Citation
Bacterial Citrus Canker: Molecular Aspects of a Compatible Plant Microbe Interaction

Material Information

Title:
Bacterial Citrus Canker: Molecular Aspects of a Compatible Plant Microbe Interaction
Creator:
EL YACOUBI, BASMA ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Cell walls ( jstor )
Cells ( jstor )
Diseases ( jstor )
Inoculation ( jstor )
Leaves ( jstor )
Phenotypes ( jstor )
Plasmids ( jstor )
Sumo ( jstor )
Symptomatology ( jstor )
Xanthomonas ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Basma El Yacoubi. Permission granted to University of Florida to digitize and display 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.
Embargo Date:
5/1/2005

Downloads

This item has the following downloads:

elyacoubi_b ( .pdf )

elyacoubi_b_Page_041.txt

elyacoubi_b_Page_012.txt

elyacoubi_b_Page_016.txt

elyacoubi_b_Page_062.txt

elyacoubi_b_Page_084.txt

elyacoubi_b_Page_017.txt

elyacoubi_b_Page_101.txt

elyacoubi_b_Page_083.txt

elyacoubi_b_Page_031.txt

elyacoubi_b_Page_095.txt

elyacoubi_b_Page_006.txt

elyacoubi_b_Page_066.txt

elyacoubi_b_Page_074.txt

elyacoubi_b_Page_092.txt

elyacoubi_b_Page_014.txt

elyacoubi_b_Page_086.txt

elyacoubi_b_Page_053.txt

elyacoubi_b_Page_047.txt

elyacoubi_b_Page_056.txt

elyacoubi_b_Page_080.txt

elyacoubi_b_Page_058.txt

elyacoubi_b_Page_073.txt

elyacoubi_b_Page_094.txt

elyacoubi_b_Page_067.txt

elyacoubi_b_Page_023.txt

elyacoubi_b_Page_077.txt

elyacoubi_b_Page_087.txt

elyacoubi_b_Page_093.txt

elyacoubi_b_Page_040.txt

elyacoubi_b_Page_045.txt

elyacoubi_b_Page_052.txt

elyacoubi_b_Page_048.txt

elyacoubi_b_Page_022.txt

elyacoubi_b_Page_001.txt

elyacoubi_b_Page_042.txt

elyacoubi_b_pdf.txt

elyacoubi_b_Page_103.txt

elyacoubi_b_Page_055.txt

elyacoubi_b_Page_028.txt

elyacoubi_b_Page_071.txt

elyacoubi_b_Page_051.txt

elyacoubi_b_Page_015.txt

elyacoubi_b_Page_018.txt

elyacoubi_b_Page_043.txt

elyacoubi_b_Page_026.txt

elyacoubi_b_Page_061.txt

elyacoubi_b_Page_098.txt

elyacoubi_b_Page_032.txt

elyacoubi_b_Page_008.txt

elyacoubi_b_Page_004.txt

elyacoubi_b_Page_091.txt

elyacoubi_b_Page_096.txt

elyacoubi_b_Page_029.txt

elyacoubi_b_Page_054.txt

elyacoubi_b_Page_063.txt

elyacoubi_b_Page_030.txt

elyacoubi_b_Page_005.txt

elyacoubi_b_Page_085.txt

elyacoubi_b_Page_059.txt

elyacoubi_b_Page_003.txt

elyacoubi_b_Page_097.txt

elyacoubi_b_Page_019.txt

elyacoubi_b_Page_049.txt

elyacoubi_b_Page_044.txt

elyacoubi_b_Page_002.txt

elyacoubi_b_Page_079.txt

elyacoubi_b_Page_100.txt

elyacoubi_b_Page_070.txt

elyacoubi_b_Page_024.txt

elyacoubi_b_Page_064.txt

elyacoubi_b_Page_102.txt

elyacoubi_b_Page_046.txt

elyacoubi_b_Page_076.txt

elyacoubi_b_Page_050.txt

elyacoubi_b_Page_078.txt

elyacoubi_b_Page_038.txt

elyacoubi_b_Page_037.txt

elyacoubi_b_Page_089.txt

elyacoubi_b_Page_075.txt

elyacoubi_b_Page_088.txt

elyacoubi_b_Page_068.txt

elyacoubi_b_Page_039.txt

elyacoubi_b_Page_057.txt

elyacoubi_b_Page_069.txt

elyacoubi_b_Page_036.txt

elyacoubi_b_Page_009.txt

elyacoubi_b_Page_010.txt

elyacoubi_b_Page_060.txt

elyacoubi_b_Page_065.txt

elyacoubi_b_Page_081.txt

elyacoubi_b_Page_034.txt

elyacoubi_b_Page_082.txt

elyacoubi_b_Page_035.txt

elyacoubi_b_Page_025.txt

elyacoubi_b_Page_033.txt

elyacoubi_b_Page_090.txt

elyacoubi_b_Page_021.txt

elyacoubi_b_Page_011.txt

elyacoubi_b_Page_099.txt

elyacoubi_b_Page_020.txt

elyacoubi_b_Page_027.txt

elyacoubi_b_Page_072.txt

elyacoubi_b_Page_007.txt

elyacoubi_b_Page_013.txt


Full Text











BACTERIAL CITRUS CANKER: MOLECULAR ASPECTS OF A COMPATIBLE
PLANT-MICROBE INTERACTION

















By

BASMA EL YACOUBI


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


2005

































Copyright 2005

by

Basma El Yacoubi

































To Souad, Kamal, Aziz, Mouma, Mami, Nemat, and ma petite Shemsi















ACKNOWLEDGMENTS

I would like to express my sincere gratitude to Dr. Dean W. Gabriel, my supervisor

and committee chair and for his constant support and guidance during my years as a

graduate student in his laboratory. I also extend my gratitude to Dr. John M. Davis,

(member of my supervisory committee) for his valuable advice and for welcoming me in

his laboratory each time I needed it. I also thank all other members of my committee, (Dr.

Alice Harmon, Dr. Kenneth Cline, Dr. Bill Gurley, Dr. Jim Preston) for their valuable

advice and guidance.

I thank my mother, Souad Benchemsi, who always supported me. Without her none

of this would have been possible. My husband, Nemat Keyhani, and my little daughter

Shemsi Aida Keyhani, gave to my graduate student life a new dimension, and I thank

them for that.
















TABLE OF CONTENTS
Page

A C K N O W L E D G M E N T S ................................................................................................. iv

L IS T O F T A B L E S ...................................................................... .......... .. ............. ...... v iii

LIST OF FIGURES ......... ......................... ...... ........ ............ ix

ABSTRACT .............. .......................................... xi

CHAPTER

1 A 37 KB PLASMID FROM A SOUTH AMERICAN CITRUS CANKER STRAIN
CARRIES A TYPE IV SECRETION SYSTEM ESSENTIAL FOR SELF-
M O B IL IZ A T IO N ............................................................................... ....................

In tro du ctio n .................................................................................................... ..... .
M materials and M methods ....................................................................................... 5
Bacterial Strains, Plasmids and Culture Media................. ............................5
M arker Integration M utagenesis....................................... ......................... 5
Plasm id Conjugal Transfer Techniques...................................... .....................6
Recom binant DN A Techniques....................................... .......................... 7
P lant Inoculation s ....................................................... 7
R esu lts .............................. .. ...................7..........
The Type IV Secretion System Found on pXcB is Required for Self-
M obilization ............. ............ ........ ....... ....... ... ......... 7
Involvement of the TFSS of pXcB in Pathogenicity of Xca B69 ......................
D iscu ssio n ....................................... ................... ............................ 10

2 IDENTIFICATION OF CITRUS GENES SPECIFICALLY RESPONSIVE TO
PATHOGENICITY GENE pthB OF Xanthomonas citri pv. aurantifolii ..................23

Introduction..................................... .................................. .......... 23
M material and M methods ................................................... ..... .............................. 26
Plant and M icrobial M aterial..................................... ........................ .. .......... 26
B bacterial C ounts ........................................ .... ....... .... ....... 27
M icroscopy ................................... ............................27
Differential Display-Reverse Transcriptase PCR.............................................28
Suppressive Subtractive Hybridization (SSH) Library Construction..................28
N northern Blots .................................... ..... .......... ...... ........ .. 29
R everse N northern B lots ............................................... ............................ 29









Statistical A analysis ...................................... .............................30
R e su lts ...................................... ........................................................ ... ............... 3 1
Macroscopic Disease Phenotype of Citrus Leaves Inoculated with X c.
aurantifolii B69 and Its Mutant Derivative BIM2 Lacking the Pathogenicity
Gene pthB .................................... ..................................... 31
PthB-Dependent Transcriptional Reprogramming Induced upon Infection with
X ca ................... ......... ...... ... ... .. .... ...... .. ........... ........... . 3 2
Construction of Two Libraries Enriched in pthB Responsive cDNAs ..............33
Transcript Analyses of CCRs ........................................ ...............33
Identity of cDNAs Identified as Up-Regulated by the Presence of
pthB in X citri G enom e ................................ ............ .... ................34
Identity of cDNAs Identified as Up-Regulated by X citri Lacking pthB ............35
Northern Blot Analysis of Representative CCRs.............. ..... .............36
Microscopic Phenotype of B69 and BIM2 Inoculated Leaves..........................36
Discussion .............. .... ... ......... .. ........... ........... ...............38
PthB Induces Cell Division and Cell Expansion in Citrus Leaves .................39
PthB Induces the Expression of Cell Wall Remodeling Enzymes.................40
Enod8 and SAH7/LAT52 are a Link Between Canker Symptoms Development
and Nodule Organogenesis and Pollen Tube Growth Respectively ................42
PthB Induces Up-Regulation of a Tonoplast Aquaporin..................................44
PthB Induces Up-Regulation of Two Components Involved in Vesicle
T trafficking ............. ............................................ ........ ........ .. ........ .... 44
Hormone Pathways are Possibly Involved in Canker Symptoms
Development .................................... .......................... .... ...... 45
Conclusions and Future Prospects .................................... ............................. ....... 47

3 CHANGES IN SUMO CONJUGATION ARE ASSOCIATED WITH CITRUS
C A N K E R D ISE A SE ......................................................................... ................... 66

Intro du action ...................................... ................................................ 6 6
M materials and M methods ............................................................... ..........................69
Plant Inoculations ................... ...... ............... ... ............ 69
Bacterial Strains and Culture M edia......................................... ............... 70
M arker Integration M utagenesis...................... ... .......................... 70
B ioinform atics ......................................................... ........................ 7 1
Protein Extraction and W western Blotting............................................... 71
R esu lts .................. ......... ....... ............................................. ..... 72
SUMO Conjugation Profiles are Altered in X citri-Infected Leaves .................72
SUMO Conjugation Profiles in Infected Leaves are Partially PthB Dependent.73
SUMO De-Conjugation Observed at 7 days Following Infection with B69 and
BIM2 is Dependent on a Functional Type III SecretionSystem.................74
D discussion ............... ........... .......................... ............................74









APPENDIX

A LIST OF PLASMID AND STRAINS..................................................................83

B NORTHERN BLOT ANALYSIS OF CCRS .................................. ...............85

L IST O F R E F E R E N C E S ........................................................................ .....................86

B IO G R A PH IC A L SK E T C H ..................................................................... ..................91
















LIST OF TABLES


Table page

2-1 List of putative CCR identified by DD-PCR. ........................ 49

2-2 List of CCRs confirmed by reverse northern blot analysis.....................................50

A-i List of strains and plasmids used in this study .....................................................83















LIST OF FIGURES


Figure p

1-1 Organization of the type four secretion system (virB operon) found on pXcB
compared to other described TFS systems. ....................................................... 13

1-2 Hybridization profiles of DNA from B69 integrative mutants interrupted in virB4
of B 69 virB clusters. .............................. ............. .. ........ .. ........ .... 14

1-3 EcoRI and BamHI restriction digest profiles of plasmid pB 13.1 and plasmid
pB13.2, derivatives of pXcB0 and pXcB, respectively, and integrated in gene
virB 4 ................................................................................ 15

1-4 PCR profiles using primers AB65 and AB66 specific of plasmid pXcB. 16

1-5 Self-mobilization of pXcB derivatives is dependent on a type IV secretion system. 17

1-6 Construction of suicide vector pBY17.1 ........... ............................ ............... 18

1-7 Scheme of FLP recombinase-mediated marker eviction............... ................... 19

1-8 PCR confirmation of suicide plasmid pBY17.1 integration in gene virB4 ............20

1-9 CR confirmation of Flp-mediated eviction of pBY17.1 ............ ................21

1-10 Pathogenicity phenotype of primary and secondary exconjugants disrupted in
virB 4. ............................ ...... ................ ......... ............. ........... ............... 22

2-2 Late B69 and BIM2 phenotypes. (A) BIM2 inoculated leaves 30 dpi and (B) B69
inoculated leaves 30 dpi. ........................... .................. ............... ...... ............53

2-3 Quantification of bacterial population two days post inoculation with B69 and
BIM2. (cfu: colony forming unit), Expl: experiment 1, Exp2: experiment 2)........54

2-4 Diagram of PCR-Select cDNA subtraction........................................................55

2-5 Distribution of potential citrus canker responsive genes. ..................................56

2-6 Distribution and origin of the clones stamped on the nitrocellulose membranes
used in reverse northern blot analysis. ....................................... ............... 57

2-7 Cluster analysis of genes differentially regulated by PthB. ............ .................58









2-8 Northern blot analysis of CCR genes found differentially regulated by reverse
northern blot analysis................................................................59

2-9 Microscopic phenotype of leaves inoculated with B69 (wt) and BIM2
(nonpathogenic mutant lacking PthB). ......................................... ............60

2-10 Microscopic phenotype of leaves inoculated with B69 (wt) and BIM2
(nonpathogenic mutant lacking PthB). ........................................ ............... 61

2-11 Microscopic phenotype of leaves inoculated with B69 (wt) and BIM2
(nonpathogenic mutant lacking PthB) at 14 dpi. ................... .................. .......... 62

2-13 Quantification of leaf thickening and cell division during B69 and BIM2 infection
on D uncan grapefruit leaves ......................................................... ..... .......... 63

2-13 Microscopic symptoms of rapidly developing canker. .........................................64

3-1 Alignment of grapefruit SUMO (partial sequence) with (PopSUMO1, gi:23997054,
and AtSUM O 1, At4g26840). .............................................................................. 77

3-2 SUMO profiles of B69- and mock-challenged grapefruit leaves. ...........................78

3-3 SUMO de-conjugation occurs 7 days after infection....................... ...............79

3-4 Split leaf inoculation of Xanthomonas citri pv. aurantifolii (B69) and derivative
B IM 2 m utant. ..................................................... ................. 80

3-5 B69 mutant derivative B23.5 lacks a functional Type III secretion system. ...........81

3-6 SUMO de-conjugation at 7 dpi requires a functional TTSS.................................82

B-l Northern blot analysis of CCR genes not found differentially regulated by reverse
northern blot. .........................................................................85















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

BACTERIAL CITRUS CANKER: MOLECULAR ASPECTS OF A COMPATIBLE
PLANT-MICROBE INTERACTION

By

Basma El Yacoubi

May 2005

Chair: Dean W. Gabriel
Major Department: Plant Pathology

Canker is an important disease affecting citrus worldwide. It is caused by two

phylogenetically distinct groups of strains ofXanthomonas citri (Xc), with all citrus

cultivars being susceptible to at least one Xc strain. It is known that canker-causing

xanthomonads carry at least one pathogenicity gene of thepthA (of Asiatic X citri pv

citri) gene family, which is required for causing canker on citrus. However little is

known of the host molecular events leading to canker. Our goal was to understand host

molecular mechanisms underlying disease development, and identify bacterial

components related to phylogeny or pathogenicity of canker-causing xanthomonads.

First we identified on plasmid pXcB of the South American strain X citri pv

aurantifolii B69, a pathogenicity island composed of previously identified pathogenicity

genepthB and a type IV secretion system (TFSS). This TFSS was shown to be required

for self-mobilization of pXcB, which led us to propose that natural horizontal transfer of

apth host-specific pathogenicity gene may account for the two phylogenetically distinct









groups of strains, (the Asiatic and the South American group of strains), causing canker

symptoms on citrus.

Second, we investigated plant responses to PthB using differential display PCR and

suppressive subtractive hybridization techniques. We identified forty-nine genes that

were differentially regulated when RNA expression profiles of leaves inoculated with

Xca B69 were compared to those of leaves inoculated with a B69 mutant carrying a

disrupted pthB. Among these were genes predicted to be involved in cell expansion,

protein modification, biotic/abiotic stress responses and cell-wall metabolism.

Finally, we focused on one canker-responsive gene with strong similarity to the

small ubiquitin like modifier (SUMO) from Arabidopsis. Analysis of B69 mutant strains

lacking PthB or the type III secretion system (TTSS) component, HrpG, revealed PthB-

dependent and TTSS dependent/PthB-independent changes in SUMO conjugation

profiles after infection with B69.

The genes and cellular processes that we identified reflect the molecular events

leading to disease development. They contribute to the general aim of understanding the

mechanisms underlying the variety of diseases caused by compatible interactions

between xanthomonads and their host plants.














CHAPTER 1
A 37 KB PLASMID FROM A SOUTH AMERICAN CITRUS CANKER STRAIN
CARRIES A TYPE IV SECRETION SYSTEM ESSENTIAL FOR SELF-
MOBILIZATION

Introduction

The genus Xanthomonas is comprised of strains that exhibit a high level of host-

specificity; over 125 different pathogenic variants (pathovars) ofX. campestris have been

described that differ primarily in host range (Bergey, 1994). Host specificity in

Xanthomonas can be due to gene-for-gene interactions involving avirulence genes that

act in a negative fashion to limit host range (Keen, 1990; Gabriel, 1999; Leach and

White, 1996), but also can be due to positive acting factors that condition host range in a

host-specific manner. For example, pthN, avrb6 ofX. campestris pv. malvacearum (Yang

and Gabriel, 1996), opsXofX. campestris pv. citrumelo (Kingsley et al, 1993) andpthA

of X. citri pv. citri (Swarup et al., 1991 and 1992) act as positive effectors of host range.

Interestingly, although a clonal population structure is observed among strains within

many pathovars (Gabriel et al, 1988), some pathovars are comprised of phylogenetically

distinct groups that have an identical host range and cause identical disease symptoms.

Examples include 1) common bean blight, caused by two groups of strains (X. phaseoli

and X campestris pv. phaseoli var. fuscans) that are only 20% related by DNA-DNA

hybridization (Hildebrand et al., 1990); 2) bacterial spot of tomato and pepper, caused by

two major groups of strains within X campestris pv. vesicatoria (Jones et al., 2000) that

are less than 50% related by DNA-DNA hybridization (Stall et al., 1994), and 3) citrus

canker disease, caused by two groups of strains that are only 62 -63% related by DNA-









DNA hybridization (Egel et al., 1991). Strains with 70% or greater DNA-DNA

relatedness are usually defined as single species (Wayne et al., 1987). The question arises

however, as to how phylogenetically diverse strains can cause identical diseases on an

identical range of hosts.

To date, all pathogenic xanthomonads examined require hrp genes (reviewed by

Alfano & Collmer, 1996, 1997; He, 1998; Cornelis & VanGijsegem, 2000) to cause

disease. These genes encode a type III secretion machine that is close contact-dependent

(Marenda et al., 1998) and used to inject highly adapted effector proteins into both host

and nonhost cells (Silhavy, 1997; Kubori et al., 1998; and Jin and He, 2001). These

effector proteins elicit the diverse programmed phenotypes of the plant hypersensitive

response (HR) and various pathogenicity responses. The hrp (hypersensitive response

and pathogenicity) injection system is thus appropriately named, and it is also highly

indiscriminate, injecting whatever effector proteins are available, even some from animal

pathogens (Anderson et al., 1999 and Rossier et al., 1999). If identical hrp effectors are

available within two phylogenetically distinct xanthomonads, they can cause the same

disease symptoms, provided both strains are compatible (able to multiply in the host) and

both carry functional hrp systems. For example, pthA was transferred from X citri to X

campestris pv. citromelo and converted the latter strain from a leaf-spotting strain to a

strain with ability to cause citrus canker disease (Swarup et al., 1991). PthA appears to be

an effector protein that is critical for citrus disease symptoms and is likely injected by X

citri into citrus cells, causing hyperplastic cankers (Duan et al, 1999).

Citrus canker disease is caused by two phylogenetically distinct and clonal groups

of Xanthomonas strains; each group contains subgroups that are distinguished on the









basis of host range (Brunings and Gabriel, 2003). The first phylogenetically distinct

group is the Asiatic group, named Xanthomonas citri pv citri ex Hasse (syn = X

campestris pv. citri Dye pathotype A and X axonopodis pv. citri Vauterin, Xca-A). The

second phylogenetically distinct group is the S. American group, named X citri pv.

aurantifolii Gabriel (syn = X campestris pv. citri Dye and X axonopodis pv. aurantifolii

Vauterin, Xca-B). Both groups cause identical citrus canker disease symptoms circular,

water soaked raised lesions, that become dark and thick as canker progresses (Graham et

al., 2004; Stall and Civerolo, 1991; Gottwald et al, 2002; Brunings and Gabriel, 2003).

Significantly, pthA or homologues are present in every Xanthomonas strain tested that

causes citrus canker disease, and have not been found present in xanthomonads isolated

from citrus that do not cause canker (Gabriel, 1999; Cubero and Graham, 2002). Prior to

this work, twopthA homologues, namedpthB, andpthBO were found on two separate

plasmids (pXcB and pXcB0, respectively) of a S. American canker strain (B69). Plasmid

pXcB carrying the functional homologuepthB, was then found to be readily cured from

B69 (Yuan and Gabriel unpublished, and Brunings, A.M., 2004 M.S. thesis University of

Florida). Readily cured plasmids are often mobilizable by conjugation. Since Asia is

considered to be the center of origin of citrus canker disease, and since Asiatic canker

strains are more widespread in S. America than S. American canker strains, it was of

interest to determine if pXcB could transfer horizontally. pXcB was found to horizontally

transfer in-vitro and in plant (Yuan and Gabriel unpublished) from the S. American

strain B69 to the Asiatic strain B21.1 lacking a functional pthA, restoring its capacity to

cause canker (Yuan and Gabriel unpublished). Presence of the type III effectorpthB on a









self-mobilizing plasmid might explain the creation of the entire S. American group of

canker strains, and why they are phylogenetically distinct from the Asiatic group.

pXcB was fully sequenced (NC_005240, gi32347275), and besides genepthB, a

complete Type IV secretion system (TFSS) was also found on the plasmid (Brunings,

A.M., 2004 M.S. Thesis, University of Florida). TFSS are defined on the basis of

homologies between the A. tumefaciens T-DNA transfer system, the conjugal transfer

system Tra, and the Bordetellapertussis toxin exporter, Ptl (Winams et al., 1996 and

Christie, 1997). Most members of the TFSS family function primarily to mobilize DNA,

either from bacteria to bacteria (bacterial conjugation system) or from bacteria to

eukaryotic cells (Agrobacterium oncogenic T-DNA transfer system) (Burs, 1999). In

addition, several bacterial pathogens utilize conjugation machines to export effector

molecules during infection. Such systems are said to be Type IV "adapted" conjugation

or secretion systems, for their involvement in pathogenicity. Many non-plant pathogens

such as Bordetella pertussis, Legionella pneumophila, Brucella spp. and Helicobacter

pylori use a type IV "adapted" conjugation system to secrete effector proteins to the

extracellular milieu or the cell cytosol (Burns, 1999; Christie, 1996; Christie and Vogel,

2000). Type IV systems are composed of products with homology to the Agrobacterium

virB operon (Vogel, 2000). Sequence similarity analysis revealed that the Type IV

secretion system of pXcB encodes twelve open reading frames, ten of which contained

high sequence similarities to genes of previously described virB operons as well as

similar relative positions within the cluster (Brunings, A.M., 2004 M.S. Thesis,

University of Florida).









In order to investigate whether the TFSS found on pXcB is involved in self-

mobilization of pXcB, a plasmid derivative lacking a functional TFSS was generated in

this study and tested for its ability to self-mobilize in vitro. In addition, a B69 derivative

lacking the TFSS was generated in a non-polar fashion to address whether this system

was required for pathogenicity of B69. It was found that the TFSS of pXcB was required

for self-mobilization of the plasmid. However pathogenicity tests involving TFSS

insertional mutants were inconclusive, and it remains unknown whether this secretion

system is involved in pathogenicity ofX. citri pv. aurantifolii.

Materials and Methods

Bacterial Strains, Plasmids and Culture Media

Bacterial strains and plasmids used in this study are listed in Table 1.

Xanthomonas spp. were cultured in PYGM medium at 30C (De Feyter et al., 1990).

Escherichia coli were grown in Luria-Bertani (LB) medium (Sambrook et al., 1989).

Antibiotics were used at the following concentrations (in [tg/mL): Chloramphenicol

(Cm), 35; Kanamycin (Kn) 12.5 or 25 (when used to grow Xanthomonas or E. coli

respectively); Spectinomycin (Sp) 35 and Streptomycin (St) 35.

Marker Integration Mutagenesis

Gene-specific knockout mutations of Xanthomonas were created by triparental

matings. An E. coli DH5a strain carrying an internal fragment of the target open reading

frame (ORF) cloned in suicide vector pUFR004 was used as donor. A DH5a strain

carrying pRK2013 was used as the helper. A single crossover in the exconjugates results

in duplication of the internal fragment at the integration site, and also results in

interrupting the target gene with the vector. To disrupt virB4, a PCR-generated, 270 bp









internal fragment of virB4 (virB4270) was cloned in pGEM-T Easy and recloned in

pUFR004 creating pBY13.

