Tris-Cl and 1.5 M NaCl, pH 7.5, and transferred by the procedure of Southern
(1975) to nylon membrane (Schleicher & Schuell, Keene, NH). Southern hybridization
and detection of the hybridized DNA was carried out by using the Genius

35
Nonradioactive DNA labeling and Detection kit (Boehringer Mannheim, Indianapolis,
IN) as specified by the manufacturer. Clones containing the desired insert of the hrp
gene cluster of X. campestris pv. vesicatoria, or in vitro amplified hrp fragments used
as probes, were labeled by the random primer (Feinberg and Vogelstein, 1983)
incorporation of digoxigenin-labeled deoxyuridine-triphosphate (DIG-UTP). Before
use, the probes were denatured by boiling for 10 min followed by chilling in an ice
ethanol slurry. Hybridization was carried out at 68C with 0.5X SSC and 0.1%
(wt/vol) SDS. The membranes were prewashed twice at room temperature for 5 min in
IX SSC containing 0.1% (wt/vol) SDS. Two final washes were completed at 65 C for
15 min in 0.5X SSC containing 0.1% (wt/vol) SDS.
DNA amplification
Three sets of oligonucleotide primers were selected from the nucleotide
sequence of the hrp region of X campestris pv. vesicatoria (Ulla Bonas, personal
communication). Primers RST2 (5'AGGCCCTGGAAGGTGCCCTGGA3') and RST3
(5'ATCGCACTGCGTACCGCGCGCGA3') delineated a 840-bp fragment, RST9
(5'GGCACTATGCAATGACTG3') and RST10 (5'AATACGCTGGAACTGCTG3')
delineated a 355-bp fragment, and RST21
(5'GCACGCTCCAGATCAGCATCGAGG3') and RST22
(5'GGCATCTGCATGCGTGCTCTCCGA3') delineated a 1,075-bp fragment. The
primers map to the complementation groups hrpB, hrpC, and hrpD of X campestris pv.
vesicatoria (Fig. 3-1). Furthermore, the sequences of the oligonucleotide primers RST3
and RST9 originate from the hrpB6, a gene for a putative ATPase that seems to be
highly conserved among different bacteria at the protein sequence level (Fenselau et al.,
1992). Oligonucleotide primers were synthesized with a model 394 DNA Synthesizer

Fig. 3-1. Structural organization of the hrp region in Xanthomonas campestris pv. vesicatoria. (A) FcoRI fragments of the
hrp region. (B) Position and orientation of the hrp loci, designated hrpA to hrpF (Bonas et al., 1991; Schulte and Bonas,
1992). (C) Position and orientation of the open reading frames (ORFs). The sizes of the loci are based on a combination of
genetic and sequence analysis from which possible open reading frames (ORF) are predicted (Ulla Bonas, personal
communication). Only the open reading frames relevant for this study, hrpB5 to hrpB8, hrpC3, and hrpD, are shown here.
For each RST oligonucleotide primer used for DNA amplification, its position in the DNA sequence is indicated by an
asterisk (*).

primers
RST RST
9 10
RST
3
RST
2
8.0 kb
2.7 kb
RST
21
RST
22
UJ
"J

38
(Applied Biosystems, Foster City, CA) by the DNA Synthesis Laboratory, University
of Florida, Gainesville.
DNA was amplified in a total volume of 50 pi. The reaction mixture contained
5 pi of 10X buffer (500 mM KC1,100 mM Tris Cl [pH 9.0 at 25C], 1% Triton X-
100), 1.5 mM MgCl2, 200 pM of each deoxynucleotide triphosphate (Boehringer
Mannheim), 25 pmol of each primer, and 1.25 units of Taq polymerase (Promega,
Madison, WI). The amount of template DNA added was 100 ng of purified total
bacterial DNA or 25 ng of a plasmid preparation, unless otherwise stated. The reaction
mixture was covered with 50 pi of light mineral oil. A total of 30 amplification cycles
were performed in an automated thermocycler PT-100-60 (MJ Research, Watertown,
MA). Each cycle consisted of 30 s of denaturation at 95C, 30 s of annealing at 62C,
and 45 s of extension at 72C for the primers RST2 and RST3; 30 s at 95C, 30 s at
52C and 45 s at 72C for the primers RST9 and RST10; and 30 s at 95C, 40 s at 61C
and 45 s at 72C, for the primers RST21 and RST22. The last extension step was
extended to 5 min. Amplified DNAs were detected by electrophoresis in 0.9% agarose
gels in TAE buffer (40 mM Tris acetate, 1 mM EDTA, pH 8.2) at 5 V/cm of gel
(Sambrook et al., 1989). The gel was stained with 0.5 pg of ethidium bromide per ml
and then photographed over a UV transilluminator (Fotodyne Inc., New Berlin, WI)
with type 55 Polaroid film (Polaroid, Cambridge, MA).
Restriction endonuclease analysis
The DNA fragments amplified from different bacterial strains were restricted
with the frequent-cutting endonucleases Cfol, HaeIII, Sau3Al, or Taql, according to
conditions specified by the manufacturer (Promega). The restricted fragments were
separated by electrophoresis in 4% agarose gels (3% NuSieve and 1% Seakem GTG

39
[FMC BioProducts, Rockland, ME]) in TAE buffer at 8 V/cm. Phage X Pstl restricted
DNA fragments were used as molecular standards. The gel was stained with 0.5 pg of
ethidium bromide per ml for 40 min and then destained in 1 mM MgSC>4 for 1 hr and
photographed over a UV transilluminator with type 55 Polaroid film.
Results
Specificity of the oligonucleotide primers to the hrp gene cluster
Three pairs of oligonucleotide primers were tested for amplification of
fragments from genomic DNA of strain 75-3 of X. campestris pv. vesicatoria and from
plasmids that contain cloned parts of the hrp region from strain 75-3. Plasmid pXV9
harbors a fragment of approximately 27 kb of strain 75-3 of X. campestris pv.
vesicatoria containing almost the entire hrp region (Bonas et al., 1991). Plasmid
pXV5.5 harbors a 5.5 kb EcoRI fragment containing part of the hrp complementation
groups hrpA and hrpB (Fig. 3-1). The 5.1 kb EcoRI insert of plasmid pXV5.1 maps to
the complementation groups hrpC and hrpD (Fig. 3-1). We amplified fragments of the
expected 355, 840, and 1,075 bp in length from total genomic DNA of strain 75-3 of A!
campestris pv. vesicatoria by using primers RST9 and RST10, RST2 and RST3, and
RST21 and RST22, respectively (Fig. 3-2). The 355- and 840-bp fragments were also
amplified from plasmids pXV9 and pXV5.5 (Fig. 3-2), whereas the 1,075-bp fragment
was amplified from plasmids pXV9 and pXV5.1 (Fig. 3-2), thus confirming the sizes
and locations of fragments predicted by the DNA sequence analysis (Ulla Bonas,
personal communication). The 355- and 840-bp sequences correspond to genes in the
hrpB operon, whereas the 1075 bp fragment occurs in the hrpC/hrpD region (Fig. 3-1).
Except for hrpB6, which encoded a putative ATPase (Fenselau et al., 1992), the

40
M 1 23 456 78 9 10 11 12
Fig. 3-2. Amplification of fragments of the hrp gene cluster from Xanthomonas
campestris pv. vesicatoria 75-3. The following DNAs were used: 75-3 (lanes 1, 5, and
9), plasmid pXV9 (lanes 2, 6, and 10), plasmid pXV5.5 (lanes 3, 7, and 11), and
plasmid pXV5.1 (lanes 4, 8, and 12). Lanes: M, phage X restricted with EcoRI and
Hindlll; 1 to 4, amplification with primers RST9 and RST10; 5 to 8, amplification with
primers RST2 and RST3; 9 to 12, amplification with primers RST21 and RST22.
Molecular sizes are given in base pairs.

41
functions of these genes are unknown. The hrpDl gene is highly similar to a sequence
of pathogenicity genes in A campestris pv. glycines (Ulla Bonas, personal
communication; Hwang et al., 1992).
The identity of the amplified fragments was further confirmed by restriction
enzyme analysis. The 355-, 840-, and 1,075-bp fragments amplified from both total
DNA of strain 75-3 of X. campestris pv. vesicatoria and from DNA of pXV9, pXV5.1,
and pXV5.5 were digested with Cfol, Haelll, Sow3AI, and Taq\. The banding patterns
were identical for each of the three sets of fragments amplified from strain 75-3 and
plasmids containing cloned parts of the hrp region (Fig. 3-3) and matched the
restriction map generated from the DNA sequence of the hrp gene cluster of X.
campestris pv. vesicatoria 75-3 (Ulla Bonas, personal communication).
Similarity of the hrp fragments amplified from X campestris pv. vesicatoria to DNA of
other bacteria
Total genomic DNA of strains of different pathovars of X. campestris and of
Xanthomonas spp. was digested with EcoRI, separated in agarose gels, blotted, and
probed with each of the three fragments amplified from DNA of X. campestris pv.
vesicatoria strain 75-3. The hybridization signals in the genomic DNA of A
campestris pv. vesicatoria 75-3 corresponded to the predicted 7.4 kb BamHl and 5.5 kb
EcoRI fragments (355- and 840-bp probes), and to the 6.0 kb BamiRl and 5.1 kb EcoRI
fragments (1,075-bp probe) (Fig. 3-4). Homology to these three hrp fragments oiX.
campestris pv. vesicatoria strain 75-3 was detected in strains of X fragariae and of 28
different pathovars of A! campestris (Fig. 3-4; Table 3-1). However, polymorphisms
were observed in the DNA from these different pathovars (Fig. 3-4). Total genomic
DNA of A albilineans, A campestris pv. secalis, and A campestris pv. translucens,
which are pathogens of monocotyledonous plants, as well as DNA of the non plant

42
Fig. 3-3. Taql restriction endonuclease analysis of fragments of the hrp gene cluster
amplified from total genomic DNA from Xanthomonas campestris pv. vesicatoria 75-3
and from plasmids containing the hrp region. Lanes: M, phage X. restricted with Pstl; 1
to 3, the 355-bp fragments from X. campestris pv. vesicatoria 75-3 and plasmids pXV9
and pXV5.5, respectively; lanes 4 to 6, the 840-bp fragments froml campestris pv.
vesicatoria 75-3 and plasmids pXV9 and pXV5.5, respectively; lanes 7 to 9, the 1,075-
bp fragments froml campestris pv. vesicatoria 75-3 and plasmids pXV9 and pXV5.1,
respectively. Molecular sizes are given in base pairs.

B
C
Fig. 3-4. Hybridization of the hrp fragments amplified from Xanthomonas campestris
pv. vesicatoria 75-3 to total genomic DNA of strains o Xanthomonas campestris.
Approximately 3 pg of £coRI-digested DNA was loaded per lane. The blots were
probed with the labeled 355-bp (A), 840-bp (B), and 1,075-bp (C) fragments amplified
froml campestris pv. vesicatoria 75-3. Lanes: l,X campestris pv. vesicatoria 75-3,
BamHl digested; 2, X campestris pv. vesicatoria 75-3; 3, X. campestris pv. vesicatoria
XV56; 4, X campestris pv. alfalfae KS; 5, X. campestris pv. begoniae XCB9; 6, X.
campestris pv. campestris 33913; 1,X. campestris pv. glycines 87-2; 8, X. campestris
pv. holcicola G-23; 9, X. campestris pv. malvacearum RIATC; 10, X. campestris pv.
phaseoli 85-6; \ \,X. campestris pv. pruni FLA1; \2,X. campestris pv. secalis
XC129C; 13, X. campestris pv. translucens 80-1. Molecular sizes are given in kilobase
pairs.

44
Table 3-1. Hybridization of the hrp fragments amplified from Xanthomonas campestris
pv. vesicatoria 75-3 to total genomic DNA and amplification of Arp-related fragments
from different bacterial strains.
Strain
Southern hybridization
DNA amplification
355 bp
840 bp
1,075 bp
355 bp
840 bp 1,075 bp
Xanthomonas
X. campestris
pv. alfalfae KS
+
+a
+
(+)
+
+
pv. armoraciae 63-27
+
+
+
-
+
+
pv. bilvae XCB
+
+
+
-
+
+
pv. begoniae XCB9
+
+
+
-
+
+
pv. campestris 33913
+
+
4-
-
+
+
pv. carotae #13
+
+
+
-
(+)
+
pv. citri 9771
+
+
+
-
+
+
pv. citrumelo FI
+
+
+
(+)
+
+
pv. dieffenbachiae 729
+
+
+
-
+
+
pv. fici X151
4-
+
4-
4-
+
+
pv. gardneri XG101
+
+
+
-
(+)
+
pv. glycines 87-2
+
+
+
-
+
+
pv. holcicola G-23
+
+
+
-
-
+
pv. incanae 9561-1
+
+
+
-
+
+
pv. maculifoliigardeniae X22j
+
+
4-
(+)
+
+
pv. malvacearum RIATC
+
+
4-
-
+
+
pv. manihotis Xml25D
+
+
+
(+)
+
+
pv. papavericola XP5
+
+
+
-
(+)
+
pv. phaseoli 85-6
+
4-
+
-
+
+
pv. phaseoli "fuscans" XP163A
+
+
+
-
+
+
pv. physalidicola XP172
+
+
+
+
+
+
pv. pelargonii XCP58
4-
+
+
-
(+)
+
pv. poinsettiicola 071-424
+
+
+
-
+
+
pv. pruni X1219L
4-
+
+
-
+
+
pv. raphani 69-2
+
+
+
-
+
+
pv. secalis XC129C
-
-
-
-
-
-
pv. taraxaci XT11A
+
+
+
-
(+)
+
pv. translucens 80-1
-
-
-
-
-
-
pv. vesicatoria
75-3
+
+
+
+
+
+
XV56
-
+
+
-
+
+
pv. vignicola 81-30
+
+
+
-
+
+
pv. vitians XVIT
+
+
+
-
+
+
X198
+
+
+
+
+
+
XCF
+
+
+
(+)
+
+
T-55
-
-
-
-
-
-
INA
-
-
-
-
-
-
Continued on following page

Table 3-1Continued
45
Strain Southern hybridization DNA amplification
355 bp
840 bp
1,075 bp
355 bp
840 bp
1,075 bp
X. albilineans 91 -065
-
-
-
-
-
-
X. fragariae XI297
+
+
+
-
+
+
X. maltophilia
-
-
-
-
-
-
Acidovorax avenae
subsp. avenae UK142-A
-
-
-
-
-
-
subsp. citrulli UK20
-
-
-
-
-
-
Agrobacterium tumefaciens
LBA 1050
-
-
-
-
-
-
Clavibacter michiganense
subsp. michiganense 69-1
-
-
-
-
-
-
Erwinia
E. carotovora subsp.
carotovora
K-SR-347
-
-
-
-
-
-
B-SR-38
-
-
-
-
-
-
E. herbicola NF-33
-
-
-
-
-
-
E. stewartii SW2
-
-
-
-
-
-
Pseudomonas
P. solanacearum K60
-
-
-
-
-
-
P. syringae
pv. syringae INB
-
-
-
-
-
-
pv. tomato 987
-
-
-
-
-
-
Xylella fastidiosa 89-1
-
-
-
-
-
-
a+, positive reaction; negative reaction; (+), weak signal.

46
pathogens X. maltophilia and the opportunistic strains of X. campestris T-55 and IN A,
did not hybridize to any of the three hrp fragments amplified from X. campestris pv.
vesicatoria 75-3 (Fig. 3-4; Table 3-1).
DNA from strains of the plant pathogens of the genera Acidovorax,
Agrobacterium, Clavibacter, Erwinia, Pseudomonas, and Xylella failed to hybridize to
the three hrp fragments (Table 3-1). Furthermore, total genomic DNA of E. herbicola,
a bacterium commonly associated with plant tissue, also did not hybridize to any of the
hrp fragments amplified from X. campestris pv. vesicatoria 75-3 under the conditions
used (Table 3-3).
Amplification of the hrp fragments from other X. campestris pathovars and related
Xanthomonas spp.
Primer pairs RST2 plus RST3 and RST21 plus RST22 were used to amplify
DNA sequences from strains representing X. fragariae and 28 different pathovars ofX.
campestris. In all cases with both sets of primers, fragments of identical sizes were
amplified from different pathovars of X. campestris and related Xanthomonas spp. (Fig.
3-5; Table 3-1). However, amplification with the primers RST2 and RST3 usually
gave low yield of DNA for strains of A. campestris pv. carotae, X. campestris pv.
gardneri, X. campestris pv. papavericola, X. campestris pv. pelargonii, and X.
campestris pv. taraxaci (Fig. 3-5; Table 3-1). Although DNA isolated from X.
campestris pv. holcicola hybridized to the 840 bp hrp fragment from X. campestris pv.
vesicatoria strain 75-3 (Fig. 3-4), primers RST2 and RST3 did not amplify the DNA
fragment from this pathovar (Table 3-1). No amplification occurred with purified total
genomic DNA from a number of bacteria, including plant pathogenic strains of the
xanthomonads X albilineans, X. campestris pv. secalis, and X. campestris pv.
translucens, and of the genera Acidovorax, Agrobacterium, Clavibacter, Erwinia,

47
Fig. 3-5. Amplification of the 355-bp (A), 840-bp (B), and 1,075-bp (C) fragments of
the hrp gene cluster from strains of Xanthomonas campestris. Lanes: M, phage A.
restricted with £coRI and Hincll; 1, X. campestris pv. vesicatoria 75-3; 2, X
campestris pv. bilvae XCB; 3, X. campestris pv. carotae #13; 4, X. campestris pv. citri
9771; 5,X campestris pv. citrumelo FI; 6,X. campestris pv. dieffenbachiae 729; 7,X
campestris pv. fci X151; 8, X. campestris pv. gardneri XG101; 9, X. campestris pv.
maculifoliigardeniae X22j; 10,X campestris pv. manihotis Xml25D; \\,X.
campestris pv. pelargonii XCP58; 12, X. campestris pv. phaseoli "fuscans" XP163A;
\2>,X. campestris pv. poinsettiicola 071-424; 14, X campestris pv. taraxaci XT11A;
\5,X. campestris pv. vignicola 81-30; 16, X. campestris pv. vitians XVIT; 17, X
campestris pv. physalidicola XP172. Molecular sizes are given in base pairs.

48
Pseudomonas, and Xylella, and from the non plant pathogens E. herbicola, X.
maltophilia, and the opportunistic strains of X. campestris T-55 and INA, when either
RST2 plus RST3 or RST21 plus RST22 were used (Table 3-1). The failure to amplify
the DNA fragments from all those bacterial strains was expected because of the lack of
hybridization to the three hrp fragments from X. campestris pv. vesicatoria 75-3 (Table
3-1).
DNA fragments delineated by the primers RST9 and RST10 were consistently
amplified only from strains of X. campestris pv. fci, X campestris pv. physalidicola,
X. campestris pv. vesicatoria, and A! campestris X198 (Fig. 3-5; Table 3-1). However,
some pathovars of X. campestris, including alfalfae, citrumelo, maculifoliigardinae,
and manihotis, as well as the strain XCF, sometimes produced low yield in the
amplification of the 355 bp fragment (Fig. 3-5; Table 3-1).
The identity of these hrp related fragments amplified from different strains of
plant pathogenic Xanthomonas spp. was further confirmed by Southern hybridization
analysis. Internal portions of the 840- and 1,075-bp DNA fragments, as well as the
entire 355 bp fragment amplified from X. campestris pv. vesicatoria 75-3, were used as
probes. The internal probe for the 840 bp fragment consisted of a 271-bp insert of the
plasmid pXV840, and the internal probe for the 1,075-bp fragment consisted of a 335-
bp insert of the plasmid pXV1075. The inserts were obtained from fragments
amplified from DNA of A campestris pv. vesicatoria strain 75-3 by cloning Sau3Al
digests into the BamHl site of the vector pBluescript II KS +/- (Stratagene, La Jolla,
CA). In all cases, the DNA fragments amplified from different strains of Xanthomonas
spp. with each set of primers hybridized with the respective probe (data not shown).

49
Restriction endonuclease analysis of amplified hrp-related DNA fragments
To address the question of degree of sequence conservation among different
strains, I examined the 840- and 1,075-bp hrp fragments amplified from strains of
different pathovars of X. campestris, as well as from X. fragariae, by restriction
endonuclease analysis with the endonucleases Cfol, HaeIII, Saw3AI, and Taql.
Restriction fragment length polymorphisms were apparent for both fragments. For
example, the 1,075-bp fragment, amplified from strains of different pathovars ofX.
campestris by using primers RST21 and RST22 and then restricted with Hae III and
Saw3AI, yielded different restriction patterns (Fig. 3-6). Although RFLPs were
observed with all four endonucleases for both the 840- and the 1,075-bp fragments,
restriction analysis with Cfo\ and HaelII produced more distinct patterns for
differentiation of the groups or pathovars of X. campestris.
Pathogenicity function of the DNA sequence of X. campestris pathovars from which
hrp related fragments were amplified
To determine if the DNA sequence from which hrp related fragments were
amplified has common functions in different strains of X campestris, I investigated the
functional homology of these fragments in X. campestris pv. vesicatoria group B and X
campestris pv. pelargonii. The cosmid clones pXV56/3-48, of a pLAFR3 library of
strain XV56 of A! campestris pv. vesicatoria group B (Beaulieu et al., 1991), and
pXCP58/2, of a pLAFR3 library of A. campestris pv. pelargonii XCP58 (Gerald V.
Minsavage, personal communication), were identified by amplification of the 1,075-bp
hrp fragment with the primers RST21 and RST22 Plasmid pXV56/3-48 was
transferred into the A! campestris pv. vesicatoria mutants &5-\0whrpA22, 85-
\0::hrpB85, S5-\0::hrpC44, 85A0::hrpD137, 85-10::hrpE75, and %5-\0::hrpF318,
which carry Tn3-gws insertions in the different hrp complementation groups (Bonas et

50
Fig. 3-6. Restriction analysis of the 1,075-bp DNA fragment of the hrp gene cluster
amplified from strains of Xanthomonas campestris and restricted with the
endonucleases HaelU (A) and Sau3Al (B). Lanes: M, phage X restricted with Pstl; 1,
X. campestris pv. vesicatoria 75-3; 2,X. campestris pv. bilvae XCB; 3, X campestris
pv. carotae #13; 4, A! campestris pv. citri 9771; 5,X. campestris pv. citrumelo FI; 6, X.
campestris pv. dieffenbachiae 729; 7, X campestris from Strelitzia reginae strain
X198; 8, X. campestris pv. fici X151; 9, X campestris pv. maculifoliigardeniae X22j;
10, X. campestris pv. manihotis Xml25D; \ \,X. campestris pv. pelargonii XCP58; 12,
X. campestris pv. phaseoli "fuscans" XP163A; 13, X campestris pv. poinsettiicola 071-
424; 14, X. campestris pv. taraxaci XT11A; 15, A campestris pv. vignicola 81-30; 16,
X. campestris pv. vitians XVIT; 17, X. campestris pv. papavericola XP5; 18, X.
campestris pv. holcicola G-23. Molecular sizes are given in base pairs.

51
al., 1991). Plasmid pXV56/3-48 fully restored the pathogenicity of mutants with
mutations in hrpB, hrpD, and hrpE, and the hypersensitive reaction-inducing ability but
not pathogenicity to the hrpC mutant. However, this plasmid failed to complement the
hrpA and hrpF mutants. Similarly, plasmid pXCP58/2 from X campestris pv.
pelargonii was also transferred into the six nonpathogenic 7n3-gus mutants ofX.
campestris pv. vesicatoria 85-10. This plasmid fully restored the wild type phenotype
to mutants with mutations in hrpC, hrpD, and hrpE but failed to complement hrpA,
hrpB, and hrpF7n3-gus mutants.
Sensitivity of the amplification of hrp fragment for detection ofX. campestris pv.
vesicatoria
The sensitivity of the amplification of specific DNA fragments in detection of
X. campestris was determined by using 10-fold dilutions of purified total bacterial
DNA of X. campestris pv. vesicatoria 75-3. The oligonucleotide primers RST9 plus
RST10 and RST2 plus RST3 were used for amplification of the 355- and 840-bp hrp,
respectively, from samples containing as little as 0.25 pg of total bacterial DNA after
30 cycles of DNA amplification (Fig. 3-7).
Discussion
Sequence homology to small hrp fragments amplified from X. campestris pv.
vesicatoria 75-3 was found among plant pathogenic strains of several pathovars of A
campestris and related Xanthomonas spp. by Southern hybridization analysis. DNA
probes representing regions of the hrpB, hrpC, and hrpD loci hybridized strongly to
total DNA from strains of 28 different pathovars ofX. campestris. These results
confirm and extend previous observations (Bonas et al., 1991; Stall and Minsavage,

52
Fig. 3-7. Amplification of the 355-bp (A) and (B) 840-bp fragments of
complementation group B of the hrp gene cluster from samples with different amounts
of DNA template of Xanthomonas campestris pv. vesicatoria 75-3. Lanes: M, phage X
restricted with £coRI and HindUl; 1, 25 ng; 2, 2.5 ng; 3, 0.25 ng; 4, 25 pg; 5,2.5 pg; 6,
0.25 pg; 7, 0.025 pg. Molecular sizes are given in base pairs.

53
1990) on the conservation of the hrp region among the plant pathogenic xanthomonads.
In contrast, the DNA of A campestris pv. secalis and X campestris pv. translucens did
not hybridize to the hrp fragments. X campestris pv. secalis is genetically only weakly
related to a few pathovars of X campestris based on DNA-DNA hybridization
experiments (Hildebrand et al., 1990). Previously, X campestris pv. translucens has
shown only weak hybridization to DNA probes representing the entire hrp gene cluster
region of X campestris pv. vesicatoria (Bonas et al., 1991; Stall and Minsavage, 1990).
Although hybridization to DNA of all strains of X. campestris included in this study
was not observed, the hrp region seems useful for detection and identification of a large
number of plant pathogenic xanthomonads. A major advantage in using hrp sequence
is the lack of homology to DNA of non plant pathogenic xanthomonads, as observed
for X maltophilia and the opportunistic X. campestris strains T-55 and INA, as well as
other plant pathogens of the genera Acidovorax, Agrobacterium, Clavibacter, Erwinia,
Pseudomonas, and Xylella.
The three pairs of oligonucleotide primers described in this study are specific
for the hrpB, hrpC, and hrpD regions of A campestris pv. vesicatoria 75-3 and were
used to amplify homologous DNA fragments from X. fragariae and from 31 of 33
plant pathogenic taxa of A campestris tested, which comprise at least 28 different
pathovars of this species. In all cases, each set of primers amplified DNA fragments
identical in size, suggesting a high degree of structural conservation between operons,
as seen with primers RST21 and RST22. Cloned regions of DNA of A campestris pv.
vesicatoria group B and A! campestris pv. pelargonii, from which the hrp related
fragments were also amplified, fully restored pathogenicity to several nonpathogenic
Tn3-gws mutants of A campestris pv. vesicatoria 85-10. This supports the contention

54
that these fragments were amplified from DNA sequences which also control the
pathogenicity in other xanthomonads.
In contrast to the narrow spectrum of oligonucleotide primers previously used
for detection and identification of only certain strains of X. campestris (Hartung et al.,
1993), the hrp specific primers RST2 plus RST3 and RST21 plus RST22 seem very
useful for the identification of a large range of plant pathogenic xanthomonads. This is
perhaps not surprising, because the hrp region seems very conserved among different
plant pathogenic xanthomonads as determined by Southern hybridization experiments
in the present and previous studies (Bonas et ah, 1991; Stall and Minsavage, 1990).
Furthermore, the nucleotide sequences of the primers RST21 and RST22 are identical
to corresponding sequences of pathogenicity genes of X. campestris pv. glycines
(Hwang et ah, 1992). In addition, the cloned fragment from X. campestris pv. glycines
complements hrpD mutants of X. campestris pv. vesicatoria (Ulla Bonas, personal
communication), suggesting functional homology between these regions from both
pathovars.
Primers RST9 and RST10, which delineate a fragment of 355 bp, allowed DNA
amplification only from a limited number of pathovars of X. campestris, despite
hybridization of the fragment to the majority of the strains of this species. It should be
noted that the sequence of RST9 originates from hrpB6, a gene for a putative ATPase
that seems to be highly conserved among different bacteria at the protein sequence
level (Fenselau et ah, 1992). These results indicate differences in the DNA sequences
of Xanthomonas spp. corresponding to one or both primers used. However, this set of
primers seems useful for specific detection of strains of X. campestris pv. vesicatoria
group A, X. campestris pv. fici, X. campestris pv. physalidicola, and X campestris
X198.

55
The identification of strains of X. campestris at the pathovar level is difficult
even by using different techniques, such as fatty acid analysis (Chase et al., 1992;
Graham et al., 1990), serology (Alvarez et al., 1991), metabolic profile (Chase et al.,
1992; Van den Mooter and Swings, 1990), and SDS-PAGE of proteins (Vauterin et al.,
1991a); thus, the restriction analysis of amplified hrp related fragments may be a
valuable tool for identification of subgroups of plant pathogenic strains and pathovars
of campestris. For example, restriction analysis with frequent-cutting endonucleases
produced a characteristic restriction pattern for the 840- or 1,075-bp DNA fragments
amplified from the Xanthomonas spp. included in this work; this appears to be highly
conserved within each group of certain plant pathogenic xanthomonads (Chapter 4). In
this way, RFLP profiles of a particular hrp region could be established for each plant
pathogenic group of xanthomonads thus facilitating the identification and classification
of these bacterial strains.
The use of oligonucleotide primers provides a sensitive and specific tool for
detection of DNA by amplification. Theoretically, the limit of detection of an
amplifiable DNA sequence is estimated to be as low as one single target cell in the
reaction mixture (Steffan and Atlas, 1991). In our studies, we were able to find a
detectable signal for as low as 0.25 pg of total bacterial DNA. This level of sensitivity
is comparable to those obtained by others without the use of any technique to enhance
the signal (Pickup, 1991; Steffan and Atlas, 1988). Furthermore, X. campestris pv.
vesicatoria could be detected in plant samples containing less than 100 CFU/ml,
without prior enrichment or cultivation of the organism (Chapter 7). In addition to the
sensitivity of the technique, the specificity of the oligonucleotide primers to plant
pathogenic xanthomonads certainly assures selectivity against background nontarget
microorganisms which are always present in the samples.

56
In conclusion, the results presented here indicate that plant pathogenic strains of
X. campestris and related xanthomonad species can be detected and may be identified
by analysis of DNA fragments amplified with hrp gene-specific primers. The
conservation of the hrp DNA sequence among a large number of pathovars ofX.
campestris, as well as in related Xanthomonas spp., but lack of the hrp DNA sequence
among non-plant pathogenic bacteria, makes this method a useful tool for detection and
identification of many plant pathogens. Consequently, hrp oligonucleotide primers
may be also useful to determine the pathogenic nature of unknown xanthomonads.
This is particularly significant for assessing the complex population of plant pathogenic
and nonpathogenic xanthomonads associated with plants and plant parts. The presence
of plant pathogenic strains in such samples may be determined by amplification of the
hrp fragments without the need of the troublesome methods of isolation of the
organism and inoculation into potential host plants. Moreover, RFLPs detected in the
genome of different strains seem valuable for the study of the relatedness of plant
pathogenic xanthomonads, particularly among Xanthomonas spp. and pathovars of X.
campestris. Of course, this has to be extended by testing larger numbers of strains for
each species, subspecies, or pathovar. The genetic methods of analyzing populations of
bacteria will provide valuable additional information for taxonomic, ecological, and
epidemiological studies of plant pathogenic xanthomonads.

CHAPTER 4
CHARACTERIZATION OF PLANT PATHOGENIC Xanthomonas
BASED ON RESTRICTION ANALYSIS OF AMPLIFIED DNA
SEQUENCES RELATED TO THE hrp GENES
Identification of plant pathogenic xanthomonads has relied on isolating the
organism of interest in pure culture and performing predetermined biochemical,
serological, and pathological tests (Bradbury, 1984; Saettler et al., 1989; Schaad,
1988). More recently, methods based on metabolic fingerprinting (Chase et ah, 1992;
Hildebrand et ah, 1993; Jones et ah, 1993a; Van den Mooter and Swings, 1990;
Vemiere et ah, 1993), protein profiling (Vauterin et ah, 199lab), and fatty acid
composition (Chase et ah, 1992; Hodge et ah, 1992; Yang et ah, 1993) have also been
used for identification of these plant pathogens. Although these methods have proved
useful for the differentiation of certain plant pathogenic xanthomonads, the specific
identification of the xanthomonads at the subgeneric level is still difficult. Further, the
difficulty is particularly evident for the infrasubspecific identification of the members
of X campestris. This species includes at least 125 different pathovars which are
usually distinguished only on the basis of their host range or by the disease they cause
(Bradbury, 1984; Dye et al., 1980; Hayward, 1993). Therefore, there is a need for more
rapid and unambiguous procedures for identification and detection of these plant
pathogens.
In recent years, nucleic acid-based techniques have been used for
characterization of the xanthomonads. DNA-DNA hybridizations were applied to
57

58
determine the intrageneric relationships among different xanthomonads (Hildebrand et
al., 1990; Palleroni et al., 1993; Vauterin et al., 1993) and 21 DNA homology groups
were established (Vauterin et al., 1993). Some DNA homology groups were delineated
which include pathovars of X. campestris causing diseases on closely related plants as
occurs with the xanthomonads pathogenic on leguminous (Hildebrand et al., 1990;
Palleroni et al., 1993; Vauterin et al., 1993), gramineous (Vauterin et al., 1993), and
cruciferous plants (Vauterin et al., 1993). Furthermore, these groups established based
on DNA homology are strongly supported by the groupings based on SDS-PAGE of
proteins and fatty acid analyses (Vauterin et al., 1991a, 1992; Yang et al., 1993).
However, the genetic relatedness among the xanthomonads is not always correlated
with pathogenicity features.
A striking feature of the plant pathogenic xanthomonads is the genetic diversity
among strains of the same pathovar or different pathovars of X. campestris that cause
diseases on the same or highly related hosts. For example, strains ofX. campestris pv.
vesicatoria groups A and B that cause diseases on solanaceous plants are only about
33% related on the basis of DNA homology (Stall et al., 1994). Further, this low
genetic similarity and the differences in phenotypic features strongly supports the
conclusion that these two groups of X. campestris pv. vesicatoria may even belong to
different species (Stall et al., 1994). In another study, Egel et al. (1991) examined the
genetic relatedness of the xanthomonads causing diseases of citrus and found that
strains of the citrus canker pathogen X. campestris pv. citri and strains of the citrus
bacterial spot agent X campestris pv. citrumelo are less than 60% related to one
another. The diverse nature of the X. campestris strains causing diseases on citrus has
also been determined on the basis of genomic fingerprinting (Hartung and Civerolo,
1987), analysis of restriction fragment length polymorphism (RFLP) (Gabriel et al.,

59
1988, 1989; Graham et al., 1990; Hartung and Civerolo, 1989), and fatty acid
composition (Graham et al., 1990; Stall and Hodge, 1989).
The genetic diversity of the xanthomonads has also been investigated by using
other nucleic acid-based techniques, such as genomic fingerprinting with frequent- and
rare-cutting restriction endonucleases (Egel et al., 1991; Hartung and Civerolo, 1987;
Pruvost et al., 1992) and RFLP analyses of plasmid and genomic DNAs (Berthier et al.,
1993; DeParasis and Roth, 1990; Gabriel et al., 1988, 1989; Garde and Bender, 1991;
Gilbertson et al., 1989; Gottwald et al., 1991; Graham et al., 1990; Hartung, 1992;
Hartung and Civerolo, 1989, 1991; Lazo and Gabriel, 1987; Lazo et al., 1987; Leach et
al., 1990, 1992; Qhobela and Claflin, 1988, 1992; Verdier et al., 1993). In an extensive
study, Lazo et al. (1987) differentiated strains representing 26 different pathovars of X.
campestris based on RFLP analysis by using random DNA probes obtained from a
cloned fragment derived from X. campestris pv. citrumelo. The RFLP analyses
revealed profiles highly conserved and characteristic for each pathovar tested. It was
possible to differentiate all the pathovars of X. campestris included in the study by
using more than one DNA probe or by digesting the genomic DNA with different
restriction endonucleases (Lazo et al., 1987). Plasmid DNA fragments have also been
examined as probes for specific identification of certain groups of X. campestris (Garde
and Bender, 1991; Gilbertson et al., 1989; Hartung, 1992; Hartung et al., 1993; Lazo
and Gabriel, 1987). However, DNA probes derived from specific regions in the
bacterial genome and conserved among plant pathogenic bacteria may be more
meaningful for specific differentiation and characterization of strains, pathovars, or
species of Xanthomonas.
The hypersensitive reaction and pathogenicity (hrp) gene cluster that has been
found in several plant pathogenic bacteria may be appropriate for specific identification

60
of plant pathogenic xanthomonads. The hrp genes are required by plant pathogens to
cause disease on susceptible hosts and hypersensitive reaction (HR) on resistant, or on
nonhost plants (Willis et al., 1991). The hrp genes have been found in different plant
pathogenic bacteria, including Erwinia amylovora (Beer et al., 1991), Pseudomonas
solanacearum (Boucher et al., 1987), P. syringae pv. phaseolicola (Lindgren et al.,
1986), and X. campestris pv. vesicatoria (Bonas et al., 1991). Furthermore, the hrp
gene sequence seems to be highly conserved among different pathogenic bacteria at the
functional level (Ulla Bonas, personal communication; Fenselau et al., 1992; Gough et
al., 1992; Hwang et al., 1992). On the contrary, nonpathogenic bacteria are unable to
cause disease or HR on plants and they apparently lack DNA sequences similar to the
hrp genes (Lindgren et al., 1986; Stall and Minsavage, 1990). In previous studies, the
hrp-genes sequences were useful for differentiating pathogenic from nonpathogenic
xanthomonads (Stall and Minsavage, 1990), as well as for differentiating strains from
different pathovars of X. campestris and related plant pathogenic Xanthomonas spp.
(Chapter 3).
The purpose of this study was to characterize the plant pathogenic
xanthomonads by restriction analysis of DNA sequences related to the hrp genes of X.
campestris pv. vesicatoria. I used oligonucleotide primers specific for the hrp genes to
amplify hrp-related DNA fragments from different plant pathogenic xanthomonads by
the polymerase chain reaction. Further, the reliability of the identification of the
xanthomonads was assessed by examining different groups of plant pathogenic
xanthomonads that comprise the species X. albilineans, X. fragariae, and at least 31
different pathovars of X. campestris.

61
Materials and Methods
Bacterial strains and culture conditions
The bacterial strains used in this study, their taxonomic designation, and their
sources are listed in the Appendix A. All the bacterial strains had previously been
identified as members of Xanthomonas spp. by fatty acid analysis (Nancy C. Hodge,
personal communication). The strains ofX. campestris were grown on nutrient agar
(Becton Dickson, Cockeysville, MD). Strains of X. albilineans and X fragariae were
grown on Wilbrink's medium (Koike, 1965). Broth cultures were grown 24 hours on a
rotatory shaker (150 rpm) at 28C. All strains were streaked on appropriate media and
cultures were obtained from single colonies. The strains were stored and maintained in
tap water at room temperature, or in 30% glycerol at -70C, or both.
DNA extraction
Total bacterial genomic DNA was prepared by using the phenol-chloroform
extraction procedure described by Ausubel et al. (1987) with minor modifications.
Bacterial cells were harvested by centrifuging for 2 min at 16,000 g. The pellet was
washed once in 1 ml of distilled water, and then resuspended in 567 pi of TE buffer (10
mM Tris-Cl, pH 8.0; 1 mM EDTA, pH 8.0). Lysis solution containing Proteinase K
(Boehringer Mannheim, Indianapolis, IN) and sodium dodecyl sulfate (SDS) (Sigma,
St. Louis, MO) were added to final concentrations of 100 pg/ml and 0.5%,
respectively. The samples were incubated for 1 h at 37C. Sodium chloride and
hexadecyltrimethyl ammonium bromide (Sigma) were added for a final concentration
of 0.7 M and 1%, respectively, and incubated for 10 min at 65C. DNA extraction was

62
performed by sequential chloroform-isoamyl alcohol (24:1) and phenol-chloroform-
isoamyl alcohol (25:24:1) extractions. For each extraction, the samples were hand
shaken continuously and gently for 10 min and centrifuged for 5 min at 16,000 g.
After isopropanol precipitation, the DNA pellet was washed with 70% ethanol, and
then dried under vacuum for 20 min. The pellet was redissolved in 100 pi of TE and
stored at 4C.
DNA amplification
Two sets of oligonucleotide primers from sequences of the hrp gene cluster of
Xanthomonas campestris pv. vesicatoria (Bonas et al., 1991) were used in this study.
The primers RST2 and RST3 delineated an 840-bp region and the primers RST21 and
RST22 delineated an 1,075-bp region of the complementation groups B and C/D of the
hrp gene cluster of X. campestris pv. vesicatoria, respectively (Chapter 3; Bonas et al.,
1991). Oligonucleotide primers were synthesized with a model 394 DNA Synthesizer
(Applied Biosystems, Foster City, CA) by the DNA Synthesis Laboratory, University
of Florida, Gainesville.
The DNA sequences were amplified in a reaction mixture of 50 pi containing 5
pi of 10X buffer (500 mM KC1, 100 mM Tris-Cl [pH 9.0 at 25C], 1% Triton X-100),
1.5 mM MgCl2, 200 pM of each deoxynucleotide triphosphate (Boehringer
Mannheim), 25 pmol of each primer, 1.25 units of Taq polymerase (Promega, Madison,
WI), and 100 ng of purified template DNA. The reaction mixture was overlaid with 50
pi of light mineral oil. Thirty amplification cycles were carried out in an automated
thermocycler PT-100-60 (MJ Research, Watertown, MA) according to the following
profiles: 30 sec of denaturation at 95C, 30 s of annealing at 62C, and 45 s of
extension at 72C for the primers RST2 and RST3 and 30 s of denaturation at 95C, 40

63
s of annealing at 61C, and 45 s of extension at 72C for the primers RST21 and
RST22. The last extension step for both profiles was extended to 5 min.
The amplified DNA sequences were detected by electrophoresis in 0.9%
agarose gels in TAE buffer (40 mM Trisacetate, 1 mM EDTA, pH 8.2) according to
standard procedures (Ausubel et al., 1987; Sambrook et al., 1989).
Restriction endonuclease analysis
The amplified DNA fragments were restricted with either endonucleases Cfol,
HaeIII, Saw3AI, or Taql, under conditions specified by the manufacturer (Promega).
The restricted fragments were separated by electrophoresis in 4% agarose gels (3%
NuSieve and 1% Seakem GTG [FMC BioProducts, Rockland, ME]) in TAE buffer at 8
V/cm. The gel was stained with 0.5 pg of ethidium bromide per ml for 40 min and
then destained in 1 mM MgS04 for 1 hour and photographed over a UV
transilluminator (Fotodyne Inc., New Berlin, WI).
Establishment of the restriction fragment patterns
The DNA restriction banding patterns were determined by visual examination
of the /irp-related fragments digested with each of the four endonucleases. The
banding patterns obtained for each fragment were used for comparison with the
patterns obtained for homologous fragment amplified from strains of a given pathovar
and also from strains of different pathovars of X. campestris and of the species X.
fragariae. A restriction profile number was assigned to each unique DNA restriction
banding pattern. This restriction profile number was then used to determine the RFLP
group of each strain or group of strains of the plant pathogenic xanthomonads included
in this study.

64
Result?
Amplification of the /irp-related fragments from strains of X. campestris and related
Xanthomonas spp.
The hrp-related fragments delineated by the primers RST2 plus RST3 and
primers RST21 plus RST22 were amplified by polymerase chain reaction from plant
pathogenic strains representing^ fragariae and 28 different pathovars ofX. campestris
(Table 4-1). The fragments amplified were of identical sizes in all cases with both sets
of primers (data not shown). Low DNA yield was obtained for the 840-bp hrp-related
fragment amplified with the primers RST2 and RST3 for all strains of X. campestris
pv. carotae, X campestris pv. gardneri, X. campestris pv. papavericola, X campestris
pv. pelargonii, X. campestris pv. taraxaci, and two strains of X. campestris pv. vitians
(Table 4-1). The 840 bp hrp fragment was not amplified from a strain ofX campestris
pv. holcicola (Table 4-1) whereas both hrp fragments were not amplified from strains
of X. albilineans,X. campestris pv. celebensis, X campestris pv. secalis, X. campestris
pv. translucens, and nonpathogenic X. campestris (Table 4-1). Moreover, both hrp
fragments were not amplified also from some strains of a few other pathovars of X.
campestris which includes X. campestris pv. fici, X campestris pv. pelargonii, X
campestris pv. phaseoli, X campestris pv. poinsettiicola, and X campestris pv. pruni
(Table 4-1).
Establishment of the restriction banding pattern profiles
The hrp-related fragments were amplified from 192 strains of plant pathogenic
xanthomonads representing X fragariae and 28 different pathovars of X campestris
(Table 4-1). Restriction fragment length polymorphism occurred in both /irp-related

65
Table 4-1. Amplification of the hrp-related fragments from strains of Xanthomonas
campestris and related Xanthomonas spp.
Species/Pathovar
No. of strains
tested
No. of strains with positive
amplification of /irp-related fragment
840 bp 1,075 bp
X campestris
pv. alfalfae
2
2
2
pv. armoraciae
3
3
3
pv. begoniae
9
9
9
pv. bilvae
1
1
1
pv. campestris
9
9
9
pv. carotae
7
7(7)a
7
pv. celebensis
1
0
0
pv. citri
canker A
6
6
6
canker B
7
7
7
canker C
6
6
6
pv. citrumelo
16
16
16
pv. dieffenbachiae
10
10
10
pv. fici
8
5
5
pv. gardneri
3
3(3)
3
pv. glycines
7
7
7
pv. holcicola
1
0
1
pv. incanae
1
1
1
pv. maculifoliigardeniae
1
1
1
pv. malvacearum
9
9
9
pv. manihotis
1
1
1
pv. papavericola
1
1(1)
1
pv. pelargonii
8
7(7)
7
pv. phaseoli
7
5
5
pv. phaseoli "fuscans"
1
1
1
pv. physalidicola
1
1
1
pv. poinsettiicola
7
4
6
pv. pruni
5
4
4
pv. raphani
4
4
4
pv. secalis
1
0
0
pv. taraxaci
1
1(D
1
Continued on following page

66
Table 4-1.Continued
Species/Pathovar
No. of strains
tested
No. of strains with positive
amplification of /zrp-related fragment
840 bp 1,075 bp
pv. translucens
1
0
0
pv. vesicatoria
group A
27
27
27
group B
11
11
11
pv. vignicola
3
3
3
pv. vitians
6
6(2)
6
undetermined and
1
1
1
isolated from Feronia
sp.
undetermined and
3
1
3
isolated from Hibiscus
sp.
undetermined and
1
1
1
isolated from Strelitzia
reginae
undetermined and
25
0
0
nonpathogenic
X. albilineans
1
0
0
X. fragariae
9
9
9
aNumber in brackets indicates number of strains which produced weak signal.

67
fragments amplified from the different plant pathogenic xanthomonads after digestion
with frequent-cutting endonucleases Cfol, HaeIII, SaulAl, and Taql. The combination
of Arp-related sequences and restriction endonucleases were useful for the
establishment of distinct restriction banding patterns for the different groups of plant
pathogenic xanthomonads included in this study.
The strains of plant pathogenic xanthomonads were distributed into fifty
different groups on the basis of the banding patterns of two Arp-related fragments
digested with the restriction endonucleases Cfol, Haelll, Sau3Al, and Taql (Table 4-2).
The restriction banding patterns were determined by visual comparisons of the DNA
bands separated by electrophoresis of agarose gels (Fig. 4-1 to 4-8). The 1,075-bp Arp-
related fragment amplified from different strains of xanthomonads and restricted with
the endonucleases Cfol and Haelll generated the largest number of restriction banding
patterns, 23 and 25, respectively (Fig. 4-5 and 4-6). The restriction analysis of the Arp-
related fragment with these two endonucleases was more discriminatory and was very
useful for differentiation of closely related pathovars, such as A campestris pv.
armoraciae and A! campestris pv. campestris (Table 4-2), or even to distinguish groups
of strains within pathovars, such as the case of strains of X. campestris pv.
dieffenbachiae (Fig. 4-11; Table 4-2) and A. campestris pv. citri (Chapter 6; Table 4-2).
On the other hand, the restriction analysis with less discriminatory endonucleases, such
as Taql, was very helpful to establish the genetic relatedness at pathovar level (Table 4-
2). For example, the cruciferous pathogens A campestris pv. armoraciae, A
campestris pv. campestris, A campestris pv. incanae, and A campestris pv. raphani all
belong to the Taql profile 1 for the 840-bp Arp-related fragment and to the Taql profile
3 for the 1,075-bp Arp-related fragment (Table 4-2).

68
Fig. 4-1. Restriction profiles established for the 840-bp /irp-related fragment amplified
from different strains of plant pathogenic Xanthomonas spp. and restricted with the
endonuclease Cfol. Lane M, phage X restricted with ft/I. Molecular sizes are given in
bases. See Table 4-2 for identification of strains of Xanthomonas spp. for each
restriction pattern.

69
Fig. 4-2. Restriction profiles established for the 840-bp hrp-related fragment amplified
from different strains of plant pathogenic Xanthomonas spp. and restricted with the
endonuclease HaeIII. Lane M, phage X restricted with Pstl. Molecular sizes are given
in bases. See Table 4-2 for identification of strains of Xanthomonas spp. for each
restriction pattern.

70
Fig. 4-3. Restriction profiles established for the 840-bp hrp-related fragment amplified
from different strains of plant pathogenic Xanthomonas spp. and restricted with the
endonuclease Sau3AI. Lane M, phage X restricted with Pstl. Molecular sizes are given
in bases. See Table 4-2 for identification of strains of Xanthomonas spp. for each
restriction pattern.

71
Fig. 4-4. Restriction profiles established for the 840-bp hrp-related fragment amplified
from different strains of plant pathogenic Xanthomonas spp. and restricted with the
endonuclease Taql. Lane M, phage X restricted with Pstl. Molecular sizes are given in
bases. See Table 4-2 for identification of strains of Xanthomonas spp. for each
restriction pattern.

72
M 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Fig. 4-5. Restriction profiles established for the 1,075-bp hrp-related fragment
amplified from different strains of plant pathogenic Xanthomonas spp. and restricted
with the endonuclease Cfo\. Lane M, phage X restricted with Pstl. Molecular sizes are
given in bases. See Table 4-2 for identification of strains of Xanthomonas spp. for each
restriction pattern.

Fig. 4-6. Restriction profiles established for the 1,075-bp Arp-related fragment
amplified from different strains of plant pathogenic Xanthomonas spp. and restricted
with the endonuclease HaeIII. Lane M, phage X restricted with Psft. Molecular sizes
are given in bases. See Table 4-2 for identification of strains of Xanthomonas spp. for
each restriction pattern.

74
Fig. 4-7. Restriction profiles established for the 1,075-bp hrp-related fragment
amplified from different strains of plant pathogenic Xanthomonas spp. and restricted
with the endonuclease Sau3Ad. Lane M, phage X restricted with Pstl. Molecular sizes
are given in bases. See Table 4-2 for identification of strains of Xanthomonas spp. for
each restriction pattern.

75
Fig. 4-8. Restriction profiles established for the 1,075-bp /irp-related fragment
amplified from different strains of plant pathogenic Xanthomonas spp. and restricted
with the endonuclease Taql. Lane M, phage X restricted with Pstl. Molecular sizes are
given in bases. See Table 4-2 for identification of strains of Xanthomonas spp. for each
restriction pattern.

Table 4-2. Restriction pattern of the 840- and 1.075-bp /zrp-related DNA fragments amplified from different strains of plant
pathogenic Xanthomonas spp. and digested with four restriction endonucleases
Species/Pathovar
Strain
840 bp
1,075 bp
Taq\
5au3AI HaelU
Cfo\
Taql
Saw3AI HaelU
Cfol
X. campestris
pv. alfalfae
G-22 (KS)
8a
6
16
12
8
11
13
8
82-1
8
6
16
12
8
13
13
8
pv. armoraciae
63-27,417, 756
1
3
3
4
3
4
3
2
pv. begoniae
XCB9, X274, X329, X627,
XI490, XI492, XI496
1
6
14
10
9
14
20
15
X610
1
6
14
10
9
14
21
15
X281
1
6
14
10
9
14
24
22
pv. bilvae
XCB
4
5
9
9
7
10
10
9
pv. campestris
33913, 62-1, 62-9a, 65-6b, 70-3,
70-5, 71-2, 83-1. 83-2
1
3
3
3
3
4
3
2
pv. carotae
#3, #5, #7, #9, #12, #13
ndb
nd
nd
nd
3
2
5
2
#16
nd
nd
nd
nd
3
4
3
2
Continued on following page

Table 4-2Continued
Species/Pathovar
Strain
840 bp
1,075 bp
Taql
Sau3Al Haelll
Cfol
Taql
Sau3Al Haelll
Cfol
pv. citri
canker A
3213,3340, 9760-2, 9771, Tl,
115A
2
5
1
8
7
9
10
7
canker B
B64, B69, B80, B84, B93, B94,
B148
4
5
8
9
7
9
11
11
canker C
70C, 338, 339, 340, 341,342
4
5
8
9
7
9
11
11
pv. citrumelo
FI, F54, F274, F361,3166
7
6
16
12
8
11
14
14
F59, F86
3
6
16
12
8
11
19
8
F94
3
6
16
12
8
11
15
17
F306
5
6
16
12
8
11
15
17
F254,F311
7
6
16
12
8
11
15
8
F6, F228
8
6
16
14
8
13
15
15
F100
3
6
16
12
8
12
15
11
F348
8
6
16
12
8
12
19
16
F378
3
6
16
14
8
12
14
8
Continued on following page

Table 4-2Continued
Species/Pathovar
Strain
840 bp
1,075 bp
Taq\
Saw3AI Haelll
Cfol
Taql
Saw3AI Haelll
Cfol
pv. dieffenbachiae A
X422, XI51, X790, X1272
2
7
12
10
9
15
22
18
X729, X736, X738, X745
2
7
12
10
9
14
22
18
B
X260, X763
3
6
13
11
10
14
23
15
pv. fici A
X125
3
6
16
12
8
11
15
8
X151
7
6
16
12
8
11
18
11
X212
5
6
16
12
8
11
18
11
B
X208, X702
1
4
4
5
5
7
8
5
pv. gardneri
XG101, XV6, 1066
nd
nd
nd
nd
3
3
4
2
pv. glycines A
G-56, 86-16, 86-17, 86-18, 86-
20, 87-2
2
5
6
7
7
9
10
7
B
1706A
nd
nd
nd
nd
5
7
8
5
pv. holcicola
G-23
nd
nd
nd
nd
11
16
25
23
pv. incanae
9561-1
1
3
3
3
3
4
3
2
pv.
maculifoliigardeniae
X22j
2
7
10
10
9
14
24
20
Continued on following page

Table 4-2Continued
Species/Pathovar
Strain
840 bp
1,075 bp
Taql
Sa3AI HaeIII
Cfol
Taql
Sau3A\ Hae III
Cfol
pv. malvacearum
RIATC, XcmA, XcmB, XcmC,
XcmD, XcmE, XcmF, XcmH,
XcmN
2
5
6
6
6
9
12
10
pv. manihotis
Xml25D
2
7
11
10
9
14
22
21
pv. papavericola
XP5
nd
nd
nd
nd
3
2
3
3
pv. pelargonii
XCP2, XCP10, XCP17, XCP36,
XCP54, XCP58, XCP60
nd
nd
nd
nd
3
2
3
2
pv. phaseoli A
81-19, 82-1, 85-6, XCPH4
2
7
12
10
9
14
22
19
B
82-17
2
5
6
7
7
9
10
7
pv. phaseoli 'Tuscans"
XP163A
4
5
15
9
7
9
11
12
pv. physalidicola
XP172
3
6
16
12
nd
nd
nd
nd
pv. poinsettiicola
B
X87, X202, X259
10
4
5
5
5
8
9
6
A
X352, 071-424
3
6
16
12
8
11
15
8
pv. pruni
X1219L, X1219S, X1220L,
X1220S
3
6
16
12
8
11
19
8
Continued on following page

Table 4-2Continued
Species/Pathovar
Strain
840 bp
1,075 bp
Taql
Saw3AI HaeIII
Cfo\
Taql
Sau3AI Hae III
Cfol
pv. raphani
69-2, 69-4B, 69-8R
1
3
3
3
4
6
7
4
16B
1
3
3
4
4
6
7
4
pv. taraxaci
XT11A
nd
nd
nd
nd
3
3
6
2
pv. vesicatoria
group A
75-3, 82-4, 85-10, 87-21, 87-
35T, 87-44T, 87-48T, 89-8, 89-
10, 90-20, 90-21, 90-27, 90-40,
91-66, 91-72, 92-11, 92-15, 92-
16, 92-17, 92-119, 1712,6107,
XV14, XV17
3
6
16
12
8
11
16
17
92-118,92-120
3
6
16
13
8
12
17
17
90-60
3
6
16
12
8
12
17
17
group B
141.0226A, 0350A, 695,853,
1062, B-3,B-20, BA28-1, BV5-
5, XV56
1
1
1
1
1
1
1
1
pv. vignicola
81-30, 82-38, G-55
6
7
16
12
7
9
11
11
Continued on following page
OO
O

Table 4-2Continued
Species/Pathovar
Strain
840 bp
1,075 bp
Taql
Sau3Al HaeIII
CM
Taql
Sau3A\ Haelll
CM
pv. vitians A
ICPB101, XVIT, X1215
2
7
12
10
9
14
22
17
B
XV2
2
5
6
6
6
9
12
10
C
ICPB164, XV3
nd
nd
nd
nd
3
2
4
2
undetermined and
isolated from Feronia
sp.
XCF
4
5
9
9
7
10
10
10
undetermined and
isolated from Hibiscus
sp.
X52
nd
nd
nd
nd
5
8
9
6
undetermined and
isolated from Strelitzia
reginae
X198
9
7
16
12
8
11
15
8
X. fragariae
GC-6259, GC-6265, X1238,
X1241, X1244, XI246, X1292,
X1298, XI426
1
2
2
2
2
5
2
13
Restriction pattern group; Number of the restriction pattern groups refer to lanes in Fig. 4-1 to 4-8.
bnd, not determined.

82
Variability of homologous hrp-related fragments at species and pathovar levels
Strains of 26 campestris pv. holcicola, X. campestris pv. vesicatoria group B,
and X. fragariae each produced characteristic restriction patterns for both /^-related
fragments digested with four different endonucleases. In the case of the 1,075-bp hrp-
related fragment amplified from strains of the three taxa and restricted with either Cfol,
HaeIII, SaulAl, or Taql, the restriction banding patterns produced were unique (Table
4-2). Furthermore, the restriction pattern obtained for strains of 26 campestris pv.
vesicatoria group B and X. fragariae were identical within each group for all four
endonucleases (Table 4-2). In regard to the 840-bp hrp-related fragment, the restriction
patterns were also unique for the strains of X. campestris pv. vesicatoria group B and X.
fragariae with the exception of the banding patterns obtained by using the restriction
endonuclease Taql (Table 4-2). In this case, the restriction pattern of strains ofX.
campestris pv. vesicatoria group B and X. fragariae were identical to each other and
also to strains of another group of plant pathogenic xanthomonads that included X.
campestris pv. armoraciae, X. campestris pv. begoniae, X campestris pv. campestris,
X campestris pv. fici group B, X. campestris pv. incanae, and V campestris pv.
raphani (Table 4-2). The 840-bp /^-related fragment was not amplified from the
strain ofX. campestris pv. holcicola (Table 4-1).
On the contrary, strains of the other pathovars of X. campestris usually
produced restriction banding patterns that were common to other groups of
xanthomonads in at least one combination of Arp-related fragment and restriction
endonuclease (Table 4-2). However, the combined profiles of four endonucleases and
two Arp-related fragments provided differentiation for almost all groups of plant
pathogenic xanthomonads included in this study (Table 4-2). Among the exceptions
are the strain 9561-1 ofX. campestris pv. incanae and strain #16 of26 campestris pv.

83
carotae that could not be differentiated from the strains ofZ campestris pv. campestris
(Table 4-2). Other cases include the strain 82-17 of2f campestris pv. phaseoli that
could not be differentiated from strains of X. campestris pv. glycines group A; strains
F59 and F86 of X. campestris pv. citrumelo were indistinguishable from strains ofX.
campestris pv. pruni; strain XI25 of X. campestris pv. fici group A was identical to
strains ofX. campestris pv. poinsettiicola group A; strain XV2 ofX. campestris pv.
vitians was identical to strains of X campestris pv. malvacearum; and strain X52
isolated from Hibiscus sp. was indistinguishable from strains of X campestris pv.
poinsettiicola group B (Table 4-2).
The restriction analysis of the /zrp-related sequences revealed some apparent
groupings among the plant pathogenic xanthomonads based on the similarity in the
restriction banding patterns of homologous hrp-related fragments. The pathogens of
cruciferous plants, X. campestris pv. armoraciae, X campestris pv. campestris, X.
campestris pv. incanae, and X campestris pv. raphani have very similar restriction
profiles for both ^-related fragments and they formed a very distinct group (Table 4-
2). Strains of X campestris pv. armoraciae could be differentiated from X. campestris
pv. campestris and X. campestris pv. incanae only by the restriction analysis of the 840-
bp hrp related fragment with the endonuclease Cfol (Table 4-2). The restriction
patterns obtained for these pathovars were usually identical or very similar with the
exception to the restriction patterns of the 1,075-bp fragment for strains of2f
campestris pv. raphani (Table 4-2). Similarly, the strains of the pathovars X.
campestris pv. carotae, X. campestris pv. gardneri, X. campestris pv. papavericola, X.
campestris pv. pelargonii, X campestris pv. taraxaci, and X. campestris pv. vitians
group B also formed a unique group, although they cause diseases on different hosts.
The restriction analysis of the 1,075-bp /^-related fragment amplified from strains of

84
these pathovars revealed very similar banding patterns (Table 4-2). All strains of the
latter pathovars belong to the restriction banding profiles 3 and 2 or 3 of the
endonucleases Taql and Sau3AI, respectively (Table 4-2). The restriction banding
patterns from the strains of these pathovars for the endonucleases Cfol and HaeIII are
also very similar and they have several DNA bands in common (Fig. 4-5 and 4-6).
Although the other pathovars of X. campestris also showed some trends in regard to the
similarity of the restriction patterns for both /i/y?-related fragments, a more
discriminating analysis by using some clustering approach may be more appropriate to
establish the precise delineation of the groupings.
Variability of the fap-related restriction patterns among strains within different taxa of
Xanthomonas spp. and pathovars of X campestris
Sequence variation in the hrp region was revealed by restriction analysis of
both /zrp-related fragments amplified from strains within the same taxa of plant
pathogenic xanthomonads. Further, this variation indicated that these plant pathogens
may differ in the genetic variability of the DNA sequence of the hrp gene cluster.
Whereas some taxa have very uniform restriction patterns for both hrp-related
sequences, others seem to be comprised of very distinct groups of strains. For instance,
the restriction banding patterns for all nine X fragariae strains, originally isolated from
diseased material in Florida, California, and Canada (Ann R. Chase, personal
communication), were identical to each other when both /irp-related fragments were
restricted with either Cfol, Hae III, Sau3Al, or Taql (Fig. 4-9; Table 4-2). Similarly,
uniform banding patterns were observed for strains within the pathovars X campestris
pv. campestris, X campestris pv. malvacearum, X. campestris pv. pelargonii, and X
campestris pv. raphani (Table 4-2). Strains within the pathovars X campestris pv.
armoraciae, X. campestris pv. gardneri, and X. campestris pv. vignicola also produced

85
Fig. 4-9. Restriction analysis of the 840-bp hrp-related fragments amplified from
strains of Xanthomonas fragariae and restricted with HaeIII (Lanes 2 to 10) and Cfol
(Lanes 12 to 20). Lanes 1 and 11, phage X restricted with Pstl. Lanes 2 and 12,
X1238; 3 and 13, X1241; 4 and 14, X1244; 5 and 15, X1246; 6 and 16, X1292; 7 and
17, X1298; 8 and 18, X1426; 9 and 19, GC6265; 10 and 20, X. campestris pv.
vesicatoria 75-3. Molecular sizes are given in bases.

86
identical restriction patterns to each other though less than four strains were examined
for each of these pathovars (Table 4-2). In the case of the pathovars X. campestris pv.
gardneri and X. campestris pv. pelargonii, only the 1075-bp fragment was examined,
because the amplification of the 840-bp hrp-related fragment produced very low DNA
yield (Table 4-1). Uniform banding patterns were produced by the strains of the
pathovars of X. campestris pv. carotae, X campestris pv. phaseoli, and X campestris
pv. pruni, with the exception of only one single strain within each pathovar (Table 4-2).
In another study, nine strains of X. campestris pv. begoniae produced identical banding
patterns for the restriction analysis of the 840-bp hrp-related fragment by using either
endonuclease Cfol, HaelII, Sau3A\, or Taql (Fig. 4-10A). However, the restriction
analysis of the 1,075-bp fragment with the endonucleases Cfol or HaeIII revealed
variability in the banding pattern for strains X281 and X610 of campestris pv.
begoniae (Fig. 4-1 OB). The restriction analysis of this hrp-related fragments with the
endonucleases Sau3A\ and Taql revealed that banding patterns were identical for all
nine strains of this pathovar (Table 4-2).
In contrast, large variability was found in the restriction analysis of the hrp-
related fragments amplified from strains of several pathovars of X. campestris. This
includes the pathovars X campestris pv. citri, X campestris pv. citrumelo, X.
campestris pv. dieffenbachiae, X. campestris pv. fici, X campestris pv. poinsettiicola,
X. campestris pv. vesicatoria, and X. campestris pv. vitians (Fig. 4-11 and 4-12; Table
4-2). Strains of the citrus pathogen X. campestris pv. citrumelo were the most
heterogeneous and the 17 strains examined were divided into nine different restriction
profile groups based on the restriction analysis of the two hrp-related sequences
(Chapter 6; Table 4-2). On the other hand, strains of the pathovars X. campestris pv.
citri, X campestris pv. dieffenbachiae, and X campestris pv. vesicatoria produced

87
Fig. 4-10. Restriction analysis of the (A) 840 bp and (B) 1,075 bp DNA fragments of
the hrp gene cluster amplified from strains of Xanthomonas campestris pv. begoniae
and restricted with HaeIII (Lanes 2 to 10) and Cfo\ (Lanes 12 to 20). Lanes 1 and 11,
phage X restricted with Pstl. Lanes 2 and 12, X274; 3 and 13, X281; 4 and 14, X329; 5
and 15, X610; 6 and 16, X627; 7 and 17, X1490; 8 and 18, X1492; 9 and 19, X1496;
10 and 20, XCB9. Molecular sizes are given in bases.

88
Fig. 4-11. Restriction analysis of the (A) 840 bp and (B) 1,075 bp DNA fragments of
the hrp gene cluster amplified from strains of Xanthomonas campestris pv.
dieffenbachiae and restricted with HaeIII. Lanes M, phage A restricted with Pstl.
Lanes 1, X422; 2, X757; 3, X790; 4, X1272; 5, X260; 6, X763; 7, X736; 8, X738; 9,
X745; 10, X729. Molecular sizes are given in bases.

