Citation
The epidemiology, specific detection, and genetic variability of Xanthomonas fragariae

Material Information

Title:
The epidemiology, specific detection, and genetic variability of Xanthomonas fragariae
Creator:
Roberts, Pamela D., 1963-
Publication Date:
Language:
English
Physical Description:
vii, 76 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Bacteria ( jstor )
Diseases ( jstor )
DNA ( jstor )
Fatty acids ( jstor )
Plants ( jstor )
Polymerase chain reaction ( jstor )
Stall ( jstor )
Strawberries ( jstor )
Symptomatology ( jstor )
Xanthomonas ( jstor )
Dissertations, Academic -- Plant Pathology -- UF
Plant Pathology thesis, Ph. D
Strawberries -- Diseases and pests ( lcsh )
Xanthomonas -- Control ( lcsh )
Xanthomonas -- Genetics ( lcsh )
Xanthomonas diseases ( lcsh )
City of Madison ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1996.
Bibliography:
Includes bibliographical references (leaves 70-75).
Additional Physical Form:
Also available online.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Pamela D. Roberts.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Pamela D. Roberts. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
35128644 ( OCLC )
023268694 ( ALEPH )

Downloads

This item has the following downloads:


Full Text











THE EPIDEMIOLOGY, SPECIFIC DETECTION, AND GENETIC VARIABILITY OF XANTHOMONAS FRAGARIAE











BY

PAMELA D. ROBERTS













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 1996















To my husband, Kenn,
Cindy,
and
my parents.














ACKNOWLEDGMENTS


The author would like to gratefully acknowledge the members of the committee:

Dr. R. D. Berger, Dr. R. E. Stall, Dr. C. K. Chandler, and Dr. Jeffrey B. Jones. The

enthusiatic support, expert advise, and guidance throughout the duration of this project were greatly appreciated. Special thanks go to Dr. Jeffrey B. Jones who provided his laboratory for a majority of the work. Thanks go to Dr. A. R. Chase who initially suggested the project and provided considerable support the first year.

Special thanks go to the members of Dr. Stall's and Dr. Jones' labs, including N. Cheri Hodges, G. V. Minsavage, Rui P. Leite, Jr., Hacene Bouzar, Gail Somodi, Jeanette Chun, and Rick Kelly, for their help and information.

Very special thanks go to the people at Dover-G.C.R.E.C. who helped me tremendously with all aspects of the field research and taught me all about the production of strawberries: Alicia Whidden, Anne Turgeau, Jim Sumler, Dale Wenzel, Larry Smith, and Mitch Boles.

Finally, thanks go to the folks who helped make the experience enjoyable: Gary Marlow, Pia D. Gavino, Morgan Wallace, Pamela Lopez, and Marion Bogart.








iii















TABLE OF CONTENTS






ACKNOWLEDGMENTS.......................... .. ........ iii

ABSTRACT .......................... ................... ................. ....................... vi

CHAPTERS

1 IN T R O D U C T IO N .................................................................................. 1

2 DISEASE PROGRESS, YIELD LOSS, AND CONTROL OF
XANTHOMONAS FRA GARIAE ON STRAWBERRY
PLAN T S ...................................................................... ............. 8

Intro d u ctio n ............... ...................................... ... ............................. 8
Materials and Methods ................ ..................... 10
R esults ......................................................... .................. 13
Discussion .............................................................................. .......... 22

3 DEVELOPMENT OF A SPECIFIC AND SENSITIVE DETECTION
TECHNIQUE BY THE POLYMERASE CHAIN REACTION
FOR XANTHOMONAS FRAGARIAE AND APPLICATION IN A STUDY OF SURVIVAL OF THE BACTERIUM ON
STRAWBERRY PLANTS.................. ............. ..... 25


Introduction .................................................... ................................. ... 25
M aterials and M ethods........................................................................... 28
Results ................ .. ......... ................... .............. ............... .... 33
D iscussion ................................... .................................. . . ...... ............ . 4 1






IV












4 GENOMIC RELATEDNESS OF XANTHOMONAS FRAGARIAE
ON STRAWBERRY BY FATTY ACID METHYL
ESTERASE AND RESTRICTION LENGTH FRAGMENT
POLYMORPHISM ANALYSES ........................................... 45

Introduction............................ .............................................. 45
M aterials and M ethods .......................................................................... 47
Results................ .................. ....... 51
D iscussion ............................................................................. ....... ... 56

5 DISCUSSION....................................... ............... 60

LIST OF REFERENCES................................................. 70

BIOGRAPHICAL SKETCH.................................. .... ........ 76









Abstract of 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

THE EPIDEMIOLOGY, SPECIFIC DETECTION, AND GENETIC VARIABILITY OF XANTHOMONAS FRA GARIAE


By

Pamela D. Roberts

May 1996


Chairman: Dr. R. D. Berger
Major Department: Plant Pathology


Angular leaf spot (ALS) on strawberry is caused by the bacterium Xanthomonas fragariae (XF). The epidemiology of ALS, development of a detection technique by the polymerase chain reaction (PCR), and the genetic variability of strains were examined. The disease severity of ALS on plants in field plots increased to 25% and 15% in two seasons. Yield was decreased 8% to 10%. Minimal spread of ALS occurred between field plots. Chemical sprays applied at the label rate of cupric hydroxide plus mancozeb at 7- to 14- day intervals decreased disease but was phytotoxic and decreased yield. A 10% rate of the mixture applied frequently slightly reduced disease but increased yield one season and significantly reduced disease and did not affect yield the next season. Three primers were specific for amplification of DNA from XF but not DNA from strains of 16 pathovars of Xanthomonas campestris or non-pathogenic xanthomonads from strawberry. Bacteria were detected at 10' colony forming units per ml by a single round of PCR. A nested PCR



vi









technique increased detection 1,000 fold. Plants were inoculated with a rifampicin-resistant strain and oversummered in the field at two locations in Florida. Bacteria from leaf and crown samples were detected by nested PCR and recovery onto selective media at two- week intervals for 92 days after planting. Daughter plants of the inoculated plants were positive for XF by nested PCR amplification. Analysis of genetic variability by fatty acid methyl esterase (FAME) profiles divided 50 strains into 9 groups based upon qualitative and quantitative differences. The majority (74%) of strains were placed into a closely related group determined by cluster analysis. The profiles from restriction fragment length polymorphism (RFLP) analysis of genomic DNA restricted by two infrequent cutting endonucleases and separation by pulsed field gel electrophoresis grouped strains into four groups (A-D). Another endonuclease subdivided the B group into three groups. The dendrogram unweighted pair analysis of FAME profiles divided the population into four groups which correlated well with the RFLP groups. Considerable diversity appears within the species.



















vii















INTRODUCTION


Xanthomonasfragariae causes angular leaf spot disease on strawberry (Fragaria species; Fragaria x aanassa Duchesne). The disease was first found in Minnesota in 1960 (Kennedy and King, 1962a) and it is currently found in many regions of strawberry production throughout the world (Maas, 1984; Ritchie et al., 1993). Angular leaf spot was first reported in Florida in 1971 (Howard, 1971). Dissemination of the bacterium occurred via the transportation of infected plants (Maas, 1984; Panagopolous et al., 1978; Dye and Wilkie, 1973). In Florida, strawberry plants which arrive from northern nurseries for transplanting in the fall frequently have leaves which exhibit symptoms of angular leaf spot. Typical pathogenic strains of X. fragariae can be isolated from the lesions. A diagnostic symptom of the disease is the translucent appearance of lesions when viewed with transmitted light (Maas, 1984). A vascular collapse of the plant from systematic invasion by the bacterium has been described in California (Hildebrand et al., 1967).

The epidemiology of angular leaf spot is mostly unknown in fields in Florida where strawberry production is an annual crop. Transplants are obtained each season from nurseries in Canada and northern states; few transplants are produced in Florida. A source of inoculum other than infected transplants has not been found. Howard (1971) was unable to determine the inoculum source for infected plants from nurseries in Florida or other states. In surveys


I









2

conducted in 1968, 1970, 1993, and 1994, plants which had symptoms of angular leaf spot in the spring did not have symptoms of the disease the following August (Howard, 1971; P. D. Roberts, unpublished). However, in 1969 mild infections on one variety were observed in mid-August. Plants transplanted to fields did not develop angular leaf spot. Kennedy and King (1962b) determined that the bacterium overwintered in infected leaves buried in the soil and caused disease symptoms on plants the next year. The bacterium did not survive free in the soil nor were any naturally occurring hosts identified in host-range studies (Kennedy and King, 1962a). For bacteria on plant refuse to serve as an inoculum source in Florida, the bacterium must oversummer. Optimal growth (=200 C) of the bacterium (Howard et al., 1985; Kennedy and King, 1962b) is at temperatures cooler than normally occurs in Florida during the summer. The survival of the bacterium on plants in summer nurseries in Florida and inoculum sources other than infected transplants have not been established.

The effect of angular leaf spot on yield is unknown. Howard (1971) accredited some yield losses due to angular leaf spot in fields in Florida but he did not quantify the losses. In Wisconsin, a decrease in yield of 70 to 80% was estimated due to the disease (Epstein, 1966). However, the production in the northern United States is a perennial, matted-row system which differs significantly from the Florida production. A significant loss in marketable fruit may occur due to infections of the calyx. The sepals become brown and dry and the fruit is unmarketable because of its unattractive appearance (Epstein, 1966; Maas, 1995).

Chemical control of bacterial diseases is difficult. Antibiotics have limited effectiveness over time since mutations to bacterium may occur and form resistant populations (Stall and Thayer, 1962). Therefore copper compounds are frequently used to









3

control a bacterial disease (Jones et al., 1991). Antibiotics and copper compounds were effective protectants against angular leaf spot but these did not eradicate the disease (Alippi et al., 1989). Marco and Stall (1983) examined chemical control of strains of Xanthomonas campestris pv. vesicatoria that differed in sensitivity to copper. The mixture of cupric hydroxide plus mancozeb was more effective than cupric hydroxide alone to control both the copper-resistant and the copper-sensitive strain. The application of copper compounds to strawberry plants is confounded by the fact that copper can be phytotoxic to strawberry plants (Howard and Albregts, 1973). The application of a chemical can reduce the infection by a pathogen and slow the rate of the epidemic, e.g. van der Plank's (1963) apparent infection rate (Fry et al., 1979; Pataky and Lim, 1981). A reduced rate of fungicide applied at frequent intervals as a protectant has been used effectively to reduce disease on crops (Fry 1975; Fry et al, 1979; Conway et al., 1987).

Identification of plants infected with X fragariae is a priority because of the ease of movement of infected but asymptomatic plants (Maas, 1995). International movement of infected plants is blamed for introduction of angular leaf spot into Greece and New Zealand (Panagopoulos et al., 1978; Dye and Wilkie, 1973). Nursery-plant producers are pressured to provide disease-free plants by foreign countries and by farmers who refuse to buy infected transplants. The European Plant Protection Organization (EPPO) lists X fragariae as a quarantine pest and has prescribed phytosanitary procedures. In the future, regulatory issues may be of greater concern. The production of disease-free plants is essential for control of angular leaf spot. Therefore, accurate identification of plants infected with the bacterium is imperative. Available detection techniques are limited in their usefulness and accuracy to









4

detect low populations of the bacterium that may exist in asymptomatic tissue. Xanthomonas fragariae may be identified in the early stages of leaf infection by the diagnostic translucent, watersoaked lesions viewed with transmitted light; however, older lesions may be confused with symptoms caused by fungal pathogens (Kennedy and King, 1962b). Diagnosis based on symptoms is very difficult and not applicable for asymptomatic plants. Identification of the disease based upon isolation and characterization of the causal agent may also be difficult because X fragariae grows slowly and may be masked by faster growing organisms (Hildebrand et al., 1967). Expression of watersoaked lesions takes 6 days or longer after inoculation. Thus, fulfillment of Koch's postulates to confirm pathogenicity is difficult and time-consuming.

Assays have been developed with improved sensitivity and specificity for the detection of plant pathogens in plant tissue. A specific, indirect, enzyme-linked immunosorbent assay (ELISA) was developed to detect X fragariae from symptomatic plant tissue (Rowhani et al., 1994). The antibody assay did not react with bacterial strains of 16 pathovars of X campestris or non-pathogenic bacteria from strawberry leaves. A single cross-reaction occurred to a strain of X. campestris isolated from Nerium oleander. The assay detected bacteria directly from a visible lesion on a strawberry leaf The level of sensitivity of the assay was at ca. 104 colony forming units (cfu) per ml. However, bacteria on asymptomatic plants may not be detected by this ELISA assay.

The polymerase chain reaction (PCR) has been used to amplify specific DNA sequences to detect and identify many plant pathogens including some members of the genus Xanthomonas (Henson and French, 1993; Louws et al., 1994). Leite et al. (1994) utilized









5

primers specific to regions of the hrp gene cluster which confers hypersensitivity and pathogenicity from Xanthomonas campestris pv. vesicatoria to detect pathogenic xanthomonads. Saprophytic and non-pathogenic bacteria lack the hrp genes (Stall and Minsavage, 1990; Lindgren et al., 1986; Willis et al., 1991) and DNA from these bacteria was not amplified by the hrp gene cluster primers. Differentiation ofX. campestris pathovars was made by restriction endonuclease analysis (REA) patterns generated by digestion of the PCR products with frequent cutting enzymes (Leite et al., 1994 and 1995). The primers were used to detect X c. vesicatoria in seed lots of naturally infected pepper and tomato (Leite et al., 1995). The sensitivity of detection by PCR is reported at 10' to 102 cfu per ml (Minsavage et al., 1994; Leite et al., 1994; Henson and French, 1993). In nested PCR, sensitivity of detection is increased by using PCR products from an amplification as target DNA in a second round of amplification by a second set of primers internal to the first (Schaad et al., 1993). McManus and Jones (1995) reported an increase in sensitivity of 1,000-fold with nested PCR over a single round of PCR amplification to detect Erwinia amylovora.

The epidemiology of X fragariae must be understood before effective control strategies can be devised. The delineation of bacterial populations is imperative to such studies. The relationship of X fragariae to other members of the genus Xanthomonas has been examined (Hildebrand et al., 1990; Vauterin et al., 1995; Hodge et al., 1992), but the genetic variability of strains within the species has not been reported (Maas, 1995). The information from studies to compare strains of X. fragariae at the genetic level could be useful to identify the origins and spread of specific strains or genetic types. This information could be applied in international tracking of pathogen populations for quarantine programs.









6

In addition, the identification of genetic types is a prerequisite to identify sources of resistance in strawberry to the pathogen. While strawberry cultivars exhibited levels of susceptibility or tolerance to X fragariae; only F. moschata Duch. appeared to be immune (Hazel, 1981; Hazel and Civerolo, 1980; Kennedy and King, 1962b). A screening program to identify genes for resistance must incorporate representatives of the genetic variants in the screening process otherwise the resistance may be overcome quickly by genetic variants.

Studies to characterize populations of bacteria have used biochemical and molecular biological techniques. Protein staining and fatty acid analysis have been useful to detect differences at the metabolic level within species of Xanthomonas (Bouzar et al., 1994; Stall et al., 1994). The polymerase chain reaction (PCR) has been used to analyze differences between pathovars and strains of bacteria (Hensen and French, 1993). Primers specific to the hrp-gene cluster ofXanthomonas campestris pv. vesicaloria amplified genomic DNA and restriction enzyme analysis of the PCR product differentiated between many pathovars and species ofXanthomonas (Leite et al., 1994 and 1995). Primers sets from conserved repetitive bacterial DNA elements generated genomic fingerprints which were used to differentiate gram-negative soil bacteria and pathovars of Xanthomonas and Psuedomonas (De Bruijn, 1992; Louws et al., 1994). Restriction length fragment polymorphisms (RFLP) of the bacterial genome digested with rare-cutting restriction endonucleases and resolution by pulsed field gel electrophoresis (PFGE) has been used to type many bacterial species and genera (Cooksey and Graham, 1989; Egel et al., 1991; Smith et al., 1995).

The objectives of these studies were to investigate epidemiological aspects of angular leaf spot on strawberry including the incidence of X fragariae on transplants from northern









7

nurseries, disease spread in strawberry production fields, the effect of angular leaf spot on yield, and the control of the disease by chemicals. To aid in epidemiological studies, a sensitive and specific technique by PCR reaction to detect X fragariae was developed. Primers were designed specific to the region of genomic DNA from X fragariae related to the hrp genes ofX c. vesicatoria. The survival ofX. fragariae on nursery strawberry plants in the field at two locations in Florida and dissemination to daughter plants was examined to understand the disease cycle of angular leaf spot in Florida. The genetic variability of a collection of strains of X fragariae from the United States and Canada was examined. Analyses were by restriction length fragment profiles of genomic DNA restricted with rarecutting endonucleases separated by pulsed field gel electrophoresis and by profiles of fatty acid methyl esters.














CHAPTER 2
DISEASE PROGRESS, YIELD LOSS, AND CONTROL OF
XANTHOMONAS FRAGARIAE ON STRAWBERRY PLANTS Introduction

Angular leaf spot of strawberry is caused by the bacterium Xanthomonasfragariae. The disease was first reported from Minnesota in 1960 (Kennedy and King, 1962a) and is currently found in many regions of strawberry production throughout the world. Dissemination of the bacterium occurred via the transportation of infected plants (Maas, 1984; Panagopolous et al., 1978; Dye and Wilkie, 1973). A diagnostic symptom of the disease is the translucent appearance of lesions when viewed with transmitted light (Maas, 1984). A vascular collapse of the plant from systematic invasion by the bacterium has been described in California (Hildebrand et al., 1967).

The epidemiology of angular leaf spot is mostly unknown in fields in Florida where strawberry production is an annual crop. Transplants are obtained each season from nurseries in Canada and northern states; few transplants are produced in Florida. A source of inoculum other than infected transplants has not been found. Howard (1971) was unable to determine the inoculum source for infected plants from nurseries in Florida or other states. Kennedy and King (1962b) determined that bacteria overwintered in infected leaves buried in the soil and caused disease symptoms on plants the next year. The bacterium did not survive free in the soil nor were any naturally occurring hosts identified in host range studies (Kennedy and


8









9

King, 1962b). For bacteria on plant refuse to serve as an inoculum source in Florida, the bacterium must oversummer. Optimal growth (z200 C) of the bacterium (Howard et al., 1985; Kennedy and King, 1962b) is at temperatures cooler than normally occurs in Florida during the summer. The bacterium survived on summer nursery plants in the field in Florida but at very low populations (Roberts, unpublished). The effect of angular leaf spot on yield is unknown. Howard (1971) accredited some yield losses due to angular leaf spot in fields in Florida but he did not quantify the losses. In Wisconsin, a decrease in yield of 70 to 80% was estimated due to the disease (Epstein, 1966). However, the production in the northern United States is a perennial, matted-row system which differs significantly from the Florida situation. A significant loss in marketable fruit may occur due to infections of the calyx. The sepals become brown and dry and the fruit is unmarketable because of its unattractive appearance (Epstein, 1966; Maas, 1995).

Chemical control of bacterial diseases on plants is difficult. Antibiotics have limited effectiveness over time since mutations may form resistant populations (Stall and Thayer, 1962). Therefore copper compounds are frequently used to control a bacterial disease (Jones et al., 1991). Antibiotics and copper compounds were effective protectants against angular leaf spot but these did not eradicate the disease (Alippi et al., 1989). Marco and Stall (1983) examined chemical control of strains of Xanthomonas campestris pv. vesicatoria which differed in sensitivity to copper. The mixture of cupric hydroxide plus mancozeb was more effective than cupric hydroxide alone to control both the copper-resistant and the coppersensitive strain. The application of copper compounds to strawberry plants is confounded by the fact that copper can be phytoxic to strawberry plants (Howard and Albregts, 1973). The









10

application of a chemical pesticide can reduce the infection by a pathogen and slow the rate of the epidemic, e. g., van der Plank's (1963) apparent infection rate (Fry et al., 1979; Pataky and Lim, 1981). A reduced rate of fungicide applied at frequent intervals as a protectant has been used effectively to reduce disease on crops (Fry 1975; Fry et al, 1979; Conway et al., 1987).

Our objectives in this study were to examine epidemiological aspects of angular leaf spot including the incidence of X fragariae on transplants from northern nurseries, the effect of angular leaf spot on yield, and the control by chemicals.



Materials and Methods



Survey of farmer's fields and storage facility. Disease incidence of angular leaf spot on plants in fields located around Dover and Plant City, FL, was examined in October 1993. Plants were assessed for symptoms of angular leaf spot within 2 to 4 days of transplanting. Leaves from symptomatic plants were collected. A 9-mm diameter of leaf tissue was removed surrounding a lesion and macerated in 200 1l of sterile water. A loopful of the suspension was dilution streaked onto Wilbrink's medium (WB) (Koike, 1962) and colonies of X fragariae were identified. The pathogenicity of these isolated strains was tested on 'Sweet Charlie' strawberry plants. In 1994, transplants were sampled while in cold storage at a facility in Dover, FL. Seven groups of plants, comprised of four cultivars from six northern growers were sampled. Three boxes were selected at random from within a group and 20 plants were removed from each box. The plants were examined for symptoms of









11

angular leaf spot and attempts to isolate the pathogen from putative lesions was done by the method described above.

Inoculation. Three strains ofX.fragariae (Xfl 13, Xfl03, and Xfl425) were used to inoculate plants used in field plots. Strains Xfl13 and Xfl03 were isolated from infected plants at Gulf Coast Research and Education Center (GCREC), Dover, FL and Xf1425 was obtained from A. Chase (Central Florida Research and Education Center, Apopka, FL). Strains were cultured on WB at 240 C and long-term storage was at -700 C in 15% glycerol. The sensitivity of the strains to copper was tested by growth on nutrient agar amended with CuSO4 (Stall et al., 1986). Three days prior to inoculation, each of the strains was streaked to ten plates of WB. Bacterial cells were collected from plates, suspended in sterile 0.01 M MgSO4, and the concentration adjusted to approximately 107 cfu per ml. Equal volumes of each suspension were combined to comprise the inoculum. Plants were inoculated by dipping bundles of 25 plants into the bacterial suspension for 30 s. Control plants were dipped into

0.01 M MgSO4. Plants were placed into plastic bags and incubated 24 h at 220 C.

Field experiments. Field experiments were located at GCREC, Dover, FL, from October 1993 through March 1994 and repeated the following season. Transplants of 'Sweet Charlie' grown in the summer nursery at GCREC-Dover were used. Raised beds were prepared and fertilized with 10N-4P-10K at a rate of 2000 lbs per acre with one-fourth banded before bed preparation and the remainder banded 5 cm deep in the bed center the first season. In the second season, three-fourths pound of N and K per acre per day were applied through the drip irrigation system. Soil was fumigated with 98% methyl bromide and 2%









12

chloropicrin at 448 kg per acre. Beds were covered with 1 mm black polyethylene mulch immediately after fumigation.

The experimental design was a randomized complete block in a 2 x 3 factorial design with four replications. The first factor had two levels: plants inoculated with either the suspension of X. fragariae or MgSO4. The three levels of the second factor were: no chemical treatment, the label (1 x) rate of cupric hydroxide (Kocide 101 at 9.08 kg of active ingredient per acre) plus mancozeb (Dithane DF at 6.81 kg of active ingredient per acre) sprayed at 7- to 14-day intervals, or the reduced (0.1 x) rate of cupric hydroxide plus mancozeb sprayed at 2- to 4-day intervals.

An individual plot contained 18 plants arranged in two rows of nine plants. The beds were spaced on 1.22 meter centers with 30 cm between rows and 30 cm within a row. Fallow area was 3.55 m within a row and 2.44 m between rows. Pesticides were applied throughout season as needed to control insects and fungal diseases. Chemical applications were made by a handheld wand attached to CO2-charged canister at 40 psi and pesticide was applied to runoff. Plants were transplanted on 15 October 1993 and on 20 October 1994. Overhead sprinkler irrigation was applied 8 hours daily for 10 to 14 days to establish transplants and applied throughout the season as needed. Drip irrigation was installed in the summer of 1994 at GCREC-Dover and was used as supplemental irrigation in 1995.

Estimates of disease severity of angular leaf spot were made at two-week intervals. Disease severity was expressed as percent leaf area diseased for the entire plant for each of six plants located in the center of each plot. Progress curves were plotted as the mean of disease severity of replicate treatments versus time. Area under the disease progress curve









13

(AUDPC)(Shaner and Finney, 1977) was calculated for each plot and used in statistical analysis. Statistical analysis was performed by orthogonal contrasts using PC-SAS.

Yield data. Fruit were harvested at 2- to 4-day intervals from initial bearing in December through 30 March. Fruit were graded as marketable, culls, or nonmarketable due to damage by fungi. Marketable fruit were those free of rot, not misshapen, and greater than 10 g in weight. Culls were non-marketable fruit due to physical imperfections such as small size (< 10 g), damage by insects, or undesirable shape. Nonmarketable fruit was damaged by fungal diseases, usually caused by anthracnose, botrytis, and phomopsis. The weight in grams for each category was recorded for individual plots. Statistical analysis of the weight of fruit in each category was performed using orthogonal contrasts using PC-SAS.



Results



Disease incidence of transplants in field and storage facility. In 1993, six of seven fields contained plants with symptoms of angular leaf spot from which X. fragariae was isolated. In 1994, three of the seven groups sampled from the cold storage facility contained plants with symptoms of angular leaf spot. In one of these groups, plants with symptoms of angular leaf spot were found in all of the boxes sampled. In the other two samples, one box contained plants which were symptomatic with angular leaf spot. Bacteria were isolated from all plants with lesions of angular leaf spot and pathogenicity of these isolated strains on strawberry was confirmed. Strains which were resistant to copper were not identified.









14

Disease progress. Symptoms of angular leaf spot were not visible on plants at planting in either 1994 or 1995. Disease progress curves of angular leaf spot on strawberry plants for all treatments are presented in Figure 2-1. In 1994, the disease severity of angular leaf spot on inoculated plants which did not receive a spray treatment increased for 89 days after transplanting (DAT), decreased until 118 DAT, and then increased to 25% by 118 DAT and leveled out to the end of the season. In 1995, the epidemic on plants inoculated with the bacterium initially increased, then decreased at 61 DAT until 111 DAT, and then increased and reached 15% by 154 DAT. The disease on inoculated plants sprayed with either the I x or 0.1 x rate of cupric hydroxide plus mancozeb followed similar disease progress curves except the amount of disease was reduced for each treatment. In contrast, the onset of the epidemic on the noninoculated plants was greatly delayed. Disease was observed on these plants after 118 DAT in 1994 and at 125 DAT in 1995. The noninoculated plants had very little disease (0-3% severity) by the end of the season, regardless of spray treatment.

Statistical analysis of the AUDPC for each treatment is shown in Figure 2-2. Noninoculated plants had significantly (P = 0.0001) lower disease than plants inoculated with the bacterium. Noninoculated plants treated with the 1 x rate did not have disease either year. Inoculated plants sprayed with the 1x rate of cupric hydroxide and mancozeb had a significantly (P = 0.0004 in 1994, P = 0.0001 in 1995) lower AUDPC compared to the control treatment in both years. The plants sprayed with the 0.1 x rate of the fungicide had reduced disease in 1994 (P = 0.17) and, in 1995, the reduction in AUDPC was highly significant (P = 0.0001) compared to plants without chemical sprays.










15

30


A
25
NonI, NoS
O
20 NonI, 0.1X

Nonl, IX S 15 O 15 Inoc,NoS

Inoc, 0. 1X
10
Inoc, 1X

5


0
19 33 47 64 75 89 105 118 132 145 157 DAT
3 17 1 7 29 12 28 10 24 9 21
Nov Dec Jan Feb March

16

14 B

12 Nonl, NoS

NonI, 0.1X
10
Noni, IX





4


2

0
18 32 48 61 77 111 125 139 154 DAT
7 21 7 20 5 8 22 8 23
Nov Dec Jan Feb March

Figure 2-1. Disease progress curves of Xanthomonasfragariae on strawberry in Florida in the 1994 (A) and 1995 (B) seasons. The mean of disease severity for four replicate blocks are plotted against days after transplanting (DAT).










16



80
A Ba
70 Ba

60
Bb
50 40

30 20

lo Aa Aa Aa


Noninoculated Inoculated NoS 0.1X IX NoS 0. X IX



so Ab
B
40


Q 30

Aa
20


10
Aa AaAa


Noninoculated Inoculated


NoS 0.1X IX NoS 0.IX IX

Figure 2-2. Area under the disease progress curve (AUDPC) of Xanthomonasfragariae on strawberry in the 1994 (A) and 1995 (B) seasons and the effect of applications of cupric hydroxide plus mancozeb at the 1 x rate at 7-to 14-day intervals or at 0.1 x (10%) rate of the chemical combination applied at 2-to 4-day intervals or no spray (NoS). A, B = significant difference (P = 0.05) within the first factor of inoculated vs. noninoculated plants; a, b = significant difference within the second factor of the spray treatments.









17

Yield loss. In both years, marketable yield on plants infected with angular leaf spot was significantly decreased (P = 0.005 in 1994 and P = 0.04 in 1995) compared to yield on non-inoculated plants (Fig. 2-3). Total and marketable yields were reduced by an average of 7% and 8%, respectively, in 1994 and by 6% and 10% in 1995. Total yield had similar reductions as marketable yield, therefore only marketable yield is presented. Yields were generally lower in 1995 than in 1994.

The effect of spray treatments on the yield from inoculated and noninoculated plants is presented in Figure 2-4. Marketable yield on noninoculated plants sprayed with the 1x treatment of cupric hydroxide plus mancozeb was significantly reduced (P = 0.04) compared to plants which received the 0.1 x or no-spray treatments in 1994. No difference in yield was detected on inoculated plants receiving the spray treatments in 1994. In 1994, the yield on inoculated plants sprayed with the 1 x rate was significantly reduced (P = 0.005) compared to the other two treatments. Yield on noninoculated plants sprayed with the I x and 0.1 x chemical sprays was reduced significantly (P =0.04) compared to yield on nonsprayed plants in 1995. The spray treatments did not have any effect on culled fruit in 1994 or 1995 (Fig. 2-5). No difference in culled fruit was detected among spray treatments. Nonmarketable fruit lost because of damage by fungal diseases was significantly different at P = 0.01 for all treatments in 1994 but the relationship to spray treatments was unclear (Fig. 2-6). In 1995, there was no significant difference in damage to fruit by fungi for either inoculated or spray treatments. Overall, the total yield lost to fungal infections was much lower in 1994 compared to 1995.









18


10
A
A AA
Bb
8 Ab BbBb
Bb





4/


2



Noninoculated Inoculated

NoS 0.1X IX NoS 0.1X IX



B Ab Aa
Aa





4



2




Noninoculated Inoculated
NoS O.iX IX NoS 0.X IX


Figure 2-3. The average marketable yield on strawberry plants which were inoculated with Xanthomonasfragariae (Inoc) or not inoculated (Noninoc) for the 1994 and 1995 seasons. Yield was the average for four replications for marketable fruit only. Different letters (A, B) represent significant difference at P= 0.05.
8 Aa Aaiiiiiiiii


B Ab Abiiiiiiiii














Fiur8-3.Th Aaveaemreal il oAtabrypat ahihwrncltdwt Xath.m....r.....a (Aob or noAncltd(onncao h 94 n 95saos Yield was the averageil foAorrpiainaormreal ri ny Dfeetltes(,B represent s11iiicant d::iffreceatP 005









19






10
A
BA 8B








4



2






Noninoc Inoc Noninoc Inoc



1994 1995






Figure 2-4. The marketable yield on strawberry in Florida in the 1994 (A) and 1995 (B) seasons. A, B = F was significant at P = 0.05 within the first factor of noninoculated vs.inoculated; a, b = F was significant at P = 0.05 within the second factor of spray treatments.











20



1.2

A Aa Aa
S Aa Aa


0o.s , 0.6


0.4 0.2


0
Noninoculated Inoculated NoS 0.1X IX NoS 0. IX IX Aa Aa
1.2 - Aa Aa Aa Aa




1


0.6 0.4


0.2


0
Noninoculated Inoculated NoS 0.1X 1X NoS 0.1X IX





Figure 2-5. The culled fruit on strawberry in Florida in the 1994 (A) and 1995 (B) seasons. A, B = Fwas significant at P = 0.05 within the first factor of noninoculated vs.inoculated; a, b = F was significant at P = 0.05 within the second factor of spray treatments.











21


0.8
A Aa
0.7

0.6 Bb
Aa
0.5 Ba

- 0.4

0.3

Ba
0.2 0.1



Noninoculated Inoculated NoS 0.1X IX NoS 0.iX IX
1.6

1.4 Aa Aa Aa Aa Aa Aa
1.2



-ol
0.8

0.6

0.4 0.2

0

Noninoculated Inoculated NoS 0.IX IX NoS 0.1X IX



Figure 2-6. The yield lost to fruit damaged by fungal diseases on strawberry in Florida in the 1994 (A) and 1995 (B) seasons. A, B = F was significant at P = 0.05 within the first factor of noninoculated vs.inoculated; a, b, c = F was significant at P = 0.05 within the second factor of spray treatments.









22

Discussion



A decrease in yield of strawberry fruit was observed both seasons due to angular leaf spot. This report is the first to quantify the reduction in yield due to this disease. The 8 to 10% loss determined in these studies is much lower than the 70 to 80% loss estimated by Epstein (1966) in fruit production fields in Wisconsin. Production in northern regions of the U. S. is usually perennial and the age of the plants was not given. The loss was reported on the cultivar 'Sparkle' which is not produced commercially in Florida. Therefore, we are unable to compare our yield loss to his report. Howard (1971) also reported unquantified yield losses due to this disease but on cultivars which are no longer in commercial production. In our studies, the average yield loss observed both years was very similar despite the differences in the disease levels and the base yields between the two seasons. In 1995, the disease severity was 60% of the amount estimated in the previous season; however, the yield loss was actually higher by 2% for marketable yield. The 10% reduction in yield represents an estimated loss of 1233 kg of berries per ha. Strawberry fruit has a very high cash value and a decrease in revenue of 8 to 10% represents a significant economic loss to producers.

Strawberry producers in Florida currently try to control angular leaf spot by application of copper compounds and by avoidance of the disease. An understanding of the epidemiology of angular leaf spot is necessary to determine the effectiveness and practicability of any control method. The survey of farmers' fields and the cold storage facility established that transplants arrive from northern nurseries infected with angular leaf spot at fairly high disease incidence. Therefore, disease is introduced into the field on infected transplants. In









23

our surveys by the sampling method used in the first season, more plants were assessed for symptoms of angular leaf spot in the field compared to the number of plants in the random sampling of boxes in cold storage. This probably accounts for identification of diseased plants in nearly all of the farmer's fields versus less than in 25% of boxed plants. In addition, plants were examined until disease was found in the field which skewed the randomness of the survey.

In field experiments, the spread from inoculated plants to noninoculated plants in nearby experimental plots was minimal. Thus, the incidence of plants with angular leaf spot in the field likely occurred from inoculum already present on the transplants. In the field experiments, if inoculum was present from another source, such as debris in the soil or an alternate host, most likely the disease on the noninoculated plants would have been more general and appeared earlier. Disease severity on noninoculated plants in the experimental plots was extremely low and most plants remained free of angular leaf spot. For the disease that occurred on noninoculated plants, the inoculum was probably transported mechanically from inoculated plants during harvests or disease readings.

Progress of angular leaf spot developed similar to curves of pathogens described on other crops (van der Plank, 1963). A decrease in disease severity was seen about midway through the season in 1994. This reduction, or negative infection rate, was due to growth by strawberry plants which diluted the amount of disease relative to the total leaf area. The dilution effect of healthy plant tissue was described by van der Plank (1963). In 1995, a negative infection rate was seen very early in the epidemic. This was because the symptoms of the disease almost disappeared from the field as the affected leaves died and were lost. The









24

difference in disease progress between the two seasons did not appear to be from differences in rainfall since mean rainfall was approximately the same both years (data not shown). Mean temperatures also were similar; however; the number of days at temperatures below 100 C in November and December 1993 were much greater than in 1994. Another difference between the seasons was that sprinkler irrigation was reduced in 1995 because of the change to drip irrigation. Overhead irrigation, fog, high humidity, and rain are positive factors in disease development and spread (Epstein, 1966; Hildebrand et al., 1967; Kennedy and King, 1962b; Lai, 1978). The trend towards drip irrigation in commercial production fields should have a positive effect to reduce the spread and survival of the bacterium.

Strawberry producers currently apply copper-based compounds to control angular leaf spot despite the phytotoxicity to plants caused by copper. Application of cupric hydroxide plus mancozeb at the I x rate was phytotoxic to strawberry plants (Howard, 1973; this study). In our tests, the greatly reduced 0.1 x rate did not harm plants and significantly reduced disease severity in 1995. The protectant application of bactericide at this rate and spray schedule was intended to reduce the total amount of inoculum and to prevent spread of the disease. The approach of frequent sprays at reduced concentrations may have potential to control this disease. More studies are needed to evaluate different rates, application intervals, and chemical mixtures to achieve maximum disease control while avoiding yield losses. An alternate approach might be to control the disease in the nursery by copper applications. The loss in yield due to phytotoxicity would not be an issue on nursery plants. The sprays would reduce the amount of initial inoculum on transplants and subsequently reduce the amount of disease in fruit production fields.














CHAPTER 3
DEVELOPMENT OF A SPECIFIC AND SENSITIVE DETECTION TECHNIQUE BY THE POLYMERASE CHAIN REACTION FOR XANTHOMONAS FRAGARIAE AND APPLICATION IN A STUDY OF SURVIVAL OF THE BACTERIUM ON STRAWBERRY PLANTS


Introduction



Angular leaf spot of strawberry (Fragaria x ananassa Duchesne), caused by the bacterium Xanthomonasfragariae, was first reported in Minnesota in 1960 and is now found in many areas of strawberry production throughout the world (Kennedy and King, 1960a; Maas, 1984; Ritchie et al., 1993). The disease is apparently disseminated by the transportation of infected plants (Maas, 1984). In Florida, strawberry plants which arrive from northern nurseries for transplanting in the fall frequently have leaves which exhibit symptoms of angular leaf spot. Typical pathogenic strains of X fragariae can be isolated from the lesions (P. D. Roberts, unpublished).

The epidemiology ofX fragariae is mostly unknown in fields of strawberry in Florida where production is from annual crops. Angular leaf spot was first reported in the state in 1971 (Howard, 1971). Howard (1971) was unable to determine the inoculum source for infected plants from nurseries in Florida or other states. Bacteria may survive on infested leaves in the soil (Kennedy and King, 1962b), but plants in Florida are usually treated with the herbicide paraquat at the end of the season and removed. The bacterium does not survive 25









26

freely in the soil (Kennedy and King, 1962b). Cool temperatures (= 20' C) are optimal for disease symptom expression (Howard et al., 1985; Kennedy and King, 1962b) and high temperatures (>280 C) such as those which occur in Florida during the summer months are unfavorable. In surveys conducted in 1968, 1970, 1993, and 1994, plants which had symptoms of angular leaf spot in the spring did not have symptoms of the disease the following August (Howard, 1971; P. D. Roberts, unpublished). However, in 1969 mild infections on one variety were observed in mid-August. Plants transplanted to fields did not develop angular leaf spot. The survival of the bacterium on plants in summer nurseries in Florida and inoculum sources other than infected transplants have not been investigated.

Identification of plants infected with X. fragariae is a priority because of the ease of transmission on infected but asymptomatic plants (Maas, 1995). International movement of infected plants is blamed for introduction of angular leaf spot into Greece and New Zealand (Panagopoulos et al., 1978; Dye and Wilkie, 1973). Nursery-plant producers are pressured to provide disease-free plants by buyers in foreign countries and by farmers who refuse to buy infected transplants. The European Plant Protection Organization (EPPO) lists X fragariae as a quarantine pest and has prescribed phytosanitary procedures (Maas, 1995). In the future, regulatory issues may be of greater concern. The production of disease-free plants is essential for control of angular leaf spot. Therefore, accurate identification of plants infected with the bacterium is imperative. Available detection techniques are limited in their usefulness and accuracy to detect low populations of the bacterium that may exist in asymptomatic tissue. Xanthomonasfragariae may be identified in the early stages of leaf infection by diagnostic translucent, watersoaked lesions viewed with transmitted light; however, older lesions may









27

be confused with symptoms caused by fungal pathogens (Kennedy and King, 1962b). Diagnosis based on symptoms is very difficult and not applicable for asymptomatic plants. Identification of the disease based upon isolation and characterization of the causal agent may also be difficult because X fragariae grows slowly and may be masked by faster growing organisms (Kennedy and King, 1962a). Expression of watersoaked lesions takes 6 days or longer after inoculation. Thus, fulfillment of Koch's postulates to confirm pathogenicity is difficult and time-consuming.

Assays have been developed with improved sensitivity and specificity for the detection of plant pathogens in plant tissue. A specific, indirect, enzyme-linked immunosorbent assay (ELISA) was developed to detect X. fragariae from symptomatic plant tissue (Rowhani et al., 1994). The antibody assay did not react with bacterial strains of 16 pathovars of X. campestris or non-pathogenic bacteria from strawberry leaves. A single cross-reaction occurred to a strain of X. campestris isolated from Nerium oleander. The assay detected bacteria directly from a visible lesion on a strawberry leaf. The level of sensitivity was at ca. 104 colony forming units (cfu) per ml. However, bacteria on asymptomatic plants may not be detected by this ELISA assay.

The polymerase chain reaction (PCR) has been used to amplify specific DNA sequences to detect and identify many plant pathogens including some members of the genus Xanthomonas (Henson and French, 1993; Louws et al., 1994). Leite et al. (1994) utilized primers specific to regions of the hrp gene cluster which confers hypersensitivity and pathogenicity from Xanthomonas campestris pv. vesicatoria to detect pathogenic xanthomonads. Saprophytic and non-pathogenic bacteria lack the hrp genes (Stall and









28

Minsavage, 1990; Lindgren et al., 1986 ) and DNA was not amplified by the hrp gene cluster primers. Differentiation ofX. campestris pathovars was made by restriction endonuclease analysis (REA) patterns generated by digestion of the PCR products with frequent-cutting enzymes (Leite et al., 1994 and1995). The primers were used to detect X c. vesicatoria in seed lots of naturally infected pepper and tomato (Leite et al., 1995). The sensitivity of detection by PCR is reported at 10' to 102 cfu per mi (Minsavage et al., 1994; Leite et al., 1994; Henson and French, 1993). In nested PCR, sensitivity of detection is increased by using PCR products from an amplification as target DNA in a second round of amplification by a second set of primers internal to the first (Schaad et al., 1993). McManus and Jones (1995) reported an increase in sensitivity of 1,000 -fold with nested PCR over a single round of PCR amplification to detect Erwinia amylovora.

Our objectives were to develop a sensitive and specific technique for detection of X fragariae. Our approach was to design primers specific to the region of genomic DNA from X fragariae related to the hrp genes of X. c. vesicatoria. The survival of X fragariae on nursery strawberry plants in the field at two locations in Florida and dissemination to daughter plants was examined to understand the disease cycle of angular leaf spot in Florida.



Materials and Methods



Bacterial strains and culture conditions. Strains ofX. fragariae and non-pathogenic xanthomonads isolated from strawberry were maintained at 240 C on Wilbrink's medium (Koike, 1965). Pathovars of Xanthomonas campestris were cultured on nutrient agar (Difco









29

Laboratories, Detroit, MI) and incubated at 280 C. Long term storage was at -700 C in 15% glycerol. Bacteria used for plant inoculations and DNA extractions were grown in 5 ml nutrient broth on a rotary shaker at 200 rpm for 16 h at 240 C. A rifampicin-resistant mutant of strain XF1425 was selected on Wilbrink's medium supplemented with 100 g/ml of rifampicin by the gradient plate technique (Szybalski, 1952).

Pathogenicity tests. Bacteria from overnight cultures in nutrient broth were centrifuged and washed three times with sterile water. The concentration of cells was adjusted in either 10 mM MgSO4-7H20 or sterile water to approximately 10' cfu per ml and sprayed to runoffon 'Sweet Charlie' strawberry plants placed under mist 24- to 48- h prior to inoculation. Inoculated plants were maintained under mist or put into growth chambers (Percival, Boone, IA) at 240 C with a 12 h photoperiod.

Sequencing and primer design. The hrp primers RST2 and RST3 from the hrp gene cluster of X. axonopodis pv. vesicaloria (Leite et al, 1994) were used to amplify genomic DNA of 49 strains of X fragariae. The PCR product from strain XF1425 of Xfragariae was isolated from agarose gel, cleaned by the Promega Wizard Kit (Promega, Madison, WI) and sequenced at the ICBR DNA Sequencing Facility, University of Florida, Gainesville, FL. The nucleotide sequence was compared to sequences of the PCR products amplified by the same primers from one X vesicatoria and two X a. pv. vesicatoria strains using the Seqaid II computer program (Rhoads and Roufa, 1991). Four primers were selected from the sequence of X. fragariae based upon unique DNA sequences and low homology compared to the DNA sequences from the other bacterial strains. The four oligonucleotide primers









30

were synthesized with a model 394 DNA synthesizer (Applied Biosystems, Foster City, CA) at the ICBR Facility, University of Florida, Gainesville, FL.

PCR amplification and nested amplification. Total genomic DNA was extracted by the method described by Ausubel et al. (1987). PCR amplification was performed using a DNA Thermal Controller PT-100 (MJ Research, Watertown, MA). Samples were in a total reaction volume of 50 Al and contained 1X amplification buffer (Promega, Madison, WI), 100 pM of each dNTP (Promega, Madison, WI), 50 )M of each primer, 1.25 U Taq DNA polymerase, and 100 ng of purified genomic DNA in 3 pl of TE (10 mM Tris and 1 mM EDTA, pH 8.0) buffer. Each reaction was overlaid with 50 pl of sterilized mineral oil (Sigma) for a total volume of 100 Ml in sterile 0.6 ml microcentrifuge tubes. In later experiments, the thermocycler was equipped with a heated lid controller (The Hot BonnetTM, MJ Research) which eliminated the need for the mineral oil overlay. Amplification of the DNA proceeded after template DNA was denatured at 950 C for 2 min followed by thirty amplification cycles and a final extension step at 720 C for 5 min. For primer set XF9 and XF 11, each amplification cycle consisted of denaturation at 950 C for 30 s, annealing at 650 C for 30 s, and extension at 720 C for 45 s. For primer XF12 with either primer XF9 or XF 10, the program was identical except the annealing temperature was 580 C.

For nested PCR, the first round of amplification was as described with primers XF9 and XF 11. In the second round of amplification, a 3 Ml sample from the first amplification mixture was used with primers XF9 and XF 12 and all other ingredients were added at the concentrations described above. The PCR cycle program for the primers XF9 and XF 12 was









31

used. In all PCR runs, including the nested assays, a water sample was used as a negative control.

Restriction endonuclease analysis (REA) of PCR products. The PCR samples with the overlay of mineral oil were cleaned by the method of Minsavage et al. (1994). Samples without the mineral oil were used directly in restriction reactions. An 8 gl sample of the PCR product was digested with restriction endonuclease Sau3AI, HaeIII or CfoI under conditions specified by the manufacturer (Promega). Restricted products were separated in 4% agarose gel (3% NuSieve, 1% SeaKem GTG, FMC BioProducts, Rockland, ME) containing 0.5 g.g/ml ethidium bromide in TAE buffer at 8 V/cm as described by Leite et al. (1994). DNA molecular weight marker XI (Boehringer Mannheim, Indianapolis, IN) was used for standard weight markers. Gels were photographed over a UV transilluminator with type 55 Polaroid film (Polaroid Corp., Cambridge, MA).

Detection of bacteria from infected plant tissue. The sensitivity of the assay in the presence of plant tissue was determined by adding known concentrations of bacteria to plant samples. A 9-mm diameter disk of plant tissue was macerated with mortar and pestle in 200 jul of phosphate buffer (pH 7.0) containing 5% polyvinylpolypyrrolidone (PVP40, Sigma Chemical Co., St. Louis, MO) and 0.02 M sodium ascorbate (PPA). The mixture was incubated at room temperature for a minimum of 1 h. The volume was adjusted to 582 Il with TE and DNA extraction proceeded as described above. A minimum of three experiments with three replications of each treatment for each primer set and nested reaction was performed. REA was performed on the final PCR products.









32

Strawberry plants with leaves exhibiting typical lesions resulting either from natural field infections or spray inoculation were assayed by removing a 9-mm diameter disk of tissue surrounding a lesion and proceeding as described above. Tomato leaf tissue was used as a negative control. For confirmation, a sample of the ground tissue suspension was plated on Wilbrink's medium amended with 0.05% Bravo 720 (chlorothalonil; ISK Biosciences, Mentor, OH). Plates were incubated at 240 C and colonies ofX. fragariae were identified by colony morphology.

Field experiments to examine oversummering survival of bacteria on nursery plants.

Six-week-old rooted transplants of 'Sweet Charlie' were spray inoculated with X. fragariae strain XF1425"f two weeks prior to planting. All plants transplanted into the field exhibited lesions of angular leaf spot. Plants were transplanted into the field on 27 June 1995 at Gulf Coast Research and Education Center-Dover, FL and on 29 June 1995 at Gulf Coast Research and Education Center-Bradenton, FL. Plants were sampled over a 14-week period through 27 September 1995. Another set of inoculated plants were placed into Percival growth chambers at 12 h photoperiod and 240 C and maintained for the duration of the experiment. Plants which were not inoculated and did not exhibit symptoms of angular leaf spot were placed in a greenhouse. At two-week intervals, five plants from the field, one plant from the growth chamber, and one plant from the greenhouse were sampled. Two samples of three daughter plants of the inoculated plants were sampled at 132 days after planting from both locations. Individual plants were assayed as follows. All the leaves from a single plant were removed and placed in a flask with 200 ml of phosphate buffer containing 0.02 % Tween 20. The samples were shaken either on a wrist action (Burrell Corp., Pittsburgh, PA)









33

or a rocker platform (Bellco Biotechnology, Vineland, NJ) for 2 to 16 h. A 200 P1 sample was plated onto Wilbrink's medium plus 100 mg/ml rifampicin plus 0.5% Bravo 720 and incubated at 240 C for 72 h. Colonies characteristic of X. fragariae were identified by amplification with the XF-specific primers and by REA. The remainder of the phosphate buffer sample was concentrated by vacuum filtration onto a 0.45 kim membrane disk (Millipore Corp, Bedford, MA). The disk was washed in 1.5 ml of TE and the suspension centrifuged 5 min at 14, 000 x g. The pellet was resuspended in 582 ,l TEPA (TE buffer containing 5% PVP40 + 0.02 M sodium ascorbate) and incubated at room temperature I h. DNA was extracted as described above. The crown of the plant was sectioned, macerated by mortar and pestle in 10 ml of PPA. The plant tissue debris and PVPP were collected by centrifugation at 1,000 x g for 1 min. The supernatant with the bacterial cells was removed to a clean centrifuge tube and centrifuged at 14,000 x g for 5 min. The pellet was resuspended in 582 ,l TE and DNA extraction proceeded as described above. A sample was plated onto Wilbrink's agar plus 100 ajg/ml rifampicin and 0.5% Bravo and incubated at 240 C for 72 h. The nested PCR reaction and REA were performed as described above.



Results



Pathogenicity assays. All strains identified as X. fragariae by colony characteristics caused disease symptoms typical of angular leaf spot on strawberry plants. Many other xanthomonads were isolated from strawberry tissue but they did not cause symptoms of angular leaf spot.









34

Specificity of primers For all strains ofX. fragariae, the PCR products amplified by primers RST2 and RST3 were ca. 840-bp (Fig. 3-1). The REA profiles that resulted from restriction of the PCR products with CfoI or HaeIII were the same for all the strains of X. fragariae (data not shown). Genomic DNA from non-pathogenic strains of Xanthomonas isolated from strawberry was not amplified by the primers.

The four primers synthesized were: XF9 (5' TGGGCCATGCCGGTGGAACTGT GTGG3'); XF10 (5' TGGAACTGTGTGGCGAGCCAG 3'); XF11 ('5 TACCCAGCCGT CGCAGACGACCGG 3'); and XF12 (5' TCCCAGCAACCCAGATCCG 3'). Primers XFIO and XF 12 were internal to the other two primers.