Plasmid Conjugal Transfer Techniques

Plasmid transfer by triparental mating from E. coli strains HB101 or DH5a to

various Xanthomonas strains, using helper strain pRK2013 were performed essentially as

described in De Feyter and Gabriel (1991). For plasmid transmission experiments on

artificial media, overnight cultures of E. coli strains grown without antibiotics were

mixed with 50X concentrated overnight, mid-log phase cultures of Xanthomonas strains,

grown without antibiotics. Drops (10 [l each) of recipient donor and helper cells were

placed on PYGM agar medium one after the other and without antibiotics. In each case

excess liquid was allowed to absorb into the plate before addition of the next cell type.

The mating plates were incubated at 30C overnight, and the spots were then streaked on

PYGM selection medium supplemented with the appropriate antibiotics.

In Xanthomonas to E. coli matings, B69 carrying pB13.2 (pBY13 integrated in

virB4 of pXcB) or B69 carrying pB13.1 (pBY13 integrated in virB4 of pXcBO) were used

as donor strains (in independent matings) with DH5a as the recipient strain. After

selection against Xanthomonas on MacConkey agar (DIFCO laboratories, Detroit MI,

USA) with 35 [tg/mL chloramphenicol, DH5a exconjugants were screened for the

presence of pBY13.2 or pBY13.1 by DNA mini-prep analysis. In E. coli to E. coli

matings, DH5a/pBY13.2, DH5a/pBY13.1 and DH5a/pBIM2 (pYY40.10 integrated in

pthB) were used as donor stains in independent matings with HB 101 as recipient.

For frequency of transfer assays from one E coli strain to another, donor and

recipient strains were grown overnight at 37 C to an O.D. 600nm of 0.5. Twenty









microliters of each culture were combined in a 1.5 ml Eppendorf tube containing 160 [tl

of LB and grown overnight at 37 C. Cells were then resuspended in 1 ml of LB, pelleted

and then serially diluted on medium containing chloramphenicol and streptomycin to

select for HB 101 transconjugants. All conjugation experiments were performed at least

twice with duplicate samples in each experiment, and the numbers were averaged.

Recombinant DNA Techniques

Plasmid and total DNA were prepared from Xanthomonas as described by Gabriel

and De Feyter (1992). E. coli plasmid preparation, restriction enzyme digestion, alkaline

phosphatase treatment, DNA ligation, and random priming reactions were performed

using standard techniques (Sambrook et al., 1989). Southern hybridization was performed

using nylon membranes as described by Lazo and Gabriel (1987).

Plant Inoculations

All citrus plants (Citrusparadisi 'Duncan', grapefruit) were grown under

greenhouse conditions. Plant inoculations involving all citrus canker strains were carried

out under quarantine at the Division of Plant Industry, Florida Department of Agriculture,

Gainesville. Bacterial cells were harvested from log phase cultures by centrifugation

(5,000 x g, 10 min.), washed once and resuspended in sterile tap water or distilled water

saturated with calcium carbonate to 108cfu/mL. Inoculations were performed by pressure-

infiltration into the abaxial leaf surface of the plants. Experimental inoculations were

repeated at least three times.

Results

The Type IV Secretion System Found on pXcB is Required for Self-Mobilization

Gene virB4 of the TFSS cluster of pXcB was chosen as target for marker

insertional mutagenesis (Figure 1-1). For that, a 270 bp integral fragment of virB4









(virB4270) was cloned in pUFR004 (pBY13) and used in triparental matings to generate

virB4 insertion mutants. Southern blots were used to verify integration events in the

resulting transconjugants. These results demonstrate the existence of two copies of virB4

in the B69 strain (using virB4270 as probe Figure 1-2). One copy was carried by pXcB (as

determined by sequencing) and was absent in the cured strain B69.4 [Rifamycin resistant

strain cured of plasmid pXcB but carrying plasmid pXcBO, Yuan and Gabriel,

unpublished (Lane 3)]. A second putative copy, carried by pXcBO, was maintained in

B69.4 (Lane 3). Marker insertion resulted in two categories of exconjugants. Exconjugant

strain B13.1 appeared to carry an interruption of the putative virB4 of pXcBO (virB4o)

(Lane 7), while exconjugants B13.2, B13.4 and B13.5 appeared to carry interruptions of

the virB4 gene of pXcB (Lanes 4, 5 and 6).

Plasmids pB13.1 and pB13.2 of strains B13.1 and B13.2 (marker interruptions in

the virB4 homologues found on pXcBO and pXcB, respectively) were further analyzed

for their ability to transfer to E. coli. Matings with and without the helper strain resulted

in DH5a exconjugants carrying plasmids that were chloramphenicol resistant, indicating

that both plasmids were still mobilizing. Restriction enzyme digests of plasmid DNA

extracted from the Xanthomonas (B13.2) to the E. coli exconjugant (DH5a/pB13.2),

corresponded to the expected profile of pXcB integrated with pBY13 (Figure 1-3).

Restriction enzyme digests of plasmid DNA extracted from DH5a/pB13.1 did not

corresponded to the profile expected for a pXcB insertional derivative. Therefore, p13.1

is a derivative of a second native plasmid of B69, smaller in size than pXcB and inserted

in a putative virB4 copy. These results were confirmed by PCR using primers specific to

pXcB. As shown in Figure 1-4, when pB13.2 was used as template with pXcB specific









primers AB65/AB66 a 2014 bp band was obtained, while non specific bands were

obtained when pB 13.1 was used as template

The ability of pB13.1 and pB13.2 to self-mobilize was then analyzed by

performing matings from DH5a to E. coli HB101. Using DH5a/pBIM2, and

DH5a/pBIM6 [pBIM6 is a derivative of pXcB where pUFR004 was inserted in a non-

ORF region, (Yuan and Gabriel, unpublished)] as a control, transfer of pBIM2 and

pBIM6 from DH5a to HB101 was found not to require the presence of a helper strain

and the transfer frequency was 7x10-03 and 6.6x10-05 per donor, respectively. By contrast,

E. coli to E. coli transconjugants harboring pB13.1 or pB13.2 were only recovered when

matings were performed in the presence of a helper strain (Figure 1-5). These results

indicated that the self-mobilization capacity of pXcB depended on the presence of an

intact virB cluster.

Involvement of the TFSS of pXcB in Pathogenicity of Xca B69

Non-polar knock out mutants of virB4 were generated using marker insertion

followed by FLP recombinase mediated marker eviction. Plasmid pBY17.1 was

generated so that a virB4 homology region was flanked by two FRT recognition sites

(See Figure 1-6 for illustration). After marker integration of suicide vector pBY17.2 into

primary transconjugants, the FLP recombinase plasmid pJR4, was used to evict the

marker, and generated non-polar secondary transconjugants (See Figure 1-7 for

illustration).

Several primary transconjugants (before FLP-mediated eviction of marker) (Figure

1-8) as well as secondary transconjugants (after FLP mediated eviction of marker) were

tested for integration events in a virB4 homologue using PCR. Bacterial cells directly









from the selection plates were used as template for PCR (Figure 1-9). PCR positive

colonies were then grown in liquid culture and tested for pathogenicity on citrus. In all

cases, primary exconjugants showed a decrease in pathogenicity while, unexpectedly,

secondary exconjugants lost their potential to trigger canker disease on citrus (Figure 1-

10). When the secondary exconjugants used in pathogenicity assays were tested by PCR

for presence of pXcB it was found that the plasmid and therefore gene pthB were lost

upon culturing.

Discussion

The putative TFSS of pXcB (Brunings A.M., M.S. Thesis, University of Florida

and Brunings and Gabriel, 2003) was functionally investigated to determine its

involvement in plasmid transfer as well as in pathogenicity of B69. To investigate the

role of this TFSS in plasmid transfer, gene virB4 was marker-interrupted and by

consequence the whole system rendered dysfunctional. Self-mobilization experiments

revealed that pXcB relied on a functional TFSS to self-mobilize. In the process a second

putative virB4 homologue was identified on a second plasmid of B69, pXcBO. pB13.1,

carrying a single insertion in virB40 of pXcBO and pB13.2, carrying a single insertion in

virB4 of pXcB were each able to mobilize from B13.1 and B13.2, respectively, to DH5ca

in biparental matings (without helper strain), indicating that the two putative virB systems

co-existing in B69 might be compensatory.

The characterization of pXcB as a self-mobilizing plasmid carrying a TFSS and

gene pthB suggests that the canker causing and phylogenetically distinct South American

strains may have arisen from horizontal gene transfer of an "ancestral" pthA member.

This horizontal transfer likely would have occurred from an Asiatic Xanthomonas citri









strain to a compatible TTS system-carrying xanthomonad residing on the same host. B69

was indeed shown to carry a functional TTS system required for pathogenicity (see

Chapter 3). The type IV secretion system together with pthB on pXcB of S. American

Xanthomonas citri strains can therefore be considered an "auto-mobile" pathogenicity

island (Hacker et al., 1997), capable of spreading among compatible bacteria by

horizontal gene transfer.

Since pXcB from the South American strain is smaller, yet very similar to pXAC64

from the Asiatic strain, pXcB could be a deletion derivative of pXAC64 (Brunings and

Gabriel 2003). However, while many genes on pXcB were found to be similar to genes

on pXAC64, there were differences significant enough to conclude that a simple deletion

cannot account for pXcB. More likely, several independent events were probably

responsible for its divergence away from pXAC64.

Horizontal gene transfer is proposed to be a major mechanism explaining rapid

genetic diversification in bacteria (Falcow, 1996; Syvanen and Kado, 1998; Lawrence

and Roth, 1999). It has been proposed to explain the apparent enigma of why pathogens

carry dispensable avirulence genes (Yang and Gabriel, 1996 and Gabriel, 1999). For

example, avrBs3 of Xanthomonas campestris pv. vesicatoria was found on a mobilizing

plasmid carrying copper resistance, and therefore wide horizontal transfer of avrBs3 to X

campestris pathovars may be due to coincidental linkage with copper resistance (Stall et

al, 1986, Yang and Gabriel, 1996).

The TFSS of pXcB was also analyzed for its involvement in pathogenicity. Primary

exconjugants carrying a marker integration in virB4 showed a decrease in pathogenicity

while non-polar secondary exconjugants, resulting from marker eviction of the suicide






12


plasmid, lost all pathogenicity. This was then found to be possibly due to a loss of pXcB

upon curing of secondary transconjugant strains. Another explanation is the presence of a

large insertion vector in the native plasmid decreasing the copy number in the population.

Further examination of the TFSS is necessary to access its role in pathogenicity if any.







13


Orf 06
B2B3 B4 B5 B7 B6 B8 B9 B10 Bll BI
X.aa (pXcB) virB .4 2 : 4 :4i. I: I4 0 ir> = c 0 Self-mobilization


A. tumefaciens (pAtC58) avhB [:::::::::
Ti-plasmid virB -:-:-:-:-:-:-:-:-:-


Conjugal transfer
Transfer of T-DNA


Figure 1-1: Organization of the type four secretion system (virB operon) found on pXcB
compared to other described TFS systems. ORF 106 shows no similarity to any
virB cluster gene ofAgrobacterium tumefaciens and is shown as an insertion.












69 69.4 13.2 13.4 13.5 13.1

9qr4 4virB40 (pXcBO)
9..-

6.6 --m -virB4(pXcB)

4.4






Figure 1-2: Hybridization profiles of DNA from B69 integrative mutants interrupted in
virB4 of B69 virB clusters. Total DNA was digested with HindIII and probed
with a 32P-labelled 270 bp internal fragment of virB4. The same fragment was
used as a homology region for integration of suicide vector pBY13. B13.2,
B13.4 and B13.5 were marker integrated in virB4 of pXcB, and B13.1 was
marker integrated in virB4 of pXcBO. Hind III digestion results in splitting
one restriction fragment harboring the targeted region into two hybridizing
fragments. Therefore there are two bands hybridizing to the virB420o probe in
the wild type strains, while there are three bands in the insertional strains. The
only band hybridizing to the virB4270 in the B69.4 lane corresponds to a
putative virB4 copy present on a second native plasmid of B69. Indeed pXcB
was lost upon curing in B69.4 and therefore one hybridizing band is lost.












EcoRI


BamHI


pBIM2 p13.2 p13.1 pBIM2


p13.2 p13.1


Figure 1-3: EcoRI and BamHI restriction digest profiles of plasmid pB 13.1 and plasmid
pB13.2, derivatives of pXcBO and pXcB, respectively, and integrated in gene
virB4.


23

9.4
6.6

4.4

2.3
2.1












B69 B69.4 pB13.1 pB13.2


Figure 1-4: PCR profiles using primers AB65 and AB66 specific of plasmid pXcB.
Plasmid DNA isolated from DH5a/pB13.1, DH5a/pB13.2, and total DNA
isolated from B69 and B69.4 were used as templates. [AB65: CAG CCG
CAA GTG TCT CAG GTC; AB66:GGC AAG AAA CCG TCC GAG TA
(Tm 560C)]. When B69.4 and pB13.1 DNA are used as template in the PCR
reaction non-specific bands of low intensity are the resulting products. When
B69 and pB13.2 (both derivatives of plasmid pXcB) are used as template, a
specific band of 2014 bp is the resulting product of the PCR reaction. ( ):
Amplification fragment specific to pXcB when AB65/AB66 primers are used.






















S1%



^-o

4 0%-


-1:
Y
I~ti
%!
~p~i
~t~i~l
i,


pB13.1


pB13.2


pBIM2


Plasmid Frequency of transfer to HB101
( in donor strain DH5a) (per input donor)
p13.1 0
p13.2 0
pBIM2 7.08E-03
pBIM6 6.57E-05


Figure 1-5: Self-mobilization of pXcB derivatives is dependent on a type IV secretion
system. (A) Mobilization of pXcB derivative, pBIM2 and pB 13.2 and pXcB0
derivative pB 13.1 from E. coli DH5a to E. coli HB101. Matings were carried
with and without helper strain carrying plasmid pRK2013, and HB 101
transconjugants were selected on LB supplemented with streptomycin and
chloramphenicol. Each selection plate was separated in two sections. Results
of matings with helper strain are shown on the left section, and results of
matings without helper are shown on the right. Matings: (a) DH5a/pB 13.1
with HB101; (b) DH5a/pB13.2 with HB 101; (c) DH5a/pBIM2 with HB 101.
(B) Frequency of transfer of pXcB derivatives, pB 13.2, pBIM2, pBIM6 and
pXcBO derivative, pB 13.1 from E. coilDH5a into E. coli HB101.










FRIKn PCR fragment

FRTKnF FRTKnR
PCR frag men t from temp late
pKD4 was cloned in pGEMT- ez
PCR fragment internal to virB4
RsrIII BclI
pBY1.1 BY13 BY14

\. PCR fragment from temp late
EcoRI HindlI pXcB was cloned in pGEMT- ez



pBY3.1 \
(pUFR012 FITKn) BY2.1

in dlI
Eco BclH RsrIII Bcll
RsrIII




pBY17.1
(pBY3.1 ::viMB4


RsrIII BclI

Figure 1-6: Construction of suicide vector pBY17.1. PCR was used to amplify an internal
virB4 fragment using primers BY13 (gatcaggatcctatgcgcctcgttgaggt) and
BY14 (cggtccgtcagtcagtcagagctctgaccaggtagtgcagga). RsrIII and BclI
restriction sites were incorporated in the primer sequences respectively in the
forward and reverse primer. The RsrIII-BclI fragment was used as the driver
for homologous recombination and was cloned between FLP sites in
pUFR012 (pUFR004 derivative carrying kanamycin resistance). FLP sites
were obtain by PCR using plasmid pKD4 (gi:15554332) (1.5 Kb fragment).
Primers FRTKn F (gaattcgctgcttcgaagttcctatac) and FRTKn R
(aagcttatcctccttagttccaattcc) carried an EcoRI and a HindIII site for subcloning
from pGEMT-ez (Promega) into pUFR012.






















virB4 F bes03 22mer M13R bes04
FLP recombinase I-
plasmid vr4 inserted vector
virB4

virB4 F bes03 bes04

800bp 700bp 800bp




Figure 1-7: Scheme of FLP recombinase-mediated marker eviction. Suicide plasmid
pBY17.1 is marker integrated in gene virB4 via homologous recombination,
generating a virB4 disruption. The light blue box represents the homology
region targeted for recombination, and is found duplicated after insertion of
the suicide plasmid pBY17.1 in pXcB. After transformation of the B69
derivative carrying pXcB::pBY17.1 (primary transconjugant) with plasmid
pJR4 carrying a FLP recombinase gene, pBY17.1 is evicted (secondary
transconjugants). pJR4 [derived from pFLP (gi: 1245114) (Ready and Gabriel,
unpublished)] is cured by culturing secondary transconjugants on PYGM
supplemented with 5% sucrose. VirB4F and bes03 and bes04 are virB4
specific primers and their locations are shown by arrows. 22mer and M13R
are primers specific to the polylinker region of suicide vector pBY17.1 and
are used to verify integration and eviction events. FRT sites recognized the
FLP recombinase are symbolized by yellow circles and flank the internal
fragment with homology to virB4 cloned in pBY17.1. The green boxes
symbolize DNA stretches carried over during sub-cloning steps.










22mer/bes03 M13R/BES04 M13R/bes03DS

1 2 3 4 5 6 7 8 9









Lane 1,4,7: B69, Lane 2,5,8: B18.12, Lane 3,6,9: B18.15
viB4 F bes03 mer M13R bes04 bes03DS

vector pBY17.1
virB4

Figure 1-8: PCR confirmation of suicide plasmid pBY17.1 integration in gene virB4. 22
mer (gttttcccagtgacgacg) and M13R ( agcggataacaatttcacac) are primer
specific to the polylinker of pBY17.1. bes03(catcttggatcgtgcgtt) bes03DS,
bes04 (catgttgctgagcatctt) and virB4 F(ggtaccacccatttgaaaacgtgtcc) are gene
specific primers. Lanes 1,4,and 7, B69 bacterial cells were used as template
source for the PCR. Lanes 2, 5 and 8, B18.12 (primary transformant with
pBY17.1 inserted in virB4), bacterial cells were used as template source for
the PCR. Lanes 3, 6 and 9, B 18.5 (primary transformant with pBY17.1
inserted in virB4) bacterial cells were used as template source for the PCR.
Primer combinations used in each lane are indicated in the figure. PCR bands
resulting from using virB4-based primers in combination with suicide vector
based primers are specific to an insertion in the targeted region and should not
appear when the wild type strain is used. The light blue box represents the
homology region targeted for recombination, and is found duplicated after
insertion of the suicide plasmid pBY17.1 in pXcB. FRT sites recognized the
FLP recombinase are symbolized by yellow circles and flank the internal
fragment with homology to virB4 cloned in pBY17.1. The green boxes
symbolize DNA stretches carried over during sub-cloning steps.












BES03/BES04 M13R/bes04 virB4F/bes04
1 2 3 1 2 3 1 2 3
Lane 1: B69
Lane 2:B18.12
Lane 3:B18.12-1





B18.12 virB4F bes03 2mer M13R bes04

suicide vector inserted
virB4 virB4 F bes bes04
B18.12-1
800bp 700bp 800bp
bes03/bes04: 470bp
M13R/bes04: 1300 bp
virB4 F/bes04: 950bp (wt), or 2500bp after FLP



Figure 1-9: PCR confirmation of Flp-mediated eviction of pBY17.1. Genes specific
primers (bes03, bes04, virB4F and vector based primers were used in
appropriate combinations. B69 and B18.12 (primary transconjugant with
pBY17.1 inserted in virB4) were used as negative and positive control for the
suicide vector integration respectively. B18.1 2-1 is the secondary
transformant resulting from suicide vector eviction from primary
transconjugant B18.12. Bacterial cells from selection plates were used as
template source for PCR. The expected size of each PCR band is indicated in
the figure. The light blue box represents the homology region targeted for
recombination, and is found duplicated after insertion of the suicide plasmid
pBY17.1 in pXcB. FRT sites recognized the FLP recombinase are symbolized
by yellow circles and flank the internal fragment with homology to virB4
cloned in pBY17.1. The green boxes symbolize DNA stretches carried over
during sub-cloning steps.

































(secondary exconjugants, after eviction ofpBY7.1) Picture on the left was
Figure 1-10: Pathogenicity phenotype of primary and secondary exconjugants disrupted
in virB4. B 18.12; primary exconjugant (virB4::pBY17.1). B 18.12-1
(secondary exconjugants, after eviction of pBY17.1) Picture on the left was
taken 7 days post inoculation. The two pictures on the right were taken 15
days after inoculation. Note the delay phenotype of primary transconjugant
B 18.12, and the total loss of pathogenicity of transconjugant B18.12-1.














CHAPTER 2
IDENTIFICATION OF CITRUS GENES SPECIFICALLY RESPONSIVE TO
PATHOGENICITY GENE pthB OF Xanthomonas citri pv. aurantifolii

Introduction

Many studies on plant-pathogen interactions have dealt with incompatible

interactions using model plant systems (for example see Malek et al., 2000). Emphasis

has been on dissecting signaling pathways of resistance mechanisms, with few studies

considering signaling pathways resulting in diseases of crop plants (Kazan et al., 2001).

Therefore, the molecular events at the origin of disease induction by microbial effectors

of pathogens remain obscure.

Many Gram-negative, phytopathogenic bacteria rely on a Type III secretion system

(TTSS) to deliver effector proteins into the plant cells (He et al., 2004). Inactivation of

the TTSS of bacterial species that utilize such a system results in loss of pathogenesis

indicating that the proteins (named type III effectors) delivered by the TTSS are required

for bacterial virulence (Rohmer et al., 2004). Most type III effectors identified to date

were originally discovered and characterized by their avirulence function (Avr), while

only few are recognized pathogenicity factors [PthA from X citri (Swarup at al., 1991),

AvrB6 from X campestris pv. malvacearum (Yang et al,. 1996), AvrXa7 from X oryzae

(Bai et al., 2000) and DspA from Erwinia amylovora (Gaudriault et al., 1997)]. A limited

number of type III effectors have been assigned proven or putative biochemical function

(Collmer et al., 2000; Rohmer et al, 2000; Chang et al., 2004) and for a subset of these

(principally avirulence effectors), a plant protein or cellular process has been identified as









a possible target for pathogenesis (Rohmer et al, 2004 and Chang et al. 2004). In two

cases, a bacterial effector-triggered plant phenotype has been shown to be required for

pathogenesis. In the case of pathogenicity factor DspA, a member of the P. syringae

AvrE family, its induction of reactive oxygen species release by the host cell has been

shown to be required for successful colonization (Venisse et al., 2003). While in the case

of pathogenicity factor PthA, a member of the Xanthomonas AvrBs3/PthA family, its

induction of cell division and /or cell expansion is required for pathogenesis (Swamp et

al., 1991)

In this study the compatible interaction between citrus and Xanthomonas citri (X.

citri pv. aurantifolii syn. X. axonopodis pv. aurantifolii) was examined. Probably

originating in Southeast Asia, citrus canker has now spread to most citrus producing areas

of the world and causes severe economical losses (Civerolo, 1994). All canker strains

induce similar disease phenotypes, including water soaked lesions, formation of large

hyperplastic erumpent pustules (cankers) on all aerial plant parts, and rupture of the

epidermis with accompanying cell death (Swamp et al., 1991 and Duan et al., 1999).

Specific members of the avrBs3/pthA gene family are required by strains of

Xanthomonas citri to cause cankers on citrus (Swarup and Gabriel, 1989; Swarup et al.,

1990, Swarup et al., 1991). Members of the avrBs3/pthA gene family are found in many

xanthomonads (Gabriel, 1999; Vivian and Arnold 2000), and all citrus canker strains

examined carry multiple members of the gene family (Gabriel, 1999). All Xanthomonas

avrBs3/pthA members described to date are 90-97% identical in DNA sequence and are

characterized by 1) a series of 12.5-25.5 almost identical 34 amino acid repeats in the

center of the protein that determines host specificity, pathogenicity and/or avirulence









phenotype (Herbers et al., 1992, Yang et al., 1994, Zhu et al., 1998, and Yang et al.,

2000), 2) three C-terminal nuclear localization signals (Yang and Gabriel, 1995; Van den

Ackerveken et al., 1996 and Szurek et al., 2001) and 3) a C- terminal acidic region

considered to function as an eukaryotic transcriptional activator (Zhu et al., 1998, Zhu et

al., 1999, Yang et al., 2000, Szurek et al, 2001).

Sequence and functional analysis of members of the avrBs3/pthA gene family

showed that these proteins are type III effectors, acting in the plant nucleus potentially as

transcriptional regulators (Yang and Gabriel, 1995, Zhu et al., 1998, Zhu et al., 1999,

Yang et al., 2000, Szurek et al., 2001). When the pathogenicity genepthA from Xcitri

was transiently expressed in susceptible plant cells (by Agrobacterium infection or

particle bombardment delivery), it elicited canker-like pustules, indicating thatpthA alone

was sufficient to trigger canker symptoms (Duan et al., 1999). Unlike pthA and its active

homologues in other X citri pv. citri and X citri pv. aurantifolii strains, avrBs3 is not

required for pathogenicity ofX c. pv. vesicatoria (Bonas, 1989). However, it was found

to induce a subtle hypertrophy in the mesophyll of leaves inoculated with slow-growing

strains ofX. c. vesicatoria, concomitant with the up-regulation of 13 plant genes (Marois

et al., 2002). Taken together these results indicate that members of this gene family are

able to induce transcriptional reprogramming in both susceptible and resistant plant cells.