89
M123456789 10
Fig. 4-12. Restriction analysis of the 1,075 bp DNA fragments of the hrp gene cluster
amplified from strains of Xanthomonas campestris pv. fici (Lanes 1 to 5) and
Xanthomonas campestris pv. poinsettiicola (Lanes 6 to 10) and restricted with Taql.
Lane M, phage X restricted with Pstl. Lanes 1, X125; 2, XI51; 3, X208; 4, X212; 5,
X702; 6, X87; 7, X202; 8, X352; 9, 071-424; \0,X. campestris pv. vesicatoria 75-3.
Molecular sizes are given in bases.

90
distinct restriction patterns that corresponded to the grouping established previously on
the basis of pathogenic features (Fig. 4-11; Table 4-2). For instance, the strains of X.
campestris pv. citri of the citrus canker groups A, B, and C each produced
characteristic restriction patterns for all the different combinations of fragment-
restriction endonuclease tested (Chapter 6; Table 4-2). Similarly, strains ofX.
campestris pv. vesicatoria groups A and B also produced characteristic restriction
banding patterns (Table 4-2). The two groups of X. campestris pv. vesicatoria could be
differentiated by any combination hrp fragment/restriction endonuclease (Table 4-2).
Strains of X. campestris pv. dieffenbachiae that cause diseases in plants of the family
Araceae also produced characteristic restriction patterns (Fig. 4-11; Table 4-2). Among
the strains of X. campestris pv. dieffenbachiae, three RFLP groups were established on
the basis of restriction analysis of the hrp related fragments that corresponded to the
host of origin. The strains X422, X790, XI51, and XI272 were isolated from
Anthurium sp., strains X260 and X763 were isolated from Syngonium sp., and strains
X736, X738, X739, and X745 were isolated from Xanthosoma sigittifolium (Table 4-
2). Further, the strains from Anthurium sp. could be distinguished from the X.
sigittifolium strains only by comparison of the 1,075-bp fragment restricted with the
endonuclease Sau3A\ (Table 4-2).
The strains of X campestris pv. fici and X. campestris pv. poinsettiicola each
were divided in two different groups based on the distinct restriction banding patterns
for both /irp-related fragments (Fig. 4-12; Table 4-2). Further, sequence variability was
also observed within the groups, particularly for the X. campestris pv. fici group A
(Table 4-2). However, the banding patterns obtained for the three strains of this group
were very similar and they have several bands in common (Fig. 4-12; Table 4-2). The
most striking feature of these two pathovars is the similarity of the restriction banding

91
patterns between X. campestris pv. fici group A and X. campestris pv. poinsettiicola
group A, and between X. campestris pv. fici group B and X. campestris pv.
poinsettiicola group B (Fig. 4-12; Table 4-2). Further, the restriction pattern for both
fragments amplified from strain X125 of A campestris pv. fici group A were identical
to the ones of the strains of X. campestris pv. poinsettiicola group A for all
combinations of hrp-related fragments and restriction endonucleases (Fig. 4-12 and
Table 4-2).
Discussion
Several methods have been examined for the differentiation of plant pathogenic
xanthomonads with different degrees of success. In the present study, the technique of
enzymatic amplification and analysis of specific regions of the bacterial genome was
evaluated for the differentiation and identification of plant pathogenic xanthomonads.
Unlike other studies where random and less stable regions of the bacterial genome were
used (Garde and Bender, 1991; Gilbertson et al., 1989; Hartung, 1992; Hartung et ah,
1993; Lazo and Gabriel, 1987; Lazo et ah, 1987), I analyzed sequences of the bacterial
genome related to the hrp gene cluster of X campestris pv. vesicatoria. Although this
region of the bacterial chromosome seems to be highly conserved among the plant
pathogenic xanthomonads, nonpathogenic xanthomonads lack similarity to the hrp
genes (Bonas et ah, 1991; Stall and Minsavage, 1990). This is certainly a major
advantage, because the nonpathogenic xanthomonads is of concern for plant health
inspection in certification programs (Gitaitis et ah, 1987, 1992). From the results
obtained in the present study the DNA amplification of /zr/?-related sequences is highly
promising for specific differentiation and identification of a large group of plant

92
pathogenic xanthomonads. DNA sequences related to the hrp genes were specifically
amplified from bacterial strains representing X. fragariae and at least 28 pathovars of
X. campestris. However, the hrp-related sequences were not amplified from strains of
the plant pathogenic xanthomonads X. albilineans, X. campestris pv. celebensis, X.
campestris pv. secalis, and X. campestris pv. translucens, and from a few members of
the pathovars X. campestris pv. fici, X. campestris pv. pelargonii, X campestris pv.
phaseoli, X. campestris pv. poinsettiicola, and X. campestris pv. pruni. Strains of these
xanthomonads may not be highly related genetically to the other pathovars ofX.
campestris included in the study as determined by DNA homology studies (Hildebrand
et al., 1990; Palleroni et al., 1993; Vauterin et al., 1993) and by the failure to hybridize
strongly to the hrp clones ofX. campestris pv. vesicatoria (Chapter 3; Bonas et al.,
1991; Stall and Minsavage, 1990). These results support the contention of the presence
of differences in the DNA sequences of Xanthomonas corresponding to one or both
primers used.
Although, no size variation was observed for the /zrp-related fragments
amplified from different plant pathogenic xanthomonads, the restriction analysis of
these fragments revealed the presence of sequence variation. Fragment length
polymorphisms in the hrp regions were further explored for differentiation of the
different groups of plant pathogenic xanthomonads. The restriction banding profile
established for the two hrp-related fragments digested with four endonucleases ranged
from 7 to 25 different profiles though only a very few groups of xanthomonads, i.e. X.
campestris pv. holcicola, X campestris pv. vesicatoria group B, and A! fragariae,
produced unique restriction banding profiles (Table 4-2). This was somewhat expected
because these three groups of xanthomonads seem to be genetically unique (Hildebrand
et al., 1990; Palleroni et al., 1993; Vauterin et al., 1993). For example, A! campestris

93
pv. vesicatoria group B is genetically and phenotypically very distinct from X.
campestris pv. vesicatoria group A, although both groups are presently lumped into the
same pathovar and cause similar disease symptoms on solanaceous plants (Stall et al.,
1994). These two groups of strains have different genetic backgrounds and they are
only about 33% similar on the basis of DNA homology (Stall et al., 1994). On the
other hand, the majority of the plant pathogenic xanthomonads produced restriction
banding patterns that were shared by strains of different groups or pathovars ofX.
campestris. Furthermore, different restriction profiles may have DNA bands in
common. The presence of similarities in the restriction banding patterns of strains of
different taxa strongly supports the presence of some degree of genetic relatedness
among them.
Several pathovars of X. campestris were determined to be genetically very
closely related to each other on the basis of DNA homology (Hildebrand et al., 1990;
Palleroni et al., 1993; Vauterin et al., 1993) and RFLP analysis (Graham et al., 1990;
Gottwald et al., 1991; Lazo et al., 1987; Verdier et al., 1993). For example, the
pathovars X campestris pv. alfalfae, X campestris pv. begoniae, X. campestris pv.
cassavae, X. campestris pv. citri, X. campestris pv. citrumelo, X campestris pv.
dieffenbachiae, X campestris pv. glycines, X. campestris pv. malvacearum, X
campestris pv. manihotis, X. campestris pv. phaseoli, X campestris pv. phaseoli
"fuscans", X. campestris pv. poinsettiicola group A, X. campestris pv. vesicatoria group
A, and X campestris pv. vignicola belong to the DNA homology group VIII
established by Vauterin et al. (1993) whereas the pathovars X. campestris pv.
armoraciae, X campestris pv. campestris, X. campestris pv. incanae, and X. campestris
pv. raphani belong to the group XII (Vauterin et al., 1993). Similar grouping was
obtained in the phylogenetic analysis of the restriction fragment data of the hrp-related

94
DNA sequences (Chapter 5). However, the restriction analysis of the two hrp
fragments with four restriction endonucleases consistently allowed the differentiation
of almost all the pathovars and groups of X. campestris, even for those genetically
closely related (Table 4-2). In a few cases strains of different groups of plant
pathogenic xanthomonads could not be distinguished by restriction analysis of hrp-
related sequence. These include a strain of X. campestris pv. incanae and one of X.
campestris pv. carotae which were identical to strains of X. campestris pv. campestris;
a strain of X campestris pv. phaseoli which was identical to strains ofX. campestris pv
glycines group A; strains F59 and F86 of X. campestris pv. citrumelo which were
indistinguishable from strains ofY campestris pv. pruni; strain X125 ofX. campestris
pv. fici group A which was identical to strains of X. campestris pv. poinsettiicola group
A; strain XV2 ofX. campestris pv. vitians which was identical to strains ofX.
campestris pv. malvacearum; and strain X52 isolated from Hibiscus sp. was
indistinguishable from strains ofX. campestris pv. poinsettiicola group B.
Strains of a few taxa of the plant pathogenic xanthomonads were homogeneous
on the basis of restriction analysis of the hrp-related sequences. Strains of X. fragariae
produced identical and almost unique restriction patterns for all combinations of hrp-
related fragment and restriction endonuclease. The genetic analysis of the hrp-related
sequences suggests that strains of X. fragariae may be formed by a clonal population
even though the strains were isolated from plant materials from different geographic
locations. The uniformity of the X. fragariae population has also been supported by
analysis of fatty acid composition (Yang et al., 1993) and by SDS-PAGE of proteins
(Vauterin et al., 1991a). Similarly, strains of some pathovars ofX. campestris were
also highly homogeneous with regard to the restriction banding patterns of the Arp-
related fragments, i.e. X. campestris pv. begoniae, X campestris pv. campestris, X.

95
campestris pv. malvacearum, and X campestris pv. pelargonii. The population
structure of X. campestris pv. begoniae and X. campestris pv. pelargonii has been
examined quite extensively by using different methods, such as DNA-DNA
hybridization, fatty acid composition, and SDS-PAGE of proteins. Both pathovars
seem to consist of uniform groups of strains with stable phenotypic features and
characteristic fatty acid and protein profiles (Vauterin et al., 1990a; Yang et ah, 1993).
In the same way, X campestris pv. malvacearum also seems to be a fairly
homogeneous group of strains on the basis of SDS-PAGE of proteins and fatty acid
analysis (Vauterin et ah, 1991a; Yang et ah, 1993), despite the fact that this pathovar
comprises a large number of physiological races (Brinkerhoff, 1970). Although the
restriction analyses of the hrp-related fragments also suggest for a clonal population
structure in other pathovars of X. campestris, i.e. X. campestris pv. armoraciae, X.
campestris pv. gardneri, X. campestris pv. pruni, and X campestris pv. vignicola, the
limited number of strains tested does not allow us to make conclusions with certainty.
In contrast, a striking feature of the different pathovars of X. campestris was the
high degree of genetic variability in the hrp-related sequences for strains within
pathovars. For example, strains ofX. campestris pv. citri were separated in two
homogeneous groups on the basis of the restriction endonuclease analysis of the hrp-
related sequences. One group comprises all strains that corresponds to the citrus canker
A of.Y campestris pv. citri whereas the other group includes strains of the citrus canker
B and C forms. Although the citrus canker groups were established on the basis of
pathogenic specialization of the strains (Leite, 1990; Stall and Civerolo, 1991) more
extensive studies on the genetic and phenotypic characteristics also support this
grouping of the strains ofX. campestris pv. citri (Egel et al., 1991; Gabriel et al., 1988,
1989; Graham et al., 1990; Hartung and Civerolo, 1987). Strains ofV campestris pv.

96
dieffenbachiae also have a high degree of genetic variability. Strains of X. campestris
pv. dieffenbachiae produced very similar restriction patterns for all combinations of hrp
fragments-restriction endonucleases and they were divided in three different groups of
restriction banding profiles. Further, the strains were accurately identified to their host
of origin based on the restriction analysis of the hrp related fragments. The genetic
heterogeneity of the strains of X. campestris pv. dieffenbachiae agrees with the
diversity determined on the basis of pathological and physiological features, fatty acid
composition, and SDS-PAGE of proteins (Chase et al., 1992; Vauterin et al., 1991a).
Moreover, strains of X campestris pv. dieffenbachiae that cause disease on Syngonium
sp. have been designated A', campestris pv. syngonii (Dickey and Zumoff, 1987)
although strong support has not been obtained yet to place these strains into a different
pathovar (Chase et al., 1992).
In the restriction analysis of the hrp-related sequences, the most heterogeneous
pathovar was2f campestris pv. citrumelo. The seventeen strains of X campestris pv.
citrumelo were separated into nine different groups on the basis of the restriction
banding profiles (Table 4-2). However, the strains of the highly aggressive group ofX.
campestris pv. citrumelo were homogeneous and comprised a single group (Egel et al.,
1991). The high uniformity of the strains of the highly aggressive group ofX.
campestris pv. citrumelo has also been determined in studies of DNA homology, RFLP
analyses, fatty acid composition, and SDS-PAGE of proteins (Chapter 3; Egel et al.,
1991; Gabriel et al., 1988, 1989; Graham et al., 1990; Hartung and Civerolo, 1987,
1989; Vauterin et al., 1991b). On the other hand, the moderately and weakly
aggressive strains of X. campestris pv. citrumelo are very diverse, and they are likely to
determine the heterogeneity of this pathovar. The diverse nature of the strains of X.

97
campestris pv. citrumelo has also been reported previously (Egel et al., 1991; Gabriel
et al., 1988, 1989; Graham et al., 1990; Hartung and Civerolo, 1987, 1989).
The restriction banding profiles of some strains o X. campestris pv. citrumelo
were very similar or even identical to the profiles of other pathovars oX. campestris.
For instance the strains F59 and F86 X. campestris pv. citrumelo have profiles for both
/irp-related fragments identical to strains of X campestris pv. pruni whereas other
strains of X. campestris pv. citrumelo were closely related to strains of X. campestris
pv. fici A, X. campestris pv. poinsettiicola A, and X campestris pv. vesicatoria A
(Table 4-2). The close genetic relatedness of strains of X. campestris pv. citrumelo to
other pathovars of X. campestris has also been determined previously (Egel et al.,
1991; Gabriel et al., 1988, 1989; Graham et al., 1990; Hartung and Civerolo, 1987,
1989).
The X. campestris pathovars, fici, poinsettiicola, vesicatoria, and vitians seem to
comprise of a very diverse group of bacteria (Stall et al., 1994; Vauterin et al., 1991a;
Yang et al., 1993). Our results on the genetic analysis of the hrp-re\a{ed sequences also
corroborate the existence of distinct groups of strains within these pathovars. Further,
the grouping obtained based on the analysis of the /wy?-related sequences agrees very
closely with the grouping established previously on the basis of genetic or phenotypic
features or both. For example, strains of X. campestris pv. vesicatoria were grouped
into two distinct groups based on the restriction banding profile of the hrp-related
sequences (Table 4-2). The groups established based on the /?r/?-related analysis are
genetically highly uniform and they correspond to the previous existing groups A and
B oiX. campestris pv. vesicatoria (Stall et al., 1994). On the other hand, the pathovars
fici and poinsettiicola have a more complex picture. Strains of both pathovars were
also divided into two distinct RFLP groups. Further, the group A oiX. campestris pv.

98
fici is genetically closely related to group A of X. campestris pv. poinsettiicola whereas
the B group ofX. campestris pv. fici is also genetically closely related to group B ofX.
campestris pv. poinsettiicola. The diversity within the pathovars X. campestris pv. fici
and X. campestris pv. poinsettiicola has also been supported by fatty acid analysis
(Nancy C. Hodge, personal communication).
The genetic diversity of strains within pathovars of X. campestris is not
surprising. The classification of plant pathogenic bacteria at pathovar level was not
based initially on the genetics or other intrinsic characteristics of the organism but
rather on the host from which the bacteria were isolated (Bradbury, 1984; Dye et al.,
1980). Furthermore, comprehensive studies on the genetics and phenotypic
characteristics have supported the existence of a high degree of diversity among strains
of a given group of plant pathogenic xanthomonads that cause diseases in the same
host. Therefore, the plant pathogenic xanthomonads comprise a very complex group of
bacteria that might not be easily distinguished. The genetic analysis of the hrp-related
sequences seems to be a very useful tool for differentiation of pathovars and pathogenic
groups of plant pathogenic xanthomonads. Furthermore, the diversity or uniformity of
the different taxa of xanthomonads assessed on the basis of restriction analysis of hrp-
related sequences apparently agrees very closely with the existing groups established
by using other methods. However, the restriction banding profiles generated for the
/^-related fragments may be an easier and more discriminating approach for
identification of plant pathogenic xanthomonads, compared to other methods such as
genomic fingerprinting or RFLP analysis by using random or specific DNA probes.
DNA fingerprinting by digestion of the entire bacterial genome with restriction
endonuclease usually produces very complex patterns that are difficult to interpret
(Graham and Cooksey, 1989; Hartung and Civerolo, 1987; Vauterin et al., 1993)

whereas RFLP analysis using DNA probes requires hybridization techniques. More
extensive work is necessary, however, to characterize the hrp sequence variation in
other groups of plant pathogenic xanthomonads. Furthermore, DNA-DNA
hybridization data is also necessary to determine the consistency of the grouping
established based on such a small region of the bacterial genome as the hrp gene
cluster. The restriction data may also be useful to establish the genetic evolutionary
relationship of the hrp genes among the different plant pathogenic xanthomonads.

CHAPTER 5
PHYLOGENETIC ANALYSIS PLANT PATHOGENIC Xanthomonas
BASED ON DNA SEQUENCES RELATED TO THE hrp GENES
The genus Xanthomonas Dowson 1939 includes bacteria that occur worldwide
and cause economically important diseases on many plants. The host range of the
xanthomonads spans over 392 mono and dicotyledonous plant species (Hayward, 1993;
Leyns et al., 1984). X. campestris is certainly the most complex species among the
xanthomonads. This species is highly diverse and comprises at least 125 different
pathovars (Bradbury, 1984; Dye et al., 1980; Hayward, 1993). The classification of the
plant pathogenic xanthomonads has been based largely on the capability of the bacterial
strain to cause a characteristic disease (Bradbury, 1984; Dye et al., 1980; Vauterin et
al., 1990a; Young et al., 1992). Consequently, bacteria with features different from
pathogenicity may be classified under the same taxonomic unit. On the other hand,
genetically similar bacteria may be placed in different taxa because they cause diseases
on different plants.
The establishment of relationships among the different pathovars and species of
Xanthomonas has been attempted by using different approaches, such as metabolic,
fatty acid, and protein profiling (Chase et al., 1992; Hildebrand et al., 1993; Hodge et
al., 1992; Van den Mooter and Swings, 1990; Vauterin et al., 1991b), and nucleic acid
analyses (Hildebrand et al., 1990; Palleroni et al., 1993; Vauterin et al., 1993).
Although the pathovars of X. campestris have almost identical biochemical and
physiological characteristics (Van den Mooter and Swings, 1990), diversity seems to
100

101
exist among the strains of X. campestris as demonstrated by protein and fatty acid
profiles (Vauterin et al., 1991a; Yang et al., 1993). Further, genetic analyses of the
xanthomonads by DNA-DNA hybridization has revealed a diverse genetic background
(Hildebrand et al., 1990; Palleroni et al., 1993; Vauterin et al., 1992).
A characteristic feature of plant pathogenic xanthomonads is the presence of a
genomic region that contains genes {hrp) for the hypersensitive and pathogenic
reactions on plants. This hrp gene cluster is required for plant pathogens to cause
disease on susceptible hosts and hypersensitive reaction (HR) on resistant, or on
nonhost plants (Willis et al., 1991). Besides the xanthomonads, the hrp genes have
also been found in other Gram negative plant pathogenic bacteria of the genera Erwinia
(Beer et al., 1991) and Pseudomonas (Boucher et al., 1987; Huang et al., 1990;
Lindgren et al., 1986). Furthermore, the hrp genes seem to be highly conserved among
different bacteria at the structural and functional levels (Fenselau et al., 1992; Gough et
al., 1992; Hwang et al., 1992). The hrp genes of plant pathogenic bacteria are also
similar at the protein level to virulence determinants of bacteria of the genus Yersinia,
pathogens of animals (Fenselau et al., 1992; Gough et al., 1992). On the other hand,
nonpathogenic bacteria that are unable to cause disease or HR on plants lack DNA
similar to the hrp genes sequence (Chapter 3; Bonas et al., 1991; Lindgren et al., 1986;
Stall and Minsavage, 1990).
Conservation of the hrp genes at the structural and functional level among
different plant pathogenic xanthomonads is well documented. The hrp genes are
organized in a large cluster in the bacterial genome (Arlat et al., 1991; Bonas et al.,
1991) although an additional small hrp region unrelated to the major hrp genes has also
been identified (Kamoun and Kado, 1990; Kamoun et al., 1992). Southern
hybridization analyses using the hrp genes of Xanthomonas campestris pv. vesicatoria

102
as probes revealed the presence of sequence similarity in the genomic DNA of several
pathovars of X. campestris and related species of Xanthomonas (Chapter 3; Bonas et
al., 1991; Stall and Minsavage, 1990). Further, the hrp genes of the xanthomonads
seem to hybridize to genomic DNA of P. solanacearum when low stringency
conditions are used (Arlat et al., 1991; Boucher et al., 1987). The hrp gene clusters of
the xanthomonads are also functionally interchangeable. Heterologous
complementation of the hrp genes has been achieved for several pathovars ofX.
campestris (Chapter 3; Arlat et al., 1991; Bonas et al., 1991). Despite all the
information accumulated regarding the hrp genes, little is known about the
evolutionary relationship among the hrp genes of the xanthomonads.
Inferences of the relationships of hrp genes of the plant pathogenic
xanthomonads based on molecular phylogeny would certainly improve the
understanding of the genetics of pathogenicity of these plant pathogens. Dissimilar hrp
gene clusters in strains with relatively divergent genetic background might indicate that
the hrp region has coevolved with the rest of the genome from a common bacterial
ancestor. On the contrary, similar hrp gene clusters in strains with relatively divergent
genetic background might indicate a more recent horizontal genetic movement between
bacterial strains. Alternatively, variability in the genetic relatedness between bacterial
strains for different regions of the hrp genes may support the hypothesis of different
origins of the hrp genes. In this study, I examined the evolutionary relationship of two
DNA sequences of the hrp genes from different plant pathogenic xanthomonads. DNA
fragments of the bacterial genome related to the hrpB and hrpC/D complementation
groups of the hrp gene cluster of X campestris pv. vesicatoria (Chapter 3; Bonas et al.,
1991) were enzymatically amplified from different plant pathogenic xanthomonads and
then digested with frequent-cutting restriction endonucleases. The restriction fragment

103
data were used to establish the genetic relationship of the hrp genes between strains and
to infer the phylogeny of this region of the bacterial genome for different plant
pathogenic xanthomonads.
Materials and Methods
Bacterial strains and culture conditions
The bacterial strains used in this study and their sources are listed in appendix
A. The identity of the bacterial strains used here was confirmed by fatty acid analysis
(N. C. Hodge, personal communication). Strains of A! campestris were grown on
nutrient agar (Becton Dickinson, Cockeysville, MD) at 28C, with the exception of
strains of X. campestris pv. citri group B that were grown on a sucrose based medium
(Canteros de Echenique et al., 1985). Strains of A! fragariae were cultivated on
Wilbrink's medium (Koike, 1965).
DNA manipulations
Total genomic DNA of each strain was isolated by phenol-chloroform
extraction and ethanol precipitation essentially as described by Ausubel et al. (1987).
The restriction endonuclease analyses of the DNA fragments amplified from different
bacterial strains were accomplished by using the frequently cutting endonucleases Cfol,
Haelll, Sau3 Al, and Taql under the conditions specified by the manufacturer
(Promega, Madison, WI). The restricted fragments were resolved by electrophoresis in
4% agarose gels (3% NuSieve and 1% SeaKem GTG [FMC BioProducts, Rockland,
ME]) in TAE buffer (40 mM Tris acetate, 1 mM EDTA, pH 8.2) at 8 V/cm.

104
DNA amplification
Two sets of oligonucleotide primers selected from the nucleotide sequence of
the hrp gene cluster of X. campestris pv. vesicatoria were used in this study. Primers
RST2 plus RST3 delineated an 840-bp fragment and RST21 plus RST22 delineated a
1,075-bp fragment of the complementation groups hrpB and hrpUD ofX. campestris
pv. vesicatoria, respectively (Chapter 3). The primers were used in polymerase chain
reaction for specific amplification of homologous /ir/?-related DNA fragments from
different plant pathogenic xanthomonads. The reaction conditions and polymerase
chain reaction cycles were previously described (Chapter 3).
Data analysis
The electrophoretic patterns of the /zrp-related DNA fragments restricted with
each of the four endonucleases were used to determine the restriction banding profile
for each bacterial strain. The codes 0 and 1 were assigned according to the absence or
presence of each DNA band, respectively. The genetic relationship between strains
was estimated based on the resulting matrix by determining the proportion of shared
DNA fragments (F). The equation proposed by Nei and Li (1979), F = 2/ (nx + ny),
where nxy is the number of fragments shared by both strains, and nx and ny are the total
number of fragments for each strain, was used to estimate the proportion of shared
fragments (F) by using a computer program (Appendix B) written for the SAS system
(SAS Institute Inc., Cary, NC). The genetic divergence between strains was
determined as the estimate of the number of nucleotide substitutions per site (5), on the
basis of the proportion of shared DNA fragments (Nei and Li, 1979). The number of
nucleotide substitutions per site (8) was calculated based on the iterative method

105
suggested by Nei and Li (1979). A program written for the SAS system was used for
the calculations (Appendix C).
The evolutionary relationship among strains was examined by using programs
from the computer package PHYLIP (Felsenstein, 1991, 1993). Phylogenetic trees
were inferred for each of the two hrp regions individually by using a parsimony
criterion and a distance matrix method. The restriction fragment data encoded 0 or 1
were used as input for reconstruction of an unrooted phylogenetic tree by the Wagner
parsimony criterion of the BOOT program (Felsenstein, 1991). The strains G-23 ofY
campestris pv. holcicola and XV56 of X campestris pv. vesicatoria group B were taken
as the outgroups to infer the topology of the phylogenetic trees for the hrpC/D and
hrpB regions, respectively. No assumptions were made for the ancestral character state
and the confidence intervals of the estimates of the inferred phylogenetic trees were
determined by analyzing a total of 100 bootstrap samples (Felsenstein, 1985, 1991).
The distance matrix method unweight pair-group method with arithmetic mean
(UPGMA) of phylogenetic reconstruction (Nei, 1987) was also applied to the data for
reconstruction of a rooted phylogenetic tree. The option UPGMA of the program
NEIGHBOR (Felsenstein, 1993) was performed for this analysis. The input data
consisted of the estimates of the number of nucleotide substitutions per site (5) between
strains determined by the method developed by Nei and Li (1979). For both methods,
the data of each strain were entered in a random order.
Results
DNA fragments related to the hrpB and hrpC/D complementation groups of the
hrp genes ofX. campestris pv. vesicatoria were amplified from strains representing the

106
species A! fragariae and 29 pathovars of X. campestris included in the study (data not
shown). However, the amplification of the hrpB-re\ated fragment from X. campestris
pv. carotae, X. campestris pv. gardneri, X. campestris pv. glycines B, X. campestris pv.
papavericola, X. campestris pv. pelargonii, X. campestris pv. vitians C, and a strain of
X. campestris from Hibiscus sp. usually produced low yield of DNA whereas no
amplification of this fragment was obtained from X campestris pv. holcicola. These
strains were not included in the phylogenetic analysis of the hrpB region. The hrp-
related fragments amplified with each set of oligonucleotide primers from different
plant pathogenic xanthomonads were of identical size.
Unrooted parsimony analyses were carried out on 100 and 63 restriction
fragment data sets produced by the digestion of the hrpC/D- and hrpB-related
fragments, respectively, with four different frequent-cutting restriction endonucleases.
The phylogenetic trees inferred for each amplified fragment are very similar in
topology to each other (Fig. 5-1 and 5-2). Ten major clades were identified in the
analysis of the hrpC/D-related fragment amplified from the plant pathogenic
xanthomonads (Fig. 5-1). Clades 1, 2, 4, 9, and 10 include only a single taxon whereas
the remaining clades are comprised by taxa representing different pathovars of X.
campestris (Fig. 5-1 and 5-2). Further, the members of each clade can be easily
identified based on the restriction banding profile of the hrpC/D-related fragment
produced with the endonuclease Taql (Table 5-1). The result of the phylogenetic
analysis of the hrpB region is consistent with the phylogeny determined for the hrpC/D
region with a few important differences. The clades 3 and 4 which contain pathovars
of2f. campestris that cause diseases on brassica plants merged in a larger single clade
whereas the single member of clade 9, X. campestris pv. dieffenbachiae B, were placed
in clade 7 (Fig. 5-2). Another difference is the shift of the strains of X. campestris pv.

Fig. 5-1. Unrooted phylogenetic tree inferred from restriction analysis data of the
1,075-bp DNA fragment related to the hrpC/D complementation group of the hrp genes
of Xanthomonas campestris pv. vesicatoria generated by the Wagner parsimony
criterion. The values on each node indicate the levels of support derived from 100
bootstrapped trees. The shaded boxes delineate the major clades identified in the
analysis of the hrpC/D-related DNA fragment. Also included are the number of strains
examined for each taxa (numbers in parenthesis).

108
pv. holcicola (I)
-pv. vesicatoria Bill)
96
X. campestris X52 (1)
pv. pomsettiicola B (3)
pv. glycines B (I)
pv. fici B (2)
pv. raphani B in
pv. raphani A (3)
pv. papavericola (1)
pv. pelargonii (7)
pv. carotae B (I)
pv. campestris (9)
pv. armoraciae (3)
pv. incanae (1)
pv. gardneri (3)
pv. taraxaci (1)
. vitinns C (2)
carotae A (6)
pruni (4)
citrurnelo
pv. citrurnelo iI)
pv. citrurnelo (1)
pv. alfalfae (I)
pv. alfalfae (I)
pv. poinsettiicola A (2)
X. campestris X198
pv. fici A (I)
pv. citrurnelo (2)
v. citrurnelo (2),
pv. vesicatoria A
pv. vesicatoria A
pv. vesicatoria A (24)
pv. citrurnelo ill
pv. citrurnelo (I)
pv. citrurnelo (I)
pv. fici A
pv. fici A .
pv. citrurnelo (5)
99
(2)
81
ill
pv. vitians B (1)
pv. malvacearum (9)
pv. phaseoli B (1)
pv. glycines A (6)
pv. citri A (6)
phaseoli Tuscans" (I)
pv. vignicola (3)
pv. citri C (6)
pv. citri B (7)
X. campestris XCF (1)
bilvae (1^
91
. pv. dieffenbachiae A (4)
pv. djefTenbachiae A (4)
pv. phaseoli A (4)
pv. vitians A (3)
pv. manihotis (1)
pv. begoniae (I)
pv. maculifoliigardeniae (I)
pv. begoniae (1)
pv. begoniae (7)
pv. dieffenbachiae B (2)

Fig. 5-2. Unrooted phylogenetic tree inferred from restriction analysis data of the 840-
bp DNA fragment related to the hrpB complementation group of the hrp genes of
Xanthomonas campestris pv. vesicatoria generated by the Wagner parsimony criterion.
The values on each node indicate the levels of support derived from 100 bootstrapped
trees. The shaded boxes delineate the major clades identified in the analysis of the
hrpC/D-rdated DNA fragment. Also included are the number of strains examined for
each taxa (numbers in parenthesis).

110
100
pv. vesicatoria B (11)
84
42
68
94
34
74
84
88
26r
54rl
pv. malvacearum (9)
I pv. vitians B (1)
i pv. phaseoli B (1)
1 pv. glycines A (6)
pv. citri A (o)
pv. citri C (6)
pv. citri B (7)
X. campestris XCF (1)
pv. bilvae (1)
64
14
42
84
64
60 36~TT_
44f~l
pv. phaseoli Tuscans" (1)
pv. citrumelo (1)
58| pv. alfalfae (1)
pv. alfalfae (1)
_ pv. citrumelo (2)
pv. citrumelo (5)
pv. citrumelo (2)
6_6r
20
pv. fici A (1)
pv. citrumelo (1)
pv. fici A (1)
4
K
18
5
24
r
14
4l
20
44r
20r
54
XOl
34~E?p~~~
78
78
72
80
pv. poinsettiicola A (2)
pv. pruni (4)
pv. fici A (1)
pv. physalidicola (1)
pv. vesicatoria A (11
pv. vesicatoria A (24)
pv. vesicatoria A (2)
pv. citrumelo (1)
pv. citrumelo (1)
pv. citrumelo (1)
pv. citrumelo (2)
_ pv. dieffenbachiae B (2)
_ pv. vignicola (3)
X. campestris X198 (1)
pv. maculifoliigardeniae (1)
pv. manihotis (1)
pv. vitians A (3)
pv. phaseoli A (4)
pv. dieffenbachiae A (4)
pv. dieffenbachiae A (4)
pv. begoniae (1)
pv. begoniae (7)
pv. begoniae (1)
68
40
80
pv. poinsettiicola B (3)
pv. fici B (2)
- X. fragariae (9)
96 | pv. armoraciae (3)
I pv. raphani B (1)
32
pv. campestris (9)
pv. incanae (1)
pv. raphani A (3)

Ill
vignicola from clade 6 to clade 7 in the phylogenetic analyses of the hrpC/D and hrpB
regions, respectively (Fig. 5-1 and 5-2).
The results of the bootstrapped resampling analyses presented in the branches
of the trees largely support the monophyletic nature of the major clades in almost all
the cases (Fig. 5-1 and 5-2). Further, the topology of these trees reconstructed by the
parsimony procedure is consistent with the phylogenetic analyses based on the distance
matrix approaches UPGMA (Fig. 5-3 and 5-4) and neighbor-joining (not shown).
Further, the UPGMA analyses also indicate that the strains G-23 ofX. campes tris pv.
holcicola and XV56 of X. campestris pv. vesicatoria group B taken to infer the
topology of the unrooted parsimony trees for the hrpC/D and hrpB regions,
respectively, were appropriate outgroups because they were basal to the remainder of
Xanthomonas spp. included in this study (Fig. 5-3 and 5-4).
The present phylogenetic framework supports the monophyletic nature of the
hrp gene sequences of plant pathogenic xanthomonads that belongs to different
pathovars of X campestris and cause diseases in different hosts. For instance, clade 6
comprises the pathovars X. campestris pv. citri, X campestris pv. bilvae, X campestris
pv. glycines A, X. campestris pv. malvacearum, X campestris pv. phaseoli B, X.
campestris pv. phaseoli "fuscans", and A", campestris pv. vitians B (Fig. 5-1 and 5-2).
Although these pathogens cause diseases on different plants, there is strong agreement
in the parsimony analyses to indicate a common bacterial ancestor for both hrp genes
regions of these pathovars. The bootstrap resampling analyses support the branching of
this clade at 78% and 94% for the hrpC/D and hrpB regions, respectively (Fig. 5-1 and
5-2). The estimates of the similarity between hrp genes for each region also indicates a
close genetic relatedness for the members within this clade. The estimates of the
similarity of the hrp genes region within clade 6 ranged from 0.65 to 1.00 and from

Fig. 5-3. Rooted phylogenetic tree inferred from restriction analysis data of the 1,075-
bp DNA fragment related to the hrpC/D complementation group of the hrp genes of
Xanthomonas campestris pv. vesicatoria generated by the method unweight pair-group
method with arithmetic mean (UPGMA). The shaded boxes delineate the major clades
identified in the analysis of the hrpC/D-related DNA fragment. Also included are the
number of strains examined for each taxa (numbers in parenthesis).

113
rd
tj
c1
£
pv. cilri C (6)
pv. citri B (7)
pv. vignicola (3)
pv. phaseoli Tuscans" (1)
i pv. malvacearum (9)
pv. vilians B (I)
pv. phaseoli B (1)
pv. glycines A (6)
pv. cilri A (6)
a. campestris XCF(I)
pv. bilvae (1)
pv. vesicatoria A
pv. vesicatoria A
pv. citmmelo (I)
pv. alfalfac I
pv. alfalfac (I)
pv. citruinelo (2)
pv. pruni (4)
pv. citrumclo (I)
pv. citrumclo (2)
pv. citrumelo (2)
pv. poinsettiicola A (2)
_ pv. lici
I
(24)
r
pv. fici A (I j
X. campestris XI98 (I)
.pv, citrumelo(I)
_Ppv. citrumelo (IJ
LpV. vesicatoria
pv. fici A (I)
pv. fici A (I)
pv. citrumelo (6)
pv. citrumelo (I)
pv. dielTenbachiae B (2)
pv. maculifblgardemae (I)
pv. begoniae (I)
pv. begoniae (I)
pv. begoniae (7)
pv. dieffenbachiae A (4)
pv. dielTenbachiae A (4)
pv. phaseoli A (4)
pv. vitians A (3)
pv. manihotis (I)
pv.incanae(1)
pv. armoraceae (3)
pv. campestris (9)
pv. carotae B (1)
-pv. pelargonii (7)
pv. papavericola (I)
I f pv. gardneri (3)
[ *-pv. vitians C (2)
j pv. carotae A (7)
' pv. taraxaci (J)
pv. vesicatoria B (11)
pv. raphani A (3)
pv. raphani B (1)
ipv. poinsettiicola B
A-, campestris
ipv. glycines B (1
Hpv. fici B (2)
_'V fragariae (9)
pv. holcicoia (I)
X52
I I I I I I I I I
0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0.00
Genetic distance

Fig. 5-4. Rooted phylogenetic tree inferred from restriction analysis data of the 840-bp
DNA fragment related to the hrpB complementation group of the hrp genes of
Xanthomonas campestris pv. vesicatoria generated by the method unweight pair-group
method with arithmetic mean (UPGMA). The shaded boxes delineate the major clades
identified in the analysis of the hrpC/D-related DNA fragment. Also included are the
number of strains examined for each taxa (numbers in parenthesis).

115
pv. vesicatoria A (1)
pv. vesicatoria A (24)
pv. citrumelo (1)
pv. citrumelo (I)
pv. poinsettiicola A (2)
pv. fici A (1)
pv. citrumelo (2)
pv. pruni(4)
pv. physalidicola (1)
pv. vesicatoria (2)
pv. citrumelo (1)
pv. vignicola (3)
pv. citrumelo (I)
pv. fici A (I)
pv. citrumelo (2)
pv. alfalfae (I)
pv. alfalfae (I)
pv. alfalfae (I)
pv. fici A (1)
pv. citrumelo (2)
pv. citrumelo (5)
X. campestris X198 (1)
pv. diefienbachiae B (2)
pv. begoniae (7)
pv. begoniae (1)
pv. begoniae (1)
pv. diefienbachiae A (4)
pv. phaseoli A (4)
pv. vitians A (3)
pv. dieffienbachiae A (4)
pv. manihotis (1)
pv. maculifoliigardeniae (I)
pv. phaseoli 'Tuscans" (1)
X. campestris XCF (I)
pv. bilvae (1)
pv. citri B (7)
pv. citri A (6)
pv. phaseoli B (1)
pv. glycines A (6)
pv. malvacearum (9)
pv. vitians B (1)
pv. citri A (6)
pv. campestris (9)
pv. raphani A (3)
pv.incanae (I)
pv.raphaniB(1)
pv. armoraceae (3)
pv. poinsettiicola B (3)
pv. fici B (2)
X. fragariae (9)
pv. vesicatoria B (11)
I I I I I I I I
0.035 0.030 0.025 0.020 0.015 0.010 0.005 0.0
Genetic distance

116
0.53 to 1.00 for the hrpC/D and hrpB regions, respectively (Table 5-1). Similarly, the
clades 3, 5, 7, and 8 also contain pathovars ofX. campestris that cause diseases on
different plants but are genetically very closely related regarding the hrp genes for both
regions examined (Table 5-1). The bootstrap resampling analyses further substantiate
the monophyletic nature of the gene cluster for these clades (Fig. 5-1 and 5-2).
On the contrary, the inferred phylogenetic analysis also indicates that strains
which are classified under the same pathovar, because they cause similar diseases on
the same host plants, may have very distinct hrp sequences. X. campestris pv.
vesicatoria that causes disease on solanaceous plants is known to contain at least two
diverse groups of strains (Stall et al., 1994). The two groups of this pathovar also show
distinct hrp gene sequences (Fig. 5-1 and 5-2). Strains ofX. campestris pv. vesicatoria
B comprise the sole taxon of the clade 1 whereas the strains of X. campestris pv.
vesicatoria A belong to the clade 7 (Fig. 5-1 and 5-2). The clade 7 is the largest one
and includes several pathovars ofX. campestris (Fig. 5-1 and 5-2). The placement of
the two groups of X. campestris pv. vesicatoria in two distant and distinct clades are
further supported by the bootstrap resampling analysis (Fig. 5-1 and 5-2). The
estimated similarity between the two groups of X. campestris pv. vesicatoria in relation
to the hrp regions ranged from 0.33 to 0.38 and from 0.50 to 0.51 for the hrpC/D and
hrpB regions, respectively. In addition, the three subgroups identified \nX. campestris
pv. vesicatoria A are monophyletic and genetically closely related for both hrp gene
regions (Fig. 5-1 and 5-2). Several other pathovars of X. campestris accommodate
groups of strains with very diverse hrp genes, i.e. X campestris pv. fici, X campestris
pv. glycines, X. campestris pv. phaseoli, X. campestris pv. poinsettiicola, andX.
campestris pv. vitians (Fig. 5-1 and 5-2; Table 5-2).

117
Table 5-1. Within clade similarity values generated by comparison of the restriction
profiles obtained by digestion of the /zrp-related fragments amplified from plant
pathogenic Xanthomonas spp. with restriction endonuclease enzymes.
Clade3
Taql
groupb
Number of taxa in the
cladec
Similarity within clade
hrpB
hrpC/D
hrpB
hrpC/D
1
1
1
1


2
2
1
1


3
3
}5
10
0.96-1.00de
0.78-1.0
4
5
2
0.96
1.00
5
5
2
4
0.87f
0.77-1.00
6
6 and 7
10
11
053-1.00
0.65-1.00
7
8
22
20
0.77-0.78
0.67-1.00
8
9
9
9
0.84-1.00
0.64-0.96
9
10
1
1


10
11
na§
1


aClade established on the basis of the phylogenetic analysis of the restriction fragment
data of the 1,075-bp fragment related to the hrpC/D region of the hrp genes of X.
campestris pv. vesicatoria.
bGroup established based on the restriction profile of the 1,075-bp fragment related to
the hrpC/D region of the hrp genes of X. campestris pv. vesicatoria restricted with the
endonuclease Taq\ (see Chapter 4).
cNumber of taxa determined based on the phylogenetic analysis of each hrp region
individually.
dValues are the similarities estimated by using the equation proposed by Nei and Li
(1979) for each /zrp-related fragment restricted with either endonuclease Cfol, HaelW,
Sau3Al, and Taql.
eOnly three taxa were compared.
fOnly two taxa were compared.
8 na, not applicable.

Table 5-2. Similarity values between clades of plant pathogenic Xanthomonas spp. generated based on the restriction
profiles of the DNA fragments related to the hrpB and hrpC/D complementation groups of the hrp genes oiX. campestris pv.
vesicatoria.
Clade3
1
2
3
4
5
6
7
8
9
10
1
0.25b
0.50-0.62
0.41
0.32-0.42
0.42-0.50
0.33-0.46
0.25-0.44
0.46
0.38
2
0.40

0.36-0.46
0.32
0.33-0.36
0.25-0.33
0.17-0.25
0.08-0.22
0.14
0.21
3
0.54-0.60
0.52-0.57

0.41-0.50
0.39-0.51
0.24-0.46
0.23-0.40
0.18-0.47
0.25-0.40
0.27-0.32
4
0.54-0.60
0.52-0.57
0.96-1.00

0.40-0.51
0.29-0.46
0.27-0.41
0.27-0.43
0.40
0.23
5
0.43-0.53
0.62-0.63
0.55-0.68
0.55-0.68

0.23-0.37
0.23-0.37
0.14-0.34
0.20-0.26
0.18-0.23
6
0.47-0.56
0.37-0.48
0.51-0.58
0.51-0.58
0.34-0.52

0.40-0.74
0.18-0.44
0.32-0.50
0.13-0.18
7
0.38-0.51
0.24-0.41
0.41-0.57
0.41-0.57
0.39-0.56
0.41-0.94

0.36-0.67
0.49-0.74
0.13-0.21
8
0.53-0.59
0.50-0.56
0.49-0.59
0.49-0.59
0.54-0.62
0.54-0.79
0.59-0.77

0.50-0.75
0.23
9
0.52
0.28
0.53-0.58
0.53-0.58
0.34-0.43
0.48-0.78
0.65-0.85
0.25-0.62

0.13-0.22
10
ndc
nd
nd
nd
nd
nd
nd
nd
nd

aClade established based on the restriction profile on the basis of the phylogenetic analysis of the restriction fragment data of
the 1,075-bp fragment related to the hrpC/D region of the hrp genes of X. campestris pv. vesicatoria, restricted with the endonuclease Taql.
^Values are the similarities estimates by using the equation proposed by Nei and Li (1979) for the 1,075-bp DNA fragment related to
the hrpC/D, upper triangular matrix, and for the 840 bp DNA fragment related to the hrpB, lower triangular matrix, restricted with either
endonuclease Cfol, HaeIII, Saw3AI, and Taql.
cnd, not determined.

119
Another evolutionary relationship revealed by the phylogenetic analysis of the
hrp genes of the xanthomonads is the variability in the relatedness between strains
depending on the hrp region examined. For example, X. campestris pv. vignicola
clustered in clade 6 in the phylogenetic analysis for the hrpC/D region (Fig. 5-1), and
the value of 78% for the bootstrap resampling supports the monophyletic nature of the
hrp region for this clade (Fig. 5-1). Further, similarity indices for the different
pathovars within this clade ranged from 0.67 to 1.00 (Tables 5-1). In comparison, the
similarity of X. campestris pv. vignicola to the members of clade 7 ranged from 0.42 to
0.69 (Table 5-2 and 5-3). On the other hand, the strains X. campestris pv. vignicola
were placed into clade 7 in the phylogenetic analysis of the hrpB region (Fig. 5-2), and
this was strongly supported by the bootstrap resampling with a value of 84% of the
bootstrapped trees (Fig. 5-2). The genetic relatedness of2f campestris pv. vignicola to
the other members of the clade 7 ranged from 0.76 to 0.94 for the hrpB region whereas
the similarity of this region between X. campestris pv. vignicola strains and the
members of the clade 6 ranged from 0.53 to 0.61 (Table 5-3).
Strains of2f campestris pv. raphani also comprise an unique case. The
phylogenetic analysis of the hrpC/D-rdaied region of X campestris pv. raphani
revealed that the two closely related groups were placed in clade 4 (Fig. 5-1). The
bootstrap resampling analysis also supports 100% the monophyletic nature of this clade
(Fig. 5-1). The genetic similarity of the strains in clade 4 to strains of clade 3 which
contains the other brassica pathogens i.e. X campestris pv. armoraciae, X. campestris
pv. campestris, and X. campestris pv. incanae ranged from 0.41 to 0.50 (Table 5-2). In
contrast, the analysis of the hrpB region showed that the hrp region for the strains of X.
campestris pv. raphani and strains of the other three pathovars of X campestris are

120
Table 5-3. Similarity values between Xanthomonas campestris pv. vignicola and
selected members of the clades 6 and 7 generated by comparison of the endonuclease
profiles of the DNA fragments related to the hrpB and hrpC/D complementation
groups of the hrp genes of X. campestris pv. vesicatoria.
Similarity of the hrp-related sequences of
X. campestris pv. vignicola
hrpB
hrpC/D
Clade 6a
X. campestris
pv. bilvae
0.61b
0.71
pv. citri A
0.53
0.75
pv. glycines A
0.58
0.75
pv. malvacearum
0.59
0.67
pv. phaseoli "fuscans"
0.58
0.83
pv. vitians B
0.59
0.67
Clade 7
X. campestris
pv. alfalfae
0.80
0.50
pv. citrumelo
0.87
0.53
pv. fici A
0.94
0.46
pv. poinsettiicola A
0.94
0.43
pv. pruni
0.94
0.46
pv. vesicatoria A
0.94
0.46
aClade established on the basis of the phylogenetic analysis of the restriction fragment
data of the 1,075-bp fragment related to the hrpC/D region of the hrp genes ofX.
campestris pv. vesicatoria.
bValues are the similarities estimated by the equation proposed by Nei and Li (1979)
for each /irp-related fragment restricted with either endonuclease Cfol, Haelll, Sau3M,
and Taql.

121
highly related and they are monophyletic in the phylogenetic analysis with bootstrap
resampling value of 80% (Fig. 5-2). Further, the similarity index between members of
the clades 3 and 4 ranged from 0.96 to 1.00 for the hrpB region (Table 5-2).
Discussion
The phylogenetic analysis of the hrp genes of the xanthomonads has revealed a
diverse evolutionary relationship for this region of the bacterial genome of these plant
pathogens. The hypothesis of coevolution of the hrp region with the rest of the genome
from a common bacterial ancestor is supported by comparison of either closely or
distantly related plant pathogenic xanthomonads. For instance, the similarity of the
two hrp regions examined in this study for the groups A and B ofX. campestris pv.
vesicatoria was less than 0.51. This genetic divergence determined for these two
groups of X. campestris pv. vesicatoria is very similar to the values obtained when the
entire genome was compared on the basis of DNA homology (Stall et al., 1994).
Although these pathogens cause similar diseases on solanaceous plants, their hrp genes
are genetically diverse at about the same extend as the rest of the genome. Moreover,
these comparisons do not support the contention that there was a higher selective
pressure for the hrp sequences than for other regions of the genome nor that a
horizontal movement of the hrp region of the genome may have occurred between
strains of the two groups of X. campestris pv. vesicatoria. Despite the genetic distance,
the hrp genes of these two groups of X. campestris pv. vesicatoria are functionally
complementary (Chapter 3). Similarly, the citrus pathogens X. campestris pv. citri and
X campestris pv. citrumelo also cause diseases on the same hosts though they are
genetically distinct (Egel et al., 1991; Gabriel et al., 1989; Vauterin et al., 1991a). The

122
analysis of the hrp genes revealed the same level of relatedness found when the entire
genome were compared. Whereas the estimates of the similarity for the two regions of
the hrp genes ranged from 0.41 to 0.61, the DNA-DNA hybridization relatedness
ranged from 0.55 to 0.63 (Egel et al., 1991). In contrast, strains of A campestris pv.
citri canker A are genetically highly related to strains of X. campestris pv.
malvacearum based on DNA-DNA hybridization studies (Egel et al., 1991). The
analysis of the hrp genes also revealed a high similarity between these two groups of
plant pathogenic xanthomonads with values ranging from 0.80 to 0.88 for both regions
of the /^-related sequences examined (Appendix D). Further, these two groups o X.
campestris are monophyletic in regard to the evolution of the hrp genes (Fig. 5-1 and
5-2).
Additional evidence to support the coevolution of the hrp genes and the rest of
the bacterial genome comes from the pathovars X. campestris pv. carotae, X.
campestris pv. gardneri, and X. campestris pv. pelargonii. Although these
xanthomonads have very distinct host ranges, they are genetically closely related based
on DNA-DNA hybridization (Hildebrand et al., 1990; Palleroni et al., 1993). The
phylogenetic analysis of the hrp-related sequences also supports the contention that
these pathovars of X. campestris are monophyletic and genetically closely related in
regard to the hrp genes. The examples discussed above not only point to a coevolution
of the hrp genes and the rest of the genome from a common bacterial ancestor for
certain plant pathogenic xanthomonads, but also substantiate the divergence between
hrp genes and host speciation.
Another major finding of this study is the indication of lateral movement of the
hrp genes between plant pathogenic xanthomonads. This hypothesis is supported by
two lines of evidences, the presence of similar hrp gene cluster sequences in strains

123
with divergent genetic background and the variability in the relatedness between strains
depending on the hrp region examined. Strains of X. campestris pv. vesicatoria A
belong to the same clade as the strains of X. campestris pv. citrumelo for both hrp
genes examined (Fig. 5-1 and 5-2). Furthermore, the estimated genetic relatedness
between strains of these two pathovars for the hrp genes regions ranged from about
0.71 to 0.96 (Appendix D), although these strains cause diseases on different hosts. On
the contrary, DNA-DNA hybridization studies indicated that these two pathovars are
not genetically closely related based on comparisons of the entire bacterial genome
which was compared (Egel, 1991). In fact, the relatedness between X. campestris pv.
citrumelo and X campestris pv. vesicatoria was about 0.58 which is close to the
similarity determined between strains of X. campestris pv. citrumelo and X. campestris
pv. citri discussed before (Egel, 1991; Egel et al., 1991). X. campestris pv. citrumelo
has been reported occurring only in Florida (Schoulties et al., 1987). Although the
origin of these strains is not very clear yet (Schoulties et al., 1987), the strains of this
pathovar may have evolved from an endemic flora of xanthomonads that is present in
Florida (Hartung and Civerolo, 1989). The analysis of the hrp genes of the strains of X.
campestris pv. citrumelo revealed some degree of variability, but there is strong
support for a monophyletic relationship among them. Furthermore, they are also
closely related and monophyletic in regard to both regions of the hrp genes to strains of
other pathovars of X campestris, i.e. X. campestris pv. alfalfae, X. campestris pv. fici,
X campestris pv. poinsettiicola, X. campestris pv. pruni, and X. campestris pv.
vesicatoria A, and the strain XI98 of campestris, which are also endemic to the same
geographic area. A more extensive study of the genetic relationships among these
pathogens also including nonpathogenic xanthomonads endemic to Florida may

124
provide some important information on the origin of the strains of X. campestris pv.
citrumelo.
Stronger evidence that supports the contention of lateral movement of the hrp
genes among the plant pathogenic xanthomonads comes from the variability in the
relatedness between strains for the two distinct hrp regions examined in this study. For
instance, the hrpC/D region of X. campestris pv. vignicola is phylogenetically
monophyletic and closely related to the homologous hrp region ofX. campestris pv.
bilvae, X campestris pv. citri, X campestris pv. glycines A, X. campestris pv.
malvacearum, X campestris pv. phaseoli B, X campestris pv. phaseoli "fuscans", and
X. campestris pv. vitians B (Fig. 5-2). The hrpB region of campestris pv. vignicola
is also monophyletic, but it is closely related to the homologous hrp region ofX.
campestris pv. begoniae, X. campestris pv. dieffenbachiae, X campestris pv.
maculifoliigardeniae, X. campestris pv. manihotis, and X. campestris pv. phaseoli A
(Fig. 5-1 and 5-2). In a similar way, the hrpC/D-relaied region ofX. campestris pv.
raphani was only 0.50 similar to the homologous region \nX. campestris pv.
campestris, X. campestris pv. armoraciae, and X. campestris pv. incanae (Table 5-2),
whereas the /ir/?5-related region was more than 0.90 similar for the strains of these two
groups (Table 5-2). Moreover, functional heterologous complementation of the hrp
genes has been demonstrated for different xanthomonads (Chapter 3; Arlat et al., 1991;
Bonas et al., 1991). Since there is no genetic and functional support regarding
selection pressure in relation to the hrp genes among the plant pathogenic
xanthomonads, the most likely hypothesis to explain the source of variability for these
two regions of the bacterial genome of X. campestris pv. vignicola remains in the
origin of the two hrp regions from distinct ancestors. The variability in the genetic
relatedness for these two regions of the hrp genes of X. campestris pv. vignicola

125
substantiate the contention that the hrp genes in some of the plant pathogenic
xanthomonads may have evolved from distinct ancestors through lateral movement of
genetic material.
An important conclusion from the phylogenetic studies is the divergence
between hrp genes and host specificity. The hrp genes are essential for the
development of disease on compatible hosts and hypersensitive reaction on both
resistant host and nonhost plants (Willis et al., 1991). Previous work has demonstrated
the functional conservation of the hrp genes and the lack of host speciation (Chapter 3;
Arlat et al. 1991; Bonas et al., 1991). Although the hrp genes are necessary in the
plant-pathogen interaction, other factors in the bacterial pathogen are likely to be
involved in host speciation (Fenselau et al., 1992; Gough et al., 1992). Our results
support the coevolution of the hrp genes with the rest of the bacterial genome through a
common bacterial ancestor. Also, there are bases to support the hypothesis of
horizontal movement of the hrp genes. Close relationship between plant pathogenic
xanthomonads with different genetic background and variability in the relatedness of
different regions of the hrp genes indicate distinct origins for different regions of the
hrp genes, instead of evolution from a single ancestor. Since the lateral movement of
genetic material between bacteria is a common and important mechanism in bacterial
evolution (Krawiec and Riley, 1990), the coexistence in the same biological niche may
have provided opportunities for the lateral transfer of the hrp region of the bacterial
genome between xanthomonads. The hrp are functionally conserved among the
xanthomonads (Chapter 3; Bonas et al., 1991; Fenselau et al., 1992), this also supports
the absence of functional selectivity. The resemblance at the protein level of the hrp
genes of the xanthomonads with genes involved in the secretion of pathogenicity
factors in genetically distant organisms such as the animal pathogens of the genus

126
Yersinia (Fenselau et al., 1992; Gough et al., 1992) is also intriguing. If this means a
convergent functional evolution or a common bacterial ancestry remains to be clarified,
Furthermore, the phylogeny of the hrp genes of plant pathogenic xanthomonads may
provide a framework and a rational basis through which origins and differentiation of
the xanthomonads may be assessed.

CHAPTER 6
GENETIC ANALYSIS OF hrp RELATED DNA SEQUENCES OF
Xanthomonas campestris STRAINS CAUSING DISEASES OF
CITRUS
Citrus canker, caused by strains of Xanthomonas campestris pv. citri group A,
represents an important problem for production of citrus worldwide (Civerolo, 1984).
This disease is characterized by raised lesions on leaves, stems, and fruits. Strains of X.
campestris pv. citri group A have a relatively wide host range and cause symptoms of
various degrees in all commercial citrus varieties (Stall and Seymour, 1983). In severe
cases, abscission of fruits and leaves may result (Civerolo, 1984; Stall and Seymour,
1983). Other xanthomonads that cause similar symptoms on citrus are strains of
groups B and C of X. campestris pv. citri. They are of less importance than strains of
X. campestris pv. citri group A and have comparatively limited host ranges. Citrus
bacterial spot is another bacterial disease of citrus caused by a xanthomonad, and
symptoms are similar to citrus canker with a few important differences (Schoulties et
al., 1987). The pathogen, referred to as X campestris pv. citrumelo (Gabriel et al.,
1989), causes flat, watersoaked lesions in young leaves. Strains ofX campestris pv.
citrumelo cause symptoms primarily on trifoliate orange (Poncirus trifoliata) and its
hybrids, such as Swingle citrumelo (Citrus paradisi X P. trifoliata) (Graham and
Gottwald, 1988).
Although X. campestris pv. citri and X campestris pv. citrumelo cause similar
diseases of citrus, there is evidence for differences between these pathovars. In
addition to the pathogenicity differences listed above, X campestris pv. citrumelo
127

128
strains appear to be quite heterogeneous both genetically (Egel, 1991; Egel et al., 1991)
and in aggressiveness (Graham and Gottwald, 1990) compared with A! campestris pv.
citri. This has resulted in questions about the relationship of the bacterial spot
pathogen to other pathovars of X. campestris. Several xanthomonads isolated from
ornamental plants cause lesions similar to bacterial spot when artificially inoculated
onto young citrus plants (Graham and Gottwald, 1991; Graham et al., 1990); they are
also genetically similar to some strains ofX. campestris pv. citrumelo (Egel et al.,
1991; Graham et al., 1990). It was suggested that strains oiX. campestris pv. citrumelo
may represent other pathovars of X. campestris incidentally isolated from citrus
(Graham et al., 1990) or strains of a xanthomonad that has a wide host range (Gabriel et
al., 1988); alternatively it was suggested that the most weakly aggressive strains may
be opportunistic strains which cause symptoms only when associated with injury (Egel,
1991).
These alternatives have not been resolved by studies of the genetics (Egel et al.,
1991; Gabriel et al., 1988, 1989; Graham et al., 1990; Hartung and Civerolo; 1987,
1989; Vauterin et al., 1991b) or pathogenicity (Graham and Gottwald, 1990; Graham et
al., 1990) of these strains. The genetics of pathogenicity, however, might favor one of
the above hypotheses. An excellent candidate for examination is the hypersensitivity
reaction and pathogenicity (hrp) gene cluster responsible for pathogenicity reaction on
susceptible hosts and a hypersensitive reaction on resistant hosts or on nonhosts plants
(Willis et al., 1991). The hrp gene cluster has been characterized in several bacterial
plant pathogens, such as Pseudomonas syringae pv. phaseolicola (Lindgren et al.,
1986), P. solanacearum (Boucher et al., 1987), Erwinia amylovora (Beer et al., 1991),
and A! campestris pv. vesicatoria (Bonas et al., 1991). The hrp gene cluster of X.
campestris pv. vesicatoria consists of at least 25 kb of genomic DNA, and

129
polymorphism of restriction fragments of a homologous DNA sequence occurs among
pathovars of A campestris (Bonas et al., 1991). Opportunistic xanthomonads, which
produce limited symptoms in susceptible hosts and no hypersensitive reaction in
nonhosts, do not possess DNA similar to an hrp gene cluster (Stall and Minsavage,
1990).
The genomic similarity of strains of X. campestris pv. citrumelo has been
investigated (Egel et al., 1991; Gabriel et al., 1988, 1989; Graham et al., 1990; Hartung
and Civerolo; 1987, 1989; Vauterin et al., 1991b), but examination of the similarity
between hrp clusters of these strains adds information on the comparative genetics of
pathogenicity. Similar hrp gene clusters among strains with relatively divergent
genetic backgrounds might indicate a similar origin of pathogenicity, and dissimilar
hrp gene clusters would support the hypothesis that many strains of X. campestris
involved in the citrus bacterial spot disease are diverse. The similarity of the hrp gene
in strains of X. campestris pv. citri and X campestris pv. citrumelo was investigated by
amplifying and restricting two DNA fragments of the hrp complementation groups B
and C/D (Chapter 3; Bonas et al., 1991), which are highly conserved among several
pathovars of A! campestris (U. Bonas, personal communication).
Materials and Methods
Culture conditions
The strains of X. campestris used in this study and their sources are listed in
Appendix A. All strains had previously been identified as members of X. campestris
by fatty acid analysis (N. C. Hodge, personal communication). Citrus bacterial spot
strains were rated for pathogenicity by Graham and Gottwald (1990) and by Graham et

130
al. (1990). All strains were streaked onto nutrient agar (Becton Dickinson,
Cockeysville, MD), and single colonies were selected. Nutrient broth cultures were
grown 24 hours on a rotatory shaker (150 rpm) at 28C. Strains ofX. campestris pv.
citri group B were grown on a sucrose based medium (Canteros de Echenique et al.,
1985). Strains were stored on lima bean agar (Difco, Detroit, MI) for short term
storage and in sterile tap water at room temperature for long term storage.
DNA isolation
The procedure described by Ausubel et al. (1987), with minor modifications,
was used to extract total genomic DNA. Briefly, bacterial cells were pelleted by
centrifuging in an Eppendorf microcentrifuge (Brinkmann Instruments Inc., Westbury,
NY) for 2 min at 16,000 g. The pellet was washed in 1 ml of distilled water, pelleted
again and resuspended in 567 pi of TE buffer (10 mM Tris-Cl, pH 8.0; 1 mM EDTA,
pH 8.0). Proteinase K (Boehringer Mannheim, Indianapolis, IN) and sodium dodecyl
sulfate (SDS) (Sigma, St. Louis, MO) were added for a final concentration of 100 p
g/ml and 0.5%, respectively. After incubation for 1 hour at 37C, sodium chloride and
hexadecyltrimethyl ammonium bromide (Sigma) were added to each preparation for a
final concentration of 0.7 M and 1%, respectively. The preparations were incubated for
10 min at 65C. DNA was extracted with chloroform-isoamyl alcohol (24:1). The
samples were hand shaken continuously and gently for 10 min and centrifuged for 5
min at 16,000 g. A second extraction was accomplished by adding phenol-chloroform-
isoamyl alcohol (25:24:1) and centrifuging as described above. DNA was precipitated
by adding 0.6 volumes of isopropanol and incubating for 30 min at -20C. The
samples were centrifuged for 20 min at 16,000 g. The DNA pellet obtained was
washed with 1 ml of 70% ethanol and centrifuged again. The DNA was dried under

131
vacuum for 20 min and then the pellet was redissolved in 100 pi of TE and stored at
4C.
DNA amplification
The two sets of oligonucleotide primers used in this study were designed based
on nucleotide sequences of the hrp gene cluster ofX. campestris pv. vesicatoria (U.
Bonas, personal communication). The two primers RST2 and RST3 delineate an 840-
bp region and primers RST21 and RST22 delineate a 1,075-bp region of the hrp
complementation groups B and C/D of X. campestris pv. vesicatoria, respectively
(Chapter 3). Oligonucleotide primers were synthesized with a model 394 DNA
Synthesizer (Applied Biosystems, Foster City, CA) by the DNA Synthesis Laboratory,
University of Florida, Gainesville.
DNA was amplified in a total volume of 50 pi. The reaction mixture contained
5 pi of 10X buffer (500 mM KC1, 100 mM TrisCl [pH 9.0 at 25C], 1% Triton X-
100), 1.5 mM MgCl2, 200 pM of each deoxynucleotide triphosphate (Boehringer
Mannheim), 25 pmol of each primer, and 2.5 units of Taq polymerase (Promega,
Madison, WI). The amount of template DNA added was 100 ng of purified total
bacterial DNA The reaction mixture was covered with 50 pi of light mineral oil. A
total of 30 amplification cycles were performed in an automated thermocycler (MJ
Research, Watertown, MA). Each cycle consisted of 30 s of denaturation at 95C, 30 s
of annealing at 62C, and 45 s of extension at 72C for the primers RST2 and RST3
and 30 s at 95C, 45 s at 61 C and 1.5 min at 72C, respectively, for the primers
RST21 and RST22. The last extension step was extended to 5 min.
Amplified DNAs were detected by electrophoresis in 0.9% agarose gels in TAE
buffer (40 mM Tris-acetate, 1 mM EDTA, pH 8.2) at 5 V/cm of gel (Sambrook et al.,

132
1989.). After being stained with 0.5 jag of ethidium bromide per ml, the gel was
photographed over a UV transilluminator (Fotodyne Inc., New Berlin, WI) with type
55 Polaroid film (Polaroid, Cambridge, MA).
Hybridization analysis
The identity of the amplified DNA fragments was further confirmed by
hybridization analysis with an internal DNA probe for each fragment. Samples were
electrophoresed in 0.7% agarose gel according to standard procedure (Sambrook et al.,
1989.). The gel was then denatured in 0.4 N NaOH and 0.6 M NaCl for 30 min and
neutralized for 30 min in 0.5 M Tris-Cl and 1.5 M NaCl. The denatured DNA was
transferred by the procedure of Southern (1975) to nylon membrane (Gene Screen Plus,
Du Pont, Boston, MA). Hybridization was carried out at 68C with 0.5X SSC, and
0.1% w/v SDS. The internal probes consisted of a 271-bp insert of the plasmid
pXV840 for the 840-bp fragment and a 335-bp insert of the plasmid pXV1075 for the
1,075-bp fragment (Chapter 3; Appendix A). Probes were labeled by the random
primed (Feinberg and Vogelstein, 1983) incorporation of digoxigenin-labeled
deoxyuridine-triphosphate (DIG-UTP) and detected by the use of the Genius
Nonradioactive DNA Labeling and Detection kit (Boehringer Mannheim) as specified
by the manufacturer.
Restriction endonuclease analysis of amplified DNA
Amplified DNAs were restricted with either endonuclease Cfol, Haelll,
Sau3Al, or Taql under conditions specified by the manufacturer (Promega). The
restriction fragments were separated by electrophoresis in 4% agarose gels (3%
NuSieve GTG and 1% Seakem GTG [FMC BioProducts, Rockland, ME]) in TAE

133
buffer at 8 V/cm. Phage X Pst I restricted DNA fragments were used as molecular
weight standards. After being stained with 0.5 jag of ethidium bromide per ml for 40
min, the gels were destained in 1 mM MgSC>4 for 1 hr and then photographed over a
UV transilluminator with type 55 Polaroid film.
Data analysis
DNA restriction fragment patterns were determined by direct comparison of the
electrophoretic patterns of the DNA restricted with each of the four endonucleases.
The codes 1 or 0 were assigned according to the presence or absence of each fragment,
respectively. The resulting matrix was used to estimate the genetic relationships
between strains based on the proportion of shared DNA fragments. The expected
proportion of shared fragments (F) was calculated by the equation proposed by Nei and
Li (1979), F = 2nxyJ(nx + ny), where nxy is the number of fragments shared between two
strains and nx and ny are the total number of fragments for each strain. A computer
program (Appendix B) written for the SAS system (SAS Institute Inc., Cary, NC) was
used to estimate the proportion of shared fragments (F). The genetic divergence
between strains was calculated as the estimate of the number of nucleotide substitutions
per site (8), based on the proportion of shared DNA fragments (Nei and Li, 1979). The
iterative method proposed by Nei (1987) was used to estimate the number of nucleotide
substitutions per site using a computer program written for the SAS system (Appendix
C).
Relationships among strains were studied based on phylogenetic analysis using
the BOOT and KITSCH programs from the PHYLIP computer package (Felsenstein,
1991). For the BOOT program, the restriction fragment data encoded 0 or 1 were used
as input for reconstruction of an unrooted phylogenetic tree by using the Wagner

134
parsimony criterion. No assumptions were made regarding the ancestral character
state, and the pathovar X. campestris pv. maculifoliigardeniae strain X22j was taken as
the outgroup to infer the topology of the phylogenetic tree. A total of 100 bootstrap
samples were analyzed to determine the confidence intervals of the estimates of the
inferred phylogenetic tree (Felsenstein, 1985, 1991). The KITSCH program was used
to infer a rooted phylogenetic tree by the Fitch-Margoliash method (Felsenstein, 1991;
Fitch, and Margoliash, 1967). The input data consisted of a distance matrix of pairwise
estimates of the number of nucleotide substitutions per site, (5), between strains, and
negative branch was not allowed.
Results
DNA amplification
The 840- and 1,075-bp fragments of the hrp gene cluster in X. campestris pv.
vesicatoria were successfully amplified with primers RST2 plus RST3 (Fig. 6-1) and
RST21 plus RST22 (Fig. 6-2), respectively. The same size fragments were also
successfully amplified from DNA of all strains of the other pathovars of X campestris
(Fig. 6-1 and 6-2). The DNA fragments were also amplified from 16 strains of A!
campestris pv. citrumelo representing the three aggressiveness groups and from 19
strains of A! campestris pv. citri groups A, B, and C without variation in size (data not
shown). The sequence similarity of the two DNA fragments amplified from strains of
different pathovars of X. campestris to the hrp gene cluster of X. campestris pv.
vesicatoria was further confirmed by Southern hybridization analysis. The amplified
DNA fragments of the different strains of X. campestris hybridized to the respective
internal probes specific for both hrp gene cluster fragments (data not shown).