Primer XF9 paired with XF11 or XF12 delineated a 537-bp or a 458-bp fragment, respectively (Fig. 3-1). Primers XFIO and XF12 delineated a 448-bp fragment. PCR products conformed to the estimated sizes based upon sequence data. All 49 strains of X fragariae were amplified with primer sets XF9 and XF 11, XF9 and XF 12, and XF 10 and XF 12. DNA from strains other than X fragariae was not amplified by primer XF9 paired with XF 11 or XF 12. Strains tested were ATCC type strains of Xanthomonas campestris pathovars begoniae, campestris, carotae, celebensis, glycines, incanae, manihotis, musacearum, papavericola, pelargonii, phaseoli, poinsettiicola, raphani, taraxaci, vignicola, vitians, and nine non-pathogenic strains of xanthomonads isolated from strawberry. Except for X c. pelargonii, genomic DNA from these strains could not be amplified with primer set XFl0 and XF12. Due to the non-specific amplification of DNA from X c. pelargonii by primer XF 10 paired with XF12, only primer XF9 paired with either XF 11 or









35

















1 2 3 4

1000 basepair

500 basepair





















Figure 3-1. Agarose gel of the PCR products generated by amplification of genomic DNA from a strain ofXanthomonasfragariae with the hrp-primers RST2 and RST3 (Lane 1) and the XF-specific primers XF9 paired with XF 11(Lane 2) and XF9 paired with XF 12 (Lane 3), and molecular weight marker XI (Boehringher-Manheim, Indianapolis, IN; Lane 4).









36

XF 12 was used in experiments to detect DNA from X fragariae in the presence of plant materials.

REA of PCR products. PCR products from DNA of X fragariae amplified with primer set XF9 and XFI2 restricted with CfoI, Sau3AI, or HaeIII produced distinct banding patterns for each enzyme (Fig. 3-2). Five bands resulted from CfoI restriction (125, 108, 90, 75, and 60 bp), and three bands each from the HaeIII restriction (258, 150, and 50 bp) and Sau3AI restriction (250, 125, and 83 bp). Polymorphisms were not observed among strains of X fragariae in the REA of the restricted PCR products for the enzymes tested.

Sensitivity of detection of X fragariae in plant tissue by PCR. X ftagariae was detected in the presence of plant tissue by amplification with primer XF9 and either XFI 11 or XF12 and by nested PCR of the two primer sets. The level of sensitivity of detection of bacterial cells by amplification with one round of PCR was ca. 105 to 10' cfu per ml. The nested PCR increased sensitivity of the assay 1000-fold (Fig. 3-3) and enabled detection of ca. 18 cfu per mi. PCR was used to detect bacteria directly from infected plant tissue both from inoculated strawberry and from plant samples collected in the field. The PCR product was generated from a single lesion on plant material ground and processed representing a bacterial population of 10' or above as determined by dilution plating. Identification of X fragariae was confirmed by REA of PCR products. The nested technique was not necessary to detect bacteria from a lesion on leaf tissue as a single round of PCR amplification was sufficient to obtain a positive result.









37














HaeIII Cf o





100 bp




















Figure 3-2. NuSieve agarose gel showing the restriction patterns generated after digestion of the PCR product amplified with the primers XF9 and XF 12 from the genomic DNA of strains of Xanthomonasfragariae. The PCR product was digested by restriction endonucleases Cfo I and HaeIII and the molecular weight marker XI from Boehringer Mannheim is shown.









38
















XF9 and XF 11 Nested XF9 and XF12 500 bp






87654321 87654321 Log cfu per ml










Figure 3-3. Agarose gel showing the sensitivity of detection by PCR and nested PCR in the presence of plant tissue. The genomic DNA from various concentrations of cells was extracted and amplified by primers XF9 and XF 11 and used in a second round of nested amplification with primers XF9 and XF 12.









39

Detection of bacteria on plants in oversummering experiments. All plants exhibited visible symptoms of angular leaf spot at planting. However, by 34 days after planting, symptoms of angular leaf spot were not visible on plants (Table 3-1). Bacteria were recovered on the selective medium and colonies of X. fragariae were identified by morphology, resistance to rifampicin, DNA amplification by the XF-specific primers, and REA. Bacteria were recovered on selective medium from a few samples throughout the experiment, including the final date, indicating that bacteria were viable on plants throughout the summer.

Xanthomonas fragariae was detected from leaf samples by nested PCR on each sample date. The percentage of leaf samples which were positive by the nested PCR assay ranged from 20% to 100% for the sample dates. From the leaves, all samples were initially positive but, by 51 days after planting, the number of positive samples declined to 20%. By 92 days after planting, all samples were again positive. A single round of amplification was occasionally sufficient to identify positive samples. In the crown of plants, bacteria were detected by PCR amplification on all sample dates except one. REA analysis of the PCR products confirmed banding patterns typical ofX. fragariae. Plants inoculated and placed in the growth chamber at the favorable temperatures for disease development were positive at each sampling date except one. Non-inoculated plants from the greenhouse were positive by this technique in approximately 20% of the samples.












Table 3-1. Percentage of strawberry plants positive for the presence of Xanthomonasfragariae detected by PCR and a selective medium in summer nurseries from June through September at two locations in Florida

Leaves Crown

Location DAPa NonIb Inocc Primerl 2 Prime2' Medium' Symptoms NonI Inoc Primerl Primer2 Medium
Dover 0 - + 60 100 TC + - + 20 40 TC
11 - + 0 40 TC + - + 0 40 TC 31 - + 0 20 0 + - + 0 40 0 40 + + 0 20 NS - - + 0 0 NS 66 - 0 60 0 - - + 0 40 0 85 - + 20 100 60 - - + 20 100 20 92 - + 40 100 40 - - + 0 80 40

Bradenton 0 - + 60 100 TC + - + 0 40 TC
20 - + 20 20 20 + + + 0 60 0 C
34 + + 0 20 NS - - + 0 40 NS 40 - + 0 20 20 - - + 0 100 20 54 - + 0 40 0 - - + 0 40 0 68 + + 20 100 60 - + + 20 80 20 82 - + 20 80 40 - + + 20 80 40

a Days after planting.
bNon-inoculated plants in greenhouse. - = Negative by PCR amplification, + = Positive by PCR amplification. c Inoculated strawberry plants in growth chamber at 22' C. - = Negative by PCR amplification, + = Positive by PCR amplification. d Percentage of samples amplified by primer set XF9 and XF 11. "Percentage of samples amplified in nested PCR with primers XF9 and XF12 with sample from first round. 'Wilbrink's agar plus 100 mg/ml rifampicin and 0.05% Bravo; TC= too contaminated to distinguish individual colonies;
NS= Not assayed; 0 - 100 = Percentage of plates with colonies of X fragariae.









41

Discussion



Leite et al. (1994) used the sequence variation within the hrpB operon among plantpathogenic xanthomonads to select primers with different specificities and this approach was successful in our studies to identify primers specific to X fragariae. The hrp-primers RST2 and RST3 amplified DNA from all strains of X fragariae and analysis of the PCR products by REA showed no polymorphism within this region. The homology of the amplified region presented a good site to select primers universal to X. fragariae. Primers XF9, XF 11, and XF12 designed from these unique sites were specific for amplification of DNA only from strains ofX fragariae. The primer XFI0 was responsible for the non-specific amplification of a strain of X c. pelargonii in preliminary tests and therefore was not used in later experiments. Interestingly, an ELISA test developed for identification of X. fragariae in vitro cross reacts to some related bacteria including X. c. pelargonii (Maas, 1995). Perhaps a relationship between X fragariae and X. c. pelargonii exists which is not reflected by traditional taxonomic techniques (Hildebrand et al., 1990; Vauterin et al., 1994; Maas, 1995).

The level of sensitivity for most detection techniques for specific bacteria (Pickup, 1991; Schaad et al., 1982 ) are comparable to the level which was achieved with a single round of PCR amplification. PCR was reported to be more sensitive than ELISA (Leite et al., 1995) and with our primers, a single round of PCR is 10 times more sensitive than the ELISA test for X fragariae developed by Rowhani et al. (1994). In addition, their antibodies cross-reacted with another xanthomonad whereas we were able to eliminate the non-specific reaction by discontinuing use of primer XF 10. Although this level of detection is adequate









42

to detect bacteria in symptomatic leaf tissue, we were concerned with the detection of bacteria from asymptomatic tissue. The X. fragariae primers were suitable for use in the nested technique and the sensitivity of the PCR reaction was increased by 1,000-fold. This level of detection was achieved in assays to detect the bacterium in the presence of plant tissue and is well below the number of cells needed to cause visible lesions. Therefore, the nested technique is applicable for detection of bacteria in association with asymptomatic tissue.

Cross-contamination among samples and contamination of PCR reagents are problems associated with the nested technique (McManus and Jones, 1995). We experienced false positives in a number of initial experiments and finally determined the cause to be contaminated mineral oil. Although we eliminated the need for mineral oil overlay by using a hot bonnet, extreme care is needed to avoid contamination and subsequent false positives. In our experiments, negative control samples were always included for each amplification round of the nested assay. Aerosols may also contribute to false positives and care must be taken to prepare the mixtures for PCR in a sterile environment and negative control samples should be included to check for aerosol contamination. In the field experiments, a problem encountered with the nested technique was amplification of bacterial DNA from the negative control plants. Although these plants were physically isolated from a contamination source during the experiments; they were initially obtained from fields at GCREC-Dover where plants infected with angular leaf spot were located. The plants may have been infested with the bacterium or cross-contamination of samples may have occurred during preparation of plant samples from handling infected material or using non-sterile instruments. Aerosols may









43

also have been a source of contamination of PCR samples (Innis et al., 1990). However, contamination from these sources would be detected by amplification in the negative controls. Our negative controls were included in every PCR run and were consistently negative. Therefore contamination of PCR reagents or aerosols would have been ruled out.

Confirmation that the PCR product is from amplification of the target DNA is possible by REA (Leite et al. 1994). The REA was used to identify PCR products amplified by the hrp-primers from cells ofX. c. vesicatoria added to seed washing of tomato and pepper (Leite et al. 1995). Similarly, we applied REA to identify PCR products amplified by the XFprimers. The profiles of the restricted PCR products were distinct for each enzyme. The profiles from REA should be different for unrelated organisms since it would be highly unlikely that the same restriction sites would exist for two heterologous pieces of DNA (Leite et al,. 1995).

The nested PCR technique was useful to detect bacteria on strawberry plants in nurseries in Florida. Symptoms and recovery of the rifampicin-marked strain on a selective medium were not useful to identify bacteria when populations were extremely low. However, recovery of the rifampicin-marked strain from some samples indicated that the bacteria were viable throughout the summer and that PCR was detecting viable bacterial cells. Visible symptoms on plants disappeared soon after placement in the field and recovery on media was difficult due to the slow-growing nature ofX. fragariae and overgrowth by contaminants. While bacterial populations were not enumerated, the levels were deduced to be below 10' cfu per ml, since bacteria were detected in the nested PCR but not by single round of amplification. This level of sensitivity would be useful to screen asymptomatic nursery plants









44

to detect plants contaminated with X fragariae. This is a concern of nursery-plant producers and regulatory agencies of plant shipment (Maas, 1995). Currently, plants from nurseries in California are being screened by this technique to determine its usefulness in such an application.

The field studies have implications regarding the disease cycle of angular leaf spot in Florida. The decline and later increase in the number of positive samples through the summer would indicate that populations of the bacteria declined throughout the summer and increased when more favorable conditions in the form of cooler weather occurred. The bacterial populations did not completely die. Therefore, to eliminate disease, the production of plants in nurseries in Florida would have to begin with plants free of the bacterium.

Researchers have reported the systemic movement of X. fragariae in plants (Hildebrand et al., 1967; Milholland et al., 1993). In our research, inoculation of strawberry plants resulted in survival of bacteria on leaves and the crown for extended periods under conditions not optimum for growth of the bacteria. In addition, bacteria were detected by PCR technique on daughter plants in the field. Dissemination to the daughter plants could have been due to either systemic movement through the vascular system of the runner or dispersal of bacterium by mechanical means.














CHAPTER 4
GENOMIC RELATEDNESS OF XANTHOMONAS FRAGARIAE ON STRAWBERRY BY FATTY ACID METHYL ESTER AND RESTRICTION LENGTH FRAGMENT POLYMORPHISM ANALYSES


Introduction



Xanthomonasfragariae causes angular leaf spot disease on strawberry (Fragaria species and Fragaria x ananassa Duch. cultivars). While historically the disease has not been a major deterrant in strawberry production, the disease is becoming more important because of an increase in prevalance of the disease in fruit production fields in Florida and the lack of effective control measures (Maas, 1995; C. Chandler, pers. comm). Losses in yield are caused by this disease (Epstein, 1966; Howard, 1971). In addition, regulatory issues regarding the transportation of infected plant material may impact the nursery plant industry. International movement of infected plants is blamed for the introduction of angular leaf spot into Greece and New Zealand (Panagopoulos et al., 1978; Dye and Wilkie, 1973). The European Plant Protection Organization listed X fragariae as a quarantine pest and the FAO/IPGRI recognized it as a potential risk in international movement of strawberry germplasm (Maas, 1995). Nursery-plant producers in the United States and Canada are pressured to provide disease-free plants by buyers from foreign markets and by farmers who refused to buy infected transplants.



45









46

The epidemiology of the X fragariae must be understood before effective control strategies can be devised. The delineation of bacterial populations is imperative to such studies. The relationship of X fragariae to other members of the genus Xanthomonas has been examined (Hildebrand et al., 1990; Hodge et al., 1992; Vauterin et al., 1994), but the genetic variability within the species has not been reported (Maas, 1995). The information from studies to compare strains ofX fragariae at the genetic level could be useful to identify the origins and spread of specific strains or genetic types. This information could be applied in international tracking of pathogen populations for quarantine programs. In addition, the identification of genetic types is a prerequisite to identify sources of resistance in strawberry. Strawberry cultivars exhibit levels of susceptibility or tolerence to X fragariae; only F. moschata Duch. appeared to be immune (Hazel, 1981; Hazel and Civerolo, 1980; Kennedy and King, 1962b). A screening program to identify genes for resistance must incoporate representives of the genetic variants in the screening process otherwise the resistance may be overcome quickly by genetic variants.

Studies to characterize populations of bacteria have used biochemical and molecular biological techniques. Protein staining and fatty acid analysis have been useful to detect differences at the metabolic level within species ofXanthomonas (Bouzar et al. 1994; Stall et al, 1994; Graham et al., 1990; Hodge et al., 1992). The polymerase chain reaction (PCR) has been used to analyze differences between pathovars and strains of bacteria (Hensen and French, 1993). Primers specific to the hrp-gene cluster ofXanthomonas campestris pv. vesicatoria amplifed genomic DNA and restriction enzyme analysis of the PCR product differentiated between many pathovars and species of Xanthomonas (Leite et al, 1994 and









47

1995). Primer sets from conserved repetitive bacterial DNA elements generated genomic fingerprints which were used to differentiate bacteria (De Bruijn, 1992; Louws et al., 1994). Restriction length fragment polymorphism (RFLP) of the bacterial genome digested with rarecutting restriction endonuclease and resolution by pulsed field gel electrophorisis (PFGE) has been used to type many bacterial strains (Cooksey and Graham, 1989; Egel et al., 1991; Smith et al., 1995).

In this study, a collection of strains ofX fragariae from the United States and Canada were examined for genetic diversity. Analyses was by RFLP profiles generated by PFGE and by fatty acid methyl ester (FAME).



Materials and Methods



Bacterial strains. Strains of X. fragariae used in this study are listed in Table 4-1. Strains were previously rated for pathogenicity on strawberry (Roberts, unpublished). Geographic origin denotes origin of infested plant material. Strains were cultured on Wilbrink's medium (WB)(Koike, 1962) at 240 C and long term storage was in 15% glycerol at -70' C. Single colonies were transfered to nutrient broth (Difco, Detroit, MI) and shaken for 16 to 20 h at 200 rpm at 240 C.

Restriction fragment length polymorphism, The method used for restriction endonuclease analysis was described by Egel et al. (1991) and Cooksey and Graham (1989) except for the following modifications. Cells (1.5 ml of 5 x 10' cfU per ml) were washed in 1 ml of TE (10 mM Tris, ImM EDTA, pH 8.0) and resuspended in 0.5 ml of TE. An equal









48

Table 4.1. The geographic source, collection date, and group determined by fatty acid methyl ester (FAME) and restriction fragment length polymorphism (RFLP) for strains used in this study.

Geographic
Strain Source' Year Sourceb FAMEC RFLP
1238 CA 1990 ARC 2 D 1240 CA 1990 ARC 2 D 1241 CA 1990 ARC 8 B 1242 CA 1990 ARC 8 B 1243 CA 1990 ARC 8 B 1245 CA 1990 ARC 4 B 1246 CA 1990 ARC 8 1249 CA 1990 ARC 4 1250 CA 1990 ARC 5 C 1290 CA 1989 ARC C 1291 CA 1989 ARC 1 B 1293 CA 1989 ARC 7 B 1295 CA 1989 ARC 8 1296 CA 1989 ARC 8 B3 1298 CA 1989 ARC 8 B 1424 FL 1992 ARC C 1425 FL 1992 ARC 3 B2 1426 FL 1992 ARC B 1427 FL 1992 ARC 3 B 1428 FL 1992 ARC 3 B2 1429 FL 1992 ARC 3 B 1431 FL 1992 ARC B 1514 CA 1993 ARC C 1515 CA 1993 ARC 6 B3 1516 NC 1993 ARC B2 1517 NC 1993 ARC 7 B3 1518 NC 1993 ARC 8 B 1519 NC 1993 ARC B 1520 CA 1993 ARC 5 C 1523 CA 1993 ARC 5 C 1524 CA 1993 ARC 5 C 1525 CA 1993 ARC C 1526 CA 1993 ARC 8 B

a Geographic origin of plant material from which bacteria were isolated. CAN=Canada. bARC= A. R. Chase; ATCC= American Type Culture Collection. 'Blank space indicates that group was not determined.









49

Table 4.1 continued. The geographic source, collection date, and groups determined by fatty acid methyl ester (FAME) and restriction fragment length polymorphism (RFLP) for strains used in this study.

Geographic
Strain Source" Year Sourceb FAME RFLP
1532 FL 1993 ARC 3 1533 WI 1993 ARC 5 B 1534 WI 1993 ARC B3 100 FL 1993 This study 1 B 101 FL 1993 This study 1 103 FL 1993 This study B 104 FL 1993 This study 1 B 105 FL 1993 This study 3 B 106 FL 1993 This study 3 107 FL 1993 This study 1 108 FL 1993 This study 1 B 113 FL 1993 This study 5 B3 114 FL 1993 This study 3 B 115 FL 1993 This study 4 B 116 CAN 1993 This study B 117 CAN 1993 This study B3 119 CAN 1993 This study 9 D 124 CAN 1993 This study B 125 CAN 1993 This study 6 126 CAN 1993 This study 4 127 CAN 1993 This study 6 128 CAN 1993 This study 4 B 129 CAN 1993 This study 6 B3 138 CAN 1993 This study 6 B 146 CAN 1993 This study 4 B1 153 CAN 1993 This study 6 B 33239 MN ATCC 9 A

a Geographic origin of plant material from which bacteria were isolated. CAN=Canada. bARC= A. R. Chase; ATCC= American Type Culture Collection. CBlank space indicates that group was not determined.









50

volume of 1% Seakem Gold agarose solution (10 mM Tris [pH 8.0], 10 mM MgCl2, 0.1 mM EDTA [pH 8.0], 1 % Seakem Gold agarose (FMC BioProducts, Rockland, ME) [wt/vol] in sterile filtered water) was added. Plugs were made, lysed, washed, and stored as described by Egel et al. (1991). Two sizes of plugs were utilized. A 4 x 8 mm slice of plug was restricted in a volume of 200 gl of restriction buffer (as recommended by manufacturer, Promega, Madison, WI) and inserted in wells made with a 10-well comb (Bio-Rad, Richmond, CA). A 4 x 4 mm square piece of plug was restricted in a volume of 100 pl and placed in wells made by a 20-well comb (Bio-Rad). Restriction enzyme was added at the following concentrations: XbaI at 40 U; SpeI at 30 U; VspI at 30 U (Promega). The pieces were placed into wells in a 1.2% GTG agarose gel made with 0.5X TBE. Wells were sealed with the 1% Seakem Gold solution. The gel was placed in a Bio-Rad CHEF DR II unit containing 1.8 liters of 0.5x TBE and run at 200 V (15V/cm of gel). Pulsed times for plugs digested with XbaI or Spel were at 4 s for I h followed by 8 sec for 18 h. Plugs digested with Vspl were pulsed at 4 s for I h and subsequently at 12 s for 17 h. Lambda DNA in 48.5 KB concatamers (FMC BioProducts) was used in the first and last lanes of each gel. Gels were stained in 0.5 mg ofethidium bromide per liter and photographed with type 55 Polaroid film.

The position of bands were assessed visually or by analysis with the Gelmeas computer program. Similarity values were calculated as described by Egel et al. (1991) with the mathematical equation proposed by Nei and Li (1979) based upon the proportion of shared DNA fragments. The estimate of the number of nucleotide substitutions per site was used to calculate the genetic divergence by the iterative method of Nei (1987) with the program in SAS described by Leite et al. (1994). The KITSCH program from the PHYLIP









51

computer pakage (Felsenstein, 1995) was used to create a rooted phylogenetic tree by the Fitch-Margoliash method (Fitch and Margoliash, 1967). The input data was as described for the combined SpeI, XbaI, and VspI digestion data (Stall et al., 1994).

Fatty acid composition. Strains of X fragariae were inoculated onto Trypticase Soy Broth agar (TSBA) and grown for 48 h at 240 C. Conditions were changed from the standard MIDI procedures which required that cells be grown on TSBA for 24 hours at 280 C because strains ofX fragariae produced insufficient growth at these conditions. Cellular fatty acids were extracted and derivatized to their fatty acid methyl esters as described (Sasser, 1990b). A library comprised of strains of X. fragariae was created using the MIDI Library Generation System (LGS), software version 3.3. FAMEs were analyzed by the MIDI Microbial Identification System, software version TSBA 3.5. The qualitative and quantitative differences in the fatty acid profiles were used to compute the Euclidian distance to each strain. Strains within six Euclidian distance units, the cut-off for subspecies (Sasser, 1990a), were grouped in the same cluster.



Results



RFLP-PFGE. Restriction endonucleases XbaI and SpeI generated genomic DNA fragments from 5 to 400 kb (Fig. 4-1). Typically, a strain profile contained ten DNA fragments greater than 100 kb. Analysis of the 52 strains resulted in four RFLP groups, designated A through D, which were identical by these two enzymes. Group A contained the ATCC strain only and represented 2% of the strains. The B, C, and D groups had 77%, 15%









52

and 6% of strains, respectively. A third endonuclease, VspI, generated fragments of DNA of the appropriate size for PFGE analysis and the profiles by this endonuclease separated strains of the A, C, and D groups identical to the previously determined groups. Analysis of the B group with the VspI endonuclease subdivided it into three subgroups, denoted B 1, B2, and B3. The BI contained 13%, B2 contained 20%, and B3 contained 67% of the B strains tested. Designation of RFLP groups for strains are summarized in Table 4-1.

The dendrogram derived from RFLP analysis of these strains showed the B groups were very closely related by within 0.006 genetic distance units (Fig. 4-2). The C group was the most closely related to the B groups, then group A, and finally group D. The D group was the most remote from the B group at 0.01864 genetic distance units. A strain ofX. c. vesicatoria was used as the statistical outlier and fell at 0.045 genetic distance units from the D group.

Fatty Acid Methyl Ester Analysis. The 50 strains of X. fragariae were placed into nine subgroups based upon the MIDI "10%" rule (Sasser, 1990b). The subgroups for individual strains are summarized in Table 4-1. The majority of strains was identified as six closely related subgroups which can be visualized quantitatively by three major acids, 16:1 w 7 cis, 15:0 Anteiso, and 15:0 iso (Fig.4-3). The 15:0 iso fatty acid comprised approximately 34% to 54% of FAME profile. The ATTC type strain and three closely related strains were qualitatively differentiated by the absence of palmitic acid, 16:1 w 7 cis.

The dengrogram unweighted pair analysis (Fig. 4-4) separated strains into four clusters at the Euclidian distance of six units. The four clusters, designated FAME groups









53









































Figure 4-1. Agarose gel showing the restriction fragment length polymorphisms of genomic DNA of strains ofXanthomonasfragariae after restriction with a rare-cutting endonuclease, VspI, and separated by pulsed-field gel electrophoresis. Lambda marker in 48.5 kb concatamers is shown in side lanes.






54


B3

BI

B2


C


A


--D

X.c. vesicatoria I I I
0.04 0.02 0 Genetic Distance



Figure 4-2. Relationship of groups of Xanthomonas fragariae analyzed by restriction fragment length polymorphisms.







55





20

% 16:1 w 7 cis







5
18

55
% 15:0 ANTEISO
8 30 % 15:o ISO











Figure 4-3. The groups of Xanthomonas fragariae distinguished by qualitative and quantitative differences in fatty acid methyl ester profiles graphed by three major acids.









56

1 through 4, contained 16%, 6%, 4%, and 74% of the strains, respectively. The largest cluster, FAME group 4, could be subdivided at the Euclidian distance of 4 units into two groups. The ATCC type strain 33239 and one other strain comprised FAME 3 group.



Discussion



This research represents the first effort to analyze genetic variants of strains within the species of X fragariae. The RFLP-PFGE and FAME analyses identified genetic variants within the population. Methods of RFLP including PFGE have distinguished between closely related strains of pathovars within the genus Xanthomonas campestris(Egel et al., 1991; Gabriel et al., 1988; Hartung and Civerolo; 1989, Graham et al., 1990). In these works, the importance of the diverse nature of the endonuclease sites as related to strain characteristics such as host range, geographic origin, and pathogenicity was discussed. Unfortunately, the relationship of pathogenicity of the X fragariae strains to the genetic groupings cannot be examined. Differences in pathogenicity among strains of X.fragariae were not detected by inoculation of the 52 strains onto two cultivars of strawberry. Studies of bacterial populations on two cultivars which appeared to have different levels of susceptibility in preliminary tests were also inconclusive (Roberts, unpublished). Nor have other researchers reported differences in pathogenicity among strains ofX fragariae. Such information might have been useful to determine significance of the RFLP-PFGE and FAME groups as related to pathogenicity.









57


Euclidian Distance
0.0 3.55 7.05 10.5 14.1 Cluster "10%" I I I




1 3



2
2
9


3 7


4
4a 5
6
7





1I
4b 5
8




Figure 4-4. Dendrogram of cluster analysis of fatty acid methyl ester profiles showing the four clusters (1-4b) of strains ofXanthomonasfragariae. The designation of strains into the groups determined by the MIDI "10%" rule are shown for comparison. The Euclidian distance of six is the cutoff point for group determination by cluster analysis.









58

The two methods used in this research might be useful to examine the evolution of the pathogen in future populations. The identification of two groups which contained only four strains total, including the ATCC type strain, from a population of 50 strains was of interest. By RFLP analysis, group A contained only the ATCC type strain from Minnesota which was collected over twenty years ago. However, by FAME cluster analysis, one other strain, XF 119, which was isolated from an infected plant from Canada in 1993, grouped with the ATCC type strain. In RFLP-PFGE analysis, the XF 119 profile was the same as the D group which contained strains XF1238 and XF1240, isolated from samples from California in 1990. Dendrogram analysis places these two groups close to each other. By FAME, these four strains lacked the 16:1 w 7 cis fatty acid which distinguished them from all the other X. fragariae. It is interesting to speculate that perhaps strain XF119 represent a 'bridge' between the A and D RFLP groups because of its intergroup relationship in that its RFLP profile is D but by FAME it is closer to the A group. It would be of further interest to examine future populations to determine the fate of the two groups represented by only four strains.

The majority of strains was represented by the B and C groups. An extensive survey of the pathogen population in the United States and Canada would be useful to determine whether the population ratios of the various genetic variants are remaining relatively the same. Likewise, it would be interesting to determine the distribution of the three subdivisions within the B group. The cultivars of strawberry in commercial production are changed frequently (C. Chandler, pers. comm.). This constant change in genotypes should influence the genetic composition of the population of X fragariae if genes for resistance to angular leaf spot are









59

'lost' or 'found' during the development of new cultivars. In the Philippines, the relative populations of two races of bacterial blight on rice were followed for 10 years. Researchers recorded a decline in the prevalence of the predominant race and a concurrent increase in another race. This change frequency of race, or bacterial genotype, occurred after the introduction of a gene for resistance in cultivated rice (Mew et al., 1992). Information regarding changes in the dominant genotype of both the plant and bacteriapopulations would be useful to plant breeders.

The relationship between geographical origin and genetic variants is unclear because of the transportation of infected plants. Plants are shipped from California to Canada where they are propagated and sent to Florida for field production each season. Infection of plants may have occurred at any point of the route, however symptomatic transplants from Canada are shipped to Florida. A single type of genetic variant was not found to be associated with plant material from a particular region of the U. S. or Canada. Likewise, international movement of infested plants will make it difficult to determine if endemic populations of the organism exists outside the U. S.. The origin of this disease appears to be the United States (Maas, 1995).














CHAPTER 5
DISCUSSION


Field experiments which examined epidemiological aspects of angular leaf spot on strawberry established a decrease in yield on strawberry plants due to angular leaf spot. This report is the first to quantify the reduction in yield due to this disease. The 8- to 10% loss determined in these studies is much lower than the 70- to 80% loss estimated by Epstein (1966) in fruit production fields in Wisconsin. Production in northern regions of the U. S. is usually perennial and the age of the plants was not given. The loss was reported on the cultivar 'Sparkle' which is not produced commercially in Florida. Therefore, we are unable to compare our yield loss to his report. Howard (1971) also reported unquantified yield losses due to this disease but on cultivars which are no longer in commercial production. In our studies, the average yield loss observed both years was very similar despite the differences in the disease levels and the base yields between the two seasons. In 1995, the disease severity was 60% of the amount estimated in the previous season; however, the yield loss was actually higher by 2% for marketable yield. The 10% reduction in yield represents an estimated loss of 1233 kg of berries per acre. Strawberry fruit has a very high cash value and a decrease in revenue of 8- to 10% represents a significant economic loss to producers.

Strawberry producers in Florida currently try to control angular leaf spot by application of copper compounds and by avoidance of the disease. An understanding of the



60









61

epidemiology of angular leaf spot is necessary to determine the effectiveness and practicability of any control method. The survey of farmers' fields and the cold storage facility established that transplants arrive from northern nurseries infected with angular leaf spot at fairly high disease incidence. Therefore, disease is introduced into the field on infected transplants. In our surveys, by the sampling method used in the first season, more plants were assessed for symptoms of angular leaf spot in the field compared to the number of plants in the random sampling of boxes in cold storage. This probably accounts for identification of diseased plants in nearly all of the farmer's fields versus less than in 25% of boxed plants. In addition, plants were examined until disease was found in the field which skewed the randomness of the survey.

In field experiments, the spread from inoculated plants to noninoculated plants in nearby experimental plots was minimal. Thus, the incidence of plants with angular leaf spot in the field likely occurred from inoculum already present on the transplants. In the field experiments, if inoculum was present from another source, such as debris in the soil or an alternate host, most likely the disease on the noninoculated plants would have been more general and appeared earlier. Disease severity on noninoculated plants in the experimental plots was extremely low and most plants remained free of angular leaf spot. For the disease that occurred on noninoculated plants, the inoculum was probably transported mechanically from inoculated plants during harvests or disease readings.

Progress of angular leaf spot developed similar to curves of pathogens described on other crops (van der Plank, 1963). A decrease in disease severity was seen about midway through the season in 1994. This reduction, or negative infection rate, was due to growth









62

by strawberry plants which diluted the amount of disease relative to the total leaf area. The dilution effect of healthy plant tissue was described by van der Plank (1963). In 1995, a negative infection rate was seen very early in the epidemic. This was because the symptoms of the disease almost disappeared from the field as the affected leaves died and were lost. The difference in disease progress between the two seasons did not appear to be from differences in rainfall since mean rainfall was approximately the same both years (data not shown). Mean temperatures also were similar; however the number of days at temperatures below 100 C in November and December 1993 were much greater than in 1994. Another difference between the seasons was that sprinkler irrigation was reduced in 1995 because of the change to drip irrigation. Overhead irrigation, fog, high humidity, and rain are positive factors in disease development and spread (Epstein, 1966; Hildebrand et al., 1967; Kennedy and King, 1962b; Lai, 1978). The trend towards drip irrigation in commercial production fields should have a positive effect to reduce the spread and survival of the bacterium.

Strawberry producers currently apply copper-based compounds to control angular leaf spot despite the phytotoxicity to plants caused by copper. Application of cupric hydroxide plus mancozeb at the lx rate was phytotoxic to strawberry plants (Howard, 1973; this study). In our tests, the greatly reduced 0.1 x rate did not harm plants and significantly reduced disease severity in 1995. The protectant application of bactericide at this rate and spray schedule was intended to reduce the total amount of inoculum and to prevent spread of the disease. The approach of frequent sprays at reduced concentrations may have potential to control this disease. More studies are needed to evaluate different rates, application intervals, and chemical mixtures to achieve maximum disease control while avoiding yield losses. An









63

alternate approach might be to control the disease in the nursery by copper applications. The loss in yield due to phytotoxicity would not be an issue on nursery plants. The sprays would reduce the amount of initial inoculum on transplants and subsequently reduce the amount of disease in fruit production fields. The effect of sprays on the categories of yield were difficult to interpret because of the lack of consistent, significant differences between years.

Leite et al. (1994) used the sequence variation within the hrpB operon among plantpathogenic xanthomonads to select primers with different specificities and this approach was successful in our studies to identify primers specific to X fragariae. The hrp-primers RST2 and RST3 amplified DNA from all strains of X fragariae and analysis of the PCR products by REA showed no polymorphism within this region. The homology of the amplified region presented a good site to select primers universal to X fragariae. Primers XF9, XF 11, and XF12 designed from these unique sites were specific for amplification of DNA only from strains of Xfragariae. The primer XF10 was responsible for the non-specific amplification of a strain of X. c. pelargonii in preliminary tests and therefore was not used in later experiments. Interestingly, an ELISA test developed for identification of X fragariae in vitro cross reacts to some related bacteria including X c. pelargonii (Maas, 1995). Perhaps a relationship between X fragariae and X c. pelargonii exists which is not reflected by traditional taxonomic techniques (Hildebrand et al., 1990; Vauterin et al., 1994; Maas, 1995).

The level of sensitivity for most detection techniques for specific bacteria (Pickup, 1991; Schaad et al., 1982 ) are comparable to the level which was achieved with a single round of PCR amplification. PCR was reported to be more sensitive than ELISA (Leite et al., 1995) and with our primers, a single round of PCR is 10 times more sensitive than the









64

ELISA test for X fragariae developed by Rowhani et al. (1994). In addition, their antibodies cross-reacted with another xanthomonad whereas we were able to eliminate the non-specific reaction by discontinuing use of primer XF10. Although this level of detection is adequate to detect bacteria in symptomatic leaf tissue, we were concerned with the detection of bacteria from asymptomatic tissue. The X. fragariae primers were suitable for use in the nested technique and the sensitivity of the PCR reaction was increased by 1,000-fold. This level of detection was achieved in assays to detect the bacterium in the presence of plant tissue and is well below the number of cells needed to cause visible lesions. Therefore, the nested technique is applicable for detection of bacteria in association with asymptomatic tissue.

Cross-contamination among samples and contamination of PCR reagents are problems associated with the nested technique (McManus and Jones, 1995). We experienced false positives in a number of initial experiments and finally determined the cause to be contaminated mineral oil. Although we eliminated the need for mineral oil overlay by using a hot bonnet, extreme care is needed to avoid contamination and subsequent false positives. In our experiments, negative control samples were always included for each amplification round of the nested assay. Aerosols may also contribute to false positives and care must be taken to prepare the mixtures for PCR in a sterile environment and negative control samples should be included to check for aerosol contamination. In the field experiments, a problem encountered with the nested technique was amplification of bacterial DNA from the negative control plants. Although these plants were physically isolated from a contamination source during the experiments; they were initially obtained from fields at GCREC-Dover where









65

plants infected with angular leaf spot were located. The plants may have been infested with the bacterium or cross-contamination of samples may have occurred during preparation of plant samples from handling infected material or using non-sterile instruments. Aerosols may also have been a source of contamination of PCR samples (Innis et al., 1990). However, contamination from these sources would be detected by amplification in the negative controls. Our negative controls were included in every PCR run and were consistently negative. Therefore contamination of PCR reagents or aerosols would have been ruled out.

Confirmation that the PCR product is from amplification of the target DNA is possible by REA (Leite et al. 1994). The REA was used to identify PCR products amplified by the hrp-primers from cells of X c. vesicatoria added to seed washing of tomato and pepper (Leite et al. 1995). Similarly, we applied REA to identify PCR products amplified by the XFprimers. The profiles of the restricted PCR products were distinct for each enzyme. The profiles from REA should be different for unrelated organisms since it would be highly unlikely that the same restriction sites would exist for two heterologous pieces of DNA (Leite et al,. 1995).

The nested PCR technique was useful to detect bacteria on strawberry plants in nurseries in Florida. Symptoms and recovery of the rifampicin-marked strain on a selective medium were not useful to identify bacteria when populations were extremely low. However, recovery of the rifampicin-marked strain from some samples indicated that the bacteria were viable throughout the summer and that PCR was detecting viable bacterial cells. Visible symptoms on plants disappeared soon after placement in the field and recovery on media was difficult due to the slow-growing nature ofX. fragariae and overgrowth by contaminants.









66

While bacterial populations were not enumerated, the levels were deduced to be below 103 cfu per ml, since bacteria were detected in the nested PCR but not by single round of amplification. This level of sensitivity would be useful to screen asymptomatic nursery plants to detect plants contaminated with X fragariae. This is a concern of nursery-plant producers and regulatory agencies of plant shipment (Maas, 1995). Currently, plants from nurseries in California are being screened by this technique to determine its usefulness in such an application.

The field studies have implications regarding the disease cycle of angular leaf spot in Florida. The decline and later increase in the number of positive samples through the summer would indicate that populations of the bacteria declined throughout the summer and increased when more favorable conditions in the form of cooler weather occurred. The bacterial populations did not completely die. Therefore, to eliminate disease, the production of plants in nurseries in Florida would have to begin with plants free of the bacterium.

Researchers have reported the systemic movement of X. fragariae in plants (Hildebrand et al., 1967; Milholland et al., 1993). In our research, inoculation of strawberry plants resulted in survival of bacteria on leaves and the crown for extended periods under conditions not optimum for growth of the bacteria. In addition, bacteria were detected by PCR technique on daughter plants in the field. Dissemination to the daughter plants could have been due to either systemic movement through the vascular system of the runner or dispersal of bacterium by mechanical means.

This research represents the first effort to analyze genetic variants of strains within the species ofX. fragariae. The RFLP-PFGE and FAME analyses identified genetic variants









67

within the population. Methods of RFLP including PFGE have distinguished between closely related strains of pathovars within the genus Xanthomonas campestris (Egel et al., 1991; Gabriel et al., 1988; Hartung and Civerolo; 1989, Graham et al., 1990). In these works, the importance of the diverse nature of the endonuclease sites as related to strain characteristics such as host range, geographic origin, and pathogenicity was discussed. Unfortunately, the relationship of pathogenicity of the X fragariae strains to the genetic groupings cannot be examined. Differences in pathogenicity among strains of X fragariae were not detected by inoculation of the 52 strains onto two cultivars of strawberry. Studies of bacterial populations on two cultivars which appeared to have different levels of susceptibility in preliminary tests were also inconclusive (Roberts, unpublished). Nor have other researchers reported differences in pathogenicity among strains ofX fragariae. Such information might have been useful to determine significance of the RFLP-PFGE and FAME groups as related to pathogenicity.

The two methods used in this research might be useful to examine the evolution of the pathogen in future populations. The identification of two groups which contained only four strains total, including the ATCC type strain, from a population of 50 strains was of interest. By RFLP analysis, group A contained only the ATCC type strain from Minnesota which was collected over twenty years ago. However, by FAME cluster analysis, one other strain, XF 119, which was isolated from an infected plant from Canada in 1993, grouped with the ATCC type strain. In RFLP-PFGE analysis, the XFI 19 profile was the same as the D group which contained strains XF1238 and XF1240, isolated from samples from California in 1990. Dendrogram analysis places these two groups close to each other. By FAME, these four









68

strains lacked the 16:1 w 7 cis fatty acid which distinguished them from all the other X. fragariae. It is interesting to speculate that perhaps strain XF119 represents a 'bridge' between the A and D RFLP groups because of its intergroup relationship in that its RFLP profile is D but by FAME it is closer to the A group. It would be of further interest to examine future populations to determine the fate of the two groups represented by only four strains.

The majority of strains was represented by the B and C groups. An extensive survey of the pathogen population in the United States and Canada would be useful to determine whether the population ratios of the various genetic variants are remaining relatively the same. Likewise, it would be interesting to determine the distribution of the three subdivisions within the B group. The cultivars of strawberry in commercial production are changed frequently (C. Chandler, pers. comm.). This constant change in genotypes should influence the genetic composition of the population of X. fagariae if genes for resistance to angular leaf spot are 'lost' or 'found' during the development of new cultivars. In the Philippines, the relative populations of two races of bacterial blight on rice were followed for 10 years. Researchers recorded a decline in the prevalence of the predominant race and a concurrent increase in another race. This change in frequency of race, or bacterial genotype, occurred after the introduction of a gene for resistance in cultivated rice (Mew et al., 1992). Information regarding changes in the dominant genotype would be useful to plant breeders.

The identification of a relationship between geographical origin and genetic variants is unclear because of the transportation of infected plants. Plants are shipped from California to Canada where they are propagated and sent to Florida for field production each season.









69

Infection of plants may have occurred at any point of the route, however symptomatic transplants from Canada are shipped to Florida. A single type of genetic variant was not found to be associated with plant material from a particular region of the U. S. or Canada. Likewise, international movement of infested plants will make it difficult to determine if endemic populations of the organism exists outside the U. S.. The origin of this disease appears to be the United States (Maas, 1995).














LIST OF REFERENCES

Alippi, A. M., Ronco, B. L., and Carranza, M. R. 1989. Angular leaf spot of strawberry, a new disease in Argentina: Comparative control with antibiotics and fungicides. Adv. Hort. Sci. 3:3-6.

Ausubel, F. M., Brent, R. Kingston, R. E., Moore, D. D., and Seidman, K. 1987. Current Protocols in Molecular Biology. John Wiley & Sons, New York.

Bouzar, H., Jones, J. B., Stall, R. E., Hodge, N.C., Minsavage, G. V., Benedict, A. A., and
Alvarez, A. M. 1994. Physiological, chemical, serological, and pathogenic analyses of a worldwide collection of Xanthomonas campestris pv. vesicatoria strains. Phytopathology 84:663-671.

Conway, K. E., Motes, J. E., Bostian, B., Fisher, C. G., and Claypool, P. L. 1987. Cercospora blight development on asparagus fern and effects of fungicides on disease severity and yield. Plant Dis. 71:254-259.

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.

De Bruijn, F. J. 1992. Use of repetitive (repetitive extragenic palindromic and enterobacterial repetitive intergeneric consensus) sequences and the polymerase chain reaction to fingerprint the genomes of Rhizobium meliloti isolates and other soil bacteria. Appl. Environ. Microbiol. 58:2180-2187.

Dye, D. W., and Wilkie, J. P. 1973. Angular leafspot of strawberry in New Zealand. N. Z. J. Agr. Res. 16:311-314.

Egel, D. S., Graham, J. H., and Stall, R. E. 1991. Genomic relatedness of Xanthomonas campestris strains causing diseases of citrus. App. Environ. Microbiol. 57:2724-2730.

Epstein, A. H. 1966. Angular leaf spot of strawberry. Plant Dis. Rep. 50:167.

Felsenstien, J. 1995. PHYLIP. Phylogeny inference package, ver.3.57. Department of Genetics, University of Washinton, Seattle, WA.



70









71

Fitch, W. C, and Margoliash, E. 1967. Construction of phylogenetic trees. Science 155: 279-284.

Fry, W. E. 1975. Integrated effects of polygenic resistance and a protective fungicide on development of potato late blight. Phytopathology 65:908-911.

Fry, W. E., Bruck, R. I., and Mundt, C. C. 1979. Retardation of potato late blight epidemics by fungicides with eradicant and protectant properties. Plant Dis. Rep. 63:970-974.

Gabriel, D. W., Hunter, J. E., Kingsley, M. T., Miller, J. M., 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.

Hartung, J. S., and Civerolo, E. L. 1989. Restriction fragment length polymorphisms dintinguish Xanthomonas campestris strains isolated from Florida citrus nurseries from X. c. pv. citri. Phytopathology 79:793-799.

Hazel, W. J. 1981. Xanthomonas fragariae, cause of strawberry angular leaf spot: Its growth, symptomatology, bacteriophages, and control. Ph.D. thesis. University of Maryland, College Park. 119 p.

Hazel, W. J., and Civerolo, E. L. 1980. Procedures for growth and inoculation of Xanthomonas fragariae, causal organism of angular leaf spot of strawberry. Plant Dis. 64:178-181.

Henson, J. M., and French, R. 1993. The polymerase chain reaction and plant disease diagnosis. Ann. Rev. Phytopathol. 31:81-109.

Hildebrand, D. C., Palleroni, N. J., and Schroth, M. N. 1990. Deoxyribonucleic acid relatedness of 24 xanthomonad strains representing 23 Xanthomonas campestris pathovars and Xanthomonasfragariae. J. Appl. Bacteriol. 68:263-269.

Hildebrand, D. C., Schroth, M. N., and Wilhelm, S. 1967. Systemic invasion of strawberry by Xanthomonasfragariae causing vascular collapse. Phytopathology 57:1260-1261.

Hodge, N. C., Chase, A. R., and Stall, R. E. 1992. Diversity of four species of Xanthomonas as determined by cellular fatty acid analyses. (Abst.) Phytopathology 82:1153.

Howard, C. M. 1971. Occurrence of strawberry angular leaf spot, Xanthomonasfragariae, in Florida. Plant Dis. Rep. 55:142.









72

Howard, C. M., and Albregts, E. E. 1973. Strawberry report. Fungicide and Nematicide Tests 29:47.

Howard, C. M., Overman, A. J., Price, J. F., and Albregts, E. E. 1985. Diseases, nematodes, mites, and insects affecting strawberries. Florida. Agric. Experiment Stations, Inst. of Food and Agric. Sci., Univ. of Fl., Bulletin 857.

Innis, M. A., Gelfand, D. H., Sninsky, J. J., and White, T. J. 1990. PCR Protocols: A Guide to Methods and Applications. Academic Press, Inc., San Diego, CA.

Jones, A. L. 1995. A stewardship program for using fungicides and antibiotics in apple disease management programs. Plant Dis. 79:427-432.

Jones, J. B., Woltz, S. S., Kelly, R. O., and Harris, G. 1991. The role of ionic copper, total copper, and select bactericides on control of bacterial spot of tomato. Proc. Fla. State Hort. Soc. 104:257-259.

Kennedy, B. W., and King, T. H. 1962a. Angular leaf spot of strawberry caused by Xanthomonasfragariae sp. nov. Phytopathology 52:873-875.

Kennedy, B. W., and King, T. H. 1962b. Studies on epidemiology of bacterial angular leaf spot on strawberry. Plant Dis. Rep. 46:360-363.

Koike, H. 1965. The aluminum-cap method for testing sugarcane varieties against leaf scald disease. Phytopathology 55:317-319.

Lai, C. M. 1978. Bacterial disease of strawberry. Plant Path. Reports. California Dept. Of Food and Agric. Lab Services. Vol. II.

Leite, R. P., Egel, D. S., and Stall, R. E. 1994. Genetic analysis of hrp-related DNA sequences of Xanthomonas campestris strains causing diseases of citrus. Appl. Environ. Microbiol. 60:1078-1086.

Leite, R. P., Jr., Jones, J. B., Somodi, G. C., Minsavage, G. V., and Stall, R. E. 1995.
Detection of Xanthomonas campestris pv. vesicatoria associated with pepper and tomato seed by DNA amplification. Plant Dis. 79:917-922.

Leite, R. P., Jr., Minsavage, G. V., Bonas, U., and Stall, R. E. 1994. Detection and identification of phytopathogenic Xanthomonas strains by amplification of DNA sequences related to the hrp genes of Xanthomonas campestris pv. vesicatoria. Appl. Environ. Microbiol. 60:1068-1077.









73

Lindgren, P. B., Peet, R. C., and Panopoulos, N. J. 1986. Gene cluster of Pseudomonas syringae pv. "phaseolicola" control pathogenicity of bean plants and hypersenstivity on nonhost plants. J. Bacteriol. 168:512-522.

Louws, F. J., Fulbright, D. W., Stephens, C. T., and Bruijn, F. J. 1994. Specific genomic fingerprints of phytopathogenic Xanthomonas and Pseudomonas pathovars and strains generated with repetitive sequences and PCR. Appl. Environ. Microbiol. 60:2286-2295.

Maas, J. L. (ed.) 1984. Angular leaf spot. Compendium of Strawberry Diseases. American Phytopathological Society, St. Paul, MN.

Maas, J. L., Pooler, M. R., and Galletta, G. J. 1995. Bacterial angular leafspot disease of strawberry: Present status and prospects for control. Adv. Strawberry Res. 14:18-24.

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.

McManus, P. S., and Jones, A. L. 1995. Detection of Erwinia amylovora by nested PCR and PCR-dot-blot and reverse blot hybridizations. Phytopathology 85:618-623.

Milholland, R. D., Ritchie, D. F., Daykin, M. E., and Gutierrez, W. A. 1993. Systemic movement of Xanthomonas fragariae in inoculated strawberry plants. (Abst.) Phytopathology 83:1408.

Minsavage, G. V., Thompson, C. M., Hopkins, D. L., Leite, R. M. V. B., and Stall, R. E.
1994. Development of a polymerase chain reaction protocol for detection of Xylella fastidiosa in plant tissue. Phytopathology 84:456-461.

Nei, M. 1987. Molecular Evolutionary Genetics. Columbia University Press, New York.

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.

Panagopoulos, C. G., Psallidas, P. G., and Alivizatos, A. S. 1978. A bacterial leafspot of strawberry in Greece caused by Xanthomonasfragariae. Phytopath. Z. 91:33-38.

Pataky, J. K., and Lim, S. M. 1981. Efficacy ofbenomyl for controlling Septoria brown spot of soybeans. Phytopatholology 71:438-442.

Pickup, R. W. 1991. Development of molecular methods for the detection of specific bacteria in the environment. J. Gen. Microbiol. 137:1009-1019.









74

Rhoads, D. D., and Roufa, D.J. 1991. Seqaid II Computer Program. Molecular Genetics Lab, Kansas State University, Manhattan, KS 66506.

Ritchie, D. F., Averre, C. W., and Milholland, R. D. 1993. First report of angular leaf spot, caused by Xanthomonasfragariae, on strawberry in North Carolina. Plant Dis. 76:1263.

Rowhani, A., Feliciano, A. J., Lips, T., and Gubler, W. D. 1994. Rapid identification of Xanthomonasfragariae in infected strawberry leaves by enzyme-linked immunosorbent assay. Plant Dis. 78:248-250.

SAS Institute Inc. 1988. SAS/STAT User's Guide, ed. 6.03. SAS Institute Inc., Cary, N.C.

Sasser, M. 1990. "Tracking" a strain using the the microbial identification system. Technical note 102. Microbial ID, Inc. Newark, DE.

Sasser, M. 1990. Identification of bacteria through fatty acid analysis. Methods in Phytobacteriology. Z. Klement, K., Rudolph, and D. Sands, eds. Akademiai Kiado, Budapest.

Schaad, N. W., Cheong, S. S., Tamaki, S., Hatziloukas, E., and Panopoulos, N. J. 1993. A viable cell enrichment, two-step, direct PCR technique for detection of Pseudomonas syringae pv. phaseolicola in bean seeds. (Abstr.) Phytopathology 83:1342.

Shaner, G., and Finney, R. E. 1977. The effect of nitrogen fertilization on the expression of slow-mildewing resistance in Knox wheat. Phytopathology 67:1051-1056.

Smith, J. J., Offord, L. C., Holderness, M., and Saddler, G. S. 1995. Genetic diversity of Burkholderia solanacearum (synonym Pseudomonas solanacearum) race 3 in Kenya. Appl. Environ. Microbiol. 4263-4268.

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. Sys. Bacteriol. 44:47-53.

Stall, R. E., Loschke, D. C., and Jones, J. B. 1986. Linkage of copper resistance and avirulence loci on a self-transmissible plasmid in Xanthomonas campestris pv. vesicatoria. Phytopathology 76:240-243.

Stall, R. E., and Minsavage, G. V. 1990. The use of hrp genes to identify opportunistic xanthomonads. Proc. 7th Int. Conf Plant Pathog. Bacteria 1989. Z. Klement, Akad6miai Kiad6, Budapest.









75

Stall, R. E., and Thayer, P. L. 1962. Streptomycin resistance of the bacterial spot pathogen and control with streptomycin. Plant Dis. Rep. 16:389-392. Szybalski, W. 1952. Microbial Selection. 1. Gradient plate technique for study of bacterial resistance. Science 116:46-48. Vauterin, L., Hoste, B., Kersters, K. and Swings, J. 1995. Reclassification of Xanthomonas. Int. J. Syst. Bacteriol. 45:472-489. van der Plank, J. E. 1963. Control of Disease by Fungicides. Plant Diseases: Epidemics and Control. Academic Press, NY. Willis, K., Rich, J. J., and Hrbak, E. M. 1991. hrp genes of phytopathogenic bacteira. Mol. Plant Microbe Interact. 4:132-138.

