In this study, the citrus canker system was used to probe the functions ofpthB,

another member of the avrBs3/pthA gene family, that is isofunctional withpthA in

eliciting host-specific symptoms. A comparative analysis of gene expression in citrus

leaves inoculated with the wild type X c. aurantifolli strain B69 (carrying pthB) and an

Xca mutant derivative carrying a defective (marker interrupted) pthB was performed.









Methodological approaches in this analysis included differential display-reverse

transcriptase PCR (Liang and Pardee, 1992), suppressive subtraction hybridization, and

microscopy. Forty-six clones of citrus canker responsive genes belonging to several

broad categories of cellular functions were identified as being specifically regulated by

pthB. These categories included genes identified to be involved in cell wall loosening and

growth, water homeostasis and vesicle trafficking. In addition evidence is presented for

the involvement of hormone signaling in canker disease development.

Material and Methods

Plant and Microbial Material

Bacterial strains and plasmids used in this study are listed in Appendix A. B69 and

BIM2 were grown on PYGM (De Feyter et al. 1990) supplemented with 35 mg/L

spectinomycin and 35 mg/L spectinomycin plus 35 mg/L Chloramphenicol. All citrus

plants (Citrusparadisi 'Duncan', grapefruit) were grown under greenhouse conditions.

Plant inoculations involving all citrus canker strains were carried out under quarantine at

the Division of Plant Industry, Florida Department of Agriculture, Gainesville. Bacterial

cells were harvested from log phase cultures by centrifugation (5,000xg, 10 min.),

washed (IX) and resuspended in sterile tap water or distilled water saturated with

calcium carbonate to an OD600nm of 0.6-0.7, unless stated otherwise. Inoculations were

performed by pressure-infiltration into the abaxial leaf surface of the plants. Experimental

inoculations were repeated at least three times.

For differential display-reverse transcriptase PCR (DD-PCR) experiments and

construction of the suppressive subtractive libraries (SSH), inoculations were performed

following a split leaf model. Strain B69 was inoculated on one side of the mid-vein;

while BIM2 was inoculated on the opposite side of the mid-vein, in order to control for









leaf to leaf variations. Tissue was harvested 0, 2 or 7 days post inoculation (dpi)

depending on the experiment.

Bacterial Counts

B69 and BIM2 bacterial cells were normalized to an OD600 of 0.7 and infiltrated as

described previously. At 0 and 2 dpi, a total of 9 discs (0.28 mm in diameter) from 3

leaves (3 discs per leaves) were harvested for each treatment and ground in lml of tap

water. After serial dilution, the bacterial populations of wild type strain B69 and mutant

strain BIM2 were counted. Bacterial cell count determinations represent the average of

three replicate experiments.

Microscopy

Fresh, tender and half-expanded leaves were inoculated with a high inoculum of

B69 or BIM2. At 0, 2, 7 and 14 dpi, leaf samples of an area of approximately 6 mm2 were

harvested and fixed in 2% glutaraldehyde in phosphate buffer saline (PBS) for 48 hr at 40

C. They were then washed three times for 15 min each and fixed in 1% buffered osmium

tetroxide overnight at 40 C. This was followed by one wash in PBS for 10 mi and by two

washes in distilled water. A stepwise dehydration was conducted after these washes using

ethanol (25%, 50%, 75%, 95% and 100%) for 10 min each step, followed by three

washes in acetone for 15 min each. Samples were then infiltrated at room temperature in

30% acetone/EMbed (Electron microscopy sciences, Pennsylvania) for 1 hr, followed by

50% acetone/EMbed for 1 hr and 70% acetone/EMbed for 2 hr. Samples were

subsequently incubated in 100% EMbed overnight at room temperature to complete the

infiltration and polymerized in fresh 100% EMbed in a 75 C oven overnight.









Differential Display-Reverse Transcriptase PCR

Two and seven days after inoculation, leaf tissue was harvested, pooled and frozen

in liquid nitrogen for total RNA extraction as described (Chang et al., 1993). Potential

canker responsive (CCR) cDNAs were cloned as fragments by differential display-

reverse transcriptase PCR (DD-PCR) of mRNA using 48 primer combinations (Liang and

Pardee, 1992) with the RNAimage kit from Genhunter (Nashville, TN, USA).

Suppressive Subtractive Hybridization (SSH) Library Construction

For polyA mRNA isolation, leaves were frozen in liquid nitrogen and stored at -800

C until extraction. PolyA mRNA was isolated from leaves using the FastTrack mRNA

isolation kit (Invitrogen) according to the manufacturer's protocol. SSH was constructed

using a cDNA subtraction kit (Clontech PCR-Select, Palo Alto, CA). For construction of

the forward subtraction library (FS), the tester was chosen to be the pool of mRNA

isolated from B69 inoculated leaves at 2 dpi while the driver was chosen to be the pool of

mRNA isolated from BIM2 inoculated leaves, and therefore, the FS was enriched in

transcripts up-regulated by pthB. For the reverse subtraction library (RS), transcripts

isolated from BIM2 inoculated leaves (2 dpi) were used as tester, and therefore, while the

driver was chosen to be the pool of mRNA isolated from B69 inoculated leaves, the RS

library was enriched in transcripts up-regulated in the absence ofpthB.

Potential differentially regulated clones were sent for sequencing to the

Interdisciplinary Center for Biotechnology Research (ICBR) core at the University of

Florida. Putative functions were assigned based on annotation derived by BLAST

analysis.









Northern Blots

For RNA sample preparation, NorthernMax Formaldehyde Load Dye was used as

recommended by the manufacturer (Abion Austin, TX) with 5-10 |tg of RNA. Samples

were loaded on a denaturing formaldehyde agarose gel (1%) and electrophoresis was

conducted at 5 V/cm. RNA was blotted on GeneScreen Plus hybridization transfer

membrane (NemTM Life Science Products, MA) using 20X SSC as transfer buffer.

Hybridization and washes were done as recommended by the manufacturer (Ultrahyb,

Ambion Austin, TX). Probes were made with DECA primeTMII (random priming),

(Ambion Austin, TX) as recommended.

Reverse Northern Blots

For reverse northern blots, cDNAs identified by DD-PCR or SSH were amplified

using vector primers and purified using Qiaquick columns in plate format (Qiagen,

Valencia CA). Membrane arrays were made essentially as described by Desprez et al.,

(1998). cDNAs were arrayed onto Hybond N+ membranes (Amersham Biosciences,

Piscataway, NJ) using a 96-pin colony replicator (V&P Scientific, San Diego CA). Six

replicate arrays were generated and used to analyze transcript abundance of a subset of

potential canker responsive genes or CCRs. Each cDNA was spotted in two locations,

and several cDNAs were represented by more than one clone. Three replicate membranes

for each treatment (B69 or BIM2 infection) were used in hybridization experiments (total

of six membranes or 3 pairs). Each membrane was probed with radiolabelled cDNA

synthesized from RNA isolated from one of three split leaf-experiments conducted, 2 dpi.

Each membrane pair was one of three biological replications. Signal intensities were

statistically compared after normalization.









For probe preparation, first strand cDNA probes were prepared from 10 |tg of total

RNA by reverse transcription using MMLV-RT (Gibco-BRL, Gaithersburg MD) in the

presence of 32P-dCTP. Unincorporated nucleotides were separated from first strand

cDNA using Sephadex G-50 columns (Amersham Pharmacia Biotech, Ithaca NY) and

quantified using a liquid scintillation counter (Beckman Coulter, Fullerton CA). Pre-

hybridization, hybridization and low and high stringency washes were carried out at

65C. Membranes were exposed to phosphorimager screen for visualization. Spot

intensities (called volumes) on the membrane arrays were quantified using a BioRad

Molecular Imager FX run with the associated Quantity One software (Bio-Rad

Laboratories, Inc. Hercules, CA). Data were imported into Microsoft Excel (Microsoft

Corp., Redmond, WA, USA) for further analysis.

Statistical Analysis

A mixed model analysis (SAS Proc Mixed) was run on the log base 2 transformed

(normalization) local background adjusted volumes. cDNAs that did not exhibit a mean

value greater than 120 from either treatment were not included in the analysis. The linear

model used included replication (three biological replications), treatment (B69 treated or

BIM2 treated) and gene (CCRs or Citrus Canker responsive clones). Least square means

for the treatment by gene interaction were saved and used to form by-gene contrasts

between treatments. Significance of these contrasts was controlled for an experiment-

wide alpha level.









Results

Macroscopic Disease Phenotype of Citrus Leaves Inoculated with X. c. aurantifolii
B69 and Its Mutant Derivative BIM2 Lacking the Pathogenicity GenepthB

Xanthomonas strains B69 (wt) and its nonpathogenic mutant derivative BIM2,

carrying a marker integration in genepthB (pthB::pUFR004), were inoculated at high

levels (OD = 0.7) on tender half-expanded leaves of new flushes of Duncan grapefruit

and the corresponding induced disease phenotype analyzed. At day two post-inoculation,

no symptoms were visible and no macroscopic differences were observed among leaves

inoculated with tap water, B69 or BIM2. By seven days post-inoculation, leaves that were

mock inoculated showed no symptoms, while leaves inoculated with the wild type strain

B69 showed a whitish canker phenotype, typical of South American canker disease. On

the abaxial side of the leaf, the entire inoculated area became raised, with a soft, velvet-

like appearance, while a few individualized pustules appeared at the margins of

inoculated areas. Pustules possibly corresponded to areas where bacteria were infiltrated

at low density (Figure 2-1, A and B). On the adaxial side of the leaf, no raising was

apparent; instead some yellowing developed. This rapid symptom development is

typically observed when a high inoculum is used on fresh, young expanding leaves.

By contrast, at 7 dpi, no major symptoms were visible on leaves inoculated with

BIM2. Limited raising of the epidermis occurred at the margins of some inoculation

zones, with development of minimal pustule-like structures reminiscent of those seen in

canker (Figure 2-1, C and D). These symptoms were not observed in mock-inoculated

leaves. BIM2 inoculated leaves ultimately displayed attenuated canker phenotypes after

thirty days (Figure 2-2). This is possibly due to the week canker-inducing activity of

pthBO, the second pthA homologue found in the B strain, B69.









PthB-Dependent Transcriptional Reprogramming Induced upon Infection with Xca

A small scale DD-PCR was conducted to compare transcript levels of leaves

inoculated with B69 to those of leaves inoculated with BIM2 at two and seven dpi. To

maximize the homogeneity and the intensity of the response, B69 and BIM2 were

inoculated at high levels (OD600 of 0.7). In order to minimize leaf-to-leaf variation, a

split-leaf inoculation strategy was used. An average of fifteen leaves (from three trees)

were inoculated with B69 on one side of the mid-vein and with BIM2 on the other side.

Two and seven days after inoculation, half-leaves were harvested, pooled into "B69

treated" or "BIM2 treated" samples and RNA extracted from both samples. Since B69

(carrying pthB and pthBO) differs from BIM2 (carrying pthB::pUFR004 and pthBO) only

by the presence of a single effector, PthB, differentially regulated transcripts (named

citrus canker responsive or CCRs) were PthB responsive. Transcripts identified by DD-

PCR appeared differentially regulated as early as two days post-inoculation despite a

complete lack of symptoms. Twenty cDNAs were identified by DD PCR (Table 2-1),

including six with homology to biotic or abiotic stress response genes (CCR20.2 to PR-1

proteins and CCR9.5, CCR15.1 to PR-5 proteins, CCR2.2, CCR17.2 to peroxidases and

CCR12.1 to catalases). One cDNA, CCR6.4 displayed homology to cell wall remodeling

enzymes of the cellulase family. CCR25.1 was homologous to the small ubiquitin like

modifier SUMO.

To remove the possibility that potential changes in transcript level were due to

differences in the number of bacteria present in B69 inoculated leaves compared to BIM2

inoculated leaves, both bacterial populations were monitored at 0 and 2 dpi. B69 and

BIM2 bacterial populations were found to be comparable with almost no growth

observed during the first two days post-inoculation (Figure 2-3). Bacterial growth at 2 dpi









will occur if bacteria are inoculated at lower initial levels (OD600 of 0.3-0.4) (data not

shown). However, when inoculated at lower levels, growth of BIM2 is very poor (data

not shown and discussed later).

Construction of Two Libraries Enriched in pthB Responsive cDNAs

Following the same split leaf scheme as for the DD-PCR experiment, forward and

reverse libraries were constructed by suppressive subtraction hybridization (see Figure 2-

4 for illustration of the methodology), extending the collection of putative CCRs. The

forward subtraction library (FS) was constructed to be enriched in transcripts up-

regulated by PthB while the reverse subtraction library was constructed to be enriched in

transcripts up-regulated in the absence of PthB (see Materials and Methods for design of

the SSH).

Approximately 500 clones were sequenced and annotated using homology based

searches. Figure 2-5 illustrates the distribution of CCRs for each of the FS and RS

libraries according to their putative function. Categories representing genes of unknown

function (8 %) and genes involved in cell growth and division (10%) were found more

frequently in the forward library (up-regulated in the presence ofpthB) compared to the

reverse library (1 % and 2 % respectively), while genes in the category representing

abiotic and biotic stress responses were found more frequently in the reverse library (15%

vs 6% in the FS)

Transcript Analyses of CCRs

cDNAs from each of the forward (131) and reverse (161) libraries, as well as 20

clones identified by DD PCR (total of 312 cDNAs) (Figure 2-6) were chosen for reverse

northern-blot analysis. CCRs homologous to genes of known function were preferentially

selected. Six replicate arrays were generated as described in materials and methods, and









used to analyze transcript abundance of a subset of potential CCRs. Three membranes per

treatment were probed with radiolabelled cDNA synthesized from RNA isolated from

three split leaf-experiments, 2dpi with B69 and BIM2 (three biological replicates). For

each experiment, inoculated leaves were sampled from new and older flushes, were half

to fully expanded and were all tender (minimal cuticle). Signal intensities were

statistically compared after normalization as described in material and methods. Forty-six

clones were identified as differentially regulated at (p < 0.05) (Figure 2-7). Only fifteen

out of forty-six clones were found up-regulated in the absence of PthB, while the

remaining thirty-two were found up-regulated by PthB. Ratios of transcript abundance

were calculated for each cDNA. Ratios ranged from -3.5 to +34.5 (- sign indicating over-

expression of the gene in the absence of PthB and + sign indicating an up-regulation in

the presence of PthB) (Table 2-2).

Identity of cDNAs Identified as Up-Regulated by the Presence ofpthB in X. citri
Genome

Of the forty-six clones identified as differentially regulated, all but four clones

showed significant (e-value >2e-03) matches with sequences in available databases (Table

2-2). Thirty CCRs out of forty-six were found up-regulated by the presence ofpthB in the

bacterial genome i.e. up-regulated in B69 infected leaves compared to BIM2 infected

leaves. These are listed in Table 2-2.

Cell growth. Twelve clones were highly similar to genes involved in cell growth

(cell wall loosening and expansion): CCR339 was similar to cellulases; CCR1511,

CCR113 were similar to expansins; CCR889 was similar to mannanendo-1,4-beta

mannosidases; CCR571 and CCR1453 were similar to pectate lyases and CCR313 was

similar to tonoplast aquaporins (TIP3). Another clone of interest, CCR575, had homology









to the early nodulin gene Enod8 (predicted cell wall localized esterase). An additional

gene represented by CCR 109, CCR959 and CCR501 had homology to a secreted cell-

wall-associated pollen-specific allergen of the ole e 1 family (SAH7).

Giberellic acid pathway. Two CCRs had homology to the GAST1 (GA

responsive genes of unknown function) family of genes.

Vesicle trafficking. Several clones had homology to proteins involved in vesicle

trafficking. For example, CCR673 had homology to a small GTPase of the Rab family

(RAB8B, Vernoud et al. 2003), and CCR1258 had homology to the beta COP protein of

the COPI complex.

Unknown function. Another eight clones found up-regulated had either no

significant homology to any sequences in available databases or had sequence homology

to genes of unknown function.

Identity of cDNAs Identified as Up-Regulated by X citri Lacking pthB

Sixteen CCRs out of forty-six were found up-regulated by X citri lacking pthB

i.e. up-regulated in BIM2 infected leaves as compared to B69 inoculated leaves. These

are listed in Table 2-2.

Cell growth. CCR243, was the only BIM2 up-regulated gene involved in cell

wall metabolism. CCR243 is homologous to caffeic acid methyl transferases and is

involved in phenylpropanoid metabolism.

GA pathway. CCR 237 was homologous to cytP450 ent-keuren oxidase and

CCR105 was homologous to another cytP450 (possibly ent-kautenoic acid oxidase).

Protein modification and stability. For example, CCR409 had homology to

RD21a, a drought responsive cysteine proteinase, and CCR915 had homology to the

small ubiquitin modifiers (SUMO).









Transport. CCR1339 and CCR1435 were homologous to a mitochondrial import

inner membrane translocase and a monosaccharide-H+ symporter, respectively.

Unknown function. Another four clones found-up regulated in BIM2 infected

leaves had either no significant homology to any sequences in available databases or had

sequence homology to genes of unknown function.

Northern Blot Analysis of Representative CCRs

Expression of several candidate CCR genes identified by reverse northern blot

analysis was evaluated by northern blot analysis. Leaf tissue from split-leaf inoculations

using B69 and BIM2 were harvested and processed for RNA extraction. Several labeled

cDNA fragments were used to probe RNA blots (Figure 2-8). As in reverse northern blot

analysis, clones corresponding to expansion, cellulase, SAH7/LAT52, GAST1, Enod8 and

pectate lyase showed high levels of induction.

Microscopic Phenotype of B69 and BIM2 Inoculated Leaves

In order to characterize the microscopic phenotype of B69 and BIM2 infected

leaves, leaf discs mock inoculated and infected with BIM2 or B69 were harvested and

processed for light microscopy analysis (Figure 2-9, 2-10, 2-11 and 2-12). Leaves were

pooled as fast-responding to canker when disease symptoms were fully developed by

seven dpi (see figure 2-1, A and B). Leaves were pooled as slow-responding to canker

when disease symptoms were fully developed by 12 to 14 dpi.

Slow-responding leaves. At 2 dpi B69, BIM2 and mock inoculated leaves looked

identical at both the macroscopic and the microscopic level. At 7 dpi, while mock and

BIM2 inoculated leaves showed no phenotypic signs at both the microscopic or

macroscopic level (data not shown and Figure 2-9, A), the first signs of canker became

visible on B69 inoculated leaves, i.e. regions of darker green color around the veins and









slight swelling. At the microscopic level, B69 leaves showed high levels of cell division

occurring across all the inoculated area (Figure 2-9, compare B, Cto A). Intense cell

expansion and cell division phase resulted in complete filling of the air spaces of the

spongy mesophyll in B69 infected leaves (Figure 2-9, compare B, C to A). The number

of mesophyll cells from the abaxial to the adaxial epidermis more than doubled compared

to the day 0 control or day 7 BIM2 inoculated leaves, while some cells almost tripled in

size (Figure 2-10, compare A, B and D to C, and Figure 2-12). At later stages (14 dpi),

increased raising of the epidermis and whitish coloration with soft or velvety appearance

were observed at the macroscopic level. These phenotypes coincided with a phase of

increased cell expansion (data not shown and Figure 2-11, compare A, B to C and D).

While areas of cell division were still visible, a significant subset of cells became much

larger and the leaf dramatically thickened (twice that of the control leaf, see Figure 2-12).

A critical preliminary conclusion from these analyses indicated that the earliest visible

canker phenotype was mainly due to cell division, with a moderate cell expansion, while

late onset phenotypes were due to scattered but dramatic increases in cell expansion.

Fast-responding leaves. Macroscopic analysis indicated that cell expansion was

the primary phenotype with very little cell division occurring (Figure 2-13, compare B, C

and D to A). Furthermore, several areas of cell lysis, were visible immediately under the

abaxial epidermis.

Bacterial growth in B69 and BIM2 infected leaves. Canker visible symptoms

(cell division, cell expansion and resulting cell death) appeared necessary for B69 growth

as very few bacteria were visible in BIM2 infected tissue 14 dpi while numerous pockets









of bacteria were seen in B69 infected tissue (Figure 2-11, compare B to C and D and data

not shown).

Discussion

In this study, we have used macroscopic and microscopic phenotypic analysis in

combination with targeted gene discovery techniques to understand how the

pathogenicity factorpthB, ofX. c. aurantifolii belonging to the avrBs3/pthA gene family

elicits host-specific citrus canker symptoms in a compatible plant microbe interaction.

The nonpathogenic mutant BIM2, lacking pthB was used in combination with the wild

type strain, Xca B69, to study the specific effects of PthB on the plant cell transcriptome.

A split-leaf inoculation experimental design was used to minimize leaf-to leaf variations

in gene expression. In addition, bacterial cells were inoculated with high inoculum to: (1)

ensure near saturation of infection sites, (2) maximize the synchronicity of the host

response, and (3) artificially normalize the levels of bacterial populations (wild type and

mutant) present during early infection stages of the plant leaves (up to 2dpi). In order to

obtain a collection of genes potentially responsive to PthB, two complementary

techniques, DD-PCR, and forward and reverse SSH, were used to enrich for: (1)

transcript up-regulated when PthB is secreted in plant cells by Xcitri and (2) those up-

regulated in the absence of a functional PthB. Transcript analysis of a subset of 312

clones was conducted using reverse northern blot technique. Statistical analysis was used

to identify a list of forty-nine PthB responsive genes and differential regulation for a

subset of these was verified by northern blot analysis. Northern blot analysis was also

conducted on several CCR that did not show differential regulation by reverse northern

blot analysis (Appendix B). Several of these showed differential regulation when

northern analysis was used suggesting a better sensitivity than with reverse northern









analysis. This implies that the subtraction libraries contain additional CCR that need

identification.

PthB Induces Cell Division and Cell Expansion in Citrus Leaves

When inoculated on citrus leaves, Xanthomonas citri pv. aurantifolii was able to

cause cell division and cell expansion, consistent with previous reports onpthA-induced

phenotypes (Duan et al. 1999). Quantification of the three visible phenotypes of canker

i.e. cell division, cell expansion and the resulting thickening of the leaves was difficult

due to (1) the heterogeneity of the cells in the spongy mesophyll and (2) the

heterogeneity in distribution of the abundant air filled spaces in citrus leaf tissue.

Therefore, as first approximation of the phenotype, quantification measurements were

performed on areas where cellular activity was the most dramatic (areas of intense cell

expansion, cell cycle activity and thicker leaf areas). Analysis of PthB induced symptoms

over time revealed that the earliest visible phenotype associated with canker was cell

division in the infected spongy mesophyll, whereas heterogeneous but massive cell

expansion was observed at later stages of the infection. Interestingly, when canker

developed rapidly, i.e advanced canker symptoms at 7 dpi versus 12 to 14 dpi, symptoms

of cell division were found to be reduced compared to slower developing canker. In

addition, cell expansion was the major phenotype, primarily affecting mesophyll cells

directly under the abaxial epidermis layer. This supports the hypothesis that the primary

cellular mechanism affected by PthB alteration of the plant cell transcriptome is the

integrity of the cell wall and the induction of cell expansion. In turn, cell division could

either be: (1) a consequence of modification associated with cell expansion (e.g. changes

in cell volume) and (2) due to a second and distinct effect of PthB. However, because cell

expansion constituted a major phenotype in both rapid and slow developing canker,









induction of cell expansion may be the primary consequence ofpthB functions in the

plant cell. Furthermore, a specific set of genes with homology to genes involved in cell

growth were identified as responsive to PthB.

PthB Induces the Expression of Cell Wall Remodeling Enzymes

In order to understand PthB-induced phenotypes on citrus leaves, we have

identified a set of forty-six genes (CCRs) specifically regulated by the presence of this

effector in the plant cell. Consistent with the PthB-induced morphological phenotypes,

several CCRs were homologous to genes involved in plant cell wall modifications.

Expansins. Among these PthB-up-regulated plant genes, two were homologous to

ca-expansins. The role of the expansion gene family in wall loosening (polymer creep) and

cell expansion has been widely documented (Cosgrove, 2000). Expansins are

extracellular proteins that facilitate cell wall expansion probably by altering hydrogen

bonds between hemicellulosic wall components and cellulose microfibrils (Coscrove,

1998). These can act alone to induce cell wall extension in vitro, however, in vivo they

act with a suite of enzymes capable of restructuring the plant cell wall (Cosgrove, 1998).

Consistent with this, several CCRs homologous to genes associated with cell wall

remodeling were also identified.

Pectate lyases. Among CCRs associated with cell wall remodeling, CCR571 and

CCR1453 were similar to pectate lyases (PLs). These enzymes are involved in hydrolysis

of wall polymers, via cleavage of de-esterfied pectin, thereby facilitating cell expansion

(Carpita and Gibeaut, 1993 and Domingo et al., 1998). Although the role of bacterial

secreted PLs in cell wall degradation is well known (Collmer and Keen, 1998), the role of

endogenous plant PLs in development has not been extensively examined. In pollen,









plant PLs are thought to initiate the loosening of the cell wall enabling the emergence and

growth of the pollen tube (Cosgrove et al., 1997). PLs also mediate cell wall breakdown

in the style's transmitting tissue, allowing penetration of the pollen (Taniguchi et al.,

1995, Wu et al., 1996). Thus, induction of plant PLs by PthB can help account for aspects

of the disease phenotype.

Cellulases. Another PthB up-regulated CCR was homologous to the cellulase

family, another class of cell wall remodeling enzymes. Cellulases catalyze the cleavage

of internal 1,4 P linkages of cellulose and are involved in several aspects of plant

development involving cell wall modifications, including abscission, fruit softening and

cell expansion (Lewis and Koehler, 1979, and Fisher and Bennet, 1991). Relevant to

PthB induced phenotypes, it has been shown that constitutive expression of a poplar

cellulase in A. thaliana led to a significant increase in cell size (Park et al., 2003).