135
Ml 2 3 4 5 6 7 B 9 10 11 12 13
Fig. 6-1. Amplification of the 840-bp fragment of the complementation group B of the
hrp gene cluster of Xanthomonas campestris pv. vesicatoria from strains of X.
campestris. Lanes: M, phage X restricted with Eco RI and Hind III; 1 to 3 strains FI,
F6, and FI 00 of A c. pv. citrumelo, respectively; 4,X c. pv. maculifoliigardeniae
strain X22j; 5, X campestris from Strelitzia reginae XI98; 6,X. c. pv. fici strain X151;
1,X. c. pv. alfalfae strain 82-1; 8, X c. pv. bilvae strain XCB; 9, X campestris from
Feronia sp. strain XCF; 10 to 12, strains 9771, B84, and 339 of A! c. pv. citri,
respectively; \3,X. c. pv. vesicatoria strain 75-3. Molecular sizes are given in bases.

136
M 1 2 3 4 5 6 7 8 9 10 11 12 13
Fig. 6-2. Amplification of the 1,075-bp fragment of the complementation groups C/D
of the hrp gene cluster of Xanthomonas campestris pv. vesicatoria from strains of A
campestris. Lanes: M, phage X restricted with Eco RI and Hind III; 1 to 3 strains FI,
F6, and FI00 of A c. pv. citrumelo, respectively; 4,X c. pv. maculifoliigardeniae
strain X22j; 5, X. campestris from Strelitzia reginae X198; 6, X. c. pv. fici strain X151;
7,X. c. pv. alfalfae strain 82-1; 8, X. c. pv. bilvae strain XCB; 9, X. campestris from
Feronia sp. strain XCF; 10 to 12, strains 9771, B84, and 339 of A c. pv. citri,
respectively; 13, A! c. pv. vesicatoria strain 75-3. Molecular sizes are given in bases.

137
Restriction analysis
The 840- and 1,075-bp hrp gene cluster fragments amplified from strains of
different pathovars of X. campestris were each restricted with either Cfol, HaeIII,
Sau3Al, or Taql. The banding patterns for each set of fragments from the pathovars of
X. campestris included in this study were variable. The banding patterns of the 1,075
bp hrp gene cluster fragment amplified from strains of different pathovars ofX.
campestris restricted with the endonucleases Cfol and Haelll are presented in Fig. 6-3.
The banding patterns of the strains of X. campestris pv. citrumelo were very similar to
the patterns obtained forX. campestris pv. vesicatoria strain 75-3 (Fig. 6-3). Also, X
campestris pv. alfalfa, X. campestris pv. fici, and the strain XI98 of X. campestris
produced patterns similar to the strains of X. campestris pv. citrumelo with certain
combinations of fragment-restriction endonucleases (Fig. 6-3). X campestris pv.
bilvae, X campestris XCF, and the strains of X. campestris pv. citri also made up a
group with very similar banding pattern (Fig. 6-3). On the other hand, X. campestris
pv. maculifoliigardeniae had a more distinct restriction fragment profile (Fig. 6-3).
Although variability was also observed in the banding patterns obtained with the
endonucleases Sau3A\ and Taql, a characteristic pattern for each group or pathovar of
X. campestris was less evident for these two endonucleases (data not shown).
Similarly, restriction analysis of the 840-bp hrp fragment also produced a pattern of
variation for the different strains of X. campestris (data not shown).
Sixteen strains of A. campestris pv. citrumelo, representing all three
aggressiveness groups, were analyzed by restriction analysis of the amplified hrp
fragments. The banding patterns of five strains of the highly aggressive group of X.
campestris pv. citrumelo were identical to each other when restricted with either Cfol
(Fig. 6-4A), Haelll (Fig. 6-4B), Sau3Al or Taql (data not shown). For certain

138
M 1 2 3 4 5 6 7 8 9 10 11 12 13
M 1 2 3 456 78 9 10 11 12 13
Fig. 6-3. Restriction analysis of the 1,075-bp DNA fragment of the complementation
groups C/D of the hrp gene cluster amplified from strains of Xanthomonas campestris
and restricted with the endonucleases (A) HaeIII and (B) Cfol. Lanes: M, phage X
restricted with Pstl; 1 to 3 strains FI, F6, and FI 00 of A! c. pv. citrumelo, respectively;
4, X. c. pv. maculifoliigardeniae strain X22j; 5, X. campestris from Strelitzia reginae
strain XI98; 6, X c. pv. fici strain X151; 7,X c. pv. alfalfae strain 82-1; 8, X c. pv.
bilvae strain XCB; 9, X. campestris from Feronia sp. strain XCF; 10 to 12, strains
9771, B84, and 339 of A! c. pv. citri, respectively; 13, A! c. pv. vesicatoria strain 75-3.
Molecular sizes are given in bases.

139
M 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Fig. 6-4. Restriction analysis of the 1,075-bp DNA fragment of the complementation
groups C/D of the hrp gene cluster amplified from strains of Xanthomonas campestris
pv. citrumelo and restricted with the endonucleases (A) HaelW and (B) Cfol.. Lanes:
M, phage X restricted with the Pst I; 1 to 5, highly aggressive strains FI, F54, F274,
F361, and 3166, respectively; 6 to 11, moderately aggressive strains F6, F228, F311,
F254, F348, and F378; 12 to 16, weakly aggressive strains F59, F86, F100, F306, and
F94; \7,X. c. pv. vesicatoria strain 75-3. Molecular sizes are given in bases.

140
combinations of fragment and endonuclease, the restriction pattern of the highly
aggressive strains were similar to some strains of the moderately or weakly aggressive
groups, these include the 1,075-bp fragment from the weakly aggressive strain FI00
restricted with Cfo\ (Fig. 6-4A) and the fragment from the moderately aggressive strain
F378 restricted with HaeIII (Fig. 6-4B). However, the overall banding patterns of the
combinations of two fragments and four endonucleases for the highly aggressive strains
were different from the patterns obtained for the strains of the moderately and weakly
aggressiveness groups of X. campestris pv. citrumelo. In contrast to the highly
aggressive group, restriction fragment polymorphism was evident for the strains within
the moderately and weakly aggressive groups of X. campestris pv. citrumelo (Fig. 6-4).
The weakly aggressive strains F94, FI00, and F306 had banding patterns almost
identical to A! campestris pv. vesicatoria strain 75-3 (Fig. 6-4). Also, strains of the
highly aggressive group of X. campestris pv. citrumelo had banding patterns similar to
X. campestris pv. fici strain XI51 (Fig. 6-3 and 6-4).
In contrast to the diversity of the moderately and weakly aggressive strains of
X campestris pv. citrumelo, strains of X. campestris pv. citri of the groups A, B, and C
each produced characteristic restriction patterns (Fig. 6-5). The banding patterns of all
strains of group A were identical when restricted with either HaelW (Fig. 6-5), Cfol,
Sau3M, or Taql (data not shown). Similarly, strains of groups B and C were also
homogeneous within each group (Fig. 6-5), as well as between the two groups for the
four endonucleases and the two hrp gene cluster fragments.
Genetic relationships of the hrp from different strains of X campestris
Differences in the number of common restriction fragments from the amplified
DNA of the hrp gene cluster indicated that there is variation in the relatedness of the

141
M 1 2 3 4 5 6 7 89 10 11 12 13 141516171819
Fig. 6-5. Restriction analysis of the 1,075-bp DNA fragment of the complementation
groups C/D of the hrp gene cluster amplified from strains of Xanthomonas campestris
pv. citri restricted with the endonuclease Haelll. Lanes: M, phage X restricted with the
Pstl; 1-6, group A strains 9771, 3340, 9760-2, 3213, Tl, and 115-A; 7 to 13, group B
strains B64, B69, B80, B84, B93, B94 and B148; 14 to 18, group C strains 338, 339,
340, 341, and 342; 19, X. c. pv. vesicatoria strain 75-3. Molecular sizes are given in
bases.

142
hrp genes of the different strains of X. campestris. The genetic divergence between
strains was estimated based on the data for 106 restriction fragments obtained from the
combination of two hrp gene cluster fragments and four endonucleases. A pairwise
matrix of the genetic distances, 5, was calculated for the 18 distinct banding patterns
(Table 6-1). X. campestris pv. citri 339 was included in the genetic analysis as a
representative of group C, although the restriction banding patterns of the hrp
fragments were identical to those of the strains of group B of X. campestris pv. citri.
The largest genetic divergence value was 0.082 nucleotide substitution per site between
X. campestris pv. maculifoliigardeniae X22j and A! campestris pv. citri 9771 of group
A (Table 6-1). However, most of the estimates of nucleotide substitutions per site are
smaller than 0.05, which is considered the upper limit to give accurate estimates of
genetic distance based on restriction fragment data (Nei, 1987).
Strains of X. campestris pv. citrumelo that represent all three aggressiveness
groups exhibited nine different restriction patterns of the hrp fragments that were
divergent from 0.002 to 0.022 nucleotide substitution per site (Table 6-1). Similarly,
strains of A! campestris pv. citri groups A, B, and C showed a low genetic divergence
for the hrp genes, ranging from 0.000 to 0.014 nucleotide substitution per site (Table 6-
1). As mentioned above, the banding patterns of the strains of X. campestris pv. citri
groups B and C were identical to each other for all combinations of hrp gene cluster
fragments and restriction endonucleases. On the other hand, the hrp genes of strains of
X campestris pv. citrumelo were very poorly related to the ones of X campestris pv.
citri, with divergence ranging from 0.050 to 0.064 nucleotide substitution per site
(Table 6-1).

Table 6-1. Genetic divergence matrix for 19 strains of Xanthomonas campestris based on the estimates of the number of nucleotide
substitution per site in two fragments related to the hrp gene cluster X. campestris pv. vesicatoria.
X. campestris pv. citrumelo3
X. campestris pv.
citrib
Other strains of X. campestris
High
Moderate
Weak
A
B
C
82-1
XCB
XCF
X22j
XI98
X151
75-3
FI
F6
F311
F348
534
F86
F94
F100
F306
9771
B84
339
FI
0.020
0.009
0.013
0.010
0.014
0.016
0.022
0.020
0.057
0.055
0.055
0.011
0.043
0.041
0.060
0.019
0.004
0.014
F6

0.009
0.015
0.019
0.012
0.012
0.018
0.010
0.057
0.063
0.063
0.007
0.056
0.057
0.052
0.015
0.018
0.018
F311

0.013
0.012
0.008
0.010
0.014
0.012
0.057
0.055
0.055
0.007
0.043
0.044
0.049
0.011
0.007
0.012
F348

0.010
0.010
0.014
0.016
0.016
0.056
0.061
0.061
0.011
0.045
0.046
0.058
0.017
0.009
0.018
534

0.011
0.017
0.017
0.015
0.051
0.050
0.050
0.014
0.036
0.037
0.046
0.018
0.010
0.015
F86

0.007
0.011
0.009
0.050
0.051
0.051
0.006
0.037
0.038
0.048
0.008
0.010
0.012
F94

0.005
0.002
0.057
0.055
0.055
0.012
0.046
0.044
0.052
0.012
0.014
0.005
F100

0.007
0.064
0.058
0.058
0.016
0.049
0.047
0.052
0.016
0.018
0.011
F306

0.057
0.055
0.055
0.014
0.046
0.044
0.052
0.014
0.016
0.023
9771

0.014
0.014
0.059
0.018
0.015
0.082
0.052
0.051
0.057
B84

0.000
0.061
0.012
0.011
0.080
0.051
0.049
0.055
339

0.061
0.012
0.011
0.080
0.051
0.049
0.055
82-1

0.048
0.049
0.065
0.011
0.011
0.018
XCB

0.002
0.074
0.039
0.038
0.046
XCF

0.075
0.040
0.038
0.044
X22j

0.051
0.060
0.055
XI98

0.017
0.018
X151
-
0.012
75-3

3 High, Moderate, and Weak denote aggressiveness groups.
b A, B, and C denote canker groups.

144
The relatedness of the hrp genes of strains of the citrus pathogens X campestris
pv. citrumelo and X. campestris pv. citri to the hrp gene cluster of some other
pathovars of X. campestris was also investigated. The hrp gene cluster of strains of X.
campestris pv. alfalfae 82-1, X. campestris pv. fici X151, and X. campestris pv.
vesicatoria 75-3 and X campestris XI98 were closely related to hrp genes ofX.
campestris pv. citrumelo, and the genetic divergence ranged from 0.004 to 0.023
nucleotide substitution per site (Table 6-1). However, strains ofX. campestris pv. citri
were much less related to those four strains of X. campestris, with genetic divergence
ranging from 0.049 to 0.061 nucleotide substitutions per site (Table 6-1). The hrp
genes of strains of X. campestris pv. citri were highly related to X campestris pv.
bilvae XCB and X. campestris XCF. The genetic divergence of the hrp genes of X.
campestris pv. citri from the genes of these strains of X. campestris ranged from 0.011
to 0.018 nucleotide substitution per site (Table 6-1). Moreover, X. campestris pv.
maculifoliigardeniae strain X22j has hrp genes not highly related to any of the
xanthomonads from citrus, with a genetic divergence ranging from 0.046 from Y
campestris pv. citrumelo strain 534 to as high as 0.082 from Y. campestris pv. citri
group A (Table 6-1).
The restriction fragment data of the hrp genes encoded 0 or 1 and the distance
matrix (Table 6-1) were used to construct phylogenetic trees based on a parsimony
criterion by using the BOOT program and a distance method by using the KITSCH
program of the PHYLIP computer package (Felsenstein, 1991), respectively. Although
the general topology is slightly different, the phylogenetic trees inferred by using two
different approaches of tree reconstruction showed very similar branching patterns for
the major clades (Fig. 6-6 and 6-7). The branching pattern obtained with the BOOT
program is unrooted, although the strain X. campestris pv. maculifoliigardeniae X22j

Fig. 6-6. Unrooted tree for 19 strains of Xanthomonas campestris inferred from
restriction analysis data of DNA fragments related to the complementation groups B
and C/D of the hrp genes of X. campestris pv. vesicatoria generated by the BOOT
procedure from the PHYLIP computer package by using the Wagner parsimony
criterion. Numbers at each node indicate the bootstrap percentages from 100 samples.
Bootstrap values less than 50 indicate that the assemblage is not well supported by the
data.

146
56
100
100
59
16
15
25
17
54
35
33
pv. citri 9771
98 | pv. citri 339
* pV. citri B84
79 | X. campes iris XCF
pv. bilvae XCB
pv. citrumelo F100
pv. vesicatoria 75-3
pv. citrumelo F306
pv. citrumelo F94
L pv. citrumelo F86
X. campes tris XI98
pv. alfalfae 82-1
pv. citrumelo F6
pv. citrumelo F311
pv. citrumelo F348
87 pv. fici X151
pv. citrumelo FI
pv. citrumleo 534
pv. maculifoliigardeniae X22j
13
441 '
t

Fig. 6-7. Rooted tree for 19 strains of X. campestris inferred from restriction analysis
data of DNA fragments related to the complementation groups B and C/D of the hrp
genes of X. campestris pv. vesicatoria generated by the KITSCH procedure from the
PHYLIP computer package by using the Fitch-Margoliash method.

148
pv. citruinelo F94
pv. citruinelo F306
pv. citrumelo F100
pv. vesicatoria 75-3
pv. citruinelo 534
pv. citruinelo F348
pv. fici X151
pv. citruinelo FI
pv. citrumelo F86
pv. alfalfae 82-1
pv. citrumelo F311
pv. citrumelo F6
X. campes tris X198
X. campestris XCF
pv. bilvae XCB
pv. citri 339
pv. citri B84
pv. citri 9771
pv. maculifoliigardeniae X22j
j
.030 .025 .020 .015 .010 .005 0
Genetic distance

149
was taken as the outgroup to infer the topology of the tree (Fig. 6-6). The strains ofX.
campestris can be divided into three major clades based on the hrp genes, with A"!
campestris pv. maculifoliigardeniae X22j as the sole member of one clade (Fig. 6-6).
The second clade is the largest one and comprises all strains oiX. campestris pv.
citrumelo, X. campestris pv. alfalfae, X. campestris pv. fici, X. campestris pv.
vesicatoria, andX. campestris from S. reginae (Fig. 6-6). The third clade is formed by
all the strains of X. campestris pv. citri, X. campestris pv. bilvae, and a strain of X.
campestris from Feronia sp. (Fig. 6-6). The assemblage of these phylogenetic clades is
highly supported by the bootstrap values (Fig. 6-6). Furthermore, X. campestris pv.
maculifoliigardeniae X22j was indeed chosen as an appropriate outgroup to evaluate
the relationships among X. campestris strains causing disease on citrus on the basis of
the hrp genes. The rooted tree obtained by the KITSCH program, selected among 1908
trees examined, also indicates that X. campestris pv. maculifoliigardeniae X22j is basal
to the remainder^ campestris strains included in this study (Fig. 6-7).
The inferred phylogenetic trees seem to support the hypothesis that the hrp gene
cluster of X. campestris pv. citrumelo strains from the three aggressiveness groups are
monophyletic and closely related to other pathovars of X. campestris, including X.
campestris pv. alfalfae, X. campestris pv. fici, and X. campestris pv. vesicatoria.
Similarly, the monophyly of the hrp genes of X. campestris pv. citri, X. campestris pv.
bilvae, andX. campestris XCF is also supported (Fig. 6-6 and 6-7).
Discussion
Strains of all pathovars of X. campestris included in this study have an hrp gene
cluster on the basis of hybridization of genomic DNA with the hrp gene cluster from X.

150
campestris pv. vesicatoria (Egel, 1991) and on the basis of amplification of hrp
fragments with oligonucleotide primers specific for the complementation groups B and
C/D of the hrp genes of X. campestris pv. vesicatoria (Chapter 3). This is particularly
significant in regard to the strains associated with citrus bacterial spot disease. Despite
differences in the pathogenic characteristics of those strains (Graham and Gottwald,
1990; Graham et al., 1990), the presence of an hrp gene cluster supports the pathogenic
nature of those bacterial strains. If they were opportunistic xanthomonads (Gitaitis et
al., 1987), they would lack an hrp region (Bonas et al., 1991; Stall and Minsavage,
1990).
Information about the similarity of the hrp genes of the bacteria causing
diseases of citrus was obtained from restriction enzyme patterns of amplified fragments
of the hrp gene cluster. Although the DNA fragments amplified with the two sets of
primers were of the same size for all the strains of X. campestris, characteristic
restriction banding patterns for each bacterial group occurred with the 840- and 1,075-
bp fragments. Complementation groups B and C/D of the hrp gene cluster, from which
the fragments were amplified, are considered to be highly conserved among plant
pathogenic xanthomonads (U. Bonas, personal communication). Therefore, the
homology of the restriction enzyme fragments from amplified hrp genes should furnish
valid relationships among these pathogens. These relationships were determined by a
phylogenetic analysis.
The phylogenetic analysis based on the hrp gene cluster showed polyphyletic
relationships of the strains of X campestris causing disease in citrus. This suggests
that the hrp gene cluster may have evolved independently in these strains of X.
campestris. This evolution could be convergent or parallel. The analysis is based on
the assumption that the differences in restriction sites in the hrp gene cluster region

151
were due to nucleotide substitutions and not to insertion or deletion of DNA sequences.
In fact, this assumption is supported by the fact that no apparent length variation was
observed in the two DNA fragments amplified from all strains ofX. campestris. In
support of the phylogenetic analysis presented here is the monophyletic nature of the
hrp gene cluster of bacterial pathogens with different genetic backgrounds causing
disease in different hosts. The phylogenetic grouping presented here also correlates
with genetic analyses based on DNA-DNA hybridization, fingerprinting, and
conventional restriction fragment length polymorphism (Egel, 1991; Egel et al., 1991;
Gabriel et al., 1988, 1989; Graham et al., 1990; Hartung and Civerolo; 1987, 1989;
Kubicek et al., 1989). The data presented supports the concept of a group of causal
microorganisms of citrus bacterial spot disease which are closely related yet represent a
variety of different genotypes.
Two different groups ofX. campestris pv. citri were distinguished by restriction
enzyme analysis of the amplified fragments of the hrp gene cluster. The strains of
group A produced identical patterns for all fragment-endonuclease combinations.
Similarly, all strains within the groups B and C also produced identical restriction
banding patterns. The banding patterns for group A strains were different from the
patterns of group B and C strains, however. This substantiates other reports of the
relative genetic uniformity of the strains of the A, B, and C groups of X campestris pv.
citri based on restriction analysis of the entire genome by using genomic fingerprinting,
pulse-field electrophoresis, or restriction fragment length polymorphisms with random
DNA probes (Egel, 1991; Egel et al., 1991; Gabriel et al., 1988, 1989; Graham et al.,
1990; Hartung and Civerolo; 1987). The two groups of X campestris pv. citri have
about 60% DNA homology (Egel et al., 1991). The major difference from previous
work is that in our studies specific homologous regions of the bacterial genome, the hrp

152
gene cluster, were compared. Nevertheless, the results of the genetic analysis of the
hrp related regions are consistent with those obtained when the entire genome was
randomly examined. From this and the previously cited research, the two groups of
strains within X. campestris pv. citri should probably be differentiated at some
taxonomic level.
Two groups of X campestris pv. citrumelo were also distinguished after
restriction enzyme digestion of the amplified hrp fragments. Strains of the highly
aggressive group were very uniform for all fragment-endonuclease combinations and
had a characteristic restriction banding pattern. On the other hand, the moderately and
weakly aggressive groups of the bacterium had diverse restriction banding patterns for
both amplified hrp fragments. This concurs with previous studies of the moderately
and weakly aggressive groups of X. campestris pv. citrumelo (Egel, 1991; Egel et al.,
1991; Gabriel et ah, 1988, 1989; Hartung and Civerolo, 1987, 1989; Kubicek et ah,
1989). Even with genetic diversity among the moderately and weakly aggressive
strains, all strains included under X. campestris pv. citrumelo are 80% similar using
DNA-DNA hybridization (Egel, 1991; Egel et ah, 1991). As with strains ofX.
campestris pv. citri, the two groups of strains of X. campestris pv. citrumelo should
also be distinguished at the taxonomic level.
The results presented in this work demonstrate that strains of X. campestris
causing disease in citrus can be reliably differentiated and identified by restriction
analysis of amplified fragments related to the hrp gene cluster. The number of strains
examined were small, but the genetic diversity within different pathovars of X.
campestris could be assessed. The reliable identification of the citrus pathogen using
DNA amplification will greatly facilitate disease diagnosis, as well as ecological and

153
epidemiological studies. Furthermore, the use of oligonucleotide primers for the hrp
gene cluster region provide certainty for identification of strains X. campestris that may
not be possible with other methods.

CHAPTER 7
EVALUATION OF A DNA AMPLIFICATION APPROACH FOR
DETECTION AND IDENTIFICATION OF PLANT PATHOGENIC
XANTHOMONADS ASSOCIATED WITH PEPPER AND TOMATO
SEEDS
Bacterial spot caused by X. campestris pv. vesicatoria is one of the most
important diseases for pepper and tomato productions worldwide (Jones et al., 1991;
Stall, 1993). The disease is characterized by necrotic lesions on leaves, stems, and
fruits. In warm and rainy weather, the disease is usually severe and may cause heavy
defoliation of the plants that results in reduced yield (Pohronezny et al., 1986).
Furthermore, diseased fruits are usually commercially depreciated and may not be
suitable for fresh-market (Cox, 1966; Pohronezny et al., 1983). Bacterial spot is also of
major concern in the certification program in the transplant industry in southern USA
where producers are expected to provide disease-free transplants (Gitaitis et al., 1987,
1992).
Control of bacterial spot has been hampered by many factors. No bactericide is
completely effective to control the disease and tolerance to bactericides, such as
streptomycin and copper compounds, is widespread among strains of X campestris pv.
vesicatoria (Marco and Stall, 1983; Thayer and Stall, 1961). This has lead to reduced
efficacy of bactericide sprays for the control of the disease in the field. Sources of
resistance to bacterial spot were identified in pepper (Sowell and Dempsey, 1977) and
tomato (Scott and Jones, 1989) and have been used in breeding programs to develop
horticulturally desirable pepper and tomato cultivars. However, sudden shifts in races
ofX campestris pv. vesicatoria can overcome the resistanceand poses a major concern
154

155
for the use of such resistance commercially (Pohronezny et al., 1992). The causal
bacterium may spread from diseased to healthy plants by wind- blown water, clipping
of plants, and aerosols (Gitaitis et al., 1992; Jones et al., 1991; Volcani, 1969).
Furthermore, the bacterium may overwinter on tomato volunteers and diseased plant
debris (Jones et al., 1986). Nevertheless, exclusion of the pathogen from pepper and
tomato growing areas is still one of the main control measures. Consequently, the use
of pathogen free seeds and transplants has become an important part of the strategy for
control of bacterial spot (Gitaitis et al., 1992; Jones et al., 1991).
X. campestris pv. vesicatoria has been reported to be in association with tomato
and pepper seeds (Bashan et al., 1982; Gardner and Kendrick, 1921, 1923; Jones et al.,
1986; Sharon et al., 1982), and symptoms of bacterial spot were observed in seedlings
from contaminated seed lots (Gardner and Kendrick, 1923; Higgis, 1922). However,
the detection and identification of X. campestris pv. vesicatoria associated with seeds is
still a problem with the methods available. The semiselective media developed
specifically for X. campestris pv. vesicatoria (Gitaitis et al., 1991; McGuire et al., 1986;
Sijam et al., 1991) may still support the growth of strains of other pathovars of ^
campestris as well as nonpathogenic xanthomonads and saprophytes associated with
seeds. These bacteria may have cultural characteristics similar to X. campestris pv.
vesicatoria (Gitaitis et al., 1987, 1991; McGuire and Jones, 1989; McGuire et al., 1986;
Sijam et al., 1991). Therefore, additional biological, biochemical, or physiological
tests are required to determine with certainty the identity of the xanthomonads
recovered (Gitaitis et al., 1991; McGuire and Jones, 1989). Serological tests with
monoclonal and polyclonal antibodies have also been examined for specific
identification of plant pathogenic xanthomonads, including X. campestris pv.
vesicatoria (Benedict et al., 1990; O'Brien et al., 1967). However, the heterogeneous

156
nature of antigenic determinants in immunogens of the xanthomonads poses a major
limitation for the use of polyclonal antibodies. For instance, polyclonal antibodies
developed for strains of X. campestris pv. vesicatoria cross-reacted with different plant
pathogenic xanthomonads (O'Brien et al., 1967). In contrast, strains ofX. campestris
pv. vesicatoria are serologically diverse and no single polyclonal or monoclonal
antibody has been obtained that reacts to all strains of this pathogen (Benedict et ah,
1990; Jones et ah, 1993b; Stall et ah, 1994). More recently, nucleic acid based
techniques have been examined for specific detection of X. campestris pv. vesicatoria
(Garde and Bender, 1991; Jones et ah, 1993b). Garde and Bender (1991) developed
DNA probes for detection of copper resistant X. campestris pv. vesicatoria, but the
copper region from X. campestris pv. vesicatoria may hybridize to other plant
pathogens, such as the tomato pathogen Pseudomonas syringae pv. tomato (Voloudakis
etah, 1993).
The objective of this study was to examine a procedure based on the
amplification of DNA fragments related to the hrp gene cluster of X. campestris pv.
vesicatoria (Chapter 3) for the specific detection and identification of plant pathogenic
xanthomonads associated with pepper and tomato seeds. The sensitivity and specificity
of the polymerase chain reaction (PCR) technique were determined by the detection of
X. campestris pv. vesicatoria in artificially and naturally contaminated seeds lots. The
reliability of the technique for detection of X. campestris pv. vesicatoria was further
investigated by comparison with the dilution plating on semiselective media and
detection by ELISA. Furthermore, plant pathogenic xanthomonads isolated from
pepper and tomato seeds were tentatively identified based on the analysis of the
restriction pattern produced from amplified DNA fragments related to the hrp genes.

157
Materials and Methods
Bacterial strains and culture conditions
The bacterial strains isolated from pepper and tomato materials which were
used in this study and their sources are listed in Table 7-1. All strains had been
identified previously as members of the Xanthomonas by fatty acid analysis (N. C.
Hodge, personal communication). All strains were streaked on nutrient agar (Becton
Dickinson) and single colonies were selected. Nutrient broth cultures were grown 24
hr on a rotatory shaker (150 rpm) at 28 C. Strains were stored in sterile tap-water at
room temperature.
Seed sources and artificial infestation
Pepper and tomato seeds were obtained from fruits or were provided by
commercial seed companies. Naturally infested seeds of pepper and tomato were
obtained by extracting seeds from diseased fruits or were provided by the Georgia
Department of Agriculture at Tilton, GA. In some experiments, bacterial suspensions
ofX. campestris pv. vesicatoria 75-3 was added to the seed wash mixture at the
appropriate concentrations before ultrasonication.
Plant material and plant inoculations
All plants were maintained in a growth chamber at 28-30C during inoculation
and incubation. The pepper cultivar. Early Calwonder and the tomato cultivar, Bonny
Best were used in the pathogenicity tests. In pathogenicity tests, fully expanded leaves
of plants were inoculated with bacterial suspensions by infiltrating the bacteria into the
intercellular spaces by using a 1 ml plastic syringe with a 27 gauge needle. The

158
concentration of the inoculum was approximately 5 X 108 CFU/ml in sterile tap water,
determined by measuring the optical density in a Spectronic 20 spectrophotometer
(Bausch and Lomb, Inc., Rochester, NY). Plant reactions were scored over a period of
several days.
Recovery of bacteria from seeds and seedlings
The bacterial fraction was recovered from seeds and seedlings by using the
ultrasonication technique (Haefele and Webb, 1982; Morris, 1985). Two and one half
grams of seeds, or ten grams of aerial parts of seedling washings were washed in 20 ml,
or 150 ml of phosphate buffer (8.5 mM K2HP04, 7.5 mM KH2PO4, 0.02% Tween 20,
pH 7.0), respectively, for 20 min in a model B-22-4 ultrasonic cleaner (Branson
Cleaning Equipment Co., Shelton, CT). Bacterial populations were determined by the
dilution plate count method (Taylor, 1962). Dilutions were made from the washings in
sterile tap-water, and 0.1 ml aliquots were plated onto yeast-extract nutrient agar
(YNA) (Schaad and Stall, 1988), or Tween medium B (McGuire et al., 1986), or both.
X. campestris pv. vesicatoria was determined by the presence of circular, raised, and
yellow colonies surrounded by a lipolytic halo on Tween medium B (McGuire et al.,
1986; McGuire and Jones, 1989). Aliquots of the seed and seedling washings were
also processed and used in the DNA amplification and ELISA assays.
Extraction of DNA from bacteria and plant washings
Total genomic DNA of each bacterial strain was isolated by phenol-chloroform
extraction and ethanol precipitation essentially as described by Ausubel et al. (1987).
Bulk DNA extracts were obtained from seed and seedling washings by using a
modification of the method for DNA extraction described by Ausubel et al. (1987).

159
Aliquots of the seed or seedling washings were transferred to 1.5 ml microfuge tubes,
and sodium ascorbate (Sigma, St. Louis, MO) and insoluble polyvinypolypyrrolidone
(PVPP) (Sigma) were added for a final concentration of 0.2 M and 0.1% respectively.
These concentrations of PVPP and sodium ascorbate were used throughout this study,
unless otherwise stated. Prior to use, the PVPP was acid washed by the procedure
described by Holben et al. (1988). The samples were homogenized by vortexing. The
homogenate was pelleted by centrifuging in an Eppendorf microcentrifuge (Brinkmann
Instruments Inc., Westbury, NY) for 2 min at 16,000 g. The pellet was resuspended in
567 pi of TE buffer (10 mM Tris-Cl, pH 8.0; 1 mM EDTA, pH 8.0). Proteinase K
(Boehringer Mannheim, Indianapolis, IN) and sodium dodecyl sulfate (SDS) (Sigma)
were added for a final concentration of 100 pg/ml and 0.5%, respectively. After
incubation for 1 hour at 37C, sodium chloride and hexadecyltrimethyl ammonium
bromide (Sigma) were added to each preparation for a final concentration of 0.7 M and
1%, respectively. The preparations were incubated for 10 min at 65C. DNA was
purified by treatment with chloroform-isoamyl alcohol (24:1). The samples were
vortexed and centrifuged for 5 min at 16,000 g. A second purification was
accomplished by adding phenol-chloroform-isoamyl alcohol (25:24:1) and centrifuging
as described above. DNA was precipitated by adding 0.6 volumes of isopropanol and
incubating for 30 min at -20C. The samples were centrifuged for 20 min at 16,000 g.
The DNA pellet obtained was washed with 70% ethanol and centrifuged again. After
drying, the pellet was redissolved in 50 pi of TE buffer and stored at 4C.
DNA amplification
Three sets of oligonucleotide primers selected from the nucleotide sequence of
the hrp gene cluster of Xanthomonas campes tris pv. vesicatoria (Ulla Bonas, personal

160
communication) were used in this study. Primers RST2 plus RST3 delineated an 840-
bp fragment, RST9 plus RST10 delineated a 355-bp fragment, and RST21 plus RST22
delineated a 1,075-bp fragment. The primers map to the complementation groups hrpB
and hrpC/D of the hrp gene cluster of X. campestris pv. vesicatoria (Chapter 3).
Oligonucleotide primers were synthesized with a model 394 DNA Synthesizer
(Applied Biosystems, Foster City, CA) by the DNA Synthesis Laboratory, University
of Florida, Gainesville.
DNA fragments were amplified in a reaction mixture of 50 pi containing 5 pi of
10X buffer (500 mM KC1, 100 mM Tris-Cl [pH 9.0 at 25C], 1% Triton X-100), 1.5
mM MgC^, 200 pM of each deoxynucleotide triphosphate (Boehringer Mannheim), 25
pmol of each primer, 1.25 units of Taq polymerase (Promega, Madison, WI). The
amount of template DNA added was 100 ng of purified bacterial DNA whereas for the
seed and seedling samples a volume of 10 pi of the DNA extract was added, unless
otherwise stated. The reaction mixture was overlaid with 50 pi of light mineral oil.
Thirty amplification cycles were performed in an automated thermocycler PT-100-60
(MJ Research, Watertown, MA) according to the following profiles: 30 s of
denaturation at 95C, 30 s of annealing at 62C, and 45 s of extension at 72C for the
primers RST2 plus RST3; 30 s of denaturation at 95C, 30 s of annealing at 52C, and
45 s of extension at 72C for the primers RST9 plus RST10; 30 s of denaturation at 95
C, 40 s of annealing at 61C, and 45 s of extension at 72C for the primers RST21 plus
RST22. For all three profiles, the initial denaturation step was 5 min at 95C, and the
last extension step was extended to 5 min. Amplified DNA fragments were detected by
electrophoresis in 0.9% agarose gels in TAE buffer (40 mM Trisacetate, 1 mM EDTA,
pH 8.2) according to standard procedures (Sambrook et al 1989).

161
Restriction endonuclease analysis
The hrp-related fragments amplified from different bacterial strains were
restricted with the frequent-cutting endonucleases Cfo\, HaeIII, Sau3 AI, or Taql,
according to conditions specified by the manufacturer (Promega). The restricted
fragments were separated by electrophoresis in 4% agarose gels (3% NuSieve and 1%
Seakem GTG [FMC BioProducts]) in TAE buffer at 8 V/cm. Phage X fid-restricted
DNA fragments were used as molecular standards. The gel was stained with 0.5 pg of
ethidium bromide per ml for 40 min and then destained in 1 mM MgS04 for 1 hr and
photographed over a UV transilluminator with type 55 Polaroid film. The restriction
pattern obtained was compared to the standard banding profiles established for the
different groups of plant pathogenic xanthomonads (Chapter 4).
ELISA assay
An ELISA procedure that was developed for detection of low populations ofX.
campests pv. vesicatoria was used (Somodi et al., 1993). Aliquots of the seed
washings were transferred to microfuge tubes and an equal amount of EDTA/lysozyme
lysis buffer (2.0 g KH2P04, 11.5 g Na2P04, 0.14 g EDTA disodium, 0.02 g thimerosal,
0.2 g lysozyme (Sigma), 1 1 deionized water) was added. The samples were
homogenized by vortexing and incubated for at least 16 hr at room temperature.
Immulon 2 (Dynatech Laboratories, Chantily, VA) flat bottom 96-well
microtiter plates were coated with a polyclonal antibody developed against X.
campestris pv. vesicatoria 75-3 and incubated at 4C overnight. The coating buffer and
phosphate buffered saline (PBS) were used as described by Clark and Adams (1977).
All further incubations were for 2 hr at 37C. The microtiter plates were washed with a
solution containing 0.8% NaCl and 0.1% Tween 20 in deionized water (NTrinse) (G.

162
C. Somodi, personal communication). Plates were then blocked with 1% bovine
albumin (Sigma, A-9647) in PBS and shaken to remove excess liquid. Aliquots of 100
pi of each sample in EDTA/lysozyme lysis buffer were added to three wells on the
microtiter plate and incubated. The microtiter plates were rinsed with NTrinse and a
monoclonal antibody (2H10) prepared against X. campestris pv. vesicatoria 75-3 was
added to all wells and the plates were incubated. The microtiter plates were rinsed
again with NT rinse, and alkaline phosphatase conjugated goat antimouse A-1047
(Sigma) was applied to the plate. The microtiter plates were incubated and then rinsed
with the final washing buffer Tris-buffered saline (1.51 g of Tris-base, 2.19 g NaCl,
final volume 250 ml with deionized water, pH 7.5) four times. Substrate and amplifier
were added according to the instructions specified by the manufacturer of the ELISA
Amplification System (Gibco BRL, Gaithersburg, MD). Readings were made at A492
15 min after addition of the amplifier with an EAR400 AT plate reader (SLT
Labinstruments, Austria).
Fatty acid analysis
The fatty acid profiles of the strains of X. campestris isolated from pepper and
tomato seeds and seedlings were determined. A single colony of each strain was
transferred from a nutrient agar culture to trypticase soy broth agar (TSBA).
Approximately 40 mg of cells of a 24 h growth at 28C was collected for analysis. The
bacterial fatty acids were derivatized to their methyl esters (Miller, 1982) and separated
by gas chromatography (Sasser, 1990) and identified with the Microbial Identification
System software (version 3.80; MIDI, Newark, DE). The similarity index and best
match for the strains based on fatty acid analysis data were determined by using the
MIDI software.

Table 7-1. Phenotypic and genetic characterization of strains of Xanthomonas campestris associated with pepper and tomato
seed and seedling.
Strain
Origin
Reaction on
plants
Amplification
of hrp fragment
Tentative identification
Source3
pepper tomato
hrpB
hrpC/D
hrp analysis3
FAAC
SP1.92
pepper seed
_d
HR
+e
+
pv. campestrisf
pv. campestris (0.820)
GOK
SP101.92
pepper seed
-
HR
+
+
pv. campestris
pv. campestris (0.463)
GOK
SP268.92
pepper seed
-
-
-
-
na§
pv. armoraciae (0.703)
GOK
SP290.92
pepper seed
-
HR
-
-
na
pv. pruni (0.852)
GOK
3118-GP
pepper seed
+
HR
+
+
pv. vesicatoria A
pv. vesicatoria (0.735)
GOK
7502-2FS1
pepper seed
-
HR
-
-
na
pv. raphani (0.677)
JD
157
pepper seed
HR
HR
+
+
nd
pv. juglandis (0.644)
WW
T1083
tomato seed
-
-
-
-
na
pv. celebensis (0.051)
WW
524A-1
pepper seed
HR
HR
+
+
pv. campestris
pv. armoraciae (0.768)
WW
524A-2
pepper seed
HR
HR
+
+
pv. campestris
pv. armoraciae (0.812)
WW
75-0-3
tomato seed
HR
+
+
+
pv. vesicatoria C
pv. manihotis (0.835)
WW
7B-0-1
tomato seed
HR
+
+
+
pv. vesicatoria C
pv. fici (0.773)
WW
T1087
tomato seed
HR
HR
+
+
pv. citrumelo
pv. poinsettiicola (0.789)
WW
l-A-0-1
pepper seed
+
+
-1-
+
pv. vesicatoria A
pv. vesicatoria (0.745)
WW
P996
pepper seed
-
-
-
-
na
pv. raphani (0.581)
WW
DM-1
pepper seed
-
+
+
+
nd
pv. juglandis (0.618)
WW
LM-1
tomato seed
-
HR
-
(+)
nd
pv. malvacearum (0.644)
WW
639-6/FS1
pepper seed
-
-
-
-
na
nd
JW
639-6/FS2
pepper seed
-
-
-
-
na
nd
JW
Continued on the following page

Table 7-1 Continued
Strain Origin Reaction on Amplification Tentative identification Source
plants of hrp fragment
pepper tomato
hrpB
hrpC/D
hrp analysis
FAA
9310
tomato seedling
HR
+
+
+
pv. vesicatoria B
pv. manihotis (0.723)
DAM
9311
tomato seedling
HR
+
+
-
nd
pv. campestris (0.778)
DAM
140A-dl
tomato seed
HR
HR
+
+
pv. raphani
pv. campestris (0.897)
DAM
140A-d2
tomato seed
HR
HR
+
+
pv. raphani
pv. campestris (0.911)
DAM
T-93-23
tomato transplant
HR
+
+
+
pv. vesicatoria B
pv. alfalfae (0.707)
DAM
724-4
tomato seed
HR
HR
+
-
nd
pv. vesicatoria (0.816)
DAM
P93-22
pepper seedling
-
-
-
-
na
pv. pruni (0.799)
GOK
T93-09
tomato seedling
HR
HR
+
+
pv. raphani
pv. campestris (0.715)
GOK
T93-12A
tomato seedling
-
-
-
-
na
pv. zinnae (0.856)
GOK
VSE069
tomato seed
HR
+
+
+
pv. vesicatoria B
pv. vesicatoria (0.698)
MD
VSE070
tomato seed
HR
+
+
+
pv. vesicatoria B
pv. vesicatoria (0.620)
MD
VSE071
tomato seed
HR
+
+
+
pv. vesicatoria B
pv. vesicatoria (0.718)
MD
BSA1
pepper seed
HR
HR
+
+
nd
pv. corylina (0.628)
JW
BSA2
pepper seed
HR
HR
+
+
nd
pv. corylina (0.619)
JW
BSA3
pepper seed
HR
HR
(+)
(+)
nd
pv. pruni (0.671)
JW
BSA4
pepper seed
HR
HR
+
+
pv. citrumelo
pv. fici (0.765)
JW
BSA5
pepper seed
HR
-
-
-
na
pv. cannae (0.598)
JW
BSA6
pepper seed
HR
-
-
-
na
pv. cannae (0.586)
JW
BSA7
pepper seed
HR
HR
-
-
na
pv. pruni (0.832)
JW
BSA8
pepper seed
HR
HR
(+)
+
pv. carotae
pv carotae (0.730)
JW
Continued on the following page

Table 7-1 Continued
Strain
Origin
Reaction on
plants
Amplification
of hrp fragment
Tentative identification
Source
pepper tomato
hrpB
hrpC/D
hrp analysis
FAA
BSA9
pepper seed
HR
HR
(+)
+
pv. carotae
pv. carotae (0.824)
JW
BSA10
pepper seed
HR
HR
(+)
+
pv. carotae
pv. carotae (0.903)
JW
BSA11
pepper seed
HR
HR
(+)
+
pv. carotae
pv. carotae (0.852)
JW
BSA12
pepper seed
HR
HR
(+)
+
pv. carotae
pv. carotae (0.767)
JW
BSA13
pepper seed
-
HR
(+)
+
pv. carotae
pv. carotae (0.753)
JW
BSA14
pepper seed
HR
HR
(+)
+
pv. carotae
pv. carotae (0.577)
JW
BSA15
pepper seed
HR
HR
(+)
+
pv. carotae
pv. carotae (0.748)
JW
BSA16
pepper seed
-
-
(+)
+
pv. carotae
pv. carotae (0.841)
JW
BSA17
pepper seed
-
-
(+)
+
pv. carotae
pv. carotae (0.868)
JW
BSA18
pepper seed
HR
HR
-
-
na
pv. malvacearum (0.736)
JW
BSA19
pepper seed
HR
HR
-
-
na
pv. malvacearum (0.720)
JW
BSA20
pepper seed
-
HR
(+)
+
pv. carotae
pv. carotae (0.789)
JW
BSA21
pepper seed
-
-
(+)
+
pv. carotae
pv. carotae (0.905)
JW
BSA22
pepper seed
-
-
(+)
+
pv. carotae
pv. carotae (0.809)
JW
BSA23
pepper seed
HR
HR
(+)
+
pv. carotae
pv. carotae (0.782)
JW
BSA24
pepper seed
-
HR
(+)
+
pv. carotae
pv. carotae (0.779)
JW
BSA25
pepper seed
-
-
(+)
+
pv. carotae
pv. carotae (0.752)
JW
Continued on the following page

Table 7-1Continued
a DAM, D. A. Maddox, Seed Testing of America, Longmont, CO; GOK, G. O'Keefe, Department of Agriculture, State of
Georgia, Tifton, GA; JD, J. Dodson, Woodland Research Station, Petoseed, Woodland, CA; JW, J. Watterson, Woodland
Reserach Station, Petoseed, Woodland, CA; MD, M. Derie, Washington State University, Pullman, WA; WW, W. Wiebe,
Rogers NK, Woodland CA.
b Identification of the strain was based on comparison of the hrp profile established to the different groups of plant
pathogenic xanthomonads (see Chapter 4).
c Identification based on the MIDI TSBA Library (version 3.80; Microbial Identification System software, Newark, DE).
Number in parenthesis is the similarity index.
d -, no /irp-related fragment amplification; +, /zr/?-related fragment amplification; (+) weak /irp-related fragment
amplification.
e -, no reaction; +, pathogenic reaction; HR, hypersensitive reaction.
f Strains of X. campestris pv. campestris could not be distinguished from strains of X. campestris pv. armoraciae by the
restriction analysis of both hrp fragments (see Chapter 4).
8 na, not applicable; nd, not determined.
On
On

167
Results
Development of a DNA amplification approach for detection of plant pathogenic
xanthomonads in seed extracts
Cells of X campestris pv. vesicatoria were detected in extracts of tomato seeds
by amplification of an 840-bp DNA fragment of the hrp gene cluster. The fragment
was amplified from bulked DNA preparations obtained from the extracts to which cells
of strain 75-3 were added (Fig. 7-1 and 7-2). The 355-bp hrpB fragment was also
amplified from the same samples (data not shown). The hrp fragments were amplified
from the samples processed with and without the presence of PVPP and sodium
ascorbate (Fig. 7-1 and 7-2). Similar results were also obtained with other pepper and
tomato seed lots (data not shown). The two reagents have been added to environmental
and plant samples in order to assure the extraction of nucleic acids with the quality
required for consistent enzymatic amplification of DNA fragments (Holben et al.,
1988; Minsavage et al., 1994). The addition of PVPP, or sodium ascorbate, or both at
different concentrations to the seed washings apparently did not interfere with the
recovery of DNA, nor the amplification of the hrp fragments by polymerase chain
reaction (Fig. 7-1 and 7-2). Other methods of sample preparation, such as the quick
approach of DNA isolation (Kawasaki, 1990), did not produced consistent results in the
amplification of hrp fragments for the detection of X. campestris pv. vesicatoria present
in pepper and tomato seed washings (data not shown).
Investigation was carried out to further examine the quality of the DNA
extracted from tomato seed washings with sodium ascorbate and PVPP added to the
final concentration of 0.2 M and 0.1% respectively, or without these reagents for
amplification of the hrp fragments. Aliquots of 5 pi, 10 pi, and 15 pi of DNA
extracted from tomato seed washings containing cells of X. campestris pv. vesicatoria

168
M 1 2 3 4 5 6 7
1 375
564-
Fig. 7-1. Amplification of the 840-bp fragment of the complementation group B of the
hrp gene cluster from seeds of tomato containing 106 CFU/ml of Xanthomonas
campestris pv. vesicatoria 75-3 and different amounts of PVPP. Lanes: M, phage X
restricted with £coRI and HindiII; 1, no PVPP nor sodium ascorbate; 2, no PVPP; 3,
1% of PVPP; 4, 2% of PVPP; 5, 3% of PVPP; 6, 4% of PVPP; 7, 5% of PVPP.
Molecular sizes are given in bases.

169
M 1 2 3 4 5
1
Fig. 7-2. Amplification of the 840-bp fragment of the complementation group B of the
hrp gene cluster from seeds of tomato containing 106 CFU/ml of Xanthomonas
campestris pv. vesicatoria 75-3 and different amounts of sodium ascorbate. Lanes: M,
phage X restricted with EcoRl and HindiII; 1, no sodium ascorbate nor PVPP; 2, no
sodium ascorbate; 3, 0.02 M of sodium ascorbate; 4, 0.2 M of sodium ascorbate; 5,2 M
sodium ascorbate. Molecular sizes are given in bases.

170
75-3 were added to the PCR reaction mixture for amplification of the hrpB fragment
with primers RST2 and RST3. The 840-bp hrpB fragment was amplified from all
different samples, and no significant inhibition of the amplification reaction was
observed (Fig. 7-3).
Sensitivity of the detection procedure
The sensitivity of the DNA amplification procedure for detection of cells of2f
campestris pv. vesicatoria in preparations of tomato seed extracts was about 102
CFU/ml when the primers RST9 and RST10 were used for amplification of the 355-bp
hrpB fragment (Fig. 7-4; Table 7-2). The minimum detection level ofX. campestris
pv. vesicatoria was usually 10 to 100 times higher when the primers RST2 and RST3
were used for amplification of the 840-bp hrpB fragment (Fig. 7-4; Table 7-2). The
lower sensitivity of the primers RST2 and RST3 for detection of X. campestris pv.
vesicatoria was also observed for other tomato and pepper seed lots (data not shown).
Despite the differences in sensitivity, both sets of primers produced specific
amplification of the hrp fragments from seed extracts containing different
concentration of A", campestris pv. vesicatoria (Fig. 7-4). In comparison to ELISA, the
DNA amplification procedure was at least 100 to 1000 times more sensitive for
detection ofX. campestris pv. vesicatoria when cells were added to the seed
preparation (Table 7-2). The detection level ofX. campestris pv. vesicatoria by ELISA
was greater than 105 CFU/ml (Fig. 7-4; Table 7-2).
Detection of plant pathogenic xanthomonads in naturally contaminated seed lots
Investigations were conducted to determine the feasibility of the DNA
amplification procedure for detection of the presence of plant pathogenic

171
M 1 2 3 4 5 6
1 3
5
Fig. 7-3. Amplification of the 840-bp fragment of the hrp gene cluster of Xanthomonas
campestris pv. vesicatoria 75-3 from seeds of tomato by using sodium ascorbate and
PVPP (lanes 1, 2, and 3) and without these reagents (lanes 4, 5, and 6). Lanes: M,
phage X restricted with £coRI and Hindlll; 1 and 4, 5 pi of DNA extract; 2 and 5, 10 pi
of DNA extract; 3 and 6, 15 pi of DNA extract. Molecular sizes are given in bases.

172
12 3 4
8 9 101112 13 14 15 16171819 20
Fig. 7-4. Amplification of the 355-bp (Lanes 2 to 10) and 840-bp (Lanes 12 to 20)
fragments of the complementation group B of the hrp gene cluster from samples of
tomato seeds containing different concentration of Xanthomonas campestris pv.
vesicatoria 75-3 added to the seed washings. Lanes: 1 and 11, phage X restricted with
EcoRl and Hindlll; 2 and 12, 2.6 X 10* CFU/ml; 3 and 13,2.9 X 107 CFU/ml; 4 and
14, 0.9 X 106 CFU/ml; 5 and 15, 3.4 X 105 CFU/ml; 6 and 16,1.0 X 104 CFU/ml; 7
and 17, 3.3 X 103, CFU/ml; 8 and 18, 3.0 X 102 CFU/ml; 9 and 19, 4.0 X 101 CFU/ml;
10 and 20, no bacteria added. Concentration ofX. campestris pv. vesicatoria in the
seed washings was determined by plating on Tween B medium. Molecular sizes are
given in bases.

173
Table 7-2. Sensitivity of ELISA and DNA amplification procedures for detection of
Xanthomonas campestris pv. vesicatoria in tomato seeds.
Sample Total population of X. campestris recovered ELISA Amplification
bacteria on YNA on Tween B medium Anm492a of hrp
(CFU/ml) (CFU/ml) fragment*5
355 bp 840 bp
1
2.9 X 108
2.0 X 108
2.79 0.01
c

2
7.3 X 107
6.7 X 107
2.09 0.24


3
5.4 X 106
4.6 X 106
0.51 0.01


4
5.6 X 105
5.5 X 105
0.21 0.01


5
6.8 X 104
6.8 X 104
0.16 0.02


6
5.8 X 103
5.2 X 103
0.16 0.01


7
1.7 X 102
1.4 X 102
0.17 0.01

-
8
2.6 X 102
7.0 X 101
0.17 0.01
-
-
9
1.5 X 102
not detected
0.16 0.01
-
-
aValues are the average of the spectrophotometer readings of three wells.
bAliquots of 300 pi and 600 pi of the seed washings were used in the ELISA assay and
for extraction of DNA, respectively. In the DNA amplification assay was used 10 pi of
DNA template.
c no hrp fragment amplification; +, hrp fragment amplification.

174
xanthomonads in naturally contaminated seeds. Plant pathogenic xanthomonads in
pepper and tomato seed lots were detected consistently by amplification of the hrpB
fragments in DNA extracted from washes of 11 seed lots out of the 15 tested ( Table 7-
3). The technique specificaly amplified the expected DNA fragments despite the
background bacterial microflora of more than 105 CFU/g of seed (Fig. 7-5; Table 7-3).
The primers RST9 and RST10 failed to produce amplification of the 355-bp hrp-
related fragment from the pepper seed lots SP 135.92 and PK-l-PF though
amplification of the 840-bp /irp-related fragment indicated the presence of plant
pathogenic xanthomonads in these seeds (Table 7-3). This is not surprising because the
set of primers RST9 and RST10 seem to allow DNA amplification only from strains of
a limited number of pathovars of A. campestris (Chapter 3).
Attempts to recover plant pathogenic xanthomonads from different seed lots by
plating on general and semiselective media was unsuccessful. Whereas no colonies
typical of xanthomonads on both media used, YNA and Tween B, were found with the
seed lots containing low bacterial population, a large background bacterial microflora
in extracts from other seed lots prevented the identification of any possible
xanthomonads that may have grown. However, the presence of viable plant pathogenic
xanthomonads in 5 seed lots was confirmed by detection of the bacteria in seedlings
from seeds (Table 7-3). The detection of the plant pathogenic xanthomonads in the
seedlings was accomplished by isolating the bacteria from lesions in the leaflets, or by
amplification of the hrp-related fragments from bulked DNA extracted from washings
of the aerial part of the seedlings grown under greenhouse conditions. Furthermore, the
presence of viable plant pathogenic xanthomonads in the pepper seed lots SP2.92,
SP66.92, SP124.92, SP133.92, SP 135.92, and SP306.92 had been determined
previously (G. O'Keefe, personal communication).

175
M1 23 45678
1 375
564
Fig. 7-5. Amplification of the 355 bp DNA fragment related to the complementation
group B of the hrp gene cluster of Xanthomonas campestris pv. vesicatoria from DNA
extracted from different pepper seed lots. Lanes: M, phage X restricted with EcoRl and
HindilI; 1, SP2.92; 2, SP66.92; 3, SP124.92; 4, SP133.92; 5, SP135.92; 6, SP306.92;
7, Greenhouse grown; 8, Jupiter. Molecular sizes are given in bases.

Table 7-3. Detection of plant pathogenic xanthomonads in naturally infected pepper and tomato seeds.
Seed sample
Total population of
bacteria on YNA
(CFU/g seed)
Recovery of
xanthomonads
Tween B Grow outc
medium
Amplification of hrp
fragment3
355 bp 840 bp
Source of the
seedb
Pepper
SP2.92
3.2 X 102
_d
-
+
+
GOK
SP66.92
1.9 X 103
-
-
+
+
GOK
SP 124.92
3.2 X 102
-
+
+
+
GOK
SP133.92
>2.4 X 105
-
+
+
+
GOK
SP135.92
1.6 X 102
-
-
-
+
GOK
SP306.92
0.8 X 102
-
+
+
+
GOK
Jupiter
4.0 X 102
-
-
-
-
RST
Greenhouse grown
5.4 X 102
-
-
(+)
(+)
RST
PK-l-PF
>4.0 X 103
-
-
-
+
JFW
PK-2-PF
>4.0 X 103
-
+
+
+
JFW
PL-3-P
>4.0 X 103
-
-
+
+
JFW
Tomato
#51-lb
4.0 X 103
-
-
-
-
MM
Diseased fruits
1.1 X 107
-
+
+
nd
RST
Marglobe #100016
4.6 X 103
-
-
-
nd
RST
Manalucie
1.0 X 103
-
-
-
nd
RST
Continued on the following page

Table 7-3Continued
Amplification of fragments related to the hrpB of Xanthomonas campestris pv. vesicatoria by 30 cycles of polymerase chain
reaction.
bGOK, G. O'Keefe, Department of Agriculture, State of Georgia, Tifton, GA; JFW, J. F. Wang, The Asian Vegetable
Research and Development Center Tainan, Taiwan; MM, M. Meadows, Rogers NK, Naples, FL; RST, R. E. Stall, University
of Florida, Gainesville.
cPresence of plant pathogenic xanthomonads was determined by isolation of bacteria or by detection by amplification of hrp-
related fragment from diseased tissue.
d- negative result; + positive result; (+), weak positive result; nd, not determined.

178
Tentative identification of A', campestris associated with seeds and seedling? of pepper
and tomato
The identity of 56 strains of Xanthomonas from seeds and seedlings of pepper and
tomato was investigated by examining the reaction on plants and by analyzing DNA
sequences related to the hrp genes and fatty acid composition (Table 7-1). The strains
seem to comprise a very diverse group of xanthomonads that includes plant pathogenic
and nonplant pathogenic bacteria. Some of the strains were pathogenic on plants of
Bonny Best tomato and were identified as X. campestris pv. vesicatoria on the basis of
the hrp fragment analysis, with the exception of the strains DM-1 and 9311 (Table 7-1).
In the case of the latter strains, /^-related fragments were amplified, but the restriction
analysis of the fragments were not done. The strains SP268.92, P996, 639-6/FS1, 639-
6/FS2, T93-12A, BSA5, and BSA6 did not show any reaction on plants and they failed to
produce the 840- and 1,075-bp hrp-related fragments in the DNA amplification assay
(Table 7-1). Although these strains have some degree of similarity to different pathovars
of* campestris on the basis of fatty acid composition (Table 7-1), they resemble the
opportunistic xanthomonads that have been reported previously in association with
tomato and pepper transplants (Gitaitis et al., 1987, 1992). The remainder of strains form
a very diverse group of potential plant pathogenic xanthomonads, as determined by the
hypersensitive reaction on plants, or by the presence in their genome of a region similar
to the hrp genes, or by both criteria (Table 7-1).
The identification of these strains of Xanthomonas based on the hrp analysis was
performed by comparison of the restriction profile generated by digestion of the hrp
fragment amplified from the strains with frequent-cutting endonucleases (Fig. 7-6) to the
pattern established for the different groups of plant pathogenic xanthomonads (Chapter
4). The restriction analyses allowed the identification of groups of strains of X.
campestris pv. vesicatoria (Table 7-1). Two strains were identified to the group A, five to

179
Fig. 7-6. Restriction analysis of the 1,075-bp DNA fragments related to the hrpC/D
complementation group and amplified from strains of Xanthomonas campestris isolated
from pepper and tomato seed and seedling, and restricted with the endonucleases Taql
(Lanes 2 to 10) and HaeIII (Lanes 12 to 20). Lanes: 1 and 11, phage k restricted with
Pstl; 2 and 12, 157; 3 and 13, 524A-1; 4 and 14, 75-0-3; 5 and 15, 7B-0-1; 6 and 16,
T1087; 7 and 17, 524A-2; 8 and 18, l-A-0-1; 9 and 19, DM-1; 10 and 20, X. campestris
pv. vesicatoria 75-3. Molecular sizes are given in number of bases.