BIOGRAPHICAL SKETCH


Pamela D. Roberts was born September 18, 1963 in Manhattan, KS. She graduated with a Bachelor of Science degree in horticulture from Kansas State University in 1987. She completed a Master of Science from the University of Hawaii in 1991 in the Department of Plant Pathology. She spent 16 months as a Research Scholar at the International Rice Research Institute in conjunction with the research at the Univerisity of Hawaii. She arrived at the University of Florida in 1991 to undertake a Doctor of Philosophy degree in the Department of Plant Pathology.






















76












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.


Richard D. Berger, ChaiF
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.


Robert E. Stall
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.


.f ftrJonei
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.


Craig K. andler
Associate Professor of Horticulture










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.


May, 1996
Dear, College of Agricult




Dean, Graduate School
















LD
1780 1996





UNIVERSITY OF FLORIDA 3 1262 08554 9185




Full Text
55
20
% 16:1 w 7 cis
Figure 4-3. The groups of Xanthomonas fragaricte distinguished by qualitative and
quantitative differences in fatty acid methyl ester profiles graphed by three major acids.


% Disease % Disease
15
7 21 7 20 5 8 22 8 23
Nov Dec Jan Feb March
Figure 2-1. Disease progress curves of Xcmthomoms fragariae on strawberry in Florida in
the 1994 (A) and 1995 (B) seasons. The mean of disease severity for four replicate blocks
are plotted against days after transplanting (DAT).



PAGE 1

THE EPIDEMIOLOGY, SPECIFIC DETECTION, AND GENETIC VARIABILITY OF XANTHOMONAS FRAGARIAE BY PAMELA D. ROBERTS 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 1996

PAGE 2

To my husband, Kenn, Cindy, and my parents.

PAGE 3

ACKNOWLEDGMENTS The author would like to gratefully acknowledge the members of the committee: Dr. R D. Berger, Dr R E. Stall, Dr C. K. Chandler, and Dr. Jeffrey B Jones The enthusiatic support, expert advise, and guidance throughout the duration of this project were greatly appreciated. Special thanks go to Dr. Jeffrey B. Jones who provided his laboratory for a majority of the work. Thanks go to Dr. A. R. Chase who initially suggested the project and provided considerable support the first year. Special thanks go to the members of Dr Stall's and Dr. Jones' labs, including N. Cheri Hodges, G V. Minsavage, Rui P. Leite, Jr., Hacene Bouzar, Gail Somodi, Jeanette Chun, and Rick Kelly, for their help and information. Very special thanks go to the people at Dover-G.C.R EC. who helped me tremendously with all aspects of the field research and taught me all about the production of strawberries: Alicia Whidden, Anne Turgeau, Jim Sumler, Dale Wenzel, Larry Smith, and Mitch Boles. Finally, thanks go to the folks who helped make the experience enjoyable: Gary Marlow, Pia D. Gavino, Morgan Wallace, Pamela Lopez, and Marion Bogart.

PAGE 4

TABLE OF CONTENTS ACKNOWLEDGMENTS Hi ABSTRACT vi CHAPTERS 1 INTRODUCTION 1 2 DISEASE PROGRESS, YIELD LOSS, AND CONTROL OF XANTHOMONAS FRAGARIAE ON STRAWBERRY PLANTS 8 Introduction 8 Materials and Methods 10 Results 13 Discussion 22 3 DEVELOPMENT OF A SPECIFIC AND SENSITIVE DETECTION TECHNIQUE BY THE POLYMERASE CHAIN REACTION FOR XANTHOMONAS FRAGARIAE AND APPLICATION IN A STUDY OF SURVIVAL OF THE BACTERIUM ON STRAWBERRY PLANTS 25 Introduction 25 Materials and Methods 28 Results 33 Discussion 41

PAGE 5

4 GENOMIC RELATEDNESS OF XANTHOMONAS FRAGARIAE ON STRAWBERRY BY FATTY ACID METHYL ESTERASE AND RESTRICTION LENGTH FRAGMENT POLYMORPHISM ANALYSES 45 Introduction 45 Materials and Methods 47 Results 51 Discussion 56 5 DISCUSSION 60 LIST OF REFERENCES 70 BIOGRAPHICAL SKETCH 76

PAGE 6

Abstract of 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 THE EPIDEMIOLOGY, SPECIFIC DETECTION, AND GENETIC VARIABILITY OF XANTHOMONAS FRAGARIAE By Pamela D. Roberts May 1996 Chairman: Dr. R. D Berger Major Department: Plant Pathology Angular leaf spot (ALS) on strawberry is caused by the bacterium Xanthomonas fragariae (XF). The epidemiology of ALS, development of a detection technique by the polymerase chain reaction (PCR), and the genetic variability of strains were examined. The disease severity of ALS on plants in field plots increased to 25% and 15% in two seasons. Yield was decreased 8% to 10%. Minimal spread of ALS occurred between field plots. Chemical sprays applied at the label rate of cupric hydroxide plus mancozeb at 7to 14day intervals decreased disease but was phytotoxic and decreased yield. A 10% rate of the mixture applied frequently slightly reduced disease but increased yield one season and significantly reduced disease and did not affect yield the next season. Three primers were specific for amplification of DNA from XF but not DNA from strains of 16 pathovars of Xanthomonas campeslris or non-pathogenic xanthomonads from strawberry. Bacteria were detected at 10 4 colony forming units per ml by a single round of PCR A nested PCR

PAGE 7

technique increased detection 1,000 fold. Plants were inoculated with a rifampicin-resistant strain and oversummered in the field at two locations in Florida. Bacteria from leaf and crown samples were detected by nested PCR and recovery onto selective media at twoweek intervals for 92 days after planting. Daughter plants of the inoculated plants were positive for XF by nested PCR amplification. Analysis of genetic variability by fatty acid methyl esterase (FAME) profiles divided 50 strains into 9 groups based upon qualitative and quantitative differences. The majority (74%) of strains were placed into a closely related group determined by cluster analysis. The profiles from restriction fragment length polymorphism (RFLP) analysis of genomic DNA restricted by two infrequent cutting endonucleases and separation by pulsed field gel electrophoresis grouped strains into four groups (A-D). Another endonuclease subdivided the B group into three groups. The dendrogram unweighted pair analysis of FAME profiles divided the population into four groups which correlated well with the RFLP groups. Considerable diversity appears within the species.

PAGE 8

INTRODUCTION Xanthomonas fragariae causes angular leaf spot disease on strawberry (Fragaria species; Fragaria x ananassa Duchesne). The disease was first found in Minnesota in 1960 (Kennedy and King, 1962a) and it is currently found in many regions of strawberry production throughout the world (Maas, 1984; Ritchie et al., 1993). Angular leaf spot was first reported in Florida in 1971 (Howard, 1971). Dissemination of the bacterium occurred via the transportation of infected plants (Maas, 1984; Panagopolous et al., 1978; Dye and Wilkie, 1973). In Florida, strawberry plants which arrive from northern nurseries for transplanting in the fall frequently have leaves which exhibit symptoms of angular leaf spot Typical pathogenic strains of A', fragariae can be isolated from the lesions. A diagnostic symptom of the disease is the translucent appearance of lesions when viewed with transmitted light (Maas, 1984). A vascular collapse of the plant from systematic invasion by the bacterium has been described in California (Hildebrand et al., 1967). The epidemiology of angular leaf spot is mostly unknown in fields in Florida where strawberry production is an annual crop. Transplants are obtained each season from nurseries in Canada and northern states; few transplants are produced in Florida. A source of inoculum other than infected transplants has not been found Howard ( 1 97 1 ) was unable to determine the inoculum source for infected plants from nurseries in Florida or other states. In surveys

PAGE 9

2 conducted in 1968, 1970, 1993, and 1994, plants which had symptoms of angular leaf spot in the spring did not have symptoms of the disease the following August (Howard, 1971; P. D. Roberts, unpublished). However, in 1969 mild infections on one variety were observed in mid-August. Plants transplanted to fields did not develop angular leaf spot. Kennedy and King (1962b) determined that the bacterium overwintered in infected leaves buried in the soil and caused disease symptoms on plants the next year. The bacterium did not survive free in the soil nor were any naturally occurring hosts identified in host-range studies (Kennedy and King, 1 962a). For bacteria on plant refuse to serve as an inoculum source in Florida, the bacterium must oversummer. Optimal growth (=20° C) of the bacterium (Howard et al., 1985; Kennedy and King, 1962b) is at temperatures cooler than normally occurs in Florida during the summer The survival of the bacterium on plants in summer nurseries in Florida and inoculum sources other than infected transplants have not been established. The effect of angular leaf spot on yield is unknown Howard ( 1 97 1 ) accredited some yield losses due to angular leaf spot in fields in Florida but he did not quantify the losses. In Wisconsin, a decrease in yield of 70 to 80% was estimated due to the disease (Epstein, 1966). However, the production in the northern United States is a perennial, matted-row system which differs significantly from the Florida production. A significant loss in marketable fruit may occur due to infections of the calyx. The sepals become brown and dry and the fruit is unmarketable because of its unattractive appearance (Epstein, 1966; Maas, 1995). Chemical control of bacterial diseases is difficult. Antibiotics have limited effectiveness over time since mutations to bacterium may occur and form resistant populations (Stall and Thayer, 1962). Therefore copper compounds are frequently used to

PAGE 10

3 control a bacterial disease (Jones et al., 1991) Antibiotics and copper compounds were effective protectants against angular leaf spot but these did not eradicate the disease (Alippi et al., 1989). Marco and Stall (1983) examined chemical control of strains of Xanthomonas campestris pv. vesicatoria that differed in sensitivity to copper. The mixture of cupric hydroxide plus mancozeb was more effective than cupric hydroxide alone to control both the copper-resistant and the copper-sensitive strain. The application of copper compounds to strawberry plants is confounded by the fact that copper can be phytotoxic to strawberry plants (Howard and Albregts, 1973). The application of a chemical can reduce the infection by a pathogen and slow the rate of the epidemic, e.g. van der Plank's (1963) apparent infection rate (Fry et al., 1979; Pataky and Lim, 1981). A reduced rate of fungicide applied at frequent intervals as a protectant has been used effectively to reduce disease on crops (Fry 1975; Fry et al, 1979; Conway et al., 1987). Identification of plants infected with X. fragahae is a priority because of the ease of movement of infected but asymptomatic plants (Maas, 1995). International movement of infected plants is blamed for introduction of angular leaf spot into Greece and New Zealand (Panagopoulos et al.,1978; Dye and Wilkie, 1973). Nursery-plant producers are pressured to provide disease-free plants by foreign countries and by farmers who refuse to buy infected transplants The European Plant Protection Organization (EPPO) lists X. fragahae as a quarantine pest and has prescribed phytosanitary procedures In the future, regulatory issues may be of greater concern. The production of disease-free plants is essential for control of angular leaf spot. Therefore, accurate identification of plants infected with the bacterium is imperative Available detection techniques are limited in their usefulness and accuracy to

PAGE 11

4 detect low populations of the bacterium that may exist in asymptomatic tissue. Xanthomonas fragariae may be identified in the early stages of leaf infection by the diagnostic translucent, watersoaked lesions viewed with transmitted light; however, older lesions may be confused with symptoms caused by fungal pathogens (Kennedy and King, 1 962b). Diagnosis based on symptoms is very difficult and not applicable for asymptomatic plants Identification of the disease based upon isolation and characterization of the causal agent may also be difficult because X. fragariae grows slowly and may be masked by faster growing organisms (Hildebrand et al., 1967). Expression of watersoaked lesions takes 6 days or longer after inoculation. Thus, fulfillment of Koch's postulates to confirm pathogenicity is difficult and time-consuming. Assays have been developed with improved sensitivity and specificity for the detection of plant pathogens in plant tissue. A specific, indirect, enzyme-linked immunosorbent assay (ELISA) was developed to detect X. fragariae from symptomatic plant tissue (Rowhani et al., 1994). The antibody assay did not react with bacterial strains of 16 pathovars of X. campestris or non-pathogenic bacteria from strawberry leaves. A single cross-reaction occurred to a strain of A , campestris isolated from Nerium oleander. The assay detected bacteria directly from a visible lesion on a strawberry leaf The level of sensitivity of the assay was at ca. 10 4 colony forming units (cfu) per ml. However, bacteria on asymptomatic plants may not be detected by this ELISA assay. The polymerase chain reaction (PCR) has been used to amplify specific DNA sequences to detect and identify many plant pathogens including some members of the genus Xanthomonas (Henson and French, 1993; Louws et al., 1994). Leite et al (1994) utilized

PAGE 12

5 primers specific to regions of the hrp gene cluster which confers hypersensitivity and pathogenicity from Xanthomonas campestris pv. vesicatoria to detect pathogenic xanthomonads. Saprophytic and non-pathogenic bacteria lack the hrp genes (Stall and Minsavage, 1990; Lindgren et al., 1986; Willis et al., 1991) and DNA from these bacteria was not amplified by the hrp gene cluster primers. Differentiation of X. campestris pathovars was made by restriction endonuclease analysis (REA) patterns generated by digestion of the PCR products with frequent cutting enzymes (Leite et al., 1994 and 1995). The primers were used to detect X. c. vesicatoria in seed lots of naturally infected pepper and tomato (Leite et al., 1995) The sensitivity of detection by PCR is reported at 10 3 to 10 2 cfu per ml (Minsavage et al., 1994; Leite et al., 1994; Henson and French, 1993). In nested PCR, sensitivity of detection is increased by using PCR products from an amplification as target DNA in a second round of amplification by a second set of primers internal to the first (Schaad et al., 1993). McManus and Jones (1995) reported an increase in sensitivity of 1,000-fold with nested PCR over a single round of PCR amplification to detect Erwinia amylovora. The epidemiology of X. fragariae must be understood before effective control strategies can be devised. The delineation of bacterial populations is imperative to such studies The relationship of X. fragariae to other members of the genus Xanthomonas has been examined (Hildebrand et al., 1990; Vauterin et al., 1995; Hodge et al., 1992), but the genetic variability of strains within the species has not been reported (Maas, 1995). The information from studies to compare strains of X. fragariae at the genetic level could be useful to identify the origins and spread of specific strains or genetic types. This information could be applied in international tracking of pathogen populations for quarantine programs.

PAGE 13

6 In addition, the identification of genetic types is a prerequisite to identify sources of resistance in strawberry to the pathogen. While strawberry cultivars exhibited levels of susceptibility or tolerance to X.fragariae; only F. moschata Duch. appeared to be immune (Hazel, 1981; Hazel and Civerolo, 1980; Kennedy and King, 1962b). A screening program to identify genes for resistance must incorporate representatives of the genetic variants in the screening process otherwise the resistance may be overcome quickly by genetic variants Studies to characterize populations of bacteria have used biochemical and molecular biological techniques. Protein staining and fatty acid analysis have been useful to detect differences at the metabolic level within species of Xanihomonas (Bouzar et al., 1994; Stall et al., 1994). The polymerase chain reaction (PCR) has been used to analyze differences between pathovars and strains of bacteria (Hensen and French, 1993). Primers specific to the hrp-gene cluster of Xanihomonas campestris pv. vesicaloria amplified genomic DNA and restriction enzyme analysis of the PCR product differentiated between many pathovars and species of Xanihomonas (Leite et al., 1994 and 1995). Primers sets from conserved repetitive bacterial DNA elements generated genomic fingerprints which were used to differentiate gram-negative soil bacteria and pathovars of Xanihomonas and Psuedomonas (De Bruijn, 1992; Louws et al., 1994). Restriction length fragment polymorphisms (RFLP) of the bacterial genome digested with rare-cutting restriction endonucleases and resolution by pulsed field gel electrophoresis (PFGE) has been used to type many bacterial species and genera (Cooksey and Graham, 1989; Egel et al., 1991; Smith et al., 1995). The objectives of these studies were to investigate epidemiological aspects of angular leaf spot on strawberry including the incidence of X. fragariae on transplants from northern

PAGE 14

7 nurseries, disease spread in strawberry production fields, the effect of angular leaf spot on yield, and the control of the disease by chemicals. To aid in epidemiological studies, a sensitive and specific technique by PCR reaction to detect X. fragariae was developed. Primers were designed specific to the region of genomic DNA from X. fragariae related to the hrp genes of X. c. vesicaloria. The survival of X. fragariae on nursery strawberry plants in the field at two locations in Florida and dissemination to daughter plants was examined to understand the disease cycle of angular leaf spot in Florida. The genetic variability of a collection of strains of X. fragariae from the United States and Canada was examined. Analyses were by restriction length fragment profiles of genomic DNA restricted with rarecutting endonucleases separated by pulsed field gel electrophoresis and by profiles of fatty acid methyl esters.

PAGE 15

CHAPTER 2 DISEASE PROGRESS, YIELD LOSS, AND CONTROL OF XANTHOMONAS FRAGAR1AE ON STRAWBERRY PLANTS Introduction Angular leaf spot of strawberry is caused by the bacterium Xaiilhomotiasfragariae. The disease was first reported from Minnesota in 1960 (Kennedy and King, 1962a) and is currently found in many regions of strawberry production throughout the world. Dissemination of the bacterium occurred via the transportation of infected plants (Maas, 1984; Panagopolous et al., 1978; Dye and Wilkie, 1973). A diagnostic symptom of the disease is the translucent appearance of lesions when viewed with transmitted light (Maas, 1984). A vascular collapse of the plant from systematic invasion by the bacterium has been described in California (Hildebrand et al., 1967). The epidemiology of angular leaf spot is mostly unknown in fields in Florida where strawberry production is an annual crop Transplants are obtained each season from nurseries in Canada and northern states; few transplants are produced in Florida. A source of inoculum other than infected transplants has not been found. Howard (1971) was unable to determine the inoculum source for infected plants from nurseries in Florida or other states Kennedy and King (1962b) determined that bacteria overwintered in infected leaves buried in the soil and caused disease symptoms on plants the next year. The bacterium did not survive free in the soil nor were any naturally occurring hosts identified in host range studies (Kennedy and 8

PAGE 16

9 King, 1962b). For bacteria on plant refuse to serve as an inoculum source in Florida, the bacterium must oversummer. Optimal growth (=20° C) of the bacterium (Howard et al., 1985; Kennedy and King, 1962b) is at temperatures cooler than normally occurs in Florida during the summer. The bacterium survived on summer nursery plants in the field in Florida but at very low populations (Roberts, unpublished) The effect of angular leaf spot on yield is unknown. Howard (1971) accredited some yield losses due to angular leaf spot in fields in Florida but he did not quantify the losses. In Wisconsin, a decrease in yield of 70 to 80% was estimated due to the disease (Epstein, 1966). However, the production in the northern United States is a perennial, matted-row system which differs significantly from the Florida situation. A significant loss in marketable fruit may occur due to infections of the calyx. The sepals become brown and dry and the fruit is unmarketable because of its unattractive appearance (Epstein, 1966; Maas, 1995). Chemical control of bacterial diseases on plants is difficult. Antibiotics have limited effectiveness over time since mutations may form resistant populations (Stall and Thayer, 1 962). Therefore copper compounds are frequently used to control a bacterial disease (Jones et al., 1991). Antibiotics and copper compounds were effective protectants against angular leaf spot but these did not eradicate the disease (Alippi et al., 1989). Marco and Stall (1983) examined chemical control of strains of Xanthomonas campestris pv. vesicaloria which differed in sensitivity to copper. The mixture of cupric hydroxide plus mancozeb was more effective than cupric hydroxide alone to control both the copper-resistant and the coppersensitive strain. The application of copper compounds to strawberry plants is confounded by the fact that copper can be phytoxic to strawberry plants (Howard and Albregts, 1973). The

PAGE 17

10 application of a chemical pesticide can reduce the infection by a pathogen and slow the rate of the epidemic, e g , van der Plank's (1963) apparent infection rate (Fry et al., 1979; Pataky and Lim, 1981). A reduced rate of fungicide applied at frequent intervals as a protectant has been used effectively to reduce disease on crops (Fry 1975; Fry et al, 1979; Conway et al., 1987). Our objectives in this study were to examine epidemiological aspects of angular leaf spot including the incidence of X. fragariae on transplants from northern nurseries, the effect of angular leaf spot on yield, and the control by chemicals. Materials and Methods Survey of farmer's fields and storage facility. Disease incidence of angular leaf spot on plants in fields located around Dover and Plant City, FL, was examined in October 1993. Plants were assessed for symptoms of angular leaf spot within 2 to 4 days of transplanting. Leaves from symptomatic plants were collected. A 9-mm diameter of leaf tissue was removed surrounding a lesion and macerated in 200 iA of sterile water A loopful of the suspension was dilution streaked onto Wilbrink's medium (WB) (Koike, 1962) and colonies of X. fragariae were identified The pathogenicity of these isolated strains was tested on 'Sweet Charlie' strawberry plants. In 1994, transplants were sampled while in cold storage at a facility in Dover, FL Seven groups of plants, comprised of four cultivars from six northern growers were sampled Three boxes were selected at random from within a group and 20 plants were removed from each box The plants were examined for symptoms of

PAGE 18

11 angular leaf spot and attempts to isolate the pathogen from putative lesions was done by the method described above. Inoculation. Three strains ofX.fragariae (Xfl 13, Xfl03, and XA425) were used to inoculate plants used in field plots Strains Xfl 13 and Xfl03 were isolated from infected plants at Gulf Coast Research and Education Center (GCREC), Dover, FL and Xfl 425 was obtained from A. Chase (Central Florida Research and Education Center, Apopka, FL). Strains were cultured on WB at 24° C and long-term storage was at -70° C in 15% glycerol. The sensitivity of the strains to copper was tested by growth on nutrient agar amended with CuS0 4 (Stall et al., 1986). Three days prior to inoculation, each of the strains was streaked to ten plates of WB Bacterial cells were collected from plates, suspended in sterile 0.01 M MgS0 4 , and the concentration adjusted to approximately 10 7 cfu per ml. Equal volumes of each suspension were combined to comprise the inoculum Plants were inoculated by dipping bundles of 25 plants into the bacterial suspension for 30 s. Control plants were dipped into 0.01 M MgS0 4 . Plants were placed into plastic bags and incubated 24 h at 22° C. Field experiments. Field experiments were located at GCREC, Dover, FL, from October 1993 through March 1994 and repeated the following season. Transplants of 'Sweet Charlie' grown in the summer nursery at GCREC-Dover were used. Raised beds were prepared and fertilized with 10N-4P-10K at a rate of 2000 lbs per acre with onefourth banded before bed preparation and the remainder banded 5 cm deep in the bed center the first season In the second season, three-fourths pound of N and K per acre per day were applied through the drip irrigation system. Soil was fumigated with 98% methyl bromide and 2%

PAGE 19

12 chloropicrin at 448 kg per acre. Beds were covered with 1 mm black polyethylene mulch immediately after fumigation. The experimental design was a randomized complete block in a 2 * 3 factorial design with four replications The first factor had two levels: plants inoculated with either the suspension of X. fragahae or MgS0 4 . The three levels of the second factor were: no chemical treatment, the label (1 *) rate of cupric hydroxide (Kocide 101 at 9.08 kg of active ingredient per acre) plus mancozeb (Dithane DF at 6 81 kg of active ingredient per acre) sprayed at 7to 14-day intervals, or the reduced (0.1*) rate of cupric hydroxide plus mancozeb sprayed at 2to 4-day intervals. An individual plot contained 18 plants arranged in two rows of nine plants. The beds were spaced on 1 .22 meter centers with 30 cm between rows and 30 cm within a row. Fallow area was 3.55 m within a row and 2.44 m between rows. Pesticides were applied throughout season as needed to control insects and fungal diseases. Chemical applications were made by a handheld wand attached to CCycharged canister at 40 psi and pesticide was applied to runoff Plants were transplanted on 15 October 1993 and on 20 October 1994 Overhead sprinkler irrigation was applied 8 hours daily for 10 to 14 days to establish transplants and applied throughout the season as needed. Drip irrigation was installed in the summer of 1994 at GCREC-Dover and was used as supplemental irrigation in 1995. Estimates of disease severity of angular leaf spot were made at two-week intervals. Disease severity was expressed as percent leaf area diseased for the entire plant for each of six plants located in the center of each plot. Progress curves were plotted as the mean of disease severity of replicate treatments versus time. Area under the disease progress curve

PAGE 20

13 (AUDPC)(Shaner and Finney, 1977) was calculated for each plot and used in statistical analysis. Statistical analysis was performed by orthogonal contrasts using PC-SAS. Yield data. Fruit were harvested at 2to 4-day intervals from initial bearing in December through 30 March Fruit were graded as marketable, culls, or nonmarketable due to damage by fungi. Marketable fruit were those free of rot, not misshapen, and greater than 10 g in weight. Culls were non-marketable fruit due to physical imperfections such as small size (< 10 g), damage by insects, or undesirable shape. Nonmarketable fruit was damaged by fungal diseases, usually caused by anthracnose, botrytis, and phomopsis. The weight in grams for each category was recorded for individual plots. Statistical analysis of the weight of fruit in each category was performed using orthogonal contrasts using PC-SAS. Results Disease incidence of transplants in field and storage facility. In 1993, six of seven fields contained plants with symptoms of angular leaf spot from which X. fragariae was isolated. In 1994, three of the seven groups sampled from the cold storage facility contained plants with symptoms of angular leaf spot. In one of these groups, plants with symptoms of angular leaf spot were found in all of the boxes sampled. In the other two samples, one box contained plants which were symptomatic with angular leaf spot Bacteria were isolated from all plants with lesions of angular leaf spot and pathogenicity of these isolated strains on strawberry was confirmed. Strains which were resistant to copper were not identified

PAGE 21

14 Disease progress. Symptoms of angular leaf spot were not visible on plants at planting in either 1994 or 1995. Disease progress curves of angular leaf spot on strawberry plants for all treatments are presented in Figure 2-1. In 1994, the disease severity of angular leaf spot on inoculated plants which did not receive a spray treatment increased for 89 days after transplanting (DAT), decreased until 118 DAT, and then increased to 25% by 1 18 DAT and leveled out to the end of the season. In 1995, the epidemic on plants inoculated with the bacterium initially increased, then decreased at 61 DAT until 1 1 1 DAT, and then increased and reached 15% by 154 DAT The disease on inoculated plants sprayed with either the 1* or 0. 1 x rate of cupric hydroxide plus mancozeb followed similar disease progress curves except the amount of disease was reduced for each treatment. In contrast, the onset of the epidemic on the noninoculated plants was greatly delayed. Disease was observed on these plants after 1 18 DAT in 1994 and at 125 DAT in 1995. The noninoculated plants had very little disease (0-3% severity) by the end of the season, regardless of spray treatment. Statistical analysis of the AUDPC for each treatment is shown in Figure 2-2. Noninoculated plants had significantly (P = 0.0001) lower disease than plants inoculated with the bacterium. Noninoculated plants treated with the 1 * rate did not have disease either year. Inoculated plants sprayed with the 1 * rate of cupric hydroxide and mancozeb had a significantly (P = 0.0004 in 1994, P = 0001 in 1995) lower AUDPC compared to the control treatment in both years. The plants sprayed with the 0. 1 x rate of the fungicide had reduced disease in 1994 (P = 0.17) and, in 1995, the reduction in AUDPC was highly significant (P = 0.0001) compared to plants without chemical sprays

PAGE 22

15 S3 s, a 19 33 47 64 75 89 105 3 17 1 7 29 12 28 IIS 10 132 24 145 9 157 DAT 21 Nov Dec Jan Feb March March Figure 2-1. Disease progress curves of Xanthomonas fragariae on strawberry in Florida in the 1994 (A) and 1995 (B) seasons. The mean of disease severity for four replicate blocks are plotted against days after transplanting (DAT).

PAGE 23

16 O a, Noninoculated NoS NoS 0.1X so id B Ab o 1 30 20 10 Aa Aa Aa Aa Aa Noninoculated NoS 0.1X IX Figure 2-2. Area under the disease progress curve (AUDPC) of Xanthomonas fragariae on strawberry in the 1994 (A) and 1995 (B) seasons and the effect of applications of cupric hydroxide plus mancozeb at the 1 * rate at 7-to 14-day intervals or at 0. 1 x (10%) rate of the chemical combination applied at 2-to 4-day intervals or no spray (NoS) A, B = significant difference (P = 0.05) within the first factor of inoculated vs. noninoculated plants; a, b = significant difference within the second factor of the spray treatments.

PAGE 24

17 Yield loss. In both years, marketable yield on plants infected with angular leaf spot was significantly decreased (P = 0.005 in 1994 and P = 0.04 in 1995) compared to yield on non-inoculated plants (Fig 2-3). Total and marketable yields were reduced by an average of 7% and 8%, respectively, in 1994 and by 6% and 10% in 1995. Total yield had similar reductions as marketable yield, therefore only marketable yield is presented. Yields were generally lower in 1995 than in 1994. The effect of spray treatments on the yield from inoculated and noninoculated plants is presented in Figure 2-4. Marketable yield on noninoculated plants sprayed with the 1 x treatment of cupric hydroxide plus mancozeb was significantly reduced (P = 0.04) compared to plants which received the 0. 1 x or no-spray treatments in 1994. No difference in yield was detected on inoculated plants receiving the spray treatments in 1994 In 1994, the yield on inoculated plants sprayed with the 1 * rate was significantly reduced (P = 0.005) compared to the other two treatments. Yield on noninoculated plants sprayed with the 1 « and 0. 1 * chemical sprays was reduced significantly (P =0.04) compared to yield on nonsprayed plants in 1995. The spray treatments did not have any effect on culled fruit in 1994 or 1995 (Fig. 2-5). No difference in culled fruit was detected among spray treatments. Nonmarketable fruit lost because of damage by fungal diseases was significantly different at P = 0.01 for all treatments in 1994 but the relationship to spray treatments was unclear (Fig. 2-6). In 1995, there was no significant difference in damage to fruit by fungi for either inoculated or spray treatments. Overall, the total yield lost to fungal infections was much lower in 1994 compared to 1995.

PAGE 25

18 3 16 13 6 AA Bb Noninoculated Inoculated NoS 0.1X IX NoS 0.1X IX a ? B Aa Ab Ab Noninoculated Inoculated NoS Figure 2-3. The average marketable yield on strawberry plants which were inoculated with Xanthomonas Jragahae (Inoc) or not inoculated (Noninoc) for the 1994 and 1995 seasons. Yield was the average for four replications for marketable fruit only. Different letters (A, B) represent significant difference at P = 0.05.

PAGE 26

10 19 O A B ' I IS :..'" ..:':: Noninoc Inoc Noninoc Inoc 1994 1995 Figure 2-4. The marketable yield on strawberry in Florida in the 1994 (A) and 1995 (B) seasons. A, B = F was significant at P ™ 0.05 within the first factor of noninoculated vs. inoculated; a, b = F was significant at P = 0.05 within the second factor of spray treatments.

PAGE 27

20 .-. 0.8 2 0.6 s 0.4 Aa NoS Aa Aa Noninoculated 0.1X IX Aa Inoculated NoS 0.1X Noninoculated Inoculated IX 0.1X IX Figure 2-5. The culled fruit on strawberry in Florida in the 1994 (A) and 1995 (B) seasons A, B = F was significant at P = 05 within the first factor of noninoculated vs. inoculated; a, b = F was significant aX P = 0.05 within the second factor of spray treatments.

PAGE 28

21 O P IS B Noninoculated NoS 0.1X IX Aa Aa Aa Noninoculated NoS 0.1 X IX NoS 0.1X IX Figure 2-6. The yield lost to fruit damaged by fungal diseases on strawberry in Florida in the 1994 (A) and 1995 (B) seasons. A, B = /-"was significant at P = 0.05 within the first factor of noninoculated vs.inoculated; a, b, c = F was significant at P = 0.05 within the second factor of spray treatments

PAGE 29

22 Discussion A decrease in yield of strawberry fruit was observed both seasons due to angular leaf spot. This report is the first to quantify the reduction in yield due to this disease. The 8 to 10% loss determined in these studies is much lower than the 70 to 80% loss estimated by Epstein (1966) in fruit production fields in Wisconsin. Production in northern regions of the U. S. is usually perennial and the age of the plants was not given. The loss was reported on the cultivar 'Sparkle' which is not produced commercially in Florida. Therefore, we are unable to compare our yield loss to his report. Howard (1971) also reported unquantified yield losses due to this disease but on cultivars which are no longer in commercial production. In our studies, the average yield loss observed both years was very similar despite the differences in the disease levels and the base yields between the two seasons. In 1995, the disease severity was 60% of the amount estimated in the previous season; however, the yield loss was actually higher by 2% for marketable yield. The 10% reduction in yield represents an estimated loss of 1233 kg of berries per ha. Strawberry fruit has a very high cash value and a decrease in revenue of 8 to 10% represents a significant economic loss to producers. Strawberry producers in Florida currently try to control angular leaf spot by application of copper compounds and by avoidance of the disease An understanding of the epidemiology of angular leaf spot is necessary to determine the effectiveness and practicability of any control method. The survey of farmers' fields and the cold storage facility established that transplants arrive from northern nurseries infected with angular leaf spot at fairly high disease incidence. Therefore, disease is introduced into the field on infected transplants. In

PAGE 30

23 our surveys by the sampling method used in the first season, more plants were assessed for symptoms of angular leaf spot in the field compared to the number of plants in the random sampling of boxes in cold storage. This probably accounts for identification of diseased plants in nearly all of the farmer's fields versus less than in 25% of boxed plants. In addition, plants were examined until disease was found in the field which skewed the randomness of the survey. In field experiments, the spread from inoculated plants to noninoculated plants in nearby experimental plots was minimal. Thus, the incidence of plants with angular leaf spot in the field likely occurred from inoculum already present on the transplants In the field experiments, if inoculum was present from another source, such as debris in the soil or an alternate host, most likely the disease on the noninoculated plants would have been more general and appeared earlier. Disease severity on noninoculated plants in the experimental plots was extremely low and most plants remained free of angular leaf spot. For the disease that occurred on noninoculated plants, the inoculum was probably transported mechanically from inoculated plants during harvests or disease readings. Progress of angular leaf spot developed similar to curves of pathogens described on other crops (van der Plank, 1963). A decrease in disease severity was seen about midway through the season in 1994. This reduction, or negative infection rate, was due to growth by strawberry plants which diluted the amount of disease relative to the total leaf area. The dilution effect of healthy plant tissue was described by van der Plank (1963). In 1995, a negative infection rate was seen very early in the epidemic. This was because the symptoms of the disease almost disappeared from the field as the affected leaves died and were lost. The

PAGE 31

24 difference in disease progress between the two seasons did not appear to be from differences in rainfall since mean rainfall was approximately the same both years (data not shown) Mean temperatures also were similar; however; the number of days at temperatures below 10° C in November and December 1993 were much greater than in 1994. Another difference between the seasons was that sprinkler irrigation was reduced in 1 995 because of the change to drip irrigation. Overhead irrigation, fog, high humidity, and rain are positive factors in disease development and spread (Epstein, 1966; Hildebrand et al., 1967, Kennedy and King, 1962b; Lai, 1978). The trend towards drip irrigation in commercial production fields should have a positive effect to reduce the spread and survival of the bacterium Strawberry producers currently apply copper-based compounds to control angular leaf spot despite the phototoxicity to plants caused by copper Application of cupric hydroxide plus mancozeb at the 1 * rate was phytotoxic to strawberry plants (Howard, 1973; this study). In our tests, the greatly reduced 0. 1 x rate did not harm plants and significantly reduced disease severity in 1995. The protectant application of bactericide at this rate and spray schedule was intended to reduce the total amount of inoculum and to prevent spread of the disease. The approach of frequent sprays at reduced concentrations may have potential to control this disease More studies are needed to evaluate different rates, application intervals, and chemical mixtures to achieve maximum disease control while avoiding yield losses. An alternate approach might be to control the disease in the nursery by copper applications The loss in yield due to phytotoxicity would not be an issue on nursery plants. The sprays would reduce the amount of initial inoculum on transplants and subsequently reduce the amount of disease in fruit production fields.

PAGE 32

CHAPTER 3 DEVELOPMENT OF A SPECIFIC AND SENSITIVE DETECTION TECHNIQUE BY THE POLYMERASE CHAIN REACTION FOR XANTHOMONAS FRAGARIAE AND APPLICATION IN A STUDY OF SURVIVAL OF THE BACTERIUM ON STRAWBERRY PLANTS Introduction Angular leaf spot of strawberry (Fragaria * ananassa Duchesne), caused by the bacterium Xanthomonas fragariae, was first reported in Minnesota in 1960 and is now found in many areas of strawberry production throughout the world (Kennedy and King, 1 960a; Maas, 1984; Ritchie et al., 1993). The disease is apparently disseminated by the transportation of infected plants (Maas, 1984). In Florida, strawberry plants which arrive from northern nurseries for transplanting in the fall frequently have leaves which exhibit symptoms of angular leaf spot. Typical pathogenic strains of X. fragariae can be isolated from the lesions (P. D. Roberts, unpublished). The epidemiology of X. fragariae is mostly unknown in fields of strawberry in Florida where production is from annual crops. Angular leaf spot was first reported in the state in 1971 (Howard, 1971). Howard (1971) was unable to determine the inoculum source for infected plants from nurseries in Florida or other states. Bacteria may survive on infested leaves in the soil (Kennedy and King, 1962b), but plants in Florida are usually treated with the herbicide paraquat at the end of the season and removed. The bacterium does not survive 25

PAGE 33

26 freely in the soil (Kennedy and King, 1962b). Cool temperatures (= 20° C) are optimal for disease symptom expression (Howard et al., 1985; Kennedy and King, 1962b) and high temperatures (>28° C) such as those which occur in Florida during the summer months are unfavorable. In surveys conducted in 1968, 1970, 1993, and 1994, plants which had symptoms of angular leaf spot in the spring did not have symptoms of the disease the following August (Howard, 1971; P. D. Roberts, unpublished). However, in 1969 mild infections on one variety were observed in mid-August. Plants transplanted to fields did not develop angular leaf spot. The survival of the bacterium on plants in summer nurseries in Florida and inoculum sources other than infected transplants have not been investigated. Identification of plants infected with X.fragariae is a priority because of the ease of transmission on infected but asymptomatic plants (Maas, 1995). International movement of infected plants is blamed for introduction of angular leaf spot into Greece and New Zealand (Panagopoulos et al., 1978; Dye and Wilkie, 1973). Nursery-plant producers are pressured to provide disease-free plants by buyers in foreign countries and by farmers who refuse to buy infected transplants The European Plant Protection Organization (EPPO) lists X fragariae as a quarantine pest and has prescribed phytosanitary procedures (Maas, 1995). In the future, regulatory issues may be of greater concern. The production of disease-free plants is essential for control of angular leaf spot. Therefore, accurate identification of plants infected with the bacterium is imperative. Available detection techniques are limited in their usefulness and accuracy to detect low populations of the bacterium that may exist in asymptomatic tissue. Xanthomonas fragariae may be identified in the early stages of leaf infection by diagnostic translucent, watersoaked lesions viewed with transmitted light; however, older lesions may

PAGE 34

27 be confused with symptoms caused by fungal pathogens (Kennedy and King, 1962b). Diagnosis based on symptoms is very difficult and not applicable for asymptomatic plants. Identification of the disease based upon isolation and characterization of the causal agent may also be difficult because X. fragariae grows slowly and may be masked by faster growing organisms (Kennedy and King, 1962a) Expression of watersoaked lesions takes 6 days or longer after inoculation. Thus, fulfillment of Koch's postulates to confirm pathogenicity is difficult and time-consuming Assays have been developed with improved sensitivity and specificity for the detection of plant pathogens in plant tissue. A specific, indirect, enzyme-linked immunosorbent assay (ELISA) was developed to detect X. fragariae from symptomatic plant tissue (Rowhani et al., 1994) The antibody assay did not react with bacterial strains of 16 pathovars of X. campestris or non-pathogenic bacteria from strawberry leaves A single cross-reaction occurred to a strain of X. campestris isolated from Nerium oleander. The assay detected bacteria directly from a visible lesion on a strawberry leaf . The level of sensitivity was at ca. 10 4 colony forming units (cfu) per ml However, bacteria on asymptomatic plants may not be detected by this ELISA assay. The polymerase chain reaction (PCR) has been used to amplify specific DNA sequences to detect and identify many plant pathogens including some members of the genus Xanthomonas (Henson and French, 1993; Louws et al., 1994). Leite et al (1994) utilized primers specific to regions of the hrp gene cluster which confers hypersensitivity and pathogenicity from Xanthomonas campestris pv. vesicatoria to detect pathogenic xanthomonads. Saprophytic and non-pathogenic bacteria lack the hrp genes (Stall and

PAGE 35

28 Minsavage, 1990; Lindgren et al., 1986 ) and DNA was not amplified by the hrp gene cluster primers. Differentiation of X. campestris pathovars was made by restriction endonuclease analysis (REA) patterns generated by digestion of the PCR products with frequent-cutting enzymes (Leite et al., 1994 andl995) The primers were used to detect X. c. vesicatoria in seed lots of naturally infected pepper and tomato (Leite et al., 1995). The sensitivity of detection by PCR is reported at 10 3 to 10 2 cfu per ml (Minsavage et al., 1994; Leite et al., 1994; Henson and French, 1993). In nested PCR sensitivity of detection is increased by using PCR products from an amplification as target DNA in a second round of amplification by a second set of primers internal to the first (Schaad et al., 1993). McManus and Jones (1995) reported an increase in sensitivity of 1,000 -fold with nested PCR over a single round of PCR amplification to detect Erwinia amylovora. Our objectives were to develop a sensitive and specific technique for detection of A". jragariae. Our approach was to design primers specific to the region of genomic DNA from X. fragariae related to the hrp genes of X. c. vesicatoria The survival of X. Jragariae on nursery strawberry plants in the field at two locations in Florida and dissemination to daughter plants was examined to understand the disease cycle of angular leaf spot in Florida. Materials and Methods Bacterial st rains and culture conditions. Strains of X. fragariae and non-pathogenic xanthomonads isolated from strawberry were maintained at 24° C on Wilbrink's medium (Koike, 1965) Pathovars of 'Xanthomonas campestris were cultured on nutrient agar (Difco

PAGE 36

29 Laboratories, Detroit, MI) and incubated at 28° C. Long term storage was at -70° C in 15% glycerol. Bacteria used for plant inoculations and DNA extractions were grown in 5 ml nutrient broth on a rotary shaker at 200 rpm for 16 h at 24° C. A rifampicin-resistant mutant of strain XF1425 was selected on Wilbrink's medium supplemented with 100 //g/ml of rifampicin by the gradient plate technique (Szybalski, 1952). Pathogenicity tests. Bacteria from overnight cultures in nutrient broth were centrifuged and washed three times with sterile water The concentration of cells was adjusted in either 10 mM MgSO 4 -7H 2 or sterile water to approximately 10 5 cfu per ml and sprayed to runoff on 'Sweet Charlie' strawberry plants placed under mist 24to 48h prior to inoculation. Inoculated plants were maintained under mist or put into growth chambers (Percival, Boone, IA) at 24° C with a 12 h photoperiod. Sequencing and primer design. The hrp primers RST2 and RST3 from the hrp gene cluster of X. axonopodis pv. vesicatoria (Leite et al, 1994) were used to amplify genomic DNA of 49 strains of X. fragariae . The PCR product from strain XF1425 of X. fragariae was isolated from agarose gel, cleaned by the Promega Wizard Kit (Promega, Madison, WI) and sequenced at the ICBR DNA Sequencing Facility, University of Florida, Gainesville, FL. The nucleotide sequence was compared to sequences of the PCR products amplified by the same primers from one X. vesicatoria and two X. a. pv. vesicatoria strains using the Seqaid II computer program (Rhoads and Roufa, 1991). Four primers were selected from the sequence of X. fragariae based upon unique DNA sequences and low homology compared to the DNA sequences from the other bacterial strains. The four oligonucleotide primers

PAGE 37

30 were synthesized with a model 394 DNA synthesizer (Applied Biosystems, Foster City, CA) at the ICBR Facility, University of Florida, Gainesville, FL. PCR amplification and nested amplification. Total genomic DNA was extracted by the method described by Ausubel et al. (1987) PCR amplification was performed using a DNA Thermal Controller PT1 00 (MJ Research, Watertown, MA). Samples were in a total reaction volume of 50 fx\ and contained IX amplification buffer (Promega, Madison, WI), 100 ^M of each dNTP (Promega, Madison, WI), 50 //M of each primer, 1 .25 U Taq DNA polymerase, and 100 ng of purified genomic DNA in 3 /A of TE (10 mM Tris and 1 mM EDTA, pH 8.0) buffer. Each reaction was overlaid with 50 ,ul of sterilized mineral oil (Sigma) for a total volume of 100 ^1 in sterile 0.6 ml microcentrifuge tubes. In later experiments, the thermocycler was equipped with a heated lid controller (The Hot Bonnet™, MJ Research) which eliminated the need for the mineral oil overlay. Amplification of the DNA proceeded after template DNA was denatured at 95° C for 2 min followed by thirty amplification cycles and a final extension step at 72° C for 5 min. For primer set XF9 and XF1 1, each amplification cycle consisted of denaturation at 95° C for 30 s, annealing at 65° C for 30 s, and extension at 72° C for 45 s. For primer XF12 with either primer XF9 or XF10, the program was identical except the annealing temperature was 58° C. For nested PCR the first round of amplification was as described with primers XF9 and XF1 1 In the second round of amplification, a 3 /A sample from the first amplification mixture was used with primers XF9 and XF12 and all other ingredients were added at the concentrations described above. The PCR cycle program for the primers XF9 and XF 1 2 was

PAGE 38

31 used. In all PCR runs, including the nested assays, a water sample was used as a negative control. Restriction endonuclease analysis fREA) of PCR products The PCR samples with the overlay of mineral oil were cleaned by the method of Minsavage et al. (1994). Samples without the mineral oil were used directly in restriction reactions An 8 /il sample of the PCR product was digested with restriction endonuclease Sau3AI, Haelll or Cfol under conditions specified by the manufacturer (Promega) Restricted products were separated in 4% agarose gel (3% NuSieve, 1% SeaKem GTG, FMC BioProducts, Rockland, ME) containing 0.5 ^g/ml ethidium bromide in TAE buffer at 8 V/cm as described by Leite et al. (1994). DNA molecular weight marker XI (Boehringer Mannheim, Indianapolis, IN) was used for standard weight markers. Gels were photographed over a UV transilluminator with type 55 Polaroid film (Polaroid Corp., Cambridge, MA). Detection of bacteria from infected plant tissue. The sensitivity of the assay in the presence of plant tissue was determined by adding known concentrations of bacteria to plant samples A 9-mm diameter disk of plant tissue was macerated with mortar and pestle in 200 //l of phosphate buffer (pH 7.0) containing 5% polyvinylpolypyrrolidone (PVP40, Sigma Chemical Co., St. Louis, MO) and 0.02 M sodium ascorbate (PPA). The mixture was incubated at room temperature for a minimum of 1 h. The volume was adjusted to 582 /A with TE and DNA extraction proceeded as described above. A minimum of three experiments with three replications of each treatment for each primer set and nested reaction was performed REA was performed on the final PCR products.

PAGE 39

32 Strawberry plants with leaves exhibiting typical lesions resulting either from natural field infections or spray inoculation were assayed by removing a 9-mm diameter disk of tissue surrounding a lesion and proceeding as described above Tomato leaf tissue was used as a negative control. For confirmation, a sample of the ground tissue suspension was plated on Wilbrink's medium amended with 0.05% Bravo 720 (chlorothalonil; ISK Biosciences, Mentor, OH). Plates were incubated at 24° C and colonies of X. fragariae were identified by colony morphology. Field experiments to examine oversummerinp survival of bacteria on nursery plants. Six-week-old rooted transplants of 'Sweet Charlie' were spray inoculated with X. fragariae strain XF1425 rif two weeks prior to planting. All plants transplanted into the field exhibited lesions of angular leaf spot. Plants were transplanted into the field on 27 June 1995 at Gulf Coast Research and Education Center-Dover, FL and on 29 June 1995 at Gulf Coast Research and Education Center-Bradenton, FL. Plants were sampled over a 14week period through 27 September 1995. Another set of inoculated plants were placed into Percival growth chambers at 12 h photoperiod and 24° C and maintained for the duration of the experiment Plants which were not inoculated and did not exhibit symptoms of angular leaf spot were placed in a greenhouse At two-week intervals, five plants from the field, one plant from the growth chamber, and one plant from the greenhouse were sampled. Two samples of three daughter plants of the inoculated plants were sampled at 132 days after planting from both locations. Individual plants were assayed as follows. All the leaves from a single plant were removed and placed in a flask with 200 ml of phosphate buffer containing 02 % Tween 20. The samples were shaken either on a wrist action (Burrell Corp., Pittsburgh, PA)

PAGE 40

33 or a rocker platform (Bellco Biotechnology, Vineland, NJ) for 2 to 16 h. A 200 fj.\ sample was plated onto Wilbrink's medium plus 100 ,ug/ml rifampicin plus 5% Bravo 720 and incubated at 24° C for 72 h. Colonies characteristic of X. fragariae were identified by amplification with the XF-specific primers and by REA. The remainder of the phosphate buffer sample was concentrated by vacuum filtration onto a 0.45 fan membrane disk (Millipore Corp, Bedford, MA). The disk was washed in 1.5 ml of TE and the suspension centrifuged 5 min at 14, 000 x g. The pellet was resuspended in 582 ^1 TEPA (TE buffer containing 5% PVP40 + 0.02 M sodium ascorbate) and incubated at room temperature 1 h. DNA was extracted as described above. The crown of the plant was sectioned, macerated by mortar and pestle in 10 ml of PPA. The plant tissue debris and PVPP were collected by centrifugation at 1,000 x g for 1 min. The supernatant with the bacterial cells was removed to a clean centrifuge tube and centrifuged at 14,000 x g for 5 min. The pellet was resuspended in 582 \A TE and DNA extraction proceeded as described above. A sample was plated onto Wilbrink's agar plus 100 /^g/ml rifampicin and 0.5% Bravo and incubated at 24° C for 72 h. The nested PCR reaction and REA were performed as described above. Results Pathogenicity assays. All strains identified as X. fragariae by colony characteristics caused disease symptoms typical of angular leaf spot on strawberry plants. Many other xanthomonads were isolated from strawberry tissue but they did not cause symptoms of angular leaf spot.

PAGE 41

34 Specificity of primers. For all strains of X. fragahae, the PCR products amplified by primers RST2 and RST3 were ca. 840-bp (Fig. 3-1). The REA profiles that resulted from restriction of the PCR products with Cfol or HaelU were the same for all the strains of A". fragahae (data not shown). Genomic DNA from non-pathogenic strains of Xanthomimas isolated from strawberry was not amplified by the primers. The four primers synthesized were: XF9 (5' TGGGCCATGCCGGTGGAACTGT GTGG3'); XF10 (5' TGGAACTGTGTGGCGAGCCAG 3'); XF1 1 ('5 TACCCAGCCGT CGCAGACGACCGG 3'); and XF12 (5 1 TCCCAGCAACCCAGATCCG 3'). Primers XF10 and XF 1 2 were internal to the other two primers. Primer XF9 paired with XFI 1 or XF12 delineated a 537-bp or a 458-bp fragment, respectively (Fig. 3-1). Primers XF10 and XF12 delineated a 448-bp fragment. PCR products conformed to the estimated sizes based upon sequence data. All 49 strains ofX. fragahae were amplified with primer sets XF9 and XFI 1, XF9 and XF12, and XF10 and XFI 2. DNA from strains other than X. fragahae was not amplified by primer XF9 paired with XFI 1 or XFI 2. Strains tested were ATCC type strains of Xanthomonas campeslhs pathovars begoniae, campeslhs, carotae, celebensis, glycines, incanae, maniholis, musaceariim, papavehcola, pelargonii, phaseoli, poinsettiicola, raphani, taraxaci, vignicola, vitians, and nine non-pathogenic strains of xanthomonads isolated from strawberry Except for X. c. pelargonii, genomic DNA from these strains could not be amplified with primer set XF10 and XFI 2. Due to the non-specific amplification of DNA from X. c. pelargonii by primer XFI paired with XFI 2, only primer XF9 paired with either XFI 1 or

PAGE 42

35 1 000 basepair 500 basepair Figure 3-1. Agarose gel of the PCR products generated by amplification of genomic DNA from a strain of Xanthomonas fragariae with the A/p-primers RST2 and RST3 (Lane 1 ) and the XF-specific primers XF9 paired with XF1 l(Lane 2) and XF9 paired with XF12 (Lane 3), and molecular weight marker XI (Boehringher-Manheim, Indianapolis, IN; Lane 4).