Beta-endo-mannanase. In addition to CCRs homologous to expansins, PL, and

cellulases, a fourth type of cell wall remodeling enzyme, a mannan endo-1,4-3D

mannosidase (endo-beta-mannanase) was also identified as up-regulated by PthB. This

enzyme catalyzes the hydrolysis of 1-4-pD mannosidic linkages in mannans,

galactomannans, glucomannans and galactoglucoomannans (Matheson and McCleary,

1985 and Matheson, 1986) and has been implicated in cell wall weakening during anther

and pollen development (Filichkin et al., 2004) and in seed ripening where it is involved

in mobilization of the mannan-containing cell walls of the tomato seed endosperm (Mo

and Bewley, 2003).

Caffeic acid methyl transferase. Only one gene involved in cell wall metabolism

was down regulated by PthB, CCR243. This clone was homologous a caffeic acid methyl









transferase (COMT), belonging to the phenylpropanoid pathway that leads to lignin

biosynthesis. Its expression has been shown to be regulated by biotic and abiotic elicitors

including infection by avirulent and virulent bacteria (Toquin et al., 2003). It is possible

that down-regulation of this enzyme relates to down-regulation of defense responses by

down-regulation of lignin deposition. This event could occur due to alterations in the

lignin content or composition. COMT down-regulation is in accord with the cell

expansion induced by PthB since mature walls lack acid-induced extension (Cosgrove,

1989). It is also interesting that fully expanded mature leaves are more resistant to canker,

whereas young leaves (one half to two-third expanded) are the most sensitive ones

(Graham et al., 2004). This is consistent with the hypothesis that PthB targets the cell

wall, inducing cell expansion ultimately resulting in disease progression.

A synthesis of our results indicates that type III effector PthB triggers the up-

regulation of an array of proteins whose combined activities induce cell wall loosening

and cell expansion. The roles of expansins, PLs and cellulases in cell wall loosening have

been shown to be complementary in other systems (Cosgrove et al.,1998; Carpita and

Gibeault; 1993, Domingo et al., 1998, Inouhe and Nevins, 1991).

Enod8 and SAH7/LAT52 are a Link Between Canker Symptoms Development and
Nodule Organogenesis and Pollen Tube Growth Respectively

Two additional classes of CCRs (CCR575 and CCR109, 959 and 501) identified

as up-regulated by PthB also support the theory that this effector targets cellular growth.

The first one, CCR575, was homologous to Enod8, an early nodulin gene associated with

the development of rhizobial nodule structures prior to nitrogen-fixation (Dickstein et al.,

1988, 1993). Enod8 has sequence similarity to exopolygalacturonase and lanatoside 15'-

O-acetylesterase (Pringle and Dickstein, 2003). Intriguingly, the up-regulation of Enod8









in response to X citri and Rhizobium suggests some common steps between nodule

formation and canker pustule formation. This is also supported by the fact that both

infections trigger cellular reprogramming events that lead to cellular growth. The

function of Enod8 is unknown, but in-vitro characterization and sequence analysis predict

that it is a cell wall localized esterase with acetylated oligo- or polysaccharides as

substrates (Pringle and Dickstein, 2004). Thus the enzymatic activity of Enod8, its cell

wall localization and involvement in both nodule and canker pustule formation point to

its involvement in modification of cell wall components during cellular growth.

The second class of CCRs reinforcing the hypothesis that the cell wall is the target

of PthB, displayed homology to SAH7 and LAT52 genes encoding for members of the ole

e I family of proteins. Originally identified as pollen allergens, members of this family

have also been found expressed in other tissues (e.g. SAH7 in leaves). A recent study of

one homologue, LAT52 (tomato), indicates that these genes may be involved in

controlling hydration and pollen tube growth (Tang et al., 2002). LAT52 interaction with

the pollen receptor kinase LePRK2 (LRR kinase) led to the hypothesis that binding of

LAT52 initiates a signal transduction pathway required for pollen germination and pollen

tube growth (Tang et al., 2002 and Johnson and Preuss, 2003). The up-regulation of a

LAT52- like gene in canker might, therefore, be part of a signaling pathway leading to

cell growth (the phenotype of both canker and pollen tube). Interestingly, pollen tube

growth, which occurs by tip extension, involves expansion and deposition of cell wall

precursors at the growing tip and requires the concerted action of endo-beta-mannanase,

expansins and pectate lyases (Marin-Rodriguez et al., 2002, Cosgrove, 1998 and

Filichkin et al., 2004), also found up-regulated during canker symptoms development.









PthB Induces Up-Regulation of a Tonoplast Aquaporin

CCR313, identified as up-regulated by PthB, displayed sequence similarity to a

tonoplast aquaporin of the TIPs family (Maurel, 1997 and 2002, and Hill et al., 2004).

Besides cell wall loosening, expansion requires extensive solute and water uptake

resulting in the formation of a prominent vacuolar compartment. This maintains the

turgor pressure that drives cell expansion (Veytsman and Cosgrove, 1998). Expansion is

thought to require high hydrolic permeability of the tonoplast in order to support water

entry into the vacuole, and tonoplast aquaporins (TIPs) play a critical role in this process

(Ludevid et al., 1992; Chaumont et al., 1998). TIPs are enriched in zones of cell

expansion (Tyerman et al. 2002) as well as in zones of active cell division where their up-

regulation is linked to vacuole biogenesis (Marty, 1997). Whether the identified tonoplast

aquaporin is indeed a marker for cell division and/or is actively involved in driving the

cellular expansion is unknown.

PthB Induces Up-Regulation of Two Components Involved in Vesicle Trafficking

Cell expansion and cell division both require deposition of new wall components

into the extending cell walls (Veytsman and Cosgrove, 1998) or into the cell plate of

dividing cells (Staehlin and Hepler, 1996 and Samuels et al, 1995). This may be achieved

via secretary processes involving vesicle trafficking. However, most genes identified here

suggest that in response to canker, cell walls are mainly extended without the building of

new cell wall components. This would imply that walls become thinner as cells expand.

This indeed has been observed at late stages of canker (Figure 2-10 compare A, B to C

and D). Although plant cell walls generally appear not to become thinner as they extend

(Veytsman and Cosgrove, 1998), expansion without new cell wall deposition could be at

the origin of the cell lysis observed in advanced canker stages.









Several CCRs identified as up-regulated by PthB are involved in vesicle

trafficking. Among these, CCR673 and CCR1258 have homology to RAB8B and beta

COP respectively. RAB8B is a member of the small GTPase gene family. The yeast

homologue of Rab8 (also named RABE see Vernoud et al., 2003) regulates membrane

trafficking to the daughter cell bud site (Salminen and Novick, 1987 and Goud et al.,

1988). Interestingly, in tomato, members of this subfamily appear to be targeted by the

Pseudomonas avirulence factor, AvrPto. This implies that in susceptible plants, AvrPto

may interfere with membrane trafficking pathways (Bogdanove and Martin, 2000). It has

been suggested that RAB8B might be involved in polarized secretion of antimicrobial

compounds (Bogdanove and Martin, 2000).

In mammals, beta COP belongs to a large complex that coats COPI vesicles

(Kreis et al. 1995). COPI vesicles transport membrane proteins and soluble molecules in

a retrograde, and possibly anterograde, direction through mammalian Golgi stacks

(Nickel and Wieland, 1997 and Harter, 1999). In plants little is known about COPI

vesicles. Recent evidence suggests that COPI-like vesicles are functional in plant

secretion and localize mainly to the Golgi apparatus as well as to the cell plate of dividing

cells (Couchy et al, 2003).

Hormone Pathways are Possibly Involved in Canker Symptoms Development

Triggering of cell expansion as well as induction of expansins and pectate lyases

constitutes a common point between the effect ofpthB on citrus (this study) and that of

the avirulence effector avrBs3 on susceptible pepper plants (Marois et al., 2002). Cell

expansion induction by both effectors share similar features; however, several plant

auxin-induced proteins of the SAUR family were found up-regulated by avrBs3 (Marois

et al. 2002). Several clones identified aspthB responsive are regulated by auxin in other









systems. These include the expansins (Catala et al., 2000, Civello et al., 1999, Hutchison

et al., 1999) and the pectate lyases (Domingo et al., 1998). In the pepper model, one of

two identified c-expansins was found up-regulated by exogenous application of auxin;

whereas a second expansion as well as a pectate lyase were not (Marois et al., 2002).

These data suggest that an auxin-independent pathway might operate under certain

conditions leading to cell expansion. In addition to a possible role of auxin in canker

disease, there is evidence for the involvement of the gibberellic acid signaling pathway in

the plant response topthB. CCRs with homology to ent-kaurenoic acid oxidase and

possibly to ent-kaurene oxidase (KO) (of GA biosynthetic pathway) (Oszewski et al.,

2002) and two clones with homology to the GAST1 family (GA induced genes) were

identified. Interestingly the GAST1 homologues were up-regulated by pthB; whereas the

putative KO and KAO were down-regulated. This may be explained by feed-back

regulation of KAO and KO expression by GA. However, feedback regulation of several

enzymes of the GAs biosynthetic pathway has been described in other systems, it has not

been reported to occur in the case of KAO and KO (Olszewski et al., 2002).

GA is known to regulate TIPs (Phillips and Huttly, 1994, Ozga et al., 2002),

expansins (Oka et al., 2001, Vogler et al., 2003, Lee and Kende, 2002, Chen and

Bradford, 2000), GAST1-like genes (Kotilainen et al., 1999 and Aubert et al., 1998),

endo-beta-mannanase (Dutta et al., 1997, Yamaguchi e al., 2001) and cellulases (Litts et

al., 1990). Therefore, PthB may act on regulatory steps upstream of GA biosynthesis. The

involvement of GA does not preclude that auxin is also involved since the latter is able to

regulate the production of the bioactive GA1 in elongating shoots (Ross et al., 2000).









Indeed, these two hormones are known to, in concert, promote cellular division and

elongation (Cleland, 2001 and Davies, 1995).

Conclusions and Future Prospects

The tight relationship between cell division and cell expansion makes it difficult

to address the question of whether cell expansion or cell division are the cellular

pathways that are altered as a downstream consequence of PthB regulating the plant cell

transcriptome. However, the following results presented here support the hypothesis that

cell wall loosening and expansion is the major plant cellular mechanism targeted by

PthB: 1) cell expansion occurs whether canker symptoms develop rapidly or slowly, 2)

genes involved in cell expansion have been identified as responsive to PthB, 3) cell

expansion is triggered by one another member of the avrBs3/pthA gene family and 4)

PthB responsive genes are involved in cell growth.

Microscopic analysis of leaves showing a slow canker symptom development

indicated that cell division is the major visible phenotype in initial infection stages, while

cell expansion remains at a moderate level. During the late infection stage however, cells

dramatically expanded leading to areas of cell lysis. It is possible that PthB induces cell

expansion and cell division by targeting several distinct cellular mechanisms. Another

hypothesis is that PthB targets cell expansion by altering cell wall composition

(loosening). This in-turn leads to a cell autonomous response that mainly involves the

triggering of cell division in the early stages and massive cell wall loosening and

expansion in later stages. The concentration of bacteria surrounding infected cells and,

therefore, the concentration of PthB protein secreted into the plant cells as well as the

physiological state of the infected tissue (for example immature expanding leaves will

readily expand) would modulate this response. When the concentration of PthB is low,









moderate expansion and the subsequent change in cell volume would lead to cell

division, while in later stages, elevated concentrations of PthB would lead to gross cell

expansion and cell lysis (Figure 2-14).

The relationship between cell expansion and cell division in plant growth and

development remains controversial. Whether growth starts by an increase in cell size,

triggering division, or whether division occurs first followed by restoration of the original

cell size (Foard, 1971 and Cleland, 2001) is mainly unknown. Studies on leaf primordial

(LP) initiation may begin to resolve this issue. Initially, since the first visible sign of a

new LP is a periclinal division in the L1 or L2 layer of the shoot apical meristem, it was

suggested that division occurs first (Steeves and Sussex, 1989). However, recent evidence

indicates that cell enlargement is the first step in LP initiation since LPs can be induced

by adding expansins either by microinjection of by up-regulation of expansion transcripts

at the shoot apical meristem (Pien et al., 2001, Fleming et al. 1997). Canker could follow

a similar pattern where cells expand first and then divide in response to expansion.

The canker phenotype is necessary for optimal growth and dispersal of X citri

(Swamp et al., 1991 and this study); therefore, induction of cell division and or expansion

are key steps in canker disease development and, unlike AvrBs3 forXcv, PthA/B confer a

benefit to X citri strains carrying it. The PTHA/B family of pathogenicity effectors may

prove to be a valuable tool in dissecting the molecular events surrounding microbe-

induced diseases since they are required for pathogenesis and can induce canker

symptoms alone. Finally, an understanding of the mechanisms by which PthB induces

canker phenotypes could help unravel the intricate relationship between cell division and

cell expansion that occurs in plant development.







49


Table 2-1: List of putative CCR identified by DD-PCR.
CCR Homology e-value
CCR23.2 Unknown protein [A. thaliana] (NP_196103.1) 2e-37
CCR24.5 Putative protein [A.thaliana] (NP_195874.1) 3e-24
CCR1.1 Putative Transposase [A. thaliana] (NP_189803.1] 4e-33
CCR22.5 3-hydroxyisobutyryl-coA hydrolase [A. thaliana] (NP_193072.1) le-22
CCR27.1 Cytochrome P450 [soybean] (T05942) 5e-41
CCR11.4 Cytochrome f [Nicotiana tabacum] (NP_054512.1) 4e-53
CCR6.2 Phosphoribosyl pyrophosphate synthase [Spinacia oleracea] 6e-25
(CAB43599.1)
CCR28.2 Putative mitochondrial carrier protein [A. thaliana] (NP_181124.1) 4e-34
CCR7.6 Copper Transport Protein [A. thaliana] (NP_200711.1) 4e33
CCR8.2 Receptor-like protein kinase-like (LRR) [A.thaliana] 6.8e-45
CCR6.4 Cellulase [sweet orange] (eC3.2.1.4) le-23
CCR28.4 Peroxidase [A. thaliana] (CAA66035.1) 2e-50
CCR12.1 Catalase [Campylobacterjejuni] (Q59296) 2e-10
CCR2.2 Bacterial-induced peroxidase [Goss hirsutum] (AF155124) 3e-26
CCR17.2 Peroxidase [Nicotiana tabacum] (BAA82306.1) 6e-63
CCR20.2 Pathogenicity-related protein la [barley] (AF245497) 2e-43
CCR15.1 Osmotin-like protein [Fagus sylvatica] (AJ298303) 2e 17
CCR9.5 Osmotin -like protein [Fragaria x ananassa] (AF1999508) 3e-61
CCR21.1 Auxin induced protein, putative [A. thaliana] (NP_176274.1) 1. 9e-3
CCR25.1 Ubiquitin-like protein [A.thaliana] (NP 194414.1) 4e-28







50


Table 2-2: List of CCRs confirmed by reverse northern blot analysis.
CCR Homology e value Ratio*
(aO.05)


Ribosomal protein
1385 30S ribosomal protein S20 (A. thaliana) gi21592469
Unknown function
1065 EST (0. sativa) gi50919279
1243 EST (A. thaliana) gi42569501
497 Putative protein (0. sativa) gi50919279
767 Putative protein (A. thaliana) gi15241855
1111 No significant homology
171 No significant homology
137 No significant homology
809 No significant homology
1139 Ring Finger Protein (A. thaliana) gi26450511
1061 Splicing factor RSZp22 (A. thaliana) gi21554419
475 Zinc finger protein (A. thaliana) gi28416541
1312 LRR receptor kinase (A. thaliana) gi42562316
Metabolism/energy
539 CytoF (N. tabacum) gi11465970
1415 RubisCO activase (malus x domestic) gi415852
1239 FIFO ATPase inhibitor protein (0. sativa) gi 52077175
1057 Hydroxymethyltransferase (A. thaliana) gi21593312
901 UMP-kinase (A. thaliana) gi2497486
1445 UMP-kinase (A. thaliana) gi2497486
33 UDP-galactose epimerase (A. thaliana) gi9758701
Transport
1339 Mitochondrial import inner membrane translocase S.U gi42568553
343 Copper T protein (A. thaliana) gi15237802
1435 Monosaccharide-H+ symporter (D. glomerata) gi30349804
Protein modification/stability
915 Small ubiquitin-like modifier (A. thaliana) gi15236885
1262 Protease inhibitor/seed storage//LTP (A. thaliana) gi42567284
1345 Aminopeptidase (A. thaliana) gi34098848
409 Putative cysteine proteinase RD21A (A. thaliana) gi22136972
GA pathway
1535 GASTI-like protein (A. thaliana) gi25406361
493 GASTI-like protein (A. thaliana) gi25406361
237 Cyt. P450 ent-keuren oxydase (Malus x domestic) gi45551401
1051 Cyt P450 (possibly ent-kaurenoic acid oxidase) (P. sativum) gi27776451
Vesicle trafficking
673 RAB 8B (Lotus corniculatus) gi1370192
1258 beta COP protein (0. sativa) gi50900798
279 Phosphatase (put. membrane trafficking factor) (A. thaliana) gi21553471


le 28

4e-1
2e-40
4e-27
3e-24





5e-07
2e 03
le07
le43

4e 53
9e-58
4e 10
4e-73
3e-38
le-10
4e 21


-3.36

4.79
-3.2
4.89
-2.41
2.95
3.60
3.09
-3.14
3.86
3.16
-2.55
3.25

-3.05
-2.60
-2.91
3.53
2.64
6.92
3.58


le-19 -2.62
4e33 4.92
2e-13 -3.36


4e-28
6e-04
2e-21
5e-32

3e-10
3e-34
7e-44
le 23


-2.53
5.35
3.09
-2.99

11.00
6.10
-3.2
-2.46


le 28 6.68
7e-20 2.38
3e-19 -2.51










Table 2-2. Continued
CCR Homology e value Ratio
(aO.05)
Cell growth (cell wall metabolism and expansion)
889 Mannan endo 1,4 beta mannosidase (0. sativa) gi34912090 4e 15 3.53
113 Alpha expansion (P. cerasus) gi13898655 4e-49 4.06
1511 Alpha expansion (P. cerasus) gi13898655 4e51 4.26
571 Pecate lyase (malus x domestic) gi 34980263 le-54 7.89
1453 Pectate lyase (A. thaliana) gi21593312 12-15 4.69
243 Caffeic acid 0-methyl transferase (C. roses) gi 18025321 6e-59 -2.39
339 Cellulase (sweet orange) gi7488904 le 23 34.53
575 Enod8 (early nodulin 8 like) (A. thaliana) gi26451820 3e-07 4.59
313 Tonoplast aquaporin gamma TIP (TIP3) (A. thaliana) gi3688799 5e-13 4.82
109 SAH7/LAT52 (ole e I allergen family) (L. esculentum) gi 295812 7e-17 3.78
959 SAH7/LAT52 (ole e I allergen family) (L. esculentum) gi 295812 5e-11 3.73
501 SAH7/LAT52 (ole e I allergen family) (L. esculentum) gi 295812 9e-21 2.75
*: A positive ratio indicates up-regulation in B69 infected tissue compared to BIM2
infected tissue. A negative ration indicates up-regulation in BIM2 infected tissue
compared to B69 infected.tissue.






52

























Figure 2-1: Phenotype of B69 and BIM2 infections on grapefruit leaves. BIM2 lacks
PthB and induces formation of very small pustule like structures, reminiscent
of canker pustules, at the edges of some inoculated areas. (A), (C) are BIM2

inoculations. Pictures were taken 7 days post inoculation.
i~ ').'- ." ., :












inoculations. Pictures were taken 7 days post inoculation.







53








.4P



A'






Figure 2-2: Late B69 and BIM2 phenotypes. (A) BIM2 inoculated leaves 30 dpi and (B)
B69 inoculated leaves 30 dpi. Note the much attenuated phenotype of BIM2
infected leaves.












1.E+06


1.E+04


1.E+02


1.E+00


Odpi 2dpi Odpi 2dpi Odpi 2dpi Odpi 2dpi


Expl


Exp2


BIM2

Expl


BIM2

Exp2


Figure 2-3: Quantification of bacterial population two days post inoculation with B69 and
BIM2. (cfu: colony forming unit), Expl: experiment 1, Exp2: experiment 2).















Citrus leaves infected Citrus leaves infected
with BIM2 with B69
Poly A+ RNA Isolation

f
AAAAAAA cDNA Synthesis AAAAAAA
Sby Reverse Transcriptase Y

AAAAAAADouble-stranded cDNA Synthesis T
-AAAAAAA AAAAAAA
Restriction Enzyme Digestion 4

Dnver withRsa I Tester

Adaptor Ligation to the Tester DNA Adaptor tor 2R
Dnver




Firnt Hybridiation
a 68 OC for 8 hrs
b m
c-- c

d Dnver d{


Second Hybridization
68 OC for 16 hrs
aIh



d{

em

SFill in the ends

b
aE-E

d_

e
PCR Amplification
using an Adaptor Pnmer

a and d No amplification
b b b No amphfication
c Linear amphfication
511 1 3'
3 5' e Exponential amphfication
3 ~ 5'

Figure 2-4: Diagram of PCR-Select cDNA subtraction. Type e molecules are formed only

if the sequence is up-regulated in the tester cDNA. Solid lines represent the

Rsa Idigested tester or driver cDNA. Solid boxes represent the outer part of

the Adaptor 1 and 2R longer strands and corresponding PCR primer 1

sequence. Green boxes represent the inner part of Adaptor 1 and the

corresponding Nested PCR primer 1 sequence; red boxes represent the inner

part of Adaptor 2R and the corresponding Nested PCR primer 2R sequence.














unknown n function
8%


metabolism/energy
9%


abiotic and biotic sires-
response
6%

/
/




no homology
24%















B) RS


ribosomal protein
20%


abiotic and biotic stress
response
1 ..







no homology
14%


Iranscription and translation
6%


cell growth and division
10%



secondary metabolism
4%
protein stability and degradation
3%


transport
4%
signaling
5%


hormone metabolism and
signaling
1%


unknown function
1%


metabolism/energy
14%


transcription and translation
6%


cell growth and division
2%

secondary metabolism
6%


protein stability/degradation
4%


ribosomal protein
28%


transport
,. 5%
signaling
3%

hor no:,ne metabolism and
signaling
2%


Figure 2-5: Distribution of potential citrus canker responsive genes.


A)FS











200


150 1

SDD
100 101
131 |SSH

50-


0





Figure 2-6: Distribution and origin of the clones stamped on the nitrocellulose
membranes used in reverse northern blot analysis.








58











539
237
1435
1385
1243
1239
1051
915
767
475
409
279
243
1415
1339
1057
959
809
171
137
113
109
33
1345
1312
1139
1061
1453
1445
1262
1065
673
575
571
497
493
343
339
313
1535
1511
901
889
501
501
1258
1111





Figure 2-7: Cluster analysis of genes differentially regulated by PthB. In green are genes
down-regulated by PthB, and in red are genes up-regulated by PthB










BIM2


B69


BIM2


B69


CCR137


GAST1


BIM2

i-? 1" .-
">0:.o J. .. .


Exp


rRNA


LAT52


S
ar


4m


yTIP



18 S


Figure 2-8: Northern blot analysis of CCR genes found differentially regulated by reverse
northern blot analysis. rRNA was used as control for loading.


2 dpi


Enod8


rRNA


2 dpi

cellulase


2 dpi


B69













































Figure 2-9: Microscopic phenotype of leaves inoculated with B69 (wt) and BIM2
(nonpathogenic mutant lacking PthB). 7dpi BIM2 infected leaves, A; 7dpi
B69 infected leaves, B, C. By In canker-infected tissue, by 7 dpi, air spaces of
the spongy mesophyll are almost inexistent. These spaces are replaced by new
divided cells as well as by cell of larger size, resulting in thickening of the
leaves.






































Figure 2-10: Microscopic phenotype of leaves inoculated with B69 (wt) and BIM2
(nonpathogenic mutant lacking PthB). 7dpi BIM2 infected leaves, C; 7dpi
B69 infected leaves, A, B, D. At 40X magnification, pockets of bacterial cells
are visible surrounding mesophyll cells of B69 infected tissue while almost no
bacteria is present in BIM2 infected tissue. Also not the areas of cell lysis in
B69 infected tissue.





































Figure 2-11: Microscopic phenotype of leaves inoculated with B69 (wt) and BIM2
(nonpathogenic mutant lacking PthB) at 14 dpi. A, B: BIM2, and C, D:B69
infected leaves. Note High levels of bacteria in B69 infected leaves compared
to BIM2 infected leaves, as well as possible wall thinning of cells in B69
infected tissue.







63




1000


C 800 -- B69
---- BIM2
600


400


200


0
0 -----------------------
0 2 7 14
30

25-

S g 20

x 15 -

il 10-
o ~o
5 -

0
0 2 7 14
Days post inoculation



Figure 2-13: Quantification of leaf thickening and cell division during B69 and BIM2
infection on Duncan grapefruit leaves. These measurements where taken on
"slow canker-developing" leaves, i.e. leaves showing high rate of cell division
when inoculated with B69. The number of cells from abaxial epidermis to
adaxial epidermis was calculated by counting the number of cells that a virtual
line perpendicular to the epidermal layers would cross. Ten lanes were used n
the analysis and the number shown are averages.



































Figure 2-13: Microscopic symptoms of rapidly developing canker. 14dpi BIM2 infected
leaves, A; 14dpi B69 infected leaves, B, C, D. Note the highly enlarged cells
the large areas of cell lysis and the absence of high rate of cell division in B69
infected tissue.