180
the group B ofX. campestris pv. vesicatoria, and two to group C (Table 7-1). The fatty
acid analysis did not identify the groups of strains and strains 75-0-3 and 9310 were
similar to the fatty acid library ofX. campestris pv. manihotis, and the strains T-93-23
and 7B-0-1 that were very similar to X campestris pv. alfalfae and X campestris pv. fici,
respectively (Table 7-1). Most of the strains that produced hypersensitive reaction on
pepper and tomato were identified as X. campestris pv. armoraciae, X. campestris pv.
campestris, X. campestris pv. carotae, or X campestris pv. raphani, with a high degree of
agreement in the identification by analyses of DNA sequences related to the hrp genes
and by fatty acid composition (Table 7-1). Fifteen strains isolated from pepper seeds
were identified as X. campestris pv. carotae by analyses of DNA sequences related to the
hrp genes and fatty acid composition, though some variability was observed on the
reaction on plants (Table 7-1). The restriction analysis of the /^-related fragments
amplified from the strains SP290.92, 157, T1083, DM-1, LM-1. 9311, 724-4, BSA1,
BSA2, and BSA3 produced pattern different from those determined for 50 different
groups of plant pathogenic xanthomonads (Chapter 4), or the yield obtained in the DNA
amplification was too low for restriction fragment length polymorphism analysis.
Therefore, the identification of these strains on the basis of the restriction profile of the
hrp-related fragments to pathovar or group of X. campestris was not possible.
Discussion
There is an increased need for specific and sensitive methods for detection of
plant pathogenic bacteria associated with seeds and other plant propagative materials
(Miller and Martin, 1988; Schaad, 1982; Saettler et al., 1989). Based on the results
presented here, the amplification and analysis of DNA sequences related to the hrp genes

181
can be a useful tool for the specific detection and identification of plant pathogenic
xanthomonads. X. campestris pv. vesicatoria was readily detected in preparations
containing cells added to pepper and tomato seed extracts and in seed washings obtained
from naturally contaminated pepper and tomato seeds. The sensitivity of the method of
DNA amplification ranged from about 102 to 103 CFU/ml of seed washings. This level
was 100 to 1000 times higher than the level obtained with ELISA, and it is certainly
comparable to the levels obtained with the most sensitive techniques available for
detection of bacteria (Pickup, 1991; Saettler et al., 1989). Furthermore, the specificity of
the method allowed the detection of plant pathogenic xanthomonads against a
background bacterial micro flora larger than 107 CFU/g of seed. This level of non target
bacterial population may have prevented the detection of the pathogen by plating in
general and semiselective media.
X. campestris pv. vesicatoria was detected in DNA extracted from seed washings
containing the reagents PVPP and sodium ascorbate and from preparations without these
reagents. Although the hrp fragments were amplified from DNA extracted from seed
samples without adding reagents to prevent potential inhibitors, the addition of PVPP and
sodium ascorbate in the seed washings before nucleic acid extraction may be needed to
assure the necessary purity of the DNA for amplification. The presence of inhibitors have
been a major concern in the extraction of high quality DNA from environmental and plant
samples (Holben et al., 1988; Minsavage et al., 1994; Steffan et al., 1988). Nevertheless,
the addition of these two reagents to the seed washings before DNA extraction apparently
did not interfere with the yield of DNA fragments obtained in the amplification of the hrp
fragments.
A basic feature of the approach for recovering bacterial DNA used here is the
initial separation of intact bacterial cells from the seed washings by centrifugation. This

182
is followed by lysis of the cells and purification of the DNA recovered. This cell
extraction method may be less efficient than the direct lysis method that involves the
release and extraction of DNA without prior separation of the cells from the original
matrix (Ogram et al., 1988; Steffan et al., 1988). However, the cell extraction method is
likely to ensure exclusion of extracellular DNA that may be present in the sample and
may produce DNA of higher quality (Holben et al., 1988; Steffan et al., 1988). In fact,
inconsistent results in the detection of X. campestris pv. vesicatoria were obtained with
the quick approach of DNA isolation which employs direct lysis of the bacterial cells in
the original extraction matrix.
The procedure of amplification of DNA sequences of the hrp genes was highly
sensitive and specific for detection of plant pathogenic xanthomonads in DNA
preparations from seeds washes. Based on dilution plating counts, X. campestris pv.
vesicatoria could be detected in preparations containing less than 102 CFU added to seed
extracts (Table 7-2). However, this level of sensitivity may be improved further by
including a step for concentration of the cells in the sample to enhance the recovery of
bacterial cells (Roth, 1989; Schaad, 1982). The concentration step concomitantly
concentrates unwanted saprophyte microflora. Although the presence of a large
background microflora is a major limitation in the detection of plant pathogenic
xanthomonads in pepper and tomato seeds by planting on general or even semiselective
media (Gitaitis et al., 1991; McGuire et al., 1986), the specificity of the DNA
amplification procedure is likely to overcome this kind of problem. In fact, X. campestris
pv. vesicatoria was detected in naturally contaminated seed lots containing background
bacterial microflora larger than 107 CFU/g of seeds, whereas plating on semiselective
medium failed to recover the xanthomonads (Table 7-2). Nevertheless, the transmission
tests confirmed the presence of viable cells of plant pathogenic xanthomonads in 5 seed

183
lots out of the 11 which produced positive results in the DNA amplification procedure.
This is not surprising because X. campestris pv. vesicatoria has the capability to survive
for long periods of time in seeds and to initiate disease (Bashan et al., 1982; Gardner and
Kendrick, 1921, 1923).
A diverse group of xanthomonads was found associated with pepper and tomato
seeds. The majority of the 56 strains examined have the basic characteristics of plant
pathogenic xanthomonads. They produced pathogenic or hypersensitive reaction on
plants in pathogenicity tests, and DNA sequences related to the hrp genes were amplified
from their genome. On the contrary, seven strains did not produce reaction on plants, and
they failed to produce DNA sequences similar to the hrp genes in DNA amplification
assay. These strains resemble the opportunistic xanthomonads that have been described
in association with pepper and tomato transplants (Gitaitis et ah, 1987, 1992). Among
the plant pathogenic xanthomonads, only nine strains were positively identified as X.
campestris pv. vesicatoria, the pepper and tomato pathogen (Table 7-1). Another group
included potential plant pathogenic strains that were not pathogenic to pepper and tomato
but instead produced a hypersensitive reaction on these plants. Based on the analyses of
DNA sequences related to hrp genes and fatty acid composition, they were highly related
to different pathovars of X. campestris, but not to X campestris pv. vesicatoria.
Furthermore, the identification of the strains based on restriction profiles of the DNA
sequences related to the hrp genes agreed very closely with the identification by fatty acid
analysis. Some of these strains were identified as X. campestris pv. campestris though
their pathogenicity to cabbage was not determined. Strains similar to X. campestris pv.
campestris and pathogenic to cabbage have been reported previously on pepper and
tomato transplants (Gitaitis et al., 1987). Other strains were similar to X. campestris pv.
raphani which has also been found commonly in association with pepper and tomato

184
seeds (Gitaitis et al., 1987). Strains of this pathovar have a large host range and may
cause disease on pepper and tomato plants (Gitaitis et al., 1987; White, 1930).
Furthermore, outbreaks of a bacterial spot on tomato transplants in southern USA have
been associated to strains that have genetic and phenotypic characteristics similar to X.
campestris pv. raphani (R. E. Stall, personal communication). Another common
intriguing point is the similarity of fifteen strains isolated from pepper seeds to X.
campestris pv. carotae. Both restriction analysis of the DNA sequences related to the hrp
genes and fatty acid composition agreed in the identification of these strains as X.
campestris pv. carotae (Table 7-1). I am unaware of any previous report of the
association of strains similar to X campestris pv. carotae with pepper seeds.
The presence of a diverse xanthomonad community on pepper and tomato seeds
certainly poses a major concern for health inspection in seed certification programs.
Whereas some of the xanthomonads are probably nonpathogenic to plants, others are
pathogenic to pepper and tomato. The latter were identified as X. campestris pv.
vesicatoria based on pathogenicity tests, analysis of DNA sequences related to the hrp
genes, and fatty acid composition. However, a third group of xanthomonads that includes
potential plant pathogenic bacteria was also identified. These strains, tentatively
identified as belonging to different pathovars of X. campestris, cause hypersensitive
reaction or no reactions on pepper and tomato plants under artificial inoculation.
However, the role that these strains may play in a bacterial disease syndrome on pepper
and tomato plants remains to be clarified. Furthermore, the association of such a diverse
group of bacterial pathogens with pepper and tomato seeds is a major problem with the
procedures currently available for their detection and identification. These strains grow
on semiselective media developed for the pepper and tomato pathogen X. campestris pv.
vesicatoria (Gitaitis et al., 1987, 1991; McGuire et al., 1986) and they may cause certain

185
reactions on plants under artificial inoculations. Therefore, additional studies are
necessary to determine the role of these xanthomonads found in the seeds of pepper and
tomato in the subsequent development of bacterial diseases on seedlings and plants.
In conclusion, a procedure based on amplification of hrp-related fragments was
examined for sensitive and specific detection of plant pathogenic xanthomonads
associated with pepper and tomato seeds. The complexity of the xanthomonads
community found on seeds was assessed, and confirmed previous results obtained by
other researchers using different methods (Gitaitis et al., 1987, 1992). Furthermore, the
combined sensitivity and specificity of the DNA amplification approach certainly makes
the procedure an indispensable tool for the detection and identification of plant
pathogenic xanthomonads in seeds and other propagative materials. Although the
different xanthomonads could not be distinguished by the size of the amplified DNA
fragment, the restriction fragment length polymorphism present in the DNA sequences
related to the hrp genes obtained from different strains of Xanthomonas was valuable for
differentiation of these plant pathogenic xanthomonads. Furthermore, the comparison of
the restriction profile to predetermined profiles of several pathovars of X. campestris
(Chapter 4) allowed the tentative and precise identification of these plant pathogenic
xanthomonads.

CHAPTER 8
SUMMARY AND CONCLUSIONS
The plant pathogenic xanthomonads form a diverse group of usually yellow-
pigmented bacteria that occur worldwide and cause disease on many plants. The
differentiation of these bacteria at the subgeneric level is by the capability of the
bacterial strain to cause characteristic disease or by reference to their host range (Dye et
al., 1980; Vauterin et al., 1990a; Young et al., 1992). In fact, it is almost impossible to
differentiate the xanthomonads by biochemical and physiological features without
knowing their plant hosts (Bradbury, 1984; Dye, 1962; Holt et al., 1994; Schaad, 1988;
Van den Mooter and Swings, 1990). On the contrary, the genomic relationship among
plant pathogenic xanthomonads has revealed a group of bacteria with diverse genetic
background (Hildebrand et al., 1990; Palleroni et al., 1993; Vauterin et al., 1993). In
this work, I investigated a specific region of the bacterial genome, the DNA sequences
related to the hypersensitive reaction and pathogenicity (hrp) genes, for differentiation
of the xanthomonads and also for the establishment of the evolutionary relationships
among them.
Three sets of oligonucleotide primers selected from the DNA sequence of the
hrp genes of X campestris pv. vesicatoria were tested for amplification of DNA
fragments from different plant pathogenic xanthomonads. Two sets of primers, RST2
plus RST3 and RST9 plus RST10, are specific for the hrpB, and the set RST21 plus
RST22 is specific for a region of the hrpC/D of X. campestris pv. vesicatoria 75-3.
DNA fragments were amplified from X fragariae and from 31 of 33 plant pathogenic
taxa oiX. campestris tested, which comprise at least 28 different pathovars of this
186

187
species. In all cases, each set of primers amplified DNA fragments identical in size,
suggesting a high degree of structural conservation between operon, as seen with the
primers RST21 and RST22. Furthermore, the pathogenicity of several nonpathogenic
Tn3-gus mutants of X campestris pv. vesicatoria 85-10 was fully restored with cloned
regions of DNA o2l campestris pv. vesicatoria group B and X. campestris pv.
pelargonii, from which the hrp-rdated fragments were amplified. This supports the
contention that the fragments were amplified from DNA sequences which also control
the pathogenicity in other xanthomonads.
In contrast to the narrow spectrum of oligonucleotide primers previously
described for detection and identification of only certain strains of X. campestris
(Garde and Bender, 1991; Hartung, 1992), the hrp-specific primer sets RST2 plus
RST3 and RST21 plus RST22 were useful for identification of a large number of plant
pathogenic xanthomonads. This is not surprising, because the hrp region seems to be
very conserved among different plant pathogenic xanthomonads as determined in the
present and previous studies (Bonas et al., 1991; Stall and Minsavage, 1990). On the
other hand, the primers RST9 and RST10 allowed amplification of DNA fragment only
from a limited number of pathovars of X. campestris, although the hrp fragment
amplified with these primers from X. campestris pv. vesicatoria 75-3 hybridized to the
majority of the plant pathogenic xanthomonads included in this study. These results
indicate differences in the DNA sequences of the xanthomonads corresponding to one
or both primers used. However, the set of primers RST9 and RST10 seems useful for
specific detection of strains ofX. campestris pv. vesicatoria group A, X. campestris pv.
ci,X. campestris pv. physalidicola, and X campestris XI98.
Although no size variation was observed for the /irp-related fragments
amplified from different plant pathogenic xanthomonads, the restriction analysis of the

188
840- and 1,075-bp DNA fragments with frequent-cutting endonucleases indicated the
presence of characteristic sequence variation. A study carried out to determine the
extension of the diversity of /irp-related sequences of 192 strains of plant pathogenic
xanthomonads revealed that only a few taxa of the xanthomonads are homogeneous.
This includes X. fragariae, X. campestris pv. begoniae, X campestris pv. campestris,
X. campestris pv. malvacearum, and X campestris pv. pelargonii. The homogeneous
population structure of these taxa has also been supported by using different methods,
such as DNA-DNA hybridization, fatty acid composition, and SDS-PAGE of proteins
(Vauterin et al., 1990a, 1991a; Yang et al., 1993). On the contrary, the majority of the
pathovars of X campestris seem to comprise an heterogeneous group of strains as
determined by the analysis of the /?rp-related sequences. Whereas strains of some
pathovars formed distinct and uniform subgroups, e.g. X. campestris pv. citri, X
campestris pv. dieffenbachiae, and X campestris pv. vesicatoria, other pathovars were
composed of a highly diverse group of strains. The latter may be well characterized by
the population structure determined for X campestris pv. citrumelo. The 16 strains of
X. campestris pv. citrumelo included in the study were separated in nine different
groups on the basis of the restriction banding profiles generated for the two hrp-related
fragments digested with four different endonucleases. The diverse nature of the strains
of X. campestris pv. citrumelo has been reported previously based on DNA homology,
restriction fragment length polymorphism analysis, fatty acid composition, and SDS-
PAGE of proteins (Egel et al., 1991; Gabriel et al., 1988, 1989; Graham et al., 1990;
Hartung and Civerolo, 1987, 1989; Vauterin et al., 1991a).
The most striking feature of the restriction analysis of the amplified /zrp-related
fragments is the possibility for differentiation of almost all the pathovars and groups of
X. campestris included in the study on the basis of the restriction banding profile of the

189
hrp-related sequences. Fifty different restriction fragment length polymorphism groups
of plant pathogenic xanthomonads were established based on the banding patterns
generated by the analysis of two hrp-related fragments digested with four
endonucleases. Although the banding pattern for a combination /zrp-related fragment
and endonuclease may be shared by more than one group, the profile of all
combinations hrp fragments and endonucleases allowed the specific differentiation and
identification of the different plant pathogenic xanthomonads included in the study.
However, some strains of different groups of plant pathogenic xanthomonads could not
be distinguished by restriction analysis of the /^-related sequences. These include a
strain of X. campestris pv. incanae and one of X. campestris pv. carotae which were
identical to strains ofX. campestris pv. campestris; a strain of A! campestris pv.
phaseoli identical to strains of X. campestris pv. glycines group A; strains F59 and F86
of A! campestris pv. citrumelo were indistinguishable from strains ofX. campestris pv.
pruni; strain XI25 ofX. campestris pv. fici group A was identical to strains ofX.
campestris pv. poinsettiicola group A; strain XV2 of X. campestris pv. vitians was
identical to strains of X. campestris pv. malvacearum; and strain X52 isolated from
Hibiscus sp. was indistinguishable from strains of X. campestris pv. poinsettiicola
group B. No further investigation was carried out with those strains, but it may be
necessary a better assessment of the identity of these strains by using different
approaches.
To illustrate the usefulness of the amplification and analysis of /np-related
fragments in the study of plant pathogenic xanthomonads, I examined the
xanthomonads associated with pepper and tomato seeds. A complex xanthomonad
community was found in association with these seeds as determined by restriction
analysis of /zr/j-related DNA sequences and confirmed by fatty acid analysis. Among

190
the 56 strains isolated from pepper and tomato seeds or seedlings, only nine strains
were positively identified as X. campestris pv. vesicatoria, the pepper and tomato
pathogen. The other strains formed two distinct groups. Seven strains did not produce
a reaction on pepper and tomato plants, and they failed to produce DNA sequences
similar to the hrp genes. These strains resemble the opportunistic xanthomonads that
have been described previously in association with pepper and tomato transplants
(Gitaitis et al., 1987, 1992). The other group of xanthomonads included potential plant
pathogenic strains that were not pathogenic to pepper and tomato, but instead caused a
hypersensitive reaction or no reaction on these plants under artificial inoculation.
These strains were tentatively identified as belonging to different pathovars of X.
campestris, but not to X campestris pv. vesicatoria, on the basis of the hrp analysis and
fatty acid composition. The presence of a diverse group of xanthomonads associated
with pepper and tomato seeds and transplants is of major significance for the health
inspection in certification programs, particularly with the procedures currently
available for the detection and identification of plant pathogenic xanthomonads (Holt,
1994; Saettler et al., 1989; Schaad, 1988). Although more investigation may be
necessary, the combined specificity and sensitivity of the DNA amplification approach
of /zrp-related fragments seems highly promising for the detection and identification of
plant pathogenic xanthomonads associated with propagative plant materials.
Differences observed in the number of common restriction fragments of the
amplified DNA sequences related to the hrp genes indicated that there is variation in
the relatedness of the hrp genes of the different strains of plant pathogenic
xanthomonads. The relatedness of the hrp genes of the different plant pathogenic
xanthomonads was further investigated by phylogenetic analysis. This analysis
revealed a diverse evolutionary relationship for the hrp genes of the plant pathogenic

191
xanthomonads. The hypothesis of coevolution of the hrp region with the rest of the
genome from a common ancestor is supported in several cases. For instance, the
similarity in the hrpB and hrpC/D regions examined in this study for the groups A and
B of X. campestris pv. vesicatoria was less than 0.51. This genetic divergence of the
hrp genes is in agreement with the results obtained when the entire genome of these
two groups of X. campestris pv. vesicatoria were compared on the basis of DNA
homology (Stall et al., 1994). Although these pathogens cause similar diseases on
solanaceous plants, their hrp genes are genetically diverse to the same extent as the rest
of the genome. Furthermore, these comparisons do not support the contention that
there was a higher selective pressure for the hrp sequences than for other regions of the
genome nor that a horizontal movement of the hrp region of the genome may have
occurred between strains of the two groups of X. campestris pv. vesicatoria. Further
evidence to support the coevolution of the hrp genes and the rest of the genome
involves the pathovars X. campestris pv. carotae, X. campestris pv. gardneri, and X
campestris pv. pelargonii. Although these xanthomonads have very distinct host
ranges, they are genetically closely related based on DNA-DNA hybridization
(Hildebrand et al., 1989; Palleroni et al., 1993). In relation to the hrp genes they are
not only monophyletic but also very closely related. This example points to a
coevolution of the hrp genes and the rest of the genome from a common bacterial
ancestor for certain plant pathogenic xanthomonads and also substantiates the
divergence between hrp genes and host speciation. An important conclusion from the
phylogenetic studies is the divergence between hrp genes and host specificity. The hrp
genes are essential for the development of disease on compatible hosts and
hypersensitive reaction on both resistant host and nonhost plants (Willis et al., 1991).
Previous work has demonstrated the functional conservation of the hrp genes and the

192
lack of host speciation (Arlat et al. 1992; Bonas et al., 1991). Although the hrp genes
are necessary in the plant-pathogen interaction, other factors in the bacterial pathogen
are likely to be involved in host specialization (Fenselau et al., 1992; Gough et al.,
1992).
On the contrary, there is also indication of horizontal movement of the hrp
genes among plant pathogenic xanthomonads. This hypothesis is supported by two
lines of evidences, the presence of similar hrp sequences in strains with divergent
genetic background and variability in the relatedness between strains depending on the
hrp region examined. In the first case, the hrp genes of the strains ofX. campestris pv.
vesicatoria A were very closely related to the homologous sequences of the strains of
X. campestris pv. citrumelo. The estimated genetic relatedness between strains of these
two pathovars for both hrp-related regions examined ranged in all from 0.71 to 0.96.
Furthermore, the phylogenetic analysis strongly supports a monophyletic relationship
for the hrp genes of the strains of both pathovars. In contrast, DNA homology studies
indicated that the strains of these two pathovars are not closely related, with genomic
similarity usually below 0.58 (Egel, 1991).
Stronger evidence to support the contention of horizontal movement of the hrp
genes among the plant pathogenic xanthomonads is the variability in the relatedness
between strains for the two distinct hrp regions examined. For instance, the hrpC/D
region ofX. campestris pv. vignicola is phylogenetically monophyletic and closely
related to the homologous hrp region of a group of pathovars of X. campestris that
includes X. campestris pv. bilvae, X campestris pv. citri, X campestris pv. glycines A,
X. campestris pv. malvacearum, X. campestris pv. phaseoli B, X. campestris pv.
phaseoli "fuscans", and X campestris pv. vitians B. On the contrary, the hrpB region
of A! campestris pv. vignicola though monophyletic is closely related to the

193
homologous hrp region of X. campestris pv. begoniae, X. campestris pv.
dieffenbachiae, X. campestris pv. maculifoliigardeniae, X. campestris pv. manihotis,
and X. campestris pv. phaseoli A. Since there is no genetic and functional indication of
selection pressure regarding the hrp genes, the most likely hypothesis to explain the
source of the variability in the relatedness for these two regions of the genome of X.
campestris pv. vignicola remains in the origin of the hrpB and hrpC/D genes from
distinct ancestors. The coexistence in the same biological niche may have provided
opportunities for the lateral movement of this region of bacterial genome between
xanthomonads. The horizontal movement of genetic material between bacteria is a
common and important mechanism in bacterial evolution (Krawiec and Riley, 1990).
Furthermore, the resemblance at the protein level of the hrp genes of the xanthomonads
with genes involved in the secretion of pathogenicity factors of genetically distant
organisms, such as the animal pathogens of the genus Yersinia (Fenselau et al., 1992;
Gough et al., 1992), is also intriguing. If this means a convergent functional evolution
or a common bacterial ancestor remains to be clarified. Nevertheless, the phylogeny of
the hrp genes of plant pathogenic xanthomonads may provide a framework and a
rational basis through which origins and differentiation of the xanthomonads may be
assessed.
The classification of plant pathogenic bacteria at pathovar level was not based
initially on the genetic or other intrinsic characteristic of the organism but rather on the
host from which the bacteria were isolated (Bradbury, 1984; Dye et al., 1980).
Furthermore, comprehensive studies on the genetic and phenotypic characteristics have
supported the existence of a high degree of diversity among strains of a given pathovar
ofX. campestris. The genetic analysis of the /2/77-related sequences also indicates a
large diversity among the plant pathogenic xanthomonads. Furthermore, the diversity

194
or uniformity of the different taxa of xanthomonads assessed on the basis of restriction
analysis of hrp-related sequences apparently agrees very closely with the structure
established by using other methods. However, the restriction banding profiles
generated for the hrp-related fragments may be an easier and a more discriminating
approach for identification of plant pathogenic xanthomonads, compared to other
nucleic acid-based approaches, such as the genomic fingerprinting or restriction
fragment length polymorphism analysis by using random or specific DNA probes.
However, more extensive work is necessary to further characterize the hrp-related
sequence variation in other groups of plant pathogenic xanthomonads. Furthermore,
DNA-DNA hybridization data is also necessary to determine the consistency of the
groupings established on such a small region of the bacterial genome as the hrp genes.

APPENDIX A
BACTERIAL STRAINS AND PLASMIDS USED IN THIS STUDY
Table A-l. List of bacterial strains used in this study and their source.
Pathovar
Strain
Source3
Xanthomonas
X. campestris
pv. alfalfae
G-22 (KS)
DLS
82-1
RES
pv. armoraciae
63-27,417, 756
JBJ
pv begoniae
XCB9, X274, X281, X329, X610, X627,
XI490, XI492, XI496
ARC
pv. bilvae
XCB
ELC
pv. campestris
33913
ATCC
62-1, 62-9a, 65-6b, 70-3, 70-5, 71-2, 83-1, 83-
2
JBJ
pv. carotae
#3, #5, #7, #9, #12, #13, #16
RES
pv. celebensis
6207
ARC
pv. citri
DPI
canker A
3213, 3340, 9760-2, 9771, Tl, 115A
canker B
B64, B69, B80, B84, B93, B94, B148
canker C
70C, 338, 339, 340, 341,342
pv. citrumelo
FI, F6, F54, F59, F86, F94, F100, F228,
F254, F274, F306, F311, F348, F361, F378,
3166
DPI
pv. dieffenbachiae
X729
RES
Continued on following page
195

196
Table A-1 Continued
Pathovar
Strain
Source8
pv. dieffenbachiae
X260, X422, X736, X738, X739, X745,
X757, X763, X790, X1238, X1241, X1246,
X1272, XI292, X1298, X1426
ARC
pv. fici
X125, X151, X208, X212, X224, X548,
X687, X702
ARC
pv. gardneri
XG101
DCH
XV6, 1066
RES
pv. glycines
G-56, 86-16, 86-17, 86-18, 86-20, 87-2
RES
pv. holcicola
G-23
JEL
pv. incanae
9561-1, UK152, UK153
RES
pv.
macul i fol i igardeniae
X22j
DPI
pv. malvacearum
RIATC
RES
XcmA, XcmB, XcmC, XcmD, XcmE, XcmF,
XcmH, XcmN
DWG
pv. manihotis
Xml25D
RES
pv. papavericola
XP5
DCH
pv. pelargonii
XCP2, XCP10, XCP17, XCP36, XCP44,
XCP54, XCP58, XCP60
JBJ
pv. phaseoli
81-19, 82-1, 82-17, 85-6, EK11, XCPH4,
XP20
RES
pv. phaseoli "ftiscans"
XP163A
DCH
pv. physalidicola
XP172
DCH
pv. poinsettiicola
071-424, X87, X202, X259, X351, X352,
X354
RES
pv. pruni
X1219L, X1219S, X1220L, X1220S
ARC
pv. raphani
16B, 69-2, 69-4B, 69-8R
RES
pv. secalis
XC129C
RES
pv. translucens
80-1
RES
Continued on following page

197
Table A-lContinued
Pathovar
Strain
Source3
pv. taraxaci
XT11A
DCH
pv. vesicatoria
RES
group A
75-3, 82-4, 85-10, 87-21, 87-35T, 87-44T, 87-
48T, 89-8, 89-10, 90-20, 90-21, 90-27, 90-40,
90-60,91-66, 91-72, 92-11, 92-15, 92-16, 92-
17, 92-119, 1712, 6107, XV14, XV17
group B
141, 0226A, 0350A, 695, 853, 1062, B-3,B-
20, BA28-1, BV5-5, XV56
RES
group C
92-118, 92-120
RES
pv. vignicola
81-30, 82-38, G-55
RES
pv. vitians
ICPB101, ICPB164, XVIT, XV2, XV3,
X1215
RES
undetermined and
isolated from Feronia
XCF
ELC
sp.
undetermined and
isolated from Hibiscus
sp.
X10, X27, X52,
DWG
undetermined and
isolated from Strelitzia
reginae
X198
ARC
undetermined and
opportunistic
1,2, 2-8, 4, 7, 8, 9, 10, 11, 12, 17, 661, 663,
AAI, CB, Danny Gay, INA, INA42, INA69,
RG-1, S52, T-55, T-56, Toad Flax, TP78
RES
X. albilineans
91-065
JCC
X. fragariae
GC-6259, GC-6265, X1238, X1241, X1244,
XI246, XI292, X1298, X1426
ARC
X. maltophilia
RES
Continued on the following page

Table A-1 Continued
198
Pathovar Strain Source3
Acidovorax avenae
A. avenae subsp. avenae
UK142-A
JBJ
A. avenae subsp. citrulii UK20
JBJ
Agrobacterium
LBA1050
BJS
tumefaciens
Clavibacter michiganense
subsp. michiganense
69-1,75-1
RES
Erwinia
E. carotovora subsp.
K-SR-347, B-SR38
JAB
carotovora
E. herbicola
NF-33
RES
E. stewartii
SW2
DLC
Pseudomonas
P. solanacearum
K60
GM1/1000
AK
P. syringae
pv. syringae
INB
RES
pv. tomato
987
RES
Xylella fastidiosa
89-1
DLH
aAK, A. Kelman, University of Wisconsin, Madison, WI; ARC, A. R. Chase,
University of Florida, Apopka, FL; ATCC, American Type Culture Collection,
Rockville, MD; BJS, B. J. Staskawicz, University of California, Berkely, CA; DCH,
D. C. Hildebrand, University of California, Berkely, CA; DLC, D. L. Coplin, Ohio
State University, Columbus, OH; DLH, D. L. Hopkins, University of Florida,
Leesburg, FL; DLS, D. L. Stuteville, Kansas State University, Manhattan, KS; DPI,
Department of Plant Industry, Gainesville, FL; DWG, D. W. Gabriel, University of
Florida, Gainesville, FL; ELC, E. L. Civerolo, U.S. Department of Agriculture,
Beltsville, MD; JAB, J. A. Bartz, University of Florida, Gainesville, FL; JBJ, J. B.
Jones, University of Florida, Bradenton, FL; JCC, J. C. Comstock, U.S. Department
of Agriculture, Canal Point, FL; JEH, J. E. Hunter, Cornell University, Geneva, NY;
JEL, J. E. Leach, Kansas State University, Manhattan, KS; RES, R. E. Stall,
University of Florida, Gainesville, FL.

199
Table A-2. List of bacterial strains and plasmids used in molecular transformation
and conjugation in
this study.
Strain
Relevant characteristics
Source or reference3
Bacteria
Xanthomonas campestris
pv. vesicatoria
22
nonpathogenic Tn3-gus insertion mutant
of strain 85-10
Bonas et al., 1991
44
nonpathogenic Tn3-gs insertion mutant
of strain 85-10
Bonas et al., 1991
75
nonpathogenic Tn3-gus insertion mutant
of strain 85-10
Bonas et al., 1991
85
nonpathogenic Tn3-gi/.v insertion mutant
of strain 85-10
Bonas et al., 1991
137
nonpathogenic Tn3-gw.v insertion mutant
of strain 85-10
Bonas et al., 1991
318
nonpathogenic Tn3-gns insertion mutant
of strain 85-10
Bonas et al., 1991
Escherichia coli
DH5a
FTecA 80dlacZM15
BRL
HB101
F'recA
BRL
Plasmids
pLAFR3
Tetrr/x+ RK2replicon
Staskawicz et al., 1987
pBluescript-KS+
Ampr, Bluescript
Stratagene
pRK2013
KmrTraRK2+Mob+ColEl replicn
Figurski and Helinski,
1979
pXV9
pLAFR3 cosmid clone from X. c. pv.
vesicatoria 75-3
Bonas et al., 1991
pXV847
pBluescript clone containing internal Sau
3 A hrp fragment amplified from X. c. pv.
vesicatoria 75-3
This study
pXVllll
pBluescript clone containing internal Sau
3 A hrp fragment amplified from X. c. pv.
vesicatoria 75-3
This study
pXV5.1
pLAFR3 Eco RI subclone from pXV9
This study
pXV5.5
pLAFR3 Eco Rl subclone from pXV9
This study
Continued on the following page

200
Table A-2Continued
Strain
Relevant characteristics
Source or reference
pXV56/3-48
pLARF3 cosmid clone from X. c. pv.
vesicatoria XV56
This study
pXCP58/2
pLARF3 cosmid clone from X. c. pv.
pelargonii
This study
aBRL, Bethesda Research Laboratories, Gaithersburg, MD; Stratagene, Stratagene, Inc., La Jolla, CA.

APPENDIX B
SAS PROGRAM TO ESTIMATE THE SIMILARITY VALUES
FROM RESTRICTION FRAGMENT DATA
options ls=80 ps=54 nodate nonumber;
data dnatype;
infile bands lrecl=835;
input bacteria $ bandl -band 100; /* Read data from file Bacteria and set number of
*/
/* fragments compared */
proc sort data=dna_type;
by bacteria;
data dna type;
set dna type;
bactl=_n_;
bact2=_n_;
proc transpose data=dna_type out=transout; /* Create a new set with rows of the */
var band 1 -band 100; /* original data matrix becoming
*/
/* columns, and columns becoming */
/* rows */
data coefmtx;
coef=;
bactl=;
bact2=;
%macro calccoef;
%do i=l %to 3;
%do j=%eval (&i+l) %to 4;
/* Invoke the macro processor and execute */
/* the calculations until j = # of strains to be */
/* compared; i =j -1 */
201

data temp; /* Compare fragment data between two */
set transout (keep=col&i col&j); /* strains and create a new data set */
columnl=col&i;
column2=col&j;
if columnl=l & column2=l then prod=l;
else prod=0;
keep column 1 column2 prod;
proc means noprint data=temp;
var prod column 1 column2;
output out=outset sum=prodsum coll sum col2sum;
data outset; /* Determine the similarity value */
set outset; /* between two strains */
coef=(2*prodsum) / (collsum+col2sum);
bactl=&i;
bact2=&j;
keep coefbactl bact2;
data coefmtx;
set coefmtx outset;
%end;
%end;
%mend calccoef;
%calccoef
data coefmtx;
set coefmtx;
if bactl=. then delete;
proc sort data=coefmtx;
by bact2;
data coefmtx;
merge coefmtx (in=i) dna type (keep=bact2 bacteria);
by bact2;
if i= 1;
bacter2=bacteria;
drop bacteria;

203
proc sort data=coefmtx; /* Sort the result by strain name */
by bactl;
data coefmtx;
merge coefmtx (in=i) dna type (keep=bactl bacteria);
by bactl;
if i=l;
bacterl=bacteria;
drop bacteria;
label bacterl-'Strain" bacter2-'Strain"
coef=" strain";
proc tabulate f=9.4 formchar=' '; /* Print out the result */
title "Similarity values for bacteria";
class bacterl bacter2;
var coef;
table bacterl,bacter2*coef;
key label sum-' ";
run;

APPENDIX C
SAS PROGRAM TO ESTIMATE THE NUCLEOTIDE
SUBSTITUTION BASED ON RESTRICTION FRAGMENT DATA
data frag;
infile fragment; /* Read data of similarity values from */
input similar @@; /* file fragment */
g0=similar**0.25;
g=(similar*(3-(2*g0)))**0.25;
gl=G0;
x=l;
do while (gl ne g);
gi=g;
if x ne 50 then do;
g=(similar*(3-(2*gl)))**0.25;
x=x+l;
end;
end;
do differe=-((2/r)*(log(g)));
end;
output;
/* Determine the probability that */
/* a restriction site has remained */
/* unaltered */
/* Determine the nucleotide substitution; */
/* r is the number of nucleotides in the */
/* recognition site of the restriction */
/* endonuclease */
proc print split-/* Print out the result */
var similar gO g differe x;
label similar='Similarity'
gO='Initial*probability'
g='Final*probability'
differe-N ucleotide substitution'
x-N umber* iteration';
title 'DNA similarity and nucleotide substitution';
title2 'Nei and Li (1979) procedure';
run;
204

APPENDIX D
ESTIMATES OF SIMILARITY VALUES FOR PLANT PATHOGENIC XANTHOMONADS BASED ON
COMPARISON OF RESTRICTION FRAGMENT DATA OF THE hrp RELATED DNA SEQUENCES
Table D-l. Similarity values for plant pathogenic strains of different pathovars of Xanthomonas campestris and X. fragariae
based on the genetic analysis of the hrpC/D related DNA sequences.
Species/Pathovar
Species/Pathovar
2
3
4
5
6
7
8
9
10
11
1. pv. alfalfae
.9583
.3182
.5000
.5217
.5217
.7083
.3182
.3636
.3182
.5833
2. pv. alfalfae
.3182
.5000
.5217
.5217
.6667
.3182
.3636
.3182
.5833
3. pv. armoraciae
.2500
.3333
.3810
.3182
1.000
.8500
1.000
.3636
4. pv. begonia A
.
.9048
.8571
.3636
.2500
.3000
.2500
.3636
5. pv. begonia B
.
.9545
.3913
.3333
.3333
.3333
.3913
6. pv. begoniae C
.3913
.3810
.3810
.3810
.3913
7. pv. bilvae
.
.3182
.4091
.3182
.7917
8. pv. campestris
.
.
.8500
1.000
.3636
9. pv. carotae A
.
.
.
.8500
.4545
10. pv. carotae B
.
.
.
.
.3636
11. pv. citri A




Continued on following page

Table D-lContinued
Species/Pathovar Species/Pathovar
12
13
14
15
1. pv. alfalfae
.5000
.5000
.8163
.9388
2. pv. alfalfae
.5000
.5000
.7755
.8980
3. pv. armoraceae
.3182
.3182
.3111
.3111
4. pv. begoniae A
.1818
.1818
.3556
.4889
5. pv. begoniae A
.2174
.2174
.3830
.5532
6. pv. begonia B
.2609
.2609
.4255
.5532
7. pv. bilvae
.7083
.7083
.6531
.7347
8. pv. campestris
.3182
.3182
.3111
.3111
9. pv. carotae A
.4091
.4091
.3556
.3556
10. pv. carotae B
.3182
.3182
.3111
.3111
11. pv. citri A
.7500
.7500
.4898
.6122
12. pv. citri B
1.000
.5306
.5306
13. pv. citri C
.5306
.5306
14. pv. citrumelo
15. pv. citrumelo
16. pv. citrumelo
17. pv. citrumelo
18. pv. citrumelo
19. pv. citrumelo
20. pv. citrumelo
21. pv. citrumelo
.7600
16
17
18
19
20
21
.8333
.8333
.9167
.8511
.7083
.8333
.7917
.7917
.8750
.8936
.6667
.7917
.2727
.2727
.2727
.2791
.2727
.3182
.5455
.5455
.5909
.6047
.5000
.5455
.5217
.5217
.5652
.5778
.4783
.6087
.5217
.5217
.5652
.5778
.5217
.6522
.6667
.6667
.7083
.6383
.6250
.6667
.2727
.2727
.2727
.2791
.2727
.3182
.3182
.3182
.3182
.3256
.3182
.3636
.2727
.2727
.2727
.2791
.2727
.3182
.5833
.5833
.5833
.5957
.4583
.5833
.4583
.4583
.4583
.4255
.5833
.5000
.4583
.4583
.4583
.4255
.5833
.5000
.7755
.7755
.7347
.6667
.6939
.6939
.8163
.8163
.8980
.8333
.6939
.8980
1.000
.9167
.8936
.7917
.7500
.9167
.8936
.7917
.7500
.
.9362
.7917
.7917
.7234
.7660
.7500
Continued on following page

Table D-lContinued
Species/Pathovar Species/Pathovar
22
23
24
25
26
27
28
29
30
31
1. pv. alfalfae
.8750
.4167
.4167
.6818
.9167
.7755
.7755
.2727
.3636
.5833
2. pv. alfalfae
.8333
.4167
.4167
.6818
.8750
.7347
.7347
.2727
.3636
.5833
3. pv. armoraceae
.2727
.1818
.1818
.2500
.2727
.3556
.3556
.4000
.9000
.3636
4. pv. begoniae A
.5455
.6364
.6818
.7500
.5909
.3556
.3556
.2000
.3500
.3636
5. pv. begoniae A
.5652
.6957
.7391
.7143
.5652
.3830
.3830
.2381
.3333
.3913
6. pv. begonia B
.5652
.6522
.6957
.7143
.5652
.4255
.4255
.2857
.3810
.3913
7. pv. bilvae
.6667
.3333
.3333
.5000
.7083
.6939
.6939
.3636
.3182
.7917
8. pv. campestris
.2727
.1818
.1818
.2500
.2727
.3556
.3556
.4000
.9000
.3636
9. pv. carotae A
.3182
.2273
.2273
.3000
.3182
.4000
.4000
.4500
.8500
.4545
10. pv. carotae B
.2727
.1818
.1818
.2500
.2727
.3556
.3556
.4000
.9000
.3636
11. pv. citri A
.5417
.2917
.2500
.5000
.5833
.5306
.5306
.2273
.3636
1.000
12. pv. citri B
.4583
.2500
.2083
.3182
.4583
.6939
.6939
.3182
.3182
.7500
13. pv. citri C
.4583
.2500
.2083
.3182
.4583
.6939
.6939
.3182
.3182
.7500
14. pv. citrumelo
.8163
.4898
.4898
.4889
.7347
.8400
.8400
.3111
.3556
.4898
15. pv. citrumelo
.8163
.4490
.4490
.6667
.8980
.8000
.8000
.2667
.3556
.6122
16. pv. citrumelo
.7083
.4167
.4167
.6818
.9167
.7347
.7347
.2273
.3636
.5833
17. pv. citrumelo
.7083
.4167
.4167
.6818
.9167
.7347
.7347
.2273
.3636
.5833
18. pv. citrumelo
.7917
.4583
.4583
.7273
1.000
.7347
.7347
.2273
.3636
.5833
19. pv. citrumelo
.7234
.4255
.4255
.7442
.9362
.6667
.6667
.2326
.3721
.5957
20. pv. citrumelo
.7083
.4167
.4583
.6364
.7917
.8163
.8163
.2727
.3636
.4583
21. pv. citrumelo
.8333
.4167
.4583
.7273
.7917
.7347
.7347
.3182
.3636
.5833
22. pv. citrumelo
.
.5000
.5417
.6818
.7917
.7347
.7347
.2727
.3182
.5417
Continued on following page
K>
o

Table D-lContinued
Species/Pathovar
Species/Pathovar
22
23
24
25
26
27
28
29
30
31
23. pv. dieffenbachiae A
.9583
.5000
.4583
.4082
.4082
.1364
.2273
.2917
24. pv. dieffenbachiae A
.5455
.4583
.4082
.4082
.1818
.2273
.2500
25. pv. dieffenbachiae B
.7273
.4889
.4889
.2000
.3500
.5000
26. pv. fici A
.7347
.7347
.2273
.3636
.5833
27. pv. fici A
1.000
.3111
.4000
.5306
28. pv. fici A
.3111
.4000
.5306
29. pv. fici B
.
.
.4000
.2273
30. pv. gardneri
.
.
.3636
31. pv. glycines A
.
.
Continued on following page
to
o
oo

Table D-lContinued
Species/Pathovar Species/Pathovar
32
33
34
35
36
37
38
39
40
41
1. pv. alfalfae
.2727
.1667
.3182
.5217
.4889
.5306
.4000
.3556
.4681
.5833
2. pv. alfalfae
.2727
.1667
.3182
.5217
.4889
.5306
.4000
.3556
.4681
.5833
3. pv. armoraceae
.4000
.3182
1.000
.3333
.3415
.2667
.8780
.9268
.1860
.3636
4. pv. begoniae A
.2000
.1818
.2500
.8095
.3415
.8444
.3415
.2927
.7442
.3636
5. pv. begoniae A
.2381
.1739
.3333
.9091
.3721
.8936
.4186
.3721
.8000
.3913
6. pv. begonia B
.2857
.2174
.3810
.9091
.3721
.8511
.4651
.4186
.7556
.3913
7. pv. bilvae
.3636
.1250
.3182
.4348
.7111
.4082
.4000
.4000
.3830
.7917
8. pv. campestris
.4000
.3182
1.000
.3333
.3415
.2667
.8780
.9268
.1860
.3636
9. pv. carotae A
.4500
.3182
.8500
.3333
.4390
.3111
.8780
.9268
.2326
.4545
10. pv. carotae B
.4000
.3182
1.000
.3333
.3415
.2667
.8780
.9268
.1860
.3636
11. pv. citri A
.2273
.2500
.3636
.3478
.8000
.4082
.4444
.4444
.2979
1.000
12. pv. citri B
.3182
.1667
.3182
.2609
.6667
.2449
.4000
.4000
.2128
.7500
13. pv. citri C
.3182
.1667
.3182
.2609
.6667
.2449
.4000
.4000
.2128
.7500
14. pv. citrumelo
.3111
.1633
.3111
.4255
.4348
.4800
.3478
.3478
.5417
.4898
15. pv. citrumelo
.2667
.1633
.3111
.5532
.5217
.5600
.3913
.3478
.5000
.6122
16. pv. citrumelo
.2273
.1667
.2727
.4783
.4444
.5306
.3111
.3111
.4681
.5833
17. pv. citrumelo
.2273
.1667
.2727
.4783
.4444
.5306
.3111
.3111
.4681
.5833
18. pv. citrumelo
.2273
.1667
.2727
.5652
.4889
.5714
.3556
.3111
.5106
.5833
19. pv. citrumelo
.2326
.1702
.2791
.5333
.4545
.5833
.3636
.3182
.4783
.5957
20. pv. citrumelo
.2727
.1667
.2727
.5217
.4000
.4898
.3111
.3111
.4681
.4583
21. pv. citrumelo
.3182
.2083
.3182
.5652
.4444
.6122
.4000
.3556
.5106
.5833
22. pv. citrumelo
.2727
.1667
.2727
.5652
.4444
.6531
.3556
.3111
.5957
.5417
Continued on following page
209

Table D-lContinued
Species/Pathovar
Species/Pathovar
32
33
34
35
36
37
38
39
40
41
23. pv. dieffenbachiae A
.1364
.1250
.1818
.7391
.3111
.8163
.2667
.2222
.8936
.2917
24. pv. dieffenbachiae A
.1818
.1250
.1818
.7826
.2667
.8571
.2667
.2222
.9362
.2500
25. pv. dieffenbachiae B
.2000
.2273
.2500
.6667
.3415
.7111
.3415
.2927
.6047
.5000
26. pv. fici A
.2273
.1667
.2727
.5652
.4889
.5714
.3556
.3111
.5106
.5833
27. pv. fici A
.3111
.1633
.3556
.4255
.4783
.4400
.3913
.3913
.4167
.5306
28. pv. fici A
.3111
.1633
.3556
.4255
.4783
.4400
.3913
.3913
.4167
.5306
29. pv. fici B
1.000
.1818
.4000
.2381
.2439
.2222
.3902
.4390
.1860
.2273
30. pv. gardneri
.4000
.3182
.9000
.3333
.3415
.3111
.8293
.8780
.2326
.3636
31. pv. glycines A
.2273
.2500
.3636
.3478
.8000
.4082
.4444
.4444
.2979
1.000
32. pv. glycines B
.1818
.4000
.2381
.2439
.2222
.3902
.4390
.1860
.2273
33. pv. holcicola
.3182
.1304
.1778
.1633
.3111
.3111
.1277
.2500
34. pv. incanae
.3333
.3415
.2667
.8780
.9268
.1860
.3636
35. pv. maculifoliigardeniae
.4186
.8085
.4186
.3721
.8000
.3478
36. pv. malvacearum
.3478
.4286
.4286
.3182
.8000
37. pv. manihotis
.
.3478
.3043
.8750
.4082
38. pv. papavericola
.
.
.9524
.2727
.4444
39. pv. pelargonii
.
.
.2273
.4444
40. pv. phaeoli A
.
.

.
.2979
41. pv. phaseoli B




Continued on following page

Table D-1 Continued
Species/Pathovar Species/Pathovar
42
44
45
46
47
48
49
50
51
52
1. pv. alfalfae
.5217
.3256
.9167
.9388
.4091
.4091
.3636
.7917
.7660
.7660
2. pv. alfalfae
.5217
.3256
.8750
.8980
.4091
.4091
.3636
.7500
.7234
.7234
3. pv. armoraceae
.2381
.4615
.2727
.3111
.4500
.4500
.8500
.2273
.2326
.2326
4. pv. begoniae A
.2381
.2564
.5909
.4889
.3500
.3500
.3500
.5000
.5116
.5116
5. pv. begoniae A
.2727
.2927
.5652
.5532
.3810
.3810
.3810
.4783
.5333
.5333
6. pv. begonia B
.2727
.3415
.5652
.5532
.3810
.3810
.4286
.4783
.5333
.5333
7. pv. bilvae
.6957
.3721
.7083
.7347
.4091
.4091
.3182
.6250
.6383
.6383
8. pv. campestris
.2381
.4615
.2727
.3111
.4500
.4500
.8500
.2273
.2326
.2326
9. pv. carotae A
.3333
.5128
.3182
.3556
.5000
.5000
.8000
.2727
.2791
.2791
10. pv. carotae B
.2381
.4615
.2727
.3111
.4500
.4500
.8500
.2273
.2326
.2326
11. pv. citri A
.8261
.2791
.5833
.6122
.4091
.4091
.3636
.5417
.5532
.5532
12. pv. citri B
.8261
.3256
.4583
.5306
.3636
.3636
.3182
.4583
.4681
.4681
13. pv. citri C
.8261
.3256
.4583
.5306
.3636
.3636
.3182
.4583
.4681
.4681
14. pv. citrumelo
.5106
.3182
.7347
.7600
.3556
.3556
.3556
.7755
.7500
.7500
15. pv. citrumelo
.5532
.3182
.8980
1.000
.4000
.4000
.3556
.7755
.7500
.7500
16. pv. citrumelo
.5652
.2791
.9167
.8163
.3182
.3182
.3636
.9583
.8511
.8511
17. pv. citrumelo
.5652
.2791
.9167
.8163
.3182
.3182
.3636
.9583
.8511
.8511
18. pv. citrumelo
.4783
.2791
1.000
.8980
.3636
.3636
.3636
.8750
.7660
.7660
19. pv. citrumelo
.4889
.2857
.9362
.8333
.3256
.3256
.3721
.8511
.7391
.7391
20. pv. citrumelo
.3913
.3256
.7917
.6939
.3636
.3636
.3636
.7500
.7660
.7660
21. pv. citrumelo
.5217
.3721
.7917
.8980
.3636
.3636
.3636
.7083
.8085
.8085
22. pv. citrumelo
.4783
.3256
.7917
.8163
.4091
.4091
.3182
.7083
.8085
.8085
Continued on following page

Table D-lContinued
Species/Pathovar Species/Pathovar
42
44
45
46
47
48
49
50
51
52
23. pv. dieffenbachiae A
.2609
.1395
.4583
.4490
.2727
.2727
.2273
.4167
.4255
.4255
24. pv. dieffenbachiae A
.2174
.1860
.4583
.4490
.3182
.3182
.2273
.4167
.4681
.4681
25. pv. dieffenbachiae B
.3810
.2564
.7273
.6667
.4000
.4000
.3500
.6364
.6977
.6977
26. pv. fici A
.4783
.2791
1.000
.8980
.3636
.3636
.3636
.8750
.7660
.7660
27. pv. fici A
.5106
.3636
.7347
.8000
.4000
.4000
.4000
.7347
.7083
.7083
28. pv. fici A
.5106
.3636
.7347
.8000
.4000
.4000
.4000
.7347
.7083
.7083
29. pv. fici B
.2381
.7692
.2273
.2667
.4000
.4000
.4000
.2273
.2791
.2791
30. pv. gardneri
.2381
.4615
.3636
.3556
.4500
.4500
.9000
.3182
.2791
.2791
31. pv. glycines A
.8261
.2791
.5833
.6122
.4091
.4091
.3636
.5417
.5532
.5532
32. pv. glycines B
.2381
.7692
.2273
.2667
.4000
.4000
.4000
.2273
.2791
.2791
33. pv. holcicola
.1739
.2326
.1667
.1633
.2273
.2273
.2727
.1250
.1277
.1277
34. pv. incanae
.2381
.4615
.2727
.3111
.4500
.4500
.8500
.2273
.2326
.2326
35. pv. maculifoliigardeniae
.2273
.2927
.5652
.5532
.4286
.4286
.3810
.4348
.4889
.4889
36. pv. malvacearum
.6512
.2500
.4889
.5217
.3902
.3902
.3415
.4000
.4091
.4091
37. pv. manihotis
.2979
.2727
.5714
.5600
.3556
.3556
.3111
.5306
.5833
.5833
38. pv. papavericola
.3256
.5000
.3556
.3913
.4878
.4878
.7805
.2667
.2727
.2727
39. pv. pelargonii
.3256
.5000
.3111
.3478
.4878
.4878
.8293
.2667
.2727
.2727
40. pv. phaeoli A
.2667
.1905
.5106
.5000
.3256
.3256
.2326
.4681
.5217
.5217
41. pv. phaseoli B
.8261
.2791
.5833
.6122
.4091
.4091
.3636
.5417
.5532
.5532
42. pv. phaseoli "fuscans"
,
.2439
.4783
.5532
.2857
.2857
.2381
.5652
.5778
.5778
44. pv. poinsettiicola B


.2791
.3182
.5128
.5128
.4615
.2791
.3333
.3333
Continued on following page

Table D-lContinued
Species/Pathovar
Species/Pathovar
42
44
45
46
47
48
49
50
51
52
45. pv. poinsettiicola A
.8980
.3636
.3636
.3636
.8750
.7660
.7660
46. pv. pruni
.4000
.4000
.3556
.7755
.7500
.7500
47. pv. raphani A
1.000
.4500
.2727
.3256
.3256
48. pv. raphani B
.4500
.2727
.3256
.3256
49. pv. taraxaci
.3182
.2791
.2791
50. pv. vesicatoria A
.
.8936
.8936
51. pv. vesicatoria A
.
.
.
.
1.000
52. pv. vesicatoria A




Continued on following page

Table D-lContinued
Species/Pathovar Species/Pathovar
53
54
55
56
57
58
59
60
61
1. pv. alfalfae
.4583
.5000
.4898
.4889
.3636
.9167
.7083
.3256
.2083
2. pv. alfalfae
.4583
.5000
.4898
.4889
.3636
.8750
.6667
.3256
.2083
3. pv. armoraceae
.5000
.3182
.2667
.3415
.8500
.2727
.3182
.4615
.4545
4. pv. begoniae A
.4091
.1818
.8000
.3415
.3500
.5909
.3636
.2564
.0909
5. pv. begoniae A
.4348
.2174
.8511
.3721
.3333
.5652
.3913
.2927
.1739
6. pv. begonia B
.4348
.2609
.8085
.3721
.3810
.5652
.3913
.3415
.2174
7. pv. bilvae
.4167
.7083
.4082
.7111
.3636
.7083
.9583
.3721
.2500
8. pv. campestris
.5000
.3182
.2667
.3415
.8500
.2727
.3182
.4615
.4545
9. pv. carotae A
.5909
.4091
.3111
.4390
.9000
.3182
.4091
.5128
.3636
10. pv. carotae B
.5000
.3182
.2667
.3415
.8500
.2727
.3182
.4615
.4545
11. pv. citri A
.5000
.7500
.4082
.8000
.4091
.5833
.8333
.2791
.2917
12. pv. citri B
.4167
1.000
.2857
.6667
.3636
.4583
.7083
.3256
.3333
13. pv. citri C
.4167
1.000
.2857
.6667
.3636
.4583
.7083
.3256
.3333
14. pv. citrumelo
.3265
.5306
.5600
.4348
.3556
.7347
.6531
.3182
.2449
15. pv. citrumelo
.4490
.5306
.5200
.5217
.3556
.8980
.7347
.3182
.2041
16. pv. citrumelo
.3750
.4583
.6122
.4444
.3636
.9167
.6667
.2791
.1667
17. pv. citrumelo
.3750
.4583
.6122
.4444
.3636
.9167
.6667
.2791
.1667
18. pv. citrumelo
.4167
.4583
.5306
.4889
.3636
1.000
.7083
.2791
.1667
19. pv. citrumelo
.4255
.4255
.5417
.4545
.3721
.9362
.6383
.2857
.1702
20. pv. citrumelo
.3333
.5833
.5306
.4000
.3636
.7917
.6250
.3256
.2083
21. pv. citrumelo
.4583
.5000
.5714
.4444
.3636
.7917
.6667
.3721
.2500
22. pv. citrumelo
.4167
.4583
.6122
.4444
.3182
.7917
.6667
.3256
.2083
Continued on following page
214

Table D-lContinued
Species/Pathovar Species/Pathovar
23. pv. dieffenbachiae A
24. pv. dieffenbachiae A
25. pv. dieffenbachiae B
26. pv. fici A
27. pv. fici A
28. pv. fici A
29. pv. fici B
30. pv. gardneri
31. pv. glycines A
32. pv. glycines B
33. pv. holcicola
34. pv. incanae
35. pv. maculifoliigardeniae
36. pv. malvacearum
37. pv. manihotis
38. pv. papavericola
39. pv. pelargonii
40. pv. phaeoli A
41. pv. phaseoli B
42. pv. phaseoli "fuscans"
44. pv. poinsettiicola B
45. pv. poinsettiicola A
53 54 55
.2500
.2500
.7755
.2917
.2083
.8163
.4545
.3182
.6667
.4167
.4583
.5306
.3265
.6939
.4800
.3265
.6939
.4800
.3182
.3182
.2222
.5455
.3182
.3111
.5000
.7500
.4082
.3182
.3182
.2222
.3750
.1667
.1633
.5000
.3182
.2667
.4348
.2609
.7660
.4444
.6667
.3478
.4082
.2449
.9200
.6222
.4000
.3043
.5778
.4000
.3043
.3404
.2128
.8750
.5000
.7500
.4082
.4348
.8261
.3830
.4186
.3256
.2727
.4167
.4583
.5306
56 57 58
.3111
.2273
.4583
.2667
.2273
.4583
.3415
.3500
.7273
.4889
.3636
1.000
.4783
.4000
.7347
.4783
.4000
.7347
.2439
.4000
.2273
.3415
.9500
.3636
.8000
.4091
.5833
.2439
.4000
.2273
.1778
.3182
.1667
.3415
.8500
.2727
.4186
.3333
.5652
1.000
.3902
.4889
.3478
.3111
.5714
.4286
.8780
.3556
.4286
.9268
.3111
.3182
.2326
.5106
.8000
.4091
.5833
.6512
.2857
.4783
.2500
.4615
.2791
.4889
.3636
1.000
59 60 61
.3333
.1395
.0833
.3333
.1860
.0833
.5000
.2564
.1364
.7083
.2791
.1667
.6939
.3636
.2449
.6939
.3636
.2449
.3182
.7692
.3636
.3182
.4615
.3636
.8333
.2791
.2917
.3182
.7692
.3636
.1667
.2326
.2083
.3182
.4615
.4545
.4348
.2927
.1739
.7556
.2500
.2667
.4082
.2727
.1224
.4000
.5000
.4000
.4000
.5000
.4444
.3830
.1905
.0851
.8333
.2791
.2917
.6957
.2439
.2609
.3256
1.000
.3256
.7083
.2791
.1667
Continued on following page

Table D-lContinued
Species/Pathovar Species/Pathovar
46. pv. pruni
47. pv. raphani A
48. pv. raphani B
49. pv. taraxaci
50. pv. vesicatoria A
51. pv. vesicatoria A
52. pv. vesicatoria A
53. pv. vesicatoria B
54. pv. vignicola
55. pv. vitians A
56. pv. vitians B
57. pv. vitians C
58. X. campe sir is XI98
59. X. campestris XCF
60. X. campestris X52
61.X. fragariae
53 54 55
.4490
.5306
.5200
.4091
.3636
.3556
.4091
.3636
.3556
.5000
.3182
.3111
.3333
.4583
.6122
.3830
.4681
.6667
.3830
.4681
.6667
.4167
.3673
.2857
56 57 58
.5217
.3556
.8980
.3902
.4500
.3636
.3902
.4500
.3636
.3415
.8500
.3636
.4000
.3182
.8750
.4091
.2791
.7660
.4091
.2791
.7660
.4444
.5909
.4167
.6667
.3636
.4583
.3478
.3111
.5306
.3902
.4889
.3636
59
60
61
.7347
.3182
.2041
.4545
.5128
.3182
.4545
.5128
.3182
.3182
.4615
.3636
.6250
.2791
.1667
.6383
.3333
.2128
.6383
.3333
.2128
.4167
.4186
.2500
.7083
.3256
.3333
.4082
.2727
.1224
.7556
.2500
.2667
.3636
.4615
.3636
.7083
.2791
.1667
.
.3256
.2500
.3256
216

Table D-2. Similarity values for plant pathogenic strains of different pathovars of Xanthomonas campestris and X. fragariae
based on the genetic analysis of the hrpB related DNA sequences.
Species/Pathovar
Species/Pathovar
2 3
4
5
6
7
8
11
12
13
1. pv. alfalfae
1.000 .4091
.6341
.6341
.6341
.4889
.4545
.4091
.4091
.4091
2. pv. alfalfae
.4091
.6341
.6341
.6341
.4889
.4545
.4091
.4091
.4091
3. pv. armoraciae
.4878
.4878
.4878
.5778
.9545
.5455
.5455
.5455
4. pv. begoniae A
1.000
1.000
.6190
.5366
.5366
.5854
.5854
5. pv. begoniae A
1.000
.6190
.5366
.5366
.5854
.5854
6. pv. begoniae B
.6190
.5366
.5366
.5854
.5854
7. pv. bilvae
.
.
.5778
.8444
.9333
.9333
8. pv. campestris
.
.
.
.5455
.5455
.5455
11. pv. citri A
.
.8182
.8182
12. pv. citri B
.
.
.
.
1.000
13. pv. citri C





Continued on following page

Table D-2Continued
Species/Pathovar
Species/Pathovar
14
15
16
17
18
19
20
21
22
23
1. pv. alfalfae
.9333
.8696
.8696
.9333
.9333
.9545
.8696
1.000
.8261
.6190
2. pv. alfalfae
.9333
.8696
.8696
.9333
.9333
.9545
.8696
1.000
.8261
.6190
3. pv. armoraciae
.4444
.5217
.5217
.4889
.4444
.4091
.5217
.4091
.5217
.5238
4. pv. begoniae A
.6667
.7442
.7442
.7143
.6667
.6341
.7442
.6341
.7442
.9231
5. pv. begoniae A
.6667
.7442
.7442
.7143
.6667
.6341
.7442
.6341
.7442
.9231
6. pv. begoniae B
.6667
.7442
.7442
.7143
.6667
.6341
.7442
.6341
.7442
.9231
7. pv. bilvae
.5217
.5957
.5957
.5652
.5217
.4889
.5957
.4889
.5957
.6047
8. pv. campestris
.4889
.5652
.5652
.5333
.4889
.4545
.5652
.4545
.5652
.5714
11. pv. citri A
.4444
.5217
.5217
.4889
.4444
.4091
.5217
.4091
.5217
.5714
12. pv. citri B
.4444
.5217
.5217
.4889
.4444
.4091
.5217
.4091
.5217
.5714
13. pv. citri C
.4444
.5217
.5217
.4889
.4444
.4091
.5217
.4091
.5217
.5714
14. pv. citrumelo
.9362
.9362
.8696
1.000
.8889
.9362
.9333
.8936
.6512
15. pv. citrumelo
1.000
.9362
.9362
.8261
1.000
.8696
.9583
.7273
16. pv. citrumelo
.9362
.9362
.8261
1.000
.8696
.9583
.7273
17. pv. ctirumelo
.8696
.8889
.9362
.9333
.8936
.6977
18. pv. citrumelo
.8889
.9362
.9333
.8936
.6512
19. pv. citrumelo
.
.8261
.9545
.8696
.6190
20. pv. citrumelo
.
.8696
.9583
.7273
21. pv. citrumelo
#
.
.8261
.6190
22. pv. citrumelo
.
#
.7273
23. pv. dieffenbachiae A



Continued on following page

Table D-2Continued
Species/Pathovar Species/Pathovar
24
25
26
27
28
29
31
34
35
36
1. pv. alfalfae
.6190
.7111
.8696
.9333
.9333
.3902
.4545
.4545
.5854
.4651
2. pv. alfalfae
.6190
.7111
.8696
.9333
.9333
.3902
.4545
.4545
.5854
.4651
3. pv. armoraciae
.5238
.5778
.5217
.4444
.4889
.6341
.5455
.9545
.5366
.5116
4. pv. begoniae A
.9231
.5714
.7442
.6667
.7143
.5789
.5366
.5366
.8421
.5500
5. pv. begoniae A
.9231
.5714
.7442
.6667
.7143
.5789
.5366
.5366
.8421
.5500
6. pv. begoniae B
.9231
.5714
.7442
.6667
.7143
.5789
.5366
.5366
.8421
.5500
7. pv. bilvae
.6047
.5217
.5957
.5217
.5652
.4286
.8444
.5778
.6190
.8182
8. pv. campestris
.5714
.5333
.5652
.4889
.5333
.6829
.5455
1.000
.5854
.5116
11. pv. citri A
.5714
.4889
.5217
.4444
.4889
.4390
.9091
.5455
.5854
.8837
12. pv. citri B
.5714
.4889
.5217
.4444
.4889
.3902
.8182
.5455
.5854
.7907
13. pv. citri C
.5714
.4889
.5217
.4444
.4889
.3902
.8182
.5455
.5854
.7907
14. pv. citrumelo
.6512
.7826
.9362
1.000
.8696
.4286
.4889
.4889
.6190
.5000
15. pv. citrumelo
.7273
.8511
1.000
.9362
.9362
.5116
.5652
.5652
.6977
.5778
16. pv. citrumelo
.7273
.8511
1.000
.9362
.9362
.5116
.5652
.5652
.6977
.5778
17. pv. ctirumelo
.6977
.7826
.9362
.8696
1.000
.4762
.5333
.5333
.6667
.5455
18. pv. citrumelo
.6512
.7826
.9362
1.000
.8696
.4286
.4889
.4889
.6190
.5000
19. pv. citrumelo
.6190
.6667
.8261
.8889
.8889
.4390
.4545
.4545
.5854
.4186
20. pv. citrumelo
.7273
.8511
1.000
.9362
.9362
.5116
.5652
.5652
.6977
.5778
21. pv. citrumelo
.6190
.7111
.8696
.9333
.9333
.3902
.4545
.4545
.5854
.4651
22. pv. citrumelo
.7273
.8085
.9583
.8936
.8936
.5581
.5652
.5652
.6977
.5333
Continued on following page

Table D-2Continued
Species/Pathovar
Species/Pathovar
24
25
26
27
28
29
31
34
35
36
23. pv. dieffenbachiae A
1.000
.6047
.7273
.6512
.6977
.6154
.5714
.5714
.9231
.5854
24. pv. dieffenbachiae A
.6047
.7273
.6512
.6977
.6154
.5714
.5714
.9231
.5854
25. pv. dieffenbachiae B
.8511
.7826
.7826
.4286
.5333
.5333
.6190
.5455
26. pv. fici A
.9362
.9362
.5116
.5652
.5652
.6977
.5778
27. pv. fici A
.8696
.4286
.4889
.4889
.6190
.5000
28 pv. fici A
.
.4762
.5333
.5333
.6667
.5455
29. pv. fici B
.
.4390
.6829
.5789
.4500
31. pv. glycines A
.
.5455
.5854
.9302
34. pv. incanae
.
.5854
.5116
35. pv. maculifoliigardeniae
.
.