PAGE 43

36 XF12 was used in experiments to detect DNA from X. fragariae in the presence of plant materials REA of PCR products. PCR products from DNA of X. fragariae amplified with primer set XF9 and XF12 restricted with Cfol, Sau3AI, or Haelll produced distinct banding patterns for each enzyme (Fig 3-2). Five bands resulted from Cfol restriction (125, 108, 90, 75, and 60 bp), and three bands each from the Haelll restriction (258, 1 50, and 50 bp) and Sati3M restriction (250, 125, and 83 bp). Polymorphisms were not observed among strains of X. fragariae in the REA of the restricted PCR products for the enzymes tested. Sensitivity of detection of X. fragariae in plant tissue by PCR. X. fragariae was detected in the presence of plant tissue by amplification with primer XF9 and either XF1 1 or XF12 and by nested PCR of the two primer sets The level of sensitivity of detection of bacterial cells by amplification with one round of PCR was ca. 10 s to 10 4 cfu per ml. The nested PCR increased sensitivity of the assay 1000-fold (Fig 3-3) and enabled detection of ca. 18 cfu per ml. PCR was used to detect bacteria directly from infected plant tissue both from inoculated strawberry and from plant samples collected in the field The PCR product was generated from a single lesion on plant material ground and processed representing a bacterial population of 10 4 or above as determined by dilution plating. Identification of X. fragariae was confirmed by REA of PCR products. The nested technique was not necessary to detect bacteria from a lesion on leaf tissue as a single round of PCR amplification was sufficient to obtain a positive result.

PAGE 44

37 lhiu\U Cfol 100 bp Figure 3-2. NuSieve agarose gel showing the restriction patterns generated after digestion of the PCR product amplified with the primers XF9 and XF12 from the genomic DNA of strains of Xanihomonas fragariae. The PCR product was digested by restriction endonucleases Cfo\ and Haelll and the molecular weight marker XI from Boehringer Mannheim is shown.

PAGE 45

.58 XF9 and XF 1 1 Nested XF9 and XF 1 2 500 bp 7654321 87654321 Log cfu per ml Figure 3-3. Agarose gel showing the sensitivity of detection by PCR and nested PCR in the presence of plant tissue The genomic DNA from various concentrations of cells was extracted and amplified by primers XF9 and XF1 1 and used in a second round of nested amplification with primers XF9 and XF12.

PAGE 46

39 Detection of bacteria on plants in oversummering experiments. All plants exhibited visible symptoms of angular leaf spot at planting. However, by 34 days after planting, symptoms of angular leaf spot were not visible on plants (Table 3-1). Bacteria were recovered on the selective medium and colonies of X. fragariae were identified by morphology, resistance to rifampicin, DNA amplification by the XF-specific primers, and REA. Bacteria were recovered on selective medium from a few samples throughout the experiment, including the final date, indicating that bacteria were viable on plants throughout the summer. Xanthomonas fragariae was detected from leaf samples by nested PCR on each sample date. The percentage of leaf samples which were positive by the nested PCR assay ranged from 20% to 100% for the sample dates. From the leaves, all samples were initially positive but, by 5 1 days after planting, the number of positive samples declined to 20%. By 92 days after planting, all samples were again positive. A single round of amplification was occasionally sufficient to identify positive samples. In the crown of plants, bacteria were detected by PCR amplification on all sample dates except one. REA analysis of the PCR products confirmed banding patterns typical of X. fragariae. Plants inoculated and placed in the growth chamber at the favorable temperatures for disease development were positive at each sampling date except one. Non-inoculated plants from the greenhouse were positive by this technique in approximately 20% of the samples.

PAGE 47

40 > re -a c re u a. >, x> -a u -^ o c« > u I sE o <£: c/3 u oo a CO C c u r-r-, U .3 H E % c O > re T3 CM U £ i> CO E 3 u u c o TO o c -J OUo«ooo H H Z (N t o o o o o o o t "* ^r -^o oo CM o o o o ^j o + + + + + + + I I l l + + + H H o co o o O VO 'tfo o o o o o o o X "t CM CM vO ^ ^ o o o o o o o nO CN ^r + + + + , + + 111 + I I I i — r— o nO i/~. CM O — C-> Tf nO 00 On > c Q U <£ o o o H ^ Z CM '-' CM TT O O O O O O O T no rr o tt oo oo o o o o o o o CM CM + + + + + + + i ~~t~~ t i i + + + + I I I I U O &5 O O O O H cm Z cm no ^r g o o o o g ~ CM CM CM -sT X o o o o o o o NO CM CM CM + + + + + + O TT O Tf OO CM o cm m -sT v) no oo C o c u -a re ic o — re U "a. E re Pti U Dm >. c 8 .2 \p to '§; o o vz a, o, li E + £0 oei § U •« Q CO " O > £ le I+ -O g.e > re re ; 0C I = g O fe CM C >n CM CO ° t3 <^ «5 u 'E _o o o "re P c « 3 E c o -s ch .2 u •« a. 2 . B ~° Q W fi 'N 6« <2 ^ S c be "51^ 2 «» II 81 ^ S to ^ oo £ CO u Z o u E — a, >. 5)-° .S S > C CO > •i 3 « o op § •E a, E "a CO CD a. D Q. § E > CO CO « CO O M O u — I-.2 -o u >. re re i-o u o Z u re 3 u c c

PAGE 48

41 Discussion Leite et al. (1994) used the sequence variation within the hrpB operon among plantpathogenic xanthomonads to select primers with different specificities and this approach was successful in our studies to identify primers specific to X. fragariae. The /irp-primers RST2 and R.ST3 amplified DNA from all strains of X. fragariae and analysis of the PCR products by REA showed no polymorphism within this region The homology of the amplified region presented a good site to select primers universal to X. fragariae. Primers XF9, XF1 1, and XF12 designed from these unique sites were specific for amplification of DNA only from strains of X. fragariae The primer XF10 was responsible for the non-specific amplification of a strain of X. c. pelargonii in preliminary tests and therefore was not used in later experiments. Interestingly, an ELISA test developed for identification of X. fragariae in vitro cross reacts to some related bacteria including X. c. pelargonii (Maas, 1995). Perhaps a relationship between X. fragariae and X. c. pelargonii exists which is not reflected by traditional taxonomic techniques (Hildebrand et al., 1990; Vauterin et al., 1994, Maas, 1995). The level of sensitivity for most detection techniques for specific bacteria (Pickup, 1991; Schaad et al., 1982 ) are comparable to the level which was achieved with a single round of PCR amplification PCR was reported to be more sensitive than ELISA (Leite et al., 1995) and with our primers, a single round of PCR is 10 times more sensitive than the ELISA test for X. fragariae developed by Rowhani et al. (1994). In addition, their antibodies cross-reacted with another xanthomonad whereas we were able to eliminate the non-specific reaction by discontinuing use of primer XF10 Although this level of detection is adequate

PAGE 49

42 to detect bacteria in symptomatic leaf tissue, we were concerned with the detection of bacteria from asymptomatic tissue. The X. fragariae primers were suitable for use in the nested technique and the sensitivity of the PCR reaction was increased by 1,000-fold. This level of detection was achieved in assays to detect the bacterium in the presence of plant tissue and is well below the number of cells needed to cause visible lesions. Therefore, the nested technique is applicable for detection of bacteria in association with asymptomatic tissue. Cross-contamination among samples and contamination of PCR reagents are problems associated with the nested technique (McManus and Jones, 1995). We experienced false positives in a number of initial experiments and finally determined the cause to be contaminated mineral oil. Although we eliminated the need for mineral oil overlay by using a hot bonnet, extreme care is needed to avoid contamination and subsequent false positives. In our experiments, negative control samples were always included for each amplification round of the nested assay Aerosols may also contribute to false positives and care must be taken to prepare the mixtures for PCR in a sterile environment and negative control samples should be included to check for aerosol contamination. In the field experiments, a problem encountered with the nested technique was amplification of bacterial DNA from the negative control plants. Although these plants were physically isolated from a contamination source during the experiments; they were initially obtained from fields at GCREC-Dover where plants infected with angular leaf spot were located. The plants may have been infested with the bacterium or cross-contamination of samples may have occurred during preparation of plant samples from handling infected material or using non-sterile instruments. Aerosols may

PAGE 50

43 also have been a source of contamination of PCR samples (Innis et al., 1990). However, contamination from these sources would be detected by amplification in the negative controls. Our negative controls were included in every PCR run and were consistently negative Therefore contamination of PCR reagents or aerosols would have been ruled out Confirmation that the PCR product is from amplification of the target DNA is possible by REA (Leite et al. 1994). The REA was used to identify PCR products amplified by the hrp-primers from cells of X. c. vesicaloria added to seed washing of tomato and pepper (Leite et al. 1995). Similarly, we applied REA to identify PCR products amplified by the XFprimers The profiles of the restricted PCR products were distinct for each enzyme. The profiles from REA should be different for unrelated organisms since it would be highly unlikely that the same restriction sites would exist for two heterologous pieces of DNA (Leite etal,. 1995). The nested PCR technique was useful to detect bacteria on strawberry plants in nurseries in Florida. Symptoms and recovery of the rifampicin-marked strain on a selective medium were not useful to identify bacteria when populations were extremely low. However, recovery of the rifampicin-marked strain from some samples indicated that the bacteria were viable throughout the summer and that PCR was detecting viable bacterial cells. Visible symptoms on plants disappeared soon after placement in the field and recovery on media was difficult due to the slow-growing nature of X. fragariae and overgrowth by contaminants. While bacterial populations were not enumerated, the levels were deduced to be below 10 3 cfu per ml, since bacteria were detected in the nested PCR but not by single round of amplification. This level of sensitivity would be useful to screen asymptomatic nursery plants

PAGE 51

44 to detect plants contaminated with X.fragariae. This is a concern of nursery-plant producers and regulatory agencies of plant shipment (Maas, 1995). Currently, plants from nurseries in California are being screened by this technique to determine its usefulness in such an application. The field studies have implications regarding the disease cycle of angular leaf spot in Florida. The decline and later increase in the number of positive samples through the summer would indicate that populations of the bacteria declined throughout the summer and increased when more favorable conditions in the form of cooler weather occurred. The bacterial populations did not completely die. Therefore, to eliminate disease, the production of plants in nurseries in Florida would have to begin with plants free of the bacterium. Researchers have reported the systemic movement of X. fragariae in plants (Hildebrand et al., 1967; Milholland et al., 1993). In our research, inoculation of strawberry plants resulted in survival of bacteria on leaves and the crown for extended periods under conditions not optimum for growth of the bacteria In addition, bacteria were detected by PCR technique on daughter plants in the field. Dissemination to the daughter plants could have been due to either systemic movement through the vascular system of the runner or dispersal of bacterium by mechanical means

PAGE 52

CHAPTER 4 GENOMIC RELATEDNESS OF XANTHOMONAS FRAGARIAE ON STRAWBERRY BY FATTY ACID METHYL ESTER AND RESTRICTION LENGTH FRAGMENT POLYMORPHISM ANALYSES Introduction Xanthomonas fragariae causes angular leaf spot disease on strawberry (Fragaria species and Fragaria x ananassa Duch. cultivars) While historically the disease has not been a major deterrent in strawberry production, the disease is becoming more important because of an increase in prevalance of the disease in fruit production fields in Florida and the lack of effective control measures (Maas, 1995; C. Chandler, pers. comm) Losses in yield are caused by this disease (Epstein, 1 966; Howard, 1971). In addition, regulatory issues regarding the transportation of infected plant material may impact the nursery plant industry. International movement of infected plants is blamed for the introduction of angular leaf spot into Greece and New Zealand (Panagopoulos et al., 1978; Dye and Wilkie, 1973). The European Plant Protection Organization listed X. fragariae as a quarantine pest and the FAO/IPGRI recognized it as a potential risk in international movement of strawberry germplasm (Maas, 1995). Nursery-plant producers in the United States and Canada are pressured to provide disease-free plants by buyers from foreign markets and by farmers who refused to buy infected transplants. 4S

PAGE 53

46 The epidemiology of the X. fragariae must be understood before effective control strategies can be devised. The delineation of bacterial populations is imperative to such studies. The relationship of X. fragariae to other members of the genus Xanthomonas has been examined (Hildebrand et al, 1990; Hodge et al., 1992; Vauterin et al, 1994), but the genetic variability within the species has not been reported (Maas, 1995). The information from studies to compare strains oiX fragariae at the genetic level could be useful to identify the origins and spread of specific strains or genetic types This information could be applied in international tracking of pathogen populations for quarantine programs. In addition, the identification of genetic types is a prerequisite to identify sources of resistance in strawberry. Strawberry cultivars exhibit levels of susceptibility or tolerence to X. fragariae; only F moschata Duch. appeared to be immune (Hazel, 1981; Hazel and Civerolo, 1980; Kennedy and King, 1962b). A screening program to identify genes for resistance must incoporate representives of the genetic variants in the screening process otherwise the resistance may be overcome quickly by genetic variants. Studies to characterize populations of bacteria have used biochemical and molecular biological techniques. Protein staining and fatty acid analysis have been useful to detect differences at the metabolic level within species of Xanthomonas (Bouzar et al. 1994; Stall et al, 1994; Graham et al, 1990; Hodge et al, 1992). The polymerase chain reaction (PCR) has been used to analyze differences between pathovars and strains of bacteria (Hensen and French, 1993) Primers specific to the hrp-gene cluster of Xanthomonas campestris pv. vesicatoria amplifed genomic DNA and restriction enzyme analysis of the PCR product differentiated between many pathovars and species of Xanthomonas (Leite et al, 1994 and

PAGE 54

47 1995). Primer sets from conserved repetitive bacterial DNA elements generated genomic fingerprints which were used to differentiate bacteria (De Bruijn, 1992; Louws et al., 1994). Restriction length fragment polymorphism (RFLP) of the bacterial genome digested with rarecutting restriction endonuclease and resolution by pulsed field gel electrophorisis (PFGE) has been used to type many bacterial strains (Cooksey and Graham, 1989; Egel et al., 1991; Smith etal., 1995). In this study, a collection of strains of X.fragariae from the United States and Canada were examined for genetic diversity. Analyses was by RFLP profiles generated by PFGE and by fatty acid methyl ester (FAME). Materials and Methods Bacterial strains. Strains of X.fragariae used in this study are listed in Table 4-1. Strains were previously rated for pathogenicity on strawberry (Roberts, unpublished). Geographic origin denotes origin of infested plant material Strains were cultured on Wilbrink's medium (WB)(Koike, 1962) at 24° C and long term storage was in 15% glycerol at -70° C. Single colonies were transfered to nutrient broth (Difco, Detroit, MI) and shaken for 16 to 20 h at 200 rpm at 24° C. Restriction fragment length polymorphism. The method used for restriction endonuclease analysis was described by Egel et al. (1991) and Cooksey and Graham (1989) except for the following modifications. Cells (1.5 ml of 5 * lO'cfu per ml) were washed in 1 ml of TE (10 mM Tris, ImM EDTA, pH 8.0) and resuspended in 0.5 ml of TE. An equal

PAGE 55

48 Table 4. 1 . The geographic source, collection date, and group determined by fatty acid methyl ester (FAME) and restriction fragment length polymorphism (RF LP ) for strains used in this study. Geographic Strain Source" Year Source b FAME' RFLP 1238 CA 1990 ARC 2 D 1240 CA 1990 ARC 2 D 1241 CA 1990 ARC 8 B 1242 CA 1990 ARC 8 B 1243 CA 1990 ARC 8 B 1245 CA 1990 ARC 4 B 1246 CA 1990 ARC 8 1249 CA 1990 ARC 4 1250 CA 1990 ARC 5 C 1290 CA 1989 ARC C 1291 CA 1989 ARC 1 B 1293 CA 1989 ARC 7 B 1295 CA 1989 ARC 8 1296 CA 1989 ARC 8 B3 1298 CA 1989 ARC 8 B 1424 FL 1992 ARC C 1425 FL 1992 ARC 3 B2 1426 FL 1992 ARC B 1427 FL 1992 ARC 3 B 1428 FL 1992 ARC 3 B2 1429 FL 1992 ARC 3 B 1431 FL 1992 ARC B 1514 CA 1993 ARC C 1515 CA 1993 ARC 6 B3 1516 NC 1993 ARC B2 1517 NC 1993 ARC 7 B3 1518 NC 1993 ARC 8 B 1519 NC 1993 ARC B 1520 CA 1993 ARC 5 C 1523 CA 1993 ARC 5 C 1524 CA 1993 ARC 5 C 1525 CA 1993 ARC C 1526 CA 1993 ARC 8 B " Geogra phic origin of pi ant material from which bacteria were isolated. CAN=Canada. b ARC= A R. Chase; ATCC= American Type Culture Collection. 'Blank space indicates that group was not determined.

PAGE 56

49 Table 4. 1 continued. The geographic source, collection date, and groups determined by fatty acid methyl ester (FAME) and restriction fragment length polymorphism (RFLP ) for strains used in this study. Geographic Strain Source" Year Source b FAME' RFLP 1532 FL 1993 ARC 3 1533 WI 1993 ARC 5 B 1534 Wl 1993 ARC B3 100 FL 1993 This study 1 B 101 FL 1993 This study 1 103 FL 1993 This study B 104 FL 1993 This study 1 B 105 FL 1993 This study 3 B 106 FL 1993 This study 3 107 FL 1993 This study 1 108 FL 1993 This study 1 B 113 FL 1993 This study 5 B3 114 FL 1993 This study 3 B 115 FL 1993 This study 4 B 116 CAN 1993 This study B 117 CAN 1993 This study B3 119 CAN 1993 This study 9 D 124 CAN 1993 This study B 125 CAN 1993 This study 6 126 CAN 1993 This study 4 127 CAN 1993 This study 6 128 CAN 1993 This study 4 B 129 CAN 1993 This study 6 B3 138 CAN 1993 This study 6 B 146 CAN 1993 This study 4 Bl 153 CAN 1993 This study 6 B 33239 MN ATCC 9 A "Geographic origin of plant material from which bacteria were isolated. CAN=Canada. b ARC= A. R. Chase; ATCC= American Type Culture Collection. 'Blank space indicates that group was not determined.

PAGE 57

50 volume of 1% Seakem Gold agarose solution (10 mM Tris [pH 8.0], 10 mM MgC12, 0.1 mM EDTA [pH 8.0], 1 % Seakem Gold agarose (FMC BioProducts, Rockland, ME) [wt/vol] in sterile filtered water) was added. Plugs were made, lysed, washed, and stored as described by Egel et al. (1991). Two sizes of plugs were utilized. A 4 * 8 mm slice of plug was restricted in a volume of 200 iA of restriction buffer (as recommended by manufacturer, Promega, Madison, WI) and inserted in wells made with a 10-well comb (Bio-Rad, Richmond, CA). A 4 * 4 mm square piece of plug was restricted in a volume of 100 ^1 and placed in wells made by a 20-well comb (Bio-Rad). Restriction enzyme was added at the following concentrations: Xbal at 40 U; Spel at 30 U, Vspl at 30 U (Promega). The pieces were placed into wells in a 1 .2% GTG agarose gel made with 0. 5X TBE. Wells were sealed with the 1% Seakem Gold solution. The gel was placed in a Bio-Rad CHEF DR II unit containing 1.8 liters of 0.5x TBE and run at 200 V (15V/cm of gel). Pulsed times for plugs digested with Xbal or Spel were at 4 s for 1 h followed by 8 sec for 1 8 h. Plugs digested with Vspl were pulsed at 4 s for 1 h and subsequently at 12 s for 17 h. Lambda DNA in 48.5 KB concatamers (FMC BioProducts) was used in the first and last lanes of each gel Gels were stained in 5 mg of ethidium bromide per liter and photographed with type 55 Polaroid film. The position of bands were assessed visually or by analysis with the Gelmeas computer program. Similarity values were calculated as described by Egel et al. (1 991) with the mathematical equation proposed by Nei and Li (1979) based upon the proportion of shared DNA fragments. The estimate of the number of nucleotide substitutions per site was used to calculate the genetic divergence by the iterative method of Nei (1987) with the program in SAS described by Leite et al. (1994). The KITSCH program from the PHYLIP

PAGE 58

51 computer pakage (Felsenstein, 1995) was used to create a rooted phylogenetic tree by the Fitch-Margoliash method (Fitch and Margoliash, 1967). The input data was as described for the combined Spel, Xbal, and Vspl digestion data (Stall et al., 1994). Fatty acid composition. Strains of X. fragariae were inoculated onto Trypticase Soy Broth agar (TSBA) and grown for 48 h at 24° C. Conditions were changed from the standard MIDI procedures which required that cells be grown on TSBA for 24 hours at 28° C because strains of X. fragariae produced insufficient growth at these conditions. Cellular fatty acids were extracted and derivatized to their fatty acid methyl esters as described (Sasser, 1990b). A library comprised of strains of X. fragariae was created using the MIDI Library Generation System (LGS), software version 3.3. FAMEs were analyzed by the MIDI Mcrobial Identification System, software version TSBA 3.5. The qualitative and quantitative differences in the fatty acid profiles were used to compute the Euclidian distance to each strain. Strains within six Euclidian distance units, the cut-off for subspecies (Sasser, 1990a), were grouped in the same cluster. Results RFLP-PFGE Restriction endonucleases Xbal and Spel generated genomic DNA fragments from 5 to 400 kb (Fig. 4-1). Typically, a strain profile contained ten DNA fragments greater than 100 kb. Analysis of the 52 strains resulted in four RFLP groups, designated A through D, which were identical by these two enzymes. Group A contained the ATCC strain only and represented 2% of the strains. The B, C, and D groups had 77%, 15%

PAGE 59

52 and 6% of strains, respectively. A third endonuclease, Vspl, generated fragments of DNA of the appropriate size for PFGE analysis and the profiles by this endonuclease separated strains of the A, C, and D groups identical to the previously determined groups. Analysis of the B group with the Vsp\ endonuclease subdivided it into three subgroups, denoted Bl, B2, and B3. The Bl contained 13%, B2 contained 20%, and B3 contained 67% of the B strains tested. Designation of RFLP groups for strains are summarized in Table 4-1. The dendrogram derived from RFLP analysis of these strains showed the B groups were very closely related by within 0.006 genetic distance units (Fig. 4-2). The C group was the most closely related to the B groups, then group A, and finally group D. The D group was the most remote from the B group at 0.01864 genetic distance units. A strain of A", c. vesicatoria was used as the statistical outlier and fell at 0.045 genetic distance units from the D group. Fatty Acid Methyl Ester Analysis. The 50 strains of X. fragariae were placed into nine subgroups based upon the MIDI "10%" rule (Sasser, 1990b). The subgroups for individual strains are summarized in Table 4-1 The majority of strains was identified as six closely related subgroups which can be visualized quantitatively by three major acids, 16:1 w 7 cis, 15:0 Anteiso, and 15:0 iso (Fig.4-3). The 15:0 iso fatty acid comprised approximately 34% to 54% of FAME profile. The ATTC type strain and three closely related strains were qualitatively differentiated by the absence of palmitic acid, 16:1 w 7 cis. The dengrogram unweighted pair analysis (Fig. 4-4) separated strains into four clusters at the Euclidian distance of six units. The four clusters, designated FAME groups

PAGE 60

53 Figure 4-1. Agarose gel showing the restriction fragment length polymorphisms of genomic DNA of strains ofXanthomoiiasfragariae after restriction with a rare-cutting endonuclease, Vspl, and separated by pulsed-field gel electrophoresis. Lambda marker in 48.5 kb concatamers is shown in side lanes.

PAGE 61

L 54 I B3 Bl B2 C A D X.c. vesicatoria J 0.04 0.02 Genetic Distance Figure 4-2. Relationship of groups of Xanthomonas fragariae analyzed by restriction fragment length polymorphisms.

PAGE 62

55 % 16:1 w7cis %15:0ANTEISO % 15:0 ISO Figure 4-3. The groups of Xanthomonas fragariae distinguished by qualitative and quantitative differences in fatty acid methyl ester profiles graphed by three major acids.

PAGE 63

56 1 through 4, contained 16%, 6%, 4%, and 74% of the strains, respectively. The largest cluster, FAME group 4, could be subdivided at the Euclidian distance of 4 units into two groups. The ATCC type strain 33239 and one other strain comprised FAME 3 group. Discussion This research represents the first effort to analyze genetic variants of strains within the species of A", fragariae. The RFLP-PFGE and FAME analyses identified genetic variants within the population Methods of RFLP including PFGE have distinguished between closely related strains of pathovars within the genus Xanlhomonas campeslris(Ege\ et al., 1991; Gabriel et al., 1988; Hartung and Civerolo; 1989, Graham et al., 1990). In these works, the importance of the diverse nature of the endonuclease sites as related to strain characteristics such as host range, geographic origin, and pathogenicity was discussed. Unfortunately, the relationship of pathogenicity of the X. fragariae strains to the genetic groupings cannot be examined. Differences in pathogenicity among strains of X.fragariae were not detected by inoculation of the 52 strains onto two cultivars of strawberry Studies of bacterial populations on two cultivars which appeared to have different levels of susceptibility in preliminary tests were also inconclusive (Roberts, unpublished). Nor have other researchers reported differences in pathogenicity among strains of X. fragariae Such information might have been useful to determine significance of the RFLP-PFGE and FAME groups as related to pathogenicity.

PAGE 64

S7 Euclidian Distance 0.0 Cluster "10%" | 3.55 _l 7.05 10.5 14.1 I 3 3 7 4 4a 5 6 7 4b 5 Figure 4-4. Dendrogram of cluster analysis of fatty acid methyl ester profiles showing the four clusters ( 1 -4b) of strains of Xanthomonas fragariae . The designation of strains into the groups determined by the MIDI "10%" rule are shown for comparison. The Euclidian distance of six is the cutoff point for group determination by cluster analysis.

PAGE 65

58 The two methods used in this research might be useful to examine the evolution of the pathogen in future populations. The identification of two groups which contained only four strains total, including the ATCC type strain, from a population of 50 strains was of interest. By RFLP analysis, group A contained only the ATCC type strain from Minnesota which was collected over twenty years ago. However, by FAME cluster analysis, one other strain, XF1 19, which was isolated from an infected plant from Canada in 1993, grouped with the ATCC type strain. In RFLP-PFGE analysis, the XF1 19 profile was the same as the D group which contained strains XF1238 and XF1240, isolated from samples from California in 1990. Dendrogram analysis places these two groups close to each other. By FAME, these four strains lacked the 16:1 w 7 cis fatty acid which distinguished them from all the other X. fragariae. It is interesting to speculate that perhaps strain XF119 represent a 'bridge' between the A and D RFLP groups because of its intergroup relationship in that its RFLP profile is D but by FAME it is closer to the A group. It would be of further interest to examine future populations to determine the fate of the two groups represented by only four strains. The majority of strains was represented by the B and C groups. An extensive survey of the pathogen population in the United States and Canada would be useful to determine whether the population ratios of the various genetic variants are remaining relatively the same. Likewise, it would be interesting to determine the distribution of the three subdivisions within the B group. The cultivars of strawberry in commercial production are changed frequently (C. Chandler, pers comm.) This constant change in genotypes should influence the genetic composition of the population of X. fragariae if genes for resistance to angular leaf spot are

PAGE 66

59 'lost' or 'found' during the development of new cultivars. In the Philippines, the relative populations of two races of bacterial blight on rice were followed for 10 years. Researchers recorded a decline in the prevalence of the predominant race and a concurrent increase in another race. This change frequency of race, or bacterial genotype, occurred after the introduction of a gene for resistance in cultivated rice (Mew et al., 1992) Information regarding changes in the dominant genotype of both the plant and bacteriapopulations would be useful to plant breeders. The relationship between geographical origin and genetic variants is unclear because of the transportation of infected plants Plants are shipped from California to Canada where they are propagated and sent to Florida for field production each season. Infection of plants may have occurred at any point of the route, however symptomatic transplants from Canada are shipped to Florida. A single type of genetic variant was not found to be associated with plant material from a particular region of the U. S. or Canada Likewise, international movement of infested plants will make it difficult to determine if endemic populations of the organism exists outside the U. S.. The origin of this disease appears to be the United States (Maas, 1995).

PAGE 67

CHAPTER 5 DISCUSSION Field experiments which examined epidemiological aspects of angular leaf spot on strawberry established a decrease in yield on strawberry plants due to angular leaf spot. This report is the first to quantify the reduction in yield due to this disease. The 8to 10% loss determined in these studies is much lower than the 70to 80% loss estimated by Epstein (1966) in fruit production fields in Wisconsin. Production in northern regions of the U. S. is usually perennial and the age of the plants was not given. The loss was reported on the cultivar 'Sparkle' which is not produced commercially in Florida. Therefore, we are unable to compare our yield loss to his report Howard (1971) also reported unquantified yield losses due to this disease but on cultivars which are no longer in commercial production In our studies, the average yield loss observed both years was very similar despite the differences in the disease levels and the base yields between the two seasons. In 1995, the disease severity was 60% of the amount estimated in the previous season, however, the yield loss was actually higher by 2% for marketable yield. The 10% reduction in yield represents an estimated loss of 1233 kg of berries per acre. Strawberry fruit has a very high cash value and a decrease in revenue of 8to 10% represents a significant economic loss to producers Strawberry producers in Florida currently try to control angular leaf spot by application of copper compounds and by avoidance of the disease. An understanding of the 60

PAGE 68

61 epidemiology of angular leaf spot is necessary to determine the effectiveness and practicability of any control method. The survey of farmers' fields and the cold storage facility established that transplants arrive from northern nurseries infected with angular leaf spot at fairly high disease incidence Therefore, disease is introduced into the field on infected transplants. In our surveys, by the sampling method used in the first season, more plants were assessed for symptoms of angular leaf spot in the field compared to the number of plants in the random sampling of boxes in cold storage. This probably accounts for identification of diseased plants in nearly all of the farmer's fields versus less than in 25% of boxed plants. In addition, plants were examined until disease was found in the field which skewed the randomness of the survey. In field experiments, the spread from inoculated plants to noninoculated plants in nearby experimental plots was minimal. Thus, the incidence of plants with angular leaf spot in the field likely occurred from inoculum already present on the transplants. In the field experiments, if inoculum was present from another source, such as debris in the soil or an alternate host, most likely the disease on the noninoculated plants would have been more general and appeared earlier. Disease severity on noninoculated plants in the experimental plots was extremely low and most plants remained free of angular leaf spot. For the disease that occurred on noninoculated plants, the inoculum was probably transported mechanically from inoculated plants during harvests or disease readings. Progress of angular leaf spot developed similar to curves of pathogens described on other crops (van der Plank, 1963). A decrease in disease severity was seen about midway through the season in 1994. This reduction, or negative infection rate, was due to growth

PAGE 69

62 by strawberry plants which diluted the amount of disease relative to the total leaf area. The dilution effect of healthy plant tissue was described by van der Plank (1963). In 1995, a negative infection rate was seen very early in the epidemic This was because the symptoms of the disease almost disappeared from the field as the affected leaves died and were lost. The difference in disease progress between the two seasons did not appear to be from differences in rainfall since mean rainfall was approximately the same both years (data not shown). Mean temperatures also were similar; however the number of days at temperatures below 10° C in November and December 1 993 were much greater than in 1 994 Another difference between the seasons was that sprinkler irrigation was reduced in 1995 because of the change to drip irrigation Overhead irrigation, fog, high humidity, and rain are positive factors in disease development and spread (Epstein, 1966; Hildebrand et al., 1967; Kennedy and King, 1962b; Lai, 1978). The trend towards drip irrigation in commercial production fields should have a positive effect to reduce the spread and survival of the bacterium. Strawberry producers currently apply copper-based compounds to control angular leaf spot despite the phytotoxicity to plants caused by copper Application of cupric hydroxide plus mancozeb at the 1 * rate was phytotoxic to strawberry plants (Howard, 1973; this study) In our tests, the greatly reduced 0. 1 * rate did not harm plants and significantly reduced disease severity in 1995. The protectant application of bactericide at this rate and spray schedule was intended to reduce the total amount of inoculum and to prevent spread of the disease. The approach of frequent sprays at reduced concentrations may have potential to control this disease. More studies are needed to evaluate different rates, application intervals, and chemical mixtures to achieve maximum disease control while avoiding yield losses. An

PAGE 70

63 alternate approach might be to control the disease in the nursery by copper applications. The loss in yield due to phytotoxicity would not be an issue on nursery plants. The sprays would reduce the amount of initial inoculum on transplants and subsequently reduce the amount of disease in fruit production fields. The effect of sprays on the categories of yield were difficult to interpret because of the lack of consistent, significant differences between years. Leite et al. (1994) used the sequence variation within the hrpB operon among plantpathogenic xanthomonads to select primers with different specificities and this approach was successful in our studies to identify primers specific to X. fragariae. The hrp-primen RST2 and RST3 amplified DNA from all strains of X. fragariae and analysis of the PCR products by REA showed no polymorphism within this region. The homology of the amplified region presented a good site to select primers universal to X. fragariae. Primers XF9, XF1 1, and XF12 designed from these unique sites were specific for amplification of DNA only from strains of X. fragariae . The primer XF10 was responsible for the non-specific amplification of a strain of X. c. pelargonii in preliminary tests and therefore was not used in later experiments. Interestingly, an ELISA test developed for identification oiX. fragariae in vitro cross reacts to some related bacteria including X. c. pelargonii (Maas, 1995). Perhaps a relationship between X. fragariae and X. c. pelargonii exists which is not reflected by traditional taxonomic techniques (Hildebrand et al., 1990; Vauterin et al., 1994; Maas, 1995). The level of sensitivity for most detection techniques for specific bacteria (Pickup, 1991; Schaad et al., 1982 ) are comparable to the level which was achieved with a single round of PCR amplification. PCR was reported to be more sensitive than ELISA (Leite et al., 1995) and with our primers, a single round of PCR is 10 times more sensitive than the

PAGE 71

64 ELISA test for X. fragariae developed by Rowhani et al. (1994). In addition, their antibodies cross-reacted with another xanthomonad whereas we were able to eliminate the non-specific reaction by discontinuing use of primer XF10. Although this level of detection is adequate to detect bacteria in symptomatic leaf tissue, we were concerned with the detection of bacteria from asymptomatic tissue. The X. fragariae primers were suitable for use in the nested technique and the sensitivity of the PCR reaction was increased by 1,000-fold. This level of detection was achieved in assays to detect the bacterium in the presence of plant tissue and is well below the number of cells needed to cause visible lesions. Therefore, the nested technique is applicable for detection of bacteria in association with asymptomatic tissue. Cross-contamination among samples and contamination of PCR reagents are problems associated with the nested technique (McManus and Jones, 1995). We experienced false positives in a number of initial experiments and finally determined the cause to be contaminated mineral oil. Although we eliminated the need for mineral oil overlay by using a hot bonnet, extreme care is needed to avoid contamination and subsequent false positives In our experiments, negative control samples were always included for each amplification round of the nested assay. Aerosols may also contribute to false positives and care must be taken to prepare the mixtures for PCR in a sterile environment and negative control samples should be included to check for aerosol contamination. In the field experiments, a problem encountered with the nested technique was amplification of bacterial DNA from the negative control plants. Although these plants were physically isolated from a contamination source during the experiments; they were initially obtained from fields at GCREC-Dover where

PAGE 72

65 plants infected with angular leaf spot were located. The plants may have been infested with the bacterium or cross-contamination of samples may have occurred during preparation of plant samples from handling infected material or using non-sterile instruments. Aerosols may also have been a source of contamination of PCR samples (Innis et al., 1990). However, contamination from these sources would be detected by amplification in the negative controls. Our negative controls were included in every PCR run and were consistently negative. Therefore contamination of PCR reagents or aerosols would have been ruled out. Confirmation that the PCR product is from amplification of the target DNA is possible by REA (Leite et al. 1994). The REA was used to identify PCR products amplified by the top-primers from cells of X. c. vesicatoria added to seed washing of tomato and pepper (Leite et al. 1995). Similarly, we applied REA to identify PCR products amplified by the XFprimers. The profiles of the restricted PCR products were distinct for each enzyme. The profiles from REA should be different for unrelated organisms since it would be highly unlikely that the same restriction sites would exist for two heterologous pieces of DNA (Leite etal,. 1995). The nested PCR technique was useful to detect bacteria on strawberry plants in nurseries in Florida. Symptoms and recovery of the rifampicin-marked strain on a selective medium were not useful to identify bacteria when populations were extremely low. However, recovery of the rifampicin-marked strain from some samples indicated that the bacteria were viable throughout the summer and that PCR was detecting viable bacterial cells. Visible symptoms on plants disappeared soon after placement in the field and recovery on media was difficult due to the slow-growing nature of X. fragariae and overgrowth by contaminants

PAGE 73

66 While bacterial populations were not enumerated, the levels were deduced to be below 10' cfu per ml, since bacteria were detected in the nested PCR but not by single round of amplification This level of sensitivity would be useful to screen asymptomatic nursery plants to detect plants contaminated with X fragariae. This is a concern of nursery-plant producers and regulatory agencies of plant shipment (Maas, 1995). Currently, plants from nurseries in California are being screened by this technique to determine its usefulness in such an application. The field studies have implications regarding the disease cycle of angular leaf spot in Florida. The decline and later increase in the number of positive samples through the summer would indicate that populations of the bacteria declined throughout the summer and increased when more favorable conditions in the form of cooler weather occurred The bacterial populations did not completely die. Therefore, to eliminate disease, the production of plants in nurseries in Florida would have to begin with plants free of the bacterium. Researchers have reported the systemic movement of X. fragariae in plants (Hildebrand et al., 1967; Milholland et al., 1993). In our research, inoculation of strawberry plants resulted in survival of bacteria on leaves and the crown for extended periods under conditions not optimum for growth of the bacteria. In addition, bacteria were detected by PCR technique on daughter plants in the field. Dissemination to the daughter plants could have been due to either systemic movement through the vascular system of the runner or dispersal of bacterium by mechanical means. This research represents the first effort to analyze genetic variants of strains within the species atX fragariae. The RFLP-PFGE and FAME analyses identified genetic variants

PAGE 74

67 within the population. Methods of RFLP including PFGE have distinguished between closely related strains of pathovars within the genus Xanlhomonas campestris (Egel et al., 1991; Gabriel et al., 1988; Hartung and Civerolo; 1989, Graham et al., 1990). In these works, the importance of the diverse nature of the endonuclease sites as related to strain characteristics such as host range, geographic origin, and pathogenicity was discussed. Unfortunately, the relationship of pathogenicity of the X. fragariae strains to the genetic groupings cannot be examined. Differences in pathogenicity among strains of X. fragariae were not detected by inoculation of the 52 strains onto two cultivars of strawberry. Studies of bacterial populations on two cultivars which appeared to have different levels of susceptibility in preliminary tests were also inconclusive (Roberts, unpublished). Nor have other researchers reported differences in pathogenicity among strains of X. fragariae . Such information might have been useful to determine significance of the RFLP-PFGE and FAME groups as related to pathogenicity. The two methods used in this research might be useful to examine the evolution of the pathogen in future populations. The identification of two groups which contained only four strains total, including the ATCC type strain, from a population of 50 strains was of interest By RFLP analysis, group A contained only the ATCC type strain from Minnesota which was collected over twenty years ago. However, by FAME cluster analysis, one other strain, XF1 19, which was isolated from an infected plant from Canada in 1993, grouped with the ATCC type strain In RFLP-PFGE analysis, the XF1 19 profile was the same as the D group which contained strains XF1238 and XF1240, isolated from samples from California in 1990. Dendrogram analysis places these two groups close to each other. By FAME, these four

PAGE 75

68 strains lacked the 16:1 w 7 cis fatty acid which distinguished them from all the other X. fragariae. It is interesting to speculate that perhaps strain XF1 19 represents a 'bridge' between the A and D RFLP groups because of its intergroup relationship in that its RFLP profile is D but by FAME it is closer to the A group. It would be of further interest to examine future populations to determine the fate of the two groups represented by only four strains The majority of strains was represented by the B and C groups. An extensive survey of the pathogen population in the United States and Canada would be useful to determine whether the population ratios of the various genetic variants are remaining relatively the same. Likewise, it would be interesting to determine the distribution of the three subdivisions within the B group. The cultivars of strawberry in commercial production are changed frequently (C. Chandler, pers. comm). This constant change in genotypes should influence the genetic composition of the population of X. fragariae if genes for resistance to angular leaf spot are 'lost' or 'found' during the development of new cultivars In the Philippines, the relative populations of two races of bacterial blight on rice were followed for 10 years. Researchers recorded a decline in the prevalence of the predominant race and a concurrent increase in another race. This change in frequency of race, or bacterial genotype, occurred after the introduction of a gene for resistance in cultivated rice (Mew et al , 1992). Information regarding changes in the dominant genotype would be useful to plant breeders The identification of a relationship between geographical origin and genetic variants is unclear because of the transportation of infected plants. Plants are shipped from California to Canada where they are propagated and sent to Florida for field production each season

PAGE 76

69 Infection of plants may have occurred at any point of the route, however symptomatic transplants from Canada are shipped to Florida. A single type of genetic variant was not found to be associated with plant material from a particular region of the U. S. or Canada. Likewise, international movement of infested plants will make it difficult to determine if endemic populations of the organism exists outside the U. S .. The origin of this disease appears to be the United States (Maas, 1995).

PAGE 77

LIST OF REFERENCES Alippi, A. M, Ronco, B. L ., and Carranza, M. R. 1989. Angular leaf spot of strawberry, a new disease in Argentina: Comparative control with antibiotics and fungicides. Adv. Hort. Sci. 3:3-6. Ausubel, F. M, Brent, R. Kingston, R. E., Moore, D. D., and Seidman, K. 1987. Current Protocols in Molecular Biology. John Wiley & Sons, New York. Bouzar, H., Jones, J. B., Stall, R. E., Hodge, N.C., Minsavage, G. V., Benedict, A. A., and Alvarez, A. M. 1994. Physiological, chemical, serological, and pathogenic analyses of a worldwide collection of Xanthomonas campestris pv. vesicatoria strains. Phytopathology 84:663-671. Conway, K. E., Motes, J. E., Bostian, B., Fisher, C. G, and Claypool, P. L. 1987. Cercospora blight development on asparagus fern and effects of fungicides on disease severity and yield. Plant Dis. 71:254-259. 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. De Bruijn, F. J 1992. Use of repetitive (repetitive extragenic palindromic and enterobacterial repetitive intergeneric consensus) sequences and the polymerase chain reaction to fingerprint the genomes of Rhizobium meliloli isolates and other soil bacteria. Appl. Environ. Microbiol. 58:2180-2187. Dye, D W , and Wilkie, J. P. 1973. Angular leafspot of strawberry in New Zealand. N. Z. J. Agr. Res. 16:311-314 Egel, D. S., Graham, J. H., and Stall, R. E. 1991. Genomic relatedness of Xanthomonas campestris strains causing diseases of citrus. App. Environ. Microbiol. 57:2724-2730. Epstein, A. H. 1966. Angular leaf spot of strawberry . Plant Dis. Rep. 50:167. Felsenstien, J. 1995. PHYLIP Phylogeny inference package, ver.3.57. Department of Genetics, University of Washinton, Seattle, WA. 70

PAGE 78

71 Fitch, W. C, and Margoliash, E. 1967. Construction of phylogenetic trees. Science 155: 279-284. Fry, W. E. 1975. Integrated effects of polygenic resistance and a protective fungicide on development of potato late blight Phytopathology 65:908-91 1. Fry, W. E , Bruck, R. I., and Mundt, C. C. 1979. Retardation of potato late blight epidemics by fungicides with eradicant and protectant properties. Plant Dis. Rep 63:970-974. Gabriel, D. W., Hunter, J. E., Kingsley, M T., Miller, J. M , 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. Hartung, J. S , and Civerolo, E. L. 1989 Restriction fragment length polymorphisms dintinguish Xatithomonas campestris strains isolated from Florida citrus nurseries from X. c. pv. cilri. Phytopathology 79:793-799. Hazel, W. J. 1981. Xanthomonas fragariae , cause of strawberry angular leaf spot: Its growth, symptomatology, bacteriophages, and control. Ph.D. thesis. University of Maryland, College Park. 1 19 p. Hazel, W. J., and Civerolo, E L 1980. Procedures for growth and inoculation of Xanthomonas fragariae, causal organism of angular leaf spot of strawberry Plant Dis. 64:178-181. Henson, J. M., and French, R. 1993. The polymerase chain reaction and plant disease diagnosis. Ann Rev Phytopathol 31:81-109. Hildebrand, D. C, Palleroni, N. J , and Schroth, M. N. 1990. Deoxyribonucleic acid relatedness of 24 xanthomonad strains representing 23 Xanthomonas campestris pathovars and Xanlhomonas fragariae . J. Appl Bacterid. 68:263-269 Hildebrand, D. C, Schroth, M. N., and Wilhelm, S. 1967. Systemic invasion of strawberry by Xanthomonas fragariae causing vascular collapse. Phytopathology 57: 12601 26 1 Hodge, N. C, Chase, A. R., and Stall, R E 1992. Diversity of four species of Xanthomonas as determined by cellular fatty acid analyses. (Abst.) Phytopathology 82:1 153. Howard, CM. 1971. Occurrence of strawberry angular leaf spot, Xanthomonas fragariae, in Florida Plant Dis. Rep. 55:142.

PAGE 79

72 Howard, C. M., and Albregts, E. E. 1973. Strawberry report. Fungicide and Nematicide Tests 29:47. Howard, C M., Overman, A. J , Price, J. F., and Albregts, E. E. 1985. Diseases, nematodes, mites, and insects affecting strawberries. Florida. Agric. Experiment Stations, Inst, of Food and Agric. Sci , Univ. of Fl., Bulletin 857. Innis, M. A., Gelfand, D. H, Sninsky, J. J., and White, T. J 1990. PCR Protocols: A Guide to Methods and Applications. Academic Press, Inc , San Diego, CA. Jones, A. L. 1995. A stewardship program for using fungicides and antibiotics in apple disease management programs. Plant Dis. 79:427-432. Jones, J B„ Woltz, S S„ Kelly, R O , and Harris, G. 1991. The role of ionic copper, total copper, and select bactericides on control of bacterial spot of tomato. Proc. Fla. State Hort. Soc. 104:257-259. Kennedy, B. W., and King, T. H. 1962a. Angular leaf spot of strawberry caused by Xanthomonas fragariae sp. nov. Phytopathology 52:873-875. Kennedy, B. W., and King, T. H. 1962b. Studies on epidemiology of bacterial angular leaf spot on strawberry Plant Dis. Rep 46:360-363. Koike, H 1965 The aluminum-cap method for testing sugarcane varieties against leaf scald disease. Phytopathology 55:317-319. Lai, CM. 1978. Bacterial disease of strawberry. Plant Path. Reports. California Dept. Of Food and Agric. Lab Services. Vol. II. Leite, R. P., Egel, D. S„ and Stall, R. E. 1994. Genetic analysis of /i/p-related DNA sequences of Xanthomonas campestris strains causing diseases of citrus. Appl. Environ. Microbiol. 60:1078-1086. Leite, R. P., Jr., Jones, J. B , Somodi, G. C, Minsavage, G V., and Stall, R E 1995. Detection of Xanthomonas campestris pv. vesicatoria associated with pepper and tomato seed by DNA amplification. Plant Dis. 79:917-922. Leite, R. P., Jr , Minsavage, G V., Bonas, U., and Stall, R. E. 1994. Detection and identification of phytopathogenic Xanthomonas strains by amplification of DNA sequences related to the hip genes of Xanthomonas campestris pv. vesicatoria. Appl. Environ. Microbiol. 60:1068-1077.

PAGE 80

73 Lindgren, P B , Peet, R. C, and Panopoulos, N. J. 1986. Gene cluster of Pseudomonas syrwgae pv. "phaseolicola" control pathogenicity of bean plants and hypersenstivity on nonhost plants. J. Bacteriol 168:512-522. Louws, F. J., Fulbright, D. W., Stephens, C. T, and Bruijn, F J 1994. Specific genomic fingerprints of phytopathogenic Xanthomonas and Pseudomonas pathovars and strains generated with repetitive sequences and PCR Appl. Environ Microbiol. 60:2286-2295. Maas, J. L. (ed.) 1984. Angular leaf spot. Compendium of Strawberry Diseases. American Phytopathological Society, St. Paul, MN Maas, J. L., Pooler, M. R., and Galletta, G. J. 1995. Bacterial angular leafspot disease of strawberry: Present status and prospects for control. Adv. Strawberry Res. 14: 18-24. Marco, CM, and Stall, RE. 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. McManus, P. S., and Jones, A. L. 1995 Detection ofErwinia amylovora by nested PCR and PCR-dot-blot and reverse blot hybridizations Phytopathology 85:618-623. Milholland, R D., Ritchie, D. F , Daykin, M. E., and Gutierrez, W. A. 1993. Systemic movement of Xanthomonas fragariae in inoculated strawberry plants. (Abst ) Phytopathology 83: 1408. Minsavage, G. V., Thompson, C M., Hopkins, D. L„ Leite, R M. V B , and Stall, R. E. 1994. Development of a polymerase chain reaction protocol for detection of Xylella fastidiosa in plant tissue. Phytopathology 84:456-46 1 . Nei, M. 1987. Molecular Evolutionary Genetics. Columbia University Press, New York. 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. Panagopoulos, C. G., Psallidas, P. G, and Alivizatos, A S. 1978. A bacterial leafspot of strawberry in Greece caused by Xanthomonas fragariae Phytopath. Z. 91:33-38 Pataky, J. K , and Lim, S. M. 1981 Efficacy of benomyl for controlling Septoria brown spot of soybeans. Phytopatholology 71:438-442. Pickup, R. W. 1991. Development of molecular methods for the detection of specific bacteria in the environment. J. Gen. Microbiol. 137:1009-1019.

PAGE 81

74 Rhoads, D. D., and Roufa, D.J. 1991. Seqaid II Computer Program. Molecular Genetics Lab, Kansas State University, Manhattan, KS 66506. Ritchie, D. F., Averre, C. W., and Milholland, R. D. 1993. First report of angular leaf spot, caused by Xanthomonas fragariae, on strawberry in North Carolina. Plant Dis. 76: 1263. Rowhani, A., Feliciano, A. J., Lips, T„ and Gubler, W. D. 1994. Rapid identification of Xanthomonas fragariae in infected strawberry leaves by enzyme-linked immunosorbent assay. Plant Dis. 78:248-250. SAS Institute Inc. 1988. SAS/STAT User's Guide, ed. 6.03. SAS Institute Inc., Cary, N.C. Sasser, M. 1990. "Tracking" a strain using the the microbial identification system. Technical note 102. Microbial ID, Inc. Newark, DE. Sasser, M 1990. Identification of bacteria through fatty acid analysis. Methods in Phytobacteriology. Z. Klement, K , Rudolph, and D. Sands, eds Akademiai Kiado, Budapest. Schaad, N. W., Cheong, S. S, Tamaki, S., Hatziloukas, E., and Panopoulos, N. J. 1993. A viable cell enrichment, two-step, direct PCR technique for detection of Pseudomonas syringae pv. phaseoiicola in bean seeds (Abstr ) Phytopathology 83 : 1 342. Shaner, G., and Finney, R E 1977. The effect of nitrogen fertilization on the expression of slow-mildewing resistance in Knox wheat Phytopathology 67:1051-1056. Smith, J. J., Offord, L. C, Holderness, M„ and Saddler, G. S. 1995. Genetic diversity of Burkholderia solanaceantm (synonym Pseudomonas solanacearum) race 3 in Kenya. Appl. Environ. Microbiol. 4263-4268 Stall, R E„ Beaulieu, C , Egel, D„ Hodge, N. C, Leite, R. P., Minsavage, G. V., Bouzar, K, 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. Sys. Bacteriol. 44:47-53. Stall, R. E., Loschke, D. C , and Jones, J. B. 1986. Linkage of copper resistance and avirulence loci on a self-transmissible plasmid in Xanthomonas campestris pv. vesicatoria. Phytopathology 76:240-243. Stall, R. E , and Minsavage, G. V. 1990. The use of hrp genes to identify opportunistic xanthomonads. Proc 7th Int. Conf. Plant Pathog. Bacteria 1989. Z Klement, Akademiai Kiado, Budapest.

PAGE 82

75 Stall, R. E., and Thayer, P. L. 1962. Streptomycin resistance of the bacterial spot pathogen and control with streptomycin. Plant Dis. Rep 16:389-392. Szybalski, W. 1952. Microbial Selection. 1. Gradient plate technique for study of bacterial resistance. Science 1 16:46-48 Vauterin, L., Hoste, B., Kersters, K. and Swings, J. 1995. Reclassification of Xanthomonas Int. J. Syst. Bacterid 45:472-489. van der Plank, J. E. 1963. Control of Disease by Fungicides. Plant Diseases: Epidemics and Control. Academic Press, NY. Willis, K., Rich, J. J, and Hrbak, E. M. 1991. hrp genes of phytopathogenic bacteira. Mol. Plant Microbe Interact. 4:132-138.