Cell division Cell expansion
Effector molecules: Effector molecules:
9 Exps, PLs, cellulase, Mannanase, TIP3,
Enod8, GAST1, RAB8B, BetaCOP



9 4......................... Cell wall loosening








V
Cell Division """"""""""""......................... Cell Expansion





Rapid multiplication ofX. citri and Canker phenotypes


Figure 2-14: Possible model for PthB effects on susceptible citrus cell showing
parallel pathways activating cell division and expansion.














CHAPTER 3
CHANGES IN SUMO CONJUGATION ARE ASSOCIATED WITH CITRUS
CANKER DISEASE

Introduction

Citrus canker is an important disease of citrus worldwide (Civerolo, E., 1984). It is

caused by several pathovars ofXanthomonas citri, which differ mainly in their host range

(Shubert et al, 2001, Verniere et al, 1998). Canker infections cause defoliation, fruit

blemishes, premature fruit drop and tree decline, resulting in severe economical losses

(Shubert et al, 2001). Considerable international regulatory efforts are implemented to

prevent the spreading of the already quarantined pathogen, with negative effects on

national and international trade of citrus (Timmer et al, 1996; Shubert et al, 2002).

Canker symptoms are characterized by erumpent corky lesions that can affect all

aerial parts of citrus trees (Shubert et al, 2002). Microscopy studies showed that canker

lesions result from hyperplasia (cell division) and hypertrophy (cell expansion) in the

spongy mesophyll tissue, where the bacteria contact plant cells (Swarup et al, 1991; Duan

et al, 1999 and Chapter 2). Ultimately, this intense increase in cellular growth ruptures

the epidermis and causes necrosis. The rupture of the epidermis is thought to be crucial

for bacterial dissemination and spread of the disease (Graham and Gottwald, 1991; Duan

et al, 1999).

A crucial step towards understanding citrus canker disease was the identification of

a pathogenicity gene, pthA, required by X citri pv. citri to cause canker on citrus (Swarup

et al., 1991). Since then, all canker-causing strains have been shown to carry at least two









members of the pthA gene family, with one copy sufficient for most or all pathogenicity

(Yang and Gabriel, 1995; Al-Saadi and Gabriel unpublished). pthA, found in X citri pv.

citri (Xcc) of the Asiatic group of strains, and pthB, found in X citri pv. aurantifolii B69

(Xca) of the South American group, have been shown to be interchangeable in their

ability to elicit canker (Yuan and Gabriel, unpublished). As forpthA ofXcc, pthB of Xca

was also shown to be required for pathogenicity on citrus (Yuan and Gabriel,

unpublished, and Chapter 2), and therefore, the B69 derivative mutant strain BIM2

lacking pthB does not elicit the typical macroscopic symptoms associated with canker

disease (Chapter 2). When transferred to other xanthomonads carrying a functional type

III secretion system (TTSS), or transiently expressed in leaf cells, pthA was found to

induce cell division, cell expansion, and rupture of the epidermis the three most

prevalent canker symptoms (Swarup et al, 1991 and 1992; Duan et al, 1999). It was

therefore concluded thatpthA alone was able to cause canker-like symptoms and that its

delivery into the plant cell relies on a functional TTSS.

Members of thepthA gene family are also found in non-canker causing strains of

Xanthomonas. Examples of genes belonging to this gene family include avrBs3 and

avrBs3-2 of Xanthomonas campestris pv. vesicatoria (Bonas et al, 1989, and Bonas et al,

1993), avrXalO and avrXa7 of Xanthomonas oryzae pv. oryzae (Hopkins et al, 1992);

along with avrB4, avrb6, and avrb7 of Xanthomonas campestris pv. malvacearum (De

Feyter and Gabriel, 1991 and 1993). Proteins encoded by members of this gene family

are 90 to 97% similar and are characterized by several structural features essential for

their function in avirulence and/or pathogenicity. Such features include 1) nearly identical

102-bp tandem repeats in their center, 2) C-terminal nuclear localization signals (NLS),









and 3) C-terminal eukaryotic acidic transcriptional activator (Herbers et al, 1992; Yang et

al, 1994; Zhu et al, 1998; Yang et al, 2000; Yang and Gabriel, 1995; Van den

Ackerveken et al, 1996, Szurek et al, 2001).

Little is known about how canker disease is initiated inplanta. In order to

understand the molecular mechanism underlying canker, a differential display PCR

experiment was conducted to identify plant genes potentially responsive to canker

(Chapter 2). At two days post inoculation (dpi), transcripts were compared between

leaves inoculated with B69 and leaves inoculated with BIM2 (B69 derivative carrying a

non-functional pthB). One clone was related to AtSUMO1 from Arabidopsis. SUMO

belongs to the ubiquitin family of proteins that are conjugated to target proteins;

however;its functions are distinct from those of ubiquitin.

SUMO conjugation has been shown to be an important regulatory step in processes

such as protein stability, subcellular localization, and response to various stresses.

SUMOylation is carried out in a ATP-dependant reaction cascade similar to the E1-E2-

E3 reactions responsible for ubiquitin conjugation (Melchior F., 2000; Kim et al, 2002;

Kurepa et al, 2003). In addition, SUMO modification has been shown to be important for

cell cycle progression in yeast. Specifically, temperature-sensitive mutants lacking a

functional SUMO conjugation pathway have been shown to arrest at the G2/M transition

(Seufert et al, 1995; Johnson and Gupta, 2001). Such work is of interest, as Xanthomonas

citri infection triggers division of mesophyll cells contacted by the bacteria.

Recent work has shown that strains of the phytopathogenic bacterium

Xanthomonas campestris pv. vesicatoria encode at least two type III effectors with

demonstrated SUMO protease activity (Hotson et al, 2003; Roden et al, 2004). Though









loss of these SUMO protease-like effectors did not affect pathogenicity on susceptible

plants, it raises the possibility that the plant SUMO conjugation pathway could be

targeted during infection by X c. vesicatoria (Hotson et al, 2003; Roden et al, 2004).

This study indicates that: 1) changes in plant protein SUMOylation profiles

occurred after host infection by Xanthomonas citri pv. aurantifolii, 2) these changes in

SUMOylation profiles were of two types, gene pthB-dependent and independent, and 3)

these changes in SUMOylation profiles did not occur following challenge with a non-

pathogenic mutant strain lacking a TTSS. Together, these data indicate that the TTSS of

Xca delivers one or more effectors that directly, or indirectly, de-conjugate SUMO from

host proteins in vivo.

Materials and Methods

Plant Inoculations

All inoculations were done with needle-less syringes on the abaxial surface of the

leaf. Plants (Citrus paradisi 'Duncan' grapefruit) were grown under greenhouse

conditions. Inoculations involving strains B69 and its derivatives were carried out in BL-

3P level containment (refer to Federal Register Vol.52 no 154, 1987) at the Division of

Plant industry, Florida Department of Agriculture, Gainesville, FL. For inoculation,

bacterial suspensions were standardized in sterile 10mM CaCO3 (mock) to an optical

density of 0.5 and pressure-infiltrated. For phenotypic observation, inoculations were

repeated at least three times. For protein extraction, a split leaf inoculation scheme was

followed to normalize differences due to physiological state of inoculated tissue. For each

combination of treatments (i.e. mock/B69 and mock/BIM2), one treatment was

inoculated on the right side of the mid-vein and the other strain on the left side of the









mid-vein. For each split-leaf experiment three trees were used, with an average of 10

leaves inoculated per tree (approximately 5 leaves per treatment combination).

Bacterial Strains and Culture Media

Bacterial strains and plasmids used in this study are listed in Table 1 Appendix A.

All Xanthomonas strains were cultured in PYGM medium at 300C (De Feyter et al.

1990). Escherichia coli were grown on Luria-Bertani (LB) medium (Sambrook et al.,

1990). For culture on solid media, agar was added at 15 g/L. Antibiotics were used at the

following concentrations: Spectinomycin (Sp), 35 mg/L; Kanamycin (Kn), 12.5 mg/L;

Chloramphenicol (Cm), 35 mg/L; Gentomycin (Gt), 3 mg/L.

Marker Integration Mutagenesis

hrpG gene knock-out mutation was generated by triparental matings (as described

in Chapter 1). Briefly, a 550 bp internal fragment of hrpG was cloned in the suicide

vector pUFR012 [derivative of pUFR004 carrying kanamycin resistance (Gabriel

laboratory, unpublished)] creating pBY23. Transconjugants resulting from E. coli

DH5 /pBY23, DH5 /pRK2013 (helper plasmid) and B69 matings were selected on

spectinomycin to select against E. coli and chloramphenicol and kanamycin to select for

plasmid insertion events. Putative transconjugants were purified to a single colony, and

Southern hybridization was used to confirm the integration of suicide vector pBY23 in

hrpG.

For complementation purposes, a HindIII to KpnI fragment was cloned out of

plasmid pXG8 (REF) and recloned in pUFR053 (Yuan and Gabriel, unpublished)

creating pBY24. DH5ca/pBY24 was used in triparental matings to create B23.5/pBY24

(B23.5c and B23.5cl). Putative exconjugants were purified to a single colony, and









Southern hybridization was used to confirm the presence of the complementation

plasmid. Total DNA extractions were performed as described in Gabriel and De Feyter

(1992). Southern hybridizations were performed as described by Lazo and Gabriel

(1987).

Bioinformatics

Alignments and box shading were carried out using Clustal W

(http://clustalw.genome.jp).

Protein Extraction and Western Blotting

Citrus leaf tissue was harvested at 0, 2 or 7 days post inoculation (dpi), depending

on the experiment, and ground to a fine powder in liquid nitrogen. Soluble proteins were

extracted in two volumes of extraction buffer (50mM Tris, pH = 8.0, 300mM sucrose,

2mM EDTA, 0.3% DIECA, 10mM N-ethylmaleimide, ltg/tl pepstatin, 1 tgg/tl

leupeptin, and 7.5% w/v PVPP). Extracts were vortexed and briefly sonicated, then

clarified by two rounds of centrifugation at 16,000 x g for 10 min at 40C. Soluble

proteins were quantified by the BCA assay (Pierce Biotechnology, Rockford, IL).

Proteins were separated by polyacryalmide electrophoresis on a 15% Tris-Tricine

gel, and transferred to PVDF membrane (Millipore, Bedford, MA). For immunoblot

analysis, membranes were probed with 1:2,500 immunopurified polyclonal PopSUMOl

(gi:23997054) antiserum (Cocalico, Reamstown, PA) diluted in phosphate buffered saline

(137 mM NaC1, 2.7 mM KC1, 1.4 mM K2HPO4, 10.1 mM Na2HPO4, pH 7.4) containing

0.1% Tween 20 (T-PBS) with 1% v/v goat serum (Sigma, St. Louis, MO). The antibodies

were raised against purified PopSUMO1, which also contained an additional N-terminal

hexahistidine tag generated by PCR (Reed, J., Master's Thesis University of Florida,

2005). For secondary antibody, the membranes were probed with 1:25,000 horseradish









peroxidase conjugated donkey anti-rabbit secondary antibodies (Amersham,

Buckinghamshire, England) diluted in IX T-PBS. Chemilluminescence was carried out

according to the manufacturer's instructions using the ECL plus (+) kit (Amersham).

Following chemilluminescence, each membrane was rinsed in IX T-PBS and stained

with Coomassie R250 as a loading control.

Results

SUMO Conjugation Profiles are Altered in X citri-Infected Leaves

The grapefruit partial cDNA, CCR915 was identified by differential display as

being canker responsive. Following reverse northern blot analysis, CCR915 which shows

homology to SUMO, was found up-regulated in leaves inoculated with BIM2 (lacking

pthB) compared to leaves inoculated with B69 (wt) (Chapter 2). To determine if shifts in

SUMO transcript abundance reflected regulation at the protein levels, a split-leaf

experiment was conducted in which Duncan grapefruit leaves were mock inoculated on

one side of the mid-vein, and Xanthomonas citri pv. aurantifolii strain B69 was

inoculated on the other side. Soluble extracts taken from canker or mock -inoculated

leaves were probed for CitSUMO and CitSUMO-conjugated proteins using PopSUMO1

antibodies. The grapefruit sequence was highly similar to poplar SUMO isoform

PopSUMO1 (gi:23997054) (Figure 3-1) and as expected, the grapefruit SUMO and its

protein conjugates cross-reacted with antibodies raised against PopSUMO 1. Using anti

PopSUMOl, it was found that at two days post inoculation, the profile of SUMO

conjugation is noticeably altered (Figure 3-2). The amounts of free CitSUMO and high

molecular weight CitSUMO conjugated proteins were higher in B69-infiltrated leaves as

compared to mock-infiltrated leaves.









SUMO Conjugation Profiles in Infected Leaves are Partially PthB Dependent

To determine if SUMOylation patterns were associated with disease symptom

development, a split leaf inoculation experiment was conducted and the effects of three

separate treatments examined over time. Split-leaves were mock infiltrated, or inoculated

with wild type strain B69, or the non-pathogenic mutant strain BIM2, which lacks pthB.

At 0, 2, and 7 dpi, half-leaves were harvested and soluble proteins examined by western

blot analysis.

SUMO profiles of leaves inoculated with B69 were compared to those of leaves

inoculated with mutant BIM2 at two dpi. There were no changes in the abundance of free

CitSUMO or SUMOylated proteins in BIM2 inoculated leaves (Figure 3-3, lane 4 and 5).

The expected changes were seen in leaves inoculated with B69, i.e. an increase in the

amount of free SUMO and SUMO-conjugated proteins (Figure 3-3, lane 7 and 8).

SUMO profiles at 7 days post inoculation revealed that the majority of the high

molecular weight conjugates seen at 2 dpi in canker infected leaves were lost (Figure 3-3,

lane 8 and 9). Interestingly, this loss of high molecular weight conjugates was also

observed in leaves inoculated with non-pathogenic mutant strain BIM2. Whether the

identities of SUMOylated proteins in canker infected leaves are similar to the ones in

BIM2 infected leaves is unknown; however, in both cases, SUMO de-conjugation

occurred 7 dpi. These findings suggest that the SUMO de-conjugation observed at 7 dpi,

in both BIM2- and B69-inoculated leaves is PthB-independent and is also independent of

the development of the macromolecular disease symptom of canker (Figure 3-4).

Conversely, the increase in the amounts of free SUMO and SUMO-conjugated proteins

seen at 2 dpi with B69 are PthB-dependent.









SUMO De-Conjugation Observed at 7 days Following Infection with B69 and BIM2
is Dependent on a Functional Type III Secretion System

To determine if the SUMO de-conjugation observed at day 7 post inoculation in

both B69- and BIM2-inoculated leaves is dependent on a functional type III secretion

system, a hrpG integrative mutant, B23.5, was generated. B23.5 was no longer

pathogenic on citrus, and the hrpG- phenotype was complemented after transformation of

B23.5 with pUFR057::XcvhrpG (Figure 3-5).

There was no SUMO de-conjugation at day 7 following inoculation with B23.5

(Figure 3-6), indicating that SUMO de-conjugation relies on a functional TTSS. In

addition, B23.5 inoculation stimulated accumulation of a 45kDa SUMO conjugate. A

SUMOylated product of similar size was observed in leaves inoculated with B69 and

BIM2, but did not accumulate (Figure 3-3).

Discussion

A great deal of effort has been directed towards investigating the mechanisms by

which plants mount defense responses towards pathogenic bacteria. Most studied cases

involve incompatible plant microbe interactions that lead to the classical hypersensitive

response or HR (Malek et al., 2000; Kazan et al, 2001). However, far less effort has been

invested in trying to elucidate the mechanisms by which a specific pathogen, or a group

of pathogens elicit a particular disease with specific sets of morphological and molecular

symptoms. In an effort to understand the processes by which different pathovars of

Xanthomonas citri trigger canker symptoms, a canker responsive gene with sequence

similarity to the SUMO gene family was identified by differential display PCR.

The SUMO conjugation pathway in canker disease was investigated using a split-

leaf inoculation experiment to normalize for leaf-to-leaf variations. It was found that at 2









dpi, X citri pv. aurantifolii infection induces an increase in free CitSUMO and an

increase in the number of high molecular weight SUMOylated proteins. These changes

were not observed in mock-inoculated leaves. SUMO conjugation in plants and other

systems has been shown to be up-regulated by various instances of biotic and abiotic

stresses (Kurepa et. al., 2003, Lois et. al., 2003 and O'Donnell et. al., 2003).

In order to test if changes in SUMO conjugation observed were specific to X citri

pv. aurantifolii infection, two mutant strains unable to cause canker on citrus were used in

this study, BIM2 (interrupted in pathogenicity gene pthB) and B23.5 (interrupted in the

TTSS regulatory gene, hrpG). Disruption of hrpG was previously shown to disable the

type III secretion system in Xanthomonas (Wengelnik et al. 1996). Using split leaf

inoculations, it was shown that in BIM2 inoculated leaves, at 2 dpi, there were no

changes in the amount of free SUMO and SUMOylated high molecular weight proteins.

Thus, the increase in free SUMO and in the number of SUMOylated proteins is likely to

be a PthB-specific plant response rather than a general stress response. A large number of

SUMO targets identified in other organisms are cell-cycle related (Melchior, 2000). It has

been shown in yeast (Saccharomyces cerevisiae) that temperature-sensitive mutants

corresponding to SUMO and the enzymes involved in its conjugation pathway arrest the

cell cycle at the G2/M transition, therefore, showing a critical role for SUMO in cell

cycle progression (Johnson and Gupta, 2001). It is possible that the observed up-

regulation of SUMOylated proteins and free SUMO reflects activation of the plant cell

cycle by X c. pv. aurantifolii in the early stages of infection. Remarkably, this increase in

free SUMO and in the amount of high molecular weight SUMOylated proteins is lost 7

dpi, potentially indicating a transition to a second disease phase. The deconjugation









phenotype observed at 7 dpi with B69 is also observable at 7 dpi with BIM2, and

therefore, the triggering factor of de-conjugation is probably independent of PthB.

The possibility that another effector could be the trigger of the de-conjugation

observed at 7 dpi came from the finding that the TTSS mutant B23.5, did not induce de-

conjugation. Therefore it is possible that another type three effector, beside PthB is

responsible for the de-conjugation observed at day 7. Alternatively, it is possible that the

second PthA homologue, PthB0 (not required for canker, Chapter 2), found in B69 and

BIM2 is also able to trigger the de-conjugation observed 7dpi.

It has been proposed that the abundance of SUMO proteases in X campestris pv.

vesicatoria could reflect an important role of theses effectors in Xcv pathogenesis (Hotson

et al. 2003 and Roden et al 2004). However, none of the identified proteases have been

implicated in disease and are, in fact, dispensable. Given the critical role of SUMO

conjugation in cell cycle processes (Melchior, 2000), and the lack of apparent SUMO

proteases encoded by another canker causing strain X citri pv. citri it is possible that the

late de-conjugation phenotype is not directly triggered by a type III effector of a protease

nature, but rather that a type III effector(s) acts to induce endogenous citrus SUMO

protease(s) leading to the de-conjugation observed in late stages of canker infection.

Both hypotheses are not mutually exclusive and characterization of additional X

citri effectors as well as citrus proteins SUMOylated in response to canker are required to

better characterize the involvement of SUMOylation in the infection process of canker

causing xanthomonads.














popSUMO1
GfSUMO/CCR915
AtSUMOI


popSUMO1
GfSUMO/CCR915
AtSUMO1


MSEATGQPQEEDKKPNDQSAHINLKVKGQDGNEVFFRIKRSTQLKKLMNAYCDRQSVEIN 60
------------------EFHINLKVKGQDGNEVFFRIKRSTQLKKLMNAYCDRQSVEIN 42
MSAN----QEEDKKPGDGGAHINLKVKGQDGNEVFFRIKRSTQLKKLMNAYCDRQSVDMN 56


SIAFLFDGRRLRGEQTPDELDMEDGDEIDAMLHQTGGAVKASDYA 105
SIAFLFDGRRLRGEQTPDELDM----------------------- 64
SIAFLFDGRRLRAEQTPDELDMEDGDEIDAMLHQTGGSGGGATA- 100


Figure 3-1: Alignment of grapefruit SUMO (partial sequence) with (PopSUMO1,
gi:23997054, and AtSUMO1, At4g26840).






78





A

206.7\
115.8
98.0-
Cu
54.6
U)
37.4

29.6-

o 20.4-


7.0

B I m




Figure 3-2: SUMO profiles of B69- and mock-challenged grapefruit leaves. 10pg of
crude protein from day 2 of the split leaf experiment was separated by
electrophoresis, blotted to PVDF and (A) probed with purified PopSUMO
antisera. Lane 1, Mock treated leaf; lane 2, B69 inoculated leaf; lane 3, 2 ng
purified recombinant PopSUMO ([]): high molecular weight SUMOylated
proteins. (->): un-conjugated SUMO. (B) The membrane was stained with
Coomassie R250 as a loading control (Shown is the small subunit of Rubisco).











A Mock
027
0 2 7


BIM2
027


=P WP a M ff"
WI, ..


B69
027


lop "
'U
-m


I m - -
29.6 m
20.4
wo e~memu.I


Treatment
DPI


Figure 3-3: SUMO de-conjugation occurs 7 days after infection. Leaves were inoculated
with Mock, BIM2, and B69 strains. 7.5 [g of crude protein from 0, 2, and 7
dpi from each treatment of the split leaf experiment was separated by
electrophoresis, blotted to PVDF and (Upper panel) probed with purified
PopSUMO1 antisera. ([]): high molecular weight SUMOylated proteins. (-):
un-conjugated SUMO. (Lower panel) The membrane was stained with
Coomassie R250 as a loading control (Shown is the small subunit of Rubisco).


206.7
115.8
98.0-
54.6
37.4














i i






Figure 3-4: Split leaf inoculation ofXanthomonas citri pv. aurantifolii (B69) and
derivative BIM2 mutant. Duncan grapefruit leaf 7 dpi with B69 (shown on the
left side of the mid-vein and BIM2 (shown on the right side of the mid-vein).
(A) adaxial side and (B) abaxial side of the leaf. Note the whitish canker
characteristic of the Xca strain and yellowing associated with the day 7 post
inoculation canker phenotype. (C) Advanced B69 canker phenotype.















X Hin dIII


B69 B23.5 B23.5c




d-om


Figure 3-5: B69 mutant derivative B23.5 lacks a functional Type III secretion system. (A)
Southern blot hybridization profiles contrast B69, B23.5 and B23.5c
(B23.5/hrpG). DNA was digested with HindIII and probed with the same
internal fragment of hrpG used as homology region for marker interruption.
(B) B69 and B23.5c inoculation on Duncan grapefruit. hrpG complemented
the hrp phenotype of B23.5










B23.5
0 2 7


B69
0 2 7


206.7 .
115.8-
98.0 &
54.6-

37.4-

29.6-

20.4-


7.0-


J
*


Figure 3-6: SUMO de-conjugation at 7 dpi requires a functional TTSS. Leaves were
inoculated with B23.5 and B69 strains. 7.5[tg of crude protein from 0, 2, and
7 dpi from each split leaf treatment was separated by electrophoresis, blotted
to PVDF and (A) probed with purified PopSUMO1 antisera. ([]): high
molecular weight SUMOylated proteins. (->): un-conjugated SUMO. (*)
novel 70kDa protein unique to B23.5 7 dpi leaves. Equal amounts of protein
was loaded in each lane.


- m




m mM dm
a*4














APPENDIX A
LIST OF PLASMIDS AND STRAINS


Table A-i: List of strains and plasmids used in this study.
Strains or plasmids Relevant characteristics Reference or source
Escherichia coli
DH5a F-, endA1, hsdR17(rk-mk), Gibco BRL, Gaithesburg,
supE44, thi-1, recA1, gyrA, MD
relAl.,80OdlacZAM15,
A(lacZYA-argF)U169
HB 101 supE44, hsdS20(rk-mk), Boyer and Roulland-
recA 13, ara-14, proA2, Dussoix
lacY1, galK2, rpsL20, xyl-5,
mtl-1, SmR
ED8767 supE44, supF58, hsdS3(rkrkr), Murray et al. 1977
recA56, galK2, galT22,
metB1l
Xanthomonas
3213T' X citri pv. citri A Gabriel et al, 1989
3213Sp X citri pv. citri A, SpR Swamp et al., 1991
derivative of 3213
B21.1 pthA::Tn5-gusA, marker Swamp et al., 1991
exchanged mutant of 3213 Sp,
SpRKnR
B69 X axonopodis pv. aurantifolii
69, ATCC, B form citrus
canker type strain
B69Sp Spntaneous SpR derivative of Unpublished
69, SpR
BIM2 pthB::CmR, marker integration Unpublished
mutant of B69Sp, SpRCmR
BIM6 Marker integration mutant of Unpublished
B69Sp, CmR integrated
upstream ofpthB, SpRCmR
B13.2 VirB4::CmR, marker This study
integration mutant of B69Sp,
SpRCmR









Table A-1. Continued
Strains or plasmids Relevant characteristics Reference or source
B 13.1 VirB4o::CmR, marker This study
integration mutant of B69Sp,
SpRCmR
B69.4 Unpublished
pRK2013 ColE1, KmR,Tra helper Figurski and Helinski, 1979
plasmid
pUFR004 ColE1, Mob+, Cmr, lacZ+ De Feyter et al, 1990
pUFR012 Derivative of pUFR004 with Unpublished
Kn resistance. ColE1, Mob+,
KnRCmR, lacZUa+
pBY13 270 bp fragment of virB4 This study
cloned in pUFR004, CmR
pB13.1 virB4::pBY13 of pXcBO, CmR This study
pB13.2, pB13.4, pB13.5 virB4::pBY13 ofpXcB, CmR This study
PXcB Natural plasmid of B69 Unpublished
carrying pthB
pXcBO Natural plasmid of B69 Unpublished
carrying pthBO
pBIM2 pthB::CmR(pYY40.10) of Unpublished
pXcB, CmR
pBIM6 pXcB:: CmR(pQY92. 1), pthB Unpublished
still functional, CmR
pBY23 550 bp fragment of hrpG This study
cloned in pUFR012, KnR
CmR
pBY23c HrpG from pXG8 (REF) This study
cloned in pUFR53
B23.5 hrpG::pBY23 of B69, KnR This study
CmR
B23.5c B23.5/pBY23c This study

















APPENDIX B
NORTHERN BLOT ANALYSIS OF CCRS


Mock BIM2 B69 Mock BIM2 B69
PR1

PR2 lrr

iss f l 'A P* w ***
2 dpi 7 dpi
Mock BIM2 B69 BIM2 B69


2 dpi B69 BIM2
CHI

rRNA


GST


rRNAMC


RD22


rRNA m

2 dpi B69
PR5 5


2 dpi B69
Frap/tor uiAW


V 1


BIM2

is


BIM2

w


rRNA


2dpi 7dpi
CaCO3 BIM2 B69 BI2
TIP mu
Ti U^^^I


rRNA


2 dpi

P Exp


BIM2


rRNA


2 dpi B69 BIM2

pip3

18 S I


Figure B-1: Northern blot analysis of CCR genes not found differentially regulated by
reverse northern blot


e


-I -.Wdk JM& -
















LIST OF REFERENCES


Alfano, J.R. and Collmer, A. 1996. Bacterial pathogens in plants: life up against the wall.
Plant Cell. 8: 1683-1698.