.6000
36. pv. malvacearum





Continued on following page
tO
to
o

Table D-2Continued
Species/Pathovar Species/Pathovar
37
40
41
42
43
44
45
46
47
48
1. pv. alfalfae
.5854
.6190
.4545
.4444
.8696
.4000
.8696
.8696
.4545
.4091
2. pv. alfalfae
.5854
.6190
.4545
.4444
.8696
.4000
.8696
.8696
.4545
.4091
3. pv. armoraciae
.4878
.5238
.5455
.5333
.5217
.5500
.5217
.5217
.9545
1.000
4. pv. begoniae A
.8947
.9231
.5366
.6667
.7442
.5405
.7442
.7442
.5366
.4878
5. pv. begoniae A
.8947
.9231
.5366
.6667
.7442
.5405
.7442
.7442
.5366
.4878
6. pv. begoniae B
.8947
.9231
.5366
.6667
.7442
.5405
.7442
.7442
.5366
.4878
7. pv. bilvae
.5714
.6047
.8444
.9565
.5957
.3902
.5957
.5957
.5778
.5778
8. pv. campestris
.5366
.5714
.5455
.5333
.5652
.6000
.5652
.5652
1.000
.9545
11. pv. citri A
.5366
.5714
.9091
.8000
.5217
.4000
.5217
.5217
.5455
.5455
12. pv. citri B
.5854
.5714
.8182
.9333
.5217
.3500
.5217
.5217
.5455
.5455
13. pv. citri C
.5854
.5714
.8182
.9333
.5217
.3500
.5217
.5217
.5455
.5455
14. pv. citrumelo
.6190
.6512
.4889
.4783
.9362
.4390
.9362
.9362
.4889
.4444
15. pv. citrumelo
.6977
.7273
.5652
.5532
1.000
.4762
1.000
1.000
.5652
.5217
16. pv. citrumelo
.6977
.7273
.5652
.5532
1.000
.4762
1.000
1.000
.5652
.5217
17. pv. ctirumelo
.6667
.6977
.5333
.5217
.9362
.4390
.9362
.9362
.5333
.4889
18. pv. citrumelo
.6190
.6512
.4889
.4783
.9362
.4390
.9362
.9362
.4889
.4444
19. pv. citrumelo
.5854
.6190
.4545
.4444
.8261
.4500
.8261
.8261
.4545
.4091
20. pv. citrumelo
.6977
.7273
.5652
.5532
1.000
.4762
1.000
1.000
.5652
.5217
21. pv. citrumelo
.5854
.6190
.4545
.4444
.8696
.4000
.8696
.8696
.4545
.4091
22. pv. citrumelo
.6977
.7273
.5652
.5532
.9583
.5238
.9583
.9583
.5652
.5217
23. pv. dieffenbachiae A
.9231
1.000
.5714
.6047
.7273
.5789
.7273
.7273
.5714
.5238
24. pv. dieffenbachiae A
.9231
1.000
.5714
.6047
.7273
.5789
.7273
.7273
.5714
.5238
Continued on following page
to
to

Table D-2Continued
Species/Pathovar
Species/Pathovar
37
40
41
42
43
44
45
46
47
48
25. pv. dieffenbachiae B
.5714
.6047
.5333
.4783
.8511
.3415
.8511
.8511
.5333
.5778
26. pv. fici A
.6977
.7273
.5652
.5532
1.000
.4762
1.000
1.000
.5652
.5217
27. pv. fici A
.6190
.6512
.4889
.4783
.9362
.4390
.9362
.9362
.4889
.4444
28 pv. fici A
.6667
.6977
.5333
.5217
.9362
.4390
.9362
.9362
.5333
.4889
29. pv. fici B
.5789
.6154
.4390
.3810
.5116
.8649
.5116
.5116
.6829
.6341
31. pv. glycines A
.5366
.5714
1.000
.8000
.5652
.4000
.5652
.5652
.5455
.5455
34. pv. incanae
.5366
.5714
.5455
.5333
.5652
.6000
.5652
.5652
1.000
.9545
35. pv. maculifoliigardeniae
.8947
.9231
.5854
.5714
.6977
.5405
.6977
.6977
.5854
.5366
36. pv. malvacearum
.5500
.5854
.9302
.7727
.5778
.4103
.5778
.5778
.5116
.5116
37. pv. manihotis
.9231
.5366
.5714
.6977
.5405
.6977
.6977
.5366
.4878
40. pv. phaseoli A
.5714
.6047
.7273
.5789
.7273
.7273
.5714
.5238
41. pv. phaseoli B
.8000
.5652
.4000
.5652
.5652
.5455
.5455
42. pv. phaseoli "fiiscans"
.5532
.3415
.5532
.5532
.5333
.5333
43. pv. physalidicola
.4762
1.000
1.000
.5652
.5217
44. pv.poinsettiicola B
.
.4762
.4762
.6000
.5500
45. pv. poinsettiicola A
.
1.000
.5652
.5217
46. pv. pruni
.
.
.5652
.5217
47. pv. raphani A
.
.
.
.9545
48. pv. raphani B




Continued on following page

Table D-2Continued
Species/Pathovar Species/Pathovar
50
51
52
53
54
55
56
58
59
61
1. pv. alfalfae
.8696
.8511
.8696
.3784
.8000
.6190
.4651
.8444
.4889
.2381
2. pv. alfalfae
.8696
.8511
.8696
.3784
.8000
.6190
.4651
.8444
.4889
.2381
3. pv. armoraciae
.5217
.5106
.5217
.5405
.5333
.5238
.5116
.4444
.5778
.5238
4. pv. begoniae A
.7442
.7727
.7442
.5294
.7619
.9231
.5500
.6667
.6190
.5128
5. pv. begoniae A
.7442
.7727
.7442
.5294
.7619
.9231
.5500
.6667
.6190
.5128
6. pv. begoniae B
.7442
.7727
.7442
.5294
.7619
.9231
.5500
.6667
.6190
.5128
7. pv. bilvae
.5957
.6250
.5957
.5263
.6087
.6047
.8182
.5217
1.000
.4186
8. pv. campestris
.5652
.5532
.5652
.5946
.5778
.5714
.5116
.4889
.5778
.5714
11. pv. citri A
.5217
.5532
.5217
.4865
.5333
.5714
.8837
.4444
.8444
.4762
12. pv. citri B
.5217
.5532
.5217
.4865
.5333
.5714
.7907
.4444
.9333
.4286
13. pv. citri C
.5217
.5532
.5217
.4865
.5333
.5714
.7907
.4444
.9333
.4286
14. pv. citrumelo
.9362
.9167
.9362
.4211
.8696
.6512
.5000
.7826
.5217
.2791
15. pv. citrumelo
1.000
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
16. pv. citrumelo
1.000
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
17. pv. ctirumelo
.9362
.9167
.9362
.4737
.8696
.6977
.5455
.8696
.5652
.3256
18. pv. citrumelo
.9362
.9167
.9362
.4211
.8696
.6512
.5000
.7826
.5217
.2791
19. pv. citrumelo
.8261
.8085
.8261
.3784
.7556
.6190
.4186
.8000
.4889
.2857
20. pv. citrumelo
1.000
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
21. pv. citrumelo
.8696
.8511
.8696
.3784
.8000
.6190
.4651
.8444
.4889
.2381
22. pv. citrumelo
.9583
.9388
.9583
.5128
.8936
.7273
.5333
.7660
.5957
.4091
23. pv. dieffenbachiae A
.7273
.7556
.7273
.5714
.7907
1.000
.5854
.6977
.6047
.5000
24. pv. dieffenbachiae A
.7273
.7556
.7273
.5714
.7907
1.000
.5854
.6977
.6047
.5000
Continued on following page

Table D-2Continued
Species/Pathovar Species/Pathovar
50
51
52
53
54
55
56
58
59
61
25. pv. dieffenbachiae B
.8511
.8333
.8511
.5263
.7826
.6047
.5455
.6522
.5217
.2791
26. pv. fici A
1.000
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
27. pv. fici A
.9362
.9167
.9362
.4211
.8696
.6512
.5000
.7826
.5217
.2791
28 pv. fici A
.9362
.9167
.9362
.4737
.8696
.6977
.5455
.8696
.5652
.3256
29. pv. fici B
.5116
.5000
.5116
.5294
.5238
.6154
.4500
.4286
.4286
.6154
31. pv. glycines A
.5652
.5532
.5652
.5405
.5778
.5714
.9302
.4889
.8444
.4762
34. pv. incanae
.5652
.5532
.5652
.5946
.5778
.5714
.5116
.4889
.5778
.5714
35. pv. maculifoliigardeniae
.6977
.7273
.6977
.5882
.7619
.9231
.6000
.6667
.6190
.5641
36. pv. malvacearum
.5778
.5652
.5778
.5556
.5909
.5854
1.000
.5000
.8182
.4390
37. pv. manihotis
.6977
.7273
.6977
.5294
.7619
.9231
.5500
.6667
.5714
.5128
40. pv. phaseoli A
.7273
.7556
.7273
.5714
.7907
1.000
.5854
.6977
.6047
.5000
41. pv. phaseoli B
.5652
.5532
.5652
.5405
.5778
.5714
.9302
.4889
.8444
.4762
42. pv. phaseoli "fiiscans"
.5532
.5833
.5532
.4737
.5652
.6047
.7727
.4783
.9565
.4186
43. pv. physalidicola
1.000
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
44. pv.poinsettiicola B
.4762
.4651
.4762
.4242
.4878
.5789
.4103
.4390
.3902
.6316
45. pv. poinsettiicola A
1.000
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
46. pv. pruni
1.000
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
47. pv. raphani A
.5652
.5532
.5652
.5946
.5778
.5714
.5116
.4889
.5778
.5714
48. pv. raphani B
.5217
.5106
.5217
.5405
.5333
.5238
.5116
.4444
.5778
.5238
Continued on following page

Table D-2Continued
Species/Pathovar
Species/Pathovar
50
51
52
53
54
55
56
58
59
61
50. pv. vesicatoria A
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
51. pv. vesicatoria A
.
.9796
.5000
.9167
.7556
.5652
.7917
.6250
.4000
52. pv. vesicatoria A
.5128
.9362
.7273
.5778
.8085
.5957
.3636
53. pv. vesicatoria B
.5263
.5714
.5556
.4211
.5263
.4000
54. pv. vignicola
.7907
.5909
.8696
.6087
.3721
55. pv. vitians A
.5854
.6977
.6047
.5000
56. pv. vitians B
.5000
.8182
.4390
58. X campestrisX 198
.5217
.2791
59. X. campestris XCF
61. A! fragariae
*


.4186
to
to
L/t

LITERATURE CITED
Alvarez, A. M., Benedict, A. A., and Gottwald, T. R. 1990. Serological relationships
among Xanthomonas campestris strains associated with citrus bacterial spot. (Abstr.)
Phytopathology 80: 964
Alvarez, A. M., Benedict, A. A., and Mizumoto, C. Y. 1985. Identification of
xanthomonads and grouping of strains of Xanthomonas campestris pv. campestris with
monoclonal antibodies. Phytopathology 75:722-728.
Alvarez, A. M., Benedict, A. A., Mizumoto, C. Y., Pollard, L. W., and Civerolo, E. L.
1991. Analysis of Xanthomonas campestris pv. citri and X. c. citrumelo with
monoclonal antibodies. Phytopathology 81:857-865.
Alvarez, A. M. and Lou, K. 1982. Rapid field identification of a bacterial pathogen by
an enzyme-linked immunosorbent assay (ELISA). (Abstr.) Phytopathology 72: 947.
Anderson, M. J., and Nameth, S. T. 1990. Development of a polyclonal antibody-
based serodiagnostic assay for the detection of Xanthomonas campestris pv. pelargonii
in geranium plants. Phytopathology 80: 357-360.
Angeles-Ramos, R., Vidaver, A. K., and Flynn, P. 1991. Characterization of epiphytic
Xanthomonas campestris pv. phaseoli and pectolytic xanthomonads recovered from
symptomless weeds in the Dominican Republic. Phytopathology 81: 677-681.
Arlat, M., Gough, C. L., Barber, C. E., Boucher, C., and Daniels, M. J. 1991.
Xanthomonas campestris contains a cluster of hrp genes related to the larger hrp cluster
of Pseudomonas solanacearum. Mol. Plant-Microbe Interact. 4:593-601.
Arlat, M., Gough, C. L., Zischek, C., Barberis, P. A., Trigalet, A., and Boucher, C. A.
1992. Transcriptional organization and expression of the large hrp gene cluster of
Pseudomonas solanacearum. Mol. Plant-Microbe Interact. 5:187-193.
Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., and Seidman, K., eds.. 1987.
Current Protocols in Molecular Biology. J. Wiley & Sons, New York.
226

227
Azad, H., and Schaad, N. W. 1988. Serological relationships among membrane
proteins of strains of Xanthomonas campestris pv. translucens. Phytopathology 78:
272-277.
Bach, E. E., Alba, A. P. C., Pereira, A. L. G., Zagatto, A. G., and Rossetti, V. 1978.
Serological studies of Xanthomonas citri (Hasse) Dowson. Arq. Inst. Biol. Sao Paulo
45: 229-236.
Bashan, Y., Okon, Y., and Henis, Y. 1982. Long-term survival of Pseudomonas
syringae pv. tomato and Xanthomonas campestris pv. vesicatoria in tomato and pepper
seeds. Phytopathology 72: 1143-1144.
Beaulieu, C., Minsavage, G. V., Canteros, B. I., and Stall, R. E. 1991. Biochemical
and genetic analysis of a pectate lyase gene from Xanthomonas campestris pv.
vesicatoria. Mol. Plant-Microbe Interact. 4:446-451.
u Beer, S. V., Bauer, D. W., Jiang, X. H., Laby, R. J., Sneath, B. J., Wei, Z.-M., Wilcox,
D. A., and Zumoff, C. H. 1991. The hrp gene cluster of Erwinia amylovora. Pages
53-60 in: Advances in Molecular Genetics of Plant-Microbe Interactions, vol. 1. H.
Hennecke and D. P. S. Verma, eds. Kluwer Academic Publishers, Dordrecht, The
Netherlands.
Benedict, A. A., Alvarez, A. M., Berestecky, J., Imanaka, W., Mizumoto, C. Y.,
Pollard, L. W., Mew, T. W., and Gonzalez, C. F. 1989. Pathovar-specific monoclonal
antibodies for Xanthomonas campestris pv. oryzae and for Xanthomonas campestris
pv. oryzicola. Phytopathology 79: 322-328.
Benedict, A. A., Alvarez, A. M., Civerolo, E. L., W., and Mizumoto, C. Y. 1985.
Delineation of Xanthomonas campestris pv. citri strains with monoclonal antibodies.
(Abstr.) Phytopathology 75:1352.
Benedict, A. A., Alvarez, A. M., and Pollard, L. W. 1990. Pathovar-specific antigens
of Xanthomonas campestris pv. begoniae and Xanthomonas campestris pv. pelargonii.
Appl. Environ. Microbiol. 56: 572-574.
Bereswill, S., Pahl, A., Bellemann, P., Zeller, W., and Geider, K. 1992. Sensitive and
species-specific detection of Erwinia amylovora by polymerase chain reaction analysis.
Appl. Environ. Microbiol. 58: 3522-3526.
Berthier, Y., Verdier, V., Guesdon, J.-L., Chevrier, D., Denis, J.-B., Decoux, G., and
Lemattre, M. 1993. Characterization of Xanthomonas campestris pathovars by rRNA
gene restriction patterns. Appl. Environ. Microbiol. 59:851-859.

228
Bochner, B. R. 1989. Sleuthing out bacterial identities. Nature 339:157-158.
v Bonas, U., Schulte, R., Fenselau, S., Minsavage, G. V., Staskawicz, B. J., and Stall, R.
E. 1991. Isolation of a gene cluster from Xanthomonas campestris pv. vesicatoria that
determines pathogenicity and the hypersensitive response on pepper and tomato. Mol.
Plant-Microbe Interact. 4:81-88.
Boucher, C. A., Van Gijsegem, F., Barberis, P. A., Arlat, M., and Zischek, C. 1987.
Pseudomonas solanacearum genes controlling both pathogenicity and hypersensitivity
on tobacco are clustered. J. Bacteriol. 169: 5626-5632.
Bouzar, H., Jones, J. B., and Stall, R. E. 1994. Silver staining reveals proteins unique
to phenotypically distinct groups of Xanthomonas campestris pv. vesicatoria.
Phytopathology 84: 39-44.
Bouzar, H., Jones, J. B., Stall, R. E., and Scott, J. W. 1992. Cluster analysis of
Xanthomonas campestris pv. vesicatoria based on carbon source utilization patterns.
Proc. 8th Int. Conf. Plant Pathog. Bacteria. INRA, Versailles, France, (in press)
Bradbury, J. F. 1984. Genus II.Xanthomonas Dowson 1939, 187. Pages 199-210 in:
Bergey's Manual of Determinative Bacteriology, 9th ed. J. C. Holt and N. R. Krieg,
eds. Williams & Wilkins Co., Baltimore, MD.
Bradbury, J. F. 1986. Guide to Plant Pathogenic Bacteria. CAB International
Mycological Institute, Slough, Great Britain.
Brinkerhoff, L. A. 1970. Variation in Xanthomonas malvacearum and its relation to
control. Annu. Rev. Phytopathol. 8:85-110.
Burkholder, W. H., and Starr, M. P. 1948. The generic and specific characters of
phytopathogenic species of Pseudomonas and Xanthomonas. Phytopathology 38: 494-
502.
Canteros de Echenique, B. I., Zagory, D., and Stall, R. E. 1985. A medium for
cultivation of the B strains of Xanthomonas campestris pv. citri, cause of cancrosis B
in Argentina and Uruguay. Plant Dis. 69:122-123.
Charudattan, R., Stall, R. E., and Batchelor, D. L. 1973. Serotypes of Xanthomonas
vesicatoria unrelated to its pathotypes. Phytopathology 63: 1260-1265.

229
Chase, A. R., Stall, R. E., Hodge, N. C., and Jones, J. B. 1992. Characterization of
Xanthomonas campestris strains from aroids using physiological, pathological, and
fatty acid analyses. Phytopathology 82: 754-759.
Civerolo, E. L. 1984. Bacterial canker disease of citrus. J. Rio Grande Val. Hortic.
Soc. 37: 127-146.
Civerolo, E L., and Fan, F. 1982. Xanthomonas campestris pv. citri detection and
identification by enzyme-linked immunosorbent assay. Plant Dis. 66: 231-236.
Clark, M. F., and Adams, A. N. 1977. Characteristic of the microplate method of
enzyme-linked immunosorbent assay for the detection of plant viruses. J. Gen. Virol.
34: 475-483.
Cooksey, D. A., Azad, H. R., Jae-Soon, C., and Chun-Keun, L. 1990. Copper
resistance gene homologs in pathogenic and saprophytic bacterial species from tomato.
Appl. Environ. Microbiol. 56: 431-435.
Cooksey, D. A., and Graham, J. H. 1989. Genomic fingerprinting of two pathovars of
phytopathogenic bacteria by rare-cutting restriction enzymes and field inversion gel
electrophoresis. Phytopathology 79: 745-750.
Cox, R. S. 1966. The role of bacterial spot in tomato production in south Florida.
Plant Dis. Rep. 50: 699-700.
Davis, M. J., French, W. J., and Schaad, N. W. 1981. Axenic culture of the bacteria
associated with phony disease of peach and plum. Curr. Microbiol. 6: 309-314.
DeParasis, J., and Roth, D.A. 1990. Nucleic acid probes for identification of
phytobacteria: Identification of genus-specific 16s rRNA sequences. Phytopathology
80: 618-621.
DeVos, P., and DeLey, J. 1983. Intra- and intergeneric similarities of Pseudomonas
dead Xanthomonas ribosomal ribonucleic acid cistrons. Int. J. Syst. Bacteriol. 33: 487-
509.
DeVos, P., Goor, M., Gillis, M., and DeLey, J. 1985. Ribosomal ribonucleic acid
cistron similarities of phytopathogenic Pseudomonas species. Int. J. Syst. Bacteriol.
35: 169-184.
Dickey, R. S., and Zumoff, C. H. 1987. Bacterial leaf blight of Syngonium caused by a
pathovar of Xanthomonas campestris. Phytopathology 77:1257-1262.

230
Ditta, G., Stanfield, S., Corbin, D., and Helinski, D. 1980. Broad host range DNA
cloning system for gram-negative bacteria: Construction of a gene bank of Rhizobium
meliloti. Proc. Natl. Acad. Sci. USA 77:7347-7351.
DosSantos, R. M. D., B., and Dianese, J. C. 1985. Comparative membrane
characterization of Xanthomonas campestris pv. cassavae and X. campestris pv.
manihotis. Phytopathology 75: 581-587.
Dye, D. W. 1962. The inadequacy of the usual determinative tests for the
identification of Xanthomonas spp.. N. Z. J. Sci. 5: 393-416.
Dye, D. W. 1966. Cutural and biochemical reactions of additional Xanthomonas spp.
N. Z.J. Sci. 9:913-919.
Dye, D. W., Bradbury, J. F., Goto, M., Hayward, A. C., Lelliot, R. A., and Schroth, M.
N. 1980. International standards for naming pathovars of phytopathogenic bacteria
and a list of pathovar names and pathotype strains. Rev. Plant Pathol. 59: 153-168.
Egel, D. S. 1991. Pathogenic and Genomic Characterization of Strains of
Xanthomonas campestris Causing Diseases of Citrus. Ph.D. dissertation. University of
Florida, Gainesville. 132 pp.
Egel, D. S., Graham, J. H., and Stall, R. E. 1991. Genomic relatedness of
Xanthomonas campestris strains causing diseases of citrus. Appl. Environ. Microbiol.
57: 2724-2730.
Elrod, R. P., and Braun, A. C. 1947. Serological studies of the genus Xanthomonas. I.
Cross-agglutination relationships. J. Bacteriol. 53:509-518.
El-Sharkawy, T. A., and Huisingh, D. 1971a. Electrophorethic analysis of esterases
and other soluble proteins from represntatives of phytopathogenic bacterial genera. J.
Gen. Microbiol. 68: 149-154.
El-Sharkawy, T. A., and Huisingh, D. 1971b. Differentiation among Xanthomonas
species by polyacrylamide gel electrophoresis of soluble proteins. J. Gen. Microbiol.
68: 155-165.
Feinberg, A. P., and Vogelstein, B. 1983. A technique for radiolabeling DNA
restriction endonuclease fragments to high specific activity. Anal. Biochem. 132:6-
13.

231
Felsenstein, J. 1985. Confidence limits on phylogenies: An approach using the
bootstrap. Evolution 39:783-791.
Felsenstein, J. 1991. Phylogeny Inference Package, Version 3.4. Department of
Genetics, University of Washington, Pullman.
Felsenstein, J. 1993. Phylogeny Inference Package, Version 3.5. Department of
Genetics, University of Washington, Pullman.
' Fenselau, S., Balbo, I., and Bonas, U. 1992. Determinants of pathogenicity in
Xanthomonas campestris pv. vesicatoria are related to proteins involved in secretion in
bacterial pathogens of animals. Mol. Plant-Microbe Interact. 5: 390-396.
Figurski, D., and Helinski, D. R. 1979. Replication of an origin containing derivative
of plasmid RK2 dependent on a plasmid function provided in trans. Proc. Natl. Acad.
Sci. USA 76: 1648-1979.
Fitch, W. M., and Margoliash, E. 1967. Construction of phylogenetic trees. Science
155:279-284.
Franken, A. A. J. M., Zilverentant, J. F., Boonekamp, P. M., and Schots, A. 1992.
Specificity of polyclonal and monoclonal antibodies for the identification of
Xanthomonas campestris pv. campestris. Neth. J. PI. Path. 98: 81-94.
Gabriel, D. W., Kingsley, M. T., Hunter, J. E., and Gottwald, T. 1989. Reinstatement
of Xanthomonas citri (ex Hasse) and X. phaseoli (ex Smith) to species and
reclassification of all A! campestris pv. citri strains. Int. J. Syst. Bacteriol. 39: 14-22.
Gabriel, D. W., Hunter, J. E., Kingsley, M. T., Miller, J. W., and Lazo, G. R. 1988.
Clonal population structure of Xanthomonas campestris and genetic diversity among
citrus canker strains. Mol. Plant-Microbe Interact. 1: 59-65.
Garde, S., and Bender, C. L. 1991. DNA probes for detection of copper resistance
genes in Xanthomonas campestris pv. vesicatoria. Appl. Environ. Microbiol. 57:
2435-2439.
Gardner, M. W., and Kendrick, J. B. 1921. Bacterial spot of tomato. J. Agr. Res. 21:
123-156.
Gardner, M. W., and Kendrick, J. B. 1923. Bacterial spot of tomato and pepper.
Phytopathology 13:307-315.

232
Gilbertson, R. L., Maxwell, D. P., Hagedom, D. J., and Leong, S. A. 1989.
Development and application of a plasmid DNA probe for detection of bacteria causing
common bacterial blight of bean. Phytopathology 79:518-525.
Gilbertson, R. L., Rand, R. E., and Hagedom, D. J. 1990. Survival of Xanthomonas
campestris pv. phaseoli and pectolytic strains o Xanthomonas campestris in bean
debris. Plant Dis. 74: 322-327.
Gitaitis, R. D., Chang, C. J., Sijam, K., and Dowler, C. C. 1991. A differential
medium for semiselective isolation of Xanthomonas campestris pv. vesicatoria and
other cellulolytic xanthomonads from various natural sources. Plant Dis. 75: 1274-
1278.
Gitaitis, R., McCarter, S., and Jones, J. 1992. Disease control in tomato transplants
produced in Georgia and Florida. Plant Dis. 76: 651-656.
Gitaitis, R. D., Sasser, M. J., Beaver, R. W., Mclnnes, T. B., and Stall, R. E. 1987.
Pectolytic xanthomonads in mixed infections with Pseudomonas syringae pv. syringae,
P. syringae pv. tomato, and Xanthomonas campestris pv. vesicatoria in tomato and
pepper transplants. Phytopathology 77:611-615.
Gottwald, T. R., Alvarez, A. M., Hartung, J. S., and Benedict, A. A. 1991. Diversity
of Xanthomonas campestris pv. citrumelo strains associated with epidemics of citrus
bacterial spot in Florida citrus nurseries: Correlation of detached leaf, monoclonal
antibody, and restriction fragment length polymorphism assays. Phytopathology 81:
749-753.
Gough, C. L., Genin, S., Zischek, C., and Boucher, C. A. 1992. hrp genes of
Pseudomonas solanacearum are homologous to pathogenicity determinants of animal
pathogenic bacteria and are conserved among plant pathogenic. Mol. Plant-Microbe
Interact. 5: 384-389.
Graham, J. H., and Gottwald, T. R. 1988. Citrus canker and citrus bacterial spot in
Florida: Research fmdings-future considerations. The Citrus Industry 69: 42-51.
Graham, J. H., and Gottwald, T. R. 1990. Variation in aggressiveness of Xanthomonas
campestris pv. citrumelo associated with citrus bacterial spot in Florida citrus nurseries.
Phytopathology 80: 190-196.
Graham, J.H., and Gottwald, T.R. 1991. Research perspectives on eradication of citrus
bacterial diseases in Florida. Plant Dis. 75:1193-1200.

233
Graham, J. H., Hartung, J. S., Stall, R. E., and Chase, A.R. 1990. Pathological,
restriction-fragment length polymorphism, and fatty acid profile relationships between
Xanthomonas campestris from citrus and noncitrus hosts. Phytopathology 80: 829-
836.
Haefele, D., and Webb, R. 1982. The use of ultrasound to facilitate the harvesting and
quantification of epiphytic and phytopathogenic microorganisms. (Abstr.)
Phytopathology 72: 947.
Hartung, J. S. 1992. Plasmid-based hybridization probes for detection and
identification of Xanthomonas campestris pv. citri. Plant Dis. 76: 889-893.
Hartung, J. S., and Civerolo, E. L. 1987. Genomic fingerprints of Xanthomonas
campestris pv. citri strains from Asia, South America, and Florida. Phytopathology
77: 282-285.
Hartung, J. S., and Civerolo, E. L. 1989. Restriction fragment length polymorphisms
distinguish Xanthomonas campestris strains isolated from Florida citrus nurseries from
X. c. pv. citri. Phytopathology 79: 793-799.
Hartung, J. S., and Civerolo, E. L. 1991. Variation among strains of Xanthomonas
campestris causing citrus bacterial spot. Plant Dis. 75:622-626.
Hartung, J. S., Daniel, J. F., and Pruvost, O. P. 1993. Detection of Xanthomonas
campestris pv. citri by the polymerase chain reaction method. Appl. Environ.
Microbiol. 59: 1143-1148.
Hayward, A. C. 1993. The hosts of Xanthomonas. Pages 1-119 in: Xanthomonas.
J.G. Swings, and E.L. Civerolo, eds. Chapman & Hall, London, United Kingdom.
Higgis, B. B. 1922. The bacterial spot of pepper. Phytopathology 12:501-516.
Hildebrand, D. C., Hendson, M., and Schroth, M. N. 1993. Usefulness of nutritional
screening for the identification of Xanthomonas campestris DNA homology groups and
pathovars. J. Appl. Bacteriol. 75: 447-455.
Hildebrand, D. C., Palleroni, N. J., and Schroth, M. N. 1990. Deoxyribonucleic acid
relatedness of 24 xanthomonad strains representing 23 Xanthomonas campestris
pathovars and Xanthomonas fragariae. J. Appl. Bacteriol. 68: 263-269.

234
Hodge, N. C., Chase, A. R., and Stall, R. E. 1992. Diversity of four species of
Xanthomonas as determined by cellular fatty acid anlyses. (Abstr.) Phytopathology
82:1153.
Holben, W. E., Jansson, J. K., Chelm, B. K., and Tiedje, J. M. 1988. DNA probe
method for the detection of specific microorganisms in the soil bacterial community.
Appl. Environ. Microbiol. 54: 703-711.
Holt, J. G., Krieg, N. R., Sneath, P. H. A., Staley, J. T., and Williams, S. T., eds. 1994.
Bergey's Manual of Determinative Bacteriology, 9th ed. Williams & Wilkins,
Baltimore, MD. 787 pp.
Huang, Y., Xu, P., and Sequeira, L. 1990. A second cluster of genes that specify
pathogenicity and host response in Pseudomonas solanacearum. Mol. Plant-Microbe
Interact. 3:48-53.
Hwang, I. Y., Lim, S. M., and Shaw, P. D. 1992. Cloning and characterization of
pathogenicity genes from Xanthomonas campestris pv. glycines. J. Bacteriol. 174:
1923-1931.
Jones, J. B., Chase, A. R., and Harris, G. K. 1993a. Evaluation of the Biolog GN
MicroPlate system for identification of some plant-pathogenic bacteria. Plant Dis. 77:
553-558.
Jones, J. B., Jones, J. P., Stall, R. E., and Zitter, T. A. 1991. Compendium of Tomato
Diseases. American Phytopathological Society Press, St. Paul, MN. 73 pp.
Jones, J. B., Minsavage, G. V., Stall, R. E., Kelly, R. O., and Bouzar, H. 1993b.
Genetic analysis of a DNA region involved in expression of two epitopes associated
with lipopolysaccharide in Xanthomonas campestris pv. vesicatoria. Phytopathology
83: 551-556.
Jones, J. B., Pohronezny, K. L., Stall, R. E., and Jones, J. P. 1986. Survival of
Xanthomonas campestris pv. vesicatoria in Florida on tomato crop residue.
Phytopathology 76: 430-434.
Kamoun, S., and Kado, C. I. 1990. A plant-inducible gene of Xanthomonas
campestris pv. campestris encodes an exocellular component required for growth in the
host and hypersensitivity on nonhosts. J. Bacteriol. 172: 5165-5172.

235
Kamoun, S., Kamdar, H. V., Tola, E., and Kado, C. I. 1992. Incompatible interactions
between crucifers and Xanthomonas campestris involve a vascular hypersensitive
response: role of the hrpXlocus. Mol. Plant-Microbe Interact. 5:22-33.
Kawasaki, E. S. 1990. Sample preparation from blood, cells, and other fluids. Pages
146-152 in: PCR Protocols: A Guide to Methods and Applications. M.A. Innis, D.H.
Gelfand, J.J. Sninsky, and T.J. White, eds. Academic Press, Inc., San Diego, CA.
Kersters, K., and DeLey, J. 1975. Identification and grouping of bacteria by numerical
analysis of their electrophoretic protein patterns. J. Gen. Microbiol. 87: 333-342.
Kersters, K., Pot, B., Hoste, B., Gillis, M., and DeLey, J. 1989. Protein
electrophoresis and DNA:DNA hybridization of xanthomonads from grasses and
cereals. EPPO Bull. 19:51-55.
Koike, H. 1965. The aluminum-cap method for testing sugarcane varieties against leaf
scald disease. Phytopathology 55:317-319.
Krawiec, S., and Riley, M. 1990. Organization of the bacterial chromosome.
Microbiol. Rev. 54: 502-539.
Krieg, N. R., and Holt, J. G., eds. 1984. Bergey's Manual of Systematic Bacteriology,
vol. 1. Williams & Wilkins, Baltimore, MD. 964 pp.
Kubicek, Q. B., Civerolo, E. L., Bonde, M. R., Hartung, J. S., and Peterson, G. L.
1989. Isozyme analysis of Xanthomonas campestris pv. citri. Phytopathology 79:
297-300.
Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head
of bacteriophage T4. Nature 227:680-685.
Lazo, G. R., and Gabriel, D. W. 1987. Conservation of plasmid DNA sequences and
pathovar identification of strains of Xanthomonas campestris. Phytopathology 77:
448-453.
Lazo, G. R., Roffey, R., and Gabriel, D. W. 1987. Pathovars of Xanthomonas
campestris are distinguishable by restriction fragment-length polymorphism. Int. J.
Syst. Bacteriol. 37:214-221.
Leach, J. E., Rhoads, M. L., Vera Cruz, C. M., White, F. F., Mew, T. W., and Leung,
H. 1992. Assessment of genetic diversity and population structure of Xanthomonas

236
oryzae pv. oryzae with a repetitive DNA element. Appl. Environ. Microbiol. 58:
2188-2195.
Leach, J. E., White, F. F., Rhoads, M. L., and Leung, H. 1990. A repetitive DNA
sequence differentiates Xanthomonas campes tris pv. oryzae from other pathovars of X.
campestris. Mol. Plant-Microbe Interact. 3:238-246.
Leite, R. P., Jr. 1990. Cancro Ctrico: Preven9ao e Controle no Paran. Funda^ao
Instituto Agronmico do Paran, Londrina, Brazil. 51 pp.
Liao, C. H., and Wells, J. M. 1987. Association of pectolytic strains of Xanthomonas
campestris with soft rots of fruits and vegetables at retail markets. Phytopathology 77:
418-422.
Lindgren, P. B., Peet, R. C., and Panopoulos, N. J. 1986. Gene cluster of
Pseudomonas syringae pv. "phaseolicola" controls hypersensitivity on nonhost plants.
J. Bacteriol. 168:512-522.
Link, G. K. K., and Sharp, C. G. 1927. Correlation of host and serological specificity
of Bact. campestre, Bact. flaccumfaciens, Bad. phaseoli, and Bad. phaseoli sojense.
Bot.Gaz. 83: 145-160.
Lipp, R. L., Alvarez, A. M., Benedict, A. A., and Berestecky, J. 1992. Use of
monoclonal antibodies and pathogenicity tests to characterize strains of Xanthomonas
campestris pv. dieffenbachiae from aroids. Phytopathology 82: 677-682.
Leyns, F., DeCleene, M., Swings, J.-G., and DeLey, J. 1984. The host range of the
gems Xanthomonas. Bot. Rev. 50:308-356.
Maas, J. L., Finney, M. M., Civerolo, E. L., and Sasser, M. 1985. Association of an
unusual strain of Xanthomonas campestris with apple. Phytopathology 75: 438-445.
Mahanta, I. C., and Addy, S. K. 1977. Serological specificity o Xanthomonas oryzae
incitant of bacterial blight of rice. Int. J. Syst. Bacteriol. 27: 383-385.
Manulis, S., Gafni, Y., Clark, E., Zutra, D., Ophir, Y., and Barash, I. 1991.
Identification of a plasmid DNA probe for detection of strains of Erwinia herbicola
pathogenic on Gypsophila paniculata. Phytopathology 81:54-57.
Marco, G. M., and Stall, R. E. 1983. Control of bacterial spot of pepper initiated by
strains of Xanthomonas campestris pv. vesicatoria that differ in sensitivity to copper.
Plant Dis. 67:779-781.

237
McGuire, R. G., and Jones, J.B. 1989. Detection of Xanthomonas campestris pv.
vesicatoria in tomato. Pages 59-62 in: Detection of Bacteria in Seed and Other
Planting Material. A.W. Saettler, N.W. Schaad, and D.A. Roth, eds. American
Phytopathological Society Press, St. Paul, MN.
McGuire, R. G., Jones, J. B., and Sasser, M. 1986. Tween media for semiselective
isolation of Xanthomonas campestris pv. vesicatoria from soil and plant material.
Plant Dis. 70:887-891.
Miller, J. H. 1972. Experiments in Molecular Genetics. Cold Spring Harbor
Laboratory, Cold Spring Harbor, NY. 466 pp.
Miller, L. T. 1982. Single derivatization method for routine analyses of bacterial
esters, including hydroxy acids. J. Clin. Microbiol. 16: 584-586.
Miller, S. A., and Martin, R. R. 1988. Molecular diagnosis of plant disease. Annu.
Rev. Phytopathol. 26: 409-432.
Ming, D., Ye, H., Schaad, N. W., and Roth, D. A. 1991. Selective recovery of
Xanthomonas spp. from rice seed. Phytopathology 81: 1358-1363.
Minsavage, G. V., Dahlbeck, D., Whalen, M. C., Kearney, B., Bonas, U., Staskawicz,
B. J., and Stall, R. E. 1990. Gene-for-gene relationships specifying disease resistance
in Xanthomonas campestris pv. vesicatoria pepper interactions. Mol. Plant-Microbe
Interact. 3: 41-47.
Minsavage, G. V., and Schaad, N. W. 1983. Characterization of membrane proteins of
Xanthomonas campestris pv. campestris. Phytopathology 73:747-755.
Minsavage, G. V., Thompson, C. M., Hopkins, D. L., Leite, R. M. V. B. C., and Stall,
R. E. 1994. Development of a polymerase chain reaction protocol for detection of
Xylella fastidiosa in plant tissue. Phytopathology 84:456-461.
Morris, C. 1985. Diversity of Epiphytic Bacteria on Snap Bean Leaflets Based on
Nutrient Utilization Abilities: Biological and Statistical Considerations. Ph.D.
dissertation. University of Wisconsin, Madison. 210 pp.
Nei, M. 1987. Molecular Evolutionary Genetics. Columbia University Press, New
York, NY.

238
Nei, M, and Li, W. H. 1979. Mathematical model for studying genetic variation in
terms of restriction endonucleases. Proc. Natl. Acad. Sci. USA 76: 5269-5273.
Obata, T., and Tsuboi, F. 1972. Studies on the serological diagnosis of bacterial plant
diseases. I. Cross-agglutination and gel-diffusion tests with some Xanthomonas nomen
species. Res. Bull. Plant Prot. Ser. Jpn. 10: 8-16.
O'Brien, L. M., Morton, D. J., Manning, W. J., and Scheetz, R. W. 1967. Serological
differences between apparently typical pepper and tomato isolates of Xanthomonas
vesicatoria. Nature 215:532-533.
Ogram, A., Sayler, G. S., and Barkay, T. 1988. DNA extraction and purification from
sediments. J. Microbiol. Methods. 7 : 57-66.
Palleroni, N. J., Hildebrand, D. C., Schroth, M. N., and Hendson, M. 1993.
Deoxyribonucleic acid relatedness of 21 strains o Xanthomonas species and pathovars.
J. Appl. Bacteriol. 75:441-446.
Palleroni, N. J., Kunisawa, R., Contopoulou, R., and Doudoroff, M. 1973. Nucleic
acid homologies in the genus Pseudomonas. Int. J. Syst. Bacteriol. 23: 333-339.
Permar, T. A., and Gottwald, T. R. 1989. Specific recognition of a Xanthomonas
campestris Florida citrus nursery strain by a monoclonal antibody probe in a
microfiltration enzyme immunoassay. Phytopathology 79: 780-783.
Pickup, R. W. 1991. Development of molecular methods for the detection of specific
bacteria in the environment. J. Gen. Microbiol. 137:1009-1019.
Pohronezny, K., Stall, R. E., Canteros, B. I., Kegley, M., Datnoff, L. E., and
Subramanya, R. 1992. Sudden shift in the prevalent race of Xanthomonas campestris
pv. vesicatoria in pepper fields in southern Florida. Plant Dis. 76: 118-120.
Pohronezny, K,. and Volin, R. B. 1983. The effect of bacterial spot on yield and
quality of fresh market tomatoes. HortScience 18:69-70.
Pohronezny, K., Waddill, V. H., Schuster, D. J., and Sonoda, R. M. 1986. Integrated
pest management for Florida tomatoes. Plant Dis. 70:96-102.
Pruvost, O., Hartung, J. S., Civerolo, E. L., Dubois, C., and Perrier, X. 1992. Plasmid
DNA fingerprints distinguish pathotypes of Xanthomonas campestris pv. citri, the
causal agent of citrus bacterial canker disease. Phytopathology 82: 485-490.

239
Qhobela, M, and Claflin, L. E. 1988. Characterization of Xanthomonas campestris
pv. pennamericanum pv. nov., causal agent of bacterial leaf streak of pearl millet. Int.
J. Syst. Bacteriol. 38: 362-366.
Qhobela, M, and Claflin, L. E. 1992. Eastern and southern african strains of
Xanthomonas campestris pv. vasculorum are distinguishable by restriction fragment
length polymorphism of DNA and polyacrylamide gel electrophoresis of membrane
proteins. Plant Pathol. 41:113-121.
Qhobela, M, Leach, J. E., Claflin, L. E., and Pearson, D. L. 1991. Characterization of
strains of Xanthomonas campestris pv. holcicola by PAGE of membrane proteins and
by REA and RFLP analysis of genomic DNA. Plant Dis. 75: 32-36.
Roth, D. A. 1989. Review of extraction and isolation methods. Pages 3-8 in:
Detection of Bacteria in Seeds and Other Planting Material. A.W. Saettler, N.W.
Schaad, and D.A. Roth, eds. American Phytopathological Society Press, St. Paul, MN.
Saettler, A. W., Schaad, N. W., and Roth, D. A., eds. 1989. Detection of Bacteria in
Seeds and Other Planting Material. American Phytopathological Society Press, St.
Paul, MN. 122pp
Sambrook, J., Fritsch, E. F., and Maniatis, T. 1989. Molecular Cloning: A Laboratory
Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 1546 pp.
SAS Institute Inc. 1988. SAS/STAT User's Guide. SAS Institute Inc., Cary, NC.
1028 pp.
Sasser, M. 1990. Identification of bacteria through fatty acid analysis. Pages 199-204
in: Methods in Phytobacteriology. Z. Klement, K. Rudolph, and D. Sands, eds.
Akadmiai Kiad, Budapest, Hungary.
Schaad, N.W. 1979. Serological identification of plant pathogenic bacteria. Annu.
Rev. Phytopathol. 17:123-147.
Schaad, N.W. 1982. Detection of seedbome bacterial plant pathogens. Plant Dis. 66:
885-890.
Schaad, N. W., ed. 1988. Laboratory Guide for Identification of Plant Pathogenic
Bacteria. American Phytopathological Society Press, St. Paul, MN. 164 pp.

240
Schaad, N. W., Azad, H., Peet, R. C., and Panopoulos, N. J. 1989. Identification of
Pseudomonas syringae pv. phaseolicola by a DNA hybridization probe.
Phytopathology 79: 903-907.
Schaad, N. W., and Stall, R. E. 1988. Xanthomonas. Pages 81-94 in: Laboratory
Guide for Identification of Plant Pathogenic Bacteria. N. W. Schaad, ed. American
Phytopathological Society Press, St. Paul, MN.
Schoulties, C. L., Civerolo, E. L., Miller, J. W., Stall, R. E., Krass, C. J., Poe, S. R.,
and DuCharme, E. P. 1987. Citrus canker in Florida. Plant Dis. 71:388-395.
Schulte, R., and Bonas, U. 1992. Expression of the Xanthomonas campestris pv.
vesicatoria hrp gene cluster, which determines pathogenicity and hypersensitivity on
pepper and tomato, is plant inducible. J. Bacteriol. 174: 815-823.
Scott, J. W., and Jones, J. B. 1989. Inheritance of resistance to foliar bacterial spot of
tomato incited by Xanthomonas campestris pv. vesicatoria. J. Amer. Soc. for Hort.
Sci. 114: 111-114.
Seal, S. E., Jackson, L. A., and Daniels, M. J. 1992. Isolation of a Pseudomonas
solanacearum specific DNA probe by subtraction hybridization and construction of
species-specific oligonucleotide primers for sensitive detection by the polymerase
chain reaction. Appl. Environ. Microbiol. 58: 3751-3758.
Sharon, E., Okon, Y., Bashan, Y., and Henis, Y. 1982. Detached leaf enrichment: A
method for detecting small numbers of Pseudomonas syringae pv. tomato and
Xanthomonas campestris pv. vesicatoria in seed and symptomless leaves of tomato and
pepper. J. Appl. Bacteriol. 53:371-377.
Sijam, K., Chang, C. J., and Gitaitis, R. D. 1991. An agar medium for the isolation
and identification of Xanthomonas campestris pv. vesicatoria from seed.
Phytopathology 81: 831-834.
Sneath, P. H. A. 1989. Analysis and interpretation of sequence data for bacterial
systematics: The view of a numerical taxonomist. Syst. Appl. Microbiol. 12: 15-31.
Somodi, G. C., Jones, J. B., and Scott, J. W. 1993. An ELISA procedure developed
for detection of low populations of Xanthomonas campestris pv. vesicatoria (Xcv) in
tomato leaves. (Abstr.) Phytopathology 83:1394
Southern, E. 1975. Detection of specific sequences among DNA fragments separated
by gel electrophoresis. J. Mol. Biol. 98:503-517.

Sowell, G., Jr., and Dempsey, A. H. 1977. Additional sources of resistance to
bacterial spot of pepper. Plant Dis. Reptr. 61: 684-686.
241
Stall, R. E. 1993. Xanthomonas campestris pv. vesicatoria: Cause of bacterial spot of
tomato and pepper. Pages 57-60 in: Xanthomonas. J. G. Swings and E. L. Civerolo,
eds. Chapman & Hall, London, United Kingdom.
Stall, R. E., Beaulieu, C., Egel, D., Hodge, N. C., Leite, R. P., Minsavage, G. V.,
Bouzar, H., Jones, J. B., Alvarez, A. M., and Benedict, A. A. 1994. Two genetically
diverse groups of strains are included in Xanthomonas campestris pv. vesicatoria. Int.
J. Syst. Bacteriol. 44: 47-53.
Stall, R. E., and Civerolo, E. L. 1991. Research relating to the recent outbreak of
citrus canker in Florida. Annu. Rev. Phytopathol. 29: 399-420.
Stall, R. E., and Hodge, N. C. 1989. Use of fatty acid profiles to identify strains of
Xanthomonas campestris pv. citri from the citrus canker epidemic in Florida. (Abstr.)
Phytopathology 79: 376
Stall, R. E., and Minsavage, G. V. 1990. The use of hrp genes to identify
opportunistic xanthomonads. Pages 369-374 in: Proceedings of the 7th International
Conference on Plant Pathogenic Bacteria. Z. Klement, eds. Akadmiai Kiad,
Budapest, Hungary.
Stall, R. E., and Seymour, C. P. 1983. Canker, a threat to citrus in the gulf-coast
states. Plant Dis. 67:581-585.
Steffan, R. J., and Atlas, R. M. 1988. DNA amplification to enhance the detection of
genetically engineered bacteria in environmental samples. Appl. Environ. Microbiol.
54:2185-2191.
Steffan, R. J., and Atlas, R. M. 1991. Polymerase chain reaction: Applications in
environmental microbiology. Annu. Rev. Microbiol. 45:137-161.
Steffan, R. J., Goksoyr, J., Bej, A. K., and Atlas, R. M. 1988. Recovery of DNA from
soils and sediments. Appl. Environ. Microbiol. 54:2908-2915.
Taylor, J. 1962. The estimation of numbers of bacteria by ten-fold dilution series. J.
Appl. Bacteriol. 25:54-61.

242
Thaveechai, N., and Schaad, N. W. 1984. Comparison of different immunogen
preparations for serological identification of Xanthomonas campestris pv. campestris.
Phytopathology 74: 1065-1070.
Thayer, P. L., and Stall, R. E. 1961. Effect of variation in the bacterial spot pathogen
of pepper and tomato on control with streptomycin. Phytopathology 51: 568-571.
Van den Mooter, M., Maraite, H., Meiresonne, L., Gillis, M., Kersters, K., and DeLey,
J. 1987a. Comparison between Xanthomonas campestris pv. manihotis (ISPP List
1980) and X campestris pv. cassavae (ISPP List 1980) by means of phenotypic,
protein electrophoretic, DNA hybridization and phytopathological techniques. J. Gen.
Microbiol. 133:57-71.
Van den Mooter, M., Steenackers, M., Maertens, C., Gossele, F., DeVos, P., Swings, J.,
Kersters, K., and DeLey, J. 1987b. Differentiation between Xanthomonas campestris
pv. graminis (ISPP List 1980), pphleipratensis (ISPP List 1980) emend., p\.poae
Egli and Schmidt 1982 and pv. arrhenatheri Egli and Schmidt 1982, by numerical
analysis of phenotypic features and protein gel electrophoregrams. J. Phytopathol.
118: 135-156.
Van den Mooter, M., and Swings, J. 1990. Numerical analysis of 295 phenotypic
features of 266 Xanthomonas strains and related strains and an improved taxonomy of
the genus. Int. J. Syst. Bacteriol. 40: 348-369.
Vauterin, L., Hoste, B., Yang, P., Alvarez, A., Kersters, K., and Swings, J. 1993.
Taxonomy of the genus Xanthomonas. Pages 157-192 in: Xanthomonas. J.G. Swings
and E.L. Civerolo, eds. Chapman & Hall, London, United Kingdom.
Vauterin, L., Swings, J., and Kersters, K. 1991a. Grouping of Xanthomonas
campestris pathovars by SDS-PAGE of proteins. J. Gen. Microbiol. 137: 1677-1687.
Vauterin, L., Swings, J., Kersters, K., Gillis, M., Mew, T. W., Schroth, M. N.,
Palleroni, N. J., Hildebrand, D. C., Stead, D. E., Civerolo, E. L., Hayward, A. C.,
Maraite, H., Stall, R. E., Vidaver, A. K., and Bradbury, J. F. 1990a. Towards an
improved taxonomy of Xanthomonas. Int. J. Syst. Bacteriol. 40: 312-316.
Vauterin, L., Vantomme, R., Pot, B., Hoste, B., Swings, J., and Kersters, K. 1990b.
Taxonomic analysis of Xanthomonas campestris pv. begoniae and X. campestris pv.
pelargonii by means of phytopathological, phenotypic, protein electrophoretic and
DNA hybridization methods. Syst. Appl. Microbiol. 13:166-176.

243
Vauterin, L., and Vauterin, P. 1992. Computer-aided objective comparison of
electrophoresis patterns for grouping and identification of microorganisms. Eur.
Microbiol. 1:37-41.
Vauterin, L., Yang, P., Hoste, B., Pot, B., Swings, J., and Kersters, K. 1992.
Taxonomy of xanthomonads from cereals and grasses based on SDS-PAGE of
proteins, fatty acid analysis and DNA hybridization. J. Gen. Microbiol. 138: 1467-
1477.
Vauterin, L., Yang, P., Hoste, B., Vancanneyt, M., Civerolo, E.L., Swings, J., and
Kersters, K. 1991b. Differentiation of Xanthomonas campestris pv. citri strains by
sodium dodecyl sulfate-polyacrylamide gel electrophoresis of proteins, fatty acid
analysis, and DNA-DNA hybridization. Int. J. Syst. Bacteriol. 41: 535-542.
Verdier, V., Dongo, P., and Boher, B. 1993. Assessment of genetic diversity among
strains o Xanthomonas campestris pv. manihotis. J. Gen. Microbiol. 139: 2591-2601.
Vemiere, C., Pruvost, O., Civerolo, E. L., Gambin, O., Jacquemoud-Collet, J. P., and
Luisetti, J. 1993. Evaluation of the Biolog substrate utilization system to identify and
assess metabolic variation among strains of Xanthomonas campestris pv. citri. Appl.
Environ. Microbiol. 59: 243-249.
Volcani, Z. 1969. The effect of mode of irrigation and wind direction on disease
severity caused by Xanthomonas vesicatoria on tomato in Israel. Plant Dis. Reptr. 53:
459-461.
Voloudakis, A. E., Bender, C. L., and Cooksey, D. A. 1993. Similarity between
copper resistance genes from Xanthomonas campestris and Pseudomonas syringae.
Appl. Environ. Microbiol. 59: 1627-1634.
Wayne, L. G., Brenner, D. J., Colwell, R. R., Grimount, P. A. D., Kandler, O.,
Krichevsky, M. I., Moore, L. H., Moore, W. E. C., Murray, R. G. E., Stackebrandt, E.,
Starr, M. P., and Truper, H. G. 1987. Report of the ad hoc committee on
reconciliation of approaches to bacterial systematics. Int. J. Syst. Bacteriol. 37: 463-
464.
Weber, K., and Osborn, M. 1960. The reliability of molecular weight determinations
by dodecyl sulfate-polyacrylamide gel electrophoresis. The Journal of Biological
Chemistry 244: 4406-4412.
White, H. E. 1930. Bacterial spot of radish and turnip. Phytopathology 20: 653-662.

244
Williams, O. B., and Glass, H. B. 1931. Agglutination studies on Phytomonas
malvaceara. Phytopathology 21:1181-1184.
/ Willis, K., Rich, J. J., and Hrabak, E. M. 1991. The hrp genes of phytopathogenic
bacteria. Mol. Plant-Microbe Interact. 4:132-138.
Yang, P., Vauterin, L., Vancanneyt, M., Swings, J., and Kersters, K. 1993.
Application of fatty acid methyl esters for the taxonomic analysis of the genus
Xanthomonas. Syst. Appl. Microbiol. 16:47-71.
Young, J. M., Takikawa, Y., Gardan, L., and Stead, D. E. 1992. Changing concepts in
the taxonomy of plant pathogenic bacteria. Annu. Rev. Phytopathol. 30: 67-105.

BIOGRAPHICAL SKETCH
Rui Pereira Leite, Jr., was bom in Sao Pedro, Sao Paulo, Brazil on May 26,
1957, to Ruy Pereira Leite and Daysy Fortes Pereria Leite. He attended Rafael de
Moura Campos Elementary and Instituto de Educado Cardoso de Almeida High
School, Botucatu, Sao Paulo, where he graduated in 1974. In 1975, Rui enrolled at the
Faculdade de Cincias Agronmicas of the Universidade Paulista Jlio de Mesquita
Filho UNESP, Botucatu, Sao Paulo, where he majored in agriculture and took courses
in plant pathology from Dr. Chukichi Kurozawa. Rui graduated with a B.S. from
Universidade Paulista Jlio de Mesquita Filho UNESP in 1978. After graduating, Rui
took the position of plant pathologist in the Department of Plant Pathology at the
Funda9o Instituto Agronmico do Paran IAPAR, Londrina, Paran. At IAPAR,
Rui carried out research on characterization and control of bacterial diseases of fruit
crops. In August 1984, Rui got a study leave and enrolled in the graduate program of
the Department of Plant Pathology, at the University of Wisconsin, Madison. In
Madison, Rui worked with Dr. Doug Rouse on the ecology of Pseudomonas syringae
pv. syringae that causes the bacterial brown spot on snap beans. Rui graduated with a
M.S. degree in plant pathology from the University of Wisconsin in 1986 and returned
to his position in the Department of Plant Pathology at IAPAR. In August 1990, Rui
enrolled in the graduate program at the University of Florida, Department of Plant
Pathology, with Dr. Robert Stall as advisor. Rui was married in 1980 to Claudia
Stulzer, and they have two sons, Ruy and Rafael.
245

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.
Robert E. Stall, Chair
Professor of Plant Pathology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.
Professor of Plant Pathology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.
Daryl R. Pring'
Professor of Plant Pathology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.
In
Professor of Microbiology and Cell
Science

This dissertation was submitted to the Graduate Faculty of the College of
Agriculture and to the Graduate School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy. /I
August 1994
Dean, College of Agriculture
Dean, Graduate School



62
performed by sequential chloroform-isoamyl alcohol (24:1) and phenol-chloroform-
isoamyl alcohol (25:24:1) extractions. For each extraction, the samples were hand
shaken continuously and gently for 10 min and centrifuged for 5 min at 16,000 g.
After isopropanol precipitation, the DNA pellet was washed with 70% ethanol, and
then dried under vacuum for 20 min. The pellet was redissolved in 100 pi of TE and
stored at 4C.
DNA amplification
Two sets of oligonucleotide primers from sequences of the hrp gene cluster of
Xanthomonas campestris pv. vesicatoria (Bonas et al., 1991) were used in this study.
The primers RST2 and RST3 delineated an 840-bp region and the primers RST21 and
RST22 delineated an 1,075-bp region of the complementation groups B and C/D of the
hrp gene cluster of X. campestris pv. vesicatoria, respectively (Chapter 3; Bonas et al.,
1991). Oligonucleotide primers were synthesized with a model 394 DNA Synthesizer
(Applied Biosystems, Foster City, CA) by the DNA Synthesis Laboratory, University
of Florida, Gainesville.
The DNA sequences were amplified in a reaction mixture of 50 pi containing 5
pi of 10X buffer (500 mM KC1, 100 mM Tris-Cl [pH 9.0 at 25C], 1% Triton X-100),
1.5 mM MgCl2, 200 pM of each deoxynucleotide triphosphate (Boehringer
Mannheim), 25 pmol of each primer, 1.25 units of Taq polymerase (Promega, Madison,
WI), and 100 ng of purified template DNA. The reaction mixture was overlaid with 50
pi of light mineral oil. Thirty amplification cycles were carried out in an automated
thermocycler PT-100-60 (MJ Research, Watertown, MA) according to the following
profiles: 30 sec of denaturation at 95C, 30 s of annealing at 62C, and 45 s of
extension at 72C for the primers RST2 and RST3 and 30 s of denaturation at 95C, 40


116
0.53 to 1.00 for the hrpC/D and hrpB regions, respectively (Table 5-1). Similarly, the
clades 3, 5, 7, and 8 also contain pathovars ofX. campestris that cause diseases on
different plants but are genetically very closely related regarding the hrp genes for both
regions examined (Table 5-1). The bootstrap resampling analyses further substantiate
the monophyletic nature of the gene cluster for these clades (Fig. 5-1 and 5-2).
On the contrary, the inferred phylogenetic analysis also indicates that strains
which are classified under the same pathovar, because they cause similar diseases on
the same host plants, may have very distinct hrp sequences. X. campestris pv.
vesicatoria that causes disease on solanaceous plants is known to contain at least two
diverse groups of strains (Stall et al., 1994). The two groups of this pathovar also show
distinct hrp gene sequences (Fig. 5-1 and 5-2). Strains ofX. campestris pv. vesicatoria
B comprise the sole taxon of the clade 1 whereas the strains of X. campestris pv.
vesicatoria A belong to the clade 7 (Fig. 5-1 and 5-2). The clade 7 is the largest one
and includes several pathovars ofX. campestris (Fig. 5-1 and 5-2). The placement of
the two groups of X. campestris pv. vesicatoria in two distant and distinct clades are
further supported by the bootstrap resampling analysis (Fig. 5-1 and 5-2). The
estimated similarity between the two groups of X. campestris pv. vesicatoria in relation
to the hrp regions ranged from 0.33 to 0.38 and from 0.50 to 0.51 for the hrpC/D and
hrpB regions, respectively. In addition, the three subgroups identified \nX. campestris
pv. vesicatoria A are monophyletic and genetically closely related for both hrp gene
regions (Fig. 5-1 and 5-2). Several other pathovars of X. campestris accommodate
groups of strains with very diverse hrp genes, i.e. X campestris pv. fici, X campestris
pv. glycines, X. campestris pv. phaseoli, X. campestris pv. poinsettiicola, andX.
campestris pv. vitians (Fig. 5-1 and 5-2; Table 5-2).


200
Table A-2Continued
Strain
Relevant characteristics
Source or reference
pXV56/3-48
pLARF3 cosmid clone from X. c. pv.
vesicatoria XV56
This study
pXCP58/2
pLARF3 cosmid clone from X. c. pv.
pelargonii
This study
aBRL, Bethesda Research Laboratories, Gaithersburg, MD; Stratagene, Stratagene, Inc., La Jolla, CA.


138
M 1 2 3 4 5 6 7 8 9 10 11 12 13
M 1 2 3 456 78 9 10 11 12 13
Fig. 6-3. Restriction analysis of the 1,075-bp DNA fragment of the complementation
groups C/D of the hrp gene cluster amplified from strains of Xanthomonas campestris
and restricted with the endonucleases (A) HaeIII and (B) Cfol. Lanes: M, phage X
restricted with Pstl; 1 to 3 strains FI, F6, and FI 00 of A! c. pv. citrumelo, respectively;
4, X. c. pv. maculifoliigardeniae strain X22j; 5, X. campestris from Strelitzia reginae
strain XI98; 6, X c. pv. fici strain X151; 7,X c. pv. alfalfae strain 82-1; 8, X c. pv.
bilvae strain XCB; 9, X. campestris from Feronia sp. strain XCF; 10 to 12, strains
9771, B84, and 339 of A! c. pv. citri, respectively; 13, A! c. pv. vesicatoria strain 75-3.
Molecular sizes are given in bases.


Table D-lContinued
Species/Pathovar Species/Pathovar
23. pv. dieffenbachiae A
24. pv. dieffenbachiae A
25. pv. dieffenbachiae B
26. pv. fici A
27. pv. fici A
28. pv. fici A
29. pv. fici B
30. pv. gardneri
31. pv. glycines A
32. pv. glycines B
33. pv. holcicola
34. pv. incanae
35. pv. maculifoliigardeniae
36. pv. malvacearum
37. pv. manihotis
38. pv. papavericola
39. pv. pelargonii
40. pv. phaeoli A
41. pv. phaseoli B
42. pv. phaseoli "fuscans"
44. pv. poinsettiicola B
45. pv. poinsettiicola A
53 54 55
.2500
.2500
.7755
.2917
.2083
.8163
.4545
.3182
.6667
.4167
.4583
.5306
.3265
.6939
.4800
.3265
.6939
.4800
.3182
.3182
.2222
.5455
.3182
.3111
.5000
.7500
.4082
.3182
.3182
.2222
.3750
.1667
.1633
.5000
.3182
.2667
.4348
.2609
.7660
.4444
.6667
.3478
.4082
.2449
.9200
.6222
.4000
.3043
.5778
.4000
.3043
.3404
.2128
.8750
.5000
.7500
.4082
.4348
.8261
.3830
.4186
.3256
.2727
.4167
.4583
.5306
56 57 58
.3111
.2273
.4583
.2667
.2273
.4583
.3415
.3500
.7273
.4889
.3636
1.000
.4783
.4000
.7347
.4783
.4000
.7347
.2439
.4000
.2273
.3415
.9500
.3636
.8000
.4091
.5833
.2439
.4000
.2273
.1778
.3182
.1667
.3415
.8500
.2727
.4186
.3333
.5652
1.000
.3902
.4889
.3478
.3111
.5714
.4286
.8780
.3556
.4286
.9268
.3111
.3182
.2326
.5106
.8000
.4091
.5833
.6512
.2857
.4783
.2500
.4615
.2791
.4889
.3636
1.000
59 60 61
.3333
.1395
.0833
.3333
.1860
.0833
.5000
.2564
.1364
.7083
.2791
.1667
.6939
.3636
.2449
.6939
.3636
.2449
.3182
.7692
.3636
.3182
.4615
.3636
.8333
.2791
.2917
.3182
.7692
.3636
.1667
.2326
.2083
.3182
.4615
.4545
.4348
.2927
.1739
.7556
.2500
.2667
.4082
.2727
.1224
.4000
.5000
.4000
.4000
.5000
.4444
.3830
.1905
.0851
.8333
.2791
.2917
.6957
.2439
.2609
.3256
1.000
.3256
.7083
.2791
.1667
Continued on following page


28
was further supported on the basis of DNA homology and phenotypic characteristics
(Egel et al., 1991; Graham et ah, 1990).
Nucleic acid based techniques have also been examined for the specific
detection of plant pathogenic xanthomonads (Garde and Bender, 1991; Gilbertson et
ah, 1989; Hartung, 1992; Hartung et ah, 1993). Garde and Bender (1991) developed
DNA probes for detection of copper resistance in X. campestris pv. vesicatoria.
Although the DNA probes seem specific to the genes conferring resistance to copper in
X. campestris pv. vesicatoria and apparently did not hybridize to DNA of copper
resistant P. syringae pv. tomato (Garde and Bender, 1991), the cop operon of P.
syringae pv. tomato not only hybridized to copper sensitive bacteria (Cooksey et ah,
1990), but also to strains of ^ campestris pv. vesicatoria (Voloudakis et ah, 1993). In
another study, Gilbertson et ah (1989) used DNA probes obtained from a cryptic
plasmid for specific detection of the bean pathogens X. campestris pv. phaseoli and X
campestris pv. phaseoli "fiiscans". Although the DNA probes did not hybridize to
nonpathogenic xanthomonads isolated from bean plants, they did hybridize to total
genomic DNA of bacterial strains of other pathovars ofX. campestris (Gilbertson et ah,
1989). However, strains of these other pathovars of A! campestris were thought
unlikely to be associated with bean plants. Plasmid based DNA probes have also been
used for detection of the citrus canker pathogen, X. campestris pv. citri (Hartung,
1992). The DNA probes were developed for specific detection of the citrus canker
pathogen, X. campestris pv. citri which include four different groups. The DNA probes
were highly specific for the citrus canker pathogens and did not hybridize to strains of
the citrus bacterial spot pathogen, X. campestris pv. citrumelo. However, cross
hybridization was observed in strains of X campestris pv. bilvae and X campestris pv.
vignicola (Hartung, 1992). Furthermore, a DNA amplification-based procedure was


58
determine the intrageneric relationships among different xanthomonads (Hildebrand et
al., 1990; Palleroni et al., 1993; Vauterin et al., 1993) and 21 DNA homology groups
were established (Vauterin et al., 1993). Some DNA homology groups were delineated
which include pathovars of X. campestris causing diseases on closely related plants as
occurs with the xanthomonads pathogenic on leguminous (Hildebrand et al., 1990;
Palleroni et al., 1993; Vauterin et al., 1993), gramineous (Vauterin et al., 1993), and
cruciferous plants (Vauterin et al., 1993). Furthermore, these groups established based
on DNA homology are strongly supported by the groupings based on SDS-PAGE of
proteins and fatty acid analyses (Vauterin et al., 1991a, 1992; Yang et al., 1993).
However, the genetic relatedness among the xanthomonads is not always correlated
with pathogenicity features.
A striking feature of the plant pathogenic xanthomonads is the genetic diversity
among strains of the same pathovar or different pathovars of X. campestris that cause
diseases on the same or highly related hosts. For example, strains ofX. campestris pv.
vesicatoria groups A and B that cause diseases on solanaceous plants are only about
33% related on the basis of DNA homology (Stall et al., 1994). Further, this low
genetic similarity and the differences in phenotypic features strongly supports the
conclusion that these two groups of X. campestris pv. vesicatoria may even belong to
different species (Stall et al., 1994). In another study, Egel et al. (1991) examined the
genetic relatedness of the xanthomonads causing diseases of citrus and found that
strains of the citrus canker pathogen X. campestris pv. citri and strains of the citrus
bacterial spot agent X campestris pv. citrumelo are less than 60% related to one
another. The diverse nature of the X. campestris strains causing diseases on citrus has
also been determined on the basis of genomic fingerprinting (Hartung and Civerolo,
1987), analysis of restriction fragment length polymorphism (RFLP) (Gabriel et al.,


Table D-lContinued
Species/Pathovar Species/Pathovar
22
23
24
25
26
27
28
29
30
31
1. pv. alfalfae
.8750
.4167
.4167
.6818
.9167
.7755
.7755
.2727
.3636
.5833
2. pv. alfalfae
.8333
.4167
.4167
.6818
.8750
.7347
.7347
.2727
.3636
.5833
3. pv. armoraceae
.2727
.1818
.1818
.2500
.2727
.3556
.3556
.4000
.9000
.3636
4. pv. begoniae A
.5455
.6364
.6818
.7500
.5909
.3556
.3556
.2000
.3500
.3636
5. pv. begoniae A
.5652
.6957
.7391
.7143
.5652
.3830
.3830
.2381
.3333
.3913
6. pv. begonia B
.5652
.6522
.6957
.7143
.5652
.4255
.4255
.2857
.3810
.3913
7. pv. bilvae
.6667
.3333
.3333
.5000
.7083
.6939
.6939
.3636
.3182
.7917
8. pv. campestris
.2727
.1818
.1818
.2500
.2727
.3556
.3556
.4000
.9000
.3636
9. pv. carotae A
.3182
.2273
.2273
.3000
.3182
.4000
.4000
.4500
.8500
.4545
10. pv. carotae B
.2727
.1818
.1818
.2500
.2727
.3556
.3556
.4000
.9000
.3636
11. pv. citri A
.5417
.2917
.2500
.5000
.5833
.5306
.5306
.2273
.3636
1.000
12. pv. citri B
.4583
.2500
.2083
.3182
.4583
.6939
.6939
.3182
.3182
.7500
13. pv. citri C
.4583
.2500
.2083
.3182
.4583
.6939
.6939
.3182
.3182
.7500
14. pv. citrumelo
.8163
.4898
.4898
.4889
.7347
.8400
.8400
.3111
.3556
.4898
15. pv. citrumelo
.8163
.4490
.4490
.6667
.8980
.8000
.8000
.2667
.3556
.6122
16. pv. citrumelo
.7083
.4167
.4167
.6818
.9167
.7347
.7347
.2273
.3636
.5833
17. pv. citrumelo
.7083
.4167
.4167
.6818
.9167
.7347
.7347
.2273
.3636
.5833
18. pv. citrumelo
.7917
.4583
.4583
.7273
1.000
.7347
.7347
.2273
.3636
.5833
19. pv. citrumelo
.7234
.4255
.4255
.7442
.9362
.6667
.6667
.2326
.3721
.5957
20. pv. citrumelo
.7083
.4167
.4583
.6364
.7917
.8163
.8163
.2727
.3636
.4583
21. pv. citrumelo
.8333
.4167
.4583
.7273
.7917
.7347
.7347
.3182
.3636
.5833
22. pv. citrumelo
.
.5000
.5417
.6818
.7917
.7347
.7347
.2727
.3182
.5417
Continued on following page
K>
o


92
pathogenic xanthomonads. DNA sequences related to the hrp genes were specifically
amplified from bacterial strains representing X. fragariae and at least 28 pathovars of
X. campestris. However, the hrp-related sequences were not amplified from strains of
the plant pathogenic xanthomonads X. albilineans, X. campestris pv. celebensis, X.
campestris pv. secalis, and X. campestris pv. translucens, and from a few members of
the pathovars X. campestris pv. fici, X. campestris pv. pelargonii, X campestris pv.
phaseoli, X. campestris pv. poinsettiicola, and X. campestris pv. pruni. Strains of these
xanthomonads may not be highly related genetically to the other pathovars ofX.
campestris included in the study as determined by DNA homology studies (Hildebrand
et al., 1990; Palleroni et al., 1993; Vauterin et al., 1993) and by the failure to hybridize
strongly to the hrp clones ofX. campestris pv. vesicatoria (Chapter 3; Bonas et al.,
1991; Stall and Minsavage, 1990). These results support the contention of the presence
of differences in the DNA sequences of Xanthomonas corresponding to one or both
primers used.
Although, no size variation was observed for the /zrp-related fragments
amplified from different plant pathogenic xanthomonads, the restriction analysis of
these fragments revealed the presence of sequence variation. Fragment length
polymorphisms in the hrp regions were further explored for differentiation of the
different groups of plant pathogenic xanthomonads. The restriction banding profile
established for the two hrp-related fragments digested with four endonucleases ranged
from 7 to 25 different profiles though only a very few groups of xanthomonads, i.e. X.
campestris pv. holcicola, X campestris pv. vesicatoria group B, and A! fragariae,
produced unique restriction banding profiles (Table 4-2). This was somewhat expected
because these three groups of xanthomonads seem to be genetically unique (Hildebrand
et al., 1990; Palleroni et al., 1993; Vauterin et al., 1993). For example, A! campestris


244
Williams, O. B., and Glass, H. B. 1931. Agglutination studies on Phytomonas
malvaceara. Phytopathology 21:1181-1184.
/ Willis, K., Rich, J. J., and Hrabak, E. M. 1991. The hrp genes of phytopathogenic
bacteria. Mol. Plant-Microbe Interact. 4:132-138.
Yang, P., Vauterin, L., Vancanneyt, M., Swings, J., and Kersters, K. 1993.
Application of fatty acid methyl esters for the taxonomic analysis of the genus
Xanthomonas. Syst. Appl. Microbiol. 16:47-71.
Young, J. M., Takikawa, Y., Gardan, L., and Stead, D. E. 1992. Changing concepts in
the taxonomy of plant pathogenic bacteria. Annu. Rev. Phytopathol. 30: 67-105.