PAGE 83

BIOGRAPHICAL SKETCH Pamela D. Roberts was born September 1 8, 1 963 in Manhattan, KS. She graduated with a Bachelor of Science degree in horticulture from Kansas State University in 1987. She completed a Master of Science from the University of Hawaii in 1991 in the Department of Plant Pathology. She spent 16 months as a Research Scholar at the International Rice Research Institute in conjunction with the research at the Univerisity of Hawaii. She arrived at the University of Florida in 1 99 1 to undertake a Doctor of Philosophy degree in the Department of Plant Pathology. 76

PAGE 84

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. Richard D. Berger, 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. Robert E. Stall 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. ?Jone* 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. Craig K. 0randler Associate Professor of Horticulture

PAGE 85

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 May, 1996 Deart/College of Agriculture' Dean, Graduate School

PAGE 86

LD 1780 1996 .RMb UNIVERSITY OF FLORIDA 3 1262 08554 9185


57
Euclidian Distance
0.0 3.55 7.05 10.5 14.1
Cluster "10%" | | | |
1
2
3
4a
4b
3
2
9
7
4
5
6
7
1
5
8
Figure 4-4. Dendrogram of cluster analysis of fatty acid methyl ester profiles showing the
four clusters (1 -4b) of strains of Xanthomonas fragariae. The designation of strains into the
groups determined by the MIDI 10% rule are shown for comparison. The Euclidian
distance of six is the cutoff point for group determination by cluster analysis.


69
Infection of plants may have occurred at any point of the route, however symptomatic
transplants from Canada are shipped to Florida A single type of genetic variant was not found
to be associated with plant material from a particular region of the U. S. or Canada.
Likewise, international movement of infested plants will make it difficult to determine if
endemic populations of the organism exists outside the U. S.. The origin of this disease
appears to be the United States (Maas, 1995).


32
Strawberry plants with leaves exhibiting typical lesions resulting either from natural
field infections or spray inoculation were assayed by removing a 9-mm diameter disk of tissue
surrounding a lesion and proceeding as described above. Tomato leaf tissue was used as a
negative control. For confirmation, a sample of the ground tissue suspension was plated on
Wilbrinks medium amended with 0.05% Bravo 720 (chlorothalonil; ISK Biosciences,
Mentor, OH). Plates were incubated at 24 C and colonies of A", fragariae were identified
by colony morphology.
Field experiments to examine oversummering survival of bacteria on nursery plants.
Six-week-old rooted transplants ofSweet Charlie were spray inoculated with X fragariae
strain XF1425lif two weeks prior to planting. All plants transplanted into the field exhibited
lesions of angular leaf spot. Plants were transplanted into the field on 27 June 1995 at Gulf
Coast Research and Education Center-Dover, FL and on 29 June 1995 at Gulf Coast
Research and Education Center-Bradenton, FL. Plants were sampled over a 14-week period
through 27 September 1995. Another set of inoculated plants were placed into Percival
growth chambers at 12 h photoperiod and 24 C and maintained for the duration of the
experiment. Plants which were not inoculated and did not exhibit symptoms of angular leaf
spot were placed in a greenhouse. At two-week intervals, five plants from the field, one plant
from the growth chamber, and one plant from the greenhouse were sampled. Two samples
of three daughter plants of the inoculated plants were sampled at 132 days after planting from
both locations. Individual plants were assayed as follows. All the leaves from a single plant
were removed and placed in a flask with 200 ml of phosphate buffer containing 0.02 %
Tween 20. The samples were shaken either on a wrist action (Burrell Corp., Pittsburgh, PA)


53
Figure 4-1. Agarose gel showing the restriction fragment length polymorphisms of genomic
DNA of strains of Xanthomonas fragariae after restriction with a rare-cutting endonuclease,
Vspl, and separated by pulsed-field gel electrophoresis. Lambda marker in 48.5 kb
concatamers is shown in side lanes.


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 EGIMTS4BR_4N07CE INGEST_TIME 2015-04-06T19:01:31Z PACKAGE AA00029744_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


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.
f)
Richard D. Berger, 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.
Robert E. Stall
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
1 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.
Associate Professor of Horticulture


TABLE OF CONTENTS
page
ACKNOWLEDGMENTS iii
ABSTRACT vi
CHAPTERS
1 INTRODUCTION 1
2 DISEASE PROGRESS, YIELD LOSS, AND CONTROL OF
XANTHOMONAS FRAGARIAE ON STRAWBERRY
PLANTS 8
Introduction 8
Materials and Methods 10
Results 13
Discussion 22
3DEVELOPMENT OF A SPECIFIC AND SENSITIVE DETECTION
TECHNIQUE BY THE POLYMERASE CHAIN REACTION
FOR XANTHOMONAS FRAGARIAE AND APPLICATION
IN A STUDY OF SURVIVAL OF THE BACTERIUM ON
STRAWBERRY PLANTS 25
Introduction 25
Materials and Methods 28
Results 33
Discussion 41
IV


LIST OF REFERENCES
Alippi, A. M., Ronco, B. L., and Carranza, M. R. 1989. Angular leaf spot of strawberry,
a new disease in Argentina: Comparative control with antibiotics and fungicides. Adv. Hort.
Sci. 3:3-6.
Ausubel, F. M., Brent, R. Kingston, R. E., Moore, D. D., and Seidman, K. 1987. Current
Protocols in Molecular Biology. John Wiley & Sons, New York.
Bouzar, H., Jones, J B., Stall, R. E., Hodge, N.C., Minsavage, G. V., Benedict, A. A., and
Alvarez, A. M. 1994. Physiological, chemical, serological, and pathogenic analyses of a
worldwide collection of Xanlhomonas campestris pv. vesicatoria strains. Phytopathology
84:663-671.
Conway, K. E., Motes, J. E., Bostian, B., Fisher, C. G., and Claypool, P. L. 1987.
Cercospora blight development on asparagus fern and effects of fungicides on disease severity
and yield. Plant Dis. 71:254-259.
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.
De Bruijn, F. J. 1992. Use of repetitive (repetitive extragenic palindromic and
enterobacterial repetitive intergeneric consensus) sequences and the polymerase chain reaction
to fingerprint the genomes of Rhizobium meliloti isolates and other soil bacteria. Appl.
Environ. Microbiol. 58:2180-2187.
Dye, D. W., and Wilkie, J. P 1973. Angular leafspot of strawberry in New Zealand N. Z.
J. Agr. Res. 16:311-314.
Egel, D. S., Graham, J. H., and Stall, R E. 1991. Genomic relatedness of Xanthomonas
campestris strains causing diseases of citrus. App. Environ. Microbiol. 57:2724-2730.
Epstein, A. H. 1966. Angular leaf spot of strawberry. Plant Dis. Rep. 50:167.
Felsenstien, J. 1995. PHYLIP. Phylogeny inference package, ver.3.57. Department of
Genetics, University of Washinton, Seattle, WA.
70


5
primers specific to regions of the hrp gene cluster which confers hypersensitivity and
pathogenicity from Xanthomonas campestris pv. vesicatoria to detect pathogenic
xanthomonads. Saprophytic and non-pathogenic bacteria lack the hrp genes (Stall and
Minsavage, 1990; Lindgren et al., 1986; Willis et al., 1991) and DNA from these bacteria was
not amplified by the htp gene cluster primers. Differentiation of X. campestris pathovars was
made by restriction endonuclease analysis (REA) patterns generated by digestion of the PCR
products with frequent cutting enzymes (Leite et al., 1994 and 1995). The primers were used
to detect X c. vesicatoria in seed lots of naturally infected pepper and tomato (Leite et al.,
1995). The sensitivity of detection by PCR is reported at 103 to 102 cfu per ml (Minsavage
et al., 1994; Leite et al., 1994; Henson and French, 1993). In nested PCR, sensitivity of
detection is increased by using PCR products from an amplification as target DNA in a
second round of amplification by a second set of primers internal to the first (Schaad et al.,
1993). McManus and Jones (1995) reported an increase in sensitivity of 1,000-fold with
nested PCR over a single round of PCR amplification to detect Erwinia amylovora.
The epidemiology of X fragariae must be understood before effective control
strategies can be devised. The delineation of bacterial populations is imperative to such
studies. The relationship of X. fragariae to other members of the genus Xanthomonas has
been examined (Hildebrand et al., 1990; Vauterin et al., 1995; Hodge et al., 1992), but the
genetic variability of strains within the species has not been reported (Maas, 1995). The
information from studies to compare strains of X. fragariae at the genetic level could be
useful to identify the origins and spread of specific strains or genetic types. This information
could be applied in international tracking of pathogen populations for quarantine programs.


16
Noninoculated
Inoculated
NoS 0.1X IX NoS 0.1X IX
Figure 2-2. Area under the disease progress curve (AUDPC) of Xanthomonas fragariae on
strawberry in the 1994 (A) and 1995 (B) seasons and the effect of applications of cupric
hydroxide plus mancozeb at the 1 x rate at 7-to 14-day intervals or at 0.1 x (10%) rate of the
chemical combination applied at 2-to 4-day intervals or no spray (NoS). A, B = significant
difference (P = 0.05) within the first factor of inoculated vs. noninoculated plants; a, b =
significant difference within the second factor of the spray treatments.


44
to detect plants contaminated with X fragariae. This is a concern of nursery-plant producers
and regulatory agencies of plant shipment (Maas, 1995). Currently, plants from nurseries in
California are being screened by this technique to determine its usefulness in such an
application.
The field studies have implications regarding the disease cycle of angular leaf spot in
Florida. The decline and later increase in the number of positive samples through the summer
would indicate that populations of the bacteria declined throughout the summer and increased
when more favorable conditions in the form of cooler weather occurred. The bacterial
populations did not completely die. Therefore, to eliminate disease, the production of plants
in nurseries in Florida would have to begin with plants free of the bacterium.
Researchers have reported the systemic movement of X. fragariae in plants
(Hildebrand et al., 1967; Milholland et al., 1993). In our research, inoculation of strawberry
plants resulted in survival of bacteria on leaves and the crown for extended periods under
conditions not optimum for growth of the bacteria. In addition, bacteria were detected by
PCR technique on daughter plants in the field. Dissemination to the daughter plants could
have been due to either systemic movement through the vascular system of the runner or
dispersal of bacterium by mechanical means.


46
The epidemiology of the X. fragariae must be understood before effective control
strategies can be devised. The delineation of bacterial populations is imperative to such
studies. The relationship of X. fragariae to other members of the genus Xanthomonas has
been examined (Hildebrand et al., 1990; Hodge et al., 1992; Vauterin et al., 1994), but the
genetic variability within the species has not been reported (Maas, 1995). The information
from studies to compare strains of X. fragariae at the genetic level could be useful to identify
the origins and spread of specific strains or genetic types. This information could be applied
in international tracking of pathogen populations for quarantine programs. In addition, the
identification of genetic types is a prerequisite to identify sources of resistance in strawberry.
Strawberry cultivars exhibit levels of susceptibility or tolerence to X. fragariae; only F.
moschata Duch. appeared to be immune (Hazel, 1981; Hazel and Civerolo, 1980; Kennedy
and King, 1962b). A screening program to identify genes for resistance must incoporate
representives of the genetic variants in the screening process otherwise the resistance may be
overcome quickly by genetic variants.
Studies to characterize populations of bacteria have used biochemical and molecular
biological techniques. Protein staining and fatty acid analysis have been useful to detect
differences at the metabolic level within species o Xanthomonas (Bouzar et al. 1994; Stall
et al, 1994; Graham et al., 1990; Hodge et al., 1992). The polymerase chain reaction (PCR)
has been used to analyze differences between pathovars and strains of bacteria (Hensen and
French, 1993). Primers specific to the hrp-gene cluster of Xanthomonas campestris pv.
vesicatoria amplifed genomic DNA and restriction enzyme analysis of the PCR product
differentiated between many pathovars and species of Xanthomonas (Leite et al, 1994 and


27
be confused with symptoms caused by fungal pathogens (Kennedy and King, 1962b).
Diagnosis based on symptoms is very difficult and not applicable for asymptomatic plants.
Identification of the disease based upon isolation and characterization of the causal agent may
also be difficult because X. fragariae grows slowly and may be masked by faster growing
organisms (Kennedy and King, 1962a). Expression of watersoaked lesions takes 6 days or
longer after inoculation. Thus, fulfillment of Kochs postulates to confirm pathogenicity is
difficult and time-consuming.
Assays have been developed with improved sensitivity and specificity for the detection
of plant pathogens in plant tissue. A specific, indirect, enzyme-linked immunosorbent assay
(ELISA) was developed to detect X. fragariae from symptomatic plant tissue (Rowhani et
al., 1994). The antibody assay did not react with bacterial strains of 16 pathovars of X
campestris or non-pathogenic bacteria from strawberry leaves. A single cross-reaction
occurred to a strain of X. campestris isolated from Nerium oleander. The assay detected
bacteria directly from a visible lesion on a strawberry leaf. The level of sensitivity was at ca.
104 colony forming units (cfu) per ml. However, bacteria on asymptomatic plants may not
be detected by this ELISA assay.
The polymerase chain reaction (PCR) has been used to amplify specific DNA
sequences to detect and identify many plant pathogens including some members of the genus
Xanthomonas (Henson and French, 1993; Louws et al., 1994). Leite et al. (1994) utilized
primers specific to regions of the hrp gene cluster which confers hypersensitivity and
pathogenicity from Xanthomonas campestris pv. vesicatoria to detect pathogenic
xanthomonads. Saprophytic and non-pathogenic bacteria lack the hrp genes (Stall and


31
used. In all PCR runs, including the nested assays, a water sample was used as a negative
control.
Restriction endonuclease analysis (REA) of PCR products. The PCR samples with
the overlay of mineral oil were cleaned by the method of Minsavage et al (1994). Samples
without the mineral oil were used directly in restriction reactions. An 8 /I sample of the PCR
product was digested with restriction endonuclease Sau3Al, Haelll or Cfol under conditions
specified by the manufacturer (Promega). Restricted products were separated in 4% agarose
gel (3% NuSieve, 1% SeaKem GTG, FMC BioProducts, Rockland, ME) containing 0.5
/g/ml ethidium bromide in TAE buffer at 8 V/cm as described by Leite et al. (1994). DNA
molecular weight marker XI (Boehringer Mannheim, Indianapolis, IN) was used for standard
weight markers. Gels were photographed over a UV transilluminator with type 55 Polaroid
film (Polaroid Corp., Cambridge, MA).
Detection of bacteria from infected plant tissue. The sensitivity of the assay in the
presence of plant tissue was determined by adding known concentrations of bacteria to plant
samples. A 9-mm diameter disk of plant tissue was macerated with mortar and pestle in 200
yul of phosphate buffer (pH 7.0) containing 5% polyvinylpolypyrrolidone (PVP40, Sigma
Chemical Co., St. Louis, MO) and 0.02 M sodium ascorbate (PPA). The mixture was
incubated at room temperature for a minimum of 1 h. The volume was adjusted to 582 //I
with TE and DNA extraction proceeded as described above. A minimum of three
experiments with three replications of each treatment for each primer set and nested reaction
was performed. REA was performed on the final PCR products.


42
to detect bacteria in symptomatic leaf tissue, we were concerned with the detection of
bacteria from asymptomatic tissue. The X. fragariae primers were suitable for use in the
nested technique and the sensitivity of the PCR reaction was increased by 1,000-fold This
level of detection was achieved in assays to detect the bacterium in the presence of plant
tissue and is well below the number of cells needed to cause visible lesions. Therefore, the
nested technique is applicable for detection of bacteria in association with asymptomatic
tissue.
Cross-contamination among samples and contamination of PCR reagents are problems
associated with the nested technique (McManus and Jones, 1995). We experienced false
positives in a number of initial experiments and finally determined the cause to be
contaminated mineral oil. Although we eliminated the need for mineral oil overlay by using
a hot bonnet, extreme care is needed to avoid contamination and subsequent false positives.
In our experiments, negative control samples were always included for each amplification
round of the nested assay. Aerosols may also contribute to false positives and care must be
taken to prepare the mixtures for PCR in a sterile environment and negative control samples
should be included to check for aerosol contamination. In the field experiments, a problem
encountered with the nested technique was amplification of bacterial DNA from the negative
control plants. Although these plants were physically isolated from a contamination source
during the experiments; they were initially obtained from fields at GCREC-Dover where
plants infected with angular leaf spot were located. The plants may have been infested with
the bacterium or cross-contamination of samples may have occurred during preparation of
plant samples from handling infected material or using non-sterile instruments. Aerosols may


43
also have been a source of contamination of PCR samples (Innis et al., 1990). However,
contamination from these sources would be detected by amplification in the negative controls.
Our negative controls were included in every PCR run and were consistently negative.
Therefore contamination of PCR reagents or aerosols would have been ruled out.
Confirmation that the PCR product is from amplification of the target DNA is possible
by REA (Leite et al. 1994). The REA was used to identify PCR products amplified by the
/vp-primers from cells of X. c. vesicatoria added to seed washing of tomato and pepper (Leite
et al. 1995). Similarly, we applied REA to identify PCR products amplified by the XF-
primers. The profiles of the restricted PCR products were distinct for each enzyme. The
profiles from REA should be different for unrelated organisms since it would be highly
unlikely that the same restriction sites would exist for two heterologous pieces of DNA (Leite
et al,. 1995).
The nested PCR technique was useful to detect bacteria on strawberry plants in
nurseries in Florida Symptoms and recovery of the rifampicin-marked strain on a selective
medium were not useful to identify bacteria when populations were extremely low. However,
recovery of the rifampicin-marked strain from some samples indicated that the bacteria were
viable throughout the summer and that PCR was detecting viable bacterial cells. Visible
symptoms on plants disappeared soon after placement in the field and recovery on media was
difficult due to the slow-growing nature of X. fragariae and overgrowth by contaminants.
While bacterial populations were not enumerated, the levels were deduced to be below 103
cfu per ml, since bacteria were detected in the nested PCR but not by single round of
amplification. This level of sensitivity would be useful to screen asymptomatic nursery plants


ACKNOWLEDGMENTS
The author would like to gratefully acknowledge the members of the committee:
Dr. R. D. Berger, Dr. R. E. Stall, Dr. C. K. Chandler, and Dr. Jeffrey B Jones. The
enthusiatic support, expert advise, and guidance throughout the duration of this project were
greatly appreciated. Special thanks go to Dr. Jeffrey B. Jones who provided his laboratory
for a majority of the work. Thanks go to Dr. A. R. Chase who initially suggested the project
and provided considerable support the first year.
Special thanks go to the members of Dr. Stalls and Dr. Jones labs, including N. Cheri
Hodges, G. V. Minsavage, Rui P Leite, Jr., Hacene Bouzar, Gail Somodi, Jeanette Chun,
and Rick Kelly, for their help and information.
Very special thanks go to the people at Dover-G.C.R.E.C. who helped me
tremendously with all aspects of the field research and taught me all about the production of
strawberries: Alicia Whidden, Anne Turgeau, Jim Sumler, Dale Wenzel, Larry Smith, and
Mitch Boles.
Finally, thanks go to the folks who helped make the experience enjoyable: Gary
Marlow, Pia D. Gavino, Morgan Wallace, Pamela Lopez, and Marion Bogart.
in


68
strains lacked the 16:1 w 7 cis fatty acid which distinguished them from all the other X.
fragariae. It is interesting to speculate that perhaps strain XF119 represents a bridge
between the A and D RFLP groups because of its intergroup relationship in that its RFLP
profile is D but by FAME it is closer to the A group. It would be of further interest to
examine future populations to determine the fate of the two groups represented by only four
strains.
The majority of strains was represented by the B and C groups. An extensive survey
of the pathogen population in the United States and Canada would be useful to determine
whether the population ratios of the various genetic variants are remaining relatively the same.
Likewise, it would be interesting to determine the distribution of the three subdivisions within
the B group. The cultivars of strawberry in commercial production are changed frequently
(C. Chandler, pers. comm ). This constant change in genotypes should influence the genetic
composition of the population of X. fragariae if genes for resistance to angular leaf spot are
lost or found during the development of new cultivars. In the Philippines, the relative
populations of two races of bacterial blight on rice were followed for 10 years. Researchers
recorded a decline in the prevalence of the predominant race and a concurrent increase in
another race. This change in frequency of race, or bacterial genotype, occurred after the
introduction of a gene for resistance in cultivated rice (Mew et al., 1992). Information
regarding changes in the dominant genotype would be useful to plant breeders.
The identification of a relationship between geographical origin and genetic variants
is unclear because of the transportation of infected plants. Plants are shipped from California
to Canada where they are propagated and sent to Florida for field production each season.


35
12 3 4
Figure 3-1. Agarose gel of the PCR products generated by amplification of genomic DNA
from a strain ofXcmthonionas fragariae with the /wp-primers RST2 and RST3 (Lane 1) and
the XF-specific primers XF9 paired with XF1 l(Lane 2) and XF9 paired with XF12 (Lane 3),
and molecular weight marker XI (Boehringher-Manheim, Indianapolis, IN; Lane 4).


73
Lindgren, P. B., Peet, R. C., and Panopoulos, N. J. 1986. Gene cluster of Pseudomonas
syringae pv. "phaseolicola" control pathogenicity of bean plants and hypersenstivity on
nonhost plants. J. Bacteriol. 168:512-522.
Louws, F. J., Fulbright, D. W., Stephens, C. T., and Bruijn, F. J. 1994. Specific genomic
fingerprints of phytopathogenic Xanthomonas and Pseudomonas pathovars and strains
generated with repetitive sequences and PCR. Appl. Environ. Microbiol. 60:2286-2295.
Maas, J. L. (ed.) 1984. Angular leaf spot. Compendium of Strawberry Diseases. American
Phytopathological Society, St. Paul, MN.
Maas, J. L., Pooler, M. R, and Galletta, G. J. 1995. Bacterial angular leafspot disease of
strawberry: Present status and prospects for control. Adv. Strawberry Res. 14:18-24.
Marco, G.M., and Stall, RE. 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.
McManus, P. S., and Jones, A. L. 1995. Detection of Erwinia amylovora by nested PCR
and PCR-dot-blot and reverse blot hybridizations. Phytopathology 85:618-623.
Milholland, R. D., Ritchie, D F., Daykin, M. E., and Gutierrez, W. A. 1993. Systemic
movement of Xanthomonas fragariae in inoculated strawberry plants. (Abst.)
Phytopathology 83:1408.
Minsavage, G. V., Thompson, C. M., Hopkins, D. L., Leite, R M. V. B., and Stall, R E.
1994. Development of a polymerase chain reaction protocol for detection of Xylella
fastidiosa in plant tissue. Phytopathology 84:456-461.
Nei, M. 1987. Molecular Evolutionary Genetics. Columbia University Press, New York.
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.
Panagopoulos, C. G., Psallidas, P. G., and Alivizatos, A. S. 1978. A bacterial leafspot of
strawberry in Greece caused by Xanthomonas fragariae. Phytopath. Z. 91:33-38.
Pataky, J. K., and Lim, S. M. 1981. Efficacy of benomyl for controlling Septoria brown spot
of soybeans. Phytopatholology 71 438-442.
Pickup, R. W. 1991. Development of molecular methods for the detection of specific bacteria
in the environment J. Gen. Microbiol. 137:1009-1019.


49
Table 4.1 continued. The geographic source, collection date, and groups determined by
fatty acid methyl ester (FAME) and restriction fragment length polymorphism (RFLP ) for
strains used in this study.
Geographic
Strain
Source3
Year
Sourceb
FAMEC
RFLP
1532
FL
1993
ARC
3
1533
WI
1993
ARC
5
B
1534
WI
1993
ARC
B3
100
FL
1993
This study
1
B
101
FL
1993
This study
1
103
FL
1993
This study
B
104
FL
1993
This study
1
B
105
FL
1993
This study
3
B
106
FL
1993
This study
3
107
FL
1993
This study
1
108
FL
1993
This study
1
B
113
FL
1993
This study
5
B3
114
FL
1993
This study
3
B
115
FL
1993
This study
4
B
116
CAN
1993
This study
B
117
CAN
1993
This study
B3
119
CAN
1993
This study
9
D
124
CAN
1993
This study
B
125
CAN
1993
This study
6
126
CAN
1993
This study
4
127
CAN
1993
This study
6
128
CAN
1993
This study
4
B
129
CAN
1993
This study
6
B3
138
CAN
1993
This study
6
B
146
CAN
1993
This study
4
B1
153
CAN
1993
This study
6
B
33239
MN
ATCC
9
A
a Geographic origin of plant material from which bacteria were isolated. CAN=Canada.
bARC= A. R. Chase; ATCC= American Type Culture Collection.
cBlank space indicates that group was not determined.


Abstract of 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
THE EPIDEMIOLOGY, SPECIFIC DETECTION, AND GENETIC VARIABILITY OF
XANTHOMONAS FRAGAR1AE
By
Pamela D. Roberts
May 1996
Chairman: Dr. R. D. Berger
Major Department: Plant Pathology
Angular leaf spot (ALS) on strawberry is caused by the bacterium Xanthomonas
fragariae (XF). The epidemiology of ALS, development of a detection technique by the
polymerase chain reaction (PCR), and the genetic variability of strains were examined. The
disease severity of ALS on plants in field plots increased to 25% and 15% in two seasons.
Yield was decreased 8% to 10%. Minimal spread of ALS occurred between field plots.
Chemical sprays applied at the label rate of cupric hydroxide plus mancozeb at 7- to 14- day
intervals decreased disease but was phytotoxic and decreased yield. A 10% rate of the
mixture applied frequently slightly reduced disease but increased yield one season and
significantly reduced disease and did not affect yield the next season. Three primers were
specific for amplification of DNA from XF but not DNA from strains of 16 pathovars of
Xanthomonas campestris or non-pathogenic xanthomonads from strawberry. Bacteria were
detected at 104 colony forming units per ml by a single round of PCR A nested PCR
vi


21
Noninoculated
Inoculated
NoS 0.1X IX
NoS 0.1X IX
Figure 2-6. The yield lost to fruit damaged by fungal diseases on strawberry in Florida in the
1994 (A) and 1995 (B) seasons. A, B = F was significant at P = 0.05 within the first factor
of noninoculated vs.inoculated; a, b, c = F'was significant at P = 0.05 within the second factor
of spray treatments.


Table 3-1. Percentage of strawberry plants positive for the presence of Xanthomonas fragariae detected by PCR and a selective
medium in summer nurseries from June through September at two locations in Florida
Leaves
Crown
Location
Dover
Bradenton
DAPa NonIb Inoc" Primer ld Primed" Medium1 Symptoms NonI Inoc Primerl Primer2 Medium
0
-
+
60
100
TC
+
-
+
20
40
TC
11
-
+
0
40
TC
+
-
+
0
40
TC
31
-
+
0
20
0
+
-
+
0
40
0
40
+
+
0
20
NS
-
-
+
0
0
NS
66
-
-
0
60
0
-
-
+
0
40
0
85
-
+
20
100
60
-
-
+
20
100
20
92
-
+
40
100
40
-
-
+
0
80
40
0
-
+
60
100
TC
+
_
+
0
40
TC
20
-
+
20
20
20
+
+
+
0
60
0
34
+
f
0
20
NS
-
-
+
0
40
NS
40
-
+
0
20
20
-
-
+
0
100
20
54
-
+
0
40
0
-
-
+
0
40
0
68
+
+
20
100
60
-
+
+
20
80
20
82
-
+
20
80
40
-
4-
+
20
80
40
a Days after planting.
bNon-inoculated plants in greenhouse. = Negative by PCR amplification, + = Positive by PCR amplification.
c Inoculated strawberry plants in growth chamber at 22 C. = Negative by PCR amplification, + = Positive by PCR amplification.
d Percentage of samples amplified by primer set XF9 and XF11.
"Percentage of samples amplified in nested PCR with primers XF9 and XF12 with sample from first round.
Wilbrink's agar plus 100 mg/ml rifampicin and 0.05% Bravo; TC= too contaminated to distinguish individual colonies;
NS= Not assayed, 0 100 = Percentage of plates with colonies of A", fragariae.


50
volume of 1% Seakem Gold agarose solution (10 mM Tris [pH 8.0], 10 mM MgC12, 0.1
mM EDTA [pH 8.0], 1 % Seakem Gold agarose (FMC BioProducts, Rockland, ME) [wt/vol]
in sterile filtered water) was added. Plugs were made, lysed, washed, and stored as described
by Egel et al. (1991). Two sizes of plugs were utilized. A 4 8 mm slice of plug was
restricted in a volume of 200 p1 of restriction buffer (as recommended by manufacturer,
Promega, Madison, WI) and inserted in wells made with a 10-well comb (Bio-Rad,
Richmond, CA). A44 mm square piece of plug was restricted in a volume of 100 p\ and
placed in wells made by a 20-well comb (Bio-Rad). Restriction enzyme was added at the
following concentrations: Xba\ at 40 U; Spel at 30 U; Vsp\ at 30 U (Promega). The pieces
were placed into wells in a 1.2% GTG agarose gel made with 0.5X TBE. Wells were sealed
with the 1% Seakem Gold solution. The gel was placed in a Bio-Rad CHEF DR II unit
containing 1.8 liters of 0.5* TBE and run at 200 V (15V/cm of gel). Pulsed times for plugs
digested with Xbal or Spel were at 4 s for 1 h followed by 8 sec for 18 h. Plugs digested with
Vspl were pulsed at 4 s for 1 h and subsequently at 12 s for 17 h. Lambda DNA in 48.5 KB
concatamers (FMC BioProducts) was used in the first and last lanes of each gel. Gels were
stained in 0.5 mg of ethidium bromide per liter and photographed with type 55 Polaroid film.
The position of bands were assessed visually or by analysis with the Gelmeas
computer program. Similarity values were calculated as described by Egel et al. (1991) with
the mathematical equation proposed by Nei and Li (1979) based upon the proportion of
shared DNA fragments. The estimate of the number of nucleotide substitutions per site was
used to calculate the genetic divergence by the iterative method of Nei (1987) with the
program in SAS described by Leite et al. (1994). The KITSCH program from the PHYLIP


58
The two methods used in this research might be useful to examine the evolution of the
pathogen in future populations. The identification of two groups which contained only four
strains total, including the ATCC type strain, from a population of 50 strains was of interest.
By RFLP analysis, group A contained only the ATCC type strain from Minnesota which was
collected over twenty years ago. However, by FAME cluster analysis, one other strain,
XF119, which was isolated from an infected plant from Canada in 1993, grouped with the
ATCC type strain. In RFLP-PFGE analysis, the XF119 profile was the same as the D group
which contained strains XF1238 and XF1240, isolated from samples from California in 1990.
Dendrogram analysis places these two groups close to each other. By FAME, these four
strains lacked the 16:1 w 7 cis fatty acid which distinguished them from all the other X.
fragariae. It is interesting to speculate that perhaps strain XF119 represent a bridge
between the A and D RFLP groups because of its intergroup relationship in that its RFLP
profile is D but by FAME it is closer to the A group. It would be of further interest to
examine future populations to determine the fate of the two groups represented by only four
strains.
The majority of strains was represented by the B and C groups. An extensive survey
of the pathogen population in the United States and Canada would be useful to determine
whether the population ratios of the various genetic variants are remaining relatively the same.
Likewise, it would be interesting to determine the distribution of the three subdivisions within
the B group. The cultivars of strawberry in commercial production are changed frequently
(C. Chandler, pers. comm ). This constant change in genotypes should influence the genetic
composition of the population of X fragariae if genes for resistance to angular leaf spot are


54
*- B2
C
A
D
X.c. vesicatoria
i i i
0.04 0.02 0
Genetic Distance
Figure 4-2. Relationship of groups of Xanthomonas fragariae analyzed by restriction
fragment length polymorphisms.


10
application of a chemical pesticide can reduce the infection by a pathogen and slow the rate
of the epidemic, e. g., van der Planks (1963) apparent infection rate (Fry et al., 1979; Pataky
and Lim, 1981). A reduced rate of fungicide applied at frequent intervals as a protectant has
been used effectively to reduce disease on crops (Fry 1975; Fry et al, 1979; Conway et al.,
1987).
Our objectives in this study were to examine epidemiological aspects of angular leaf
spot including the incidence of A*, fragariae on transplants from northern nurseries, the effect
of angular leaf spot on yield, and the control by chemicals.
Materials and Methods
Survey of farmers fields and storage facility. Disease incidence of angular leaf spot
on plants in fields located around Dover and Plant City, FL, was examined in October 1993.
Plants were assessed for symptoms of angular leaf spot within 2 to 4 days of transplanting.
Leaves from symptomatic plants were collected. A 9-mm diameter of leaf tissue was
removed surrounding a lesion and macerated in 200 /il of sterile water. A loopful of the
suspension was dilution streaked onto Wilbrinks medium (WB) (Koike, 1962) and colonies
of X. fragariae were identified. The pathogenicity of these isolated strains was tested on
Sweet Charlie strawberry plants. In 1994, transplants were sampled while in cold storage
at a facility in Dover, FL. Seven groups of plants, comprised of four cultivars from six
northern growers were sampled. Three boxes were selected at random from within a group
and 20 plants were removed from each box. The plants were examined for symptoms of


64
ELISA test for X. fragariae developed by Rowhani et al. (1994). In addition, their antibodies
cross-reacted with another xanthomonad whereas we were able to eliminate the non-specific
reaction by discontinuing use of primer XF10. Although this level of detection is adequate
to detect bacteria in symptomatic leaf tissue, we were concerned with the detection of
bacteria from asymptomatic tissue. The X fragariae primers were suitable for use in the
nested technique and the sensitivity of the PCR reaction was increased by 1,000-fold. This
level of detection was achieved in assays to detect the bacterium in the presence of plant
tissue and is well below the number of cells needed to cause visible lesions. Therefore, the
nested technique is applicable for detection of bacteria in association with asymptomatic
tissue.
Cross-contamination among samples and contamination of PCR reagents are problems
associated with the nested technique (McManus and Jones, 1995). We experienced false
positives in a number of initial experiments and finally determined the cause to be
contaminated mineral oil Although we eliminated the need for mineral oil overlay by using
a hot bonnet, extreme care is needed to avoid contamination and subsequent false positives.
In our experiments, negative control samples were always included for each amplification
round of the nested assay. Aerosols may also contribute to false positives and care must be
taken to prepare the mixtures for PCR in a sterile environment and negative control samples
should be included to check for aerosol contamination. In the field experiments, a problem
encountered with the nested technique was amplification of bacterial DNA from the negative
control plants. Although these plants were physically isolated from a contamination source
during the experiments; they were initially obtained from fields at GCREC-Dover where


CHAPTER 3
DEVELOPMENT OF A SPECIFIC AND SENSITIVE DETECTION TECHNIQUE
BY THE POLYMERASE CHAIN REACTION FOR XANTHOMONAS FRAGAR1AE
AND APPLICATION IN A STUDY OF SURVIVAL OF THE BACTERIUM ON
STRAWBERRY PLANTS
Introduction
Angular leaf spot of strawberry (,Fragaria x ananassa Duchesne), caused by the
bacterium Xanthomonas fragariae, was first reported in Minnesota in 1960 and is now found
in many areas of strawberry production throughout the world (Kennedy and King, 1960a;
Maas, 1984; Ritchie et al., 1993). The disease is apparently disseminated by the
transportation of infected plants (Maas, 1984). In Florida, strawberry plants which arrive
from northern nurseries for transplanting in the fall frequently have leaves which exhibit
symptoms of angular leaf spot. Typical pathogenic strains of X. fragariae can be isolated
from the lesions (P. D. Roberts, unpublished).
The epidemiology of X fragariae is mostly unknown in fields of strawberry in Florida
where production is from annual crops. Angular leaf spot was first reported in the state in
1971 (Howard, 1971). Howard (1971) was unable to determine the inoculum source for
infected plants from nurseries in Florida or other states. Bacteria may survive on infested
leaves in the soil (Kennedy and King, 1962b), but plants in Florida are usually treated with
the herbicide paraquat at the end of the season and removed. The bacterium does not survive
25


59
lost or found during the development of new cultivars. In the Philippines, the relative
populations of two races of bacterial blight on rice were followed for 10 years. Researchers
recorded a decline in the prevalence of the predominant race and a concurrent increase in
another race. This change frequency of race, or bacterial genotype, occurred after the
introduction of a gene for resistance in cultivated rice (Mew et al., 1992). Information
regarding changes in the dominant genotype of both the plant and bacteriapopulations would
be useful to plant breeders.
The relationship between geographical origin and genetic variants is unclear because
of the transportation of infected plants. Plants are shipped from California to Canada where
they are propagated and sent to Florida for field production each season. Infection of plants
may have occurred at any point of the route, however symptomatic transplants from Canada
are shipped to Florida. A single type of genetic variant was not found to be associated with
plant material from a particular region of the U. S. or Canada. Likewise, international
movement of infested plants will make it difficult to determine if endemic populations of the
organism exists outside the U. S .. The origin of this disease appears to be the United States
(Maas, 1995).


62
by strawberry plants which diluted the amount of disease relative to the total leaf area. The
dilution effect of healthy plant tissue was described by van der Plank (1963). In 1995, a
negative infection rate was seen very early in the epidemic. This was because the symptoms
of the disease almost disappeared from the field as the affected leaves died and were lost. The
difference in disease progress between the two seasons did not appear to be from differences
in rainfall since mean rainfall was approximately the same both years (data not shown). Mean
temperatures also were similar; however the number of days at temperatures below 10 C in
November and December 1993 were much greater than in 1994. Another difference between
the seasons was that sprinkler irrigation was reduced in 1995 because of the change to drip
irrigation. Overhead irrigation, fog, high humidity, and rain are positive factors in disease
development and spread (Epstein, 1966; Hildebrand et al., 1967; Kennedy and King, 1962b;
Lai, 1978). The trend towards drip irrigation in commercial production fields should have a
positive effect to reduce the spread and survival of the bacterium.
Strawberry producers currently apply copper-based compounds to control angular leaf
spot despite the phytotoxicity to plants caused by copper. Application of cupric hydroxide
plus mancozeb at the lx rate was phytotoxic to strawberry plants (Howard, 1973; this study).
In our tests, the greatly reduced 0.1 rate did not harm plants and significantly reduced
disease severity in 1995. The protectant application of bactericide at this rate and spray
schedule was intended to reduce the total amount of inoculum and to prevent spread of the
disease. The approach of frequent sprays at reduced concentrations may have potential to
control this disease. More studies are needed to evaluate different rates, application intervals,
and chemical mixtures to achieve maximum disease control while avoiding yield losses. An


67
within the population. Methods of RFLP including PFGE have distinguished between closely
related strains of pathovars within the genus Xanthomonas campestris (Egel et al., 1991;
Gabriel et al., 1988; Hartung and Civerolo; 1989, Graham et al., 1990). In these works, the
importance of the diverse nature of the endonuclease sites as related to strain characteristics
such as host range, geographic origin, and pathogenicity was discussed. Unfortunately, the
relationship of pathogenicity of the X fragariae strains to the genetic groupings cannot be
examined. Differences in pathogenicity among strains of X. fragariae were not detected by
inoculation of the 52 strains onto two cultivars of strawberry. Studies of bacterial populations
on two cultivars which appeared to have different levels of susceptibility in preliminary tests
were also inconclusive (Roberts, unpublished). Nor have other researchers reported
differences in pathogenicity among strains ofAi fragariae. Such information might have been
useful to determine significance of the RFLP-PFGE and FAME groups as related to
pathogenicity.
The two methods used in this research might be useful to examine the evolution of the
pathogen in future populations. The identification of two groups which contained only four
strains total, including the ATCC type strain, from a population of 50 strains was of interest.
By RFLP analysis, group A contained only the ATCC type strain from Minnesota which was
collected over twenty years ago. However, by FAME cluster analysis, one other strain,
XF119, which was isolated from an infected plant from Canada in 1993, grouped with the
ATCC type strain. In RFLP-PFGE analysis, the XF119 profile was the same as the D group
which contained strains XF1238 and XF1240, isolated from samples from California in 1990.
Dendrogram analysis places these two groups close to each other. By FAME, these four


39
Detection of bacteria on plants in oversummering experiments. All plants exhibited
visible symptoms of angular leaf spot at planting. However, by 34 days after planting,
symptoms of angular leaf spot were not visible on plants (Table 3-1). Bacteria were
recovered on the selective medium and colonies of X. fragariae were identified by
morphology, resistance to rifampicin, DNA amplification by the XF-specific primers, and
REA. Bacteria were recovered on selective medium from a few samples throughout the
experiment, including the final date, indicating that bacteria were viable on plants throughout
the summer.
Xanthomonas fragariae was detected from leaf samples by nested PCR on each
sample date. The percentage of leaf samples which were positive by the nested PCR assay
ranged from 20% to 100% for the sample dates. From the leaves, all samples were initially
positive but, by 51 days after planting, the number of positive samples declined to 20%. By
92 days after planting, all samples were again positive. A single round of amplification was
occasionally sufficient to identify positive samples. In the crown of plants, bacteria were
detected by PCR amplification on all sample dates except one. REA analysis of the PCR
products confirmed banding patterns typical of X. fragariae. Plants inoculated and placed in
the growth chamber at the favorable temperatures for disease development were positive at
each sampling date except one. Non-inoculated plants from the greenhouse were positive by
this technique in approximately 20% of the samples.


65
plants infected with angular leaf spot were located. The plants may have been infested with
the bacterium or cross-contamination of samples may have occurred during preparation of
plant samples from handling infected material or using non-sterile instruments. Aerosols may
also have been a source of contamination of PCR samples (Innis et al., 1990). However,
contamination from these sources would be detected by amplification in the negative controls.
Our negative controls were included in every PCR run and were consistently negative.
Therefore contamination of PCR reagents or aerosols would have been ruled out.
Confirmation that the PCR product is from amplification of the target DNA is possible
by REA (Leite et al. 1994). The REA was used to identify PCR products amplified by the
/wp-primers from cells of X. c. vesicatoria added to seed washing of tomato and pepper (Leite
et al. 1995). Similarly, we applied REA to identify PCR products amplified by the XF-
primers. The profiles of the restricted PCR products were distinct for each enzyme. The
profiles from REA should be different for unrelated organisms since it would be highly
unlikely that the same restriction sites would exist for two heterologous pieces of DNA (Leite
et al,. 1995).
The nested PCR technique was useful to detect bacteria on strawberry plants in
nurseries in Florida Symptoms and recovery of the rifampicin-marked strain on a selective
medium were not useful to identify bacteria when populations were extremely low. However,
recovery of the rifampicin-marked strain from some samples indicated that the bacteria were
viable throughout the summer and that PCR was detecting viable bacterial cells. Visible
symptoms on plants disappeared soon after placement in the field and recovery on media was
difficult due to the slow-growing nature of X. fragariae and overgrowth by contaminants.


13
(AUDPC)(Shaner and Finney, 1977) was calculated for each plot and used in statistical
analysis. Statistical analysis was performed by orthogonal contrasts using PC-SAS.
Yield data. Fruit were harvested at 2- to 4-day intervals from initial bearing in
December through 30 March. Fruit were graded as marketable, culls, or nonmarketable due
to damage by fungi. Marketable fruit were those free of rot, not misshapen, and greater than
10 g in weight. Culls were non-marketable fruit due to physical imperfections such as small
size (< 10 g), damage by insects, or undesirable shape. Nonmarketable fruit was damaged
by fungal diseases, usually caused by anthracnose, botrytis, and phomopsis. The weight in
grams for each category was recorded for individual plots. Statistical analysis of the weight
of fruit in each category was performed using orthogonal contrasts using PC-SAS.
Results
Disease incidence of transplants in field and storage facility. In 1993, six of seven
fields contained plants with symptoms of angular leaf spot from which X. frcigariae was
isolated. In 1994, three of the seven groups sampled from the cold storage facility contained
plants with symptoms of angular leaf spot. In one of these groups, plants with symptoms of
angular leaf spot were found in all of the boxes sampled. In the other two samples, one box
contained plants which were symptomatic with angular leaf spot. Bacteria were isolated from
all plants with lesions of angular leaf spot and pathogenicity of these isolated strains on
strawberry was confirmed. Strains which were resistant to copper were not identified


22
Discussion
A decrease in yield of strawberry fruit was observed both seasons due to angular leaf
spot. This report is the first to quantify the reduction in yield due to this disease. The 8 to
10% loss determined in these studies is much lower than the 70 to 80% loss estimated by
Epstein (1966) in fruit production fields in Wisconsin. Production in northern regions of the
U. S. is usually perennial and the age of the plants was not given. The loss was reported on
the cultivar Sparkle which is not produced commercially in Florida. Therefore, we are
unable to compare our yield loss to his report. Howard (1971) also reported unquantified
yield losses due to this disease but on cultivars which are no longer in commercial production.
In our studies, the average yield loss observed both years was very similar despite the
differences in the disease levels and the base yields between the two seasons. In 1995, the
disease severity was 60% of the amount estimated in the previous season; however, the yield
loss was actually higher by 2% for marketable yield. The 10% reduction in yield represents
an estimated loss of 1233 kg of berries per ha. Strawberry fruit has a very high cash value
and a decrease in revenue of 8 to 10% represents a significant economic loss to producers.
Strawberry producers in Florida currently try to control angular leaf spot by
application of copper compounds and by avoidance of the disease. An understanding of the
epidemiology of angular leaf spot is necessary to determine the effectiveness and practicability
of any control method. The survey of farmers fields and the cold storage facility established
that transplants arrive from northern nurseries infected with angular leaf spot at fairly high
disease incidence. Therefore, disease is introduced into the field on infected transplants. In


51
computer pakage (Felsenstein, 1995) was used to create a rooted phylogenetic tree by the
Fitch-Margoliash method (Fitch and Margoliash, 1967). The input data was as described for
the combined SpeI, Xbal, and Vspl digestion data (Stall et al., 1994).
Fatty acid composition. Strains of X. fragariae were inoculated onto Trypticase Soy
Broth agar (TSBA) and grown for 48 h at 24 C. Conditions were changed from the
standard MIDI procedures which required that cells be grown on TSBA for 24 hours at 28
C because strains of X fragariae produced insufficient growth at these conditions. Cellular
fatty acids were extracted and derivatized to their fatty acid methyl esters as described
(Sasser, 1990b). A library comprised of strains ofX. fragariae was created using the MIDI
Library Generation System (LGS), software version 3.3. FAMEs were analyzed by the MIDI
Microbial Identification System, software version TSBA 3.5. The qualitative and quantitative
differences in the fatty acid profiles were used to compute the Euclidian distance to each
strain. Strains within six Euclidian distance units, the cut-off for subspecies (Sasser, 1990a),
were grouped in the same cluster.
Results
RFLP-PFGE. Restriction endonucleases Xbal and Spel generated genomic DNA
fragments from 5 to 400 kb (Fig. 4-1). Typically, a strain profile contained ten DNA
fragments greater than 100 kb. Analysis of the 52 strains resulted in four RFLP groups,
designated A through D, which were identical by these two enzymes. Group A contained the
ATCC strain only and represented 2% of the strains. The B, C, and D groups had 77%, 15%


33
or a rocker platform (Blico Biotechnology, Vineland, NJ) for 2 to 16 h. A 200 /I sample
was plated onto Wilbrinks medium plus 100 /g/ml rifampicin plus 0.5% Bravo 720 and
incubated at 24 C for 72 h. Colonies characteristic of X. fragariae were identified by
amplification with the XF-specific primers and by REA. The remainder of the phosphate
buffer sample was concentrated by vacuum filtration onto a 0.45 m membrane disk
(Millipore Corp, Bedford, MA). The disk was washed in 1.5 ml of TE and the suspension
centrifuged 5 min at 14, 000 x g. The pellet was resuspended in 582 ^1 TEPA (TE buffer
containing 5% PVP40 + 0.02 M sodium ascorbate) and incubated at room temperature 1 h.
DNA was extracted as described above. The crown of the plant was sectioned, macerated
by mortar and pestle in 10 ml of PPA. The plant tissue debris and PVPP were collected by
centrifugation at 1,000 x g for 1 min. The supernatant with the bacterial cells was removed
to a clean centrifuge tube and centrifuged at 14,000 x g for 5 min. The pellet was
resuspended in 582 TE and DNA extraction proceeded as described above. A sample was
plated onto Wilbrinks agar plus 100 /ug/ml rifampicin and 0.5% Bravo and incubated at 24
C for 72 h. The nested PCR reaction and REA were performed as described above.
Results
Pathogenicity assays. All strains identified as X. fragariae by colony characteristics
caused disease symptoms typical of angular leaf spot on strawberry plants. Many other
xanthomonads were isolated from strawberry tissue but they did not cause symptoms of
angular leaf spot


36
XF12 was used in experiments to detect DNA from X fragariae in the presence of plant
materials.
REA of PCR products. PCR products from DNA of X. fragariae amplified with
primer set XF9 and XF12 restricted with Cfol, SaulM, or HaeIII produced distinct banding
patterns for each enzyme (Fig. 3-2). Five bands resulted from Cfo\ restriction (125, 108, 90,
75, and 60 bp), and three bands each from the Hae III restriction (258, 150, and 50 bp) and
Sau3M restriction (250, 125, and 83 bp). Polymorphisms were not observed among strains
of X. fragariae in the REA of the restricted PCR products for the enzymes tested
Sensitivity of detection of X. fragariae in plant tissue bv PCR. X. fragariae was
detected in the presence of plant tissue by amplification with primer XF9 and either XF11
or XF12 and by nested PCR of the two primer sets. The level of sensitivity of detection of
bacterial cells by amplification with one round of PCR was ca. 105 to 104 cfii per ml. The
nested PCR increased sensitivity of the assay 1000-fold (Fig. 3-3) and enabled detection of
ca. 18 cfu per ml. PCR was used to detect bacteria directly from infected plant tissue both
from inoculated strawberry and from plant samples collected in the field. The PCR product
was generated from a single lesion on plant material ground and processed representing a
bacterial population of 104 or above as determined by dilution plating. Identification of X.
fragariae was confirmed by REA of PCR products. The nested technique was not necessary
to detect bacteria from a lesion on leaf tissue as a single round of PCR amplification was
sufficient to obtain a positive result.


4
detect low populations of the bacterium that may exist in asymptomatic tissue. Xanthomonas
fragariae may be identified in the early stages of leaf infection by the diagnostic translucent,
watersoaked lesions viewed with transmitted light; however, older lesions may be confused
with symptoms caused by fungal pathogens (Kennedy and King, 1962b). Diagnosis based on
symptoms is very difficult and not applicable for asymptomatic plants. Identification of the
disease based upon isolation and characterization of the causal agent may also be difficult
because X. fragariae grows slowly and may be masked by faster growing organisms
(Hildebrand et al., 1967). Expression of watersoaked lesions takes 6 days or longer after
inoculation. Thus, fulfillment of Kochs postulates to confirm pathogenicity is difficult and
time-consuming.
Assays have been developed with improved sensitivity and specificity for the detection
of plant pathogens in plant tissue. A specific, indirect, enzyme-linked immunosorbent assay
(ELISA) was developed to detect X. fragariae from symptomatic plant tissue (Rowhani et
al., 1994). The antibody assay did not react with bacterial strains of 16 pathovars of X.
campestris or non-pathogenic bacteria from strawberry leaves. A single cross-reaction
occurred to a strain of X. campestris isolated from Nerium oleander. The assay detected
bacteria directly from a visible lesion on a strawberry leaf. The level of sensitivity of the assay
was at ca. 104 colony forming units (cfu) per ml. However, bacteria on asymptomatic plants
may not be detected by this ELISA assay.
The polymerase chain reaction (PCR) has been used to amplify specific DNA
sequences to detect and identify many plant pathogens including some members of the genus
Xanthomonas (Henson and French, 1993; Louws et al., 1994). Leite et al. (1994) utilized


CHAPTER 2
DISEASE PROGRESS, YIELD LOSS, AND CONTROL OF
XANTHOMONASFRAGARIAE ON STRAWBERRY PLANTS
Introduction
Angular leaf spot of strawberry is caused by the bacterium Xanthomonas fragariae.
The disease was first reported from Minnesota in 1960 (Kennedy and King, 1962a) and is
currently found in many regions of strawberry production throughout the world.
Dissemination of the bacterium occurred via the transportation of infected plants (Maas,
1984; Panagopolous et al., 1978; Dye and Wilkie, 1973). A diagnostic symptom of the
disease is the translucent appearance of lesions when viewed with transmitted light (Maas,
1984). A vascular collapse of the plant from systematic invasion by the bacterium has been
described in California (Hildebrand et al., 1967).
The epidemiology of angular leaf spot is mostly unknown in fields in Florida where
strawberry production is an annual crop. Transplants are obtained each season from nurseries
in Canada and northern states; few transplants are produced in Florida. A source of inoculum
other than infected transplants has not been found Howard (1971) was unable to determine
the inoculum source for infected plants from nurseries in Florida or other states. Kennedy and
King (1962b) determined that bacteria overwintered in infected leaves buried in the soil and
caused disease symptoms on plants the next year. The bacterium did not survive free in the
soil nor were any naturally occurring hosts identified in host range studies (Kennedy and
8


To my husband, Kenn,
Cindy,
and
my parents.