Alfano, J.R. and Collmer, A. 1997. The type III (Hrp) secretion pathway of plant
pathogenic bacteria: trafficking harpins, Avr proteins and death. J. Bacteriol.
179:5655-5662.

Anderson, D.M., Fouts, D.E., Collmer, A. and Schneewind, O. 1999. Reciprocal
secretion of proteins by the bacterial type III machines of plant and animal
pathogens suggests recognition of mRNA targeting signals. Proc. Nat. Acad. Sci.
USA. 96:12839-12843.

Bergey's Mannual of Determinative Bacteriology, 9th Eddition, JG Holt (ed.), Williams
and Wilkins, Baltimore, MD, USA.

Bonas, U., Stall, R.E. and Staskawicz, B. 1989. Genetic and structural characterization of
the avirulence gene, avrBs3 from Xanthomonas campestris pv. vesicatoria. Mol. &
Gen. Genet. 218:127-136.

Boyer, H. W., and Roulland-Dussoix, D. 1969. A complementation analysis of the
restriction and modification of DNA in Escherichia coli. J. Mol. Biol. 41:459-465.

Brunings A.M. and Gabriel D.W. 2003. Xanthomonas citri: breaking the surface. Molec.
Plant Pathol. 4(3):141-157.

Burns, D.L., 1999. Biochemistry of type IV secretion. Curr. Opin. Microbiol. 1999.
2(1):25-29.

Christie, P.J. 1997. Agrobacterium T-Complex transport apparatus: a paradigm for a new
family of multifunctional transporters in Eubacteria. J. Bacteriol. 179:3085-3094.

Christie, P.J. 2001. Type IV secretion: intracellular transfer of macromolecules by
systems ancestrally related to conjugation machines. Mol. Microbiol. 40:294-305.

Christie, P.J. and Vogel, J.P. 2000. Bacterial type IV secretion: conjugation systems
adapted to deliver effector molecules to host cells. Trends Microbiol. 8:354-360.

Cornelis, G.R. and VanGijsegem, F. 2000. Assembly and function of type III secretary
systems. Annu. Rev. Microbiol. 54:735-774.









Cubero, J. and Graham, J.H. 2002. Genetic relationship among worldwide strains of
Xanthomonas causing canker in citrus species and design of new primers for their
identification by PCR. Appl. Environ. Microbiol. 68:1257-1264.

da Silva, A. C., J. A. Ferro, et al. (2002). "Comparison of the genomes of two
Xanthomonas pathogens with differing host specificities." Nature 417(6887): 459-
63.

De Feyter, R., Kado, C. I., and Gabriel, D. W. 1990. Small stable shuttle vectors for use
in Xanthomonas. Gene 88:65-72.

De Feyter, R., and Gabriel, D. W. 1991. At least six avirulence genes are clustered on a
90-kilobase plasmid in Xanthomonas campestris pv. malvacearum. Mol. Plant-
Microbe Interact. 4:423-432.

De Feyter, R., Yang, Y., and Gabriel, D. W. 1993. Gene-for-genes interactions between
cotton R genes and Xanthomonas campestris pv. malvacearum avr genes. Mol.
Plant-Microbe Interact. 6:225-237.

Duan, Y.P., Castaneda, A.L., Zaho, G., Erdos, G. and Gabriel, D.W. 1999. Expression of
a single, host-specific, bacterial pathogenicity gene in plant cells elicits division,
enlargement and cell death. Mol. Plant-Microbe Interact. 12:556-560.

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

El Yacoubi, B., Brunings,A., Yuan, Q. and Gabriel, D.W. 2001. A self-transmissible
plasmid isolated from Xanthomonas campestris carries a member of the avr/pth
gene family and additional factors) required for pathogenicity. Abstract of the 10th
International Congress of Molecular Plant-Microbe Interactions, Madison, WI, 10-
14 July 2000, #650.

Falcow, S. 1996. The evolution of pathogenicity in Escherichia, ./ngel//l and Salmonela;
in Cellular and Molecular biology (ed.) F.C. Neidhadz (Washington DC: American
Society for Microbiology). 2723-2729.

Figurski, D. H., and Helinski, D. R. 1979. Replication of an origin-containing derivatives
of plasmid RK2 dependent on a plasmid function provided in trans. Proc. Natl.
Acad. Sci. USA 76:1648-1652.

Gabriel, D. W., Kingsley, M., Hunter, J. E., and Gottwald, T. R. 1989. Reinstatement of
Xanthomonas citri (ex Hasse) and X phaseoli (ex Smith) and reclassification of all
X campestris pv. citri strains. Int. J. Syst. Bacteriol. 39:14-22.









Gabriel, D. W., and De Feyter, R. 1992. RFLP analyses and gene tagging for bacterial
identification and taxonomy. Pages 51-66 in: Molecular Plant Pathology: A
Practical Approach. Vol. 1. S. J. Gurr, M. J. McPherson, and D. J. Bowles, eds.
IRL Press, Oxford.

Gabriel, D.W. 1999. Why do plant pathogens carry avirulence genes? Physiol. Mol. Plant
Pathol. 55: 205-214.

Gottwald, T.R., Graham, J.H., Schubert, T.S. 2002. Citris canker: the pathogen and its
inpact. Online. Plant health Progress. doi:10.1094/PHP-2002-0812-01-
RV.http://plant managementnetwork. org/pub/php/review/citruscanker/.

Graham, J.H., Gottwald, T.R., Cubero, J., Achor, D.S. 2004. Xanthomonas axonopodis
pv citri factors affecting successful eradication of citrus canker. Molec. Plant
Pathol. 5:1-15.

He, S.Y. 1998. Type III protein secretion system in plant and animal pathogenic bacteria.
Annu. Rev. Phytopathol. 36: 363-392.

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.

Jin, Q. and S.Y. He (2001). "Role of the Hrp pilus in type III protein secretion in
Pseudomonas syringae." Science 294(5551): 2556-2558.

Jones, J.B., Bouzar, H., Stall, R.E., Almira, E.C., Roberts, P.D., Bowen, B.W., Subderry,
J., Strickler, P.M., and Chun, J. 2000. Systematic analysis of Xanthomonads
(Xanthomonas spp.) associated with pepper and tomato lesions. Int. J. Syst. Evol.
Microbiol. 50:1211-1219.

Keen N.T. 1990. Gene for gene complementrity in plant-pathogen interactions. Annu.
Rev. Genet. 24:447-63.

Kingsley, M.T., Gabriel, D.W., Marlow G.C. and Roberts, P.D. 1993. The opsXlocus of
Xanthomonas campestris affects host range and biosynthesis of lipopolysaccharide
and extracellular polysaccharide. J. Bacteriol. 175:5839-5850.

Kubori, T. Matsushima, Y., Nakamura, D., Uralil, J., Lara-Tajero, M., Sukhan, A., Galan,
J.E., and Aizawa, S. 1998. Supramolecular structure of the Salmonella typhimurium
type III pretein secretion system. Science. 280:602-605.

Lawrance, J.G. and Roth, J.R. 1996. Selfish operons: horizontal transfer may drive the
evolution of gene clusters. Genetics. 143: 1843-9417.

Lazo, G. R., and Gabriel, D. W. 1987. Conservation of plasmid DNA sequences and
pathovar identification of strains ofXanthomonas campestris. Phytopathology 77:
448-453.




Full Text

PAGE 1

BACTERIAL CITRUS CANKER: MOLECUL AR ASPECTS OF A COMPATIBLE PLANT-MICROBE INTERACTION By BASMA EL YACOUBI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Basma El Yacoubi

PAGE 3

To Souad, Kamal, Aziz, Mouma, Ma mi, Nemat, and ma petite Shemsi

PAGE 4

iv ACKNOWLEDGMENTS I would like to express my sincere gratitude to Dr. Dean W. Gabriel, my supervisor and committee chair and for his constant s upport and guidance during my years as a graduate student in his laboratory. I also extend my gratitude to Dr. John M. Davis, (member of my supervisory committee) for hi s valuable advice and for welcoming me in his laboratory each time I needed it. I also thank all other members of my committee, (Dr. Alice Harmon, Dr. Kenneth Cline, Dr. Bill Gurl ey, Dr. Jim Preston) for their valuable advice and guidance. I thank my mother, Souad Benchemsi, w ho always supported me. Without her none of this would have been possible. My husband, Nemat Keyhani, and my little daughter Shemsi Aida Keyhani, gave to my graduate student life a new dimension, and I thank them for that.

PAGE 5

v TABLE OF CONTENTS Page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES...........................................................................................................viii LIST OF FIGURES...........................................................................................................ix ABSTRACT....................................................................................................................... xi CHAPTER 1 A 37 KB PLASMID FROM A SOUTH AMERICAN CITRUS CANKER STRAIN CARRIES A TYPE IV SECRETION SYSTEM ESSENTIAL FOR SELFMOBILIZATION.........................................................................................................1 Introduction................................................................................................................... 1 Materials and Methods.................................................................................................5 Bacterial Strains, Plasmids and Culture Media.....................................................5 Marker Integration Mutagenesis............................................................................5 Plasmid Conjugal Transfer Techniques.................................................................6 Recombinant DNA Techniques.............................................................................7 Plant Inoculations..................................................................................................7 Results........................................................................................................................ ...7 The Type IV Secretion System Fo und on pXcB is Required for SelfMobilization.......................................................................................................7 Involvement of the TFSS of pXcB in Pathogenicity of Xca B69.........................9 Discussion...................................................................................................................10 2 IDENTIFICATION OF CITRUS GENES SPECIFICALLY RESPONSIVE TO PATHOGENICITY GENE pthB OF Xanthomonas citri pv. aurantifolii..................23 Introduction.................................................................................................................23 Material and Methods.................................................................................................26 Plant and Microbial Material...............................................................................26 Bacterial Counts..................................................................................................27 Microscopy..........................................................................................................27 Differential Display-Reve rse Transcriptase PCR................................................28 Suppressive Subtractive Hybridiza tion (SSH) Library Construction..................28 Northern Blots.....................................................................................................29 Reverse Northern Blots.......................................................................................29

PAGE 6

vi Statistical Analysis..............................................................................................30 Results........................................................................................................................ .31 Macroscopic Disease Phenotype of Citrus Leaves Inoculated with X. c. aurantifolii B69 and Its Mutant Deriva tive BIM2 Lacking the Pathogenicity Gene pthB .........................................................................................................31 PthB-Dependent Transcriptional Reprogr amming Induced upon Infection with Xca ...................................................................................................................32 Construction of Two Libraries Enriched in pthB Responsive cDNAs................33 Transcript Analyses of CCRs..............................................................................33 Identity of cDNAs Identified as Up -Regulated by the Presence of pthB in X. citri Genome..................................................................................34 Identity of cDNAs Identified as Up-Regulated by X citri Lacking pthB ............35 Northern Blot Analysis of Representative CCRs................................................36 Microscopic Phenotype of B69 and BIM2 Inoculated Leaves............................36 Discussion...................................................................................................................38 PthB Induces Cell Division and Ce ll Expansion in Citrus Leaves......................39 PthB Induces the Expression of Cell Wall Remodeling Enzymes......................40 Enod8 and SAH7/LAT52 are a Link Between Canker Symptoms Development and Nodule Organogenesis and Pollen Tube Growth Respectively................42 PthB Induces Up-Regulation of a Tonoplast Aquaporin.....................................44 PthB Induces Up-Regulation of Tw o Components Involved in Vesicle Trafficking.......................................................................................................44 Hormone Pathways are Possibly I nvolved in Canker Symptoms Development...................................................................................................45 Conclusions and Future Prospects..............................................................................47 3 CHANGES IN SUMO CONJUGATION ARE ASSOCIATED WITH CITRUS CANKER DISEASE..................................................................................................66 Introduction.................................................................................................................66 Materials and Methods...............................................................................................69 Plant Inoculations................................................................................................69 Bacterial Strains and Culture Media....................................................................70 Marker Integration Mutagenesis..........................................................................70 Bioinformatics.....................................................................................................71 Protein Extraction a nd Western Blotting.............................................................71 Results........................................................................................................................ .72 SUMO Conjugation Profiles are Altered in X. citri -Infected Leaves.................72 SUMO Conjugation Profiles in Infected Leaves are Partially PthB Dependent.73 SUMO De-Conjugation Observed at 7 days Following Infection with B69 and BIM2 is Dependent on a Functi onal Type III SecretionSystem......................74 Discussion...................................................................................................................74

PAGE 7

vii APPENDIX A LIST OF PLASMID AND STRAINS........................................................................83 B NORTHERN BLOT ANALYSIS OF CCRS.............................................................85 LIST OF REFERENCES...................................................................................................86 BIOGRAPHICAL SKETCH.............................................................................................91

PAGE 8

viii LIST OF TABLES Table page 2-1 List of putative CCR identified by DD-PCR...........................................................49 2-2 List of CCRs confirmed by reve rse northern blot analysis......................................50 A-1 List of strains and plasmids used in this study.........................................................83

PAGE 9

ix LIST OF FIGURES Figure page 1-1 Organization of the type four secretion system ( virB operon) found on pXcB compared to other described TFS systems...............................................................13 1-2 Hybridization profiles of DNA from B69 integrativ e mutants interrupted in virB4 of B69 virB clusters. ................................................................................................14 1-3 Eco RI and Bam HI restriction digest profiles of plasmid pB13.1 and plasmid pB13.2, derivatives of pXcB0 and pXcB, re spectively, and integrated in gene virB4 .........................................................................................................................15 1-4 PCR profiles using primers AB65 and AB 66 specific of plasmid pXcB. 16 1-5 Self-mobilization of pXcB derivatives is dependent on a type IV secretion system.17 1-6 Construction of suicide vector pBY17.1..................................................................18 1-7 Scheme of FLP recombinase-mediated marker eviction..........................................19 1-8 PCR confirmation of suicide pl asmid pBY17.1 integration in gene virB4 ..............20 1-9 CR confirmation of Flp-mediated eviction of pBY17.1..........................................21 1-10 Pathogenicity phenotype of primary a nd secondary exconjugants disrupted in virB4 .........................................................................................................................22 2-2 Late B69 and BIM2 phenotypes. (A) BIM2 inoculated leaves 30 dpi and (B) B69 inoculated leaves 30 dpi...........................................................................................53 2-3 Quantification of bacterial population two days post inoculation with B69 and BIM2. (cfu: colony forming unit), Exp1: experiment 1, Exp2: experiment 2)........54 2-4 Diagram of PCR-Sel ect cDNA subtraction..............................................................55 2-5 Distribution of potential ci trus canker responsive genes.........................................56 2-6 Distribution and origin of the clones stamped on the nitrocellulose membranes used in reverse northern blot analysis.....................................................................57 2-7 Cluster analysis of genes di fferentially regulated by PthB. ....................................58

PAGE 10

x 2-8 Northern blot analysis of CCR gene s found differentially regulated by reverse northern blot analysis...............................................................................................59 2-9 Microscopic phenotype of leaves in oculated with B69 (wt) and BIM2 (nonpathogenic mutant lacking PthB). ....................................................................60 2-10 Microscopic phenotype of leaves in oculated with B69 (wt) and BIM2 (nonpathogenic mutant lacking PthB). ....................................................................61 2-11 Microscopic phenotype of leaves in oculated with B69 (wt) and BIM2 (nonpathogenic mutant lacking PthB) at 14 dpi. .....................................................62 2-13 Quantification of leaf thickening and cell division dur ing B69 and BIM2 infection on Duncan grapefruit leaves.....................................................................................63 2-13 Microscopic symptoms of rapidly developing canker. ...........................................64 3-1 Alignment of grapefruit SUMO (par tial sequence) with (PopSUMO1, gi:23997054, and AtSUMO1, At4g26840)....................................................................................77 3-2 SUMO profiles of B69and mo ck-challenged grap efruit leaves. ...........................78 3-3 SUMO de-conjugation occurs 7 days after infection...............................................79 3-4 Split leaf inoculation of Xanthomonas citri pv. aurantifolii (B69) and derivative BIM2 mutant. ..........................................................................................................80 3-5 B69 mutant derivative B23.5 lacks a functional Type III secretion system............81 3-6 SUMO de-conjugation at 7 dpi requires a functional TTSS....................................82 B-1 Northern blot analysis of CCR genes not found differentially regulated by reverse northern blot.............................................................................................................85

PAGE 11

xi Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy BACTERIAL CITRUS CANKER: MOLECUL AR ASPECTS OF A COMPATIBLE PLANT-MICROBE INTERACTION By Basma El Yacoubi May 2005 Chair: Dean W. Gabriel Major Department: Plant Pathology Canker is an important disease affecting citrus worldwide. It is caused by two phylogenetically distinct groups of strains of Xanthomonas citri (Xc), with all citrus cultivars being susceptible to at least one Xc strain. It is known that canker-causing xanthomonads carry at least one pathogenicity gene of the pthA (of Asiatic X. citri pv citri) gene family, which is required for causing canker on citrus. However little is known of the host molecular even ts leading to canker. Our go al was to understand host molecular mechanisms underlying disease development, and identify bacterial components related to phylogeny or pathoge nicity of canker-causing xanthomonads. First we identified on plasmid pXcB of the South American strain X. citri pv aurantifolii B69, a pathogenicit y island composed of previous ly identified pathogenicity gene pthB and a type IV secretion system (TFSS) This TFSS was shown to be required for self-mobilization of pXcB, which led us to propose that natural hor izontal transfer of a pth host-specific pathogenicity gene may acc ount for the two phylogenetically distinct

PAGE 12

xii groups of strains, (the Asiatic and the Sout h American group of strains), causing canker symptoms on citrus. Second, we investigated plant responses to PthB using differen tial display PCR and suppressive subtractive hybridi zation techniques. We identified forty-nine genes that were differentially regulated when RNA expr ession profiles of leaves inoculated with Xca B69 were compared to those of leaves inoculated with a B69 mutant carrying a disrupted pthB Among these were genes predicted to be involved in cell expansion, protein modification, biotic /abiotic stress responses and cell-wall metabolism. Finally, we focused on one canker-responsiv e gene with strong similarity to the small ubiquitin like modifier (SUMO) from Arabidopsis Analysis of B69 mutant strains lacking PthB or the type III secretion system (TTSS) component, HrpG, revealed PthBdependent and TTSS dependent/PthB–inde pendent changes in SUMO conjugation profiles after inf ection with B69. The genes and cellular processes that we identified reflect the molecular events leading to disease development. They contri bute to the general aim of understanding the mechanisms underlying the variety of diseas es caused by compatible interactions between xanthomonads and their host plants.

PAGE 13

1 CHAPTER 1 A 37 KB PLASMID FROM A SOUTH AMERICAN CITRUS CANKER STRAIN CARRIES A TYPE IV SECRETION SYSTEM ESSENTIAL FOR SELFMOBILIZATION Introduction The genus Xanthomonas is comprised of strains that exhibit a high level of hostspecificity; over 125 different path ogenic variants (pathovars) of X. campestris have been described that differ primarily in host range (Bergey, 1994). Host specificity in Xanthomonas can be due to gene-for-gene interac tions involving avirulence genes that act in a negative fashion to limit host ra nge (Keen, 1990; Gabriel, 1999; Leach and White, 1996), but also can be due to positive ac ting factors that condition host range in a host-specific manner. For example, pthN avrb6 of X. campestris pv. malvacearum (Yang and Gabriel, 1996), opsX of X. campestris pv. citrumelo (Kingsley et al, 1993) and pthA of X. citri pv. citri (Swarup et al., 1991 and 1992) act as positive effectors of host range. Interestingly, although a clonal population st ructure is observed among strains within many pathovars (Gabriel et al, 1988), some pa thovars are comprised of phylogenetically distinct groups that have an identical host range and caus e identical disease symptoms. Examples include 1) common bean bli ght, caused by two groups of strains ( X. phaseoli and X. campestris pv. phaseoli var. fuscans) that are only 20% related by DNA-DNA hybridization (Hildebrand et al ., 1990); 2) bacterial spot of tomato and pepper, caused by two major groups of strains within X. campestris pv. vesicatoria (Jone s et al., 2000) that are less than 50% related by DNA-DNA hybridiza tion (Stall et al., 1994 ), and 3) citrus canker disease, caused by two groups of st rains that are only 62 -63% related by DNA-

PAGE 14

2 DNA hybridization (Egel et al., 1991). Strains with 70% or greater DNA-DNA relatedness are usually defined as single spec ies (Wayne et al., 1987) The question arises however, as to how phylogenetical ly diverse strains can cause identical diseases on an identical range of hosts. To date, all pathogenic xanthomonads examined require hrp genes (reviewed by Alfano & Collmer, 1996, 1997; He, 1998; Cornelis & VanGijsegem, 2000) to cause disease. These genes encode a type III secret ion machine that is close contact-dependent (Marenda et al., 1998) and used to inject high ly adapted effector proteins into both host and nonhost cells (Silhavy, 1997; Kubori et al., 1998; and Jin and He, 2001). These effector proteins elicit the diverse programmed phenotypes of the plant hypersensitive response (HR) and various pa thogenicity responses. The hrp (h ypersensitive r esponse and p athogenicity) injection system is thus a ppropriately named, and it is also highly indiscriminate, injecting whatever effector proteins are available, even some from animal pathogens (Anderson et al., 1999 and Ro ssier et al., 1999). If identical hrp effectors are available within two phylogenetically distin ct xanthomonads, they can cause the same disease symptoms, provided both strains are comp atible (able to multiply in the host) and both carry functional hrp systems. For example, pthA was transferred from X. citri to X. campestris pv. citromelo and converted the latter strain from a leaf-s potting strain to a strain with ability to cause citrus canker di sease (Swarup et al., 1991). PthA appears to be an effector protein that is cr itical for citrus disease sympto ms and is likely injected by X. citri into citrus cells, cau sing hyperplastic cankers (Duan et al, 1999). Citrus canker disease is caused by two phylogenetically distinct and clonal groups of Xanthomonas strains; each group contains subgr oups that are distinguished on the

PAGE 15

3 basis of host range (Brunings and Gabriel, 2003). The first phylogenetically distinct group is the Asiatic group, named Xanthomonas citri pv citri ex Hasse (syn = X. campestris pv. citri Dye pathotype A and X. axonopodis pv. citri Vauterin, Xca -A). The second phylogenetically distinct gro up is the S. American group, named X. citri pv. aurantifolii Gabriel (syn = X. campestris pv. citri Dye and X. axonopodis pv. aurantifolii Vauterin, Xca -B). Both groups cause identical citrus canker disease symptoms circular, water soaked raised lesions, that become da rk and thick as canker progresses (Graham et al., 2004; Stall and Civerolo, 1991; Gottwald et al, 2002; Brunings and Gabriel, 2003). Significantly, pthA or homologues are present in every Xanthomonas strain tested that causes citrus canker disease, and have not been found present in xanthomonads isolated from citrus that do not cause canker (Gabri el, 1999; Cubero and Graham, 2002). Prior to this work, two pthA homologues, named pthB and pthB0 were found on two separate plasmids (pXcB and pXcB0, respectively) of a S. American canker strain (B69). Plasmid pXcB carrying the functional homologue pthB was then found to be readily cured from B69 (Yuan and Gabriel unpublished, and Bruning s, A.M., 2004 M.S. thesis University of Florida). Readily cured plasmids are ofte n mobilizable by conjugation. Since Asia is considered to be the center of origin of c itrus canker disease, and since Asiatic canker strains are more widespread in S. America than S. American canker strains, it was of interest to determine if pXcB could transfer horizontally. pX cB was found to horizontally transfer in-vitro and in planta (Yuan and Gabriel unpublished) from the S. American strain B69 to the Asiatic st rain B21.1 lacking a functional pthA restoring its capacity to cause canker (Yuan and Gabriel unpublished) Presence of the type III effector pthB on a

PAGE 16

4 self-mobilizing plasmid might explain the creat ion of the entire S. American group of canker strains, and why they are phylogene tically distinct from the Asiatic group. pXcB was fully sequenced (NC_005240, gi32347275), and besides gene pthB a complete Type IV secretion system (TFSS) was also found on the plasmid (Brunings, A.M., 2004 M.S. Thesis, University of Fl orida). TFSS are defined on the basis of homologies between the A. tumefaciens T-DNA transfer system, the conjugal transfer system Tra, and the Bordetella pertussis toxin exporter, Ptl (Winams et al., 1996 and Christie, 1997). Most members of the TFSS fa mily function primarily to mobilize DNA, either from bacteria to bacteria (bacterial conjugation system) or from bacteria to eukaryotic cells ( Agrobacterium oncogenic T-DNA transfer system) (Burns, 1999). In addition, several bacterial pat hogens utilize conjugation mach ines to export effector molecules during infection. Such systems are said to be Type IV “adapted” conjugation or secretion systems, for their involvement in pathogenicity. Many non-plant pathogens such as Bordetella pertussis Legionella pneumophila, Brucella spp. and Helicobacter pylori use a type IV “adapted” conjugation syst em to secrete effector proteins to the extracellular milieu or the cel l cytosol (Burns, 1999; Christie, 1996; Christie and Vogel, 2000). Type IV systems are composed of products with homology to the Agrobacterium virB operon (Vogel, 2000). Sequence similarity analysis revealed that the Type IV secretion system of pXcB encodes twelve open reading frames, te n of which contained high sequence similarities to ge nes of previously described virB operons as well as similar relative positions within the cluster (Bruni ngs, A.M., 2004 M.S. Thesis, University of Florida).