88
Fig. 4-11. Restriction analysis of the (A) 840 bp and (B) 1,075 bp DNA fragments of
the hrp gene cluster amplified from strains of Xanthomonas campestris pv.
dieffenbachiae and restricted with HaeIII. Lanes M, phage A restricted with Pstl.
Lanes 1, X422; 2, X757; 3, X790; 4, X1272; 5, X260; 6, X763; 7, X736; 8, X738; 9,
X745; 10, X729. Molecular sizes are given in bases.


97
campestris pv. citrumelo has also been reported previously (Egel et al., 1991; Gabriel
et al., 1988, 1989; Graham et al., 1990; Hartung and Civerolo, 1987, 1989).
The restriction banding profiles of some strains o X. campestris pv. citrumelo
were very similar or even identical to the profiles of other pathovars oX. campestris.
For instance the strains F59 and F86 X. campestris pv. citrumelo have profiles for both
/irp-related fragments identical to strains of X campestris pv. pruni whereas other
strains of X. campestris pv. citrumelo were closely related to strains of X. campestris
pv. fici A, X. campestris pv. poinsettiicola A, and X campestris pv. vesicatoria A
(Table 4-2). The close genetic relatedness of strains of X. campestris pv. citrumelo to
other pathovars of X. campestris has also been determined previously (Egel et al.,
1991; Gabriel et al., 1988, 1989; Graham et al., 1990; Hartung and Civerolo, 1987,
1989).
The X. campestris pathovars, fici, poinsettiicola, vesicatoria, and vitians seem to
comprise of a very diverse group of bacteria (Stall et al., 1994; Vauterin et al., 1991a;
Yang et al., 1993). Our results on the genetic analysis of the hrp-re\a{ed sequences also
corroborate the existence of distinct groups of strains within these pathovars. Further,
the grouping obtained based on the analysis of the /wy?-related sequences agrees very
closely with the grouping established previously on the basis of genetic or phenotypic
features or both. For example, strains of X. campestris pv. vesicatoria were grouped
into two distinct groups based on the restriction banding profile of the hrp-related
sequences (Table 4-2). The groups established based on the /?r/?-related analysis are
genetically highly uniform and they correspond to the previous existing groups A and
B oiX. campestris pv. vesicatoria (Stall et al., 1994). On the other hand, the pathovars
fici and poinsettiicola have a more complex picture. Strains of both pathovars were
also divided into two distinct RFLP groups. Further, the group A oiX. campestris pv.


59
1988, 1989; Graham et al., 1990; Hartung and Civerolo, 1989), and fatty acid
composition (Graham et al., 1990; Stall and Hodge, 1989).
The genetic diversity of the xanthomonads has also been investigated by using
other nucleic acid-based techniques, such as genomic fingerprinting with frequent- and
rare-cutting restriction endonucleases (Egel et al., 1991; Hartung and Civerolo, 1987;
Pruvost et al., 1992) and RFLP analyses of plasmid and genomic DNAs (Berthier et al.,
1993; DeParasis and Roth, 1990; Gabriel et al., 1988, 1989; Garde and Bender, 1991;
Gilbertson et al., 1989; Gottwald et al., 1991; Graham et al., 1990; Hartung, 1992;
Hartung and Civerolo, 1989, 1991; Lazo and Gabriel, 1987; Lazo et al., 1987; Leach et
al., 1990, 1992; Qhobela and Claflin, 1988, 1992; Verdier et al., 1993). In an extensive
study, Lazo et al. (1987) differentiated strains representing 26 different pathovars of X.
campestris based on RFLP analysis by using random DNA probes obtained from a
cloned fragment derived from X. campestris pv. citrumelo. The RFLP analyses
revealed profiles highly conserved and characteristic for each pathovar tested. It was
possible to differentiate all the pathovars of X. campestris included in the study by
using more than one DNA probe or by digesting the genomic DNA with different
restriction endonucleases (Lazo et al., 1987). Plasmid DNA fragments have also been
examined as probes for specific identification of certain groups of X. campestris (Garde
and Bender, 1991; Gilbertson et al., 1989; Hartung, 1992; Hartung et al., 1993; Lazo
and Gabriel, 1987). However, DNA probes derived from specific regions in the
bacterial genome and conserved among plant pathogenic bacteria may be more
meaningful for specific differentiation and characterization of strains, pathovars, or
species of Xanthomonas.
The hypersensitive reaction and pathogenicity (hrp) gene cluster that has been
found in several plant pathogenic bacteria may be appropriate for specific identification


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EP7GRA2LC_YXQOOV INGEST_TIME 2017-07-24T20:23:00Z PACKAGE AA00029802_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


6
and identification of this group of bacteria (Link and Sharp, 1927; Williams and Glass,
1931). Elrod and Braun (1947) carried out an extensive study which included different
species of Xanthomonas and several pathovars of X. campestris, thus identifying five
immunological groups which have a high degree of serological homogeneity within
them. However, the serogroups did not show good or significant correlation with
nomenspecies and pathotypes. Subsequent work has confirmed the complex nature of
the serological relationships among different taxonomic groups of the plant pathogenic
xanthomonads.
Polyclonal antibodies produced against whole cells ofX. campestris pv.
campestris cross-reacted with strains of X. campestris pv. armoraciae, X. campestris pv.
phaseoli "fuscans" and X campestris pv. vesicatoria (Franken et al., 1992). In
immunofluorescence microscopy, these polyclonal antibodies also cross-reacted with
non-xanthomonad bacteria. In a similar study, polyclonal antibodies produced to
whole cells of X campestris pv. pelargonii also reacted to strains of seven different
pathovars of campestris (Anderson and Nameth, 1990). Conflicting data have also
emerged regarding the relationship between serotypes and pathotypes. O'Brien et al.
(1967) distinguished pepper strains from tomato strains ofX. campestris pv. vesicatoria
by serological reaction, but no correlation between serotype and pathotype of X.
campestris pv. vesicatoria was found by Charudattan et al. (1973). They found that
tomato and pepper strains of the bacterium included both serotypes. Further, no
correlation was observed for hydrolysis of starch and resistance to streptomycin with
serotypes. These contrasting results are not exclusive to X. campestris pv. vesicatoria,
but have also been observed for other pathovars of X. campestris (Bach et ah, 1978;
Civerolo and Fan, 1982) and even to other plant pathogenic species of Xanthomonas
(Obata and Tsuboi, 1972; Mahanta and Addy, 1977). Furthermore, the heterogeneous


APPENDIX A
BACTERIAL STRAINS AND PLASMIDS USED IN THIS STUDY
Table A-l. List of bacterial strains used in this study and their source.
Pathovar
Strain
Source3
Xanthomonas
X. campestris
pv. alfalfae
G-22 (KS)
DLS
82-1
RES
pv. armoraciae
63-27,417, 756
JBJ
pv begoniae
XCB9, X274, X281, X329, X610, X627,
XI490, XI492, XI496
ARC
pv. bilvae
XCB
ELC
pv. campestris
33913
ATCC
62-1, 62-9a, 65-6b, 70-3, 70-5, 71-2, 83-1, 83-
2
JBJ
pv. carotae
#3, #5, #7, #9, #12, #13, #16
RES
pv. celebensis
6207
ARC
pv. citri
DPI
canker A
3213, 3340, 9760-2, 9771, Tl, 115A
canker B
B64, B69, B80, B84, B93, B94, B148
canker C
70C, 338, 339, 340, 341,342
pv. citrumelo
FI, F6, F54, F59, F86, F94, F100, F228,
F254, F274, F306, F311, F348, F361, F378,
3166
DPI
pv. dieffenbachiae
X729
RES
Continued on following page
195


140
combinations of fragment and endonuclease, the restriction pattern of the highly
aggressive strains were similar to some strains of the moderately or weakly aggressive
groups, these include the 1,075-bp fragment from the weakly aggressive strain FI00
restricted with Cfo\ (Fig. 6-4A) and the fragment from the moderately aggressive strain
F378 restricted with HaeIII (Fig. 6-4B). However, the overall banding patterns of the
combinations of two fragments and four endonucleases for the highly aggressive strains
were different from the patterns obtained for the strains of the moderately and weakly
aggressiveness groups of X. campestris pv. citrumelo. In contrast to the highly
aggressive group, restriction fragment polymorphism was evident for the strains within
the moderately and weakly aggressive groups of X. campestris pv. citrumelo (Fig. 6-4).
The weakly aggressive strains F94, FI00, and F306 had banding patterns almost
identical to A! campestris pv. vesicatoria strain 75-3 (Fig. 6-4). Also, strains of the
highly aggressive group of X. campestris pv. citrumelo had banding patterns similar to
X. campestris pv. fici strain XI51 (Fig. 6-3 and 6-4).
In contrast to the diversity of the moderately and weakly aggressive strains of
X campestris pv. citrumelo, strains of X. campestris pv. citri of the groups A, B, and C
each produced characteristic restriction patterns (Fig. 6-5). The banding patterns of all
strains of group A were identical when restricted with either HaelW (Fig. 6-5), Cfol,
Sau3M, or Taql (data not shown). Similarly, strains of groups B and C were also
homogeneous within each group (Fig. 6-5), as well as between the two groups for the
four endonucleases and the two hrp gene cluster fragments.
Genetic relationships of the hrp from different strains of X campestris
Differences in the number of common restriction fragments from the amplified
DNA of the hrp gene cluster indicated that there is variation in the relatedness of the


GENETIC AND EVOLUTIONARY CHARACTERIZATION OF
PLANT PATHOGENIC XANTHOMONADS BASED ON DNA
SEQUENCES RELATED TO THE hrp GENES
By
RUI PEREIRA LEITE, Jr.
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1994


64
Result?
Amplification of the /irp-related fragments from strains of X. campestris and related
Xanthomonas spp.
The hrp-related fragments delineated by the primers RST2 plus RST3 and
primers RST21 plus RST22 were amplified by polymerase chain reaction from plant
pathogenic strains representing^ fragariae and 28 different pathovars ofX. campestris
(Table 4-1). The fragments amplified were of identical sizes in all cases with both sets
of primers (data not shown). Low DNA yield was obtained for the 840-bp hrp-related
fragment amplified with the primers RST2 and RST3 for all strains of X. campestris
pv. carotae, X campestris pv. gardneri, X. campestris pv. papavericola, X campestris
pv. pelargonii, X. campestris pv. taraxaci, and two strains of X. campestris pv. vitians
(Table 4-1). The 840 bp hrp fragment was not amplified from a strain ofX campestris
pv. holcicola (Table 4-1) whereas both hrp fragments were not amplified from strains
of X. albilineans,X. campestris pv. celebensis, X campestris pv. secalis, X. campestris
pv. translucens, and nonpathogenic X. campestris (Table 4-1). Moreover, both hrp
fragments were not amplified also from some strains of a few other pathovars of X.
campestris which includes X. campestris pv. fici, X campestris pv. pelargonii, X
campestris pv. phaseoli, X campestris pv. poinsettiicola, and X campestris pv. pruni
(Table 4-1).
Establishment of the restriction banding pattern profiles
The hrp-related fragments were amplified from 192 strains of plant pathogenic
xanthomonads representing X fragariae and 28 different pathovars of X campestris
(Table 4-1). Restriction fragment length polymorphism occurred in both /irp-related


Table 7-1. Phenotypic and genetic characterization of strains of Xanthomonas campestris associated with pepper and tomato
seed and seedling.
Strain
Origin
Reaction on
plants
Amplification
of hrp fragment
Tentative identification
Source3
pepper tomato
hrpB
hrpC/D
hrp analysis3
FAAC
SP1.92
pepper seed
_d
HR
+e
+
pv. campestrisf
pv. campestris (0.820)
GOK
SP101.92
pepper seed
-
HR
+
+
pv. campestris
pv. campestris (0.463)
GOK
SP268.92
pepper seed
-
-
-
-
na§
pv. armoraciae (0.703)
GOK
SP290.92
pepper seed
-
HR
-
-
na
pv. pruni (0.852)
GOK
3118-GP
pepper seed
+
HR
+
+
pv. vesicatoria A
pv. vesicatoria (0.735)
GOK
7502-2FS1
pepper seed
-
HR
-
-
na
pv. raphani (0.677)
JD
157
pepper seed
HR
HR
+
+
nd
pv. juglandis (0.644)
WW
T1083
tomato seed
-
-
-
-
na
pv. celebensis (0.051)
WW
524A-1
pepper seed
HR
HR
+
+
pv. campestris
pv. armoraciae (0.768)
WW
524A-2
pepper seed
HR
HR
+
+
pv. campestris
pv. armoraciae (0.812)
WW
75-0-3
tomato seed
HR
+
+
+
pv. vesicatoria C
pv. manihotis (0.835)
WW
7B-0-1
tomato seed
HR
+
+
+
pv. vesicatoria C
pv. fici (0.773)
WW
T1087
tomato seed
HR
HR
+
+
pv. citrumelo
pv. poinsettiicola (0.789)
WW
l-A-0-1
pepper seed
+
+
-1-
+
pv. vesicatoria A
pv. vesicatoria (0.745)
WW
P996
pepper seed
-
-
-
-
na
pv. raphani (0.581)
WW
DM-1
pepper seed
-
+
+
+
nd
pv. juglandis (0.618)
WW
LM-1
tomato seed
-
HR
-
(+)
nd
pv. malvacearum (0.644)
WW
639-6/FS1
pepper seed
-
-
-
-
na
nd
JW
639-6/FS2
pepper seed
-
-
-
-
na
nd
JW
Continued on the following page


52
Fig. 3-7. Amplification of the 355-bp (A) and (B) 840-bp fragments of
complementation group B of the hrp gene cluster from samples with different amounts
of DNA template of Xanthomonas campestris pv. vesicatoria 75-3. Lanes: M, phage X
restricted with £coRI and HindUl; 1, 25 ng; 2, 2.5 ng; 3, 0.25 ng; 4, 25 pg; 5,2.5 pg; 6,
0.25 pg; 7, 0.025 pg. Molecular sizes are given in base pairs.


229
Chase, A. R., Stall, R. E., Hodge, N. C., and Jones, J. B. 1992. Characterization of
Xanthomonas campestris strains from aroids using physiological, pathological, and
fatty acid analyses. Phytopathology 82: 754-759.
Civerolo, E. L. 1984. Bacterial canker disease of citrus. J. Rio Grande Val. Hortic.
Soc. 37: 127-146.
Civerolo, E L., and Fan, F. 1982. Xanthomonas campestris pv. citri detection and
identification by enzyme-linked immunosorbent assay. Plant Dis. 66: 231-236.
Clark, M. F., and Adams, A. N. 1977. Characteristic of the microplate method of
enzyme-linked immunosorbent assay for the detection of plant viruses. J. Gen. Virol.
34: 475-483.
Cooksey, D. A., Azad, H. R., Jae-Soon, C., and Chun-Keun, L. 1990. Copper
resistance gene homologs in pathogenic and saprophytic bacterial species from tomato.
Appl. Environ. Microbiol. 56: 431-435.
Cooksey, D. A., and Graham, J. H. 1989. Genomic fingerprinting of two pathovars of
phytopathogenic bacteria by rare-cutting restriction enzymes and field inversion gel
electrophoresis. Phytopathology 79: 745-750.
Cox, R. S. 1966. The role of bacterial spot in tomato production in south Florida.
Plant Dis. Rep. 50: 699-700.
Davis, M. J., French, W. J., and Schaad, N. W. 1981. Axenic culture of the bacteria
associated with phony disease of peach and plum. Curr. Microbiol. 6: 309-314.
DeParasis, J., and Roth, D.A. 1990. Nucleic acid probes for identification of
phytobacteria: Identification of genus-specific 16s rRNA sequences. Phytopathology
80: 618-621.
DeVos, P., and DeLey, J. 1983. Intra- and intergeneric similarities of Pseudomonas
dead Xanthomonas ribosomal ribonucleic acid cistrons. Int. J. Syst. Bacteriol. 33: 487-
509.
DeVos, P., Goor, M., Gillis, M., and DeLey, J. 1985. Ribosomal ribonucleic acid
cistron similarities of phytopathogenic Pseudomonas species. Int. J. Syst. Bacteriol.
35: 169-184.
Dickey, R. S., and Zumoff, C. H. 1987. Bacterial leaf blight of Syngonium caused by a
pathovar of Xanthomonas campestris. Phytopathology 77:1257-1262.


B
C
Fig. 3-4. Hybridization of the hrp fragments amplified from Xanthomonas campestris
pv. vesicatoria 75-3 to total genomic DNA of strains o Xanthomonas campestris.
Approximately 3 pg of £coRI-digested DNA was loaded per lane. The blots were
probed with the labeled 355-bp (A), 840-bp (B), and 1,075-bp (C) fragments amplified
froml campestris pv. vesicatoria 75-3. Lanes: l,X campestris pv. vesicatoria 75-3,
BamHl digested; 2, X campestris pv. vesicatoria 75-3; 3, X. campestris pv. vesicatoria
XV56; 4, X campestris pv. alfalfae KS; 5, X. campestris pv. begoniae XCB9; 6, X.
campestris pv. campestris 33913; 1,X. campestris pv. glycines 87-2; 8, X. campestris
pv. holcicola G-23; 9, X. campestris pv. malvacearum RIATC; 10, X. campestris pv.
phaseoli 85-6; \ \,X. campestris pv. pruni FLA1; \2,X. campestris pv. secalis
XC129C; 13, X. campestris pv. translucens 80-1. Molecular sizes are given in kilobase
pairs.


47
Fig. 3-5. Amplification of the 355-bp (A), 840-bp (B), and 1,075-bp (C) fragments of
the hrp gene cluster from strains of Xanthomonas campestris. Lanes: M, phage A.
restricted with £coRI and Hincll; 1, X. campestris pv. vesicatoria 75-3; 2, X
campestris pv. bilvae XCB; 3, X. campestris pv. carotae #13; 4, X. campestris pv. citri
9771; 5,X campestris pv. citrumelo FI; 6,X. campestris pv. dieffenbachiae 729; 7,X
campestris pv. fci X151; 8, X. campestris pv. gardneri XG101; 9, X. campestris pv.
maculifoliigardeniae X22j; 10,X campestris pv. manihotis Xml25D; \\,X.
campestris pv. pelargonii XCP58; 12, X. campestris pv. phaseoli "fuscans" XP163A;
\2>,X. campestris pv. poinsettiicola 071-424; 14, X campestris pv. taraxaci XT11A;
\5,X. campestris pv. vignicola 81-30; 16, X. campestris pv. vitians XVIT; 17, X
campestris pv. physalidicola XP172. Molecular sizes are given in base pairs.


227
Azad, H., and Schaad, N. W. 1988. Serological relationships among membrane
proteins of strains of Xanthomonas campestris pv. translucens. Phytopathology 78:
272-277.
Bach, E. E., Alba, A. P. C., Pereira, A. L. G., Zagatto, A. G., and Rossetti, V. 1978.
Serological studies of Xanthomonas citri (Hasse) Dowson. Arq. Inst. Biol. Sao Paulo
45: 229-236.
Bashan, Y., Okon, Y., and Henis, Y. 1982. Long-term survival of Pseudomonas
syringae pv. tomato and Xanthomonas campestris pv. vesicatoria in tomato and pepper
seeds. Phytopathology 72: 1143-1144.
Beaulieu, C., Minsavage, G. V., Canteros, B. I., and Stall, R. E. 1991. Biochemical
and genetic analysis of a pectate lyase gene from Xanthomonas campestris pv.
vesicatoria. Mol. Plant-Microbe Interact. 4:446-451.
u Beer, S. V., Bauer, D. W., Jiang, X. H., Laby, R. J., Sneath, B. J., Wei, Z.-M., Wilcox,
D. A., and Zumoff, C. H. 1991. The hrp gene cluster of Erwinia amylovora. Pages
53-60 in: Advances in Molecular Genetics of Plant-Microbe Interactions, vol. 1. H.
Hennecke and D. P. S. Verma, eds. Kluwer Academic Publishers, Dordrecht, The
Netherlands.
Benedict, A. A., Alvarez, A. M., Berestecky, J., Imanaka, W., Mizumoto, C. Y.,
Pollard, L. W., Mew, T. W., and Gonzalez, C. F. 1989. Pathovar-specific monoclonal
antibodies for Xanthomonas campestris pv. oryzae and for Xanthomonas campestris
pv. oryzicola. Phytopathology 79: 322-328.
Benedict, A. A., Alvarez, A. M., Civerolo, E. L., W., and Mizumoto, C. Y. 1985.
Delineation of Xanthomonas campestris pv. citri strains with monoclonal antibodies.
(Abstr.) Phytopathology 75:1352.
Benedict, A. A., Alvarez, A. M., and Pollard, L. W. 1990. Pathovar-specific antigens
of Xanthomonas campestris pv. begoniae and Xanthomonas campestris pv. pelargonii.
Appl. Environ. Microbiol. 56: 572-574.
Bereswill, S., Pahl, A., Bellemann, P., Zeller, W., and Geider, K. 1992. Sensitive and
species-specific detection of Erwinia amylovora by polymerase chain reaction analysis.
Appl. Environ. Microbiol. 58: 3522-3526.
Berthier, Y., Verdier, V., Guesdon, J.-L., Chevrier, D., Denis, J.-B., Decoux, G., and
Lemattre, M. 1993. Characterization of Xanthomonas campestris pathovars by rRNA
gene restriction patterns. Appl. Environ. Microbiol. 59:851-859.


data temp; /* Compare fragment data between two */
set transout (keep=col&i col&j); /* strains and create a new data set */
columnl=col&i;
column2=col&j;
if columnl=l & column2=l then prod=l;
else prod=0;
keep column 1 column2 prod;
proc means noprint data=temp;
var prod column 1 column2;
output out=outset sum=prodsum coll sum col2sum;
data outset; /* Determine the similarity value */
set outset; /* between two strains */
coef=(2*prodsum) / (collsum+col2sum);
bactl=&i;
bact2=&j;
keep coefbactl bact2;
data coefmtx;
set coefmtx outset;
%end;
%end;
%mend calccoef;
%calccoef
data coefmtx;
set coefmtx;
if bactl=. then delete;
proc sort data=coefmtx;
by bact2;
data coefmtx;
merge coefmtx (in=i) dna type (keep=bact2 bacteria);
by bact2;
if i= 1;
bacter2=bacteria;
drop bacteria;


133
buffer at 8 V/cm. Phage X Pst I restricted DNA fragments were used as molecular
weight standards. After being stained with 0.5 jag of ethidium bromide per ml for 40
min, the gels were destained in 1 mM MgSC>4 for 1 hr and then photographed over a
UV transilluminator with type 55 Polaroid film.
Data analysis
DNA restriction fragment patterns were determined by direct comparison of the
electrophoretic patterns of the DNA restricted with each of the four endonucleases.
The codes 1 or 0 were assigned according to the presence or absence of each fragment,
respectively. The resulting matrix was used to estimate the genetic relationships
between strains based on the proportion of shared DNA fragments. The expected
proportion of shared fragments (F) was calculated by the equation proposed by Nei and
Li (1979), F = 2nxyJ(nx + ny), where nxy is the number of fragments shared between two
strains and nx and ny are the total number of fragments for each strain. A computer
program (Appendix B) written for the SAS system (SAS Institute Inc., Cary, NC) was
used to estimate the proportion of shared fragments (F). The genetic divergence
between strains was calculated as the estimate of the number of nucleotide substitutions
per site (8), based on the proportion of shared DNA fragments (Nei and Li, 1979). The
iterative method proposed by Nei (1987) was used to estimate the number of nucleotide
substitutions per site using a computer program written for the SAS system (Appendix
C).
Relationships among strains were studied based on phylogenetic analysis using
the BOOT and KITSCH programs from the PHYLIP computer package (Felsenstein,
1991). For the BOOT program, the restriction fragment data encoded 0 or 1 were used
as input for reconstruction of an unrooted phylogenetic tree by using the Wagner


124
provide some important information on the origin of the strains of X. campestris pv.
citrumelo.
Stronger evidence that supports the contention of lateral movement of the hrp
genes among the plant pathogenic xanthomonads comes from the variability in the
relatedness between strains for the two distinct hrp regions examined in this study. For
instance, the hrpC/D region of X. campestris pv. vignicola is phylogenetically
monophyletic and closely related to the homologous hrp region ofX. campestris pv.
bilvae, X campestris pv. citri, X campestris pv. glycines A, X. campestris pv.
malvacearum, X campestris pv. phaseoli B, X campestris pv. phaseoli "fuscans", and
X. campestris pv. vitians B (Fig. 5-2). The hrpB region of campestris pv. vignicola
is also monophyletic, but it is closely related to the homologous hrp region ofX.
campestris pv. begoniae, X. campestris pv. dieffenbachiae, X campestris pv.
maculifoliigardeniae, X. campestris pv. manihotis, and X. campestris pv. phaseoli A
(Fig. 5-1 and 5-2). In a similar way, the hrpC/D-relaied region ofX. campestris pv.
raphani was only 0.50 similar to the homologous region \nX. campestris pv.
campestris, X. campestris pv. armoraciae, and X. campestris pv. incanae (Table 5-2),
whereas the /ir/?5-related region was more than 0.90 similar for the strains of these two
groups (Table 5-2). Moreover, functional heterologous complementation of the hrp
genes has been demonstrated for different xanthomonads (Chapter 3; Arlat et al., 1991;
Bonas et al., 1991). Since there is no genetic and functional support regarding
selection pressure in relation to the hrp genes among the plant pathogenic
xanthomonads, the most likely hypothesis to explain the source of variability for these
two regions of the bacterial genome of X. campestris pv. vignicola remains in the
origin of the two hrp regions from distinct ancestors. The variability in the genetic
relatedness for these two regions of the hrp genes of X. campestris pv. vignicola


46
pathogens X. maltophilia and the opportunistic strains of X. campestris T-55 and IN A,
did not hybridize to any of the three hrp fragments amplified from X. campestris pv.
vesicatoria 75-3 (Fig. 3-4; Table 3-1).
DNA from strains of the plant pathogens of the genera Acidovorax,
Agrobacterium, Clavibacter, Erwinia, Pseudomonas, and Xylella failed to hybridize to
the three hrp fragments (Table 3-1). Furthermore, total genomic DNA of E. herbicola,
a bacterium commonly associated with plant tissue, also did not hybridize to any of the
hrp fragments amplified from X. campestris pv. vesicatoria 75-3 under the conditions
used (Table 3-3).
Amplification of the hrp fragments from other X. campestris pathovars and related
Xanthomonas spp.
Primer pairs RST2 plus RST3 and RST21 plus RST22 were used to amplify
DNA sequences from strains representing X. fragariae and 28 different pathovars ofX.
campestris. In all cases with both sets of primers, fragments of identical sizes were
amplified from different pathovars of X. campestris and related Xanthomonas spp. (Fig.
3-5; Table 3-1). However, amplification with the primers RST2 and RST3 usually
gave low yield of DNA for strains of A. campestris pv. carotae, X. campestris pv.
gardneri, X. campestris pv. papavericola, X. campestris pv. pelargonii, and X.
campestris pv. taraxaci (Fig. 3-5; Table 3-1). Although DNA isolated from X.
campestris pv. holcicola hybridized to the 840 bp hrp fragment from X. campestris pv.
vesicatoria strain 75-3 (Fig. 3-4), primers RST2 and RST3 did not amplify the DNA
fragment from this pathovar (Table 3-1). No amplification occurred with purified total
genomic DNA from a number of bacteria, including plant pathogenic strains of the
xanthomonads X albilineans, X. campestris pv. secalis, and X. campestris pv.
translucens, and of the genera Acidovorax, Agrobacterium, Clavibacter, Erwinia,


67
fragments amplified from the different plant pathogenic xanthomonads after digestion
with frequent-cutting endonucleases Cfol, HaeIII, SaulAl, and Taql. The combination
of Arp-related sequences and restriction endonucleases were useful for the
establishment of distinct restriction banding patterns for the different groups of plant
pathogenic xanthomonads included in this study.
The strains of plant pathogenic xanthomonads were distributed into fifty
different groups on the basis of the banding patterns of two Arp-related fragments
digested with the restriction endonucleases Cfol, Haelll, Sau3Al, and Taql (Table 4-2).
The restriction banding patterns were determined by visual comparisons of the DNA
bands separated by electrophoresis of agarose gels (Fig. 4-1 to 4-8). The 1,075-bp Arp-
related fragment amplified from different strains of xanthomonads and restricted with
the endonucleases Cfol and Haelll generated the largest number of restriction banding
patterns, 23 and 25, respectively (Fig. 4-5 and 4-6). The restriction analysis of the Arp-
related fragment with these two endonucleases was more discriminatory and was very
useful for differentiation of closely related pathovars, such as A campestris pv.
armoraciae and A! campestris pv. campestris (Table 4-2), or even to distinguish groups
of strains within pathovars, such as the case of strains of X. campestris pv.
dieffenbachiae (Fig. 4-11; Table 4-2) and A. campestris pv. citri (Chapter 6; Table 4-2).
On the other hand, the restriction analysis with less discriminatory endonucleases, such
as Taql, was very helpful to establish the genetic relatedness at pathovar level (Table 4-
2). For example, the cruciferous pathogens A campestris pv. armoraciae, A
campestris pv. campestris, A campestris pv. incanae, and A campestris pv. raphani all
belong to the Taql profile 1 for the 840-bp Arp-related fragment and to the Taql profile
3 for the 1,075-bp Arp-related fragment (Table 4-2).


160
communication) were used in this study. Primers RST2 plus RST3 delineated an 840-
bp fragment, RST9 plus RST10 delineated a 355-bp fragment, and RST21 plus RST22
delineated a 1,075-bp fragment. The primers map to the complementation groups hrpB
and hrpC/D of the hrp gene cluster of X. campestris pv. vesicatoria (Chapter 3).
Oligonucleotide primers were synthesized with a model 394 DNA Synthesizer
(Applied Biosystems, Foster City, CA) by the DNA Synthesis Laboratory, University
of Florida, Gainesville.
DNA fragments were amplified in a reaction mixture of 50 pi containing 5 pi of
10X buffer (500 mM KC1, 100 mM Tris-Cl [pH 9.0 at 25C], 1% Triton X-100), 1.5
mM MgC^, 200 pM of each deoxynucleotide triphosphate (Boehringer Mannheim), 25
pmol of each primer, 1.25 units of Taq polymerase (Promega, Madison, WI). The
amount of template DNA added was 100 ng of purified bacterial DNA whereas for the
seed and seedling samples a volume of 10 pi of the DNA extract was added, unless
otherwise stated. The reaction mixture was overlaid with 50 pi of light mineral oil.
Thirty amplification cycles were performed in an automated thermocycler PT-100-60
(MJ Research, Watertown, MA) according to the following profiles: 30 s of
denaturation at 95C, 30 s of annealing at 62C, and 45 s of extension at 72C for the
primers RST2 plus RST3; 30 s of denaturation at 95C, 30 s of annealing at 52C, and
45 s of extension at 72C for the primers RST9 plus RST10; 30 s of denaturation at 95
C, 40 s of annealing at 61C, and 45 s of extension at 72C for the primers RST21 plus
RST22. For all three profiles, the initial denaturation step was 5 min at 95C, and the
last extension step was extended to 5 min. Amplified DNA fragments were detected by
electrophoresis in 0.9% agarose gels in TAE buffer (40 mM Trisacetate, 1 mM EDTA,
pH 8.2) according to standard procedures (Sambrook et al 1989).


This dissertation was submitted to the Graduate Faculty of the College of
Agriculture and to the Graduate School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy. /I
August 1994
Dean, College of Agriculture
Dean, Graduate School


184
seeds (Gitaitis et al., 1987). Strains of this pathovar have a large host range and may
cause disease on pepper and tomato plants (Gitaitis et al., 1987; White, 1930).
Furthermore, outbreaks of a bacterial spot on tomato transplants in southern USA have
been associated to strains that have genetic and phenotypic characteristics similar to X.
campestris pv. raphani (R. E. Stall, personal communication). Another common
intriguing point is the similarity of fifteen strains isolated from pepper seeds to X.
campestris pv. carotae. Both restriction analysis of the DNA sequences related to the hrp
genes and fatty acid composition agreed in the identification of these strains as X.
campestris pv. carotae (Table 7-1). I am unaware of any previous report of the
association of strains similar to X campestris pv. carotae with pepper seeds.
The presence of a diverse xanthomonad community on pepper and tomato seeds
certainly poses a major concern for health inspection in seed certification programs.
Whereas some of the xanthomonads are probably nonpathogenic to plants, others are
pathogenic to pepper and tomato. The latter were identified as X. campestris pv.
vesicatoria based on pathogenicity tests, analysis of DNA sequences related to the hrp
genes, and fatty acid composition. However, a third group of xanthomonads that includes
potential plant pathogenic bacteria was also identified. These strains, tentatively
identified as belonging to different pathovars of X. campestris, cause hypersensitive
reaction or no reactions on pepper and tomato plants under artificial inoculation.
However, the role that these strains may play in a bacterial disease syndrome on pepper
and tomato plants remains to be clarified. Furthermore, the association of such a diverse
group of bacterial pathogens with pepper and tomato seeds is a major problem with the
procedures currently available for their detection and identification. These strains grow
on semiselective media developed for the pepper and tomato pathogen X. campestris pv.
vesicatoria (Gitaitis et al., 1987, 1991; McGuire et al., 1986) and they may cause certain


182
is followed by lysis of the cells and purification of the DNA recovered. This cell
extraction method may be less efficient than the direct lysis method that involves the
release and extraction of DNA without prior separation of the cells from the original
matrix (Ogram et al., 1988; Steffan et al., 1988). However, the cell extraction method is
likely to ensure exclusion of extracellular DNA that may be present in the sample and
may produce DNA of higher quality (Holben et al., 1988; Steffan et al., 1988). In fact,
inconsistent results in the detection of X. campestris pv. vesicatoria were obtained with
the quick approach of DNA isolation which employs direct lysis of the bacterial cells in
the original extraction matrix.
The procedure of amplification of DNA sequences of the hrp genes was highly
sensitive and specific for detection of plant pathogenic xanthomonads in DNA
preparations from seeds washes. Based on dilution plating counts, X. campestris pv.
vesicatoria could be detected in preparations containing less than 102 CFU added to seed
extracts (Table 7-2). However, this level of sensitivity may be improved further by
including a step for concentration of the cells in the sample to enhance the recovery of
bacterial cells (Roth, 1989; Schaad, 1982). The concentration step concomitantly
concentrates unwanted saprophyte microflora. Although the presence of a large
background microflora is a major limitation in the detection of plant pathogenic
xanthomonads in pepper and tomato seeds by planting on general or even semiselective
media (Gitaitis et al., 1991; McGuire et al., 1986), the specificity of the DNA
amplification procedure is likely to overcome this kind of problem. In fact, X. campestris
pv. vesicatoria was detected in naturally contaminated seed lots containing background
bacterial microflora larger than 107 CFU/g of seeds, whereas plating on semiselective
medium failed to recover the xanthomonads (Table 7-2). Nevertheless, the transmission
tests confirmed the presence of viable cells of plant pathogenic xanthomonads in 5 seed


170
75-3 were added to the PCR reaction mixture for amplification of the hrpB fragment
with primers RST2 and RST3. The 840-bp hrpB fragment was amplified from all
different samples, and no significant inhibition of the amplification reaction was
observed (Fig. 7-3).
Sensitivity of the detection procedure
The sensitivity of the DNA amplification procedure for detection of cells of2f
campestris pv. vesicatoria in preparations of tomato seed extracts was about 102
CFU/ml when the primers RST9 and RST10 were used for amplification of the 355-bp
hrpB fragment (Fig. 7-4; Table 7-2). The minimum detection level ofX. campestris
pv. vesicatoria was usually 10 to 100 times higher when the primers RST2 and RST3
were used for amplification of the 840-bp hrpB fragment (Fig. 7-4; Table 7-2). The
lower sensitivity of the primers RST2 and RST3 for detection of X. campestris pv.
vesicatoria was also observed for other tomato and pepper seed lots (data not shown).
Despite the differences in sensitivity, both sets of primers produced specific
amplification of the hrp fragments from seed extracts containing different
concentration of A", campestris pv. vesicatoria (Fig. 7-4). In comparison to ELISA, the
DNA amplification procedure was at least 100 to 1000 times more sensitive for
detection ofX. campestris pv. vesicatoria when cells were added to the seed
preparation (Table 7-2). The detection level ofX. campestris pv. vesicatoria by ELISA
was greater than 105 CFU/ml (Fig. 7-4; Table 7-2).
Detection of plant pathogenic xanthomonads in naturally contaminated seed lots
Investigations were conducted to determine the feasibility of the DNA
amplification procedure for detection of the presence of plant pathogenic


125
substantiate the contention that the hrp genes in some of the plant pathogenic
xanthomonads may have evolved from distinct ancestors through lateral movement of
genetic material.
An important conclusion from the phylogenetic studies is the divergence
between hrp genes and host specificity. The hrp genes are essential for the
development of disease on compatible hosts and hypersensitive reaction on both
resistant host and nonhost plants (Willis et al., 1991). Previous work has demonstrated
the functional conservation of the hrp genes and the lack of host speciation (Chapter 3;
Arlat et al. 1991; Bonas et al., 1991). Although the hrp genes are necessary in the
plant-pathogen interaction, other factors in the bacterial pathogen are likely to be
involved in host speciation (Fenselau et al., 1992; Gough et al., 1992). Our results
support the coevolution of the hrp genes with the rest of the bacterial genome through a
common bacterial ancestor. Also, there are bases to support the hypothesis of
horizontal movement of the hrp genes. Close relationship between plant pathogenic
xanthomonads with different genetic background and variability in the relatedness of
different regions of the hrp genes indicate distinct origins for different regions of the
hrp genes, instead of evolution from a single ancestor. Since the lateral movement of
genetic material between bacteria is a common and important mechanism in bacterial
evolution (Krawiec and Riley, 1990), the coexistence in the same biological niche may
have provided opportunities for the lateral transfer of the hrp region of the bacterial
genome between xanthomonads. The hrp are functionally conserved among the
xanthomonads (Chapter 3; Bonas et al., 1991; Fenselau et al., 1992), this also supports
the absence of functional selectivity. The resemblance at the protein level of the hrp
genes of the xanthomonads with genes involved in the secretion of pathogenicity
factors in genetically distant organisms such as the animal pathogens of the genus


150
campestris pv. vesicatoria (Egel, 1991) and on the basis of amplification of hrp
fragments with oligonucleotide primers specific for the complementation groups B and
C/D of the hrp genes of X. campestris pv. vesicatoria (Chapter 3). This is particularly
significant in regard to the strains associated with citrus bacterial spot disease. Despite
differences in the pathogenic characteristics of those strains (Graham and Gottwald,
1990; Graham et al., 1990), the presence of an hrp gene cluster supports the pathogenic
nature of those bacterial strains. If they were opportunistic xanthomonads (Gitaitis et
al., 1987), they would lack an hrp region (Bonas et al., 1991; Stall and Minsavage,
1990).
Information about the similarity of the hrp genes of the bacteria causing
diseases of citrus was obtained from restriction enzyme patterns of amplified fragments
of the hrp gene cluster. Although the DNA fragments amplified with the two sets of
primers were of the same size for all the strains of X. campestris, characteristic
restriction banding patterns for each bacterial group occurred with the 840- and 1,075-
bp fragments. Complementation groups B and C/D of the hrp gene cluster, from which
the fragments were amplified, are considered to be highly conserved among plant
pathogenic xanthomonads (U. Bonas, personal communication). Therefore, the
homology of the restriction enzyme fragments from amplified hrp genes should furnish
valid relationships among these pathogens. These relationships were determined by a
phylogenetic analysis.
The phylogenetic analysis based on the hrp gene cluster showed polyphyletic
relationships of the strains of X campestris causing disease in citrus. This suggests
that the hrp gene cluster may have evolved independently in these strains of X.
campestris. This evolution could be convergent or parallel. The analysis is based on
the assumption that the differences in restriction sites in the hrp gene cluster region


203
proc sort data=coefmtx; /* Sort the result by strain name */
by bactl;
data coefmtx;
merge coefmtx (in=i) dna type (keep=bactl bacteria);
by bactl;
if i=l;
bacterl=bacteria;
drop bacteria;
label bacterl-'Strain" bacter2-'Strain"
coef=" strain";
proc tabulate f=9.4 formchar=' '; /* Print out the result */
title "Similarity values for bacteria";
class bacterl bacter2;
var coef;
table bacterl,bacter2*coef;
key label sum-' ";
run;


242
Thaveechai, N., and Schaad, N. W. 1984. Comparison of different immunogen
preparations for serological identification of Xanthomonas campestris pv. campestris.
Phytopathology 74: 1065-1070.
Thayer, P. L., and Stall, R. E. 1961. Effect of variation in the bacterial spot pathogen
of pepper and tomato on control with streptomycin. Phytopathology 51: 568-571.
Van den Mooter, M., Maraite, H., Meiresonne, L., Gillis, M., Kersters, K., and DeLey,
J. 1987a. Comparison between Xanthomonas campestris pv. manihotis (ISPP List
1980) and X campestris pv. cassavae (ISPP List 1980) by means of phenotypic,
protein electrophoretic, DNA hybridization and phytopathological techniques. J. Gen.
Microbiol. 133:57-71.
Van den Mooter, M., Steenackers, M., Maertens, C., Gossele, F., DeVos, P., Swings, J.,
Kersters, K., and DeLey, J. 1987b. Differentiation between Xanthomonas campestris
pv. graminis (ISPP List 1980), pphleipratensis (ISPP List 1980) emend., p\.poae
Egli and Schmidt 1982 and pv. arrhenatheri Egli and Schmidt 1982, by numerical
analysis of phenotypic features and protein gel electrophoregrams. J. Phytopathol.
118: 135-156.
Van den Mooter, M., and Swings, J. 1990. Numerical analysis of 295 phenotypic
features of 266 Xanthomonas strains and related strains and an improved taxonomy of
the genus. Int. J. Syst. Bacteriol. 40: 348-369.
Vauterin, L., Hoste, B., Yang, P., Alvarez, A., Kersters, K., and Swings, J. 1993.
Taxonomy of the genus Xanthomonas. Pages 157-192 in: Xanthomonas. J.G. Swings
and E.L. Civerolo, eds. Chapman & Hall, London, United Kingdom.
Vauterin, L., Swings, J., and Kersters, K. 1991a. Grouping of Xanthomonas
campestris pathovars by SDS-PAGE of proteins. J. Gen. Microbiol. 137: 1677-1687.
Vauterin, L., Swings, J., Kersters, K., Gillis, M., Mew, T. W., Schroth, M. N.,
Palleroni, N. J., Hildebrand, D. C., Stead, D. E., Civerolo, E. L., Hayward, A. C.,
Maraite, H., Stall, R. E., Vidaver, A. K., and Bradbury, J. F. 1990a. Towards an
improved taxonomy of Xanthomonas. Int. J. Syst. Bacteriol. 40: 312-316.
Vauterin, L., Vantomme, R., Pot, B., Hoste, B., Swings, J., and Kersters, K. 1990b.
Taxonomic analysis of Xanthomonas campestris pv. begoniae and X. campestris pv.
pelargonii by means of phytopathological, phenotypic, protein electrophoretic and
DNA hybridization methods. Syst. Appl. Microbiol. 13:166-176.


13
bacterial cells varied considerably and indicated the occurrence of taxon-specific
protein profiles that could be used for differentiation and identification of plant
pathogenic xanthomonads. The presence of distinct and specific patterns of esterases at
the pathovar level was particularly valuable in differentiating and identifying the
different groups of plant pathogenic Xanthomonas spp. (El-Sharkawy and Huisingh,
1971a). However, a more reliable and reproducible techniques were required to resolve
the protein profile at subspecific level. In this respect, sodium dodecyl sulphate-
polyacrylamide gel electrophoresis (SDS-PAGE) of total cellular proteins has become a
key technique for the separation and characterization of a wide variety of proteins
(Laemmli, 1970; Weber and Osborn, 1960). Furthermore, this technique has been
useful for differentiating the plant pathogenic xanthomonads on the basis of whole-cell
protein (Bouzar et al., 1994; Kersters and De Ley, 1975; Van den Mooter et al.,
1987ab; Vauterin et al., 1990a, 199lab, 1992). Variation in the electrophoretic patterns
of specific proteins of the bacterial cell, such as total cell envelope and total membrane
proteins, has also been investigated by using SDS-PAGE for differentiation of some
pathovars ofX. campestris (Dos Santos and Dianese, 1985; Minsavage and Schaad,
1983; Qhobela and Claflin, 1992; Qhobela et al., 1991).
Vauterin et al. (1991a) analyzed whole-cell proteins of an extensive collection
of plant pathogenic bacteria of the genus Xanthomonas, comprising all species and 27
pathovars of X campestris, by using SDS-PAGE analysis. On the basis of SDS-PAGE
protein profiles, the plant pathogenic species X. albilineans, X. axonopodis, X.
fragariae, and X. oryzae could be distinguished from each other and also from X.
campestris pathovars, whereas X populi was rather similar to X. campestris. Several
distinct protein profiles were observed within X. campestris, and some of the groups
established corresponded to described pathovars of this species. Certain pathovars


16
analyses and phenotypic characteristics (Stall et al., 1994; Vauterin et al., 1990a). Two
major distinct groups have been delineated by analysis of whole-cell protein profiles
(Bouzar et al., 1994; Vauterin et al., 1991a), utilization of carbon sources (Bouzar et
al., 1992), fatty acid composition (Stall et al., 1994; Vauterin et al., 1990b), and DNA-
DNA hybridization (Stall et al., 1994). Furthermore, these two groups can be easily
distinguished by the presence of unique protein bands on the basis of silver staining of
SDS-lysed cell proteins (Bouzar et al., 1994).
A comparative study was carried out of the Xanthomonas species and X.
campestris pathovars that cause diseases on cereals and grasses by using SDS-PAGE of
whole-cell proteins (Vauterin et al., 1992). The pathovars of the "translucens" group,
which includes X. campestris pv. cerealis, X campestris pv. hordei, X. campestris pv.
secalis, X. campestris pv. translucens, and X. campestris pv. undulosa, could not be
distinguished from one another. The pathogens of grasses which formed another
homogeneous group were X. campestris pv. arrhenatheri, X. campestris pv. phlei, and
X. campestris pv. poae, and had protein profiles very similar to those of the
"translucens" group. DNA-DNA hybridization data, phenotypic features, and fatty acid
composition also support the close relationship between the two groups (Van den
Mooter et al., 1987b; Vauterin et al., 1992). On the other hand, X. axonopodis and X.
campestris pv. vasculorum were heterogeneous, with two distinct groups being found
in each taxon. The type A group of X. axonopodis was very closely related to one
group ofX. campestris pv. vasculorum, whereas the type B group of A! axonopodis had
atypical protein and fatty acid profiles which do not resemble xanthomonads. X
campestris pv. vasculorum type B group was indistinguishable from X. campestris pv.
holcicola based on protein profile, as well as on fatty acid composition and DNA-DNA
hybridization (Vauterin et al., 1992).


CHAPTER 2
REVIEW OF LITERATURE
The genus Xanthomonas includes six species containing plant pathogenic
bacteria (Bradbury, 1984, 1986) that occur worldwide and cause economically
important diseases on several plant crops. The species X. campestris is the largest and
consists of at least 125 different pathovars (Bradbury, 1984, 1986). The basis for
differentiation and identification of these plant pathogenic bacteria at the pathovar level
is by means of the capability of the bacterial strain to cause characteristic disease or by
their host range (Dye et al., 1980; Vauterin et al., 1990a; Young et al., 1992). The
identification of the xanthomonads at the genus and species level does not present a
significant problem, and can be achieved by using different techniques, such as
biochemical and physiological tests, serology, protein profiling, and fatty acid and
nucleic acid analysis (Bradbury, 1984; Holt et al., 1994; Schaad, 1988; Vauterin et al.,
1990ab). Despite a range of different techniques available, the differentiation at
infrasubspecific level is still difficult, and it is not uncommon that strains belonging to
different pathovars may be genetically and phenotypically closely related.
Serology
Several attempts have been made to distinguish between pathovars of X.
campestris based on serological reactions. Early work on the use of serology for
differentiation of plant pathogenic xanthomonads by using agglutination and
precipitation tests suggested that this technique would be specific for differentiation
5


191
xanthomonads. The hypothesis of coevolution of the hrp region with the rest of the
genome from a common ancestor is supported in several cases. For instance, the
similarity in the hrpB and hrpC/D regions examined in this study for the groups A and
B of X. campestris pv. vesicatoria was less than 0.51. This genetic divergence of the
hrp genes is in agreement with the results obtained when the entire genome of these
two groups of X. campestris pv. vesicatoria were compared on the basis of DNA
homology (Stall et al., 1994). Although these pathogens cause similar diseases on
solanaceous plants, their hrp genes are genetically diverse to the same extent as the rest
of the genome. Furthermore, these comparisons do not support the contention that
there was a higher selective pressure for the hrp sequences than for other regions of the
genome nor that a horizontal movement of the hrp region of the genome may have
occurred between strains of the two groups of X. campestris pv. vesicatoria. Further
evidence to support the coevolution of the hrp genes and the rest of the genome
involves the pathovars X. campestris pv. carotae, X. campestris pv. gardneri, and X
campestris pv. pelargonii. Although these xanthomonads have very distinct host
ranges, they are genetically closely related based on DNA-DNA hybridization
(Hildebrand et al., 1989; Palleroni et al., 1993). In relation to the hrp genes they are
not only monophyletic but also very closely related. This example points to a
coevolution of the hrp genes and the rest of the genome from a common bacterial
ancestor for certain plant pathogenic xanthomonads and also substantiates the
divergence between hrp genes and host speciation. An important conclusion from the
phylogenetic studies is the divergence between hrp genes and host specificity. The hrp
genes are essential for the development of disease on compatible hosts and
hypersensitive reaction on both resistant host and nonhost plants (Willis et al., 1991).
Previous work has demonstrated the functional conservation of the hrp genes and the


49
Restriction endonuclease analysis of amplified hrp-related DNA fragments
To address the question of degree of sequence conservation among different
strains, I examined the 840- and 1,075-bp hrp fragments amplified from strains of
different pathovars of X. campestris, as well as from X. fragariae, by restriction
endonuclease analysis with the endonucleases Cfol, HaeIII, Saw3AI, and Taql.
Restriction fragment length polymorphisms were apparent for both fragments. For
example, the 1,075-bp fragment, amplified from strains of different pathovars ofX.
campestris by using primers RST21 and RST22 and then restricted with Hae III and
Saw3AI, yielded different restriction patterns (Fig. 3-6). Although RFLPs were
observed with all four endonucleases for both the 840- and the 1,075-bp fragments,
restriction analysis with Cfo\ and HaelII produced more distinct patterns for
differentiation of the groups or pathovars of X. campestris.
Pathogenicity function of the DNA sequence of X. campestris pathovars from which
hrp related fragments were amplified
To determine if the DNA sequence from which hrp related fragments were
amplified has common functions in different strains of X campestris, I investigated the
functional homology of these fragments in X. campestris pv. vesicatoria group B and X
campestris pv. pelargonii. The cosmid clones pXV56/3-48, of a pLAFR3 library of
strain XV56 of A! campestris pv. vesicatoria group B (Beaulieu et al., 1991), and
pXCP58/2, of a pLAFR3 library of A. campestris pv. pelargonii XCP58 (Gerald V.
Minsavage, personal communication), were identified by amplification of the 1,075-bp
hrp fragment with the primers RST21 and RST22 Plasmid pXV56/3-48 was
transferred into the A! campestris pv. vesicatoria mutants &5-\0whrpA22, 85-
\0::hrpB85, S5-\0::hrpC44, 85A0::hrpD137, 85-10::hrpE75, and %5-\0::hrpF318,
which carry Tn3-gws insertions in the different hrp complementation groups (Bonas et


121
highly related and they are monophyletic in the phylogenetic analysis with bootstrap
resampling value of 80% (Fig. 5-2). Further, the similarity index between members of
the clades 3 and 4 ranged from 0.96 to 1.00 for the hrpB region (Table 5-2).
Discussion
The phylogenetic analysis of the hrp genes of the xanthomonads has revealed a
diverse evolutionary relationship for this region of the bacterial genome of these plant
pathogens. The hypothesis of coevolution of the hrp region with the rest of the genome
from a common bacterial ancestor is supported by comparison of either closely or
distantly related plant pathogenic xanthomonads. For instance, the similarity of the
two hrp regions examined in this study for the groups A and B ofX. campestris pv.
vesicatoria was less than 0.51. This genetic divergence determined for these two
groups of X. campestris pv. vesicatoria is very similar to the values obtained when the
entire genome was compared on the basis of DNA homology (Stall et al., 1994).
Although these pathogens cause similar diseases on solanaceous plants, their hrp genes
are genetically diverse at about the same extend as the rest of the genome. Moreover,
these comparisons do not support the contention that there was a higher selective
pressure for the hrp sequences than for other regions of the genome nor that a
horizontal movement of the hrp region of the genome may have occurred between
strains of the two groups of X. campestris pv. vesicatoria. Despite the genetic distance,
the hrp genes of these two groups of X. campestris pv. vesicatoria are functionally
complementary (Chapter 3). Similarly, the citrus pathogens X. campestris pv. citri and
X campestris pv. citrumelo also cause diseases on the same hosts though they are
genetically distinct (Egel et al., 1991; Gabriel et al., 1989; Vauterin et al., 1991a). The


159
Aliquots of the seed or seedling washings were transferred to 1.5 ml microfuge tubes,
and sodium ascorbate (Sigma, St. Louis, MO) and insoluble polyvinypolypyrrolidone
(PVPP) (Sigma) were added for a final concentration of 0.2 M and 0.1% respectively.
These concentrations of PVPP and sodium ascorbate were used throughout this study,
unless otherwise stated. Prior to use, the PVPP was acid washed by the procedure
described by Holben et al. (1988). The samples were homogenized by vortexing. The
homogenate was pelleted by centrifuging in an Eppendorf microcentrifuge (Brinkmann
Instruments Inc., Westbury, NY) for 2 min at 16,000 g. The pellet was resuspended in
567 pi of TE buffer (10 mM Tris-Cl, pH 8.0; 1 mM EDTA, pH 8.0). Proteinase K
(Boehringer Mannheim, Indianapolis, IN) and sodium dodecyl sulfate (SDS) (Sigma)
were added for a final concentration of 100 pg/ml and 0.5%, respectively. After
incubation for 1 hour at 37C, sodium chloride and hexadecyltrimethyl ammonium
bromide (Sigma) were added to each preparation for a final concentration of 0.7 M and
1%, respectively. The preparations were incubated for 10 min at 65C. DNA was
purified by treatment with chloroform-isoamyl alcohol (24:1). The samples were
vortexed and centrifuged for 5 min at 16,000 g. A second purification was
accomplished by adding phenol-chloroform-isoamyl alcohol (25:24:1) and centrifuging
as described above. DNA was precipitated by adding 0.6 volumes of isopropanol and
incubating for 30 min at -20C. The samples were centrifuged for 20 min at 16,000 g.
The DNA pellet obtained was washed with 70% ethanol and centrifuged again. After
drying, the pellet was redissolved in 50 pi of TE buffer and stored at 4C.
DNA amplification
Three sets of oligonucleotide primers selected from the nucleotide sequence of
the hrp gene cluster of Xanthomonas campes tris pv. vesicatoria (Ulla Bonas, personal


71
Fig. 4-4. Restriction profiles established for the 840-bp hrp-related fragment amplified
from different strains of plant pathogenic Xanthomonas spp. and restricted with the
endonuclease Taql. Lane M, phage X restricted with Pstl. Molecular sizes are given in
bases. See Table 4-2 for identification of strains of Xanthomonas spp. for each
restriction pattern.


228
Bochner, B. R. 1989. Sleuthing out bacterial identities. Nature 339:157-158.
v Bonas, U., Schulte, R., Fenselau, S., Minsavage, G. V., Staskawicz, B. J., and Stall, R.
E. 1991. Isolation of a gene cluster from Xanthomonas campestris pv. vesicatoria that
determines pathogenicity and the hypersensitive response on pepper and tomato. Mol.
Plant-Microbe Interact. 4:81-88.
Boucher, C. A., Van Gijsegem, F., Barberis, P. A., Arlat, M., and Zischek, C. 1987.
Pseudomonas solanacearum genes controlling both pathogenicity and hypersensitivity
on tobacco are clustered. J. Bacteriol. 169: 5626-5632.
Bouzar, H., Jones, J. B., and Stall, R. E. 1994. Silver staining reveals proteins unique
to phenotypically distinct groups of Xanthomonas campestris pv. vesicatoria.
Phytopathology 84: 39-44.
Bouzar, H., Jones, J. B., Stall, R. E., and Scott, J. W. 1992. Cluster analysis of
Xanthomonas campestris pv. vesicatoria based on carbon source utilization patterns.
Proc. 8th Int. Conf. Plant Pathog. Bacteria. INRA, Versailles, France, (in press)
Bradbury, J. F. 1984. Genus II.Xanthomonas Dowson 1939, 187. Pages 199-210 in:
Bergey's Manual of Determinative Bacteriology, 9th ed. J. C. Holt and N. R. Krieg,
eds. Williams & Wilkins Co., Baltimore, MD.
Bradbury, J. F. 1986. Guide to Plant Pathogenic Bacteria. CAB International
Mycological Institute, Slough, Great Britain.
Brinkerhoff, L. A. 1970. Variation in Xanthomonas malvacearum and its relation to
control. Annu. Rev. Phytopathol. 8:85-110.
Burkholder, W. H., and Starr, M. P. 1948. The generic and specific characters of
phytopathogenic species of Pseudomonas and Xanthomonas. Phytopathology 38: 494-
502.
Canteros de Echenique, B. I., Zagory, D., and Stall, R. E. 1985. A medium for
cultivation of the B strains of Xanthomonas campestris pv. citri, cause of cancrosis B
in Argentina and Uruguay. Plant Dis. 69:122-123.
Charudattan, R., Stall, R. E., and Batchelor, D. L. 1973. Serotypes of Xanthomonas
vesicatoria unrelated to its pathotypes. Phytopathology 63: 1260-1265.


Table D-2Continued
Species/Pathovar
Species/Pathovar
24
25
26
27
28
29
31
34
35
36
23. pv. dieffenbachiae A
1.000
.6047
.7273
.6512
.6977
.6154
.5714
.5714
.9231
.5854
24. pv. dieffenbachiae A
.6047
.7273
.6512
.6977
.6154
.5714
.5714
.9231
.5854
25. pv. dieffenbachiae B
.8511
.7826
.7826
.4286
.5333
.5333
.6190
.5455
26. pv. fici A
.9362
.9362
.5116
.5652
.5652
.6977
.5778
27. pv. fici A
.8696
.4286
.4889
.4889
.6190
.5000
28 pv. fici A
.
.4762
.5333
.5333
.6667
.5455
29. pv. fici B
.
.4390
.6829
.5789
.4500
31. pv. glycines A
.
.5455
.5854
.9302
34. pv. incanae
.
.5854
.5116
35. pv. maculifoliigardeniae
.
.

.6000
36. pv. malvacearum





Continued on following page
tO
to
o


119
Another evolutionary relationship revealed by the phylogenetic analysis of the
hrp genes of the xanthomonads is the variability in the relatedness between strains
depending on the hrp region examined. For example, X. campestris pv. vignicola
clustered in clade 6 in the phylogenetic analysis for the hrpC/D region (Fig. 5-1), and
the value of 78% for the bootstrap resampling supports the monophyletic nature of the
hrp region for this clade (Fig. 5-1). Further, similarity indices for the different
pathovars within this clade ranged from 0.67 to 1.00 (Tables 5-1). In comparison, the
similarity of X. campestris pv. vignicola to the members of clade 7 ranged from 0.42 to
0.69 (Table 5-2 and 5-3). On the other hand, the strains X. campestris pv. vignicola
were placed into clade 7 in the phylogenetic analysis of the hrpB region (Fig. 5-2), and
this was strongly supported by the bootstrap resampling with a value of 84% of the
bootstrapped trees (Fig. 5-2). The genetic relatedness of2f campestris pv. vignicola to
the other members of the clade 7 ranged from 0.76 to 0.94 for the hrpB region whereas
the similarity of this region between X. campestris pv. vignicola strains and the
members of the clade 6 ranged from 0.53 to 0.61 (Table 5-3).
Strains of2f campestris pv. raphani also comprise an unique case. The
phylogenetic analysis of the hrpC/D-rdaied region of X campestris pv. raphani
revealed that the two closely related groups were placed in clade 4 (Fig. 5-1). The
bootstrap resampling analysis also supports 100% the monophyletic nature of this clade
(Fig. 5-1). The genetic similarity of the strains in clade 4 to strains of clade 3 which
contains the other brassica pathogens i.e. X campestris pv. armoraciae, X. campestris
pv. campestris, and X. campestris pv. incanae ranged from 0.41 to 0.50 (Table 5-2). In
contrast, the analysis of the hrpB region showed that the hrp region for the strains of X.
campestris pv. raphani and strains of the other three pathovars of X campestris are


128
strains appear to be quite heterogeneous both genetically (Egel, 1991; Egel et al., 1991)
and in aggressiveness (Graham and Gottwald, 1990) compared with A! campestris pv.
citri. This has resulted in questions about the relationship of the bacterial spot
pathogen to other pathovars of X. campestris. Several xanthomonads isolated from
ornamental plants cause lesions similar to bacterial spot when artificially inoculated
onto young citrus plants (Graham and Gottwald, 1991; Graham et al., 1990); they are
also genetically similar to some strains ofX. campestris pv. citrumelo (Egel et al.,
1991; Graham et al., 1990). It was suggested that strains oiX. campestris pv. citrumelo
may represent other pathovars of X. campestris incidentally isolated from citrus
(Graham et al., 1990) or strains of a xanthomonad that has a wide host range (Gabriel et
al., 1988); alternatively it was suggested that the most weakly aggressive strains may
be opportunistic strains which cause symptoms only when associated with injury (Egel,
1991).
These alternatives have not been resolved by studies of the genetics (Egel et al.,
1991; Gabriel et al., 1988, 1989; Graham et al., 1990; Hartung and Civerolo; 1987,
1989; Vauterin et al., 1991b) or pathogenicity (Graham and Gottwald, 1990; Graham et
al., 1990) of these strains. The genetics of pathogenicity, however, might favor one of
the above hypotheses. An excellent candidate for examination is the hypersensitivity
reaction and pathogenicity (hrp) gene cluster responsible for pathogenicity reaction on
susceptible hosts and a hypersensitive reaction on resistant hosts or on nonhosts plants
(Willis et al., 1991). The hrp gene cluster has been characterized in several bacterial
plant pathogens, such as Pseudomonas syringae pv. phaseolicola (Lindgren et al.,
1986), P. solanacearum (Boucher et al., 1987), Erwinia amylovora (Beer et al., 1991),
and A! campestris pv. vesicatoria (Bonas et al., 1991). The hrp gene cluster of X.
campestris pv. vesicatoria consists of at least 25 kb of genomic DNA, and


Table A-1 Continued
198
Pathovar Strain Source3
Acidovorax avenae
A. avenae subsp. avenae
UK142-A
JBJ
A. avenae subsp. citrulii UK20
JBJ
Agrobacterium
LBA1050
BJS
tumefaciens
Clavibacter michiganense
subsp. michiganense
69-1,75-1
RES
Erwinia
E. carotovora subsp.
K-SR-347, B-SR38
JAB
carotovora
E. herbicola
NF-33
RES
E. stewartii
SW2
DLC
Pseudomonas
P. solanacearum
K60
GM1/1000
AK
P. syringae
pv. syringae
INB
RES
pv. tomato
987
RES
Xylella fastidiosa
89-1
DLH
aAK, A. Kelman, University of Wisconsin, Madison, WI; ARC, A. R. Chase,
University of Florida, Apopka, FL; ATCC, American Type Culture Collection,
Rockville, MD; BJS, B. J. Staskawicz, University of California, Berkely, CA; DCH,
D. C. Hildebrand, University of California, Berkely, CA; DLC, D. L. Coplin, Ohio
State University, Columbus, OH; DLH, D. L. Hopkins, University of Florida,
Leesburg, FL; DLS, D. L. Stuteville, Kansas State University, Manhattan, KS; DPI,
Department of Plant Industry, Gainesville, FL; DWG, D. W. Gabriel, University of
Florida, Gainesville, FL; ELC, E. L. Civerolo, U.S. Department of Agriculture,
Beltsville, MD; JAB, J. A. Bartz, University of Florida, Gainesville, FL; JBJ, J. B.
Jones, University of Florida, Bradenton, FL; JCC, J. C. Comstock, U.S. Department
of Agriculture, Canal Point, FL; JEH, J. E. Hunter, Cornell University, Geneva, NY;
JEL, J. E. Leach, Kansas State University, Manhattan, KS; RES, R. E. Stall,
University of Florida, Gainesville, FL.