14
Disease progress. Symptoms of angular leaf spot were not visible on plants at
planting in either 1994 or 1995. Disease progress curves of angular leaf spot on strawberry
plants for all treatments are presented in Figure 2-1. In 1994, the disease severity of angular
leaf spot on inoculated plants which did not receive a spray treatment increased for 89 days
after transplanting (DAT), decreased until 118 DAT, and then increased to 25% by 118 DAT
and leveled out to the end of the season. In 1995, the epidemic on plants inoculated with the
bacterium initially increased, then decreased at 61 DAT until 111 DAT, and then increased
and reached 15% by 154 DAT. The disease on inoculated plants sprayed with either the 1 x
or 0.1 x rate of cupric hydroxide plus mancozeb followed similar disease progress curves
except the amount of disease was reduced for each treatment. In contrast, the onset of the
epidemic on the noninoculated plants was greatly delayed. Disease was observed on these
plants after 118 DAT in 1994 and at 125 DAT in 1995. The noninoculated plants had very
little disease (0-3% severity) by the end of the season, regardless of spray treatment.
Statistical analysis of the AUDPC for each treatment is shown in Figure 2-2.
Noninoculated plants had significantly (P = 0.0001) lower disease than plants inoculated with
the bacterium. Noninoculated plants treated with the 1 rate did not have disease either year.
Inoculated plants sprayed with the lx rate of cupric hydroxide and mancozeb had a
significantly (P = 0.0004 in 1994, P 0.0001 in 1995) lower AUDPC compared to the
control treatment in both years. The plants sprayed with the 0.1 x rate of the fungicide had
reduced disease in 1994 (P = 0.17) and, in 1995, the reduction in AUDPC was highly
significant (P = 0.0001) compared to plants without chemical sprays.


47
1995). Primer sets from conserved repetitive bacterial DNA elements generated genomic
fingerprints which were used to differentiate bacteria (De Bruijn, 1992; Louws et al., 1994).
Restriction length fragment polymorphism (RFLP) of the bacterial genome digested with rare-
cutting restriction endonuclease and resolution by pulsed field gel electrophorisis (PFGE) has
been used to type many bacterial strains (Cooksey and Graham, 1989; Egel et al., 1991; Smith
et al., 1995).
In this study, a collection of strains of X. fragariae from the United States and Canada
were examined for genetic diversity. Analyses was by RFLP profiles generated by PFGE and
by fatty acid methyl ester (FAME).
Materials and Methods
Bacterial strains. Strains of X fragariae used in this study are listed in Table 4-1.
Strains were previously rated for pathogenicity on strawberry (Roberts, unpublished).
Geographic origin denotes origin of infested plant material. Strains were cultured on
Wilbrinks medium (WB)(Koike, 1962) at 24 C and long term storage was in 15% glycerol
at -70 C. Single colonies were transfered to nutrient broth (Difco, Detroit, MI) and shaken
for 16 to 20 h at 200 rpm at 24 C.
Restriction fragment length polymorphism. The method used for restriction
endonuclease analysis was described by Egel et al. (1991) and Cooksey and Graham (1989)
except for the following modifications. Cells (1.5 ml of 5 109cfu per ml) were washed in
1 ml of TE (10 mM Tris, ImM EDTA, pH 8.0) and resuspended in 0.5 ml of TE. An equal


41
Discussion
Leite et al. (1994) used the sequence variation within the hrpB operon among plant-
pathogenic xanthomonads to select primers with different specificities and this approach was
successful in our studies to identify primers specific to X fragariae. The hrp-primers RST2
and RST3 amplified DNA from all strains of X fragariae and analysis of the PCR products
by REA showed no polymorphism within this region. The homology of the amplified region
presented a good site to select primers universal to X fragariae. Primers XF9, XF11, and
XF12 designed from these unique sites were specific for amplification of DNA only from
strains ofA^ fragariae. The primer XF10 was responsible for the non-specific amplification
of a strain of X. c. pelargonii in preliminary tests and therefore was not used in later
experiments. Interestingly, an ELISA test developed for identification of X fragariae in vitro
cross reacts to some related bacteria including X. c. pelargonii (Maas, 1995). Perhaps a
relationship between X. fragariae and X. c. pelargonii exists which is not reflected by
traditional taxonomic techniques (Hildebrand et al., 1990; Vauterin et al., 1994; Maas, 1995).
The level of sensitivity for most detection techniques for specific bacteria (Pickup,
1991; Schaad et al., 1982 ) are comparable to the level which was achieved with a single
round of PCR amplification. PCR was reported to be more sensitive than ELISA (Leite et
al., 1995) and with our primers, a single round of PCR is 10 times more sensitive than the
ELISA test for A", fragariae developed by Rowhani et al. (1994). In addition, their antibodies
cross-reacted with another xanthomonad whereas we were able to eliminate the non-specific
reaction by discontinuing use of primer XF10. Although this level of detection is adequate


37
100 bp
Figure 3-2. NuSieve agarose gel showing the restriction patterns generated after
digestion of the PCR product amplified with the primers XF9 and XF12 from the genomic
DNA of strains of Xanthomonas fragariae. The PCR product was digested by restriction
endonucleases Cfo\ and HaelU and the molecular weight marker XI from Boehringer
Mannheim is shown.


72
Howard, C. M and Albregts, E. E. 1973. Strawberry report. Fungicide and Nematicide
Tests 29:47.
Howard, C. M, Overman, A. J., Price, J. F., and Albregts, E. E. 1985. Diseases, nematodes,
mites, and insects affecting strawberries. Florida. Agrie. Experiment Stations, Inst, of Food
and Agrie. Sci., Univ. of FI., Bulletin 857.
Innis, M. A., Gelfand, D. H., Sninsky, J. J., and White, T. J. 1990. PCR Protocols: A Guide
to Methods and Applications. Academic Press, Inc., San Diego, CA.
Jones, A. L. 1995. A stewardship program for using fungicides and antibiotics in apple
disease management programs. Plant Dis. 79:427-432.
Jones, J. B, Woltz, S. S., Kelly, R. O., and Harris, G. 1991. The role of ionic copper, total
copper, and select bactericides on control of bacterial spot of tomato. Proc. Fla. State Hort.
Soc. 104:257-259.
Kennedy, B. W., and King, T. H. 1962a. Angular leaf spot of strawberry caused by
Xanthomonasfragariae sp. nov. Phytopathology 52:873-875.
Kennedy, B. W., and King, T. H. 1962b Studies on epidemiology of bacterial angular leaf
spot on strawberry. Plant Dis. Rep. 46:360-363.
Koike, H. 1965. The aluminum-cap method for testing sugarcane varieties against leaf scald
disease. Phytopathology 55:317-319.
Lai, C. M. 1978. Bacterial disease of strawberry. Plant Path. Reports. California Dept. Of
Food and Agrie. Lab Services. Vol. II.
Leite, R. P., Egel, D. S., and Stall, R. E. 1994. Genetic analysis of hrp-related DNA
sequences of Xanthomonas campestris strains causing diseases of citrus. Appl Environ.
Microbiol. 60:1078-1086.
Leite, R. P., Jr., Jones, J. B., Somodi, G. C., Minsavage, G. V., and Stall, R E. 1995.
Detection of Xanthomonas campestris pv. vesicatoria associated with pepper and tomato
seed by DNA amplification. Plant Dis. 79:917-922.
Leite, R. P., Jr., Minsavage, G. V., Bonas, U., and Stall, R E. 1994. Detection and
identification of phytopathogenic Xanthomonas strains by amplification of DNA sequences
related to the hrp genes of Xanthomonas campestris pv. vesicatoria. Appl. Environ.
Microbiol. 60:1068-1077.


71
Fitch, W. C, and Margoliash, E. 1967. Construction of phylogenetic trees. Science 155:
279-284.
Fry, W E. 1975. Integrated effects of polygenic resistance and a protective fungicide on
development of potato late blight. Phytopathology 65:908-911.
Fry, W. E., Bruck, R. I., and Mundt, C. C. 1979. Retardation of potato late blight epidemics
by fungicides with eradicant and protectant properties. Plant Dis. Rep. 63:970-974.
Gabriel, D. W., Hunter, J. E., Kingsley, M. T., Miller, J. M., 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.
Hartung, J. S., and Civerolo, E. L. 1989. Restriction fragment length polymorphisms
dintinguish Xanthomonas campestris strains isolated from Florida citrus nurseries from X. c.
pv. citri. Phytopathology 79:793-799.
Hazel, W. J. 1981. Xanthomonas fragariae, cause of strawberry angular leaf spot: Its
growth, symptomatology, bacteriophages, and control. Ph D. thesis. University of Maryland,
College Park. 119 p.
Hazel, W. J., and Civerolo, E. L 1980. Procedures for growth and inoculation of
Xanthomonas fragariae, causal organism of angular leaf spot of strawberry. Plant Dis.
64:178-181.
Henson, J. M., and French, R. 1993. The polymerase chain reaction and plant disease
diagnosis. Ann. Rev. Phytopathol 31:81-109.
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.
Hildebrand, D. C., Schroth, M N., and Wilhelm, S. 1967. Systemic invasion of strawberry
by Xanthomonas fragariae causing vascular collapse Phytopathology 57:1260-1261.
Hodge, N. C., Chase, A. R., and Stall, R. E. 1992. Diversity of four species of Xanthomonas
as determined by cellular fatty acid analyses. (Abst.) Phytopathology 82:1 153.
Howard, CM. 1971. Occurrence of strawberry angular leaf spot, Xanthomonas fragariae,
in Florida Plant Dis. Rep. 55:142.


3
control a bacterial disease (Jones et al., 1991). Antibiotics and copper compounds were
effective protectants against angular leaf spot but these did not eradicate the disease (Alippi
et al., 1989). Marco and Stall (1983) examined chemical control of strains of Xanthomonas
campestris pv. vesicatoria that differed in sensitivity to copper. The mixture of cupric
hydroxide plus mancozeb was more effective than cupric hydroxide alone to control both the
copper-resistant and the copper-sensitive strain. The application of copper compounds to
strawberry plants is confounded by the fact that copper can be phytotoxic to strawberry plants
(Howard and Albregts, 1973). The application of a chemical can reduce the infection by a
pathogen and slow the rate of the epidemic, e g. van der Planks (1963) apparent infection
rate (Fry et al., 1979; Pataky and Lim, 1981). A reduced rate of fungicide applied at frequent
intervals as a protectant has been used effectively to reduce disease on crops (Fry 1975; Fry
et al, 1979; Conway et al., 1987).
Identification of plants infected with X fragariae is a priority because of the ease of
movement of infected but asymptomatic plants (Maas, 1995). International movement of
infected plants is blamed for introduction of angular leaf spot into Greece and New Zealand
(Panagopoulos et al., 1978; Dye and Wilkie, 1973). Nursery-plant producers are pressured
to provide disease-free plants by foreign countries and by farmers who refuse to buy infected
transplants. The European Plant Protection Organization (EPPO) lists X. fragariae as a
quarantine pest and has prescribed phytosanitary procedures. In the future, regulatory issues
may be of greater concern. The production of disease-free plants is essential for control of
angular leaf spot. Therefore, accurate identification of plants infected with the bacterium is
imperative. Available detection techniques are limited in their usefulness and accuracy to


75
Stall, R. E., and Thayer, P. L. 1962. Streptomycin resistance of the bacterial spot pathogen
and control with streptomycin. Plant Dis. Rep. 16:389-392.
Szybalski, W. 1952. Microbial Selection. 1. Gradient plate technique for study of bacterial
resistance. Science 116:46-48.
Vauterin, L., Hoste, B., Kersters, K. and Swings, J. 1995. Reclassification ofXanthomonas.
Int. J. Syst. Bacteriol. 45:472-489.
van der Plank, J. E. 1963. Control of Disease by Fungicides. Plant Diseases: Epidemics and
Control. Academic Press, NY.
Willis, K., Rich, J. J., and Hrbak, E M. 1991. hrp genes of phytopathogenic bacteira. Mol.
Plant Microbe Interact. 4:132-138


LO
1780
1996
, Rm
UNIVERSITY OF FLORIDA
3 1262 08554 9185


38
XF9 and XF11 Nested XF9 and XF12
500 bp
87654321 87654321
Log cfu per ml
Figure 3-3. Agarose gel showing the sensitivity of detection by PCR and nested PCR in
the presence of plant tissue. The genomic DNA from various concentrations of cells was
extracted and amplified by primers XF9 and XF11 and used in a second round of nested
amplification with primers XF9 and XF12.


63
alternate approach might be to control the disease in the nursery by copper applications. The
loss in yield due to phytotoxicity would not be an issue on nursery plants. The sprays would
reduce the amount of initial inoculum on transplants and subsequently reduce the amount of
disease in fruit production fields. The effect of sprays on the categories of yield were difficult
to interpret because of the lack of consistent, significant differences between years.
Leite et al. (1994) used the sequence variation within the hrpB operon among plant-
pathogenic xanthomonads to select primers with different specificities and this approach was
successful in our studies to identify primers specific to X. fragariae. The hrp-primers RST2
and RST3 amplified DNA from all strains ofJf. fragariae and analysis of the PCR products
by REA showed no polymorphism within this region. The homology of the amplified region
presented a good site to select primers universal to X. fragariae. Primers XF9, XF11, and
XF12 designed from these unique sites were specific for amplification of DNA only from
strains ofA^ fragariae. The primer XF10 was responsible for the non-specific amplification
of a strain of X. c. pelargonii in preliminary tests and therefore was not used in later
experiments. Interestingly, an ELISA test developed for identification of X. fragariae in vitro
cross reacts to some related bacteria including X. c. pelargonii (Maas, 1995). Perhaps a
relationship between X fragariae and X. c. pelargonii exists which is not reflected by
traditional taxonomic techniques (Hildebrand et al., 1990; Vauterin et al., 1994; Maas, 1995).
The level of sensitivity for most detection techniques for specific bacteria (Pickup,
1991; Schaad et al., 1982 ) are comparable to the level which was achieved with a single
round of PCR amplification. PCR was reported to be more sensitive than ELISA (Leite et
al., 1995) and with our primers, a single round of PCR is 10 times more sensitive than the


74
Rhoads, D. D., and Roufa, D.J. 1991. Seqaid II Computer Program. Molecular Genetics
Lab, Kansas State University, Manhattan, KS 66506.
Ritchie, D. F., Averre, C. W., and Milholland, R D. 1993. First report of angular leaf spot,
caused by Xanthomonasfragariae, on strawberry in North Carolina. Plant Dis. 76:1263.
Rowhani, A., Feliciano, A. J., Lips, T., and Gubler, W. D. 1994. Rapid identification of
Xanthomonasfragariae in infected strawberry leaves by enzyme-linked immunosorbent assay.
Plant Dis. 78:248-250.
SAS Institute Inc. 1988. SAS/STAT User's Guide, ed. 6.03. SAS Institute Inc., Cary, N.C.
Sasser, M. 1990. Tracking a strain using the the microbial identification system. Technical
note 102. Microbial ID, Inc. Newark, DE.
Sasser, M. 1990. Identification of bacteria through fatty acid analysis. Methods in
Phytobacteriology. Z. Klement, K., Rudolph, and D. Sands, eds. Akademiai Kiado,
Budapest.
Schaad, N. W., Cheong, S. S., Tamaki, S., Hatziloukas, E., and Panopoulos, N. J. 1993. A
viable cell enrichment, two-step, direct PCR technique for detection of Pseudomonas
syringae pv. phaseolicola in bean seeds. (Abstr.) Phytopathology 83:1342.
Shaner, G., and Finney, R E. 1977. The effect of nitrogen fertilization on the expression of
slow-mildewing resistance in Knox wheat. Phytopathology 67:1051-1056.
Smith, J. J., Offord, L. C., Holderness, M., and Saddler, G. S. 1995. Genetic diversity of
Burkholderia solanacean/m (synonym Pseudomonas solanacearum) race 3 in Kenya. Appl.
Environ. Microbiol. 4263-4268.
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. Sys. Bacteriol.
44:47-53.
Stall, R. E., Loschke, D. C., and Jones, J. B. 1986. Linkage of copper resistance and
avirulence loci on a self-transmissible plasmid in Xanthomonas campestris pv. vesicatoria.
Phytopathology 76:240-243.
Stall, R. E., and Minsavage, G. V. 1990. The use of hrp genes to identify opportunistic
xanthomonads. Proc. 7th Int. Conf. Plant Pathog. Bacteria 1989. Z. Klement, Akadmiai
Kiad, Budapest.


61
epidemiology of angular leaf spot is necessary to determine the effectiveness and practicability
of any control method. The survey of farmers fields and the cold storage facility established
that transplants arrive from northern nurseries infected with angular leaf spot at fairly high
disease incidence. Therefore, disease is introduced into the field on infected transplants. In
our surveys, by the sampling method used in the first season, more plants were assessed for
symptoms of angular leaf spot in the field compared to the number of plants in the random
sampling of boxes in cold storage. This probably accounts for identification of diseased plants
in nearly all of the farmers fields versus less than in 25% of boxed plants. In addition, plants
were examined until disease was found in the field which skewed the randomness of the
survey.
In field experiments, the spread from inoculated plants to noninoculated plants in
nearby experimental plots was minimal. Thus, the incidence of plants with angular leaf spot
in the field likely occurred from inoculum already present on the transplants. In the field
experiments, if inoculum was present from another source, such as debris in the soil or an
alternate host, most likely the disease on the noninoculated plants would have been more
general and appeared earlier. Disease severity on noninoculated plants in the experimental
plots was extremely low and most plants remained free of angular leaf spot. For the disease
that occurred on noninoculated plants, the inoculum was probably transported mechanically
from inoculated plants during harvests or disease readings.
Progress of angular leaf spot developed similar to curves of pathogens described on
other crops (van der Plank, 1963). A decrease in disease severity was seen about midway
through the season in 1994. This reduction, or negative infection rate, was due to growth


28
Minsavage, 1990; Lindgren et al., 1986 ) and DNA was not amplified by the hrp gene cluster
primers. Differentiation of X. campestris pathovars was made by restriction endonuclease
analysis (REA) patterns generated by digestion of the PCR products with frequent-cutting
enzymes (Leite et al., 1994 and 1995). The primers were used to detect X. c. vesicatoria in
seed lots of naturally infected pepper and tomato (Leite et al., 1995). The sensitivity of
detection by PCR is reported at 103 to 102 cfu per ml (Minsavage et al., 1994; Leite et al.,
1994; Henson and French, 1993). In nested PCR, sensitivity of detection is increased by
using PCR products from an amplification as target DNA in a second round of amplification
by a second set of primers internal to the first (Schaad et al., 1993). McManus and Jones
(1995) reported an increase in sensitivity of 1,000 -fold with nested PCR over a single round
of PCR amplification to detect Erwinia amylovora.
Our objectives were to develop a sensitive and specific technique for detection of X
fragariae. Our approach was to design primers specific to the region of genomic DNA from
X. fragariae related to the hrp genes of X c. vesicatoria. The survival of X. fragariae on
nursery strawberry plants in the field at two locations in Florida and dissemination to daughter
plants was examined to understand the disease cycle of angular leaf spot in Florida
Materials and Methods
Bacterial strains and culture conditions. Strains of X fragariae and non-pathogenic
xanthomonads isolated from strawberry were maintained at 24 C on Wilbrink's medium
(Koike, 1965). Pathovars of Xanthomonas campestris were cultured on nutrient agar (Difco


17
Yield loss. In both years, marketable yield on plants infected with angular leaf spot
was significantly decreased (P = 0.005 in 1994 and P = 0.04 in 1995) compared to yield on
non-inoculated plants (Fig. 2-3). Total and marketable yields were reduced by an average of
7% and 8%, respectively, in 1994 and by 6% and 10% in 1995. Total yield had similar
reductions as marketable yield, therefore only marketable yield is presented. Yields were
generally lower in 1995 than in 1994.
The effect of spray treatments on the yield from inoculated and noninoculated plants
is presented in Figure 2-4. Marketable yield on noninoculated plants sprayed with the 1*
treatment of cupric hydroxide plus mancozeb was significantly reduced (P = 0.04) compared
to plants which received the 0.1 x or no-spray treatments in 1994. No difference in yield was
detected on inoculated plants receiving the spray treatments in 1994. In 1994, the yield on
inoculated plants sprayed with the 1 x rate was significantly reduced (P = 0.005) compared
to the other two treatments. Yield on noninoculated plants sprayed with the lx and 0. lx
chemical sprays was reduced significantly (P =0.04) compared to yield on nonsprayed plants
in 1995. The spray treatments did not have any effect on culled fruit in 1994 or 1995 (Fig.
2-5). No difference in culled fruit was detected among spray treatments. Nonmarketable fruit
lost because of damage by fungal diseases was significantly different at P = 0.01 for all
treatments in 1994 but the relationship to spray treatments was unclear (Fig. 2-6). In 1995,
there was no significant difference in damage to fruit by fungi for either inoculated or spray
treatments. Overall, the total yield lost to fungal infections was much lower in 1994
compared to 1995.


6
In addition, the identification of genetic types is a prerequisite to identify sources of resistance
in strawberry to the pathogen. While strawberry cultivars exhibited levels of susceptibility or
tolerance to X. fragariae; only F. moschata Duch. appeared to be immune (Hazel, 1981;
Hazel and Civerolo, 1980; Kennedy and King, 1962b). A screening program to identify genes
for resistance must incorporate representatives of the genetic variants in the screening process
otherwise the resistance may be overcome quickly by genetic variants.
Studies to characterize populations of bacteria have used biochemical and molecular
biological techniques. Protein staining and fatty acid analysis have been useful to detect
differences at the metabolic level within species of Xanthomonas (Bouzar et al., 1994; Stall
et al., 1994). The polymerase chain reaction (PCR) has been used to analyze differences
between pathovars and strains of bacteria (Hensen and French, 1993). Primers specific to the
hrp-gene cluster of Xanthomonas campestris pv. vesicatoria amplified genomic DNA and
restriction enzyme analysis of the PCR product differentiated between many pathovars and
species of Xanthomonas (Leite et al., 1994 and 1995). Primers sets from conserved repetitive
bacterial DNA elements generated genomic fingerprints which were used to differentiate
gram-negative soil bacteria and pathovars of Xanthomonas and Psuedomonas (De Bruijn,
1992; Louws et al., 1994). Restriction length fragment polymorphisms (RFLP) of the
bacterial genome digested with rare-cutting restriction endonucleases and resolution by pulsed
field gel electrophoresis (PFGE) has been used to type many bacterial species and genera
(Cooksey and Graham, 1989; Egel et al., 1991; Smith et al., 1995).
The objectives of these studies were to investigate epidemiological aspects of angular
leaf spot on strawberry including the incidence of X. fragariae on transplants from northern


19
Noninoc Inoc
Noninoc Inoc
1994
1995
Figure 2-4. The marketable yield on strawberry in Florida in the 1994 (A) and 1995 (B)
seasons. A, B = F was significant at P = 0.05 within the first factor of noninoculated
vs.inoculated; a, b = F was significant at P = 0.05 within the second factor of spray
treatments.


9
King, 1962b). For bacteria on plant refuse to serve as an inoculum source in Florida, the
bacterium must oversummer. Optimal growth (=20 C) of the bacterium (Howard et al.,
1985; Kennedy and King, 1962b) is at temperatures cooler than normally occurs in Florida
during the summer. The bacterium survived on summer nursery plants in the field in Florida
but at very low populations (Roberts, unpublished) The effect of angular leaf spot on yield
is unknown. Howard (1971) accredited some yield losses due to angular leaf spot in fields
in Florida but he did not quantify the losses. In Wisconsin, a decrease in yield of 70 to 80%
was estimated due to the disease (Epstein, 1966). However, the production in the northern
United States is a perennial, matted-row system which differs significantly from the Florida
situation. A significant loss in marketable fruit may occur due to infections of the calyx. The
sepals become brown and dry and the fruit is unmarketable because of its unattractive
appearance (Epstein, 1966; Maas, 1995).
Chemical control of bacterial diseases on plants is difficult. Antibiotics have limited
effectiveness over time since mutations may form resistant populations (Stall and Thayer,
1962). Therefore copper compounds are frequently used to control a bacterial disease (Jones
et al., 1991). Antibiotics and copper compounds were effective protectants against angular
leaf spot but these did not eradicate the disease (Alippi et al., 1989). Marco and Stall (1983)
examined chemical control of strains of Xanthomonas campeslris pv. vesicatoria which
differed in sensitivity to copper. The mixture of cupric hydroxide plus mancozeb was more
effective than cupric hydroxide alone to control both the copper-resistant and the copper-
sensitive strain. The application of copper compounds to strawberry plants is confounded by
the fact that copper can be phytoxic to strawberry plants (Howard and Albregts, 1973). The


56
1 through 4, contained 16%, 6%, 4%, and 74% of the strains, respectively. The largest
cluster, FAME group 4, could be subdivided at the Euclidian distance of 4 units into two
groups. The ATCC type strain 33239 and one other strain comprised FAME 3 group.
Discussion
This research represents the first effort to analyze genetic variants of strains within the
species of X. fragariae. The RFLP-PFGE and FAME analyses identified genetic variants
within the population. Methods of RFLP including PFGE have distinguished between closely
related strains of pathovars within the genus Xanthomonas campestris(Ege\ et al., 1991;
Gabriel et al., 1988; Hartung and Civerolo; 1989, Graham et al., 1990). In these works, the
importance of the diverse nature of the endonuclease sites as related to strain characteristics
such as host range, geographic origin, and pathogenicity was discussed. Unfortunately, the
relationship of pathogenicity of the X. fragariae strains to the genetic groupings cannot be
examined Differences in pathogenicity among strains of X.fragariae were not detected by
inoculation of the 52 strains onto two cultivars of strawberry. Studies of bacterial populations
on two cultivars which appeared to have different levels of susceptibility in preliminary tests
were also inconclusive (Roberts, unpublished). Nor have other researchers reported
differences in pathogenicity among strains of X. fragariae Such information might have been
useful to determine significance of the RFLP-PFGE and FAME groups as related to
pathogenicity.


48
Table 4.1. The geographic source, collection date, and group determined by fatty acid
methyl ester (FAME) and restriction fragment length polymorphism (RFLP ) for strains
used in this study.
Strain
Geographic
Source3
Year
1238
CA
1990
1240
CA
1990
1241
CA
1990
1242
CA
1990
1243
CA
1990
1245
CA
1990
1246
CA
1990
1249
CA
1990
1250
CA
1990
1290
CA
1989
1291
CA
1989
1293
CA
1989
1295
CA
1989
1296
CA
1989
1298
CA
1989
1424
FL
1992
1425
FL
1992
1426
FL
1992
1427
FL
1992
1428
FL
1992
1429
FL
1992
1431
FL
1992
1514
CA
1993
1515
CA
1993
1516
NC
1993
1517
NC
1993
1518
NC
1993
1519
NC
1993
1520
CA
1993
1523
CA
1993
1524
CA
1993
1525
CA
1993
1526
CA
1993
Sourceb FAMEC RFLP
ARC
2
D
ARC
2
D
ARC
8
B
ARC
8
B
ARC
8
B
ARC
4
B
ARC
8
ARC
4
ARC
5
C
ARC
C
ARC
1
B
ARC
7
B
ARC
8
ARC
8
B3
ARC
8
B
ARC
C
ARC
3
B2
ARC
B
ARC
3
B
ARC
3
B2
ARC
3
B
ARC
B
ARC
C
ARC
6
B3
ARC
B2
ARC
7
B3
ARC
8
B
ARC
B
ARC
5
C
ARC
5
C
ARC
5
c
ARC
c
ARC
8
B
Geographic origin of plant material from which bacteria were isolated. CAN=Canada.
bARC= A. R. Chase; ATCC= American Type Culture Collection.
Blank space indicates that group was not determined.


THE EPIDEMIOLOGY, SPECIFIC DETECTION, AND GENETIC VARIABILITY OF
XANTHOMONAS FRAGARIAE
BY
PAMELA D. ROBERTS
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
1996

To my husband, Kenn,
Cindy,
and
my parents.

ACKNOWLEDGMENTS
The author would like to gratefully acknowledge the members of the committee:
Dr. R. D. Berger, Dr. R. E. Stall, Dr. C. K. Chandler, and Dr. Jeffrey B Jones. The
enthusiatic support, expert advise, and guidance throughout the duration of this project were
greatly appreciated. Special thanks go to Dr. Jeffrey B. Jones who provided his laboratory
for a majority of the work. Thanks go to Dr. A. R. Chase who initially suggested the project
and provided considerable support the first year.
Special thanks go to the members of Dr. Stalls and Dr. Jones labs, including N. Cheri
Hodges, G. V. Minsavage, Rui P Leite, Jr., Hacene Bouzar, Gail Somodi, Jeanette Chun,
and Rick Kelly, for their help and information.
Very special thanks go to the people at Dover-G.C.R.E.C. who helped me
tremendously with all aspects of the field research and taught me all about the production of
strawberries: Alicia Whidden, Anne Turgeau, Jim Sumler, Dale Wenzel, Larry Smith, and
Mitch Boles.
Finally, thanks go to the folks who helped make the experience enjoyable: Gary
Marlow, Pia D. Gavino, Morgan Wallace, Pamela Lopez, and Marion Bogart.
in

TABLE OF CONTENTS
page
ACKNOWLEDGMENTS iii
ABSTRACT vi
CHAPTERS
1 INTRODUCTION 1
2 DISEASE PROGRESS, YIELD LOSS, AND CONTROL OF
XANTHOMONAS FRAGARIAE ON STRAWBERRY
PLANTS 8
Introduction 8
Materials and Methods 10
Results 13
Discussion 22
3DEVELOPMENT OF A SPECIFIC AND SENSITIVE DETECTION
TECHNIQUE BY THE POLYMERASE CHAIN REACTION
FOR XANTHOMONAS FRAGARIAE AND APPLICATION
IN A STUDY OF SURVIVAL OF THE BACTERIUM ON
STRAWBERRY PLANTS 25
Introduction 25
Materials and Methods 28
Results 33
Discussion 41
IV

4 GENOMIC RELATEDNESS OF XANTHOMONASFRAGARIAE
ON STRAWBERRY BY FATTY ACID METHYL
ESTERASE AND RESTRICTION LENGTH FRAGMENT
POLYMORPHISM ANALYSES 45
Introduction 45
Materials and Methods 47
Results 51
Discussion 56
5 DISCUSSION 60
LIST OF REFERENCES 70
BIOGRAPHICAL SKETCH 76
v

Abstract of 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
THE EPIDEMIOLOGY, SPECIFIC DETECTION, AND GENETIC VARIABILITY OF
XANTHOMONAS FRAGAR1AE
By
Pamela D. Roberts
May 1996
Chairman: Dr. R. D. Berger
Major Department: Plant Pathology
Angular leaf spot (ALS) on strawberry is caused by the bacterium Xanthomonas
fragariae (XF). The epidemiology of ALS, development of a detection technique by the
polymerase chain reaction (PCR), and the genetic variability of strains were examined. The
disease severity of ALS on plants in field plots increased to 25% and 15% in two seasons.
Yield was decreased 8% to 10%. Minimal spread of ALS occurred between field plots.
Chemical sprays applied at the label rate of cupric hydroxide plus mancozeb at 7- to 14- day
intervals decreased disease but was phytotoxic and decreased yield. A 10% rate of the
mixture applied frequently slightly reduced disease but increased yield one season and
significantly reduced disease and did not affect yield the next season. Three primers were
specific for amplification of DNA from XF but not DNA from strains of 16 pathovars of
Xanthomonas campestris or non-pathogenic xanthomonads from strawberry. Bacteria were
detected at 104 colony forming units per ml by a single round of PCR A nested PCR
vi

technique increased detection 1,000 fold. Plants were inoculated with a rifampicin-resistant
strain and oversummered in the field at two locations in Florida. Bacteria from leaf and
crown samples were detected by nested PCR and recovery onto selective media at two- week
intervals for 92 days after planting. Daughter plants of the inoculated plants were positive
for XF by nested PCR amplification Analysis of genetic variability by fatty acid methyl
esterase (FAME) profiles divided 50 strains into 9 groups based upon qualitative and
quantitative differences. The majority (74%) of strains were placed into a closely related
group determined by cluster analysis. The profiles from restriction fragment length
polymorphism (RFLP) analysis of genomic DNA restricted by two infrequent cutting
endonucleases and separation by pulsed field gel electrophoresis grouped strains into four
groups (A-D). Another endonuclease subdivided the B group into three groups. The
dendrogram unweighted pair analysis of FAME profiles divided the population into four
groups which correlated well with the RFLP groups. Considerable diversity appears within
the species.
vii

INTRODUCTION
Xanthomonas fragariae causes angular leaf spot disease on strawberry (Fragaria
species; Fragaria ananassa Duchesne). The disease was first found in Minnesota in 1960
(Kennedy and King, 1962a) and it is currently found in many regions of strawberry
production throughout the world (Maas, 1984; Ritchie et al., 1993). Angular leaf spot was
first reported in Florida in 1971 (Howard, 1971). Dissemination of the bacterium occurred
via the transportation of infected plants (Maas, 1984; Panagopolous et al., 1978; Dye and
Wilkie, 1973). In Florida, strawberry plants which arrive from northern nurseries for
transplanting in the fall frequently have leaves which exhibit symptoms of angular leaf spot.
Typical pathogenic strains of X. fragariae can be isolated from the lesions. A diagnostic
symptom of the disease is the translucent appearance of lesions when viewed with transmitted
light (Maas, 1984). A vascular collapse of the plant from systematic invasion by the
bacterium has been described in California (Hildebrand et al., 1967).
The epidemiology of angular leaf spot is mostly unknown in fields in Florida where
strawberry production is an annual crop. Transplants are obtained each season from nurseries
in Canada and northern states; few transplants are produced in Florida. A source of inoculum
other than infected transplants has not been found Howard (1971) was unable to determine
the inoculum source for infected plants from nurseries in Florida or other states. In surveys

2
conducted in 1968, 1970, 1993, and 1994, plants which had symptoms of angular leaf spot
in the spring did not have symptoms of the disease the following August (Howard, 1971; P.
D. Roberts, unpublished). However, in 1969 mild infections on one variety were observed
in mid-August. Plants transplanted to fields did not develop angular leaf spot. Kennedy and
King (1962b) determined that the bacterium overwintered in infected leaves buried in the soil
and caused disease symptoms on plants the next year. The bacterium did not survive free in
the soil nor were any naturally occurring hosts identified in host-range studies (Kennedy and
King, 1962a). For bacteria on plant refuse to serve as an inoculum source in Florida, the
bacterium must oversummer. Optimal growth (-20 C) of the bacterium (Howard et al.,
1985; Kennedy and King, 1962b) is at temperatures cooler than normally occurs in Florida
during the summer. The survival of the bacterium on plants in summer nurseries in Florida
and inoculum sources other than infected transplants have not been established.
The effect of angular leaf spot on yield is unknown. Howard (1971) accredited some
yield losses due to angular leaf spot in fields in Florida but he did not quantify the losses. In
Wisconsin, a decrease in yield of 70 to 80% was estimated due to the disease (Epstein, 1966).
However, the production in the northern United States is a perennial, matted-row system
which differs significantly from the Florida production. A significant loss in marketable fruit
may occur due to infections of the calyx. The sepals become brown and dry and the fruit is
unmarketable because of its unattractive appearance (Epstein, 1966; Maas, 1995).
Chemical control of bacterial diseases is difficult. Antibiotics have limited
effectiveness over time since mutations to bacterium may occur and form resistant
populations (Stall and Thayer, 1962). Therefore copper compounds are frequently used to

3
control a bacterial disease (Jones et al., 1991). Antibiotics and copper compounds were
effective protectants against angular leaf spot but these did not eradicate the disease (Alippi
et al., 1989). Marco and Stall (1983) examined chemical control of strains of Xanthomonas
campestris pv. vesicatoria that differed in sensitivity to copper. The mixture of cupric
hydroxide plus mancozeb was more effective than cupric hydroxide alone to control both the
copper-resistant and the copper-sensitive strain. The application of copper compounds to
strawberry plants is confounded by the fact that copper can be phytotoxic to strawberry plants
(Howard and Albregts, 1973). The application of a chemical can reduce the infection by a
pathogen and slow the rate of the epidemic, e g. van der Planks (1963) apparent infection
rate (Fry et al., 1979; Pataky and Lim, 1981). A reduced rate of fungicide applied at frequent
intervals as a protectant has been used effectively to reduce disease on crops (Fry 1975; Fry
et al, 1979; Conway et al., 1987).
Identification of plants infected with X fragariae is a priority because of the ease of
movement of infected but asymptomatic plants (Maas, 1995). International movement of
infected plants is blamed for introduction of angular leaf spot into Greece and New Zealand
(Panagopoulos et al., 1978; Dye and Wilkie, 1973). Nursery-plant producers are pressured
to provide disease-free plants by foreign countries and by farmers who refuse to buy infected
transplants. The European Plant Protection Organization (EPPO) lists X. fragariae as a
quarantine pest and has prescribed phytosanitary procedures. In the future, regulatory issues
may be of greater concern. The production of disease-free plants is essential for control of
angular leaf spot. Therefore, accurate identification of plants infected with the bacterium is
imperative. Available detection techniques are limited in their usefulness and accuracy to

4
detect low populations of the bacterium that may exist in asymptomatic tissue. Xanthomonas
fragariae may be identified in the early stages of leaf infection by the diagnostic translucent,
watersoaked lesions viewed with transmitted light; however, older lesions may be confused
with symptoms caused by fungal pathogens (Kennedy and King, 1962b). Diagnosis based on
symptoms is very difficult and not applicable for asymptomatic plants. Identification of the
disease based upon isolation and characterization of the causal agent may also be difficult
because X. fragariae grows slowly and may be masked by faster growing organisms
(Hildebrand et al., 1967). Expression of watersoaked lesions takes 6 days or longer after
inoculation. Thus, fulfillment of Kochs postulates to confirm pathogenicity is difficult and
time-consuming.
Assays have been developed with improved sensitivity and specificity for the detection
of plant pathogens in plant tissue. A specific, indirect, enzyme-linked immunosorbent assay
(ELISA) was developed to detect X. fragariae from symptomatic plant tissue (Rowhani et
al., 1994). The antibody assay did not react with bacterial strains of 16 pathovars of X.
campestris or non-pathogenic bacteria from strawberry leaves. A single cross-reaction
occurred to a strain of X. campestris isolated from Nerium oleander. The assay detected
bacteria directly from a visible lesion on a strawberry leaf. The level of sensitivity of the assay
was at ca. 104 colony forming units (cfu) per ml. However, bacteria on asymptomatic plants
may not be detected by this ELISA assay.
The polymerase chain reaction (PCR) has been used to amplify specific DNA
sequences to detect and identify many plant pathogens including some members of the genus
Xanthomonas (Henson and French, 1993; Louws et al., 1994). Leite et al. (1994) utilized

5
primers specific to regions of the hrp gene cluster which confers hypersensitivity and
pathogenicity from Xanthomonas campestris pv. vesicatoria to detect pathogenic
xanthomonads. Saprophytic and non-pathogenic bacteria lack the hrp genes (Stall and
Minsavage, 1990; Lindgren et al., 1986; Willis et al., 1991) and DNA from these bacteria was
not amplified by the htp gene cluster primers. Differentiation of X. campestris pathovars was
made by restriction endonuclease analysis (REA) patterns generated by digestion of the PCR
products with frequent cutting enzymes (Leite et al., 1994 and 1995). The primers were used
to detect X c. vesicatoria in seed lots of naturally infected pepper and tomato (Leite et al.,
1995). The sensitivity of detection by PCR is reported at 103 to 102 cfu per ml (Minsavage
et al., 1994; Leite et al., 1994; Henson and French, 1993). In nested PCR, sensitivity of
detection is increased by using PCR products from an amplification as target DNA in a
second round of amplification by a second set of primers internal to the first (Schaad et al.,
1993). McManus and Jones (1995) reported an increase in sensitivity of 1,000-fold with
nested PCR over a single round of PCR amplification to detect Erwinia amylovora.
The epidemiology of X fragariae must be understood before effective control
strategies can be devised. The delineation of bacterial populations is imperative to such
studies. The relationship of X. fragariae to other members of the genus Xanthomonas has
been examined (Hildebrand et al., 1990; Vauterin et al., 1995; Hodge et al., 1992), but the
genetic variability of strains within the species has not been reported (Maas, 1995). The
information from studies to compare strains of X. fragariae at the genetic level could be
useful to identify the origins and spread of specific strains or genetic types. This information
could be applied in international tracking of pathogen populations for quarantine programs.

6
In addition, the identification of genetic types is a prerequisite to identify sources of resistance
in strawberry to the pathogen. While strawberry cultivars exhibited levels of susceptibility or
tolerance to X. fragariae; only F. moschata Duch. appeared to be immune (Hazel, 1981;
Hazel and Civerolo, 1980; Kennedy and King, 1962b). A screening program to identify genes
for resistance must incorporate representatives of the genetic variants in the screening process
otherwise the resistance may be overcome quickly by genetic variants.
Studies to characterize populations of bacteria have used biochemical and molecular
biological techniques. Protein staining and fatty acid analysis have been useful to detect
differences at the metabolic level within species of Xanthomonas (Bouzar et al., 1994; Stall
et al., 1994). The polymerase chain reaction (PCR) has been used to analyze differences
between pathovars and strains of bacteria (Hensen and French, 1993). Primers specific to the
hrp-gene cluster of Xanthomonas campestris pv. vesicatoria amplified genomic DNA and
restriction enzyme analysis of the PCR product differentiated between many pathovars and
species of Xanthomonas (Leite et al., 1994 and 1995). Primers sets from conserved repetitive
bacterial DNA elements generated genomic fingerprints which were used to differentiate
gram-negative soil bacteria and pathovars of Xanthomonas and Psuedomonas (De Bruijn,
1992; Louws et al., 1994). Restriction length fragment polymorphisms (RFLP) of the
bacterial genome digested with rare-cutting restriction endonucleases and resolution by pulsed
field gel electrophoresis (PFGE) has been used to type many bacterial species and genera
(Cooksey and Graham, 1989; Egel et al., 1991; Smith et al., 1995).
The objectives of these studies were to investigate epidemiological aspects of angular
leaf spot on strawberry including the incidence of X. fragariae on transplants from northern

7
nurseries, disease spread in strawberry production fields, the effect of angular leaf spot on
yield, and the control of the disease by chemicals. To aid in epidemiological studies, a
sensitive and specific technique by PCR reaction to detect X. fragariae was developed.
Primers were designed specific to the region of genomic DNA from X. fragariae related to
the hrp genes of X. c. vesicatoria. The survival of X fragariae on nursery strawberry plants
in the field at two locations in Florida and dissemination to daughter plants was examined to
understand the disease cycle of angular leaf spot in Florida. The genetic variability of a
collection of strains of X. fragariae from the United States and Canada was examined.
Analyses were by restriction length fragment profiles of genomic DNA restricted with rare-
cutting endonucleases separated by pulsed field gel electrophoresis and by profiles of fatty
acid methyl esters.

CHAPTER 2
DISEASE PROGRESS, YIELD LOSS, AND CONTROL OF
XANTHOMONASFRAGARIAE ON STRAWBERRY PLANTS
Introduction
Angular leaf spot of strawberry is caused by the bacterium Xanthomonas fragariae.
The disease was first reported from Minnesota in 1960 (Kennedy and King, 1962a) and is
currently found in many regions of strawberry production throughout the world.
Dissemination of the bacterium occurred via the transportation of infected plants (Maas,
1984; Panagopolous et al., 1978; Dye and Wilkie, 1973). A diagnostic symptom of the
disease is the translucent appearance of lesions when viewed with transmitted light (Maas,
1984). A vascular collapse of the plant from systematic invasion by the bacterium has been
described in California (Hildebrand et al., 1967).
The epidemiology of angular leaf spot is mostly unknown in fields in Florida where
strawberry production is an annual crop. Transplants are obtained each season from nurseries
in Canada and northern states; few transplants are produced in Florida. A source of inoculum
other than infected transplants has not been found Howard (1971) was unable to determine
the inoculum source for infected plants from nurseries in Florida or other states. Kennedy and
King (1962b) determined that bacteria overwintered in infected leaves buried in the soil and
caused disease symptoms on plants the next year. The bacterium did not survive free in the
soil nor were any naturally occurring hosts identified in host range studies (Kennedy and
8

9
King, 1962b). For bacteria on plant refuse to serve as an inoculum source in Florida, the
bacterium must oversummer. Optimal growth (=20 C) of the bacterium (Howard et al.,
1985; Kennedy and King, 1962b) is at temperatures cooler than normally occurs in Florida
during the summer. The bacterium survived on summer nursery plants in the field in Florida
but at very low populations (Roberts, unpublished) The effect of angular leaf spot on yield
is unknown. Howard (1971) accredited some yield losses due to angular leaf spot in fields
in Florida but he did not quantify the losses. In Wisconsin, a decrease in yield of 70 to 80%
was estimated due to the disease (Epstein, 1966). However, the production in the northern
United States is a perennial, matted-row system which differs significantly from the Florida
situation. A significant loss in marketable fruit may occur due to infections of the calyx. The
sepals become brown and dry and the fruit is unmarketable because of its unattractive
appearance (Epstein, 1966; Maas, 1995).
Chemical control of bacterial diseases on plants is difficult. Antibiotics have limited
effectiveness over time since mutations may form resistant populations (Stall and Thayer,
1962). Therefore copper compounds are frequently used to control a bacterial disease (Jones
et al., 1991). Antibiotics and copper compounds were effective protectants against angular
leaf spot but these did not eradicate the disease (Alippi et al., 1989). Marco and Stall (1983)
examined chemical control of strains of Xanthomonas campeslris pv. vesicatoria which
differed in sensitivity to copper. The mixture of cupric hydroxide plus mancozeb was more
effective than cupric hydroxide alone to control both the copper-resistant and the copper-
sensitive strain. The application of copper compounds to strawberry plants is confounded by
the fact that copper can be phytoxic to strawberry plants (Howard and Albregts, 1973). The

10
application of a chemical pesticide can reduce the infection by a pathogen and slow the rate
of the epidemic, e. g., van der Planks (1963) apparent infection rate (Fry et al., 1979; Pataky
and Lim, 1981). A reduced rate of fungicide applied at frequent intervals as a protectant has
been used effectively to reduce disease on crops (Fry 1975; Fry et al, 1979; Conway et al.,
1987).
Our objectives in this study were to examine epidemiological aspects of angular leaf
spot including the incidence of A*, fragariae on transplants from northern nurseries, the effect
of angular leaf spot on yield, and the control by chemicals.
Materials and Methods
Survey of farmers fields and storage facility. Disease incidence of angular leaf spot
on plants in fields located around Dover and Plant City, FL, was examined in October 1993.
Plants were assessed for symptoms of angular leaf spot within 2 to 4 days of transplanting.
Leaves from symptomatic plants were collected. A 9-mm diameter of leaf tissue was
removed surrounding a lesion and macerated in 200 /il of sterile water. A loopful of the
suspension was dilution streaked onto Wilbrinks medium (WB) (Koike, 1962) and colonies
of X. fragariae were identified. The pathogenicity of these isolated strains was tested on
Sweet Charlie strawberry plants. In 1994, transplants were sampled while in cold storage
at a facility in Dover, FL. Seven groups of plants, comprised of four cultivars from six
northern growers were sampled. Three boxes were selected at random from within a group
and 20 plants were removed from each box. The plants were examined for symptoms of

11
angular leaf spot and attempts to isolate the pathogen from putative lesions was done by the
method described above.
Inoculation. Three strains of A", fragariae (Xfl 13, Xfl03, and Xfl425) were used to
inoculate plants used in field plots. Strains Xfl 13 and Xfl03 were isolated from infected
plants at Gulf Coast Research and Education Center (GCREC), Dover, FL and Xfl 425 was
obtained from A. Chase (Central Florida Research and Education Center, Apopka, FL).
Strains were cultured on WB at 24 C and long-term storage was at -70 C in 15% glycerol.
The sensitivity of the strains to copper was tested by growth on nutrient agar amended with
CuS04 (Stall et al ., 1986). Three days prior to inoculation, each of the strains was streaked
to ten plates of WB. Bacterial cells were collected from plates, suspended in sterile 0.01 M
MgS04, and the concentration adjusted to approximately 107 cfu per ml. Equal volumes of
each suspension were combined to comprise the inoculum. Plants were inoculated by dipping
bundles of 25 plants into the bacterial suspension for 30 s. Control plants were dipped into
0.01 M MgS04. Plants were placed into plastic bags and incubated 24 h at 22 C.
Field experiments. Field experiments were located at GCREC, Dover, FL, from
October 1993 through March 1994 and repeated the following season. Transplants of'Sweet
Charlie' grown in the summer nursery at GCREC-Dover were used. Raised beds were
prepared and fertilized with 10N-4P-10K at a rate of 2000 lbs per acre with one-fourth
banded before bed preparation and the remainder banded 5 cm deep in the bed center the first
season. In the second season, three-fourths pound of N and K per acre per day were applied
through the drip irrigation system. Soil was fumigated with 98% methyl bromide and 2%

12
chloropicrin at 448 kg per acre. Beds were covered with 1 mm black polyethylene mulch
immediately after fumigation.
The experimental design was a randomized complete block in a 2 3 factorial design
with four replications. The first factor had two levels: plants inoculated with either the
suspension of X. fragariae or MgS04. The three levels of the second factor were: no
chemical treatment, the label (1 *) rate of cupric hydroxide (Kocide 101 at 9.08 kg of active
ingredient per acre) plus mancozeb (Dithane DF at 6.81 kg of active ingredient per acre)
sprayed at 7- to 14-day intervals, or the reduced (0.lx) rate of cupric hydroxide plus
mancozeb sprayed at 2- to 4-day intervals.
An individual plot contained 18 plants arranged in two rows of nine plants. The beds
were spaced on 1.22 meter centers with 30 cm between rows and 30 cm within a row. Fallow
area was 3.55 m within a row and 2.44 m between rows. Pesticides were applied throughout
season as needed to control insects and fungal diseases. Chemical applications were made
by a handheld wand attached to C02-charged canister at 40 psi and pesticide was applied to
runoff. Plants were transplanted on 15 October 1993 and on 20 October 1994. Overhead
sprinkler irrigation was applied 8 hours daily for 10 to 14 days to establish transplants and
applied throughout the season as needed. Drip irrigation was installed in the summer of 1994
at GCREC-Dover and was used as supplemental irrigation in 1995.
Estimates of disease severity of angular leaf spot were made at two-week intervals.
Disease severity was expressed as percent leaf area diseased for the entire plant for each of
six plants located in the center of each plot. Progress curves were plotted as the mean of
disease severity of replicate treatments versus time. Area under the disease progress curve

13
(AUDPC)(Shaner and Finney, 1977) was calculated for each plot and used in statistical
analysis. Statistical analysis was performed by orthogonal contrasts using PC-SAS.
Yield data. Fruit were harvested at 2- to 4-day intervals from initial bearing in
December through 30 March. Fruit were graded as marketable, culls, or nonmarketable due
to damage by fungi. Marketable fruit were those free of rot, not misshapen, and greater than
10 g in weight. Culls were non-marketable fruit due to physical imperfections such as small
size (< 10 g), damage by insects, or undesirable shape. Nonmarketable fruit was damaged
by fungal diseases, usually caused by anthracnose, botrytis, and phomopsis. The weight in
grams for each category was recorded for individual plots. Statistical analysis of the weight
of fruit in each category was performed using orthogonal contrasts using PC-SAS.
Results
Disease incidence of transplants in field and storage facility. In 1993, six of seven
fields contained plants with symptoms of angular leaf spot from which X. frcigariae was
isolated. In 1994, three of the seven groups sampled from the cold storage facility contained
plants with symptoms of angular leaf spot. In one of these groups, plants with symptoms of
angular leaf spot were found in all of the boxes sampled. In the other two samples, one box
contained plants which were symptomatic with angular leaf spot. Bacteria were isolated from
all plants with lesions of angular leaf spot and pathogenicity of these isolated strains on
strawberry was confirmed. Strains which were resistant to copper were not identified

14
Disease progress. Symptoms of angular leaf spot were not visible on plants at
planting in either 1994 or 1995. Disease progress curves of angular leaf spot on strawberry
plants for all treatments are presented in Figure 2-1. In 1994, the disease severity of angular
leaf spot on inoculated plants which did not receive a spray treatment increased for 89 days
after transplanting (DAT), decreased until 118 DAT, and then increased to 25% by 118 DAT
and leveled out to the end of the season. In 1995, the epidemic on plants inoculated with the
bacterium initially increased, then decreased at 61 DAT until 111 DAT, and then increased
and reached 15% by 154 DAT. The disease on inoculated plants sprayed with either the 1 x
or 0.1 x rate of cupric hydroxide plus mancozeb followed similar disease progress curves
except the amount of disease was reduced for each treatment. In contrast, the onset of the
epidemic on the noninoculated plants was greatly delayed. Disease was observed on these
plants after 118 DAT in 1994 and at 125 DAT in 1995. The noninoculated plants had very
little disease (0-3% severity) by the end of the season, regardless of spray treatment.
Statistical analysis of the AUDPC for each treatment is shown in Figure 2-2.
Noninoculated plants had significantly (P = 0.0001) lower disease than plants inoculated with
the bacterium. Noninoculated plants treated with the 1 rate did not have disease either year.
Inoculated plants sprayed with the lx rate of cupric hydroxide and mancozeb had a
significantly (P = 0.0004 in 1994, P 0.0001 in 1995) lower AUDPC compared to the
control treatment in both years. The plants sprayed with the 0.1 x rate of the fungicide had
reduced disease in 1994 (P = 0.17) and, in 1995, the reduction in AUDPC was highly
significant (P = 0.0001) compared to plants without chemical sprays.