PAGE 17

5 In order to investigate whether the T FSS found on pXcB is involved in selfmobilization of pXcB, a plas mid derivative lacking a functional TFSS was generated in this study and tested for its ability to self-mobilize in vitro In addition, a B69 derivative lacking the TFSS was generated in a non-polar fashion to address whether this system was required for pathogenicity of B69. It wa s found that the TFSS of pXcB was required for self-mobilization of the plasmid. Howe ver pathogenicity tests involving TFSS insertional mutants were inconclusive, a nd it remains unknown whet her this secretion system is involved in pathogenicity of X. citri pv. aurantifolii. Materials and Methods Bacterial Strains, Plasmids and Culture Media Bacterial strains and plasmids used in this study are listed in Table 1. Xanthomonas spp. were cultured in PYGM medium at 30oC (De Feyter et al., 1990). Escherichia coli were grown in Luria-Bertani (L B) medium (Sambrook et al., 1989). Antibiotics were used at the following concentrations (in g/mL): Chloramphenicol (Cm), 35; Kanamycin (Kn) 12.5 or 25 (when used to grow Xanthomonas or E. coli respectively); Spectinomycin (Sp) 35 and Streptomycin (St) 35. Marker Integration Mutagenesis Gene-specific knockout mutations of Xanthomonas were created by triparental matings. An E. coli DH5 strain carrying an internal frag ment of the target open reading frame (ORF) cloned in suicide vect or pUFR004 was used as donor. A DH5 strain carrying pRK2013 was used as the helper. A si ngle crossover in the exconjugates results in duplication of the internal fragment at the integration site, and also results in interrupting the target gene with the vector. To disrupt virB4 a PCR-generated, 270 bp

PAGE 18

6 internal fragment of virB4 (virB4 270 ) was cloned in pGEM–T Easy and recloned in pUFR004 creating pBY13. Plasmid Conjugal Transfer Techniques Plasmid transfer by triparental mating from E. coli strains HB101 or DH5 to various Xanthomonas strains, using helper strain pR K2013 were performed essentially as described in De Feyter a nd Gabriel (1991). For plasmid transmission experiments on artificial media, ove rnight cultures of E. coli strains grown without antibiotics were mixed with 50X concentrated overn ight, mid-log phase cultures of Xanthomonas strains, grown without antibiotics. Drops (10 l each) of recipient donor and helper cells were placed on PYGM agar medium one after the ot her and without antibiotics. In each case excess liquid was allowed to absorb into the plate before addition of the next cell type. The mating plates were incubated at 30oC overnight, and the spots were then streaked on PYGM selection medium supplemented with the appropriate antibiotics. In Xanthomonas to E. coli matings, B69 carrying pB13.2 (pBY13 integrated in virB4 of pXcB) or B69 carrying pB13.1 (pBY13 integrated in virB4 of pXcB0) were used as donor strains (in inde pendent matings) with DH5 as the recipient strain. After selection against Xanthomonas on MacConkey agar (DIFCO laboratories, Detroit MI, USA) with 35 g/mL chloramphenicol, DH5 exconjugants were screened for the presence of pBY13.2 or pBY13.1 by DNA mini-prep analysis. In E. coli to E. coli matings, DH5 /pBY13.2, DH5 /pBY13.1 and DH5 /pBIM2 (pYY40.10 integrated in pthB ) were used as donor stains in indepe ndent matings with HB101 as recipient. For frequency of transfer assays from one E coli strain to another, donor and recipient strains were grown overnight at 37 oC to an O.D. 600nm of 0.5. Twenty

PAGE 19

7 microliters of each culture were combined in a 1.5 ml Eppendorf tube containing 160 l of LB and grown overnight at 37 oC. Cells were then resuspended in 1 ml of LB, pelleted and then serially diluted on medium containing chloramphe nicol and streptomycin to select for HB101 transconjugants. All conjugati on experiments were performed at least twice with duplicate samples in each expe riment, and the numbers were averaged. Recombinant DNA Techniques Plasmid and total DNA were prepared from Xanthomonas as described by Gabriel and De Feyter (1992). E. coli plasmid preparation, restrictio n enzyme digestion, alkaline phosphatase treatment, DNA ligation, and ra ndom priming reactions were performed using standard techniques (Sambrook et al., 1989). Southern hybrid ization was performed using nylon membranes as descri bed by Lazo and Gabriel (1987). Plant Inoculations All citrus plants ( Citrus paradisi ‘Duncan’, grapefru it) were grown under greenhouse conditions. Plant inoculations involvi ng all citrus canker strains were carried out under quarantine at the Divi sion of Plant Industry, Florid a Department of Agriculture, Gainesville. Bacterial cells were harvested from log phase cultures by centrifugation (5,000 x g, 10 min.), washed once and resuspended in sterile tap water or distilled water saturated with calcium carbonate to 108cfu/mL. Inoculations were performed by pressureinfiltration into the abaxial leaf surface of the plants. Experimental inoculations were repeated at least three times. Results The Type IV Secretion System Found on pXcB is Required for Self-Mobilization Gene virB4 of the TFSS cluster of pXcB wa s chosen as target for marker insertional mutagenesis (Figure 1-1). Fo r that, a 270 bp integral fragment of virB4

PAGE 20

8 (virB4270) was cloned in pUFR004 (pBY13) and used in triparental matings to generate virB4 insertion mutants. Southern blots were us ed to verify integration events in the resulting transconjugants. These results dem onstrate the existence of two copies of virB4 in the B69 strain (using virB4270 as probe Figure 1-2). One copy was carried by pXcB (as determined by sequencing) and was absent in the cured strain B69.4 [Rifamycin resistant strain cured of plasmid pXcB but carrying plasmid pXcB0, Yuan and Gabriel, unpublished (Lane 3)]. A second putative copy carried by pXcB0, was maintained in B69.4 (Lane 3). Marker insertion resulted in two categories of exconjugants. Exconjugant strain B13.1 appeared to carry an interruption of the putative virB4 of pXcB0 ( virB40) (Lane 7), while exconjugants B13.2, B13.4 and B13.5 appeared to carry interruptions of the virB4 gene of pXcB (Lanes 4, 5 and 6). Plasmids pB13.1 and pB13.2 of strains B 13.1 and B13.2 (marker interruptions in the virB4 homologues found on pXcB0 and pXcB, re spectively) were further analyzed for their ability to transfer to E. coli Matings with and without the helper strain resulted in DH5 exconjugants carrying plasmids that were chloramphenicol resistant, indicating that both plasmids were still mobilizing. Re striction enzyme digests of plasmid DNA extracted from the Xanthomonas (B13.2) to the E. coli exconjugant (DH5 /pB13.2), corresponded to the expected profile of pXcB integrated with pBY13 (Figure 1-3). Restriction enzyme digests of plasmid DNA extracted from DH5 /pB13.1 did not corresponded to the profile expected for a pXcB insertional deri vative. Therefore, p13.1 is a derivative of a second native plasmid of B6 9, smaller in size than pXcB and inserted in a putative virB4 copy. Th ese results were confirmed by PCR using primers specific to pXcB. As shown in Figure 1-4, when pB13.2 wa s used as template with pXcB specific

PAGE 21

9 primers AB65/AB66 a 2014 bp band was obtained, while non specific bands were obtained when pB13.1 was used as template The ability of pB13.1 and pB13.2 to self-mobilize was then analyzed by performing matings from DH5 to E. coli HB101. Using DH5 /pBIM2, and DH5 /pBIM6 [pBIM6 is a derivative of pX cB where pUFR004 was inserted in a nonORF region, (Yuan and Gabriel, unpublished)] as a control, transfer of pBIM2 and pBIM6 from DH5 to HB101 was found not to require the presence of a helper strain and the transfer frequency was 7x10-03 and 6.6x10-05 per donor, respectively. By contrast, E. coli to E. coli transconjugants harboring pB13.1 or pB13.2 were only recovered when matings were performed in the presence of a helper strain (Figur e 1-5). These results indicated that the self-mobilization capacity of pXcB depended on the presence of an intact virB cluster. Involvement of the TFSS of pXcB in Pathogenicity of Xca B69 Non-polar knock out mutants of virB4 were generated using marker insertion followed by FLP recombinase mediated marker eviction. Plasmid pBY17.1 was generated so that a virB4 homology region was flanked by two FRT recognition sites (See Figure 1-6 for illustration) After marker integration of suicide vector pBY17.2 into primary transconjugants, the FLP recombinas e plasmid pJR4, was used to evict the marker, and generated non-polar seconda ry transconjugants (See Figure 1-7 for illustration). Several primary transconjugants (before FL P-mediated eviction of marker) (Figure 1-8) as well as secondary tr ansconjugants (after FLP mediat ed eviction of marker) were tested for integration events in a virB4 homologue using PCR. Bacterial cells directly

PAGE 22

10 from the selection plates were used as template for PCR (Figure 1-9). PCR positive colonies were then grown in liquid culture a nd tested for pathogenicity on citrus. In all cases, primary exconjugants showed a decrea se in pathogenicity while, unexpectedly, secondary exconjugants lost th eir potential to trig ger canker disease on citrus (Figure 110). When the secondary exconjugants used in pathogenicity assays were tested by PCR for presence of pXcB it was found th at the plasmid and therefore gene pthB were lost upon culturing. Discussion The putative TFSS of pXcB (Brunings A.M ., M.S. Thesis, University of Florida and Brunings and Gabriel, 2003) was functi onally investigated to determine its involvement in plasmid transfer as well as in pathogenicity of B 69. To investigate the role of this TFSS in plasmid transfer, gene virB4 was marker-interrupted and by consequence the whole system rendered dysfunctional. Self-mobilization experiments revealed that pXcB relied on a functional T FSS to self–mobilize. In the process a second putative virB4 homologue was identified on a se cond plasmid of B69, pXcB0. pB13.1, carrying a single insertion in virB40 of pXcB0 and pB13.2, carrying a single insertion in virB4 of pXcB were each able to mobilize from B13.1 and B13.2, respectively, to DH5 in biparental matings (without helper st rain), indicating that the two putative virB systems co-existing in B69 might be compensatory. The characterization of pXcB as a se lf-mobilizing plasmid carrying a TFSS and gene pthB suggests that the canker causing and p hylogenetically distinct South American strains may have arisen from horizontal gene transfer of an “ancestral” pthA member. This horizontal transfer likely w ould have occurred from an Asiatic Xanthomonas citri

PAGE 23

11 strain to a compatible TTS system-carryi ng xanthomonad residing on the same host. B69 was indeed shown to carry a functional TTS system required for pathogenicity (see Chapter 3). The type IV secr etion system together with pthB on pXcB of S. American Xanthomonas citri strains can therefore be considered an “aut o-mobile” pathogenicity island (Hacker et al., 1997) capable of spreading am ong compatible bacteria by horizontal gene transfer. Since pXcB from the South American strain is smaller, yet very similar to pXAC64 from the Asiatic strain, pXcB could be a deletion derivative of pXAC64 (Brunings and Gabriel 2003). However, while many genes on pXcB were found to be similar to genes on pXAC64, there were differen ces significant enough to concl ude that a simple deletion cannot account for pXcB. More likely, seve ral independent events were probably responsible for its divergence away from pXAC64. Horizontal gene transfer is proposed to be a major mechanism explaining rapid genetic diversificati on in bacteria (Falcow, 1996; Syvanen and Kado, 1998; Lawrence and Roth, 1999). It has been proposed to expl ain the apparent enigma of why pathogens carry dispensable avirulence genes (Yang and Gabriel, 1996 and Gabriel, 1999). For example, avrBs3 of Xanthomonas campestris pv. vesicatoria was found on a mobilizing plasmid carrying copper resistance, and th erefore wide horizontal transfer of avrBs3 to X. campestris pathovars may be due to coincidental linkage with copper resistance (Stall et al, 1986, Yang and Gabriel, 1996). The TFSS of pXcB was also analyzed for its involvement in pathogenicity. Primary exconjugants carrying a marker integration in virB4 showed a decrease in pathogenicity while non-polar secondary exconjugants, resul ting from marker eviction of the suicide

PAGE 24

12 plasmid, lost all pathogenicity. This was then found to be possibly due to a loss of pXcB upon curing of secondary transconju gant strains. Another explan ation is the presence of a large insertion vector in the native plasmi d decreasing the copy number in the population. Further examination of the TFSS is necessary to access its role in pathogenicity if any.

PAGE 25

13 Figure 1-1: Organization of the type four secretion system ( virB operon) found on pXcB compared to other described TFS system s. ORF106 shows no similarity to any virB cluster gene of Agrobacterium tumefaciens and is shown as an insertion. Orf106 B2B3B4B5 B7B6B8B9 B10 B11 B1 X.a.a ( pXcB) virB virB A tumefaciens (pAtC58) avhB Orf106 B2B3B4B5 B7B6B8B9 B10 B11 B1 Self-mobilization Transfer of T-DNA Ti-plasmid Conjugal transferOrf106 B2B3B4B5 B7B6B8B9 B10 B11 B1 X.a.a ( pXcB) virB virB A tumefaciens (pAtC58) avhB Orf106 B2B3B4B5 B7B6B8B9 B10 B11 B1 Self-mobilization Transfer of T-DNA Ti-plasmid Conjugal transfer

PAGE 26

14 6969.413.213.413.513.1 9.4 6.6 4.4 virB40 ( pXcB0) virB4( pXcB) 6969.413.213.413.513.1 9.4 6.6 4.4 virB40 ( pXcB0) virB4( pXcB) Figure 1-2: Hybridization prof iles of DNA from B69 integrat ive mutants interrupted in virB4 of B69 virB clusters. Total DNA was digested with Hin dIII and probed with a 32P-labelled 270 bp internal fragment of virB4 The same fragment was used as a homology region for integra tion of suicide vector pBY13. B13.2, B13.4 and B13.5 were marker integrated in virB4 of pXcB, and B13.1 was marker integrated in virB4 of pXcB0. Hin d III digestion results in splitting one restriction fragment harboring th e targeted region in to two hybridizing fragments. Therefore there ar e two bands hybridizing to the virB4270 probe in the wild type strains, while there are th ree bands in the insertional strains. The only band hybridizing to the virB4270 in the B69.4 lane corresponds to a putative virB4 copy present on a second native plasmid of B69. Indeed pXcB was lost upon curing in B69.4 and ther efore one hybridizing band is lost.

PAGE 27

15 pBIM2 p13.2p13.1Eco RI Bam HI pBIM2p13.2p13.1 23 9.4 6.6 4.4 2.1 2.3 pBIM2 p13.2p13.1Eco RI Bam HI pBIM2p13.2p13.1 23 9.4 6.6 4.4 2.1 2.3 Figure 1-3: Eco RI and Bam HI restriction digest profiles of plasmid pB13.1 and plasmid pB13.2, derivatives of pXcB0 and pXcB, re spectively, and integrated in gene virB4

PAGE 28

16 B69 B69.4 pB13.1 pB13.2 B69 B69.4 pB13.1 pB13.2 Figure 1-4: PCR profiles using primers AB65 and AB66 specific of plasmid pXcB. Plasmid DNA isolated from DH5 /pB13.1, DH5 /pB13.2, and total DNA isolated from B69 and B69.4 were us ed as templates. [AB65: CAG CCG CAA GTG TCT CAG GTC; AB66: GGC AAG AAA CCG TCC GAG TA (Tm 56 C)]. When B69.4 and pB13.1 DNA are used as template in the PCR reaction non-specific bands of low inte nsity are the resulting products. When B69 and pB13.2 (both derivatives of plas mid pXcB) are used as template, a specific band of 2014 bp is the resu lting product of the PCR reaction. ( ): Amplification fragment specific to pX cB when AB65/AB66 primers are used.

PAGE 29

17 0% 0.5% 1% pB13.1pB13.2pBIM2 Plasmid ( in donor strain DH5 ) Frequency of transfer to HB101 (per input donor) p13.1 0 p13.2 0 pBIM2 7.08E-03 pBIM6 6.57E-05 a b cA BFrequency of transfer (per input donor) 0% 0.5% 1% pB13.1pB13.2pBIM2 Plasmid ( in donor strain DH5 ) Frequency of transfer to HB101 (per input donor) p13.1 0 p13.2 0 pBIM2 7.08E-03 pBIM6 6.57E-05 a b c a a b cA BFrequency of transfer (per input donor) Figure 1-5: Self-mobilization of pXcB deriva tives is dependent on a type IV secretion system. ( A ) Mobilization of pXcB deriva tive, pBIM2 and pB13.2 and pXcB0 derivative pB13.1 from E. coli DH5 to E. coli HB101. Matings were carried with and without helper strain carrying plasmid pRK2013, and HB101 transconjugants were selected on LB supplemented with streptomycin and chloramphenicol. Each selection plate was separated in two sections. Results of matings with helper strain are sh own on the left section, and results of matings without helper are show n on the right. Matings: (a) DH5 /pB13.1 with HB101; (b) DH5 /pB13.2 with HB101; (c) DH5 /pBIM2 with HB101. ( B) Frequency of transfer of pXcB derivatives, pB13.2, pBIM2, pBIM6 and pXcB0 derivative, pB13.1 from E. coil DH5 into E. coli HB101.

PAGE 30

18 Figure 1-6: Construction of su icide vector pBY17.1. PCR was used to amplify an internal virB4 fragment using primers BY13 (g atcaggatcctatgcgcctcgttgaggt) and BY14 (cggtccgtcagtcagtcagagctctgaccaggtagtgcagga). Rsr III and Bcl I restriction sites were incorporated in the primer sequences respectively in the forward and reverse primer. The Rsr IIIBcl I fragment was used as the driver for homologous recombination and was cloned between FLP sites in pUFR012 (pUFR004 derivative carrying kanamycin resistance). FLP sites were obtain by PCR using plasmid p KD4 (gi:15554332) (1.5 Kb fragment). Primers FRTKn F (gaattcgctgct tcgaagttcctatac) and FRTKn R (aagcttatcctccttagttccaattcc) carried an Eco RI and a Hin dIII site for subcloning from pGEMT-ez (Promega) into pUFR012. FRTKn PCR fragment FRTKn F FRTKn R FRTKn PCR fragment FRTKn F FRTKn R PCR fragment from template pKD4 was cloned in pGEMT-ez BY13BY14PCR fragment from template pXcB was cloned in pGEMT-ez RsrIIIBclI pBY3.1 (pUFR012::FRTKn) Eco RI Hin dIIIRsrIII BclI Eco RI pBY1.1 Hin dIII pBY2.1RsrIIIBclI pBY17.1 (pBY3.1:: virB4 ) RsrIIIBclIPCR fragment internal to virB4 FRTKn PCR fragment FRTKn F FRTKn R FRTKn PCR fragment FRTKn F FRTKn R PCR fragment from template pKD4 was cloned in pGEMT-ez BY13BY14PCR fragment from template pXcB was cloned in pGEMT-ez RsrIIIBclI pBY3.1 (pUFR012::FRTKn) Eco RI Hin dIIIRsrIII BclI pBY3.1 (pUFR012::FRTKn) Eco RI Hin dIIIRsrIII BclI Eco RI pBY1.1 Hin dIII Eco RI pBY1.1 Hin dIII pBY2.1RsrIIIBclI RsrIIIBclI pBY17.1 (pBY3.1:: virB4 ) RsrIIIBclI RsrIIIBclIPCR fragment internal to virB4 FRTKn PCR fragment FRTKn F FRTKn R FRTKn PCR fragment FRTKn F FRTKn R PCR fragment from template pKD4 was cloned in pGEMT-ez BY13BY14PCR fragment from template pXcB was cloned in pGEMT-ez RsrIIIBclI pBY3.1 (pUFR012::FRTKn) Eco RI Hin dIIIRsrIII BclI Eco RI pBY1.1 Hin dIII pBY2.1RsrIIIBclI pBY17.1 (pBY3.1:: virB4 ) RsrIIIBclIPCR fragment internal to virB4 FRTKn PCR fragment FRTKn F FRTKn R FRTKn PCR fragment FRTKn F FRTKn R PCR fragment from template pKD4 was cloned in pGEMT-ez BY13BY14PCR fragment from template pXcB was cloned in pGEMT-ez RsrIIIBclI pBY3.1 (pUFR012::FRTKn) Eco RI Hin dIIIRsrIII BclI pBY3.1 (pUFR012::FRTKn) Eco RI Hin dIIIRsrIII BclI Eco RI pBY1.1 Hin dIII Eco RI pBY1.1 Hin dIII pBY2.1RsrIIIBclI RsrIIIBclI pBY17.1 (pBY3.1:: virB4 ) RsrIIIBclI RsrIIIBclIPCR fragment internal to virB4

PAGE 31

19 FRT FRThomology to virB4 pBY17.1 (Kn, Cm) 800bp800bp 700bp bes03 virB4 F bes04 inserted vector virB4 M13R bes03 bes04 virB4 F 22mer FLP recombinase plasmid Figure 1-7: Scheme of FLP recombinase-me diated marker eviction. Suicide plasmid pBY17.1 is marker integrated in gene virB4 via homologous recombination, generating a virB4 disruption. The light blue box represents the homology region targeted for recombination, and is found duplicated af ter insertion of the suicide plasmid pBY17.1 in pXcB After transformation of the B69 derivative carrying pXcB::pBY17.1 (prima ry transconjugant) with plasmid pJR4 carrying a FLP recombinase gene, pBY17.1 is evicted (secondary transconjugants). pJR4 [derived from pFLP (gi:1245114) (Ready and Gabriel, unpublished)] is cured by culturing secondary transconjugants on PYGM supplemented with 5% sucrose. VirB4F and bes03 and bes04 are virB4 specific primers and their locations ar e shown by arrows. 22mer and M13R are primers specific to the polylinker region of suicide vector pBY17.1 and are used to verify integration and ev iction events. FRT sites recognized the FLP recombinase are symbolized by yellow circles and flank the internal fragment with homology to virB4 cloned in pBY17.1. The green boxes symbolize DNA stretches carried over during sub-cloning steps.

PAGE 32

20 123456789 22mer/bes03M13R/BES04M13R/bes03DS Lane 1,4,7: B69, Lane 2,5,8: B18.12, Lane 3,6,9: B18.15 123456789 22mer/bes03M13R/BES04M13R/bes03DS Lane 1,4,7: B69, Lane 2,5,8: B18.12, Lane 3,6,9: B18.15 vector pBY17.1 virB4 M13R bes03 bes04 virB4 F 22mer vector pBY17.1 virB4 M13R bes03 bes04 virB4 F 22mer bes03DS Figure 1-8: PCR confirmation of suicid e plasmid pBY17.1 integration in gene virB4 22 mer (gttttcccagtgacgacg) and M13R ( agcggataacaatttcacac) are primer specific to the polylinker of pBY17.1. bes03(catcttggatcgtgcgtt) bes03DS, bes04 (catgttgctgagcatctt) and virB4 F(ggtaccacccatttgaaaacgtgtcc) are gene specific primers. Lanes 1,4,and 7, B69 bact erial cells were used as template source for the PCR. Lanes 2, 5 and 8, B18.12 (primary transformant with pBY17.1 inserted in virB4 ), bacterial cells were used as template source for the PCR. Lanes 3, 6 and 9, B18.5 (pri mary transformant with pBY17.1 inserted in virB4 ) bacterial cells were used as template source for the PCR. Primer combinations used in each lane are indicated in the figure. PCR bands resulting from using v irB4based primers in combination with suicide vector based primers are specific to an insertion in the targ eted region and should not appear when the wild type strain is used. The light blue box represents the homology region targeted for recombin ation, and is found duplicated after insertion of the suicide plasmid pBY1 7.1 in pXcB. FRT sites recognized the FLP recombinase are symbolized by yellow circles and flank the internal fragment with homology to virB4 cloned in pBY17.1. The green boxes symbolize DNA stretches carried over during sub-cloning steps.