120
Table 5-3. Similarity values between Xanthomonas campestris pv. vignicola and
selected members of the clades 6 and 7 generated by comparison of the endonuclease
profiles of the DNA fragments related to the hrpB and hrpC/D complementation
groups of the hrp genes of X. campestris pv. vesicatoria.
Similarity of the hrp-related sequences of
X. campestris pv. vignicola
hrpB
hrpC/D
Clade 6a
X. campestris
pv. bilvae
0.61b
0.71
pv. citri A
0.53
0.75
pv. glycines A
0.58
0.75
pv. malvacearum
0.59
0.67
pv. phaseoli "fuscans"
0.58
0.83
pv. vitians B
0.59
0.67
Clade 7
X. campestris
pv. alfalfae
0.80
0.50
pv. citrumelo
0.87
0.53
pv. fici A
0.94
0.46
pv. poinsettiicola A
0.94
0.43
pv. pruni
0.94
0.46
pv. vesicatoria A
0.94
0.46
aClade established on the basis of the phylogenetic analysis of the restriction fragment
data of the 1,075-bp fragment related to the hrpC/D region of the hrp genes ofX.
campestris pv. vesicatoria.
bValues are the similarities estimated by the equation proposed by Nei and Li (1979)
for each /irp-related fragment restricted with either endonuclease Cfol, Haelll, Sau3M,
and Taql.


I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.
Robert E. Stall, Chair
Professor of Plant Pathology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.
Professor of Plant Pathology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.
Daryl R. Pring'
Professor of Plant Pathology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and
quality, as a dissertation for the degree of Doctor of Philosophy.
In
Professor of Microbiology and Cell
Science


51
al., 1991). Plasmid pXV56/3-48 fully restored the pathogenicity of mutants with
mutations in hrpB, hrpD, and hrpE, and the hypersensitive reaction-inducing ability but
not pathogenicity to the hrpC mutant. However, this plasmid failed to complement the
hrpA and hrpF mutants. Similarly, plasmid pXCP58/2 from X campestris pv.
pelargonii was also transferred into the six nonpathogenic 7n3-gus mutants ofX.
campestris pv. vesicatoria 85-10. This plasmid fully restored the wild type phenotype
to mutants with mutations in hrpC, hrpD, and hrpE but failed to complement hrpA,
hrpB, and hrpF7n3-gus mutants.
Sensitivity of the amplification of hrp fragment for detection ofX. campestris pv.
vesicatoria
The sensitivity of the amplification of specific DNA fragments in detection of
X. campestris was determined by using 10-fold dilutions of purified total bacterial
DNA of X. campestris pv. vesicatoria 75-3. The oligonucleotide primers RST9 plus
RST10 and RST2 plus RST3 were used for amplification of the 355- and 840-bp hrp,
respectively, from samples containing as little as 0.25 pg of total bacterial DNA after
30 cycles of DNA amplification (Fig. 3-7).
Discussion
Sequence homology to small hrp fragments amplified from X. campestris pv.
vesicatoria 75-3 was found among plant pathogenic strains of several pathovars of A
campestris and related Xanthomonas spp. by Southern hybridization analysis. DNA
probes representing regions of the hrpB, hrpC, and hrpD loci hybridized strongly to
total DNA from strains of 28 different pathovars ofX. campestris. These results
confirm and extend previous observations (Bonas et al., 1991; Stall and Minsavage,


44
Table 3-1. Hybridization of the hrp fragments amplified from Xanthomonas campestris
pv. vesicatoria 75-3 to total genomic DNA and amplification of Arp-related fragments
from different bacterial strains.
Strain
Southern hybridization
DNA amplification
355 bp
840 bp
1,075 bp
355 bp
840 bp 1,075 bp
Xanthomonas
X. campestris
pv. alfalfae KS
+
+a
+
(+)
+
+
pv. armoraciae 63-27
+
+
+
-
+
+
pv. bilvae XCB
+
+
+
-
+
+
pv. begoniae XCB9
+
+
+
-
+
+
pv. campestris 33913
+
+
4-
-
+
+
pv. carotae #13
+
+
+
-
(+)
+
pv. citri 9771
+
+
+
-
+
+
pv. citrumelo FI
+
+
+
(+)
+
+
pv. dieffenbachiae 729
+
+
+
-
+
+
pv. fici X151
4-
+
4-
4-
+
+
pv. gardneri XG101
+
+
+
-
(+)
+
pv. glycines 87-2
+
+
+
-
+
+
pv. holcicola G-23
+
+
+
-
-
+
pv. incanae 9561-1
+
+
+
-
+
+
pv. maculifoliigardeniae X22j
+
+
4-
(+)
+
+
pv. malvacearum RIATC
+
+
4-
-
+
+
pv. manihotis Xml25D
+
+
+
(+)
+
+
pv. papavericola XP5
+
+
+
-
(+)
+
pv. phaseoli 85-6
+
4-
+
-
+
+
pv. phaseoli "fuscans" XP163A
+
+
+
-
+
+
pv. physalidicola XP172
+
+
+
+
+
+
pv. pelargonii XCP58
4-
+
+
-
(+)
+
pv. poinsettiicola 071-424
+
+
+
-
+
+
pv. pruni X1219L
4-
+
+
-
+
+
pv. raphani 69-2
+
+
+
-
+
+
pv. secalis XC129C
-
-
-
-
-
-
pv. taraxaci XT11A
+
+
+
-
(+)
+
pv. translucens 80-1
-
-
-
-
-
-
pv. vesicatoria
75-3
+
+
+
+
+
+
XV56
-
+
+
-
+
+
pv. vignicola 81-30
+
+
+
-
+
+
pv. vitians XVIT
+
+
+
-
+
+
X198
+
+
+
+
+
+
XCF
+
+
+
(+)
+
+
T-55
-
-
-
-
-
-
INA
-
-
-
-
-
-
Continued on following page


83
carotae that could not be differentiated from the strains ofZ campestris pv. campestris
(Table 4-2). Other cases include the strain 82-17 of2f campestris pv. phaseoli that
could not be differentiated from strains of X. campestris pv. glycines group A; strains
F59 and F86 of X. campestris pv. citrumelo were indistinguishable from strains ofX.
campestris pv. pruni; strain XI25 of X. campestris pv. fici group A was identical to
strains ofX. campestris pv. poinsettiicola group A; strain XV2 ofX. campestris pv.
vitians was identical to strains of X campestris pv. malvacearum; and strain X52
isolated from Hibiscus sp. was indistinguishable from strains of X campestris pv.
poinsettiicola group B (Table 4-2).
The restriction analysis of the /zrp-related sequences revealed some apparent
groupings among the plant pathogenic xanthomonads based on the similarity in the
restriction banding patterns of homologous hrp-related fragments. The pathogens of
cruciferous plants, X. campestris pv. armoraciae, X campestris pv. campestris, X.
campestris pv. incanae, and X campestris pv. raphani have very similar restriction
profiles for both ^-related fragments and they formed a very distinct group (Table 4-
2). Strains of X campestris pv. armoraciae could be differentiated from X. campestris
pv. campestris and X. campestris pv. incanae only by the restriction analysis of the 840-
bp hrp related fragment with the endonuclease Cfol (Table 4-2). The restriction
patterns obtained for these pathovars were usually identical or very similar with the
exception to the restriction patterns of the 1,075-bp fragment for strains of2f
campestris pv. raphani (Table 4-2). Similarly, the strains of the pathovars X.
campestris pv. carotae, X. campestris pv. gardneri, X. campestris pv. papavericola, X.
campestris pv. pelargonii, X campestris pv. taraxaci, and X. campestris pv. vitians
group B also formed a unique group, although they cause diseases on different hosts.
The restriction analysis of the 1,075-bp /^-related fragment amplified from strains of


158
concentration of the inoculum was approximately 5 X 108 CFU/ml in sterile tap water,
determined by measuring the optical density in a Spectronic 20 spectrophotometer
(Bausch and Lomb, Inc., Rochester, NY). Plant reactions were scored over a period of
several days.
Recovery of bacteria from seeds and seedlings
The bacterial fraction was recovered from seeds and seedlings by using the
ultrasonication technique (Haefele and Webb, 1982; Morris, 1985). Two and one half
grams of seeds, or ten grams of aerial parts of seedling washings were washed in 20 ml,
or 150 ml of phosphate buffer (8.5 mM K2HP04, 7.5 mM KH2PO4, 0.02% Tween 20,
pH 7.0), respectively, for 20 min in a model B-22-4 ultrasonic cleaner (Branson
Cleaning Equipment Co., Shelton, CT). Bacterial populations were determined by the
dilution plate count method (Taylor, 1962). Dilutions were made from the washings in
sterile tap-water, and 0.1 ml aliquots were plated onto yeast-extract nutrient agar
(YNA) (Schaad and Stall, 1988), or Tween medium B (McGuire et al., 1986), or both.
X. campestris pv. vesicatoria was determined by the presence of circular, raised, and
yellow colonies surrounded by a lipolytic halo on Tween medium B (McGuire et al.,
1986; McGuire and Jones, 1989). Aliquots of the seed and seedling washings were
also processed and used in the DNA amplification and ELISA assays.
Extraction of DNA from bacteria and plant washings
Total genomic DNA of each bacterial strain was isolated by phenol-chloroform
extraction and ethanol precipitation essentially as described by Ausubel et al. (1987).
Bulk DNA extracts were obtained from seed and seedling washings by using a
modification of the method for DNA extraction described by Ausubel et al. (1987).


17
Phenotypic Characteristics
The differentiation and identification of Xanthomonas spp. have largely
depended on phenotypic characteristics (Bradbury, 1984; Holt et al., 1994; Schaad,
1988). Although members of the xanthomonads can be distinguished from the bacteria
of other genera on the basis of phenotypic features (Bradbury, 1984; Holt et al., 1994;
Schaad, 1988), it has been almost impossible to differentiate within the Xanthomonas
by biochemical and physiological features alone without knowing their plant hosts
(Bradbury, 1984; Dye, 1962; Holt et al., 1994; Schaad, 1988; Van den Mooter and
Swings, 1990). In fact, this was the major reason for the inclusion of the majority of
the former species of Xanthomonas into a single species, X. campestris (Bradbury,
1984; Dye et al., 1980).
Extensive studies have been carried out to compare the phenotypic
characteristics of strains representing the different species of the genus Xanthomonas,
pathovars of X campestris, and other bacteria genetically related to the xanthomonads
(Dye, 1962; Hildebrand et al., 1993; Van den Mooter and Swings, 1990). Van den
Mooter and Swings (1990) examined 295 morphological, biochemical, and
physiological features of 266 strains of Xanthomonas spp. and related bacteria. The
plant pathogenic xanthomonads were placed in seven different groups by cluster
analysis. The groups were delineated by strains of the taxaX. albilineans, X.
axonopodis, X campestris, X. campestris pv. graminis, X fragariae, X. oryzae, and X
populi, which largely correlate with the existing species of the genus Xanthomonas.
Strains ofX. campestris pv. graminis did not cluster with other pathovars ofV
campestris, and they were considered a homogeneous and different group on the basis
of phenotypic characteristics. However, other studies of phenotypic features (Van den


Table D-2Continued
Species/Pathovar Species/Pathovar
50
51
52
53
54
55
56
58
59
61
25. pv. dieffenbachiae B
.8511
.8333
.8511
.5263
.7826
.6047
.5455
.6522
.5217
.2791
26. pv. fici A
1.000
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
27. pv. fici A
.9362
.9167
.9362
.4211
.8696
.6512
.5000
.7826
.5217
.2791
28 pv. fici A
.9362
.9167
.9362
.4737
.8696
.6977
.5455
.8696
.5652
.3256
29. pv. fici B
.5116
.5000
.5116
.5294
.5238
.6154
.4500
.4286
.4286
.6154
31. pv. glycines A
.5652
.5532
.5652
.5405
.5778
.5714
.9302
.4889
.8444
.4762
34. pv. incanae
.5652
.5532
.5652
.5946
.5778
.5714
.5116
.4889
.5778
.5714
35. pv. maculifoliigardeniae
.6977
.7273
.6977
.5882
.7619
.9231
.6000
.6667
.6190
.5641
36. pv. malvacearum
.5778
.5652
.5778
.5556
.5909
.5854
1.000
.5000
.8182
.4390
37. pv. manihotis
.6977
.7273
.6977
.5294
.7619
.9231
.5500
.6667
.5714
.5128
40. pv. phaseoli A
.7273
.7556
.7273
.5714
.7907
1.000
.5854
.6977
.6047
.5000
41. pv. phaseoli B
.5652
.5532
.5652
.5405
.5778
.5714
.9302
.4889
.8444
.4762
42. pv. phaseoli "fiiscans"
.5532
.5833
.5532
.4737
.5652
.6047
.7727
.4783
.9565
.4186
43. pv. physalidicola
1.000
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
44. pv.poinsettiicola B
.4762
.4651
.4762
.4242
.4878
.5789
.4103
.4390
.3902
.6316
45. pv. poinsettiicola A
1.000
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
46. pv. pruni
1.000
.9796
1.000
.5128
.9362
.7273
.5778
.8085
.5957
.3636
47. pv. raphani A
.5652
.5532
.5652
.5946
.5778
.5714
.5116
.4889
.5778
.5714
48. pv. raphani B
.5217
.5106
.5217
.5405
.5333
.5238
.5116
.4444
.5778
.5238
Continued on following page


42
Fig. 3-3. Taql restriction endonuclease analysis of fragments of the hrp gene cluster
amplified from total genomic DNA from Xanthomonas campestris pv. vesicatoria 75-3
and from plasmids containing the hrp region. Lanes: M, phage X. restricted with Pstl; 1
to 3, the 355-bp fragments from X. campestris pv. vesicatoria 75-3 and plasmids pXV9
and pXV5.5, respectively; lanes 4 to 6, the 840-bp fragments froml campestris pv.
vesicatoria 75-3 and plasmids pXV9 and pXV5.5, respectively; lanes 7 to 9, the 1,075-
bp fragments froml campestris pv. vesicatoria 75-3 and plasmids pXV9 and pXV5.1,
respectively. Molecular sizes are given in base pairs.


126
Yersinia (Fenselau et al., 1992; Gough et al., 1992) is also intriguing. If this means a
convergent functional evolution or a common bacterial ancestry remains to be clarified,
Furthermore, the phylogeny of the hrp genes of plant pathogenic xanthomonads may
provide a framework and a rational basis through which origins and differentiation of
the xanthomonads may be assessed.


CHAPTER 3
OLIGONUCLEOTIDE PRIMERS FOR DETECTION AND
IDENTIFICATION OF PLANT PATHOGENIC STRAINS OF
Xanthomonas BY AMPLIFICATION OF DNA SEQUENCES
RELATED TO THE hrp GENES OF Xanthomonas campestris PV.
VESICATORIA
The genus Xanthomonas Dowson 1939 includes Gram-negative, usually
yellow-pigmented bacteria that occur worldwide and cause plant diseases. Over 124
monocotyledonous and 268 dicotyledonous plant species are hosts of Xanthomonas
(Leyns et al., 1984). Among the species o Xanthomonas, X. campestris comprises at
least 125 different pathovars that are distinguished by the diseases they cause
(Bradbury, 1984). The genus Xanthomonas also includes strains which may be
associated with plant material, but are not pathogenic to the plants from which they
were isolated (Angeles-Ramos et al., 1991; Gitaitis et al., 1987; Liao and Wells, 1987;
Maas et al., 1985). These opportunistic bacteria can be identified as xanthomonads by
the presence of xanthomonadins and by unique fatty acid profiles. Although the
identification of bacteria in the genus Xanthomonas presents no great problem, sub
generic identification of xanthomonads is still difficult.
Traditional methods for the detection and identification of plant pathogenic
xanthomonads rely on isolating the organism of interest in pure culture and performing
predetermined biochemical, serological, and pathological tests (Saettler et al., 1989;
Schaad, 1988). Sometimes, non-selective or selective enrichments are required to
increase the sensitivity of the isolation, which may be complicated by the presence of
fast-growing contaminating bacteria associated with plant tissue (Saettler et al., 1989).
30


161
Restriction endonuclease analysis
The hrp-related fragments amplified from different bacterial strains were
restricted with the frequent-cutting endonucleases Cfo\, HaeIII, Sau3 AI, or Taql,
according to conditions specified by the manufacturer (Promega). The restricted
fragments were separated by electrophoresis in 4% agarose gels (3% NuSieve and 1%
Seakem GTG [FMC BioProducts]) in TAE buffer at 8 V/cm. Phage X fid-restricted
DNA fragments were used as molecular standards. The gel was stained with 0.5 pg of
ethidium bromide per ml for 40 min and then destained in 1 mM MgS04 for 1 hr and
photographed over a UV transilluminator with type 55 Polaroid film. The restriction
pattern obtained was compared to the standard banding profiles established for the
different groups of plant pathogenic xanthomonads (Chapter 4).
ELISA assay
An ELISA procedure that was developed for detection of low populations ofX.
campests pv. vesicatoria was used (Somodi et al., 1993). Aliquots of the seed
washings were transferred to microfuge tubes and an equal amount of EDTA/lysozyme
lysis buffer (2.0 g KH2P04, 11.5 g Na2P04, 0.14 g EDTA disodium, 0.02 g thimerosal,
0.2 g lysozyme (Sigma), 1 1 deionized water) was added. The samples were
homogenized by vortexing and incubated for at least 16 hr at room temperature.
Immulon 2 (Dynatech Laboratories, Chantily, VA) flat bottom 96-well
microtiter plates were coated with a polyclonal antibody developed against X.
campestris pv. vesicatoria 75-3 and incubated at 4C overnight. The coating buffer and
phosphate buffered saline (PBS) were used as described by Clark and Adams (1977).
All further incubations were for 2 hr at 37C. The microtiter plates were washed with a
solution containing 0.8% NaCl and 0.1% Tween 20 in deionized water (NTrinse) (G.


14
constituted distinct and homogeneous groups based on protein profiles, such as X.
campestris pv. campestris, X campestris pv. graminis, X. campestris pv. hyacinthi, X.
campestris pv. pelargonii, X. campestris pv. pruni, and X. campestris pv. theicola. On
the other hand, the pathovars X. campestris pv. alfalfae, X. campestris pv. cajani, X.
campestris pv. glycines, X campestris pv. phaseoli, X campestris pv. phaseoli
"fuscans", and X. campestris pv. vignicola which cause diseases on leguminous plants
could not be differentiated from one another. The relatedness of this group of
pathogens of leguminous plants has also been confirmed by DNA-DNA hybridization
(Hildebrand et al., 1990; Vauterin et al., 1990ab) and by fatty acid composition (Yang
et al., 1993). Moreover, some pathovars, including X. campestris pv. alfalfae, X.
campestris pv. citri, X campestris pv. dieffenbachiae, X campestris pv. poinsettiicola,
X campestris pv. vesicatoria, X campestris pv. vignicola, and X. campestris pv.
vitians, were heterogeneous on the basis of protein profiles, consisting of two or more
distinct protein groups (Vauterin et al., 1991b). The diversity on the basis of protein
profile for some of these pathovars, i.e. X campestris pv. citri, X. campestris pv.
dieffenbachiae, and X. campestris pv. vesicatoria, correlates with the heterogeneity
based on genetic analysis and fatty acid composition (Chase at al., 1992; Egel et al.,
1991; Graham et al., 1990; Stall et al., 1994; Vauterin et al., 1991ab).
The complexity of the grouping of strains of X campestris that cause diseases
on citrus on the basis of whole-cell protein profiles (Vauterin et al., 1991ab) is in
agreement with the diversity determined by genetic analysis (Egel et al., 1991; Gabriel
et al. 1989; Hartung and Civerolo, 1987) and fatty acid composition (Graham et al.,
1990; Vauterin et al., 1991b). Based on protein profile, the strains of the citrus canker
A of campestris pv. citri form a well defined and homogeneous group (Vauterin et
al., 1991b). The uniformity of this group was also revealed by DNA-DNA


178
Tentative identification of A', campestris associated with seeds and seedling? of pepper
and tomato
The identity of 56 strains of Xanthomonas from seeds and seedlings of pepper and
tomato was investigated by examining the reaction on plants and by analyzing DNA
sequences related to the hrp genes and fatty acid composition (Table 7-1). The strains
seem to comprise a very diverse group of xanthomonads that includes plant pathogenic
and nonplant pathogenic bacteria. Some of the strains were pathogenic on plants of
Bonny Best tomato and were identified as X. campestris pv. vesicatoria on the basis of
the hrp fragment analysis, with the exception of the strains DM-1 and 9311 (Table 7-1).
In the case of the latter strains, /^-related fragments were amplified, but the restriction
analysis of the fragments were not done. The strains SP268.92, P996, 639-6/FS1, 639-
6/FS2, T93-12A, BSA5, and BSA6 did not show any reaction on plants and they failed to
produce the 840- and 1,075-bp hrp-related fragments in the DNA amplification assay
(Table 7-1). Although these strains have some degree of similarity to different pathovars
of* campestris on the basis of fatty acid composition (Table 7-1), they resemble the
opportunistic xanthomonads that have been reported previously in association with
tomato and pepper transplants (Gitaitis et al., 1987, 1992). The remainder of strains form
a very diverse group of potential plant pathogenic xanthomonads, as determined by the
hypersensitive reaction on plants, or by the presence in their genome of a region similar
to the hrp genes, or by both criteria (Table 7-1).
The identification of these strains of Xanthomonas based on the hrp analysis was
performed by comparison of the restriction profile generated by digestion of the hrp
fragment amplified from the strains with frequent-cutting endonucleases (Fig. 7-6) to the
pattern established for the different groups of plant pathogenic xanthomonads (Chapter
4). The restriction analyses allowed the identification of groups of strains of X.
campestris pv. vesicatoria (Table 7-1). Two strains were identified to the group A, five to


235
Kamoun, S., Kamdar, H. V., Tola, E., and Kado, C. I. 1992. Incompatible interactions
between crucifers and Xanthomonas campestris involve a vascular hypersensitive
response: role of the hrpXlocus. Mol. Plant-Microbe Interact. 5:22-33.
Kawasaki, E. S. 1990. Sample preparation from blood, cells, and other fluids. Pages
146-152 in: PCR Protocols: A Guide to Methods and Applications. M.A. Innis, D.H.
Gelfand, J.J. Sninsky, and T.J. White, eds. Academic Press, Inc., San Diego, CA.
Kersters, K., and DeLey, J. 1975. Identification and grouping of bacteria by numerical
analysis of their electrophoretic protein patterns. J. Gen. Microbiol. 87: 333-342.
Kersters, K., Pot, B., Hoste, B., Gillis, M., and DeLey, J. 1989. Protein
electrophoresis and DNA:DNA hybridization of xanthomonads from grasses and
cereals. EPPO Bull. 19:51-55.
Koike, H. 1965. The aluminum-cap method for testing sugarcane varieties against leaf
scald disease. Phytopathology 55:317-319.
Krawiec, S., and Riley, M. 1990. Organization of the bacterial chromosome.
Microbiol. Rev. 54: 502-539.
Krieg, N. R., and Holt, J. G., eds. 1984. Bergey's Manual of Systematic Bacteriology,
vol. 1. Williams & Wilkins, Baltimore, MD. 964 pp.
Kubicek, Q. B., Civerolo, E. L., Bonde, M. R., Hartung, J. S., and Peterson, G. L.
1989. Isozyme analysis of Xanthomonas campestris pv. citri. Phytopathology 79:
297-300.
Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head
of bacteriophage T4. Nature 227:680-685.
Lazo, G. R., and Gabriel, D. W. 1987. Conservation of plasmid DNA sequences and
pathovar identification of strains of Xanthomonas campestris. Phytopathology 77:
448-453.
Lazo, G. R., Roffey, R., and Gabriel, D. W. 1987. Pathovars of Xanthomonas
campestris are distinguishable by restriction fragment-length polymorphism. Int. J.
Syst. Bacteriol. 37:214-221.
Leach, J. E., Rhoads, M. L., Vera Cruz, C. M., White, F. F., Mew, T. W., and Leung,
H. 1992. Assessment of genetic diversity and population structure of Xanthomonas


104
DNA amplification
Two sets of oligonucleotide primers selected from the nucleotide sequence of
the hrp gene cluster of X. campestris pv. vesicatoria were used in this study. Primers
RST2 plus RST3 delineated an 840-bp fragment and RST21 plus RST22 delineated a
1,075-bp fragment of the complementation groups hrpB and hrpUD ofX. campestris
pv. vesicatoria, respectively (Chapter 3). The primers were used in polymerase chain
reaction for specific amplification of homologous /ir/?-related DNA fragments from
different plant pathogenic xanthomonads. The reaction conditions and polymerase
chain reaction cycles were previously described (Chapter 3).
Data analysis
The electrophoretic patterns of the /zrp-related DNA fragments restricted with
each of the four endonucleases were used to determine the restriction banding profile
for each bacterial strain. The codes 0 and 1 were assigned according to the absence or
presence of each DNA band, respectively. The genetic relationship between strains
was estimated based on the resulting matrix by determining the proportion of shared
DNA fragments (F). The equation proposed by Nei and Li (1979), F = 2/ (nx + ny),
where nxy is the number of fragments shared by both strains, and nx and ny are the total
number of fragments for each strain, was used to estimate the proportion of shared
fragments (F) by using a computer program (Appendix B) written for the SAS system
(SAS Institute Inc., Cary, NC). The genetic divergence between strains was
determined as the estimate of the number of nucleotide substitutions per site (5), on the
basis of the proportion of shared DNA fragments (Nei and Li, 1979). The number of
nucleotide substitutions per site (8) was calculated based on the iterative method


156
nature of antigenic determinants in immunogens of the xanthomonads poses a major
limitation for the use of polyclonal antibodies. For instance, polyclonal antibodies
developed for strains of X. campestris pv. vesicatoria cross-reacted with different plant
pathogenic xanthomonads (O'Brien et al., 1967). In contrast, strains ofX. campestris
pv. vesicatoria are serologically diverse and no single polyclonal or monoclonal
antibody has been obtained that reacts to all strains of this pathogen (Benedict et ah,
1990; Jones et ah, 1993b; Stall et ah, 1994). More recently, nucleic acid based
techniques have been examined for specific detection of X. campestris pv. vesicatoria
(Garde and Bender, 1991; Jones et ah, 1993b). Garde and Bender (1991) developed
DNA probes for detection of copper resistant X. campestris pv. vesicatoria, but the
copper region from X. campestris pv. vesicatoria may hybridize to other plant
pathogens, such as the tomato pathogen Pseudomonas syringae pv. tomato (Voloudakis
etah, 1993).
The objective of this study was to examine a procedure based on the
amplification of DNA fragments related to the hrp gene cluster of X. campestris pv.
vesicatoria (Chapter 3) for the specific detection and identification of plant pathogenic
xanthomonads associated with pepper and tomato seeds. The sensitivity and specificity
of the polymerase chain reaction (PCR) technique were determined by the detection of
X. campestris pv. vesicatoria in artificially and naturally contaminated seeds lots. The
reliability of the technique for detection of X. campestris pv. vesicatoria was further
investigated by comparison with the dilution plating on semiselective media and
detection by ELISA. Furthermore, plant pathogenic xanthomonads isolated from
pepper and tomato seeds were tentatively identified based on the analysis of the
restriction pattern produced from amplified DNA fragments related to the hrp genes.


page
5 PHYLOGENETIC ANALYSIS OF PLANT
PATHOGENIC Xanthomonas BASED ON DNA
SEQUENCES RELATED TO THE hrp GENES 100
Material and Methods 103
Results 105
Discussion 121
6 GENETIC ANALYSIS OF hrp RELATED DNA
SEQUENCES OF Xanthomonas campestris STRAINS
CAUSING DISEASES OF CITRUS 127
Material and Methods 129
Results 134
Discussion 149
7 EVALUATION OF A DNA AMPLIFICATION
APPROACH FOR DETECTION AND IDENTIFICATION
OF PLANT PATHOGENIC XANTHOMONADS
ASSOCIATED WITH PEPPER AND TOMATO SEEDS 154
Material and Methods 157
Results 167
Discussion 180
8 SUMMARY AND CONCLUSIONS 186
APPENDICES
A BACTERIAL STRAINS AND PLASMIDS USED IN
THIS STUDY 195
B SAS PROGRAM TO ESTIMATE THE SIMILARITY
VALUES FROM RESTRICTION FRAGMENT DATA 201
C SAS PROGRAM TO ESTIMATE THE NUCLEOTIDE
SUBSTITUTION BASED ON RESTRICTION
FRAGMENT DATA 204
v


APPENDIX C
SAS PROGRAM TO ESTIMATE THE NUCLEOTIDE
SUBSTITUTION BASED ON RESTRICTION FRAGMENT DATA
data frag;
infile fragment; /* Read data of similarity values from */
input similar @@; /* file fragment */
g0=similar**0.25;
g=(similar*(3-(2*g0)))**0.25;
gl=G0;
x=l;
do while (gl ne g);
gi=g;
if x ne 50 then do;
g=(similar*(3-(2*gl)))**0.25;
x=x+l;
end;
end;
do differe=-((2/r)*(log(g)));
end;
output;
/* Determine the probability that */
/* a restriction site has remained */
/* unaltered */
/* Determine the nucleotide substitution; */
/* r is the number of nucleotides in the */
/* recognition site of the restriction */
/* endonuclease */
proc print split-/* Print out the result */
var similar gO g differe x;
label similar='Similarity'
gO='Initial*probability'
g='Final*probability'
differe-N ucleotide substitution'
x-N umber* iteration';
title 'DNA similarity and nucleotide substitution';
title2 'Nei and Li (1979) procedure';
run;
204


Table D-2Continued
Species/Pathovar Species/Pathovar
37
40
41
42
43
44
45
46
47
48
1. pv. alfalfae
.5854
.6190
.4545
.4444
.8696
.4000
.8696
.8696
.4545
.4091
2. pv. alfalfae
.5854
.6190
.4545
.4444
.8696
.4000
.8696
.8696
.4545
.4091
3. pv. armoraciae
.4878
.5238
.5455
.5333
.5217
.5500
.5217
.5217
.9545
1.000
4. pv. begoniae A
.8947
.9231
.5366
.6667
.7442
.5405
.7442
.7442
.5366
.4878
5. pv. begoniae A
.8947
.9231
.5366
.6667
.7442
.5405
.7442
.7442
.5366
.4878
6. pv. begoniae B
.8947
.9231
.5366
.6667
.7442
.5405
.7442
.7442
.5366
.4878
7. pv. bilvae
.5714
.6047
.8444
.9565
.5957
.3902
.5957
.5957
.5778
.5778
8. pv. campestris
.5366
.5714
.5455
.5333
.5652
.6000
.5652
.5652
1.000
.9545
11. pv. citri A
.5366
.5714
.9091
.8000
.5217
.4000
.5217
.5217
.5455
.5455
12. pv. citri B
.5854
.5714
.8182
.9333
.5217
.3500
.5217
.5217
.5455
.5455
13. pv. citri C
.5854
.5714
.8182
.9333
.5217
.3500
.5217
.5217
.5455
.5455
14. pv. citrumelo
.6190
.6512
.4889
.4783
.9362
.4390
.9362
.9362
.4889
.4444
15. pv. citrumelo
.6977
.7273
.5652
.5532
1.000
.4762
1.000
1.000
.5652
.5217
16. pv. citrumelo
.6977
.7273
.5652
.5532
1.000
.4762
1.000
1.000
.5652
.5217
17. pv. ctirumelo
.6667
.6977
.5333
.5217
.9362
.4390
.9362
.9362
.5333
.4889
18. pv. citrumelo
.6190
.6512
.4889
.4783
.9362
.4390
.9362
.9362
.4889
.4444
19. pv. citrumelo
.5854
.6190
.4545
.4444
.8261
.4500
.8261
.8261
.4545
.4091
20. pv. citrumelo
.6977
.7273
.5652
.5532
1.000
.4762
1.000
1.000
.5652
.5217
21. pv. citrumelo
.5854
.6190
.4545
.4444
.8696
.4000
.8696
.8696
.4545
.4091
22. pv. citrumelo
.6977
.7273
.5652
.5532
.9583
.5238
.9583
.9583
.5652
.5217
23. pv. dieffenbachiae A
.9231
1.000
.5714
.6047
.7273
.5789
.7273
.7273
.5714
.5238
24. pv. dieffenbachiae A
.9231
1.000
.5714
.6047
.7273
.5789
.7273
.7273
.5714
.5238
Continued on following page
to
to


Fig. 6-7. Rooted tree for 19 strains of X. campestris inferred from restriction analysis
data of DNA fragments related to the complementation groups B and C/D of the hrp
genes of X. campestris pv. vesicatoria generated by the KITSCH procedure from the
PHYLIP computer package by using the Fitch-Margoliash method.


Table D-lContinued
Species/Pathovar
Species/Pathovar
32
33
34
35
36
37
38
39
40
41
23. pv. dieffenbachiae A
.1364
.1250
.1818
.7391
.3111
.8163
.2667
.2222
.8936
.2917
24. pv. dieffenbachiae A
.1818
.1250
.1818
.7826
.2667
.8571
.2667
.2222
.9362
.2500
25. pv. dieffenbachiae B
.2000
.2273
.2500
.6667
.3415
.7111
.3415
.2927
.6047
.5000
26. pv. fici A
.2273
.1667
.2727
.5652
.4889
.5714
.3556
.3111
.5106
.5833
27. pv. fici A
.3111
.1633
.3556
.4255
.4783
.4400
.3913
.3913
.4167
.5306
28. pv. fici A
.3111
.1633
.3556
.4255
.4783
.4400
.3913
.3913
.4167
.5306
29. pv. fici B
1.000
.1818
.4000
.2381
.2439
.2222
.3902
.4390
.1860
.2273
30. pv. gardneri
.4000
.3182
.9000
.3333
.3415
.3111
.8293
.8780
.2326
.3636
31. pv. glycines A
.2273
.2500
.3636
.3478
.8000
.4082
.4444
.4444
.2979
1.000
32. pv. glycines B
.1818
.4000
.2381
.2439
.2222
.3902
.4390
.1860
.2273
33. pv. holcicola
.3182
.1304
.1778
.1633
.3111
.3111
.1277
.2500
34. pv. incanae
.3333
.3415
.2667
.8780
.9268
.1860
.3636
35. pv. maculifoliigardeniae
.4186
.8085
.4186
.3721
.8000
.3478
36. pv. malvacearum
.3478
.4286
.4286
.3182
.8000
37. pv. manihotis
.
.3478
.3043
.8750
.4082
38. pv. papavericola
.
.
.9524
.2727
.4444
39. pv. pelargonii
.
.
.2273
.4444
40. pv. phaeoli A
.
.

.
.2979
41. pv. phaseoli B




Continued on following page


31
More recently, methods based on metabolic and protein profiling (Chase et al., 1992;
Van den Mooter and Swings, 1990; Vauterin et al., 199lab), and fatty acid analysis
(Chase et al., 1992; Yang et al., 1993) have been used for identification, but isolation
and purification of the bacterial strain is still required. Polyclonal or monoclonal
antisera produced against strains of X. campestris have been used for detection and
identification, but they have provided variable results. Polyclonal antisera may cross-
react with other bacteria and may be unable to differentiate specific strains or pathovars
ofX. campestris (Alvarez and Lou, 1982). Several monoclonal antisera were produced
which reacted specifically with all strains of some pathovars ofX. campestris that
infect relatively few genera of hosts (Alvarez et al., 1991; Benedict et al., 1990).
However, for certain X. campestris pathovars, mainly those that infect several hosts
from different genera, no pathovar-specific monoclonal antisera that react with all
strains of the respective pathovar have been found (Alvarez et al., 1991; Jones et al.,
1993b).
Nucleic acid-based techniques have also been applied for detection and
identification of plant pathogenic bacteria (Bereswill et al., 1992; Manulis et al., 1991;
Schaad et al., 1989; Seal et al., 1992), including some members of the xanthomonads
(Garde and Bender, 1991; Gilbertson et al., 1989; Hartung, 1992; Hartung et al., 1993;
Lazo and Gabriel, 1987; Lazo et al., 1987). The techniques developed for detection
and identification of xanthomonads were based on random probes (Lazo et al., 1987),
or on plasmid DNA fragments specific for a few pathovars of X campestris (Gilbertson
et al., 1989; Hartung, 1992; Hartung et al., 1993; Lazo and Gabriel, 1987), or even for a
group of strains (Garde and Bender, 1991). Highly conserved regions in the bacterial
genome of plant pathogenic bacteria could be more useful for the selection of specific


103
data were used to establish the genetic relationship of the hrp genes between strains and
to infer the phylogeny of this region of the bacterial genome for different plant
pathogenic xanthomonads.
Materials and Methods
Bacterial strains and culture conditions
The bacterial strains used in this study and their sources are listed in appendix
A. The identity of the bacterial strains used here was confirmed by fatty acid analysis
(N. C. Hodge, personal communication). Strains of A! campestris were grown on
nutrient agar (Becton Dickinson, Cockeysville, MD) at 28C, with the exception of
strains of X. campestris pv. citri group B that were grown on a sucrose based medium
(Canteros de Echenique et al., 1985). Strains of A! fragariae were cultivated on
Wilbrink's medium (Koike, 1965).
DNA manipulations
Total genomic DNA of each strain was isolated by phenol-chloroform
extraction and ethanol precipitation essentially as described by Ausubel et al. (1987).
The restriction endonuclease analyses of the DNA fragments amplified from different
bacterial strains were accomplished by using the frequently cutting endonucleases Cfol,
Haelll, Sau3 Al, and Taql under the conditions specified by the manufacturer
(Promega, Madison, WI). The restricted fragments were resolved by electrophoresis in
4% agarose gels (3% NuSieve and 1% SeaKem GTG [FMC BioProducts, Rockland,
ME]) in TAE buffer (40 mM Tris acetate, 1 mM EDTA, pH 8.2) at 8 V/cm.


32
DNA probes for detection and identification of a larger number of strains, pathovars, or
species of Xanthomonas.
The hrp gene clusters that determine hypersensitivity and pathogenicity may be
appropriate for selection of probes for detection and identification of plant pathogenic
bacteria. The hrp gene cluster is required by bacterial plant pathogens to produce
symptoms on susceptible hosts and a hypersensitive reaction (HR) on resistant hosts, or
on nonhosts (Willis et al., 1991), and has been found in several plant pathogenic
bacteria, such as Erwinia amylovora (Beer et al., 1991), Pseudomonas solanacearum
(Boucher et al., 1987), P. syringae pv. phaseolicola (Lindgren et al., 1986), and A
campestris pv. vesicatoria (Bonas et al., 1991). Furthermore, hrp functions seem to be
highly conserved among a number of plant pathogenic bacteria (Ulla Bonas, personal
communication; Fenselau et al., 1992; Gough et al., 1992; Hwang et al., 1992). The
hrp genes of plant pathogenic bacteria are also very similar at the protein level to genes
that are involved in the secretion of pathogenicity factors by bacterial pathogens of
mammals (Fenselau et al., 1992; Gough et al., 1992). By contrast, nonpathogenic
bacteria are unable to produce symptoms on susceptible hosts and HR on nonhosts,
apparently because they do not possess DNA sequences similar to hrp genes (Lindgren
et al., 1986; Stall and Minsavage, 1990). Physically and functionally similar hrp
sequences occur among several pathovars of X. campestris, but not in opportunistic
xanthomonads (Bonas et al., 1991; Stall and Minsavage, 1990).
The objective of this study was to examine sequences of the hrp genes of A
campestris pv. vesicatoria for detection and identification of plant pathogenic
xanthomonads. Oligonucleotide primers specific for hrp genes were tested for their
suitability for identification of these xanthomonads by the polymerase chain reaction


using other methods. The restriction banding profiles generated for the hrp fragments
may be an easy and a discriminating approach for identification of plant pathogenic
xanthomonads.
IX


Table D-2. Similarity values for plant pathogenic strains of different pathovars of Xanthomonas campestris and X. fragariae
based on the genetic analysis of the hrpB related DNA sequences.
Species/Pathovar
Species/Pathovar
2 3
4
5
6
7
8
11
12
13
1. pv. alfalfae
1.000 .4091
.6341
.6341
.6341
.4889
.4545
.4091
.4091
.4091
2. pv. alfalfae
.4091
.6341
.6341
.6341
.4889
.4545
.4091
.4091
.4091
3. pv. armoraciae
.4878
.4878
.4878
.5778
.9545
.5455
.5455
.5455
4. pv. begoniae A
1.000
1.000
.6190
.5366
.5366
.5854
.5854
5. pv. begoniae A
1.000
.6190
.5366
.5366
.5854
.5854
6. pv. begoniae B
.6190
.5366
.5366
.5854
.5854
7. pv. bilvae
.
.
.5778
.8444
.9333
.9333
8. pv. campestris
.
.
.
.5455
.5455
.5455
11. pv. citri A
.
.8182
.8182
12. pv. citri B
.
.
.
.
1.000
13. pv. citri C





Continued on following page


129
polymorphism of restriction fragments of a homologous DNA sequence occurs among
pathovars of A campestris (Bonas et al., 1991). Opportunistic xanthomonads, which
produce limited symptoms in susceptible hosts and no hypersensitive reaction in
nonhosts, do not possess DNA similar to an hrp gene cluster (Stall and Minsavage,
1990).
The genomic similarity of strains of X. campestris pv. citrumelo has been
investigated (Egel et al., 1991; Gabriel et al., 1988, 1989; Graham et al., 1990; Hartung
and Civerolo; 1987, 1989; Vauterin et al., 1991b), but examination of the similarity
between hrp clusters of these strains adds information on the comparative genetics of
pathogenicity. Similar hrp gene clusters among strains with relatively divergent
genetic backgrounds might indicate a similar origin of pathogenicity, and dissimilar
hrp gene clusters would support the hypothesis that many strains of X. campestris
involved in the citrus bacterial spot disease are diverse. The similarity of the hrp gene
in strains of X. campestris pv. citri and X campestris pv. citrumelo was investigated by
amplifying and restricting two DNA fragments of the hrp complementation groups B
and C/D (Chapter 3; Bonas et al., 1991), which are highly conserved among several
pathovars of A! campestris (U. Bonas, personal communication).
Materials and Methods
Culture conditions
The strains of X. campestris used in this study and their sources are listed in
Appendix A. All strains had previously been identified as members of X. campestris
by fatty acid analysis (N. C. Hodge, personal communication). Citrus bacterial spot
strains were rated for pathogenicity by Graham and Gottwald (1990) and by Graham et


238
Nei, M, and Li, W. H. 1979. Mathematical model for studying genetic variation in
terms of restriction endonucleases. Proc. Natl. Acad. Sci. USA 76: 5269-5273.
Obata, T., and Tsuboi, F. 1972. Studies on the serological diagnosis of bacterial plant
diseases. I. Cross-agglutination and gel-diffusion tests with some Xanthomonas nomen
species. Res. Bull. Plant Prot. Ser. Jpn. 10: 8-16.
O'Brien, L. M., Morton, D. J., Manning, W. J., and Scheetz, R. W. 1967. Serological
differences between apparently typical pepper and tomato isolates of Xanthomonas
vesicatoria. Nature 215:532-533.
Ogram, A., Sayler, G. S., and Barkay, T. 1988. DNA extraction and purification from
sediments. J. Microbiol. Methods. 7 : 57-66.
Palleroni, N. J., Hildebrand, D. C., Schroth, M. N., and Hendson, M. 1993.
Deoxyribonucleic acid relatedness of 21 strains o Xanthomonas species and pathovars.
J. Appl. Bacteriol. 75:441-446.
Palleroni, N. J., Kunisawa, R., Contopoulou, R., and Doudoroff, M. 1973. Nucleic
acid homologies in the genus Pseudomonas. Int. J. Syst. Bacteriol. 23: 333-339.
Permar, T. A., and Gottwald, T. R. 1989. Specific recognition of a Xanthomonas
campestris Florida citrus nursery strain by a monoclonal antibody probe in a
microfiltration enzyme immunoassay. Phytopathology 79: 780-783.
Pickup, R. W. 1991. Development of molecular methods for the detection of specific
bacteria in the environment. J. Gen. Microbiol. 137:1009-1019.
Pohronezny, K., Stall, R. E., Canteros, B. I., Kegley, M., Datnoff, L. E., and
Subramanya, R. 1992. Sudden shift in the prevalent race of Xanthomonas campestris
pv. vesicatoria in pepper fields in southern Florida. Plant Dis. 76: 118-120.
Pohronezny, K,. and Volin, R. B. 1983. The effect of bacterial spot on yield and
quality of fresh market tomatoes. HortScience 18:69-70.
Pohronezny, K., Waddill, V. H., Schuster, D. J., and Sonoda, R. M. 1986. Integrated
pest management for Florida tomatoes. Plant Dis. 70:96-102.
Pruvost, O., Hartung, J. S., Civerolo, E. L., Dubois, C., and Perrier, X. 1992. Plasmid
DNA fingerprints distinguish pathotypes of Xanthomonas campestris pv. citri, the
causal agent of citrus bacterial canker disease. Phytopathology 82: 485-490.


9
geographical distribution of the strains. In a similar study, six MABs were used to
group 323 strains of A campestris pv. dieffenbachiae isolated from different plants of
the Araceae family (Lipp et al., 1992). The strains of A! campestris pv. dieffenbachiae
were serologically heterogeneous and 12 major serogroups were identified. The
serotype did not correspond strongly to the host of origin. Strains of A campestris that
cause diseases on citrus are also serologically diverse (Alvarez et ah, 1990, 1991;
Benedict et ah, 1985; Gottwald et ah, 1991; Permar and Gottwald, 1989). An MAB
developed to A campestris pv. citri of the group A of citrus canker did not react with
strains of B, C, and D forms of citrus canker (Alvarez et ah, 1991). Although this
antibody also did not react with 130 other Aanthomonas pathovars and species, it
reacted with some strains of A campestris pv. citrumelo, the citrus bacterial spot agent,
and with a strain of A campestris pv. manihotis. However, in serological studies with
monoclonal and polyclonal antibodies a close antigenic relationship existed between
the strains of the groups B, C, and D forms of citrus canker (Alvarez et ah, 1991;
Civerolo and Fan, 1982). The citrus bacterial spot pathogen, A campestris pv.
citrumelo, is also serologically heterogeneous, and no single MAB reacted with all
strains of this pathogen (Alvarez et ah, 1991; Gottwald et ah, 1991). Although A
campestris pv. citrumelo is serologically distinct from strains of A campestris pv. citri,
some strains of the citrus bacterial spot pathogen share a common epitope with
members of other pathovars of A campestris (Alvarez et ah, 1991; Gottwald et ah,
1991). The diversity of A campestris pv. citrumelo is also reflected genetically and in
differential host reactions (Alvarez et ah, 1991; Egel et ah, 1991; Gottwald et ah, 1991;
Graham and Gottwald, 1990).
An interesting point emerging from these studies on serological relationships
among different pathovars of A campestris by using MABs is the inverse correlation


Table D-lContinued
Species/Pathovar
Species/Pathovar
42
44
45
46
47
48
49
50
51
52
45. pv. poinsettiicola A
.8980
.3636
.3636
.3636
.8750
.7660
.7660
46. pv. pruni
.4000
.4000
.3556
.7755
.7500
.7500
47. pv. raphani A
1.000
.4500
.2727
.3256
.3256
48. pv. raphani B
.4500
.2727
.3256
.3256
49. pv. taraxaci
.3182
.2791
.2791
50. pv. vesicatoria A
.
.8936
.8936
51. pv. vesicatoria A
.
.
.
.
1.000
52. pv. vesicatoria A




Continued on following page


101
exist among the strains of X. campestris as demonstrated by protein and fatty acid
profiles (Vauterin et al., 1991a; Yang et al., 1993). Further, genetic analyses of the
xanthomonads by DNA-DNA hybridization has revealed a diverse genetic background
(Hildebrand et al., 1990; Palleroni et al., 1993; Vauterin et al., 1992).
A characteristic feature of plant pathogenic xanthomonads is the presence of a
genomic region that contains genes {hrp) for the hypersensitive and pathogenic
reactions on plants. This hrp gene cluster is required for plant pathogens to cause
disease on susceptible hosts and hypersensitive reaction (HR) on resistant, or on
nonhost plants (Willis et al., 1991). Besides the xanthomonads, the hrp genes have
also been found in other Gram negative plant pathogenic bacteria of the genera Erwinia
(Beer et al., 1991) and Pseudomonas (Boucher et al., 1987; Huang et al., 1990;
Lindgren et al., 1986). Furthermore, the hrp genes seem to be highly conserved among
different bacteria at the structural and functional levels (Fenselau et al., 1992; Gough et
al., 1992; Hwang et al., 1992). The hrp genes of plant pathogenic bacteria are also
similar at the protein level to virulence determinants of bacteria of the genus Yersinia,
pathogens of animals (Fenselau et al., 1992; Gough et al., 1992). On the other hand,
nonpathogenic bacteria that are unable to cause disease or HR on plants lack DNA
similar to the hrp genes sequence (Chapter 3; Bonas et al., 1991; Lindgren et al., 1986;
Stall and Minsavage, 1990).
Conservation of the hrp genes at the structural and functional level among
different plant pathogenic xanthomonads is well documented. The hrp genes are
organized in a large cluster in the bacterial genome (Arlat et al., 1991; Bonas et al.,
1991) although an additional small hrp region unrelated to the major hrp genes has also
been identified (Kamoun and Kado, 1990; Kamoun et al., 1992). Southern
hybridization analyses using the hrp genes of Xanthomonas campestris pv. vesicatoria


Fig. 4-6. Restriction profiles established for the 1,075-bp Arp-related fragment
amplified from different strains of plant pathogenic Xanthomonas spp. and restricted
with the endonuclease HaeIII. Lane M, phage X restricted with Psft. Molecular sizes
are given in bases. See Table 4-2 for identification of strains of Xanthomonas spp. for
each restriction pattern.


Table 4-2Continued
Species/Pathovar
Strain
840 bp
1,075 bp
Taql
Sau3Al Haelll
Cfol
Taql
Sau3Al Haelll
Cfol
pv. citri
canker A
3213,3340, 9760-2, 9771, Tl,
115A
2
5
1
8
7
9
10
7
canker B
B64, B69, B80, B84, B93, B94,
B148
4
5
8
9
7
9
11
11
canker C
70C, 338, 339, 340, 341,342
4
5
8
9
7
9
11
11
pv. citrumelo
FI, F54, F274, F361,3166
7
6
16
12
8
11
14
14
F59, F86
3
6
16
12
8
11
19
8
F94
3
6
16
12
8
11
15
17
F306
5
6
16
12
8
11
15
17
F254,F311
7
6
16
12
8
11
15
8
F6, F228
8
6
16
14
8
13
15
15
F100
3
6
16
12
8
12
15
11
F348
8
6
16
12
8
12
19
16
F378
3
6
16
14
8
12
14
8
Continued on following page


144
The relatedness of the hrp genes of strains of the citrus pathogens X campestris
pv. citrumelo and X. campestris pv. citri to the hrp gene cluster of some other
pathovars of X. campestris was also investigated. The hrp gene cluster of strains of X.
campestris pv. alfalfae 82-1, X. campestris pv. fici X151, and X. campestris pv.
vesicatoria 75-3 and X campestris XI98 were closely related to hrp genes ofX.
campestris pv. citrumelo, and the genetic divergence ranged from 0.004 to 0.023
nucleotide substitution per site (Table 6-1). However, strains ofX. campestris pv. citri
were much less related to those four strains of X. campestris, with genetic divergence
ranging from 0.049 to 0.061 nucleotide substitutions per site (Table 6-1). The hrp
genes of strains of X. campestris pv. citri were highly related to X campestris pv.
bilvae XCB and X. campestris XCF. The genetic divergence of the hrp genes of X.
campestris pv. citri from the genes of these strains of X. campestris ranged from 0.011
to 0.018 nucleotide substitution per site (Table 6-1). Moreover, X. campestris pv.
maculifoliigardeniae strain X22j has hrp genes not highly related to any of the
xanthomonads from citrus, with a genetic divergence ranging from 0.046 from Y
campestris pv. citrumelo strain 534 to as high as 0.082 from Y. campestris pv. citri
group A (Table 6-1).
The restriction fragment data of the hrp genes encoded 0 or 1 and the distance
matrix (Table 6-1) were used to construct phylogenetic trees based on a parsimony
criterion by using the BOOT program and a distance method by using the KITSCH
program of the PHYLIP computer package (Felsenstein, 1991), respectively. Although
the general topology is slightly different, the phylogenetic trees inferred by using two
different approaches of tree reconstruction showed very similar branching patterns for
the major clades (Fig. 6-6 and 6-7). The branching pattern obtained with the BOOT
program is unrooted, although the strain X. campestris pv. maculifoliigardeniae X22j


115
pv. vesicatoria A (1)
pv. vesicatoria A (24)
pv. citrumelo (1)
pv. citrumelo (I)
pv. poinsettiicola A (2)
pv. fici A (1)
pv. citrumelo (2)
pv. pruni(4)
pv. physalidicola (1)
pv. vesicatoria (2)
pv. citrumelo (1)
pv. vignicola (3)
pv. citrumelo (I)
pv. fici A (I)
pv. citrumelo (2)
pv. alfalfae (I)
pv. alfalfae (I)
pv. alfalfae (I)
pv. fici A (1)
pv. citrumelo (2)
pv. citrumelo (5)
X. campestris X198 (1)
pv. diefienbachiae B (2)
pv. begoniae (7)
pv. begoniae (1)
pv. begoniae (1)
pv. diefienbachiae A (4)
pv. phaseoli A (4)
pv. vitians A (3)
pv. dieffienbachiae A (4)
pv. manihotis (1)
pv. maculifoliigardeniae (I)
pv. phaseoli 'Tuscans" (1)
X. campestris XCF (I)
pv. bilvae (1)
pv. citri B (7)
pv. citri A (6)
pv. phaseoli B (1)
pv. glycines A (6)
pv. malvacearum (9)
pv. vitians B (1)
pv. citri A (6)
pv. campestris (9)
pv. raphani A (3)
pv.incanae (I)
pv.raphaniB(1)
pv. armoraceae (3)
pv. poinsettiicola B (3)
pv. fici B (2)
X. fragariae (9)
pv. vesicatoria B (11)
I I I I I I I I
0.035 0.030 0.025 0.020 0.015 0.010 0.005 0.0
Genetic distance


CHAPTER 4
CHARACTERIZATION OF PLANT PATHOGENIC Xanthomonas
BASED ON RESTRICTION ANALYSIS OF AMPLIFIED DNA
SEQUENCES RELATED TO THE hrp GENES
Identification of plant pathogenic xanthomonads has relied on isolating the
organism of interest in pure culture and performing predetermined biochemical,
serological, and pathological tests (Bradbury, 1984; Saettler et al., 1989; Schaad,
1988). More recently, methods based on metabolic fingerprinting (Chase et ah, 1992;
Hildebrand et ah, 1993; Jones et ah, 1993a; Van den Mooter and Swings, 1990;
Vemiere et ah, 1993), protein profiling (Vauterin et ah, 199lab), and fatty acid
composition (Chase et ah, 1992; Hodge et ah, 1992; Yang et ah, 1993) have also been
used for identification of these plant pathogens. Although these methods have proved
useful for the differentiation of certain plant pathogenic xanthomonads, the specific
identification of the xanthomonads at the subgeneric level is still difficult. Further, the
difficulty is particularly evident for the infrasubspecific identification of the members
of X campestris. This species includes at least 125 different pathovars which are
usually distinguished only on the basis of their host range or by the disease they cause
(Bradbury, 1984; Dye et al., 1980; Hayward, 1993). Therefore, there is a need for more
rapid and unambiguous procedures for identification and detection of these plant
pathogens.
In recent years, nucleic acid-based techniques have been used for
characterization of the xanthomonads. DNA-DNA hybridizations were applied to
57


Table 7-1 Continued
Strain Origin Reaction on Amplification Tentative identification Source
plants of hrp fragment
pepper tomato
hrpB
hrpC/D
hrp analysis
FAA
9310
tomato seedling
HR
+
+
+
pv. vesicatoria B
pv. manihotis (0.723)
DAM
9311
tomato seedling
HR
+
+
-
nd
pv. campestris (0.778)
DAM
140A-dl
tomato seed
HR
HR
+
+
pv. raphani
pv. campestris (0.897)
DAM
140A-d2
tomato seed
HR
HR
+
+
pv. raphani
pv. campestris (0.911)
DAM
T-93-23
tomato transplant
HR
+
+
+
pv. vesicatoria B
pv. alfalfae (0.707)
DAM
724-4
tomato seed
HR
HR
+
-
nd
pv. vesicatoria (0.816)
DAM
P93-22
pepper seedling
-
-
-
-
na
pv. pruni (0.799)
GOK
T93-09
tomato seedling
HR
HR
+
+
pv. raphani
pv. campestris (0.715)
GOK
T93-12A
tomato seedling
-
-
-
-
na
pv. zinnae (0.856)
GOK
VSE069
tomato seed
HR
+
+
+
pv. vesicatoria B
pv. vesicatoria (0.698)
MD
VSE070
tomato seed
HR
+
+
+
pv. vesicatoria B
pv. vesicatoria (0.620)
MD
VSE071
tomato seed
HR
+
+
+
pv. vesicatoria B
pv. vesicatoria (0.718)
MD
BSA1
pepper seed
HR
HR
+
+
nd
pv. corylina (0.628)
JW
BSA2
pepper seed
HR
HR
+
+
nd
pv. corylina (0.619)
JW
BSA3
pepper seed
HR
HR
(+)
(+)
nd
pv. pruni (0.671)
JW
BSA4
pepper seed
HR
HR
+
+
pv. citrumelo
pv. fici (0.765)
JW
BSA5
pepper seed
HR
-
-
-
na
pv. cannae (0.598)
JW
BSA6
pepper seed
HR
-
-
-
na
pv. cannae (0.586)
JW
BSA7
pepper seed
HR
HR
-
-
na
pv. pruni (0.832)
JW
BSA8
pepper seed
HR
HR
(+)
+
pv. carotae
pv carotae (0.730)
JW
Continued on the following page


Table 6-1. Genetic divergence matrix for 19 strains of Xanthomonas campestris based on the estimates of the number of nucleotide
substitution per site in two fragments related to the hrp gene cluster X. campestris pv. vesicatoria.
X. campestris pv. citrumelo3
X. campestris pv.
citrib
Other strains of X. campestris
High
Moderate
Weak
A
B
C
82-1
XCB
XCF
X22j
XI98
X151
75-3
FI
F6
F311
F348
534
F86
F94
F100
F306
9771
B84
339
FI
0.020
0.009
0.013
0.010
0.014
0.016
0.022
0.020
0.057
0.055
0.055
0.011
0.043
0.041
0.060
0.019
0.004
0.014
F6

0.009
0.015
0.019
0.012
0.012
0.018
0.010
0.057
0.063
0.063
0.007
0.056
0.057
0.052
0.015
0.018
0.018
F311

0.013
0.012
0.008
0.010
0.014
0.012
0.057
0.055
0.055
0.007
0.043
0.044
0.049
0.011
0.007
0.012
F348

0.010
0.010
0.014
0.016
0.016
0.056
0.061
0.061
0.011
0.045
0.046
0.058
0.017
0.009
0.018
534

0.011
0.017
0.017
0.015
0.051
0.050
0.050
0.014
0.036
0.037
0.046
0.018
0.010
0.015
F86

0.007
0.011
0.009
0.050
0.051
0.051
0.006
0.037
0.038
0.048
0.008
0.010
0.012
F94

0.005
0.002
0.057
0.055
0.055
0.012
0.046
0.044
0.052
0.012
0.014
0.005
F100

0.007
0.064
0.058
0.058
0.016
0.049
0.047
0.052
0.016
0.018
0.011
F306

0.057
0.055
0.055
0.014
0.046
0.044
0.052
0.014
0.016
0.023
9771

0.014
0.014
0.059
0.018
0.015
0.082
0.052
0.051
0.057
B84

0.000
0.061
0.012
0.011
0.080
0.051
0.049
0.055
339

0.061
0.012
0.011
0.080
0.051
0.049
0.055
82-1

0.048
0.049
0.065
0.011
0.011
0.018
XCB

0.002
0.074
0.039
0.038
0.046
XCF

0.075
0.040
0.038
0.044
X22j

0.051
0.060
0.055
XI98

0.017
0.018
X151
-
0.012
75-3

3 High, Moderate, and Weak denote aggressiveness groups.
b A, B, and C denote canker groups.


135
Ml 2 3 4 5 6 7 B 9 10 11 12 13
Fig. 6-1. Amplification of the 840-bp fragment of the complementation group B of the
hrp gene cluster of Xanthomonas campestris pv. vesicatoria from strains of X.
campestris. Lanes: M, phage X restricted with Eco RI and Hind III; 1 to 3 strains FI,
F6, and FI 00 of A c. pv. citrumelo, respectively; 4,X c. pv. maculifoliigardeniae
strain X22j; 5, X campestris from Strelitzia reginae XI98; 6,X. c. pv. fici strain X151;
1,X. c. pv. alfalfae strain 82-1; 8, X c. pv. bilvae strain XCB; 9, X campestris from
Feronia sp. strain XCF; 10 to 12, strains 9771, B84, and 339 of A! c. pv. citri,
respectively; \3,X. c. pv. vesicatoria strain 75-3. Molecular sizes are given in bases.


86
identical restriction patterns to each other though less than four strains were examined
for each of these pathovars (Table 4-2). In the case of the pathovars X. campestris pv.
gardneri and X. campestris pv. pelargonii, only the 1075-bp fragment was examined,
because the amplification of the 840-bp hrp-related fragment produced very low DNA
yield (Table 4-1). Uniform banding patterns were produced by the strains of the
pathovars of X. campestris pv. carotae, X campestris pv. phaseoli, and X campestris
pv. pruni, with the exception of only one single strain within each pathovar (Table 4-2).
In another study, nine strains of X. campestris pv. begoniae produced identical banding
patterns for the restriction analysis of the 840-bp hrp-related fragment by using either
endonuclease Cfol, HaelII, Sau3A\, or Taql (Fig. 4-10A). However, the restriction
analysis of the 1,075-bp fragment with the endonucleases Cfol or HaeIII revealed
variability in the banding pattern for strains X281 and X610 of campestris pv.
begoniae (Fig. 4-1 OB). The restriction analysis of this hrp-related fragments with the
endonucleases Sau3A\ and Taql revealed that banding patterns were identical for all
nine strains of this pathovar (Table 4-2).
In contrast, large variability was found in the restriction analysis of the hrp-
related fragments amplified from strains of several pathovars of X. campestris. This
includes the pathovars X campestris pv. citri, X campestris pv. citrumelo, X.
campestris pv. dieffenbachiae, X. campestris pv. fici, X campestris pv. poinsettiicola,
X. campestris pv. vesicatoria, and X. campestris pv. vitians (Fig. 4-11 and 4-12; Table
4-2). Strains of the citrus pathogen X. campestris pv. citrumelo were the most
heterogeneous and the 17 strains examined were divided into nine different restriction
profile groups based on the restriction analysis of the two hrp-related sequences
(Chapter 6; Table 4-2). On the other hand, strains of the pathovars X. campestris pv.
citri, X campestris pv. dieffenbachiae, and X campestris pv. vesicatoria produced


230
Ditta, G., Stanfield, S., Corbin, D., and Helinski, D. 1980. Broad host range DNA
cloning system for gram-negative bacteria: Construction of a gene bank of Rhizobium
meliloti. Proc. Natl. Acad. Sci. USA 77:7347-7351.
DosSantos, R. M. D., B., and Dianese, J. C. 1985. Comparative membrane
characterization of Xanthomonas campestris pv. cassavae and X. campestris pv.
manihotis. Phytopathology 75: 581-587.
Dye, D. W. 1962. The inadequacy of the usual determinative tests for the
identification of Xanthomonas spp.. N. Z. J. Sci. 5: 393-416.
Dye, D. W. 1966. Cutural and biochemical reactions of additional Xanthomonas spp.
N. Z.J. Sci. 9:913-919.
Dye, D. W., Bradbury, J. F., Goto, M., Hayward, A. C., Lelliot, R. A., and Schroth, M.
N. 1980. International standards for naming pathovars of phytopathogenic bacteria
and a list of pathovar names and pathotype strains. Rev. Plant Pathol. 59: 153-168.
Egel, D. S. 1991. Pathogenic and Genomic Characterization of Strains of
Xanthomonas campestris Causing Diseases of Citrus. Ph.D. dissertation. University of
Florida, Gainesville. 132 pp.
Egel, D. S., Graham, J. H., and Stall, R. E. 1991. Genomic relatedness of
Xanthomonas campestris strains causing diseases of citrus. Appl. Environ. Microbiol.
57: 2724-2730.
Elrod, R. P., and Braun, A. C. 1947. Serological studies of the genus Xanthomonas. I.
Cross-agglutination relationships. J. Bacteriol. 53:509-518.
El-Sharkawy, T. A., and Huisingh, D. 1971a. Electrophorethic analysis of esterases
and other soluble proteins from represntatives of phytopathogenic bacterial genera. J.
Gen. Microbiol. 68: 149-154.
El-Sharkawy, T. A., and Huisingh, D. 1971b. Differentiation among Xanthomonas
species by polyacrylamide gel electrophoresis of soluble proteins. J. Gen. Microbiol.
68: 155-165.
Feinberg, A. P., and Vogelstein, B. 1983. A technique for radiolabeling DNA
restriction endonuclease fragments to high specific activity. Anal. Biochem. 132:6-
13.