% Disease % Disease
15
7 21 7 20 5 8 22 8 23
Nov Dec Jan Feb March
Figure 2-1. Disease progress curves of Xcmthomoms fragariae on strawberry in Florida in
the 1994 (A) and 1995 (B) seasons. The mean of disease severity for four replicate blocks
are plotted against days after transplanting (DAT).

16
Noninoculated
Inoculated
NoS 0.1X IX NoS 0.1X IX
Figure 2-2. Area under the disease progress curve (AUDPC) of Xanthomonas fragariae on
strawberry in the 1994 (A) and 1995 (B) seasons and the effect of applications of cupric
hydroxide plus mancozeb at the 1 x rate at 7-to 14-day intervals or at 0.1 x (10%) rate of the
chemical combination applied at 2-to 4-day intervals or no spray (NoS). A, B = significant
difference (P = 0.05) within the first factor of inoculated vs. noninoculated plants; a, b =
significant difference within the second factor of the spray treatments.

17
Yield loss. In both years, marketable yield on plants infected with angular leaf spot
was significantly decreased (P = 0.005 in 1994 and P = 0.04 in 1995) compared to yield on
non-inoculated plants (Fig. 2-3). Total and marketable yields were reduced by an average of
7% and 8%, respectively, in 1994 and by 6% and 10% in 1995. Total yield had similar
reductions as marketable yield, therefore only marketable yield is presented. Yields were
generally lower in 1995 than in 1994.
The effect of spray treatments on the yield from inoculated and noninoculated plants
is presented in Figure 2-4. Marketable yield on noninoculated plants sprayed with the 1*
treatment of cupric hydroxide plus mancozeb was significantly reduced (P = 0.04) compared
to plants which received the 0.1 x or no-spray treatments in 1994. No difference in yield was
detected on inoculated plants receiving the spray treatments in 1994. In 1994, the yield on
inoculated plants sprayed with the 1 x rate was significantly reduced (P = 0.005) compared
to the other two treatments. Yield on noninoculated plants sprayed with the lx and 0. lx
chemical sprays was reduced significantly (P =0.04) compared to yield on nonsprayed plants
in 1995. The spray treatments did not have any effect on culled fruit in 1994 or 1995 (Fig.
2-5). No difference in culled fruit was detected among spray treatments. Nonmarketable fruit
lost because of damage by fungal diseases was significantly different at P = 0.01 for all
treatments in 1994 but the relationship to spray treatments was unclear (Fig. 2-6). In 1995,
there was no significant difference in damage to fruit by fungi for either inoculated or spray
treatments. Overall, the total yield lost to fungal infections was much lower in 1994
compared to 1995.

18
Figure 2-3. The average marketable yield on strawberry plants which were inoculated with
Xauthomonasfragariae (Inoc) or not inoculated (Noninoc) for the 1994 and 1995 seasons.
Yield was the average for four replications for marketable fruit only. Different letters (A, B)
represent significant difference at P = 0.05.

19
Noninoc Inoc
Noninoc Inoc
1994
1995
Figure 2-4. The marketable yield on strawberry in Florida in the 1994 (A) and 1995 (B)
seasons. A, B = F was significant at P = 0.05 within the first factor of noninoculated
vs.inoculated; a, b = F was significant at P = 0.05 within the second factor of spray
treatments.

Yield (KG) Yield (KG)
20
1.2
Noninoculated Inoculated
NoS 0.1X IX
NoS 0.1X IX
Noninoculated
Inoculated
NoS 0.1X IX
NoS 0.1X IX
Figure 2-5. The culled fruit on strawberry in Florida in the 1994 (A) and 1995 (B) seasons.
A, B = F was significant at P = 0.05 within the first factor of noninoculated vs.inoculated; a,
b = F was significant at P = 0.05 within the second factor of spray treatments.

21
Noninoculated
Inoculated
NoS 0.1X IX
NoS 0.1X IX
Figure 2-6. The yield lost to fruit damaged by fungal diseases on strawberry in Florida in the
1994 (A) and 1995 (B) seasons. A, B = F was significant at P = 0.05 within the first factor
of noninoculated vs.inoculated; a, b, c = F'was significant at P = 0.05 within the second factor
of spray treatments.

22
Discussion
A decrease in yield of strawberry fruit was observed both seasons due to angular leaf
spot. This report is the first to quantify the reduction in yield due to this disease. The 8 to
10% loss determined in these studies is much lower than the 70 to 80% loss estimated by
Epstein (1966) in fruit production fields in Wisconsin. Production in northern regions of the
U. S. is usually perennial and the age of the plants was not given. The loss was reported on
the cultivar Sparkle which is not produced commercially in Florida. Therefore, we are
unable to compare our yield loss to his report. Howard (1971) also reported unquantified
yield losses due to this disease but on cultivars which are no longer in commercial production.
In our studies, the average yield loss observed both years was very similar despite the
differences in the disease levels and the base yields between the two seasons. In 1995, the
disease severity was 60% of the amount estimated in the previous season; however, the yield
loss was actually higher by 2% for marketable yield. The 10% reduction in yield represents
an estimated loss of 1233 kg of berries per ha. Strawberry fruit has a very high cash value
and a decrease in revenue of 8 to 10% represents a significant economic loss to producers.
Strawberry producers in Florida currently try to control angular leaf spot by
application of copper compounds and by avoidance of the disease. An understanding of the
epidemiology of angular leaf spot is necessary to determine the effectiveness and practicability
of any control method. The survey of farmers fields and the cold storage facility established
that transplants arrive from northern nurseries infected with angular leaf spot at fairly high
disease incidence. Therefore, disease is introduced into the field on infected transplants. In

23
our surveys by the sampling method used in the first season, more plants were assessed for
symptoms of angular leaf spot in the field compared to the number of plants in the random
sampling of boxes in cold storage. This probably accounts for identification of diseased plants
in nearly all of the farmers fields versus less than in 25% of boxed plants. In addition, plants
were examined until disease was found in the field which skewed the randomness of the
survey.
In field experiments, the spread from inoculated plants to noninoculated plants in
nearby experimental plots was minimal. Thus, the incidence of plants with angular leaf spot
in the field likely occurred from inoculum already present on the transplants. In the field
experiments, if inoculum was present from another source, such as debris in the soil or an
alternate host, most likely the disease on the noninoculated plants would have been more
general and appeared earlier. Disease severity on noninoculated plants in the experimental
plots was extremely low and most plants remained free of angular leaf spot. For the disease
that occurred on noninoculated plants, the inoculum was probably transported mechanically
from inoculated plants during harvests or disease readings.
Progress of angular leaf spot developed similar to curves of pathogens described on
other crops (van der Plank, 1963). A decrease in disease severity was seen about midway
through the season in 1994. This reduction, or negative infection rate, was due to growth
by strawberry plants which diluted the amount of disease relative to the total leaf area The
dilution effect of healthy plant tissue was described by van der Plank (1963). In 1995, a
negative infection rate was seen very early in the epidemic. This was because the symptoms
of the disease almost disappeared from the field as the affected leaves died and were lost. The

24
difference in disease progress between the two seasons did not appear to be from differences
in rainfall since mean rainfall was approximately the same both years (data not shown). Mean
temperatures also were similar; however; the number of days at temperatures below 10 C
in November and December 1993 were much greater than in 1994. Another difference
between the seasons was that sprinkler irrigation was reduced in 1995 because of the change
to drip irrigation. Overhead irrigation, fog, high humidity, and rain are positive factors in
disease development and spread (Epstein, 1966; Hildebrand et al., 1967; Kennedy and King,
1962b; Lai, 1978). The trend towards drip irrigation in commercial production fields should
have a positive effect to reduce the spread and survival of the bacterium.
Strawberry producers currently apply copper-based compounds to control angular leaf
spot despite the phytotoxicity to plants caused by copper. Application of cupric hydroxide
plus mancozeb at the 1* rate was phytotoxic to strawberry plants (Howard, 1973; this study).
In our tests, the greatly reduced 0.1 x rate did not harm plants and significantly reduced
disease severity in 1995. The protectant application of bactericide at this rate and spray
schedule was intended to reduce the total amount of inoculum and to prevent spread of the
disease. The approach of frequent sprays at reduced concentrations may have potential to
control this disease. More studies are needed to evaluate different rates, application intervals,
and chemical mixtures to achieve maximum disease control while avoiding yield losses. An
alternate approach might be to control the disease in the nursery by copper applications. The
loss in yield due to phytotoxicity would not be an issue on nursery plants. The sprays would
reduce the amount of initial inoculum on transplants and subsequently reduce the amount of
disease in fruit production fields.

CHAPTER 3
DEVELOPMENT OF A SPECIFIC AND SENSITIVE DETECTION TECHNIQUE
BY THE POLYMERASE CHAIN REACTION FOR XANTHOMONAS FRAGAR1AE
AND APPLICATION IN A STUDY OF SURVIVAL OF THE BACTERIUM ON
STRAWBERRY PLANTS
Introduction
Angular leaf spot of strawberry (,Fragaria x ananassa Duchesne), caused by the
bacterium Xanthomonas fragariae, was first reported in Minnesota in 1960 and is now found
in many areas of strawberry production throughout the world (Kennedy and King, 1960a;
Maas, 1984; Ritchie et al., 1993). The disease is apparently disseminated by the
transportation of infected plants (Maas, 1984). In Florida, strawberry plants which arrive
from northern nurseries for transplanting in the fall frequently have leaves which exhibit
symptoms of angular leaf spot. Typical pathogenic strains of X. fragariae can be isolated
from the lesions (P. D. Roberts, unpublished).
The epidemiology of X fragariae is mostly unknown in fields of strawberry in Florida
where production is from annual crops. Angular leaf spot was first reported in the state in
1971 (Howard, 1971). Howard (1971) was unable to determine the inoculum source for
infected plants from nurseries in Florida or other states. Bacteria may survive on infested
leaves in the soil (Kennedy and King, 1962b), but plants in Florida are usually treated with
the herbicide paraquat at the end of the season and removed. The bacterium does not survive
25

26
freely in the soil (Kennedy and King, 1962b). Cool temperatures (~ 20 C) are optimal for
disease symptom expression (Howard et al., 1985; Kennedy and King, 1962b) and high
temperatures (>28 C) such as those which occur in Florida during the summer months are
unfavorable. In surveys conducted in 1968, 1970, 1993, and 1994, plants which had
symptoms of angular leaf spot in the spring did not have symptoms of the disease the
following August (Howard, 1971; P. D. Roberts, unpublished). However, in 1969 mild
infections on one variety were observed in mid-August. Plants transplanted to fields did not
develop angular leaf spot. The survival of the bacterium on plants in summer nurseries in
Florida and inoculum sources other than infected transplants have not been investigated.
Identification of plants infected with X fragariae is a priority because of the ease of
transmission on infected but asymptomatic plants (Maas, 1995). International movement of
infected plants is blamed for introduction of angular leaf spot into Greece and New Zealand
(Panagopoulos et al., 1978; Dye and Wilkie, 1973). Nursery-plant producers are pressured
to provide disease-free plants by buyers in foreign countries and by farmers who refuse to buy
infected transplants. The European Plant Protection Organization (EPPO) lists X. fragariae
as a quarantine pest and has prescribed phytosanitary procedures (Maas, 1995). In the future,
regulatory issues may be of greater concern. The production of disease-free plants is essential
for control of angular leaf spot. Therefore, accurate identification of plants infected with the
bacterium is imperative. Available detection techniques are limited in their usefulness and
accuracy to detect low populations of the bacterium that may exist in asymptomatic tissue.
Xanthomonas fragariae may be identified in the early stages of leaf infection by diagnostic
translucent, watersoaked lesions viewed with transmitted light; however, older lesions may

27
be confused with symptoms caused by fungal pathogens (Kennedy and King, 1962b).
Diagnosis based on symptoms is very difficult and not applicable for asymptomatic plants.
Identification of the disease based upon isolation and characterization of the causal agent may
also be difficult because X. fragariae grows slowly and may be masked by faster growing
organisms (Kennedy and King, 1962a). Expression of watersoaked lesions takes 6 days or
longer after inoculation. Thus, fulfillment of Kochs postulates to confirm pathogenicity is
difficult and time-consuming.
Assays have been developed with improved sensitivity and specificity for the detection
of plant pathogens in plant tissue. A specific, indirect, enzyme-linked immunosorbent assay
(ELISA) was developed to detect X. fragariae from symptomatic plant tissue (Rowhani et
al., 1994). The antibody assay did not react with bacterial strains of 16 pathovars of X
campestris or non-pathogenic bacteria from strawberry leaves. A single cross-reaction
occurred to a strain of X. campestris isolated from Nerium oleander. The assay detected
bacteria directly from a visible lesion on a strawberry leaf. The level of sensitivity was at ca.
104 colony forming units (cfu) per ml. However, bacteria on asymptomatic plants may not
be detected by this ELISA assay.
The polymerase chain reaction (PCR) has been used to amplify specific DNA
sequences to detect and identify many plant pathogens including some members of the genus
Xanthomonas (Henson and French, 1993; Louws et al., 1994). Leite et al. (1994) utilized
primers specific to regions of the hrp gene cluster which confers hypersensitivity and
pathogenicity from Xanthomonas campestris pv. vesicatoria to detect pathogenic
xanthomonads. Saprophytic and non-pathogenic bacteria lack the hrp genes (Stall and

28
Minsavage, 1990; Lindgren et al., 1986 ) and DNA was not amplified by the hrp gene cluster
primers. Differentiation of X. campestris pathovars was made by restriction endonuclease
analysis (REA) patterns generated by digestion of the PCR products with frequent-cutting
enzymes (Leite et al., 1994 and 1995). The primers were used to detect X. c. vesicatoria in
seed lots of naturally infected pepper and tomato (Leite et al., 1995). The sensitivity of
detection by PCR is reported at 103 to 102 cfu per ml (Minsavage et al., 1994; Leite et al.,
1994; Henson and French, 1993). In nested PCR, sensitivity of detection is increased by
using PCR products from an amplification as target DNA in a second round of amplification
by a second set of primers internal to the first (Schaad et al., 1993). McManus and Jones
(1995) reported an increase in sensitivity of 1,000 -fold with nested PCR over a single round
of PCR amplification to detect Erwinia amylovora.
Our objectives were to develop a sensitive and specific technique for detection of X
fragariae. Our approach was to design primers specific to the region of genomic DNA from
X. fragariae related to the hrp genes of X c. vesicatoria. The survival of X. fragariae on
nursery strawberry plants in the field at two locations in Florida and dissemination to daughter
plants was examined to understand the disease cycle of angular leaf spot in Florida
Materials and Methods
Bacterial strains and culture conditions. Strains of X fragariae and non-pathogenic
xanthomonads isolated from strawberry were maintained at 24 C on Wilbrink's medium
(Koike, 1965). Pathovars of Xanthomonas campestris were cultured on nutrient agar (Difco

29
Laboratories, Detroit, MI) and incubated at 28 C. Long term storage was at -70 C in 15%
glycerol. Bacteria used for plant inoculations and DNA extractions were grown in 5 ml
nutrient broth on a rotary shaker at 200 rpm for 16 h at 24 C. A rifampicin-resistant mutant
of strain XF1425 was selected on Wilbrinks medium supplemented with 100 ,ug/ml of
rifampicin by the gradient plate technique (Szybalski, 1952).
Pathogenicity tests. Bacteria from overnight cultures in nutrient broth were
centrifuged and washed three times with sterile water. The concentration of cells was
adjusted in either 10 mM MgSO4-7H20 or sterile water to approximately 105 cfu per ml and
sprayed to runoff on 'Sweet Charlie strawberry plants placed under mist 24- to 48- h prior
to inoculation. Inoculated plants were maintained under mist or put into growth chambers
(Percival, Boone, IA) at 24 C with a 12 h photoperiod.
Sequencing and primer design. The hrp primers RST2 and RST3 from the hrp gene
cluster of X. axonopodis pv. vesicatoria (Leite et al, 1994) were used to amplify genomic
DNA of 49 strains oX. fragariae. The PCR product from strain XF1425 of X. fragariae
was isolated from agarose gel, cleaned by the Promega Wizard Kit (Promega, Madison, WI)
and sequenced at the ICBR DNA Sequencing Facility, University of Florida, Gainesville, FL.
The nucleotide sequence was compared to sequences of the PCR products amplified by the
same primers from one X vesicatoria and two X a. pv. vesicatoria strains using the Seqaid
II computer program (Rhoads and Roufa, 1991). Four primers were selected from the
sequence of X. fragariae based upon unique DNA sequences and low homology compared
to the DNA sequences from the other bacterial strains. The four oligonucleotide primers

30
were synthesized with a model 394 DNA synthesizer (Applied Biosystems, Foster City, CA)
at the ICBR Facility, University of Florida, Gainesville, FL.
PCR amplification and nested amplification. Total genomic DNA was extracted by
the method described by Ausubel et al. (1987). PCR amplification was performed using a
DNA Thermal Controller PT-100 (MJ Research, Watertown, MA). Samples were in a total
reaction volume of 50 /ft and contained IX amplification buffer (Promega, Madison, WI), 100
M of each dNTP (Promega, Madison, WI), 50 /M of each primer, 1.25 U Taq DNA
polymerase, and 100 ng of purified genomic DNA in 3 /ft of TE (10 mM Tris and 1 mM
EDTA, pH 8.0) buffer. Each reaction was overlaid with 50 /ft of sterilized mineral oil
(Sigma) for a total volume of 100 /ft in sterile 0.6 ml microcentrifuge tubes. In later
experiments, the thermocycler was equipped with a heated lid controller (The Hot Bonnet,
MJ Research) which eliminated the need for the mineral oil overlay. Amplification of the
DNA proceeded after template DNA was denatured at 95 C for 2 min followed by thirty
amplification cycles and a final extension step at 72 C for 5 min. For primer set XF9 and
XF11, each amplification cycle consisted of denaturation at 95 C for 30 s, annealing at 65
C for 30 s, and extension at 72 C for 45 s. For primer XF12 with either primer XF9 or
XF10, the program was identical except the annealing temperature was 58 C.
For nested PCR, the first round of amplification was as described with primers XF9
and XF11. In the second round of amplification, a 3 ^ul sample from the first amplification
mixture was used with primers XF9 and XF12 and all other ingredients were added at the
concentrations described above. The PCR cycle program for the primers XF9 and XF12 was

31
used. In all PCR runs, including the nested assays, a water sample was used as a negative
control.
Restriction endonuclease analysis (REA) of PCR products. The PCR samples with
the overlay of mineral oil were cleaned by the method of Minsavage et al (1994). Samples
without the mineral oil were used directly in restriction reactions. An 8 /I sample of the PCR
product was digested with restriction endonuclease Sau3Al, Haelll or Cfol under conditions
specified by the manufacturer (Promega). Restricted products were separated in 4% agarose
gel (3% NuSieve, 1% SeaKem GTG, FMC BioProducts, Rockland, ME) containing 0.5
/g/ml ethidium bromide in TAE buffer at 8 V/cm as described by Leite et al. (1994). DNA
molecular weight marker XI (Boehringer Mannheim, Indianapolis, IN) was used for standard
weight markers. Gels were photographed over a UV transilluminator with type 55 Polaroid
film (Polaroid Corp., Cambridge, MA).
Detection of bacteria from infected plant tissue. The sensitivity of the assay in the
presence of plant tissue was determined by adding known concentrations of bacteria to plant
samples. A 9-mm diameter disk of plant tissue was macerated with mortar and pestle in 200
yul of phosphate buffer (pH 7.0) containing 5% polyvinylpolypyrrolidone (PVP40, Sigma
Chemical Co., St. Louis, MO) and 0.02 M sodium ascorbate (PPA). The mixture was
incubated at room temperature for a minimum of 1 h. The volume was adjusted to 582 //I
with TE and DNA extraction proceeded as described above. A minimum of three
experiments with three replications of each treatment for each primer set and nested reaction
was performed. REA was performed on the final PCR products.

32
Strawberry plants with leaves exhibiting typical lesions resulting either from natural
field infections or spray inoculation were assayed by removing a 9-mm diameter disk of tissue
surrounding a lesion and proceeding as described above. Tomato leaf tissue was used as a
negative control. For confirmation, a sample of the ground tissue suspension was plated on
Wilbrinks medium amended with 0.05% Bravo 720 (chlorothalonil; ISK Biosciences,
Mentor, OH). Plates were incubated at 24 C and colonies of A", fragariae were identified
by colony morphology.
Field experiments to examine oversummering survival of bacteria on nursery plants.
Six-week-old rooted transplants ofSweet Charlie were spray inoculated with X fragariae
strain XF1425lif two weeks prior to planting. All plants transplanted into the field exhibited
lesions of angular leaf spot. Plants were transplanted into the field on 27 June 1995 at Gulf
Coast Research and Education Center-Dover, FL and on 29 June 1995 at Gulf Coast
Research and Education Center-Bradenton, FL. Plants were sampled over a 14-week period
through 27 September 1995. Another set of inoculated plants were placed into Percival
growth chambers at 12 h photoperiod and 24 C and maintained for the duration of the
experiment. Plants which were not inoculated and did not exhibit symptoms of angular leaf
spot were placed in a greenhouse. At two-week intervals, five plants from the field, one plant
from the growth chamber, and one plant from the greenhouse were sampled. Two samples
of three daughter plants of the inoculated plants were sampled at 132 days after planting from
both locations. Individual plants were assayed as follows. All the leaves from a single plant
were removed and placed in a flask with 200 ml of phosphate buffer containing 0.02 %
Tween 20. The samples were shaken either on a wrist action (Burrell Corp., Pittsburgh, PA)

33
or a rocker platform (Blico Biotechnology, Vineland, NJ) for 2 to 16 h. A 200 /I sample
was plated onto Wilbrinks medium plus 100 /g/ml rifampicin plus 0.5% Bravo 720 and
incubated at 24 C for 72 h. Colonies characteristic of X. fragariae were identified by
amplification with the XF-specific primers and by REA. The remainder of the phosphate
buffer sample was concentrated by vacuum filtration onto a 0.45 m membrane disk
(Millipore Corp, Bedford, MA). The disk was washed in 1.5 ml of TE and the suspension
centrifuged 5 min at 14, 000 x g. The pellet was resuspended in 582 ^1 TEPA (TE buffer
containing 5% PVP40 + 0.02 M sodium ascorbate) and incubated at room temperature 1 h.
DNA was extracted as described above. The crown of the plant was sectioned, macerated
by mortar and pestle in 10 ml of PPA. The plant tissue debris and PVPP were collected by
centrifugation at 1,000 x g for 1 min. The supernatant with the bacterial cells was removed
to a clean centrifuge tube and centrifuged at 14,000 x g for 5 min. The pellet was
resuspended in 582 TE and DNA extraction proceeded as described above. A sample was
plated onto Wilbrinks agar plus 100 /ug/ml rifampicin and 0.5% Bravo and incubated at 24
C for 72 h. The nested PCR reaction and REA were performed as described above.
Results
Pathogenicity assays. All strains identified as X. fragariae by colony characteristics
caused disease symptoms typical of angular leaf spot on strawberry plants. Many other
xanthomonads were isolated from strawberry tissue but they did not cause symptoms of
angular leaf spot

34
Specificity of primers. For all strains of X. fragariae, the PCR products amplified by
primers RST2 and RST3 were ca. 840-bp (Fig. 3-1). The REA profiles that resulted from
restriction of the PCR products with Cfo\ or HaeIII were the same for all the strains of X.
fragariae (data not shown). Genomic DNA from non-pathogenic strains of Xanthomonas
isolated from strawberry was not amplified by the primers
The four primers synthesized were: XF9 (5' TGGGCCATGCCGGTGGAACTGT
GTGG3'); XF10 (5' TGGAACTGTGTGGCGAGCCAG 3); XF11 ('5 TACCCAGCCGT
CGCAGACGACCGG 3'); and XF12 (5' TCCCAGCAACCCAGATCCG 3'). Primers XF10
and XF12 were internal to the other two primers
Primer XF9 paired with XF11 or XF12 delineated a 537-bp or a 458-bp fragment,
respectively (Fig. 3-1). Primers XF10 and XF12 delineated a 448-bp fragment. PCR
products conformed to the estimated sizes based upon sequence data. All 49 strains of X.
fragariae were amplified with primer sets XF9 and XF11, XF9 and XF12, and XF10 and
XF12. DNA from strains other than X. fragariae was not amplified by primer XF9 paired
with XF11 or XF12. Strains tested were ATCC type strains of Xanthomonas campestris
pathovars begoniae, campestris, carotae, celebensis, glycines, incanae, manihotis,
musacearum, papavericola, pelargonii, phaseoli, poinsettiicola, raphani, taraxaci,
vignicola, vitians, and nine non-pathogenic strains of xanthomonads isolated from strawberry.
Except for X. c. pelargonii, genomic DNA from these strains could not be amplified with
primer set XF10 and XF12. Due to the non-specific amplification of DNA from X. c.
pelargonii by primer XF10 paired with XF12, only primer XF9 paired with either XF11 or

35
12 3 4
Figure 3-1. Agarose gel of the PCR products generated by amplification of genomic DNA
from a strain ofXcmthonionas fragariae with the /wp-primers RST2 and RST3 (Lane 1) and
the XF-specific primers XF9 paired with XF1 l(Lane 2) and XF9 paired with XF12 (Lane 3),
and molecular weight marker XI (Boehringher-Manheim, Indianapolis, IN; Lane 4).

36
XF12 was used in experiments to detect DNA from X fragariae in the presence of plant
materials.
REA of PCR products. PCR products from DNA of X. fragariae amplified with
primer set XF9 and XF12 restricted with Cfol, SaulM, or HaeIII produced distinct banding
patterns for each enzyme (Fig. 3-2). Five bands resulted from Cfo\ restriction (125, 108, 90,
75, and 60 bp), and three bands each from the Hae III restriction (258, 150, and 50 bp) and
Sau3M restriction (250, 125, and 83 bp). Polymorphisms were not observed among strains
of X. fragariae in the REA of the restricted PCR products for the enzymes tested
Sensitivity of detection of X. fragariae in plant tissue bv PCR. X. fragariae was
detected in the presence of plant tissue by amplification with primer XF9 and either XF11
or XF12 and by nested PCR of the two primer sets. The level of sensitivity of detection of
bacterial cells by amplification with one round of PCR was ca. 105 to 104 cfii per ml. The
nested PCR increased sensitivity of the assay 1000-fold (Fig. 3-3) and enabled detection of
ca. 18 cfu per ml. PCR was used to detect bacteria directly from infected plant tissue both
from inoculated strawberry and from plant samples collected in the field. The PCR product
was generated from a single lesion on plant material ground and processed representing a
bacterial population of 104 or above as determined by dilution plating. Identification of X.
fragariae was confirmed by REA of PCR products. The nested technique was not necessary
to detect bacteria from a lesion on leaf tissue as a single round of PCR amplification was
sufficient to obtain a positive result.

37
100 bp
Figure 3-2. NuSieve agarose gel showing the restriction patterns generated after
digestion of the PCR product amplified with the primers XF9 and XF12 from the genomic
DNA of strains of Xanthomonas fragariae. The PCR product was digested by restriction
endonucleases Cfo\ and HaelU and the molecular weight marker XI from Boehringer
Mannheim is shown.

38
XF9 and XF11 Nested XF9 and XF12
500 bp
87654321 87654321
Log cfu per ml
Figure 3-3. Agarose gel showing the sensitivity of detection by PCR and nested PCR in
the presence of plant tissue. The genomic DNA from various concentrations of cells was
extracted and amplified by primers XF9 and XF11 and used in a second round of nested
amplification with primers XF9 and XF12.

39
Detection of bacteria on plants in oversummering experiments. All plants exhibited
visible symptoms of angular leaf spot at planting. However, by 34 days after planting,
symptoms of angular leaf spot were not visible on plants (Table 3-1). Bacteria were
recovered on the selective medium and colonies of X. fragariae were identified by
morphology, resistance to rifampicin, DNA amplification by the XF-specific primers, and
REA. Bacteria were recovered on selective medium from a few samples throughout the
experiment, including the final date, indicating that bacteria were viable on plants throughout
the summer.
Xanthomonas fragariae was detected from leaf samples by nested PCR on each
sample date. The percentage of leaf samples which were positive by the nested PCR assay
ranged from 20% to 100% for the sample dates. From the leaves, all samples were initially
positive but, by 51 days after planting, the number of positive samples declined to 20%. By
92 days after planting, all samples were again positive. A single round of amplification was
occasionally sufficient to identify positive samples. In the crown of plants, bacteria were
detected by PCR amplification on all sample dates except one. REA analysis of the PCR
products confirmed banding patterns typical of X. fragariae. Plants inoculated and placed in
the growth chamber at the favorable temperatures for disease development were positive at
each sampling date except one. Non-inoculated plants from the greenhouse were positive by
this technique in approximately 20% of the samples.

Table 3-1. Percentage of strawberry plants positive for the presence of Xanthomonas fragariae detected by PCR and a selective
medium in summer nurseries from June through September at two locations in Florida
Leaves
Crown
Location
Dover
Bradenton
DAPa NonIb Inoc" Primer ld Primed" Medium1 Symptoms NonI Inoc Primerl Primer2 Medium
0
-
+
60
100
TC
+
-
+
20
40
TC
11
-
+
0
40
TC
+
-
+
0
40
TC
31
-
+
0
20
0
+
-
+
0
40
0
40
+
+
0
20
NS
-
-
+
0
0
NS
66
-
-
0
60
0
-
-
+
0
40
0
85
-
+
20
100
60
-
-
+
20
100
20
92
-
+
40
100
40
-
-
+
0
80
40
0
-
+
60
100
TC
+
_
+
0
40
TC
20
-
+
20
20
20
+
+
+
0
60
0
34
+
f
0
20
NS
-
-
+
0
40
NS
40
-
+
0
20
20
-
-
+
0
100
20
54
-
+
0
40
0
-
-
+
0
40
0
68
+
+
20
100
60
-
+
+
20
80
20
82
-
+
20
80
40
-
4-
+
20
80
40
a Days after planting.
bNon-inoculated plants in greenhouse. = Negative by PCR amplification, + = Positive by PCR amplification.
c Inoculated strawberry plants in growth chamber at 22 C. = Negative by PCR amplification, + = Positive by PCR amplification.
d Percentage of samples amplified by primer set XF9 and XF11.
"Percentage of samples amplified in nested PCR with primers XF9 and XF12 with sample from first round.
Wilbrink's agar plus 100 mg/ml rifampicin and 0.05% Bravo; TC= too contaminated to distinguish individual colonies;
NS= Not assayed, 0 100 = Percentage of plates with colonies of A", fragariae.

41
Discussion
Leite et al. (1994) used the sequence variation within the hrpB operon among plant-
pathogenic xanthomonads to select primers with different specificities and this approach was
successful in our studies to identify primers specific to X fragariae. The hrp-primers RST2
and RST3 amplified DNA from all strains of X fragariae and analysis of the PCR products
by REA showed no polymorphism within this region. The homology of the amplified region
presented a good site to select primers universal to X fragariae. Primers XF9, XF11, and
XF12 designed from these unique sites were specific for amplification of DNA only from
strains ofA^ fragariae. The primer XF10 was responsible for the non-specific amplification
of a strain of X. c. pelargonii in preliminary tests and therefore was not used in later
experiments. Interestingly, an ELISA test developed for identification of X fragariae in vitro
cross reacts to some related bacteria including X. c. pelargonii (Maas, 1995). Perhaps a
relationship between X. fragariae and X. c. pelargonii exists which is not reflected by
traditional taxonomic techniques (Hildebrand et al., 1990; Vauterin et al., 1994; Maas, 1995).
The level of sensitivity for most detection techniques for specific bacteria (Pickup,
1991; Schaad et al., 1982 ) are comparable to the level which was achieved with a single
round of PCR amplification. PCR was reported to be more sensitive than ELISA (Leite et
al., 1995) and with our primers, a single round of PCR is 10 times more sensitive than the
ELISA test for A", fragariae developed by Rowhani et al. (1994). In addition, their antibodies
cross-reacted with another xanthomonad whereas we were able to eliminate the non-specific
reaction by discontinuing use of primer XF10. Although this level of detection is adequate

42
to detect bacteria in symptomatic leaf tissue, we were concerned with the detection of
bacteria from asymptomatic tissue. The X. fragariae primers were suitable for use in the
nested technique and the sensitivity of the PCR reaction was increased by 1,000-fold This
level of detection was achieved in assays to detect the bacterium in the presence of plant
tissue and is well below the number of cells needed to cause visible lesions. Therefore, the
nested technique is applicable for detection of bacteria in association with asymptomatic
tissue.
Cross-contamination among samples and contamination of PCR reagents are problems
associated with the nested technique (McManus and Jones, 1995). We experienced false
positives in a number of initial experiments and finally determined the cause to be
contaminated mineral oil. Although we eliminated the need for mineral oil overlay by using
a hot bonnet, extreme care is needed to avoid contamination and subsequent false positives.
In our experiments, negative control samples were always included for each amplification
round of the nested assay. Aerosols may also contribute to false positives and care must be
taken to prepare the mixtures for PCR in a sterile environment and negative control samples
should be included to check for aerosol contamination. In the field experiments, a problem
encountered with the nested technique was amplification of bacterial DNA from the negative
control plants. Although these plants were physically isolated from a contamination source
during the experiments; they were initially obtained from fields at GCREC-Dover where
plants infected with angular leaf spot were located. The plants may have been infested with
the bacterium or cross-contamination of samples may have occurred during preparation of
plant samples from handling infected material or using non-sterile instruments. Aerosols may

43
also have been a source of contamination of PCR samples (Innis et al., 1990). However,
contamination from these sources would be detected by amplification in the negative controls.
Our negative controls were included in every PCR run and were consistently negative.
Therefore contamination of PCR reagents or aerosols would have been ruled out.
Confirmation that the PCR product is from amplification of the target DNA is possible
by REA (Leite et al. 1994). The REA was used to identify PCR products amplified by the
/vp-primers from cells of X. c. vesicatoria added to seed washing of tomato and pepper (Leite
et al. 1995). Similarly, we applied REA to identify PCR products amplified by the XF-
primers. The profiles of the restricted PCR products were distinct for each enzyme. The
profiles from REA should be different for unrelated organisms since it would be highly
unlikely that the same restriction sites would exist for two heterologous pieces of DNA (Leite
et al,. 1995).
The nested PCR technique was useful to detect bacteria on strawberry plants in
nurseries in Florida Symptoms and recovery of the rifampicin-marked strain on a selective
medium were not useful to identify bacteria when populations were extremely low. However,
recovery of the rifampicin-marked strain from some samples indicated that the bacteria were
viable throughout the summer and that PCR was detecting viable bacterial cells. Visible
symptoms on plants disappeared soon after placement in the field and recovery on media was
difficult due to the slow-growing nature of X. fragariae and overgrowth by contaminants.
While bacterial populations were not enumerated, the levels were deduced to be below 103
cfu per ml, since bacteria were detected in the nested PCR but not by single round of
amplification. This level of sensitivity would be useful to screen asymptomatic nursery plants

44
to detect plants contaminated with X fragariae. This is a concern of nursery-plant producers
and regulatory agencies of plant shipment (Maas, 1995). Currently, plants from nurseries in
California are being screened by this technique to determine its usefulness in such an
application.
The field studies have implications regarding the disease cycle of angular leaf spot in
Florida. The decline and later increase in the number of positive samples through the summer
would indicate that populations of the bacteria declined throughout the summer and increased
when more favorable conditions in the form of cooler weather occurred. The bacterial
populations did not completely die. Therefore, to eliminate disease, the production of plants
in nurseries in Florida would have to begin with plants free of the bacterium.
Researchers have reported the systemic movement of X. fragariae in plants
(Hildebrand et al., 1967; Milholland et al., 1993). In our research, inoculation of strawberry
plants resulted in survival of bacteria on leaves and the crown for extended periods under
conditions not optimum for growth of the bacteria. In addition, bacteria were detected by
PCR technique on daughter plants in the field. Dissemination to the daughter plants could
have been due to either systemic movement through the vascular system of the runner or
dispersal of bacterium by mechanical means.

CHAPTER 4
GENOMIC RELATEDNESS OF XANTHOMONAS FRAGARIAE ON STRAWBERRY
BY FATTY ACID METHYL ESTER AND RESTRICTION LENGTH FRAGMENT
POLYMORPHISM ANALYSES
Introduction
Xanthomonas fragariae causes angular leaf spot disease on strawberry (Fragaria
species and Fragaria x ananassa Duch. cultivars). While historically the disease has not been
a major deterrant in strawberry production, the disease is becoming more important because
of an increase in prevalance of the disease in fruit production fields in Florida and the lack of
effective control measures (Maas, 1995; C. Chandler, pers. comm). Losses in yield are
caused by this disease (Epstein, 1966; Howard, 1971). In addition, regulatory issues regarding
the transportation of infected plant material may impact the nursery plant industry.
International movement of infected plants is blamed for the introduction of angular leaf spot
into Greece and New Zealand (Panagopoulos et al., 1978; Dye and Wilkie, 1973). The
European Plant Protection Organization listed X. fragariae as a quarantine pest and the
FAO/IPGRI recognized it as a potential risk in international movement of strawberry
germplasm (Maas, 1995). Nursery-plant producers in the United States and Canada are
pressured to provide disease-free plants by buyers from foreign markets and by farmers who
refused to buy infected transplants.
45

46
The epidemiology of the X. fragariae must be understood before effective control
strategies can be devised. The delineation of bacterial populations is imperative to such
studies. The relationship of X. fragariae to other members of the genus Xanthomonas has
been examined (Hildebrand et al., 1990; Hodge et al., 1992; Vauterin et al., 1994), but the
genetic variability within the species has not been reported (Maas, 1995). The information
from studies to compare strains of X. fragariae at the genetic level could be useful to identify
the origins and spread of specific strains or genetic types. This information could be applied
in international tracking of pathogen populations for quarantine programs. In addition, the
identification of genetic types is a prerequisite to identify sources of resistance in strawberry.
Strawberry cultivars exhibit levels of susceptibility or tolerence to X. fragariae; only F.
moschata Duch. appeared to be immune (Hazel, 1981; Hazel and Civerolo, 1980; Kennedy
and King, 1962b). A screening program to identify genes for resistance must incoporate
representives of the genetic variants in the screening process otherwise the resistance may be
overcome quickly by genetic variants.
Studies to characterize populations of bacteria have used biochemical and molecular
biological techniques. Protein staining and fatty acid analysis have been useful to detect
differences at the metabolic level within species o Xanthomonas (Bouzar et al. 1994; Stall
et al, 1994; Graham et al., 1990; Hodge et al., 1992). The polymerase chain reaction (PCR)
has been used to analyze differences between pathovars and strains of bacteria (Hensen and
French, 1993). Primers specific to the hrp-gene cluster of Xanthomonas campestris pv.
vesicatoria amplifed genomic DNA and restriction enzyme analysis of the PCR product
differentiated between many pathovars and species of Xanthomonas (Leite et al, 1994 and

47
1995). Primer sets from conserved repetitive bacterial DNA elements generated genomic
fingerprints which were used to differentiate bacteria (De Bruijn, 1992; Louws et al., 1994).
Restriction length fragment polymorphism (RFLP) of the bacterial genome digested with rare-
cutting restriction endonuclease and resolution by pulsed field gel electrophorisis (PFGE) has
been used to type many bacterial strains (Cooksey and Graham, 1989; Egel et al., 1991; Smith
et al., 1995).
In this study, a collection of strains of X. fragariae from the United States and Canada
were examined for genetic diversity. Analyses was by RFLP profiles generated by PFGE and
by fatty acid methyl ester (FAME).
Materials and Methods
Bacterial strains. Strains of X fragariae used in this study are listed in Table 4-1.
Strains were previously rated for pathogenicity on strawberry (Roberts, unpublished).
Geographic origin denotes origin of infested plant material. Strains were cultured on
Wilbrinks medium (WB)(Koike, 1962) at 24 C and long term storage was in 15% glycerol
at -70 C. Single colonies were transfered to nutrient broth (Difco, Detroit, MI) and shaken
for 16 to 20 h at 200 rpm at 24 C.
Restriction fragment length polymorphism. The method used for restriction
endonuclease analysis was described by Egel et al. (1991) and Cooksey and Graham (1989)
except for the following modifications. Cells (1.5 ml of 5 109cfu per ml) were washed in
1 ml of TE (10 mM Tris, ImM EDTA, pH 8.0) and resuspended in 0.5 ml of TE. An equal

48
Table 4.1. The geographic source, collection date, and group determined by fatty acid
methyl ester (FAME) and restriction fragment length polymorphism (RFLP ) for strains
used in this study.
Strain
Geographic
Source3
Year
1238
CA
1990
1240
CA
1990
1241
CA
1990
1242
CA
1990
1243
CA
1990
1245
CA
1990
1246
CA
1990
1249
CA
1990
1250
CA
1990
1290
CA
1989
1291
CA
1989
1293
CA
1989
1295
CA
1989
1296
CA
1989
1298
CA
1989
1424
FL
1992
1425
FL
1992
1426
FL
1992
1427
FL
1992
1428
FL
1992
1429
FL
1992
1431
FL
1992
1514
CA
1993
1515
CA
1993
1516
NC
1993
1517
NC
1993
1518
NC
1993
1519
NC
1993
1520
CA
1993
1523
CA
1993
1524
CA
1993
1525
CA
1993
1526
CA
1993
Sourceb FAMEC RFLP
ARC
2
D
ARC
2
D
ARC
8
B
ARC
8
B
ARC
8
B
ARC
4
B
ARC
8
ARC
4
ARC
5
C
ARC
C
ARC
1
B
ARC
7
B
ARC
8
ARC
8
B3
ARC
8
B
ARC
C
ARC
3
B2
ARC
B
ARC
3
B
ARC
3
B2
ARC
3
B
ARC
B
ARC
C
ARC
6
B3
ARC
B2
ARC
7
B3
ARC
8
B
ARC
B
ARC
5
C
ARC
5
C
ARC
5
c
ARC
c
ARC
8
B
Geographic origin of plant material from which bacteria were isolated. CAN=Canada.
bARC= A. R. Chase; ATCC= American Type Culture Collection.
Blank space indicates that group was not determined.

49
Table 4.1 continued. The geographic source, collection date, and groups determined by
fatty acid methyl ester (FAME) and restriction fragment length polymorphism (RFLP ) for
strains used in this study.
Geographic
Strain
Source3
Year
Sourceb
FAMEC
RFLP
1532
FL
1993
ARC
3
1533
WI
1993
ARC
5
B
1534
WI
1993
ARC
B3
100
FL
1993
This study
1
B
101
FL
1993
This study
1
103
FL
1993
This study
B
104
FL
1993
This study
1
B
105
FL
1993
This study
3
B
106
FL
1993
This study
3
107
FL
1993
This study
1
108
FL
1993
This study
1
B
113
FL
1993
This study
5
B3
114
FL
1993
This study
3
B
115
FL
1993
This study
4
B
116
CAN
1993
This study
B
117
CAN
1993
This study
B3
119
CAN
1993
This study
9
D
124
CAN
1993
This study
B
125
CAN
1993
This study
6
126
CAN
1993
This study
4
127
CAN
1993
This study
6
128
CAN
1993
This study
4
B
129
CAN
1993
This study
6
B3
138
CAN
1993
This study
6
B
146
CAN
1993
This study
4
B1
153
CAN
1993
This study
6
B
33239
MN
ATCC
9
A
a Geographic origin of plant material from which bacteria were isolated. CAN=Canada.
bARC= A. R. Chase; ATCC= American Type Culture Collection.
cBlank space indicates that group was not determined.

50
volume of 1% Seakem Gold agarose solution (10 mM Tris [pH 8.0], 10 mM MgC12, 0.1
mM EDTA [pH 8.0], 1 % Seakem Gold agarose (FMC BioProducts, Rockland, ME) [wt/vol]
in sterile filtered water) was added. Plugs were made, lysed, washed, and stored as described
by Egel et al. (1991). Two sizes of plugs were utilized. A 4 8 mm slice of plug was
restricted in a volume of 200 p1 of restriction buffer (as recommended by manufacturer,
Promega, Madison, WI) and inserted in wells made with a 10-well comb (Bio-Rad,
Richmond, CA). A44 mm square piece of plug was restricted in a volume of 100 p\ and
placed in wells made by a 20-well comb (Bio-Rad). Restriction enzyme was added at the
following concentrations: Xba\ at 40 U; Spel at 30 U; Vsp\ at 30 U (Promega). The pieces
were placed into wells in a 1.2% GTG agarose gel made with 0.5X TBE. Wells were sealed
with the 1% Seakem Gold solution. The gel was placed in a Bio-Rad CHEF DR II unit
containing 1.8 liters of 0.5* TBE and run at 200 V (15V/cm of gel). Pulsed times for plugs
digested with Xbal or Spel were at 4 s for 1 h followed by 8 sec for 18 h. Plugs digested with
Vspl were pulsed at 4 s for 1 h and subsequently at 12 s for 17 h. Lambda DNA in 48.5 KB
concatamers (FMC BioProducts) was used in the first and last lanes of each gel. Gels were
stained in 0.5 mg of ethidium bromide per liter and photographed with type 55 Polaroid film.
The position of bands were assessed visually or by analysis with the Gelmeas
computer program. Similarity values were calculated as described by Egel et al. (1991) with
the mathematical equation proposed by Nei and Li (1979) based upon the proportion of
shared DNA fragments. The estimate of the number of nucleotide substitutions per site was
used to calculate the genetic divergence by the iterative method of Nei (1987) with the
program in SAS described by Leite et al. (1994). The KITSCH program from the PHYLIP

51
computer pakage (Felsenstein, 1995) was used to create a rooted phylogenetic tree by the
Fitch-Margoliash method (Fitch and Margoliash, 1967). The input data was as described for
the combined SpeI, Xbal, and Vspl digestion data (Stall et al., 1994).
Fatty acid composition. Strains of X. fragariae were inoculated onto Trypticase Soy
Broth agar (TSBA) and grown for 48 h at 24 C. Conditions were changed from the
standard MIDI procedures which required that cells be grown on TSBA for 24 hours at 28
C because strains of X fragariae produced insufficient growth at these conditions. Cellular
fatty acids were extracted and derivatized to their fatty acid methyl esters as described
(Sasser, 1990b). A library comprised of strains ofX. fragariae was created using the MIDI
Library Generation System (LGS), software version 3.3. FAMEs were analyzed by the MIDI
Microbial Identification System, software version TSBA 3.5. The qualitative and quantitative
differences in the fatty acid profiles were used to compute the Euclidian distance to each
strain. Strains within six Euclidian distance units, the cut-off for subspecies (Sasser, 1990a),
were grouped in the same cluster.
Results
RFLP-PFGE. Restriction endonucleases Xbal and Spel generated genomic DNA
fragments from 5 to 400 kb (Fig. 4-1). Typically, a strain profile contained ten DNA
fragments greater than 100 kb. Analysis of the 52 strains resulted in four RFLP groups,
designated A through D, which were identical by these two enzymes. Group A contained the
ATCC strain only and represented 2% of the strains. The B, C, and D groups had 77%, 15%

52
and 6% of strains, respectively. A third endonuclease, Vsp\, generated fragments of DNA of
the appropriate size for PFGE analysis and the profiles by this endonuclease separated strains
of the A, C, and D groups identical to the previously determined groups. Analysis of the B
group with the Vsp\ endonuclease subdivided it into three subgroups, denoted B1, B2, and
B3. The B1 contained 13%, B2 contained 20%, and B3 contained 67% of the B strains
tested. Designation of RFLP groups for strains are summarized in Table 4-1.
The dendrogram derived from RFLP analysis of these strains showed the B groups
were very closely related by within 0.006 genetic distance units (Fig. 4-2). The C group was
the most closely related to the B groups, then group A, and finally group D. The D group
was the most remote from the B group at 0.01864 genetic distance units. A strain of A', c.
vesicatoria was used as the statistical outlier and fell at 0.045 genetic distance units from the
D group.
Fatty Acid Methyl Ester Analysis. The 50 strains of X. fragariae were placed into
nine subgroups based upon the MIDI 10% rule (Sasser, 1990b). The subgroups for
individual strains are summarized in Table 4-1. The majority of strains was identified as six
closely related subgroups which can be visualized quantitatively by three major acids, 16:1
w 7 cis, 15:0 Anteiso, and 15:0 iso (Fig.4-3). The 15:0 iso fatty acid comprised
approximately 34% to 54% of FAME profile. The ATTC type strain and three closely related
strains were qualitatively differentiated by the absence of palmitic acid, 16:1 w 7 cis.
The dengrogram unweighted pair analysis (Fig. 4-4) separated strains into four
clusters at the Euclidian distance of six units. The four clusters, designated FAME groups

53
Figure 4-1. Agarose gel showing the restriction fragment length polymorphisms of genomic
DNA of strains of Xanthomonas fragariae after restriction with a rare-cutting endonuclease,
Vspl, and separated by pulsed-field gel electrophoresis. Lambda marker in 48.5 kb
concatamers is shown in side lanes.

54
*- B2
C
A
D
X.c. vesicatoria
i i i
0.04 0.02 0
Genetic Distance
Figure 4-2. Relationship of groups of Xanthomonas fragariae analyzed by restriction
fragment length polymorphisms.

55
20
% 16:1 w 7 cis
Figure 4-3. The groups of Xanthomonas fragaricte distinguished by qualitative and
quantitative differences in fatty acid methyl ester profiles graphed by three major acids.

56
1 through 4, contained 16%, 6%, 4%, and 74% of the strains, respectively. The largest
cluster, FAME group 4, could be subdivided at the Euclidian distance of 4 units into two
groups. The ATCC type strain 33239 and one other strain comprised FAME 3 group.
Discussion
This research represents the first effort to analyze genetic variants of strains within the
species of X. fragariae. The RFLP-PFGE and FAME analyses identified genetic variants
within the population. Methods of RFLP including PFGE have distinguished between closely
related strains of pathovars within the genus Xanthomonas campestris(Ege\ et al., 1991;
Gabriel et al., 1988; Hartung and Civerolo; 1989, Graham et al., 1990). In these works, the
importance of the diverse nature of the endonuclease sites as related to strain characteristics
such as host range, geographic origin, and pathogenicity was discussed. Unfortunately, the
relationship of pathogenicity of the X. fragariae strains to the genetic groupings cannot be
examined Differences in pathogenicity among strains of X.fragariae were not detected by
inoculation of the 52 strains onto two cultivars of strawberry. Studies of bacterial populations
on two cultivars which appeared to have different levels of susceptibility in preliminary tests
were also inconclusive (Roberts, unpublished). Nor have other researchers reported
differences in pathogenicity among strains of X. fragariae Such information might have been
useful to determine significance of the RFLP-PFGE and FAME groups as related to
pathogenicity.

57
Euclidian Distance
0.0 3.55 7.05 10.5 14.1
Cluster "10%" | | | |
1
2
3
4a
4b
3
2
9
7
4
5
6
7
1
5
8
Figure 4-4. Dendrogram of cluster analysis of fatty acid methyl ester profiles showing the
four clusters (1 -4b) of strains of Xanthomonas fragariae. The designation of strains into the
groups determined by the MIDI 10% rule are shown for comparison. The Euclidian
distance of six is the cutoff point for group determination by cluster analysis.