Table 7-3. Detection of plant pathogenic xanthomonads in naturally infected pepper and tomato seeds.
Seed sample
Total population of
bacteria on YNA
(CFU/g seed)
Recovery of
xanthomonads
Tween B Grow outc
medium
Amplification of hrp
fragment3
355 bp 840 bp
Source of the
seedb
Pepper
SP2.92
3.2 X 102
_d
-
+
+
GOK
SP66.92
1.9 X 103
-
-
+
+
GOK
SP 124.92
3.2 X 102
-
+
+
+
GOK
SP133.92
>2.4 X 105
-
+
+
+
GOK
SP135.92
1.6 X 102
-
-
-
+
GOK
SP306.92
0.8 X 102
-
+
+
+
GOK
Jupiter
4.0 X 102
-
-
-
-
RST
Greenhouse grown
5.4 X 102
-
-
(+)
(+)
RST
PK-l-PF
>4.0 X 103
-
-
-
+
JFW
PK-2-PF
>4.0 X 103
-
+
+
+
JFW
PL-3-P
>4.0 X 103
-
-
+
+
JFW
Tomato
#51-lb
4.0 X 103
-
-
-
-
MM
Diseased fruits
1.1 X 107
-
+
+
nd
RST
Marglobe #100016
4.6 X 103
-
-
-
nd
RST
Manalucie
1.0 X 103
-
-
-
nd
RST
Continued on the following page


181
can be a useful tool for the specific detection and identification of plant pathogenic
xanthomonads. X. campestris pv. vesicatoria was readily detected in preparations
containing cells added to pepper and tomato seed extracts and in seed washings obtained
from naturally contaminated pepper and tomato seeds. The sensitivity of the method of
DNA amplification ranged from about 102 to 103 CFU/ml of seed washings. This level
was 100 to 1000 times higher than the level obtained with ELISA, and it is certainly
comparable to the levels obtained with the most sensitive techniques available for
detection of bacteria (Pickup, 1991; Saettler et al., 1989). Furthermore, the specificity of
the method allowed the detection of plant pathogenic xanthomonads against a
background bacterial micro flora larger than 107 CFU/g of seed. This level of non target
bacterial population may have prevented the detection of the pathogen by plating in
general and semiselective media.
X. campestris pv. vesicatoria was detected in DNA extracted from seed washings
containing the reagents PVPP and sodium ascorbate and from preparations without these
reagents. Although the hrp fragments were amplified from DNA extracted from seed
samples without adding reagents to prevent potential inhibitors, the addition of PVPP and
sodium ascorbate in the seed washings before nucleic acid extraction may be needed to
assure the necessary purity of the DNA for amplification. The presence of inhibitors have
been a major concern in the extraction of high quality DNA from environmental and plant
samples (Holben et al., 1988; Minsavage et al., 1994; Steffan et al., 1988). Nevertheless,
the addition of these two reagents to the seed washings before DNA extraction apparently
did not interfere with the yield of DNA fragments obtained in the amplification of the hrp
fragments.
A basic feature of the approach for recovering bacterial DNA used here is the
initial separation of intact bacterial cells from the seed washings by centrifugation. This


96
dieffenbachiae also have a high degree of genetic variability. Strains of X. campestris
pv. dieffenbachiae produced very similar restriction patterns for all combinations of hrp
fragments-restriction endonucleases and they were divided in three different groups of
restriction banding profiles. Further, the strains were accurately identified to their host
of origin based on the restriction analysis of the hrp related fragments. The genetic
heterogeneity of the strains of X. campestris pv. dieffenbachiae agrees with the
diversity determined on the basis of pathological and physiological features, fatty acid
composition, and SDS-PAGE of proteins (Chase et al., 1992; Vauterin et al., 1991a).
Moreover, strains of X campestris pv. dieffenbachiae that cause disease on Syngonium
sp. have been designated A', campestris pv. syngonii (Dickey and Zumoff, 1987)
although strong support has not been obtained yet to place these strains into a different
pathovar (Chase et al., 1992).
In the restriction analysis of the hrp-related sequences, the most heterogeneous
pathovar was2f campestris pv. citrumelo. The seventeen strains of X campestris pv.
citrumelo were separated into nine different groups on the basis of the restriction
banding profiles (Table 4-2). However, the strains of the highly aggressive group ofX.
campestris pv. citrumelo were homogeneous and comprised a single group (Egel et al.,
1991). The high uniformity of the strains of the highly aggressive group ofX.
campestris pv. citrumelo has also been determined in studies of DNA homology, RFLP
analyses, fatty acid composition, and SDS-PAGE of proteins (Chapter 3; Egel et al.,
1991; Gabriel et al., 1988, 1989; Graham et al., 1990; Hartung and Civerolo, 1987,
1989; Vauterin et al., 1991b). On the other hand, the moderately and weakly
aggressive strains of X. campestris pv. citrumelo are very diverse, and they are likely to
determine the heterogeneity of this pathovar. The diverse nature of the strains of X.


TABLE OF CONTENTS
page
ACKNOWLEDGMENTS ii
ABSTRACT vii
CHAPTERS
1 INTRODUCTION 1
2 REVIEW OF LITERATURE 5
Serology 5
Fatty Acids Composition 10
Protein Profiles 12
Phenotypic Characteristics 17
Nucleic Acids Analysis 21
3 OLIGONUCLEOTIDE PRIMERS FOR DETECTION AND
IDENTIFICATION OF PLANT PATHOGENIC STRAINS OF
Xanthomonas BY AMPLIFICATION OF DNA SEQUENCES
RELATED TO THE hrp GENES OF Xanthomonas campes tris
PV. VESICATORIA 30
Material and Methods 33
Results 39
Discussion 51
4 CHARACTERIZATION OF PLANT PATHOGENIC
Xanthomonas BASED ON RESTRICTION ANALYSIS OF
AMPLIFIED DNA SEQUENCES RELATED TO THE hrp
GENES 57
Material and Methods 61
Results 64
Discussion 91
iv


Table D-lContinued
Species/Pathovar Species/Pathovar
32
33
34
35
36
37
38
39
40
41
1. pv. alfalfae
.2727
.1667
.3182
.5217
.4889
.5306
.4000
.3556
.4681
.5833
2. pv. alfalfae
.2727
.1667
.3182
.5217
.4889
.5306
.4000
.3556
.4681
.5833
3. pv. armoraceae
.4000
.3182
1.000
.3333
.3415
.2667
.8780
.9268
.1860
.3636
4. pv. begoniae A
.2000
.1818
.2500
.8095
.3415
.8444
.3415
.2927
.7442
.3636
5. pv. begoniae A
.2381
.1739
.3333
.9091
.3721
.8936
.4186
.3721
.8000
.3913
6. pv. begonia B
.2857
.2174
.3810
.9091
.3721
.8511
.4651
.4186
.7556
.3913
7. pv. bilvae
.3636
.1250
.3182
.4348
.7111
.4082
.4000
.4000
.3830
.7917
8. pv. campestris
.4000
.3182
1.000
.3333
.3415
.2667
.8780
.9268
.1860
.3636
9. pv. carotae A
.4500
.3182
.8500
.3333
.4390
.3111
.8780
.9268
.2326
.4545
10. pv. carotae B
.4000
.3182
1.000
.3333
.3415
.2667
.8780
.9268
.1860
.3636
11. pv. citri A
.2273
.2500
.3636
.3478
.8000
.4082
.4444
.4444
.2979
1.000
12. pv. citri B
.3182
.1667
.3182
.2609
.6667
.2449
.4000
.4000
.2128
.7500
13. pv. citri C
.3182
.1667
.3182
.2609
.6667
.2449
.4000
.4000
.2128
.7500
14. pv. citrumelo
.3111
.1633
.3111
.4255
.4348
.4800
.3478
.3478
.5417
.4898
15. pv. citrumelo
.2667
.1633
.3111
.5532
.5217
.5600
.3913
.3478
.5000
.6122
16. pv. citrumelo
.2273
.1667
.2727
.4783
.4444
.5306
.3111
.3111
.4681
.5833
17. pv. citrumelo
.2273
.1667
.2727
.4783
.4444
.5306
.3111
.3111
.4681
.5833
18. pv. citrumelo
.2273
.1667
.2727
.5652
.4889
.5714
.3556
.3111
.5106
.5833
19. pv. citrumelo
.2326
.1702
.2791
.5333
.4545
.5833
.3636
.3182
.4783
.5957
20. pv. citrumelo
.2727
.1667
.2727
.5217
.4000
.4898
.3111
.3111
.4681
.4583
21. pv. citrumelo
.3182
.2083
.3182
.5652
.4444
.6122
.4000
.3556
.5106
.5833
22. pv. citrumelo
.2727
.1667
.2727
.5652
.4444
.6531
.3556
.3111
.5957
.5417
Continued on following page
209


132
1989.). After being stained with 0.5 jag of ethidium bromide per ml, the gel was
photographed over a UV transilluminator (Fotodyne Inc., New Berlin, WI) with type
55 Polaroid film (Polaroid, Cambridge, MA).
Hybridization analysis
The identity of the amplified DNA fragments was further confirmed by
hybridization analysis with an internal DNA probe for each fragment. Samples were
electrophoresed in 0.7% agarose gel according to standard procedure (Sambrook et al.,
1989.). The gel was then denatured in 0.4 N NaOH and 0.6 M NaCl for 30 min and
neutralized for 30 min in 0.5 M Tris-Cl and 1.5 M NaCl. The denatured DNA was
transferred by the procedure of Southern (1975) to nylon membrane (Gene Screen Plus,
Du Pont, Boston, MA). Hybridization was carried out at 68C with 0.5X SSC, and
0.1% w/v SDS. The internal probes consisted of a 271-bp insert of the plasmid
pXV840 for the 840-bp fragment and a 335-bp insert of the plasmid pXV1075 for the
1,075-bp fragment (Chapter 3; Appendix A). Probes were labeled by the random
primed (Feinberg and Vogelstein, 1983) incorporation of digoxigenin-labeled
deoxyuridine-triphosphate (DIG-UTP) and detected by the use of the Genius
Nonradioactive DNA Labeling and Detection kit (Boehringer Mannheim) as specified
by the manufacturer.
Restriction endonuclease analysis of amplified DNA
Amplified DNAs were restricted with either endonuclease Cfol, Haelll,
Sau3Al, or Taql under conditions specified by the manufacturer (Promega). The
restriction fragments were separated by electrophoresis in 4% agarose gels (3%
NuSieve GTG and 1% Seakem GTG [FMC BioProducts, Rockland, ME]) in TAE


Table D-lContinued
Species/Pathovar Species/Pathovar
53
54
55
56
57
58
59
60
61
1. pv. alfalfae
.4583
.5000
.4898
.4889
.3636
.9167
.7083
.3256
.2083
2. pv. alfalfae
.4583
.5000
.4898
.4889
.3636
.8750
.6667
.3256
.2083
3. pv. armoraceae
.5000
.3182
.2667
.3415
.8500
.2727
.3182
.4615
.4545
4. pv. begoniae A
.4091
.1818
.8000
.3415
.3500
.5909
.3636
.2564
.0909
5. pv. begoniae A
.4348
.2174
.8511
.3721
.3333
.5652
.3913
.2927
.1739
6. pv. begonia B
.4348
.2609
.8085
.3721
.3810
.5652
.3913
.3415
.2174
7. pv. bilvae
.4167
.7083
.4082
.7111
.3636
.7083
.9583
.3721
.2500
8. pv. campestris
.5000
.3182
.2667
.3415
.8500
.2727
.3182
.4615
.4545
9. pv. carotae A
.5909
.4091
.3111
.4390
.9000
.3182
.4091
.5128
.3636
10. pv. carotae B
.5000
.3182
.2667
.3415
.8500
.2727
.3182
.4615
.4545
11. pv. citri A
.5000
.7500
.4082
.8000
.4091
.5833
.8333
.2791
.2917
12. pv. citri B
.4167
1.000
.2857
.6667
.3636
.4583
.7083
.3256
.3333
13. pv. citri C
.4167
1.000
.2857
.6667
.3636
.4583
.7083
.3256
.3333
14. pv. citrumelo
.3265
.5306
.5600
.4348
.3556
.7347
.6531
.3182
.2449
15. pv. citrumelo
.4490
.5306
.5200
.5217
.3556
.8980
.7347
.3182
.2041
16. pv. citrumelo
.3750
.4583
.6122
.4444
.3636
.9167
.6667
.2791
.1667
17. pv. citrumelo
.3750
.4583
.6122
.4444
.3636
.9167
.6667
.2791
.1667
18. pv. citrumelo
.4167
.4583
.5306
.4889
.3636
1.000
.7083
.2791
.1667
19. pv. citrumelo
.4255
.4255
.5417
.4545
.3721
.9362
.6383
.2857
.1702
20. pv. citrumelo
.3333
.5833
.5306
.4000
.3636
.7917
.6250
.3256
.2083
21. pv. citrumelo
.4583
.5000
.5714
.4444
.3636
.7917
.6667
.3721
.2500
22. pv. citrumelo
.4167
.4583
.6122
.4444
.3182
.7917
.6667
.3256
.2083
Continued on following page
214


The population structure of some taxa, i.e. X. jragariae, X. campestris pv.
begoniae, X campestris pv. campestris, X. campestris pv. malvacearum, and X.
campestris pv. pelargonii, was homogeneous. However, the majority of the pathovars
of X campestris presented a high degree of sequence variability in the /^-related
fragments for the strains within pathovars. For instance, the citrus pathogens X.
campestris pv. citri strains in groups A, B, and C, which cause citrus canker A, B, and
C, respectively, and X. campestris pv. citrumelo strains in the highly aggressive group,
which cause the citrus bacterial spot, had a characteristic and homogeneous restriction
banding pattern within each group. In contrast, restriction fragment polymorphism was
evident among strains of the moderately and weakly aggressive groups of X. campestris
pv. citrumelo.
The phylogenetic analysis of the hrp genes revealed a diverse evolutionary
relationship for this region of the bacterial genome of the xanthomonads. Whereas the
hypothesis of coevolution of the hrp region with the rest of the genome from a common
ancestor is supported, there is also evidence of horizontal movement of the hrp genes
between the different plant pathogenic xanthomonads. Nevertheless, the results
obtained do not substantiate the existence of genetic and functional selection pressure
regarding the hrp genes within the xanthomonads. One of the major findings of this
study is the divergence between hrp gene sequence and host specificity, which
indicates that factors other than the hrp genes are likely to be involved in host
speciation.
Strains of plant pathogenic xanthomonads could be detected by amplification of
/irp-related DNA sequences from washings of pepper and tomato seeds. The diversity
of the xanthomonads found on seeds assessed on the basis of restriction analysis of
/irp-related sequences agreed very closely with the population structure determined by
viii


171
M 1 2 3 4 5 6
1 3
5
Fig. 7-3. Amplification of the 840-bp fragment of the hrp gene cluster of Xanthomonas
campestris pv. vesicatoria 75-3 from seeds of tomato by using sodium ascorbate and
PVPP (lanes 1, 2, and 3) and without these reagents (lanes 4, 5, and 6). Lanes: M,
phage X restricted with £coRI and Hindlll; 1 and 4, 5 pi of DNA extract; 2 and 5, 10 pi
of DNA extract; 3 and 6, 15 pi of DNA extract. Molecular sizes are given in bases.


98
fici is genetically closely related to group A of X. campestris pv. poinsettiicola whereas
the B group ofX. campestris pv. fici is also genetically closely related to group B ofX.
campestris pv. poinsettiicola. The diversity within the pathovars X. campestris pv. fici
and X. campestris pv. poinsettiicola has also been supported by fatty acid analysis
(Nancy C. Hodge, personal communication).
The genetic diversity of strains within pathovars of X. campestris is not
surprising. The classification of plant pathogenic bacteria at pathovar level was not
based initially on the genetics or other intrinsic characteristics of the organism but
rather on the host from which the bacteria were isolated (Bradbury, 1984; Dye et al.,
1980). Furthermore, comprehensive studies on the genetics and phenotypic
characteristics have supported the existence of a high degree of diversity among strains
of a given group of plant pathogenic xanthomonads that cause diseases in the same
host. Therefore, the plant pathogenic xanthomonads comprise a very complex group of
bacteria that might not be easily distinguished. The genetic analysis of the hrp-related
sequences seems to be a very useful tool for differentiation of pathovars and pathogenic
groups of plant pathogenic xanthomonads. Furthermore, the diversity or uniformity of
the different taxa of xanthomonads assessed on the basis of restriction analysis of hrp-
related sequences apparently agrees very closely with the existing groups established
by using other methods. However, the restriction banding profiles generated for the
/^-related fragments may be an easier and more discriminating approach for
identification of plant pathogenic xanthomonads, compared to other methods such as
genomic fingerprinting or RFLP analysis by using random or specific DNA probes.
DNA fingerprinting by digestion of the entire bacterial genome with restriction
endonuclease usually produces very complex patterns that are difficult to interpret
(Graham and Cooksey, 1989; Hartung and Civerolo, 1987; Vauterin et al., 1993)


33
(PCR). Furthermore, the reliability of identification of the xanthomonads to subgeneric
classification was assessed by restriction analysis of amplified DNA fragments.
Materials and Methods
Bacterial strains, plasmids, and culture conditions
The bacterial strains and plasmids used in this study and their sources are listed
in Appendix A. The identity of bacterial strains used here was confirmed by fatty acid
analysis (Nancy C. Hodge, personal communication). All strains of Acidovorax
avenae, Agrobacterium tumefaciens, Erwinia spp., Clavibacter michiganense,
Pseudomonas spp., X. campes tris, and X. maltophilia were grown on nutrient agar
(Becton Dickinson, Cockeysville, MD). Nutrient broth cultures were grown 24 hours
on a rotatory shaker (150 rpm) at 28C. Strains of A albilineans andX. fragariae were
cultivated on Wilbrink's medium (Koike, 1965). Strains of Xylella fastidiosa were
grown on PW medium (Davis et al., 1981). Strains of Escherichia coli were cultivated
on Luria-Bertani medium at 37C (Miller, 1972). All strains were stored in sterile tap
water at room temperature or in 30% glycerol at -70C, or both. Antibiotics were used
to maintain selection for resistance markers at the following final concentrations:
ampicillin, 100 pg/ml; tetracycline, 10 pg/ml; rifampicin, 100 pg/ml; and
spectinomycin, 50 pg/ml.
Plant material and plant inoculations
All plants were maintained in a growth chamber at 28-30C during inoculation
and incubation. The pepper cultivar Early Calwonder (ECW) and the near-isogenic
lines ECW-1 OR, ECW-20R, and ECW-30R have been described elsewhere (Minsavage


136
M 1 2 3 4 5 6 7 8 9 10 11 12 13
Fig. 6-2. Amplification of the 1,075-bp fragment of the complementation groups C/D
of the hrp gene cluster of Xanthomonas campestris pv. vesicatoria from strains of A
campestris. Lanes: M, phage X restricted with Eco RI and Hind III; 1 to 3 strains FI,
F6, and FI00 of A c. pv. citrumelo, respectively; 4,X c. pv. maculifoliigardeniae
strain X22j; 5, X. campestris from Strelitzia reginae X198; 6, X. c. pv. fici strain X151;
7,X. c. pv. alfalfae strain 82-1; 8, X. c. pv. bilvae strain XCB; 9, X. campestris from
Feronia sp. strain XCF; 10 to 12, strains 9771, B84, and 339 of A c. pv. citri,
respectively; 13, A! c. pv. vesicatoria strain 75-3. Molecular sizes are given in bases.


72
M 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Fig. 4-5. Restriction profiles established for the 1,075-bp hrp-related fragment
amplified from different strains of plant pathogenic Xanthomonas spp. and restricted
with the endonuclease Cfo\. Lane M, phage X restricted with Pstl. Molecular sizes are
given in bases. See Table 4-2 for identification of strains of Xanthomonas spp. for each
restriction pattern.


34
et al., 1990). These lines provided a susceptible reaction or a hypersensitive reaction,
depending on the strain used.
Fully expanded leaves of plants were inoculated with bacterial suspensions by
infiltrating the bacteria into the intercellular spaces with a 1 ml plastic syringe with a
27 gauge needle. The concentration of the inoculum was approximately 5 X 108
colony forming units (CFU) per milliliter in sterile tap water, determined by measuring
the optical density in a Spectronic 20 spectrophotometer (Bausch and Lomb, Inc.,
Rochester, NY). Plant reactions were scored over a period of several days.
DNA manipulations
Total genomic DNA was isolated by phenol extraction and ethanol precipitation
essentially as described by Ausubel et al. (1987). Plasmid mini-prep, preparation of
competent cells, ligation, and transformation of E. coli cells were performed by
standard procedures (Ausubel et al., 1987; Sambrook et al., 1989). For
complementation analysis, the helper plasmid pRK2013 was used in triparental mating
to mobilize pLAFR3 clones from E. coli into Xanthomonas cells (Ditta et al., 1980;
Figurski and Helinski, 1979).
Hybridization analysis
Total genomic DNA and amplified DNA fragments were electrophoresed in
0.7% agarose according to standard procedures (Sambrook et al., 1989). The DNA was
then denatured in 0.4 N NaOH and 0.6 M NaCl for 30 min, neutralized for 30 min in
0.5 M Tris-Cl and 1.5 M NaCl, pH 7.5, and transferred by the procedure of Southern
(1975) to nylon membrane (Schleicher & Schuell, Keene, NH). Southern hybridization
and detection of the hybridized DNA was carried out by using the Genius


95
campestris pv. malvacearum, and X campestris pv. pelargonii. The population
structure of X. campestris pv. begoniae and X. campestris pv. pelargonii has been
examined quite extensively by using different methods, such as DNA-DNA
hybridization, fatty acid composition, and SDS-PAGE of proteins. Both pathovars
seem to consist of uniform groups of strains with stable phenotypic features and
characteristic fatty acid and protein profiles (Vauterin et al., 1990a; Yang et ah, 1993).
In the same way, X campestris pv. malvacearum also seems to be a fairly
homogeneous group of strains on the basis of SDS-PAGE of proteins and fatty acid
analysis (Vauterin et ah, 1991a; Yang et ah, 1993), despite the fact that this pathovar
comprises a large number of physiological races (Brinkerhoff, 1970). Although the
restriction analyses of the hrp-related fragments also suggest for a clonal population
structure in other pathovars of X. campestris, i.e. X. campestris pv. armoraciae, X.
campestris pv. gardneri, X. campestris pv. pruni, and X campestris pv. vignicola, the
limited number of strains tested does not allow us to make conclusions with certainty.
In contrast, a striking feature of the different pathovars of X. campestris was the
high degree of genetic variability in the hrp-related sequences for strains within
pathovars. For example, strains ofX. campestris pv. citri were separated in two
homogeneous groups on the basis of the restriction endonuclease analysis of the hrp-
related sequences. One group comprises all strains that corresponds to the citrus canker
A of.Y campestris pv. citri whereas the other group includes strains of the citrus canker
B and C forms. Although the citrus canker groups were established on the basis of
pathogenic specialization of the strains (Leite, 1990; Stall and Civerolo, 1991) more
extensive studies on the genetic and phenotypic characteristics also support this
grouping of the strains ofX. campestris pv. citri (Egel et al., 1991; Gabriel et al., 1988,
1989; Graham et al., 1990; Hartung and Civerolo, 1987). Strains ofV campestris pv.


162
C. Somodi, personal communication). Plates were then blocked with 1% bovine
albumin (Sigma, A-9647) in PBS and shaken to remove excess liquid. Aliquots of 100
pi of each sample in EDTA/lysozyme lysis buffer were added to three wells on the
microtiter plate and incubated. The microtiter plates were rinsed with NTrinse and a
monoclonal antibody (2H10) prepared against X. campestris pv. vesicatoria 75-3 was
added to all wells and the plates were incubated. The microtiter plates were rinsed
again with NT rinse, and alkaline phosphatase conjugated goat antimouse A-1047
(Sigma) was applied to the plate. The microtiter plates were incubated and then rinsed
with the final washing buffer Tris-buffered saline (1.51 g of Tris-base, 2.19 g NaCl,
final volume 250 ml with deionized water, pH 7.5) four times. Substrate and amplifier
were added according to the instructions specified by the manufacturer of the ELISA
Amplification System (Gibco BRL, Gaithersburg, MD). Readings were made at A492
15 min after addition of the amplifier with an EAR400 AT plate reader (SLT
Labinstruments, Austria).
Fatty acid analysis
The fatty acid profiles of the strains of X. campestris isolated from pepper and
tomato seeds and seedlings were determined. A single colony of each strain was
transferred from a nutrient agar culture to trypticase soy broth agar (TSBA).
Approximately 40 mg of cells of a 24 h growth at 28C was collected for analysis. The
bacterial fatty acids were derivatized to their methyl esters (Miller, 1982) and separated
by gas chromatography (Sasser, 1990) and identified with the Microbial Identification
System software (version 3.80; MIDI, Newark, DE). The similarity index and best
match for the strains based on fatty acid analysis data were determined by using the
MIDI software.


Table D-2Continued
Species/Pathovar
Species/Pathovar
37
40
41
42
43
44
45
46
47
48
25. pv. dieffenbachiae B
.5714
.6047
.5333
.4783
.8511
.3415
.8511
.8511
.5333
.5778
26. pv. fici A
.6977
.7273
.5652
.5532
1.000
.4762
1.000
1.000
.5652
.5217
27. pv. fici A
.6190
.6512
.4889
.4783
.9362
.4390
.9362
.9362
.4889
.4444
28 pv. fici A
.6667
.6977
.5333
.5217
.9362
.4390
.9362
.9362
.5333
.4889
29. pv. fici B
.5789
.6154
.4390
.3810
.5116
.8649
.5116
.5116
.6829
.6341
31. pv. glycines A
.5366
.5714
1.000
.8000
.5652
.4000
.5652
.5652
.5455
.5455
34. pv. incanae
.5366
.5714
.5455
.5333
.5652
.6000
.5652
.5652
1.000
.9545
35. pv. maculifoliigardeniae
.8947
.9231
.5854
.5714
.6977
.5405
.6977
.6977
.5854
.5366
36. pv. malvacearum
.5500
.5854
.9302
.7727
.5778
.4103
.5778
.5778
.5116
.5116
37. pv. manihotis
.9231
.5366
.5714
.6977
.5405
.6977
.6977
.5366
.4878
40. pv. phaseoli A
.5714
.6047
.7273
.5789
.7273
.7273
.5714
.5238
41. pv. phaseoli B
.8000
.5652
.4000
.5652
.5652
.5455
.5455
42. pv. phaseoli "fiiscans"
.5532
.3415
.5532
.5532
.5333
.5333
43. pv. physalidicola
.4762
1.000
1.000
.5652
.5217
44. pv.poinsettiicola B
.
.4762
.4762
.6000
.5500
45. pv. poinsettiicola A
.
1.000
.5652
.5217
46. pv. pruni
.
.
.5652
.5217
47. pv. raphani A
.
.
.
.9545
48. pv. raphani B




Continued on following page


Table 3-1Continued
45
Strain Southern hybridization DNA amplification
355 bp
840 bp
1,075 bp
355 bp
840 bp
1,075 bp
X. albilineans 91 -065
-
-
-
-
-
-
X. fragariae XI297
+
+
+
-
+
+
X. maltophilia
-
-
-
-
-
-
Acidovorax avenae
subsp. avenae UK142-A
-
-
-
-
-
-
subsp. citrulli UK20
-
-
-
-
-
-
Agrobacterium tumefaciens
LBA 1050
-
-
-
-
-
-
Clavibacter michiganense
subsp. michiganense 69-1
-
-
-
-
-
-
Erwinia
E. carotovora subsp.
carotovora
K-SR-347
-
-
-
-
-
-
B-SR-38
-
-
-
-
-
-
E. herbicola NF-33
-
-
-
-
-
-
E. stewartii SW2
-
-
-
-
-
-
Pseudomonas
P. solanacearum K60
-
-
-
-
-
-
P. syringae
pv. syringae INB
-
-
-
-
-
-
pv. tomato 987
-
-
-
-
-
-
Xylella fastidiosa 89-1
-
-
-
-
-
-
a+, positive reaction; negative reaction; (+), weak signal.


Fig. 5-1. Unrooted phylogenetic tree inferred from restriction analysis data of the
1,075-bp DNA fragment related to the hrpC/D complementation group of the hrp genes
of Xanthomonas campestris pv. vesicatoria generated by the Wagner parsimony
criterion. The values on each node indicate the levels of support derived from 100
bootstrapped trees. The shaded boxes delineate the major clades identified in the
analysis of the hrpC/D-related DNA fragment. Also included are the number of strains
examined for each taxa (numbers in parenthesis).


whereas RFLP analysis using DNA probes requires hybridization techniques. More
extensive work is necessary, however, to characterize the hrp sequence variation in
other groups of plant pathogenic xanthomonads. Furthermore, DNA-DNA
hybridization data is also necessary to determine the consistency of the grouping
established based on such a small region of the bacterial genome as the hrp gene
cluster. The restriction data may also be useful to establish the genetic evolutionary
relationship of the hrp genes among the different plant pathogenic xanthomonads.


91
patterns between X. campestris pv. fici group A and X. campestris pv. poinsettiicola
group A, and between X. campestris pv. fici group B and X. campestris pv.
poinsettiicola group B (Fig. 4-12; Table 4-2). Further, the restriction pattern for both
fragments amplified from strain X125 of A campestris pv. fici group A were identical
to the ones of the strains of X. campestris pv. poinsettiicola group A for all
combinations of hrp-related fragments and restriction endonucleases (Fig. 4-12 and
Table 4-2).
Discussion
Several methods have been examined for the differentiation of plant pathogenic
xanthomonads with different degrees of success. In the present study, the technique of
enzymatic amplification and analysis of specific regions of the bacterial genome was
evaluated for the differentiation and identification of plant pathogenic xanthomonads.
Unlike other studies where random and less stable regions of the bacterial genome were
used (Garde and Bender, 1991; Gilbertson et al., 1989; Hartung, 1992; Hartung et ah,
1993; Lazo and Gabriel, 1987; Lazo et ah, 1987), I analyzed sequences of the bacterial
genome related to the hrp gene cluster of X campestris pv. vesicatoria. Although this
region of the bacterial chromosome seems to be highly conserved among the plant
pathogenic xanthomonads, nonpathogenic xanthomonads lack similarity to the hrp
genes (Bonas et ah, 1991; Stall and Minsavage, 1990). This is certainly a major
advantage, because the nonpathogenic xanthomonads is of concern for plant health
inspection in certification programs (Gitaitis et ah, 1987, 1992). From the results
obtained in the present study the DNA amplification of /zr/?-related sequences is highly
promising for specific differentiation and identification of a large group of plant


240
Schaad, N. W., Azad, H., Peet, R. C., and Panopoulos, N. J. 1989. Identification of
Pseudomonas syringae pv. phaseolicola by a DNA hybridization probe.
Phytopathology 79: 903-907.
Schaad, N. W., and Stall, R. E. 1988. Xanthomonas. Pages 81-94 in: Laboratory
Guide for Identification of Plant Pathogenic Bacteria. N. W. Schaad, ed. American
Phytopathological Society Press, St. Paul, MN.
Schoulties, C. L., Civerolo, E. L., Miller, J. W., Stall, R. E., Krass, C. J., Poe, S. R.,
and DuCharme, E. P. 1987. Citrus canker in Florida. Plant Dis. 71:388-395.
Schulte, R., and Bonas, U. 1992. Expression of the Xanthomonas campestris pv.
vesicatoria hrp gene cluster, which determines pathogenicity and hypersensitivity on
pepper and tomato, is plant inducible. J. Bacteriol. 174: 815-823.
Scott, J. W., and Jones, J. B. 1989. Inheritance of resistance to foliar bacterial spot of
tomato incited by Xanthomonas campestris pv. vesicatoria. J. Amer. Soc. for Hort.
Sci. 114: 111-114.
Seal, S. E., Jackson, L. A., and Daniels, M. J. 1992. Isolation of a Pseudomonas
solanacearum specific DNA probe by subtraction hybridization and construction of
species-specific oligonucleotide primers for sensitive detection by the polymerase
chain reaction. Appl. Environ. Microbiol. 58: 3751-3758.
Sharon, E., Okon, Y., Bashan, Y., and Henis, Y. 1982. Detached leaf enrichment: A
method for detecting small numbers of Pseudomonas syringae pv. tomato and
Xanthomonas campestris pv. vesicatoria in seed and symptomless leaves of tomato and
pepper. J. Appl. Bacteriol. 53:371-377.
Sijam, K., Chang, C. J., and Gitaitis, R. D. 1991. An agar medium for the isolation
and identification of Xanthomonas campestris pv. vesicatoria from seed.
Phytopathology 81: 831-834.
Sneath, P. H. A. 1989. Analysis and interpretation of sequence data for bacterial
systematics: The view of a numerical taxonomist. Syst. Appl. Microbiol. 12: 15-31.
Somodi, G. C., Jones, J. B., and Scott, J. W. 1993. An ELISA procedure developed
for detection of low populations of Xanthomonas campestris pv. vesicatoria (Xcv) in
tomato leaves. (Abstr.) Phytopathology 83:1394
Southern, E. 1975. Detection of specific sequences among DNA fragments separated
by gel electrophoresis. J. Mol. Biol. 98:503-517.


8
Monoclonal antibodies (MABs) have been examined quite extensively for
specific identification of plant pathogenic xanthomonads. MABs have been produced
for various groups of these plant pathogens, including X albilineans, X. oryzae, and
several pathovars of X. campestris (Vauterin et al., 1993). Further, MABs have been
developed for identification of xanthomonads at different taxonomic levels (Alvarez et
al., 1985; Benedict et al., 1989, 1990). Alvarez et al. (1985) obtained a MAB that
identifies the genus Xanthomonas. This MAB was specific for all 436 xanthomonads
tested, but did not react with bacteria of 12 other genera. MABs were produced that
were specific for X oryzae pv. oryzae and for X. oryzae pv. oryzicola at pathovar level
(Benedict et al., 1989). These pathovar-specific antibodies reacted with all tested
strains of each pathovar, but neither of the MABs reacted with strains of other
pathovars and species, or with strains of other bacterial genera. Also, for MABs
specific for epitopes on the lipopolysaccharide of X. campestris pv. begoniae and X.
campestris pv. pelargonii were produced that were pathovar-specific and reacted only
with their respective strains (Benedict et al., 1990). These monoclonal antibodies did
not react with strains of other xanthomonads or non-xanthomonads tested.
Contrary to the pathovar-specific MABs mentioned above, no pathovar specific
MABs have been found that could be used to specifically detect all strains within
certain pathovars of X campestris, such as X. campestris pv. campestris (Alvarez et al.,
1985; Franken et al., 1992), X. campestris pv. citri (Alvarez et al., 1991), X. campestris
pv. citrumelo (Alvarez et al., 1991; Gottwald et al., 1991; Permar and Gottwald, 1989),
and X campestris pv. dieffenbachiae (Lipp et al., 1992). Based on the reactions of six
antibodies, 200 strains of A! campestris pv. campestris were grouped in six serogroups
(Alvarez et al., 1991). However, this grouping did not imply closer genetic
relationships among members of a group than with members of other groups or


29
developed for specific and sensitive detection of the citrus canker pathogen based on
the DNA sequence of these plasmid probes (Hartung et al., 1993).


197
Table A-lContinued
Pathovar
Strain
Source3
pv. taraxaci
XT11A
DCH
pv. vesicatoria
RES
group A
75-3, 82-4, 85-10, 87-21, 87-35T, 87-44T, 87-
48T, 89-8, 89-10, 90-20, 90-21, 90-27, 90-40,
90-60,91-66, 91-72, 92-11, 92-15, 92-16, 92-
17, 92-119, 1712, 6107, XV14, XV17
group B
141, 0226A, 0350A, 695, 853, 1062, B-3,B-
20, BA28-1, BV5-5, XV56
RES
group C
92-118, 92-120
RES
pv. vignicola
81-30, 82-38, G-55
RES
pv. vitians
ICPB101, ICPB164, XVIT, XV2, XV3,
X1215
RES
undetermined and
isolated from Feronia
XCF
ELC
sp.
undetermined and
isolated from Hibiscus
sp.
X10, X27, X52,
DWG
undetermined and
isolated from Strelitzia
reginae
X198
ARC
undetermined and
opportunistic
1,2, 2-8, 4, 7, 8, 9, 10, 11, 12, 17, 661, 663,
AAI, CB, Danny Gay, INA, INA42, INA69,
RG-1, S52, T-55, T-56, Toad Flax, TP78
RES
X. albilineans
91-065
JCC
X. fragariae
GC-6259, GC-6265, X1238, X1241, X1244,
XI246, XI292, X1298, X1426
ARC
X. maltophilia
RES
Continued on the following page


172
12 3 4
8 9 101112 13 14 15 16171819 20
Fig. 7-4. Amplification of the 355-bp (Lanes 2 to 10) and 840-bp (Lanes 12 to 20)
fragments of the complementation group B of the hrp gene cluster from samples of
tomato seeds containing different concentration of Xanthomonas campestris pv.
vesicatoria 75-3 added to the seed washings. Lanes: 1 and 11, phage X restricted with
EcoRl and Hindlll; 2 and 12, 2.6 X 10* CFU/ml; 3 and 13,2.9 X 107 CFU/ml; 4 and
14, 0.9 X 106 CFU/ml; 5 and 15, 3.4 X 105 CFU/ml; 6 and 16,1.0 X 104 CFU/ml; 7
and 17, 3.3 X 103, CFU/ml; 8 and 18, 3.0 X 102 CFU/ml; 9 and 19, 4.0 X 101 CFU/ml;
10 and 20, no bacteria added. Concentration ofX. campestris pv. vesicatoria in the
seed washings was determined by plating on Tween B medium. Molecular sizes are
given in bases.


12
the fatty acid profile, the grouping obtained by fatty acid analysis does not agree
entirely with the grouping obtained by using other methods, such as by genetic analysis
and phenotypic features. Strains of X. campestris pv. citri that cause the citrus canker
disease can be clearly differentiated from strains ofX. campestris pv. citrumelo, the
citrus bacterial spot pathogen (Vauterin et al., 1991b; Yang et al., 1993). However,
there are some discrepancies between the delineation of the citrus canker groups by
fatty acid profile compared to genetic analysis and protein profile (Vauterin et al.,
1991b). Based on fatty acid analysis, the strains of citrus canker B were
indistinguishable from strains of the citrus canker A, but they were distinct from the
strains of groups C and D of citrus canker (Vauterin et al., 1991b). However, the citrus
canker A strains are considered a homogeneous and distinct group from the strains of
citrus canker B, C, and D based on DNA-DNA hybridization, RFLP analysis, and
protein profiles (Egel et al., 1991; Gabriel et al., 1989; Hartung and Civerolo, 1987;
Vauterin et al., 1991b). Contrasting results on fatty acid analysis were also obtained
for the citrus bacterial pathogen, X. campestris pv. citrumelo. Vauterin et al. (1991b)
found that strains within X. campestris pv. citrumelo constituted a distinct and
homogeneous group based on fatty acid composition. On the other hand, Graham et al.
(1990) observed considerable diversity among the X. campestris pv. citrumelo strains,
and some of these strains were closely related to non citrus strains of X. campestris on
the basis of fatty acid composition.
Protein Profiles
El-Sharkawy and Huisingh (197lab) were the first to demonstrate the
usefulness of electrophoresis of proteins for differentiation and identification of
Xanthomonas. The patterns of native negatively charged proteins extracted from whole


151
were due to nucleotide substitutions and not to insertion or deletion of DNA sequences.
In fact, this assumption is supported by the fact that no apparent length variation was
observed in the two DNA fragments amplified from all strains ofX. campestris. In
support of the phylogenetic analysis presented here is the monophyletic nature of the
hrp gene cluster of bacterial pathogens with different genetic backgrounds causing
disease in different hosts. The phylogenetic grouping presented here also correlates
with genetic analyses based on DNA-DNA hybridization, fingerprinting, and
conventional restriction fragment length polymorphism (Egel, 1991; Egel et al., 1991;
Gabriel et al., 1988, 1989; Graham et al., 1990; Hartung and Civerolo; 1987, 1989;
Kubicek et al., 1989). The data presented supports the concept of a group of causal
microorganisms of citrus bacterial spot disease which are closely related yet represent a
variety of different genotypes.
Two different groups ofX. campestris pv. citri were distinguished by restriction
enzyme analysis of the amplified fragments of the hrp gene cluster. The strains of
group A produced identical patterns for all fragment-endonuclease combinations.
Similarly, all strains within the groups B and C also produced identical restriction
banding patterns. The banding patterns for group A strains were different from the
patterns of group B and C strains, however. This substantiates other reports of the
relative genetic uniformity of the strains of the A, B, and C groups of X campestris pv.
citri based on restriction analysis of the entire genome by using genomic fingerprinting,
pulse-field electrophoresis, or restriction fragment length polymorphisms with random
DNA probes (Egel, 1991; Egel et al., 1991; Gabriel et al., 1988, 1989; Graham et al.,
1990; Hartung and Civerolo; 1987). The two groups of X campestris pv. citri have
about 60% DNA homology (Egel et al., 1991). The major difference from previous
work is that in our studies specific homologous regions of the bacterial genome, the hrp


50
Fig. 3-6. Restriction analysis of the 1,075-bp DNA fragment of the hrp gene cluster
amplified from strains of Xanthomonas campestris and restricted with the
endonucleases HaelU (A) and Sau3Al (B). Lanes: M, phage X restricted with Pstl; 1,
X. campestris pv. vesicatoria 75-3; 2,X. campestris pv. bilvae XCB; 3, X campestris
pv. carotae #13; 4, A! campestris pv. citri 9771; 5,X. campestris pv. citrumelo FI; 6, X.
campestris pv. dieffenbachiae 729; 7, X campestris from Strelitzia reginae strain
X198; 8, X. campestris pv. fici X151; 9, X campestris pv. maculifoliigardeniae X22j;
10, X. campestris pv. manihotis Xml25D; \ \,X. campestris pv. pelargonii XCP58; 12,
X. campestris pv. phaseoli "fuscans" XP163A; 13, X campestris pv. poinsettiicola 071-
424; 14, X. campestris pv. taraxaci XT11A; 15, A campestris pv. vignicola 81-30; 16,
X. campestris pv. vitians XVIT; 17, X. campestris pv. papavericola XP5; 18, X.
campestris pv. holcicola G-23. Molecular sizes are given in base pairs.


113
rd
tj
c1
£
pv. cilri C (6)
pv. citri B (7)
pv. vignicola (3)
pv. phaseoli Tuscans" (1)
i pv. malvacearum (9)
pv. vilians B (I)
pv. phaseoli B (1)
pv. glycines A (6)
pv. cilri A (6)
a. campestris XCF(I)
pv. bilvae (1)
pv. vesicatoria A
pv. vesicatoria A
pv. citmmelo (I)
pv. alfalfac I
pv. alfalfac (I)
pv. citruinelo (2)
pv. pruni (4)
pv. citrumclo (I)
pv. citrumclo (2)
pv. citrumelo (2)
pv. poinsettiicola A (2)
_ pv. lici
I
(24)
r
pv. fici A (I j
X. campestris XI98 (I)
.pv, citrumelo(I)
_Ppv. citrumelo (IJ
LpV. vesicatoria
pv. fici A (I)
pv. fici A (I)
pv. citrumelo (6)
pv. citrumelo (I)
pv. dielTenbachiae B (2)
pv. maculifblgardemae (I)
pv. begoniae (I)
pv. begoniae (I)
pv. begoniae (7)
pv. dieffenbachiae A (4)
pv. dielTenbachiae A (4)
pv. phaseoli A (4)
pv. vitians A (3)
pv. manihotis (I)
pv.incanae(1)
pv. armoraceae (3)
pv. campestris (9)
pv. carotae B (1)
-pv. pelargonii (7)
pv. papavericola (I)
I f pv. gardneri (3)
[ *-pv. vitians C (2)
j pv. carotae A (7)
' pv. taraxaci (J)
pv. vesicatoria B (11)
pv. raphani A (3)
pv. raphani B (1)
ipv. poinsettiicola B
A-, campestris
ipv. glycines B (1
Hpv. fici B (2)
_'V fragariae (9)
pv. holcicoia (I)
X52
I I I I I I I I I
0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0.00
Genetic distance


84
these pathovars revealed very similar banding patterns (Table 4-2). All strains of the
latter pathovars belong to the restriction banding profiles 3 and 2 or 3 of the
endonucleases Taql and Sau3AI, respectively (Table 4-2). The restriction banding
patterns from the strains of these pathovars for the endonucleases Cfol and HaeIII are
also very similar and they have several DNA bands in common (Fig. 4-5 and 4-6).
Although the other pathovars of X. campestris also showed some trends in regard to the
similarity of the restriction patterns for both /i/y?-related fragments, a more
discriminating analysis by using some clustering approach may be more appropriate to
establish the precise delineation of the groupings.
Variability of the fap-related restriction patterns among strains within different taxa of
Xanthomonas spp. and pathovars of X campestris
Sequence variation in the hrp region was revealed by restriction analysis of
both /zrp-related fragments amplified from strains within the same taxa of plant
pathogenic xanthomonads. Further, this variation indicated that these plant pathogens
may differ in the genetic variability of the DNA sequence of the hrp gene cluster.
Whereas some taxa have very uniform restriction patterns for both hrp-related
sequences, others seem to be comprised of very distinct groups of strains. For instance,
the restriction banding patterns for all nine X fragariae strains, originally isolated from
diseased material in Florida, California, and Canada (Ann R. Chase, personal
communication), were identical to each other when both /irp-related fragments were
restricted with either Cfol, Hae III, Sau3Al, or Taql (Fig. 4-9; Table 4-2). Similarly,
uniform banding patterns were observed for strains within the pathovars X campestris
pv. campestris, X campestris pv. malvacearum, X. campestris pv. pelargonii, and X
campestris pv. raphani (Table 4-2). Strains within the pathovars X campestris pv.
armoraciae, X. campestris pv. gardneri, and X. campestris pv. vignicola also produced


130
al. (1990). All strains were streaked onto nutrient agar (Becton Dickinson,
Cockeysville, MD), and single colonies were selected. Nutrient broth cultures were
grown 24 hours on a rotatory shaker (150 rpm) at 28C. Strains ofX. campestris pv.
citri group B were grown on a sucrose based medium (Canteros de Echenique et al.,
1985). Strains were stored on lima bean agar (Difco, Detroit, MI) for short term
storage and in sterile tap water at room temperature for long term storage.
DNA isolation
The procedure described by Ausubel et al. (1987), with minor modifications,
was used to extract total genomic DNA. Briefly, bacterial cells were pelleted by
centrifuging in an Eppendorf microcentrifuge (Brinkmann Instruments Inc., Westbury,
NY) for 2 min at 16,000 g. The pellet was washed in 1 ml of distilled water, pelleted
again and resuspended in 567 pi of TE buffer (10 mM Tris-Cl, pH 8.0; 1 mM EDTA,
pH 8.0). Proteinase K (Boehringer Mannheim, Indianapolis, IN) and sodium dodecyl
sulfate (SDS) (Sigma, St. Louis, MO) were added for a final concentration of 100 p
g/ml and 0.5%, respectively. After incubation for 1 hour at 37C, sodium chloride and
hexadecyltrimethyl ammonium bromide (Sigma) were added to each preparation for a
final concentration of 0.7 M and 1%, respectively. The preparations were incubated for
10 min at 65C. DNA was extracted with chloroform-isoamyl alcohol (24:1). The
samples were hand shaken continuously and gently for 10 min and centrifuged for 5
min at 16,000 g. A second extraction was accomplished by adding phenol-chloroform-
isoamyl alcohol (25:24:1) and centrifuging as described above. DNA was precipitated
by adding 0.6 volumes of isopropanol and incubating for 30 min at -20C. The
samples were centrifuged for 20 min at 16,000 g. The DNA pellet obtained was
washed with 1 ml of 70% ethanol and centrifuged again. The DNA was dried under


41
functions of these genes are unknown. The hrpDl gene is highly similar to a sequence
of pathogenicity genes in A campestris pv. glycines (Ulla Bonas, personal
communication; Hwang et al., 1992).
The identity of the amplified fragments was further confirmed by restriction
enzyme analysis. The 355-, 840-, and 1,075-bp fragments amplified from both total
DNA of strain 75-3 of X. campestris pv. vesicatoria and from DNA of pXV9, pXV5.1,
and pXV5.5 were digested with Cfol, Haelll, Sow3AI, and Taq\. The banding patterns
were identical for each of the three sets of fragments amplified from strain 75-3 and
plasmids containing cloned parts of the hrp region (Fig. 3-3) and matched the
restriction map generated from the DNA sequence of the hrp gene cluster of X.
campestris pv. vesicatoria 75-3 (Ulla Bonas, personal communication).
Similarity of the hrp fragments amplified from X campestris pv. vesicatoria to DNA of
other bacteria
Total genomic DNA of strains of different pathovars of X. campestris and of
Xanthomonas spp. was digested with EcoRI, separated in agarose gels, blotted, and
probed with each of the three fragments amplified from DNA of X. campestris pv.
vesicatoria strain 75-3. The hybridization signals in the genomic DNA of A
campestris pv. vesicatoria 75-3 corresponded to the predicted 7.4 kb BamHl and 5.5 kb
EcoRI fragments (355- and 840-bp probes), and to the 6.0 kb BamiRl and 5.1 kb EcoRI
fragments (1,075-bp probe) (Fig. 3-4). Homology to these three hrp fragments oiX.
campestris pv. vesicatoria strain 75-3 was detected in strains of X fragariae and of 28
different pathovars of A! campestris (Fig. 3-4; Table 3-1). However, polymorphisms
were observed in the DNA from these different pathovars (Fig. 3-4). Total genomic
DNA of A albilineans, A campestris pv. secalis, and A campestris pv. translucens,
which are pathogens of monocotyledonous plants, as well as DNA of the non plant


39
[FMC BioProducts, Rockland, ME]) in TAE buffer at 8 V/cm. Phage X Pstl restricted
DNA fragments were used as molecular standards. The gel was stained with 0.5 pg of
ethidium bromide per ml for 40 min and then destained in 1 mM MgSC>4 for 1 hr and
photographed over a UV transilluminator with type 55 Polaroid film.
Results
Specificity of the oligonucleotide primers to the hrp gene cluster
Three pairs of oligonucleotide primers were tested for amplification of
fragments from genomic DNA of strain 75-3 of X. campestris pv. vesicatoria and from
plasmids that contain cloned parts of the hrp region from strain 75-3. Plasmid pXV9
harbors a fragment of approximately 27 kb of strain 75-3 of X. campestris pv.
vesicatoria containing almost the entire hrp region (Bonas et al., 1991). Plasmid
pXV5.5 harbors a 5.5 kb EcoRI fragment containing part of the hrp complementation
groups hrpA and hrpB (Fig. 3-1). The 5.1 kb EcoRI insert of plasmid pXV5.1 maps to
the complementation groups hrpC and hrpD (Fig. 3-1). We amplified fragments of the
expected 355, 840, and 1,075 bp in length from total genomic DNA of strain 75-3 of A!
campestris pv. vesicatoria by using primers RST9 and RST10, RST2 and RST3, and
RST21 and RST22, respectively (Fig. 3-2). The 355- and 840-bp fragments were also
amplified from plasmids pXV9 and pXV5.5 (Fig. 3-2), whereas the 1,075-bp fragment
was amplified from plasmids pXV9 and pXV5.1 (Fig. 3-2), thus confirming the sizes
and locations of fragments predicted by the DNA sequence analysis (Ulla Bonas,
personal communication). The 355- and 840-bp sequences correspond to genes in the
hrpB operon, whereas the 1075 bp fragment occurs in the hrpC/hrpD region (Fig. 3-1).
Except for hrpB6, which encoded a putative ATPase (Fenselau et al., 1992), the


54
that these fragments were amplified from DNA sequences which also control the
pathogenicity in other xanthomonads.
In contrast to the narrow spectrum of oligonucleotide primers previously used
for detection and identification of only certain strains of X. campestris (Hartung et al.,
1993), the hrp specific primers RST2 plus RST3 and RST21 plus RST22 seem very
useful for the identification of a large range of plant pathogenic xanthomonads. This is
perhaps not surprising, because the hrp region seems very conserved among different
plant pathogenic xanthomonads as determined by Southern hybridization experiments
in the present and previous studies (Bonas et ah, 1991; Stall and Minsavage, 1990).
Furthermore, the nucleotide sequences of the primers RST21 and RST22 are identical
to corresponding sequences of pathogenicity genes of X. campestris pv. glycines
(Hwang et ah, 1992). In addition, the cloned fragment from X. campestris pv. glycines
complements hrpD mutants of X. campestris pv. vesicatoria (Ulla Bonas, personal
communication), suggesting functional homology between these regions from both
pathovars.
Primers RST9 and RST10, which delineate a fragment of 355 bp, allowed DNA
amplification only from a limited number of pathovars of X. campestris, despite
hybridization of the fragment to the majority of the strains of this species. It should be
noted that the sequence of RST9 originates from hrpB6, a gene for a putative ATPase
that seems to be highly conserved among different bacteria at the protein sequence
level (Fenselau et ah, 1992). These results indicate differences in the DNA sequences
of Xanthomonas spp. corresponding to one or both primers used. However, this set of
primers seems useful for specific detection of strains of X. campestris pv. vesicatoria
group A, X. campestris pv. fici, X. campestris pv. physalidicola, and X campestris
X198.


2
1987; Maas et al., 1985; Liao and Wells, 1987; Stall and Minsavage, 1990). These
opportunistic bacteria can be identified as xanthomonads by the presence of
xanthomonadins and by unique fatty acid profiles.
Attempts have been made to differentiate the pathovars and strains of A
campestris with several techniques. Traditional methods for the identification and
characterization of plant pathogenic xanthomonads rely upon performing
predetermined biochemical, serological, and pathological tests on pure cultures of the
bacteria (Bradbury, 1984; Holt et al., 1994; Schaad, 1988). More recently, methods
based on metabolic and protein profiles (Chase et al., 1992; Hildebrand et al., 1993;
Van den Mooter and Swings, 1990; Vauterin et al., 1991ab), and fatty acid analysis
(Chase et al., 1992; Yang et al., 1993) have been used for taxonomic and identification
purposes. Polyclonal and monoclonal antisera produced against strains ofX.
campestris have been used for detection and identification, but they have provided
variable results (Alvarez and Lou, 1982; Alvarez et al., 1991; Benedict et al., 1990;
Jones et al., 1993b). Nucleic acid based techniques have become a premier approach
for detection and identification of plant pathogenic bacteria (Bereswill et al., 1992;
Manulis et al., 1991; Schaad et al., 1989; Seal et al., 1992), including members of the
xanthomonads (Berthier et al., 1993; Garde and Bender, 1991; Gilbertson et al., 1989;
Hartung, 1992; Hartung et al., 1993; Hildebrand et al., 1990; Lazo and Gabriel, 1987;
Lazo et al., 1987; Leach et al., 1990,1992).
Genetic relationships among plant pathogenic xanthomonads have been
established based on DNA-DNA homology, total genomic fingerprinting, and analysis
of restriction fragment length polymorphism (RFLP) in their DNAs (Vauterin et al.,
1990a). DNA-DNA hybridization studies have shown great variation in the relatedness
ofX. campestris pathovars (Hildebrand et al., 1990; Palleroni et al., 1993), and the


38
(Applied Biosystems, Foster City, CA) by the DNA Synthesis Laboratory, University
of Florida, Gainesville.
DNA was amplified in a total volume of 50 pi. The reaction mixture contained
5 pi of 10X buffer (500 mM KC1,100 mM Tris Cl [pH 9.0 at 25C], 1% Triton X-
100), 1.5 mM MgCl2, 200 pM of each deoxynucleotide triphosphate (Boehringer
Mannheim), 25 pmol of each primer, and 1.25 units of Taq polymerase (Promega,
Madison, WI). The amount of template DNA added was 100 ng of purified total
bacterial DNA or 25 ng of a plasmid preparation, unless otherwise stated. The reaction
mixture was covered with 50 pi of light mineral oil. A total of 30 amplification cycles
were performed in an automated thermocycler PT-100-60 (MJ Research, Watertown,
MA). Each cycle consisted of 30 s of denaturation at 95C, 30 s of annealing at 62C,
and 45 s of extension at 72C for the primers RST2 and RST3; 30 s at 95C, 30 s at
52C and 45 s at 72C for the primers RST9 and RST10; and 30 s at 95C, 40 s at 61C
and 45 s at 72C, for the primers RST21 and RST22. The last extension step was
extended to 5 min. Amplified DNAs were detected by electrophoresis in 0.9% agarose
gels in TAE buffer (40 mM Tris acetate, 1 mM EDTA, pH 8.2) at 5 V/cm of gel
(Sambrook et al., 1989). The gel was stained with 0.5 pg of ethidium bromide per ml
and then photographed over a UV transilluminator (Fotodyne Inc., New Berlin, WI)
with type 55 Polaroid film (Polaroid, Cambridge, MA).
Restriction endonuclease analysis
The DNA fragments amplified from different bacterial strains were restricted
with the frequent-cutting endonucleases Cfol, HaeIII, Sau3Al, or Taql, according to
conditions specified by the manufacturer (Promega). The restricted fragments were
separated by electrophoresis in 4% agarose gels (3% NuSieve and 1% Seakem GTG


Table 4-2Continued
Species/Pathovar
Strain
840 bp
1,075 bp
Taq\
Saw3AI Haelll
Cfol
Taql
Saw3AI Haelll
Cfol
pv. dieffenbachiae A
X422, XI51, X790, X1272
2
7
12
10
9
15
22
18
X729, X736, X738, X745
2
7
12
10
9
14
22
18
B
X260, X763
3
6
13
11
10
14
23
15
pv. fici A
X125
3
6
16
12
8
11
15
8
X151
7
6
16
12
8
11
18
11
X212
5
6
16
12
8
11
18
11
B
X208, X702
1
4
4
5
5
7
8
5
pv. gardneri
XG101, XV6, 1066
nd
nd
nd
nd
3
3
4
2
pv. glycines A
G-56, 86-16, 86-17, 86-18, 86-
20, 87-2
2
5
6
7
7
9
10
7
B
1706A
nd
nd
nd
nd
5
7
8
5
pv. holcicola
G-23
nd
nd
nd
nd
11
16
25
23
pv. incanae
9561-1
1
3
3
3
3
4
3
2
pv.
maculifoliigardeniae
X22j
2
7
10
10
9
14
24
20
Continued on following page


15
hybridization, DNA fingerprinting, RFLP analyses, and fatty acid composition (Egel et
al., 1991; Gabriel et al., 1987; Graham et al., 1990; Hartung and Civerolo, 1987;
Vauterin et ah, 1991b). The strains of the citrus canker B and D groups ofX.
campestris pv. citri constituted one group, and strains of citrus canker C formed
another distinct group on the basis of protein profile (Vauterin et ah, 1991b). However,
the groupings based on DNA-DNA hybridization, fatty acid composition, and RFLP
analysis do not entirely agree with the groups determined by protein analysis (Vauterin
et ah, 1991b). Although distinct from the citrus canker A strains ofX. campestris pv.
citri, strains of X. campestris pv. citrumelo also showed a quite uniform protein
patterns on the SDS-PAGE (Vauterin et ah, 1991b). This contrasts with the diversity
determined for the strains of this pathovar based on fatty acid composition (Graham et
ah, 1990), RFLP analyses (Gabriel et ah, 1989; Hartung and Civerolo, 1987), and
DNA-DNA hybridization (Egel et ah, 1991).
SDS-PAGE of whole-cell proteins was also used in comparative studies of the
xanthomonads that cause diseases on cassava, X. campestris pv. manihotis and A!
campestris pv. cassavae (Van den Mooter et ah, 1987a). Strains of X. campestris pv.
manihotis constituted a distinct and homogeneous group on the basis of protein
profiles, whereas the strains of X. campestris pv. cassavae formed two groups. These
results also agree with the groupings obtained by numerical analyses of phenotypic
features (Van den Mooter et ah, 1987a), DNA-DNA hybridization (Van den Mooter et
ah, 1987a), and electrophoresis of total membrane proteins (Dos Santos and Dianese,
1985).
X. campestris pv. vesicatoria, the cause of the bacterial spot disease of tomato
and pepper, comprises heterogeneous groups in relation to protein profiles (Bouzar et
ah, 1994; Vauterin et ah, 1991a). This diversity has also been confirmed by genetic


139
M 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Fig. 6-4. Restriction analysis of the 1,075-bp DNA fragment of the complementation
groups C/D of the hrp gene cluster amplified from strains of Xanthomonas campestris
pv. citrumelo and restricted with the endonucleases (A) HaelW and (B) Cfol.. Lanes:
M, phage X restricted with the Pst I; 1 to 5, highly aggressive strains FI, F54, F274,
F361, and 3166, respectively; 6 to 11, moderately aggressive strains F6, F228, F311,
F254, F348, and F378; 12 to 16, weakly aggressive strains F59, F86, F100, F306, and
F94; \7,X. c. pv. vesicatoria strain 75-3. Molecular sizes are given in bases.


60
of plant pathogenic xanthomonads. The hrp genes are required by plant pathogens to
cause disease on susceptible hosts and hypersensitive reaction (HR) on resistant, or on
nonhost plants (Willis et al., 1991). The hrp genes have been found in different plant
pathogenic bacteria, including Erwinia amylovora (Beer et al., 1991), Pseudomonas
solanacearum (Boucher et al., 1987), P. syringae pv. phaseolicola (Lindgren et al.,
1986), and X. campestris pv. vesicatoria (Bonas et al., 1991). Furthermore, the hrp
gene sequence seems to be highly conserved among different pathogenic bacteria at the
functional level (Ulla Bonas, personal communication; Fenselau et al., 1992; Gough et
al., 1992; Hwang et al., 1992). On the contrary, nonpathogenic bacteria are unable to
cause disease or HR on plants and they apparently lack DNA sequences similar to the
hrp genes (Lindgren et al., 1986; Stall and Minsavage, 1990). In previous studies, the
hrp-genes sequences were useful for differentiating pathogenic from nonpathogenic
xanthomonads (Stall and Minsavage, 1990), as well as for differentiating strains from
different pathovars of X. campestris and related plant pathogenic Xanthomonas spp.
(Chapter 3).
The purpose of this study was to characterize the plant pathogenic
xanthomonads by restriction analysis of DNA sequences related to the hrp genes of X.
campestris pv. vesicatoria. I used oligonucleotide primers specific for the hrp genes to
amplify hrp-related DNA fragments from different plant pathogenic xanthomonads by
the polymerase chain reaction. Further, the reliability of the identification of the
xanthomonads was assessed by examining different groups of plant pathogenic
xanthomonads that comprise the species X. albilineans, X. fragariae, and at least 31
different pathovars of X. campestris.


Table 5-2. Similarity values between clades of plant pathogenic Xanthomonas spp. generated based on the restriction
profiles of the DNA fragments related to the hrpB and hrpC/D complementation groups of the hrp genes oiX. campestris pv.
vesicatoria.
Clade3
1
2
3
4
5
6
7
8
9
10
1
0.25b
0.50-0.62
0.41
0.32-0.42
0.42-0.50
0.33-0.46
0.25-0.44
0.46
0.38
2
0.40

0.36-0.46
0.32
0.33-0.36
0.25-0.33
0.17-0.25
0.08-0.22
0.14
0.21
3
0.54-0.60
0.52-0.57

0.41-0.50
0.39-0.51
0.24-0.46
0.23-0.40
0.18-0.47
0.25-0.40
0.27-0.32
4
0.54-0.60
0.52-0.57
0.96-1.00

0.40-0.51
0.29-0.46
0.27-0.41
0.27-0.43
0.40
0.23
5
0.43-0.53
0.62-0.63
0.55-0.68
0.55-0.68

0.23-0.37
0.23-0.37
0.14-0.34
0.20-0.26
0.18-0.23
6
0.47-0.56
0.37-0.48
0.51-0.58
0.51-0.58
0.34-0.52

0.40-0.74
0.18-0.44
0.32-0.50
0.13-0.18
7
0.38-0.51
0.24-0.41
0.41-0.57
0.41-0.57
0.39-0.56
0.41-0.94

0.36-0.67
0.49-0.74
0.13-0.21
8
0.53-0.59
0.50-0.56
0.49-0.59
0.49-0.59
0.54-0.62
0.54-0.79
0.59-0.77

0.50-0.75
0.23
9
0.52
0.28
0.53-0.58
0.53-0.58
0.34-0.43
0.48-0.78
0.65-0.85
0.25-0.62

0.13-0.22
10
ndc
nd
nd
nd
nd
nd
nd
nd
nd

aClade established based on the restriction profile on the basis of the phylogenetic analysis of the restriction fragment data of
the 1,075-bp fragment related to the hrpC/D region of the hrp genes of X. campestris pv. vesicatoria, restricted with the endonuclease Taql.
^Values are the similarities estimates by using the equation proposed by Nei and Li (1979) for the 1,075-bp DNA fragment related to
the hrpC/D, upper triangular matrix, and for the 840 bp DNA fragment related to the hrpB, lower triangular matrix, restricted with either
endonuclease Cfol, HaeIII, Saw3AI, and Taql.
cnd, not determined.


Abstract of Dissertation Presented to the Graduate School of the University of Florida
in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
GENETIC AND EVOLUTIONARY CHARACTERIZATION OF
PLANT PATHOGENIC XANTHOMONADS BASED ON DNA
SEQUENCES RELATED TO THE hrp GENES
By
Rui Pereira Leite, Jr.
August 1994
Chairman: Robert E. Stall
Major Department: Plant Pathology
Three pairs of oligonucleotide primers specific for different regions of the
hypersensitive reaction and pathogenicity (hrp) genes of Xanthomonas campestris pv.
vesicatoria were designed and used for amplification of homologous DNA sequences
from X fragariae and from 28 pathovars of X campestris. No amplification occurred
with genomic DNA from plant pathogenic strains of X. campestris pv. secalis, X.
campestris pv. translucens, and X albilineans, from nonpathogenic xanthomonads, or
from plant pathogenic strains of the genera Acidovorax, Agrobacterium, Clavibacter,
Erwinia, Pseudomonas, and Xylella. DNA fragments amplified with a particular
primer pair were of identical size from each of the different plant pathogenic
xanthomonads. However, the restriction analysis of these fragments by using frequent-
cutting endonucleases revealed variation in the banding pattern of these hrp-related
fragments. The banding patterns allowed distinction of the strains representing a
pathovar or species of plant pathogenic xanthomonads.
vii


66
Table 4-1.Continued
Species/Pathovar
No. of strains
tested
No. of strains with positive
amplification of /zrp-related fragment
840 bp 1,075 bp
pv. translucens
1
0
0
pv. vesicatoria
group A
27
27
27
group B
11
11
11
pv. vignicola
3
3
3
pv. vitians
6
6(2)
6
undetermined and
1
1
1
isolated from Feronia
sp.
undetermined and
3
1
3
isolated from Hibiscus
sp.
undetermined and
1
1
1
isolated from Strelitzia
reginae
undetermined and
25
0
0
nonpathogenic
X. albilineans
1
0
0
X. fragariae
9
9
9
aNumber in brackets indicates number of strains which produced weak signal.


155
for the use of such resistance commercially (Pohronezny et al., 1992). The causal
bacterium may spread from diseased to healthy plants by wind- blown water, clipping
of plants, and aerosols (Gitaitis et al., 1992; Jones et al., 1991; Volcani, 1969).
Furthermore, the bacterium may overwinter on tomato volunteers and diseased plant
debris (Jones et al., 1986). Nevertheless, exclusion of the pathogen from pepper and
tomato growing areas is still one of the main control measures. Consequently, the use
of pathogen free seeds and transplants has become an important part of the strategy for
control of bacterial spot (Gitaitis et al., 1992; Jones et al., 1991).
X. campestris pv. vesicatoria has been reported to be in association with tomato
and pepper seeds (Bashan et al., 1982; Gardner and Kendrick, 1921, 1923; Jones et al.,
1986; Sharon et al., 1982), and symptoms of bacterial spot were observed in seedlings
from contaminated seed lots (Gardner and Kendrick, 1923; Higgis, 1922). However,
the detection and identification of X. campestris pv. vesicatoria associated with seeds is
still a problem with the methods available. The semiselective media developed
specifically for X. campestris pv. vesicatoria (Gitaitis et al., 1991; McGuire et al., 1986;
Sijam et al., 1991) may still support the growth of strains of other pathovars of ^
campestris as well as nonpathogenic xanthomonads and saprophytes associated with
seeds. These bacteria may have cultural characteristics similar to X. campestris pv.
vesicatoria (Gitaitis et al., 1987, 1991; McGuire and Jones, 1989; McGuire et al., 1986;
Sijam et al., 1991). Therefore, additional biological, biochemical, or physiological
tests are required to determine with certainty the identity of the xanthomonads
recovered (Gitaitis et al., 1991; McGuire and Jones, 1989). Serological tests with
monoclonal and polyclonal antibodies have also been examined for specific
identification of plant pathogenic xanthomonads, including X. campestris pv.
vesicatoria (Benedict et al., 1990; O'Brien et al., 1967). However, the heterogeneous


70
Fig. 4-3. Restriction profiles established for the 840-bp hrp-related fragment amplified
from different strains of plant pathogenic Xanthomonas spp. and restricted with the
endonuclease Sau3AI. Lane M, phage X restricted with Pstl. Molecular sizes are given
in bases. See Table 4-2 for identification of strains of Xanthomonas spp. for each
restriction pattern.