58
The two methods used in this research might be useful to examine the evolution of the
pathogen in future populations. The identification of two groups which contained only four
strains total, including the ATCC type strain, from a population of 50 strains was of interest.
By RFLP analysis, group A contained only the ATCC type strain from Minnesota which was
collected over twenty years ago. However, by FAME cluster analysis, one other strain,
XF119, which was isolated from an infected plant from Canada in 1993, grouped with the
ATCC type strain. In RFLP-PFGE analysis, the XF119 profile was the same as the D group
which contained strains XF1238 and XF1240, isolated from samples from California in 1990.
Dendrogram analysis places these two groups close to each other. By FAME, these four
strains lacked the 16:1 w 7 cis fatty acid which distinguished them from all the other X.
fragariae. It is interesting to speculate that perhaps strain XF119 represent a bridge
between the A and D RFLP groups because of its intergroup relationship in that its RFLP
profile is D but by FAME it is closer to the A group. It would be of further interest to
examine future populations to determine the fate of the two groups represented by only four
strains.
The majority of strains was represented by the B and C groups. An extensive survey
of the pathogen population in the United States and Canada would be useful to determine
whether the population ratios of the various genetic variants are remaining relatively the same.
Likewise, it would be interesting to determine the distribution of the three subdivisions within
the B group. The cultivars of strawberry in commercial production are changed frequently
(C. Chandler, pers. comm ). This constant change in genotypes should influence the genetic
composition of the population of X fragariae if genes for resistance to angular leaf spot are

59
lost or found during the development of new cultivars. In the Philippines, the relative
populations of two races of bacterial blight on rice were followed for 10 years. Researchers
recorded a decline in the prevalence of the predominant race and a concurrent increase in
another race. This change frequency of race, or bacterial genotype, occurred after the
introduction of a gene for resistance in cultivated rice (Mew et al., 1992). Information
regarding changes in the dominant genotype of both the plant and bacteriapopulations would
be useful to plant breeders.
The relationship between geographical origin and genetic variants is unclear because
of the transportation of infected plants. Plants are shipped from California to Canada where
they are propagated and sent to Florida for field production each season. Infection of plants
may have occurred at any point of the route, however symptomatic transplants from Canada
are shipped to Florida. A single type of genetic variant was not found to be associated with
plant material from a particular region of the U. S. or Canada. Likewise, international
movement of infested plants will make it difficult to determine if endemic populations of the
organism exists outside the U. S .. The origin of this disease appears to be the United States
(Maas, 1995).

CHAPTER 5
DISCUSSION
Field experiments which examined epidemiological aspects of angular leaf spot on
strawberry established a decrease in yield on strawberry plants due to angular leaf spot. This
report is the first to quantify the reduction in yield due to this disease. The 8- to 10% loss
determined in these studies is much lower than the 70- to 80% loss estimated by Epstein
(1966) in fruit production fields in Wisconsin. Production in northern regions of the U. S is
usually perennial and the age of the plants was not given. The loss was reported on the
cultivar Sparkle which is not produced commercially in Florida. Therefore, we are unable
to compare our yield loss to his report Howard (1971) also reported unquantified yield
losses due to this disease but on cultivars which are no longer in commercial production In
our studies, the average yield loss observed both years was very similar despite the differences
in the disease levels and the base yields between the two seasons. In 1995, the disease
severity was 60% of the amount estimated in the previous season; however, the yield loss was
actually higher by 2% for marketable yield The 10% reduction in yield represents an
estimated loss of 1233 kg of berries per acre. Strawberry fruit has a very high cash value and
a decrease in revenue of 8- to 10% represents a significant economic loss to producers.
Strawberry producers in Florida currently try to control angular leaf spot by
application of copper compounds and by avoidance of the disease. An understanding of the
60

61
epidemiology of angular leaf spot is necessary to determine the effectiveness and practicability
of any control method. The survey of farmers fields and the cold storage facility established
that transplants arrive from northern nurseries infected with angular leaf spot at fairly high
disease incidence. Therefore, disease is introduced into the field on infected transplants. In
our surveys, by the sampling method used in the first season, more plants were assessed for
symptoms of angular leaf spot in the field compared to the number of plants in the random
sampling of boxes in cold storage. This probably accounts for identification of diseased plants
in nearly all of the farmers fields versus less than in 25% of boxed plants. In addition, plants
were examined until disease was found in the field which skewed the randomness of the
survey.
In field experiments, the spread from inoculated plants to noninoculated plants in
nearby experimental plots was minimal. Thus, the incidence of plants with angular leaf spot
in the field likely occurred from inoculum already present on the transplants. In the field
experiments, if inoculum was present from another source, such as debris in the soil or an
alternate host, most likely the disease on the noninoculated plants would have been more
general and appeared earlier. Disease severity on noninoculated plants in the experimental
plots was extremely low and most plants remained free of angular leaf spot. For the disease
that occurred on noninoculated plants, the inoculum was probably transported mechanically
from inoculated plants during harvests or disease readings.
Progress of angular leaf spot developed similar to curves of pathogens described on
other crops (van der Plank, 1963). A decrease in disease severity was seen about midway
through the season in 1994. This reduction, or negative infection rate, was due to growth

62
by strawberry plants which diluted the amount of disease relative to the total leaf area. The
dilution effect of healthy plant tissue was described by van der Plank (1963). In 1995, a
negative infection rate was seen very early in the epidemic. This was because the symptoms
of the disease almost disappeared from the field as the affected leaves died and were lost. The
difference in disease progress between the two seasons did not appear to be from differences
in rainfall since mean rainfall was approximately the same both years (data not shown). Mean
temperatures also were similar; however the number of days at temperatures below 10 C in
November and December 1993 were much greater than in 1994. Another difference between
the seasons was that sprinkler irrigation was reduced in 1995 because of the change to drip
irrigation. Overhead irrigation, fog, high humidity, and rain are positive factors in disease
development and spread (Epstein, 1966; Hildebrand et al., 1967; Kennedy and King, 1962b;
Lai, 1978). The trend towards drip irrigation in commercial production fields should have a
positive effect to reduce the spread and survival of the bacterium.
Strawberry producers currently apply copper-based compounds to control angular leaf
spot despite the phytotoxicity to plants caused by copper. Application of cupric hydroxide
plus mancozeb at the lx rate was phytotoxic to strawberry plants (Howard, 1973; this study).
In our tests, the greatly reduced 0.1 rate did not harm plants and significantly reduced
disease severity in 1995. The protectant application of bactericide at this rate and spray
schedule was intended to reduce the total amount of inoculum and to prevent spread of the
disease. The approach of frequent sprays at reduced concentrations may have potential to
control this disease. More studies are needed to evaluate different rates, application intervals,
and chemical mixtures to achieve maximum disease control while avoiding yield losses. An

63
alternate approach might be to control the disease in the nursery by copper applications. The
loss in yield due to phytotoxicity would not be an issue on nursery plants. The sprays would
reduce the amount of initial inoculum on transplants and subsequently reduce the amount of
disease in fruit production fields. The effect of sprays on the categories of yield were difficult
to interpret because of the lack of consistent, significant differences between years.
Leite et al. (1994) used the sequence variation within the hrpB operon among plant-
pathogenic xanthomonads to select primers with different specificities and this approach was
successful in our studies to identify primers specific to X. fragariae. The hrp-primers RST2
and RST3 amplified DNA from all strains ofJf. fragariae and analysis of the PCR products
by REA showed no polymorphism within this region. The homology of the amplified region
presented a good site to select primers universal to X. fragariae. Primers XF9, XF11, and
XF12 designed from these unique sites were specific for amplification of DNA only from
strains ofA^ fragariae. The primer XF10 was responsible for the non-specific amplification
of a strain of X. c. pelargonii in preliminary tests and therefore was not used in later
experiments. Interestingly, an ELISA test developed for identification of X. fragariae in vitro
cross reacts to some related bacteria including X. c. pelargonii (Maas, 1995). Perhaps a
relationship between X fragariae and X. c. pelargonii exists which is not reflected by
traditional taxonomic techniques (Hildebrand et al., 1990; Vauterin et al., 1994; Maas, 1995).
The level of sensitivity for most detection techniques for specific bacteria (Pickup,
1991; Schaad et al., 1982 ) are comparable to the level which was achieved with a single
round of PCR amplification. PCR was reported to be more sensitive than ELISA (Leite et
al., 1995) and with our primers, a single round of PCR is 10 times more sensitive than the

64
ELISA test for X. fragariae developed by Rowhani et al. (1994). In addition, their antibodies
cross-reacted with another xanthomonad whereas we were able to eliminate the non-specific
reaction by discontinuing use of primer XF10. Although this level of detection is adequate
to detect bacteria in symptomatic leaf tissue, we were concerned with the detection of
bacteria from asymptomatic tissue. The X fragariae primers were suitable for use in the
nested technique and the sensitivity of the PCR reaction was increased by 1,000-fold. This
level of detection was achieved in assays to detect the bacterium in the presence of plant
tissue and is well below the number of cells needed to cause visible lesions. Therefore, the
nested technique is applicable for detection of bacteria in association with asymptomatic
tissue.
Cross-contamination among samples and contamination of PCR reagents are problems
associated with the nested technique (McManus and Jones, 1995). We experienced false
positives in a number of initial experiments and finally determined the cause to be
contaminated mineral oil Although we eliminated the need for mineral oil overlay by using
a hot bonnet, extreme care is needed to avoid contamination and subsequent false positives.
In our experiments, negative control samples were always included for each amplification
round of the nested assay. Aerosols may also contribute to false positives and care must be
taken to prepare the mixtures for PCR in a sterile environment and negative control samples
should be included to check for aerosol contamination. In the field experiments, a problem
encountered with the nested technique was amplification of bacterial DNA from the negative
control plants. Although these plants were physically isolated from a contamination source
during the experiments; they were initially obtained from fields at GCREC-Dover where

65
plants infected with angular leaf spot were located. The plants may have been infested with
the bacterium or cross-contamination of samples may have occurred during preparation of
plant samples from handling infected material or using non-sterile instruments. Aerosols may
also have been a source of contamination of PCR samples (Innis et al., 1990). However,
contamination from these sources would be detected by amplification in the negative controls.
Our negative controls were included in every PCR run and were consistently negative.
Therefore contamination of PCR reagents or aerosols would have been ruled out.
Confirmation that the PCR product is from amplification of the target DNA is possible
by REA (Leite et al. 1994). The REA was used to identify PCR products amplified by the
/wp-primers from cells of X. c. vesicatoria added to seed washing of tomato and pepper (Leite
et al. 1995). Similarly, we applied REA to identify PCR products amplified by the XF-
primers. The profiles of the restricted PCR products were distinct for each enzyme. The
profiles from REA should be different for unrelated organisms since it would be highly
unlikely that the same restriction sites would exist for two heterologous pieces of DNA (Leite
et al,. 1995).
The nested PCR technique was useful to detect bacteria on strawberry plants in
nurseries in Florida Symptoms and recovery of the rifampicin-marked strain on a selective
medium were not useful to identify bacteria when populations were extremely low. However,
recovery of the rifampicin-marked strain from some samples indicated that the bacteria were
viable throughout the summer and that PCR was detecting viable bacterial cells. Visible
symptoms on plants disappeared soon after placement in the field and recovery on media was
difficult due to the slow-growing nature of X. fragariae and overgrowth by contaminants.

66
While bacterial populations were not enumerated, the levels were deduced to be below 103
cfu per ml, since bacteria were detected in the nested PCR but not by single round of
amplification. This level of sensitivity would be useful to screen asymptomatic nursery plants
to detect plants contaminated with X. fragariae. This is a concern of nursery-plant producers
and regulatory agencies of plant shipment (Maas, 1995). Currently, plants from nurseries in
California are being screened by this technique to determine its usefulness in such an
application.
The field studies have implications regarding the disease cycle of angular leaf spot in
Florida The decline and later increase in the number of positive samples through the summer
would indicate that populations of the bacteria declined throughout the summer and increased
when more favorable conditions in the form of cooler weather occurred. The bacterial
populations did not completely die. Therefore, to eliminate disease, the production of plants
in nurseries in Florida would have to begin with plants free of the bacterium.
Researchers have reported the systemic movement of X. fragariae in plants
(Hildebrand et al., 1967; Milholland et al., 1993). In our research, inoculation of strawberry
plants resulted in survival of bacteria on leaves and the crown for extended periods under
conditions not optimum for growth of the bacteria. In addition, bacteria were detected by
PCR technique on daughter plants in the field. Dissemination to the daughter plants could
have been due to either systemic movement through the vascular system of the runner or
dispersal of bacterium by mechanical means.
This research represents the first effort to analyze genetic variants of strains within the
species of X. fragariae. The RFLP-PFGE and FAME analyses identified genetic variants

67
within the population. Methods of RFLP including PFGE have distinguished between closely
related strains of pathovars within the genus Xanthomonas campestris (Egel et al., 1991;
Gabriel et al., 1988; Hartung and Civerolo; 1989, Graham et al., 1990). In these works, the
importance of the diverse nature of the endonuclease sites as related to strain characteristics
such as host range, geographic origin, and pathogenicity was discussed. Unfortunately, the
relationship of pathogenicity of the X fragariae strains to the genetic groupings cannot be
examined. Differences in pathogenicity among strains of X. fragariae were not detected by
inoculation of the 52 strains onto two cultivars of strawberry. Studies of bacterial populations
on two cultivars which appeared to have different levels of susceptibility in preliminary tests
were also inconclusive (Roberts, unpublished). Nor have other researchers reported
differences in pathogenicity among strains ofAi fragariae. Such information might have been
useful to determine significance of the RFLP-PFGE and FAME groups as related to
pathogenicity.
The two methods used in this research might be useful to examine the evolution of the
pathogen in future populations. The identification of two groups which contained only four
strains total, including the ATCC type strain, from a population of 50 strains was of interest.
By RFLP analysis, group A contained only the ATCC type strain from Minnesota which was
collected over twenty years ago. However, by FAME cluster analysis, one other strain,
XF119, which was isolated from an infected plant from Canada in 1993, grouped with the
ATCC type strain. In RFLP-PFGE analysis, the XF119 profile was the same as the D group
which contained strains XF1238 and XF1240, isolated from samples from California in 1990.
Dendrogram analysis places these two groups close to each other. By FAME, these four

68
strains lacked the 16:1 w 7 cis fatty acid which distinguished them from all the other X.
fragariae. It is interesting to speculate that perhaps strain XF119 represents a bridge
between the A and D RFLP groups because of its intergroup relationship in that its RFLP
profile is D but by FAME it is closer to the A group. It would be of further interest to
examine future populations to determine the fate of the two groups represented by only four
strains.
The majority of strains was represented by the B and C groups. An extensive survey
of the pathogen population in the United States and Canada would be useful to determine
whether the population ratios of the various genetic variants are remaining relatively the same.
Likewise, it would be interesting to determine the distribution of the three subdivisions within
the B group. The cultivars of strawberry in commercial production are changed frequently
(C. Chandler, pers. comm ). This constant change in genotypes should influence the genetic
composition of the population of X. fragariae if genes for resistance to angular leaf spot are
lost or found during the development of new cultivars. In the Philippines, the relative
populations of two races of bacterial blight on rice were followed for 10 years. Researchers
recorded a decline in the prevalence of the predominant race and a concurrent increase in
another race. This change in frequency of race, or bacterial genotype, occurred after the
introduction of a gene for resistance in cultivated rice (Mew et al., 1992). Information
regarding changes in the dominant genotype would be useful to plant breeders.
The identification of a relationship between geographical origin and genetic variants
is unclear because of the transportation of infected plants. Plants are shipped from California
to Canada where they are propagated and sent to Florida for field production each season.

69
Infection of plants may have occurred at any point of the route, however symptomatic
transplants from Canada are shipped to Florida A single type of genetic variant was not found
to be associated with plant material from a particular region of the U. S. or Canada.
Likewise, international movement of infested plants will make it difficult to determine if
endemic populations of the organism exists outside the U. S.. The origin of this disease
appears to be the United States (Maas, 1995).

LIST OF REFERENCES
Alippi, A. M., Ronco, B. L., and Carranza, M. R. 1989. Angular leaf spot of strawberry,
a new disease in Argentina: Comparative control with antibiotics and fungicides. Adv. Hort.
Sci. 3:3-6.
Ausubel, F. M., Brent, R. Kingston, R. E., Moore, D. D., and Seidman, K. 1987. Current
Protocols in Molecular Biology. John Wiley & Sons, New York.
Bouzar, H., Jones, J B., Stall, R. E., Hodge, N.C., Minsavage, G. V., Benedict, A. A., and
Alvarez, A. M. 1994. Physiological, chemical, serological, and pathogenic analyses of a
worldwide collection of Xanlhomonas campestris pv. vesicatoria strains. Phytopathology
84:663-671.
Conway, K. E., Motes, J. E., Bostian, B., Fisher, C. G., and Claypool, P. L. 1987.
Cercospora blight development on asparagus fern and effects of fungicides on disease severity
and yield. Plant Dis. 71:254-259.
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.
De Bruijn, F. J. 1992. Use of repetitive (repetitive extragenic palindromic and
enterobacterial repetitive intergeneric consensus) sequences and the polymerase chain reaction
to fingerprint the genomes of Rhizobium meliloti isolates and other soil bacteria. Appl.
Environ. Microbiol. 58:2180-2187.
Dye, D. W., and Wilkie, J. P 1973. Angular leafspot of strawberry in New Zealand N. Z.
J. Agr. Res. 16:311-314.
Egel, D. S., Graham, J. H., and Stall, R E. 1991. Genomic relatedness of Xanthomonas
campestris strains causing diseases of citrus. App. Environ. Microbiol. 57:2724-2730.
Epstein, A. H. 1966. Angular leaf spot of strawberry. Plant Dis. Rep. 50:167.
Felsenstien, J. 1995. PHYLIP. Phylogeny inference package, ver.3.57. Department of
Genetics, University of Washinton, Seattle, WA.
70

71
Fitch, W. C, and Margoliash, E. 1967. Construction of phylogenetic trees. Science 155:
279-284.
Fry, W E. 1975. Integrated effects of polygenic resistance and a protective fungicide on
development of potato late blight. Phytopathology 65:908-911.
Fry, W. E., Bruck, R. I., and Mundt, C. C. 1979. Retardation of potato late blight epidemics
by fungicides with eradicant and protectant properties. Plant Dis. Rep. 63:970-974.
Gabriel, D. W., Hunter, J. E., Kingsley, M. T., Miller, J. M., 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.
Hartung, J. S., and Civerolo, E. L. 1989. Restriction fragment length polymorphisms
dintinguish Xanthomonas campestris strains isolated from Florida citrus nurseries from X. c.
pv. citri. Phytopathology 79:793-799.
Hazel, W. J. 1981. Xanthomonas fragariae, cause of strawberry angular leaf spot: Its
growth, symptomatology, bacteriophages, and control. Ph D. thesis. University of Maryland,
College Park. 119 p.
Hazel, W. J., and Civerolo, E. L 1980. Procedures for growth and inoculation of
Xanthomonas fragariae, causal organism of angular leaf spot of strawberry. Plant Dis.
64:178-181.
Henson, J. M., and French, R. 1993. The polymerase chain reaction and plant disease
diagnosis. Ann. Rev. Phytopathol 31:81-109.
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.
Hildebrand, D. C., Schroth, M N., and Wilhelm, S. 1967. Systemic invasion of strawberry
by Xanthomonas fragariae causing vascular collapse Phytopathology 57:1260-1261.
Hodge, N. C., Chase, A. R., and Stall, R. E. 1992. Diversity of four species of Xanthomonas
as determined by cellular fatty acid analyses. (Abst.) Phytopathology 82:1 153.
Howard, CM. 1971. Occurrence of strawberry angular leaf spot, Xanthomonas fragariae,
in Florida Plant Dis. Rep. 55:142.

72
Howard, C. M and Albregts, E. E. 1973. Strawberry report. Fungicide and Nematicide
Tests 29:47.
Howard, C. M, Overman, A. J., Price, J. F., and Albregts, E. E. 1985. Diseases, nematodes,
mites, and insects affecting strawberries. Florida. Agrie. Experiment Stations, Inst, of Food
and Agrie. Sci., Univ. of FI., Bulletin 857.
Innis, M. A., Gelfand, D. H., Sninsky, J. J., and White, T. J. 1990. PCR Protocols: A Guide
to Methods and Applications. Academic Press, Inc., San Diego, CA.
Jones, A. L. 1995. A stewardship program for using fungicides and antibiotics in apple
disease management programs. Plant Dis. 79:427-432.
Jones, J. B, Woltz, S. S., Kelly, R. O., and Harris, G. 1991. The role of ionic copper, total
copper, and select bactericides on control of bacterial spot of tomato. Proc. Fla. State Hort.
Soc. 104:257-259.
Kennedy, B. W., and King, T. H. 1962a. Angular leaf spot of strawberry caused by
Xanthomonasfragariae sp. nov. Phytopathology 52:873-875.
Kennedy, B. W., and King, T. H. 1962b Studies on epidemiology of bacterial angular leaf
spot on strawberry. Plant Dis. Rep. 46:360-363.
Koike, H. 1965. The aluminum-cap method for testing sugarcane varieties against leaf scald
disease. Phytopathology 55:317-319.
Lai, C. M. 1978. Bacterial disease of strawberry. Plant Path. Reports. California Dept. Of
Food and Agrie. Lab Services. Vol. II.
Leite, R. P., Egel, D. S., and Stall, R. E. 1994. Genetic analysis of hrp-related DNA
sequences of Xanthomonas campestris strains causing diseases of citrus. Appl Environ.
Microbiol. 60:1078-1086.
Leite, R. P., Jr., Jones, J. B., Somodi, G. C., Minsavage, G. V., and Stall, R E. 1995.
Detection of Xanthomonas campestris pv. vesicatoria associated with pepper and tomato
seed by DNA amplification. Plant Dis. 79:917-922.
Leite, R. P., Jr., Minsavage, G. V., Bonas, U., and Stall, R E. 1994. Detection and
identification of phytopathogenic Xanthomonas strains by amplification of DNA sequences
related to the hrp genes of Xanthomonas campestris pv. vesicatoria. Appl. Environ.
Microbiol. 60:1068-1077.

73
Lindgren, P. B., Peet, R. C., and Panopoulos, N. J. 1986. Gene cluster of Pseudomonas
syringae pv. "phaseolicola" control pathogenicity of bean plants and hypersenstivity on
nonhost plants. J. Bacteriol. 168:512-522.
Louws, F. J., Fulbright, D. W., Stephens, C. T., and Bruijn, F. J. 1994. Specific genomic
fingerprints of phytopathogenic Xanthomonas and Pseudomonas pathovars and strains
generated with repetitive sequences and PCR. Appl. Environ. Microbiol. 60:2286-2295.
Maas, J. L. (ed.) 1984. Angular leaf spot. Compendium of Strawberry Diseases. American
Phytopathological Society, St. Paul, MN.
Maas, J. L., Pooler, M. R, and Galletta, G. J. 1995. Bacterial angular leafspot disease of
strawberry: Present status and prospects for control. Adv. Strawberry Res. 14:18-24.
Marco, G.M., and Stall, RE. 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.
McManus, P. S., and Jones, A. L. 1995. Detection of Erwinia amylovora by nested PCR
and PCR-dot-blot and reverse blot hybridizations. Phytopathology 85:618-623.
Milholland, R. D., Ritchie, D F., Daykin, M. E., and Gutierrez, W. A. 1993. Systemic
movement of Xanthomonas fragariae in inoculated strawberry plants. (Abst.)
Phytopathology 83:1408.
Minsavage, G. V., Thompson, C. M., Hopkins, D. L., Leite, R M. V. B., and Stall, R E.
1994. Development of a polymerase chain reaction protocol for detection of Xylella
fastidiosa in plant tissue. Phytopathology 84:456-461.
Nei, M. 1987. Molecular Evolutionary Genetics. Columbia University Press, New York.
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.
Panagopoulos, C. G., Psallidas, P. G., and Alivizatos, A. S. 1978. A bacterial leafspot of
strawberry in Greece caused by Xanthomonas fragariae. Phytopath. Z. 91:33-38.
Pataky, J. K., and Lim, S. M. 1981. Efficacy of benomyl for controlling Septoria brown spot
of soybeans. Phytopatholology 71 438-442.
Pickup, R. W. 1991. Development of molecular methods for the detection of specific bacteria
in the environment J. Gen. Microbiol. 137:1009-1019.

74
Rhoads, D. D., and Roufa, D.J. 1991. Seqaid II Computer Program. Molecular Genetics
Lab, Kansas State University, Manhattan, KS 66506.
Ritchie, D. F., Averre, C. W., and Milholland, R D. 1993. First report of angular leaf spot,
caused by Xanthomonasfragariae, on strawberry in North Carolina. Plant Dis. 76:1263.
Rowhani, A., Feliciano, A. J., Lips, T., and Gubler, W. D. 1994. Rapid identification of
Xanthomonasfragariae in infected strawberry leaves by enzyme-linked immunosorbent assay.
Plant Dis. 78:248-250.
SAS Institute Inc. 1988. SAS/STAT User's Guide, ed. 6.03. SAS Institute Inc., Cary, N.C.
Sasser, M. 1990. Tracking a strain using the the microbial identification system. Technical
note 102. Microbial ID, Inc. Newark, DE.
Sasser, M. 1990. Identification of bacteria through fatty acid analysis. Methods in
Phytobacteriology. Z. Klement, K., Rudolph, and D. Sands, eds. Akademiai Kiado,
Budapest.
Schaad, N. W., Cheong, S. S., Tamaki, S., Hatziloukas, E., and Panopoulos, N. J. 1993. A
viable cell enrichment, two-step, direct PCR technique for detection of Pseudomonas
syringae pv. phaseolicola in bean seeds. (Abstr.) Phytopathology 83:1342.
Shaner, G., and Finney, R E. 1977. The effect of nitrogen fertilization on the expression of
slow-mildewing resistance in Knox wheat. Phytopathology 67:1051-1056.
Smith, J. J., Offord, L. C., Holderness, M., and Saddler, G. S. 1995. Genetic diversity of
Burkholderia solanacean/m (synonym Pseudomonas solanacearum) race 3 in Kenya. Appl.
Environ. Microbiol. 4263-4268.
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. Sys. Bacteriol.
44:47-53.
Stall, R. E., Loschke, D. C., and Jones, J. B. 1986. Linkage of copper resistance and
avirulence loci on a self-transmissible plasmid in Xanthomonas campestris pv. vesicatoria.
Phytopathology 76:240-243.
Stall, R. E., and Minsavage, G. V. 1990. The use of hrp genes to identify opportunistic
xanthomonads. Proc. 7th Int. Conf. Plant Pathog. Bacteria 1989. Z. Klement, Akadmiai
Kiad, Budapest.

75
Stall, R. E., and Thayer, P. L. 1962. Streptomycin resistance of the bacterial spot pathogen
and control with streptomycin. Plant Dis. Rep. 16:389-392.
Szybalski, W. 1952. Microbial Selection. 1. Gradient plate technique for study of bacterial
resistance. Science 116:46-48.
Vauterin, L., Hoste, B., Kersters, K. and Swings, J. 1995. Reclassification ofXanthomonas.
Int. J. Syst. Bacteriol. 45:472-489.
van der Plank, J. E. 1963. Control of Disease by Fungicides. Plant Diseases: Epidemics and
Control. Academic Press, NY.
Willis, K., Rich, J. J., and Hrbak, E M. 1991. hrp genes of phytopathogenic bacteira. Mol.
Plant Microbe Interact. 4:132-138

BIOGRAPHICAL SKETCH
Pamela D. Roberts was born September 18, 1963 in Manhattan, KS. She graduated
with a Bachelor of Science degree in horticulture from Kansas State University in 1987. She
completed a Master of Science from the University of Hawaii in 1991 in the Department of
Plant Pathology. She spent 16 months as a Research Scholar at the International Rice
Research Institute in conjunction with the research at the Univerisity of Hawaii She arrived
at the University of Florida in 1991 to undertake a Doctor of Philosophy degree in the
Department of Plant Pathology.
76

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.
f)
Richard D. Berger, 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.
Robert E. Stall
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
1 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.
Associate Professor of Horticulture

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.
May, 1996
Deart/College of Agricultii^
Dean, Graduate School

LO
1780
1996
, Rm
UNIVERSITY OF FLORIDA
3 1262 08554 9185



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.
May, 1996
Deart/College of Agricultii^
Dean, Graduate School


34
Specificity of primers. For all strains of X. fragariae, the PCR products amplified by
primers RST2 and RST3 were ca. 840-bp (Fig. 3-1). The REA profiles that resulted from
restriction of the PCR products with Cfo\ or HaeIII were the same for all the strains of X.
fragariae (data not shown). Genomic DNA from non-pathogenic strains of Xanthomonas
isolated from strawberry was not amplified by the primers
The four primers synthesized were: XF9 (5' TGGGCCATGCCGGTGGAACTGT
GTGG3'); XF10 (5' TGGAACTGTGTGGCGAGCCAG 3); XF11 ('5 TACCCAGCCGT
CGCAGACGACCGG 3'); and XF12 (5' TCCCAGCAACCCAGATCCG 3'). Primers XF10
and XF12 were internal to the other two primers
Primer XF9 paired with XF11 or XF12 delineated a 537-bp or a 458-bp fragment,
respectively (Fig. 3-1). Primers XF10 and XF12 delineated a 448-bp fragment. PCR
products conformed to the estimated sizes based upon sequence data. All 49 strains of X.
fragariae were amplified with primer sets XF9 and XF11, XF9 and XF12, and XF10 and
XF12. DNA from strains other than X. fragariae was not amplified by primer XF9 paired
with XF11 or XF12. Strains tested were ATCC type strains of Xanthomonas campestris
pathovars begoniae, campestris, carotae, celebensis, glycines, incanae, manihotis,
musacearum, papavericola, pelargonii, phaseoli, poinsettiicola, raphani, taraxaci,
vignicola, vitians, and nine non-pathogenic strains of xanthomonads isolated from strawberry.
Except for X. c. pelargonii, genomic DNA from these strains could not be amplified with
primer set XF10 and XF12. Due to the non-specific amplification of DNA from X. c.
pelargonii by primer XF10 paired with XF12, only primer XF9 paired with either XF11 or


30
were synthesized with a model 394 DNA synthesizer (Applied Biosystems, Foster City, CA)
at the ICBR Facility, University of Florida, Gainesville, FL.
PCR amplification and nested amplification. Total genomic DNA was extracted by
the method described by Ausubel et al. (1987). PCR amplification was performed using a
DNA Thermal Controller PT-100 (MJ Research, Watertown, MA). Samples were in a total
reaction volume of 50 /ft and contained IX amplification buffer (Promega, Madison, WI), 100
M of each dNTP (Promega, Madison, WI), 50 /M of each primer, 1.25 U Taq DNA
polymerase, and 100 ng of purified genomic DNA in 3 /ft of TE (10 mM Tris and 1 mM
EDTA, pH 8.0) buffer. Each reaction was overlaid with 50 /ft of sterilized mineral oil
(Sigma) for a total volume of 100 /ft in sterile 0.6 ml microcentrifuge tubes. In later
experiments, the thermocycler was equipped with a heated lid controller (The Hot Bonnet,
MJ Research) which eliminated the need for the mineral oil overlay. Amplification of the
DNA proceeded after template DNA was denatured at 95 C for 2 min followed by thirty
amplification cycles and a final extension step at 72 C for 5 min. For primer set XF9 and
XF11, each amplification cycle consisted of denaturation at 95 C for 30 s, annealing at 65
C for 30 s, and extension at 72 C for 45 s. For primer XF12 with either primer XF9 or
XF10, the program was identical except the annealing temperature was 58 C.
For nested PCR, the first round of amplification was as described with primers XF9
and XF11. In the second round of amplification, a 3 ^ul sample from the first amplification
mixture was used with primers XF9 and XF12 and all other ingredients were added at the
concentrations described above. The PCR cycle program for the primers XF9 and XF12 was


7
nurseries, disease spread in strawberry production fields, the effect of angular leaf spot on
yield, and the control of the disease by chemicals. To aid in epidemiological studies, a
sensitive and specific technique by PCR reaction to detect X. fragariae was developed.
Primers were designed specific to the region of genomic DNA from X. fragariae related to
the hrp genes of X. c. vesicatoria. The survival of X fragariae on nursery strawberry plants
in the field at two locations in Florida and dissemination to daughter plants was examined to
understand the disease cycle of angular leaf spot in Florida. The genetic variability of a
collection of strains of X. fragariae from the United States and Canada was examined.
Analyses were by restriction length fragment profiles of genomic DNA restricted with rare-
cutting endonucleases separated by pulsed field gel electrophoresis and by profiles of fatty
acid methyl esters.


THE EPIDEMIOLOGY, SPECIFIC DETECTION, AND GENETIC VARIABILITY OF
XANTHOMONAS FRAGARIAE
BY
PAMELA D. ROBERTS
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
1996


4 GENOMIC RELATEDNESS OF XANTHOMONASFRAGARIAE
ON STRAWBERRY BY FATTY ACID METHYL
ESTERASE AND RESTRICTION LENGTH FRAGMENT
POLYMORPHISM ANALYSES 45
Introduction 45
Materials and Methods 47
Results 51
Discussion 56
5 DISCUSSION 60
LIST OF REFERENCES 70
BIOGRAPHICAL SKETCH 76
v


Yield (KG) Yield (KG)
20
1.2
Noninoculated Inoculated
NoS 0.1X IX
NoS 0.1X IX
Noninoculated
Inoculated
NoS 0.1X IX
NoS 0.1X IX
Figure 2-5. The culled fruit on strawberry in Florida in the 1994 (A) and 1995 (B) seasons.
A, B = F was significant at P = 0.05 within the first factor of noninoculated vs.inoculated; a,
b = F was significant at P = 0.05 within the second factor of spray treatments.


66
While bacterial populations were not enumerated, the levels were deduced to be below 103
cfu per ml, since bacteria were detected in the nested PCR but not by single round of
amplification. This level of sensitivity would be useful to screen asymptomatic nursery plants
to detect plants contaminated with X. fragariae. This is a concern of nursery-plant producers
and regulatory agencies of plant shipment (Maas, 1995). Currently, plants from nurseries in
California are being screened by this technique to determine its usefulness in such an
application.
The field studies have implications regarding the disease cycle of angular leaf spot in
Florida The decline and later increase in the number of positive samples through the summer
would indicate that populations of the bacteria declined throughout the summer and increased
when more favorable conditions in the form of cooler weather occurred. The bacterial
populations did not completely die. Therefore, to eliminate disease, the production of plants
in nurseries in Florida would have to begin with plants free of the bacterium.
Researchers have reported the systemic movement of X. fragariae in plants
(Hildebrand et al., 1967; Milholland et al., 1993). In our research, inoculation of strawberry
plants resulted in survival of bacteria on leaves and the crown for extended periods under
conditions not optimum for growth of the bacteria. In addition, bacteria were detected by
PCR technique on daughter plants in the field. Dissemination to the daughter plants could
have been due to either systemic movement through the vascular system of the runner or
dispersal of bacterium by mechanical means.
This research represents the first effort to analyze genetic variants of strains within the
species of X. fragariae. The RFLP-PFGE and FAME analyses identified genetic variants


29
Laboratories, Detroit, MI) and incubated at 28 C. Long term storage was at -70 C in 15%
glycerol. Bacteria used for plant inoculations and DNA extractions were grown in 5 ml
nutrient broth on a rotary shaker at 200 rpm for 16 h at 24 C. A rifampicin-resistant mutant
of strain XF1425 was selected on Wilbrinks medium supplemented with 100 ,ug/ml of
rifampicin by the gradient plate technique (Szybalski, 1952).
Pathogenicity tests. Bacteria from overnight cultures in nutrient broth were
centrifuged and washed three times with sterile water. The concentration of cells was
adjusted in either 10 mM MgSO4-7H20 or sterile water to approximately 105 cfu per ml and
sprayed to runoff on 'Sweet Charlie strawberry plants placed under mist 24- to 48- h prior
to inoculation. Inoculated plants were maintained under mist or put into growth chambers
(Percival, Boone, IA) at 24 C with a 12 h photoperiod.
Sequencing and primer design. The hrp primers RST2 and RST3 from the hrp gene
cluster of X. axonopodis pv. vesicatoria (Leite et al, 1994) were used to amplify genomic
DNA of 49 strains oX. fragariae. The PCR product from strain XF1425 of X. fragariae
was isolated from agarose gel, cleaned by the Promega Wizard Kit (Promega, Madison, WI)
and sequenced at the ICBR DNA Sequencing Facility, University of Florida, Gainesville, FL.
The nucleotide sequence was compared to sequences of the PCR products amplified by the
same primers from one X vesicatoria and two X a. pv. vesicatoria strains using the Seqaid
II computer program (Rhoads and Roufa, 1991). Four primers were selected from the
sequence of X. fragariae based upon unique DNA sequences and low homology compared
to the DNA sequences from the other bacterial strains. The four oligonucleotide primers


24
difference in disease progress between the two seasons did not appear to be from differences
in rainfall since mean rainfall was approximately the same both years (data not shown). Mean
temperatures also were similar; however; the number of days at temperatures below 10 C
in November and December 1993 were much greater than in 1994. Another difference
between the seasons was that sprinkler irrigation was reduced in 1995 because of the change
to drip irrigation. Overhead irrigation, fog, high humidity, and rain are positive factors in
disease development and spread (Epstein, 1966; Hildebrand et al., 1967; Kennedy and King,
1962b; Lai, 1978). The trend towards drip irrigation in commercial production fields should
have a positive effect to reduce the spread and survival of the bacterium.
Strawberry producers currently apply copper-based compounds to control angular leaf
spot despite the phytotoxicity to plants caused by copper. Application of cupric hydroxide
plus mancozeb at the 1* rate was phytotoxic to strawberry plants (Howard, 1973; this study).
In our tests, the greatly reduced 0.1 x rate did not harm plants and significantly reduced
disease severity in 1995. The protectant application of bactericide at this rate and spray
schedule was intended to reduce the total amount of inoculum and to prevent spread of the
disease. The approach of frequent sprays at reduced concentrations may have potential to
control this disease. More studies are needed to evaluate different rates, application intervals,
and chemical mixtures to achieve maximum disease control while avoiding yield losses. An
alternate approach might be to control the disease in the nursery by copper applications. The
loss in yield due to phytotoxicity would not be an issue on nursery plants. The sprays would
reduce the amount of initial inoculum on transplants and subsequently reduce the amount of
disease in fruit production fields.


CHAPTER 4
GENOMIC RELATEDNESS OF XANTHOMONAS FRAGARIAE ON STRAWBERRY
BY FATTY ACID METHYL ESTER AND RESTRICTION LENGTH FRAGMENT
POLYMORPHISM ANALYSES
Introduction
Xanthomonas fragariae causes angular leaf spot disease on strawberry (Fragaria
species and Fragaria x ananassa Duch. cultivars). While historically the disease has not been
a major deterrant in strawberry production, the disease is becoming more important because
of an increase in prevalance of the disease in fruit production fields in Florida and the lack of
effective control measures (Maas, 1995; C. Chandler, pers. comm). Losses in yield are
caused by this disease (Epstein, 1966; Howard, 1971). In addition, regulatory issues regarding
the transportation of infected plant material may impact the nursery plant industry.
International movement of infected plants is blamed for the introduction of angular leaf spot
into Greece and New Zealand (Panagopoulos et al., 1978; Dye and Wilkie, 1973). The
European Plant Protection Organization listed X. fragariae as a quarantine pest and the
FAO/IPGRI recognized it as a potential risk in international movement of strawberry
germplasm (Maas, 1995). Nursery-plant producers in the United States and Canada are
pressured to provide disease-free plants by buyers from foreign markets and by farmers who
refused to buy infected transplants.
45


11
angular leaf spot and attempts to isolate the pathogen from putative lesions was done by the
method described above.
Inoculation. Three strains of A", fragariae (Xfl 13, Xfl03, and Xfl425) were used to
inoculate plants used in field plots. Strains Xfl 13 and Xfl03 were isolated from infected
plants at Gulf Coast Research and Education Center (GCREC), Dover, FL and Xfl 425 was
obtained from A. Chase (Central Florida Research and Education Center, Apopka, FL).
Strains were cultured on WB at 24 C and long-term storage was at -70 C in 15% glycerol.
The sensitivity of the strains to copper was tested by growth on nutrient agar amended with
CuS04 (Stall et al ., 1986). Three days prior to inoculation, each of the strains was streaked
to ten plates of WB. Bacterial cells were collected from plates, suspended in sterile 0.01 M
MgS04, and the concentration adjusted to approximately 107 cfu per ml. Equal volumes of
each suspension were combined to comprise the inoculum. Plants were inoculated by dipping
bundles of 25 plants into the bacterial suspension for 30 s. Control plants were dipped into
0.01 M MgS04. Plants were placed into plastic bags and incubated 24 h at 22 C.
Field experiments. Field experiments were located at GCREC, Dover, FL, from
October 1993 through March 1994 and repeated the following season. Transplants of'Sweet
Charlie' grown in the summer nursery at GCREC-Dover were used. Raised beds were
prepared and fertilized with 10N-4P-10K at a rate of 2000 lbs per acre with one-fourth
banded before bed preparation and the remainder banded 5 cm deep in the bed center the first
season. In the second season, three-fourths pound of N and K per acre per day were applied
through the drip irrigation system. Soil was fumigated with 98% methyl bromide and 2%


52
and 6% of strains, respectively. A third endonuclease, Vsp\, generated fragments of DNA of
the appropriate size for PFGE analysis and the profiles by this endonuclease separated strains
of the A, C, and D groups identical to the previously determined groups. Analysis of the B
group with the Vsp\ endonuclease subdivided it into three subgroups, denoted B1, B2, and
B3. The B1 contained 13%, B2 contained 20%, and B3 contained 67% of the B strains
tested. Designation of RFLP groups for strains are summarized in Table 4-1.
The dendrogram derived from RFLP analysis of these strains showed the B groups
were very closely related by within 0.006 genetic distance units (Fig. 4-2). The C group was
the most closely related to the B groups, then group A, and finally group D. The D group
was the most remote from the B group at 0.01864 genetic distance units. A strain of A', c.
vesicatoria was used as the statistical outlier and fell at 0.045 genetic distance units from the
D group.
Fatty Acid Methyl Ester Analysis. The 50 strains of X. fragariae were placed into
nine subgroups based upon the MIDI 10% rule (Sasser, 1990b). The subgroups for
individual strains are summarized in Table 4-1. The majority of strains was identified as six
closely related subgroups which can be visualized quantitatively by three major acids, 16:1
w 7 cis, 15:0 Anteiso, and 15:0 iso (Fig.4-3). The 15:0 iso fatty acid comprised
approximately 34% to 54% of FAME profile. The ATTC type strain and three closely related
strains were qualitatively differentiated by the absence of palmitic acid, 16:1 w 7 cis.
The dengrogram unweighted pair analysis (Fig. 4-4) separated strains into four
clusters at the Euclidian distance of six units. The four clusters, designated FAME groups


BIOGRAPHICAL SKETCH
Pamela D. Roberts was born September 18, 1963 in Manhattan, KS. She graduated
with a Bachelor of Science degree in horticulture from Kansas State University in 1987. She
completed a Master of Science from the University of Hawaii in 1991 in the Department of
Plant Pathology. She spent 16 months as a Research Scholar at the International Rice
Research Institute in conjunction with the research at the Univerisity of Hawaii She arrived
at the University of Florida in 1991 to undertake a Doctor of Philosophy degree in the
Department of Plant Pathology.
76


12
chloropicrin at 448 kg per acre. Beds were covered with 1 mm black polyethylene mulch
immediately after fumigation.
The experimental design was a randomized complete block in a 2 3 factorial design
with four replications. The first factor had two levels: plants inoculated with either the
suspension of X. fragariae or MgS04. The three levels of the second factor were: no
chemical treatment, the label (1 *) rate of cupric hydroxide (Kocide 101 at 9.08 kg of active
ingredient per acre) plus mancozeb (Dithane DF at 6.81 kg of active ingredient per acre)
sprayed at 7- to 14-day intervals, or the reduced (0.lx) rate of cupric hydroxide plus
mancozeb sprayed at 2- to 4-day intervals.
An individual plot contained 18 plants arranged in two rows of nine plants. The beds
were spaced on 1.22 meter centers with 30 cm between rows and 30 cm within a row. Fallow
area was 3.55 m within a row and 2.44 m between rows. Pesticides were applied throughout
season as needed to control insects and fungal diseases. Chemical applications were made
by a handheld wand attached to C02-charged canister at 40 psi and pesticide was applied to
runoff. Plants were transplanted on 15 October 1993 and on 20 October 1994. Overhead
sprinkler irrigation was applied 8 hours daily for 10 to 14 days to establish transplants and
applied throughout the season as needed. Drip irrigation was installed in the summer of 1994
at GCREC-Dover and was used as supplemental irrigation in 1995.
Estimates of disease severity of angular leaf spot were made at two-week intervals.
Disease severity was expressed as percent leaf area diseased for the entire plant for each of
six plants located in the center of each plot. Progress curves were plotted as the mean of
disease severity of replicate treatments versus time. Area under the disease progress curve


technique increased detection 1,000 fold. Plants were inoculated with a rifampicin-resistant
strain and oversummered in the field at two locations in Florida. Bacteria from leaf and
crown samples were detected by nested PCR and recovery onto selective media at two- week
intervals for 92 days after planting. Daughter plants of the inoculated plants were positive
for XF by nested PCR amplification Analysis of genetic variability by fatty acid methyl
esterase (FAME) profiles divided 50 strains into 9 groups based upon qualitative and
quantitative differences. The majority (74%) of strains were placed into a closely related
group determined by cluster analysis. The profiles from restriction fragment length
polymorphism (RFLP) analysis of genomic DNA restricted by two infrequent cutting
endonucleases and separation by pulsed field gel electrophoresis grouped strains into four
groups (A-D). Another endonuclease subdivided the B group into three groups. The
dendrogram unweighted pair analysis of FAME profiles divided the population into four
groups which correlated well with the RFLP groups. Considerable diversity appears within
the species.
vii


18
Figure 2-3. The average marketable yield on strawberry plants which were inoculated with
Xauthomonasfragariae (Inoc) or not inoculated (Noninoc) for the 1994 and 1995 seasons.
Yield was the average for four replications for marketable fruit only. Different letters (A, B)
represent significant difference at P = 0.05.


23
our surveys by the sampling method used in the first season, more plants were assessed for
symptoms of angular leaf spot in the field compared to the number of plants in the random
sampling of boxes in cold storage. This probably accounts for identification of diseased plants
in nearly all of the farmers fields versus less than in 25% of boxed plants. In addition, plants
were examined until disease was found in the field which skewed the randomness of the
survey.
In field experiments, the spread from inoculated plants to noninoculated plants in
nearby experimental plots was minimal. Thus, the incidence of plants with angular leaf spot
in the field likely occurred from inoculum already present on the transplants. In the field
experiments, if inoculum was present from another source, such as debris in the soil or an
alternate host, most likely the disease on the noninoculated plants would have been more
general and appeared earlier. Disease severity on noninoculated plants in the experimental
plots was extremely low and most plants remained free of angular leaf spot. For the disease
that occurred on noninoculated plants, the inoculum was probably transported mechanically
from inoculated plants during harvests or disease readings.
Progress of angular leaf spot developed similar to curves of pathogens described on
other crops (van der Plank, 1963). A decrease in disease severity was seen about midway
through the season in 1994. This reduction, or negative infection rate, was due to growth
by strawberry plants which diluted the amount of disease relative to the total leaf area The
dilution effect of healthy plant tissue was described by van der Plank (1963). In 1995, a
negative infection rate was seen very early in the epidemic. This was because the symptoms
of the disease almost disappeared from the field as the affected leaves died and were lost. The


CHAPTER 5
DISCUSSION
Field experiments which examined epidemiological aspects of angular leaf spot on
strawberry established a decrease in yield on strawberry plants due to angular leaf spot. This
report is the first to quantify the reduction in yield due to this disease. The 8- to 10% loss
determined in these studies is much lower than the 70- to 80% loss estimated by Epstein
(1966) in fruit production fields in Wisconsin. Production in northern regions of the U. S is
usually perennial and the age of the plants was not given. The loss was reported on the
cultivar Sparkle which is not produced commercially in Florida. Therefore, we are unable
to compare our yield loss to his report Howard (1971) also reported unquantified yield
losses due to this disease but on cultivars which are no longer in commercial production In
our studies, the average yield loss observed both years was very similar despite the differences
in the disease levels and the base yields between the two seasons. In 1995, the disease
severity was 60% of the amount estimated in the previous season; however, the yield loss was
actually higher by 2% for marketable yield The 10% reduction in yield represents an
estimated loss of 1233 kg of berries per acre. Strawberry fruit has a very high cash value and
a decrease in revenue of 8- to 10% represents a significant economic loss to producers.
Strawberry producers in Florida currently try to control angular leaf spot by
application of copper compounds and by avoidance of the disease. An understanding of the
60


26
freely in the soil (Kennedy and King, 1962b). Cool temperatures (~ 20 C) are optimal for
disease symptom expression (Howard et al., 1985; Kennedy and King, 1962b) and high
temperatures (>28 C) such as those which occur in Florida during the summer months are
unfavorable. In surveys conducted in 1968, 1970, 1993, and 1994, plants which had
symptoms of angular leaf spot in the spring did not have symptoms of the disease the
following August (Howard, 1971; P. D. Roberts, unpublished). However, in 1969 mild
infections on one variety were observed in mid-August. Plants transplanted to fields did not
develop angular leaf spot. The survival of the bacterium on plants in summer nurseries in
Florida and inoculum sources other than infected transplants have not been investigated.
Identification of plants infected with X fragariae is a priority because of the ease of
transmission on infected but asymptomatic plants (Maas, 1995). International movement of
infected plants is blamed for introduction of angular leaf spot into Greece and New Zealand
(Panagopoulos et al., 1978; Dye and Wilkie, 1973). Nursery-plant producers are pressured
to provide disease-free plants by buyers in foreign countries and by farmers who refuse to buy
infected transplants. The European Plant Protection Organization (EPPO) lists X. fragariae
as a quarantine pest and has prescribed phytosanitary procedures (Maas, 1995). In the future,
regulatory issues may be of greater concern. The production of disease-free plants is essential
for control of angular leaf spot. Therefore, accurate identification of plants infected with the
bacterium is imperative. Available detection techniques are limited in their usefulness and
accuracy to detect low populations of the bacterium that may exist in asymptomatic tissue.
Xanthomonas fragariae may be identified in the early stages of leaf infection by diagnostic
translucent, watersoaked lesions viewed with transmitted light; however, older lesions may


INTRODUCTION
Xanthomonas fragariae causes angular leaf spot disease on strawberry (Fragaria
species; Fragaria ananassa Duchesne). The disease was first found in Minnesota in 1960
(Kennedy and King, 1962a) and it is currently found in many regions of strawberry
production throughout the world (Maas, 1984; Ritchie et al., 1993). Angular leaf spot was
first reported in Florida in 1971 (Howard, 1971). Dissemination of the bacterium occurred
via the transportation of infected plants (Maas, 1984; Panagopolous et al., 1978; Dye and
Wilkie, 1973). In Florida, strawberry plants which arrive from northern nurseries for
transplanting in the fall frequently have leaves which exhibit symptoms of angular leaf spot.
Typical pathogenic strains of X. fragariae can be isolated from the lesions. A diagnostic
symptom of the disease is the translucent appearance of lesions when viewed with transmitted
light (Maas, 1984). A vascular collapse of the plant from systematic invasion by the
bacterium has been described in California (Hildebrand et al., 1967).
The epidemiology of angular leaf spot is mostly unknown in fields in Florida where
strawberry production is an annual crop. Transplants are obtained each season from nurseries
in Canada and northern states; few transplants are produced in Florida. A source of inoculum
other than infected transplants has not been found Howard (1971) was unable to determine
the inoculum source for infected plants from nurseries in Florida or other states. In surveys


2
conducted in 1968, 1970, 1993, and 1994, plants which had symptoms of angular leaf spot
in the spring did not have symptoms of the disease the following August (Howard, 1971; P.
D. Roberts, unpublished). However, in 1969 mild infections on one variety were observed
in mid-August. Plants transplanted to fields did not develop angular leaf spot. Kennedy and
King (1962b) determined that the bacterium overwintered in infected leaves buried in the soil
and caused disease symptoms on plants the next year. The bacterium did not survive free in
the soil nor were any naturally occurring hosts identified in host-range studies (Kennedy and
King, 1962a). For bacteria on plant refuse to serve as an inoculum source in Florida, the
bacterium must oversummer. Optimal growth (-20 C) of the bacterium (Howard et al.,
1985; Kennedy and King, 1962b) is at temperatures cooler than normally occurs in Florida
during the summer. The survival of the bacterium on plants in summer nurseries in Florida
and inoculum sources other than infected transplants have not been established.
The effect of angular leaf spot on yield is unknown. Howard (1971) accredited some
yield losses due to angular leaf spot in fields in Florida but he did not quantify the losses. In
Wisconsin, a decrease in yield of 70 to 80% was estimated due to the disease (Epstein, 1966).
However, the production in the northern United States is a perennial, matted-row system
which differs significantly from the Florida production. A significant loss in marketable fruit
may occur due to infections of the calyx. The sepals become brown and dry and the fruit is
unmarketable because of its unattractive appearance (Epstein, 1966; Maas, 1995).
Chemical control of bacterial diseases is difficult. Antibiotics have limited
effectiveness over time since mutations to bacterium may occur and form resistant
populations (Stall and Thayer, 1962). Therefore copper compounds are frequently used to