PAGE 33

21 Lane 1: B69 Lane 2:B18.12 Lane 3:B18.12-1 123123123 BES03/BES04M13R/bes04virB4F/bes04 123 123123 123123 123 BES03/BES04M13R/bes04virB4F/bes04bes03/bes04: 470bp M13R/bes04: 1300 bp virB4 F/bes04: 950bp (wt), or 2500bp after FLP 800bp800bp 700bp bes03 virB4 F bes04 800bp800bp 700bp bes03 virB4 F bes04suicide vector inserted virB4 M13R bes03 bes04 virB4 F 22mer M13R bes03 bes04 virB4 F 22merB18.12-1 B18.12 Figure 1-9: PCR confirmation of Flp-medi ated eviction of pBY17.1. Genes specific primers (bes03, bes04, virB4F and vect or based primers were used in appropriate combinations. B69 and B18.12 (primary transconjugant with pBY17.1 inserted in virB4) were used as negative and positive control for the suicide vector integration respec tively. B18.1 2-1 is the secondary transformant resulting from suicid e vector eviction from primary transconjugant B18.12. Bacter ial cells from selection plates were used as template source for PCR. The expected size of each PCR band is indicated in the figure. The light blue box represen ts the homology region targeted for recombination, and is found duplicated af ter insertion of the suicide plasmid pBY17.1 in pXcB. FRT sites recognized the FLP recombinase are symbolized by yellow circles and flank the internal fragment with homology to virB4 cloned in pBY17.1. The green boxes symbolize DNA stretches carried over during sub-cloning steps.

PAGE 34

22 Figure 1-10: Pathogenicity phenotype of prim ary and secondary exconjugants disrupted in virB4 B18.12; primary exconjugant ( virB4 ::pBY17.1). B18.12-1 (secondary exconjugants, after eviction of pBY17.1) Picture on the left was taken 7 days post inoculation. The two pi ctures on the right were taken 15 days after inoculation. Note the dela y phenotype of primary transconjugant B18.12, and the total loss of pathoge nicity of transconjugant B18.12-1. B69 B18.12 B18.12-1 B69 B69 B18.12 B18.12-1 B18.12 B69 B18.12 B18.12-1 B69 B69 B18.12 B18.12-1 B18.12

PAGE 35

23 CHAPTER 2 IDENTIFICATION OF CITRUS GENES SPECIFICALLY RESPONSIVE TO PATHOGENICITY GENE pthB OF Xanthomonas citri pv. aurantifolii Introduction Many studies on plant-pathogen interacti ons have dealt with incompatible interactions using model plant systems (for example see Malek et al., 2000). Emphasis has been on dissecting signali ng pathways of resistance mechanisms, with few studies considering signaling pathways resulting in di seases of crop plants (Kazan et al., 2001). Therefore, the molecular events at the origin of disease induction by microbial effectors of pathogens remain obscure. Many Gram-negative, phytopathogenic bacter ia rely on a Type III secretion system (TTSS) to deliver effector pr oteins into the plant cells (H e et al., 2004). Inactivation of the TTSS of bacterial species th at utilize such a system results in loss of pathogenesis indicating that the proteins (named type III effectors) delivered by the TTSS are required for bacterial virulence (Rohmer et al., 2004). Most type III effectors identified to date were originally discovered and characterized by their avirulence f unction (Avr), while only few are recognized pathoge nicity factors [PthA from X. citri (Swarup at al., 1991), AvrB6 from X. campestris pv. malvacearum (Yang et al,. 1996), AvrXa7 from X. oryzae (Bai et al., 2000) and DspA from Erwinia amylovora (Gaudriault et al., 1997)]. A limited number of type III effectors have been assi gned proven or putative biochemical function (Collmer et al., 2000; Rohmer et al, 2000; Cha ng et al., 2004) and for a subset of these (principally avirulence effectors), a plant protein or cellu lar process has been identified as

PAGE 36

24 a possible target for pathoge nesis (Rohmer et al, 2004 and Chang et al. 2004). In two cases, a bacterial effector-triggered plant phe notype has been shown to be required for pathogenesis. In the case of pathogeni city factor DspA, a member of the P. syringae AvrE family, its induction of reactive oxygen species rele ase by the host cell has been shown to be required for successful colonizati on (Venisse et al., 2003) While in the case of pathogenicity factor PthA, a member of the Xanthomonas AvrBs3/PthA family, its induction of cell division and /or cell expans ion is required for pa thogenesis (Swarup et al., 1991) In this study the compatible interaction between citrus and Xanthomonas citri ( X. citri pv. aurantifolii syn X axonopodis pv. aurantifolii) was examined. Probably originating in Southeast Asia, citrus canker ha s now spread to most citrus producing areas of the world and causes severe economical losses (Civerolo, 1994). All canker strains induce similar disease phenotypes, including water soaked lesions, formation of large hyperplastic erumpent pustules (cankers) on a ll aerial plant parts, and rupture of the epidermis with accompanying cell death (Swa rup et al., 1991 and Duan et al., 1999). Specific members of the avrBs3/pthA gene family are required by strains of Xanthomonas citri to cause cankers on citrus (Swa rup and Gabriel, 1989; Swarup et al., 1990, Swarup et al., 1991). Members of the avrBs3 / pthA gene family are found in many xanthomonads (Gabriel, 1999; Vivian and Ar nold 2000), and all citrus canker strains examined carry multiple members of the gene family (Gabriel, 1999). All Xanthomonas avrBs3/pthA members described to date are 9097% identical in DNA sequence and are characterized by 1) a series of 12.5-25.5 almo st identical 34 amino acid repeats in the center of the protein that determines host specificity, pathogenici ty and/or avirulence

PAGE 37

25 phenotype (Herbers et al., 1992, Yang et al ., 1994, Zhu et al., 1998, and Yang et al., 2000), 2) three C-terminal nucl ear localization signals (Ya ng and Gabriel, 1995; Van den Ackerveken et al., 1996 and Sz urek et al., 2001) and 3) a Cterminal acidic region considered to function as an eukaryotic tr anscriptional activator (Zhu et al., 1998, Zhu et al., 1999, Yang et al., 2000, Szurek et al, 2001). Sequence and functional anal ysis of members of the avrBs3/pthA gene family showed that these proteins are type III effect ors, acting in the plant nucleus potentially as transcriptional regula tors (Yang and Gabriel, 1995, Zhu et al., 1998, Zhu et al., 1999, Yang et al., 2000, Szurek et al., 2001) When the pathogenicity gene pthA from X.citri was transiently expressed in susceptible plant cells (by Agrobacterium infection or particle bombardment delivery), it elic ited canker-like pustules, indicating that pthA alone was sufficient to trigger canker sy mptoms (Duan et al., 1999). Unlike pthA and its active homologues in other X. citri pv. citri and X. citri pv. aurantifolii strains, avrBs3 is not required for pathogenicity of X. c. pv. vesicatoria (Bonas, 1989). However, it was found to induce a subtle hypertrophy in the mesophyll of leaves inoc ulated with slow-growing strains of X. c vesicatoria, concomitant with the up-regulation of 13 plant genes (Marois et al., 2002). Taken together these results indi cate that members of this gene family are able to induce transcriptional reprogramming in both suscepti ble and resistant plant cells. In this study, the citrus canker system was used to probe the functions of pthB another member of the avrBs3/pthA gene family, that is isofunctional with pthA in eliciting host-specific symptoms. A comparativ e analysis of gene expression in citrus leaves inoculated with the wild type X. c aurantifolli strain B69 (carrying pthB ) and an Xca mutant derivative carrying a defective (marker interrupted) pthB was performed.

PAGE 38

26 Methodological approaches in this analys is included differential display-reverse transcriptase PCR (Liang and Pardee, 1992), suppressive subtrac tion hybridization, and microscopy. Forty-six clones of citrus canke r responsive genes belonging to several broad categories of cellular functions were identified as being sp ecifically regulated by pthB These categories included genes identified to be involved in cell wall loosening and growth, water homeostasis and vesicle traffi cking. In addition evidence is presented for the involvement of hormone signaling in canker disease development. Material and Methods Plant and Microbial Material Bacterial strains and plasmids used in th is study are listed in Appendix A. B69 and BIM2 were grown on PYGM (De Feyter et al. 1990) supplemented with 35 mg/L spectinomycin and 35 mg/L spectinomycin pl us 35 mg/L Chloramphenicol. All citrus plants ( Citrus paradisi ‘Duncan’, grapefruit) were gr own under greenhouse conditions. Plant inoculations involving all citrus canker strains were carried out under quarantine at the Division of Plant Industry, Florida Depart ment of Agriculture, Gainesville. Bacterial cells were harvested from log phase cultures by centrifugation (5,000xg, 10 min.), washed (1X) and resuspended in sterile ta p water or distilled water saturated with calcium carbonate to an OD600nm of 0.6-0.7, unless stated othe rwise. Inoculations were performed by pressure-infiltration into the abaxial leaf surface of the plants. Experimental inoculations were repeated at least three times. For differential display-reverse transc riptase PCR (DD-PCR) experiments and construction of the suppressive subtractive lib raries (SSH), inoculations were performed following a split leaf model. Strain B69 was inoculated on one side of the mid-vein; while BIM2 was inoculated on the opposite side of the mid-vein, in order to control for

PAGE 39

27 leaf to leaf variations. Tissue was harves ted 0, 2 or 7 days post inoculation (dpi) depending on the experiment. Bacterial Counts B69 and BIM2 bacterial cells were normalized to an OD600 of 0.7 and infiltrated as described previously. At 0 and 2 dpi, a tota l of 9 discs (0.28 mm in diameter) from 3 leaves (3 discs per leaves) were harveste d for each treatment and ground in 1ml of tap water. After serial dilution, the bacterial populations of wild type strain B69 and mutant strain BIM2 were counted. B acterial cell count determinati ons represent th e average of three replicate experiments. Microscopy Fresh, tender and half-expanded leaves were inoculated with a high inoculum of B69 or BIM2. At 0, 2, 7 and 14 dpi, leaf samples of an area of approximately 6 mm2 were harvested and fixed in 2% glutaraldehyde in phosphate buffer saline (PBS) for 48 hr at 4 C. They were then washed three times for 15 min each and fixed in 1% buffered osmium tetroxide overnight at 4 C. This was followed by one wash in PBS for 10 mi and by two washes in distilled water. A stepwise dehydr ation was conducted after these washes using ethanol (25%, 50%, 75%, 95% and 100%) for 10 min each step, followed by three washes in acetone for 15 min each. Samples were then infiltrated at room temperature in 30% acetone/EMbed (Electron microscopy scie nces, Pennsylvania) for 1 hr, followed by 50% acetone/EMbed for 1 hr and 70% acetone/EMbed for 2 hr. Samples were subsequently incubated in 100% EMbed overnig ht at room temperature to complete the infiltration and polymerized in fresh 100% EMbed in a 75 C oven overnight.

PAGE 40

28 Differential Display-Reverse Transcriptase PCR Two and seven days after inoculation, l eaf tissue was harvested, pooled and frozen in liquid nitrogen for total RNA extraction as described (Chang et al., 1993). Potential canker responsive (CCR) cDNAs were clone d as fragments by differential displayreverse transcriptase PCR (DD-PCR) of mRNA using 48 primer combin ations (Liang and Pardee, 1992) with the R NAimage kit from Genhunter (Nashville, TN, USA). Suppressive Subtractive Hybridizatio n (SSH) Library Construction For polyA mRNA isolation, leaves were fro zen in liquid nitrogen and stored at -80 C until extraction. PolyA mRNA was isolated from leaves using the FastTrack mRNA isolation kit (Invitrogen) according to the ma nufacturerÂ’s protocol. SSH was constructed using a cDNA subtraction kit (Clontech PCR-Se lect, Palo Alto, CA). For construction of the forward subtraction library (FS), the te ster was chosen to be the pool of mRNA isolated from B69 inoculated l eaves at 2 dpi while the driver was chosen to be the pool of mRNA isolated from BIM2 inoculated leav es, and therefore, the FS was enriched in transcripts up-regulated by pthB For the reverse subtracti on library (RS), transcripts isolated from BIM2 inoculated leaves (2 dpi) were used as tester, and therefore, while the driver was chosen to be the pool of mRNA is olated from B69 inoculated leaves. the RS library was enriched in transcripts up-regulated in the absence of pthB Potential differentially regulated clon es were sent for sequencing to the Interdisciplinary Center for Biotechnology Research (ICBR) co re at the University of Florida. Putative functions were assigne d based on annotation derived by BLAST analysis.

PAGE 41

29 Northern Blots For RNA sample prepartion, NorthernMax Formaldehyde Load Dye was used as recommended by the manufacturer (Abion Austin, TX) with 5-10 g of RNA. Samples were loaded on a denaturing formaldehyde agarose gel (1%) and electrophoresis was conducted at 5 V/cm. RNA was blotted on Ge neScreen Plus hybridization transfer membrane (NemTM Life Science Products, MA) using 20X SSC as transfer buffer. Hybridization and washes were done as recommended by the manufacturer (Ultrahyb, Ambion Austin, TX). Probes were made w ith DECA primeTMII (random priming), (Ambion Austin, TX) as recommended. Reverse Northern Blots For reverse northern blots, cDNAs iden tified by DD-PCR or SSH were amplified using vector primers and purified using Qi aquick columns in plate format (Qiagen, Valencia CA). Membrane arrays were made essentially as described by Desprez et al., (1998). cDNAs were arrayed onto Hybond N+ membranes (Amersham Biosciences, Piscataway, NJ) using a 96-pin colony repli cator (V&P Scientific, San Diego CA). Six replicate arrays were generated and used to analyze transcript abunda nce of a subset of potential canker responsive genes or CCRs. Each cDNA was spotted in two locations, and several cDNAs were represented by more than one clone. Three replicate membranes for each treatment (B69 or BIM2 infection) we re used in hybridization experiments (total of six membranes or 3 pairs). Each memb rane was probed with radiolabelled cDNA synthesized from RNA isolated from one of three split leaf-experiments conducted, 2 dpi. Each membrane pair was one of three biol ogical replications. Signal intensities were statistically compared after normalization.

PAGE 42

30 For probe preparation, first strand cDNA probes were prepared from 10 g of total RNA by reverse transcription using MMLV-RT (Gibco-BRL, Gaithersburg MD) in the presence of 32P-dCTP. Unincorporated nucleotides were separated from first strand cDNA using Sephadex G-50 columns (Amers ham Pharmacia Biotech, Ithaca NY) and quantified using a liquid scinti llation counter (Beckman C oulter, Fullerton CA). Prehybridization, hybridization and low and high stringency washes were carried out at 65 C. Membranes were exposed to phosphorim ager screen for visualization. Spot intensities (called volumes) on the membrane arrays were quantified using a BioRad Molecular Imager FX run with the associ ated Quantity One software (Bio-Rad Laboratories, Inc. Hercules, CA). Data were imported into Microsoft Excel (Microsoft Corp., Redmond, WA, USA) for further analysis. Statistical Analysis A mixed model analysis (SAS Proc Mixed) was run on th e log base 2 transformed (normalization) local background adjusted volumes. cDNAs th at did not exhibit a mean value greater than 120 from either treatment we re not included in the analysis. The linear model used included replication (three biologi cal replications), trea tment (B69 treated or BIM2 treated) and gene (CCRs or Citrus Canker responsive clones). Least square means for the treatment by gene interaction were saved and used to form by-gene contrasts between treatments. Significance of these c ontrasts was controlled for an experimentwide alpha level.

PAGE 43

31 Results Macroscopic Disease Phenotype of Citrus Leaves Inoculated with X. c. aurantifolii B69 and Its Mutant Derivative BIM2 Lacking the Pathogenicity Gene pthB Xanthomonas strains B69 (wt) and its nonpathog enic mutant derivative BIM2, carrying a marker integration in gene pthB ( pthB ::pUFR004), were i noculated at high levels (OD = 0.7) on tender half-expanded leav es of new flushes of Duncan grapefruit and the corresponding induced disease phenotyp e analyzed. At day two post-inoculation, no symptoms were visible and no macroscopi c differences were observed among leaves inoculated with tap water, B69 or BIM2. By se ven days post-inoculati on, leaves that were mock inoculated showed no symp toms, while leaves inoculated with the wild type strain B69 showed a whitish canker phenotype, typical of South American canker disease. On the abaxial side of the leaf, the entire inocul ated area became raised, with a soft, velvetlike appearance, while a few individualized pustules appeared at the margins of inoculated areas. Pustules po ssibly corresponded to areas wher e bacteria were infiltrated at low density (Figure 2-1, A and B). On th e adaxial side of th e leaf, no raising was apparent; instead some yellowing develope d. This rapid symptom development is typically observed when a high inoculum is used on fresh, young expanding leaves. By contrast, at 7 dpi, no ma jor symptoms were visible on leaves inoculated with BIM2. Limited raising of the epidermis occu rred at the margins of some inoculation zones, with development of mi nimal pustule-like structures re miniscent of those seen in canker (Figure 2-1, C and D). These symptoms were not observed in mock-inoculated leaves. BIM2 inoculated leaves ultimately displayed attenuated canker phenotypes after thirty days (Figure 2-2). This is possibl y due to the week can ker-inducing activity of pthB0 the second pthA homologue found in the B strain, B69.

PAGE 44

32 PthB-Dependent Transcriptional Reprogr amming Induced upon Infection with Xca A small scale DD-PCR was conducted to co mpare transcript levels of leaves inoculated with B69 to those of leaves inoculated with BI M2 at two and seven dpi. To maximize the homogeneity and the intensity of the response, B69 and BIM2 were inoculated at high levels (OD600 of 0.7). In order to minimi ze leaf-to-leaf variation, a split-leaf inoculation strategy wa s used. An average of fifteen leaves (from three trees) were inoculated with B69 on one side of th e mid-vein and with BIM2 on the other side. Two and seven days after inoculation, ha lf-leaves were harvested, pooled into “B69 treated” or “BIM2 treated” samples and R NA extracted from both samples. Since B69 (carrying pthB and pthB0 ) differs from BIM2 (carrying pthB ::pUFR004 and pthB0 ) only by the presence of a single e ffector, PthB, differentially re gulated transcripts (named c itrus c anker r esponsive or CCRs) were PthB respons ive. Transcripts identified by DDPCR appeared differentially re gulated as early as two days post-inoculation despite a complete lack of symptoms. Twenty cDNAs were identified by DD PCR (Table 2-1), including six with homology to biotic or abio tic stress response genes (CCR20.2 to PR-1 proteins and CCR9.5, CCR15.1 to PR-5 prot eins, CCR2.2, CCR17.2 to peroxidases and CCR12.1 to catalases). One cDNA, CCR6.4 di splayed homology to cell wall remodeling enzymes of the cellulase family. CCR25.1 wa s homologous to the small ubiquitin like modifier SUMO. To remove the possibility that potential changes in transcript level were due to differences in the number of bacteria present in B69 inoculated leaves compared to BIM2 inoculated leaves, both bacter ial populations were monito red at 0 and 2 dpi. B69 and BIM2 bacterial populations were found to be comparable with almost no growth observed during the first two days post-inocula tion (Figure 2-3). Bacter ial growth at 2 dpi

PAGE 45

33 will occur if bacteria are inoculated at lower initial levels (OD600 of 0.3-0.4) (data not shown). However, when inoculated at lower levels, growth of BIM2 is very poor (data not shown and discussed later). Construction of Two Libraries Enriched in pthB Responsive cDNAs Following the same split leaf scheme as for the DD-PCR experiment, forward and reverse libraries were constructed by suppres sive subtraction hybridiz ation (see Figure 24 for illustration of the methodology), extend ing the collection of putative CCRs. The forward subtraction library ( FS) was constructed to be enriched in transcripts upregulated by PthB while the re verse subtraction library was co nstructed to be enriched in transcripts up-regulated in the absence of Pt hB (see Materials and Methods for design of the SSH). Approximately 500 clones were sequenced and annotated using homology based searches. Figure 2-5 illustra tes the distribution of CCRs for each of the FS and RS libraries according to their putative function. Categories representing genes of unknown function (8 %) and genes involved in cell gr owth and division (10%) were found more frequently in the forward library (up-regulated in the presence of pthB ) compared to the reverse library (1 % and 2 % respectively) while genes in the category representing abiotic and biotic stress responses were found mo re frequently in the reverse library (15% vs 6% in the FS) Transcript Analyses of CCRs cDNAs from each of the forward (131) a nd reverse (161) libra ries, as well as 20 clones identified by DD PCR (total of 312 cDNAs ) (Figure 2-6) were chosen for reverse northern-blot analysis. CCRs homologous to genes of known function were preferentially selected. Six replicate arrays were generated as described in materials and methods, and

PAGE 46

34 used to analyze transcript abundance of a subset of potential CCRs. Three membranes per treatment were probed with radiolabelled cDNA synthesized from RNA isolated from three split leaf-experiments, 2dpi with B69 and BIM2 (three biological replicates). For each experiment, inoculated leaves were samp led from new and older flushes, were half to fully expanded and were all tender (min imal cuticle). Signal intensities were statistically compared after normalization as described in material and methods. Forty-six clones were identified as diffe rentially regulated at (p < 0.05 ) (Figure 2-7). Only fifteen out of forty-six clones were found up-regul ated in the absence of PthB, while the remaining thirty-two were found up-regulated by PthB. Ratios of transcript abundance were calculated for each cDNA. Ratios range d from -3.5 to +34.5 (sign indicating overexpression of the gene in the absence of PthB and + sign indicating an up-regulation in the presence of PthB) (Table 2-2). Identity of cDNAs Identified as Up-Regulated by the Presence of pthB in X. citri Genome Of the forty-six clones identified as di fferentially regulated, all but four clones showed significant (e-value >2e-03) matches with sequences in available databases (Table 2-2). Thirty CCRs out of forty-six were found up-regulated by the presence of pthB in the bacterial genome i.e up-regulated in B69 infected leav es compared to BIM2 infected leaves. These are listed in Table 2-2. Cell growth. Twelve clones were highly similar to genes involved in cell growth (cell wall loosening and expansion): CCR 339 was similar to cellulases; CCR1511, CCR113 were similar to expansins; CCR 889 was similar to mannanendo-1,4-beta mannosidases; CCR571 and CCR1453 were sim ilar to pectate lyases and CCR313 was similar to tonoplast aquaporin s (TIP3). Another clone of interest, CCR575, had homology

PAGE 47

35 to the early nodulin gene Enod8 (predicted cel l wall localized estera se). An additional gene represented by CCR 109, CCR959 and CC R501 had homology to a secreted cellwall-associated pollen-specific allerg en of the ole e 1 family (SAH7). Giberellic acid pathway Two CCRs had homology to the GAST1 (GA responsive genes of unknown function) family of genes. Vesicle trafficking Several clones had homology to proteins involved in vesicle trafficking. For example, CCR673 had homology to a small GTPase of the Rab family (RAB8B, Vernoud et al. 2003), and CCR1258 had homology to the beta COP protein of the COPI complex. Unknown function Another eight clones found up-regulated had either no significant homology to any sequences in av ailable databases or had sequence homology to genes of unknown function. Identity of cDNAs Identified as Up-Regulated by X citri Lacking pthB Sixteen CCRs out of forty-six were found up-regulated by X. citri lacking pthB i.e up-regulated in BIM2 infected leaves as compared to B69 inoculated leaves. These are listed in Table 2-2. Cell growth CCR243, was the only BIM2 up-regul ated gene involved in cell wall metabolism. CCR243 is homologous to ca ffeic acid methyl transferases and is involved in phenylpropanoid metabolism. GA pathway CCR 237 was homologous to cytP450 ent-keuren oxidase and CCR105 was homologous to another cytP450 (p ossibly ent-kautenoi c acid oxidase). Protein modification and stability For example, CCR409 had homology to RD21a, a drought responsive cysteine pr oteinase, and CCR915 had homology to the small ubiquitin modifiers (SUMO).

PAGE 48

36 Transport. CCR1339 and CCR1435 were homologous to a mitochondrial import inner membrane translocase and a monos accharide-H+ symporter, respectively. Unknown function Another four clones found-up regulated in BIM2 infected leaves had either no significant homology to a ny sequences in availa ble databases or had sequence homology to genes of unknown function. Northern Blot Analysis of Representative CCRs Expression of several candidate CCR ge nes identified by reverse northern blot analysis was evaluated by northern blot analys is. Leaf tissue from split-leaf inoculations using B69 and BIM2 were harvested and pr ocessed for RNA extraction. Several labeled cDNA fragments were used to probe RNA blots (Figure 2-8). As in reverse northern blot analysis, clones corresponding to expans in, cellulase, SAH7/LAT52, GAST1, Enod8 and pectate lyase showed high levels of induction. Microscopic Phenotype of B69 and BIM2 Inoculated Leaves In order to characterize the microscopi c phenotype of B69 and BIM2 infected leaves, leaf discs mock inoculated and inf ected with BIM2 or B69 were harvested and processed for light microscopy analysis (Fig ure 2-9, 2-10, 2-11 and 2-12). Leaves were pooled as fast-responding to canker when di sease symptoms were fully developed by seven dpi (see figure 2-1, A and B). Leaves were pooled as slow-responding to canker when disease symptoms were fu lly developed by 12 to 14 dpi. Slow-responding leaves At 2 dpi B69, BIM2 and mock inoculated leaves looked identical at both the macroscopic and the mi croscopic level. At 7 dpi, while mock and BIM2 inoculated leaves showed no phenotypic signs at both the microscopic or macroscopic level (data not shown and Figure 2-9, A), the first signs of canker became visible on B69 inoculated leaves, i.e regions of darker green color around the veins and

PAGE 49

37 slight swelling. At the microsco pic level, B69 leaves showed high levels of cell division occurring across all the inoc ulated area (Figure 2-9, compare B, Cto A). Intense cell expansion and cell division phase resulted in complete filling of the air spaces of the spongy mesophyll in B69 infected leaves (Figure 2-9, compare B, C to A). The number of mesophyll cells from the abax ial to the adaxial epidermis more than doubled compared to the day 0 control or day 7 BIM2 inoculated leaves, while some cells almost tripled in size (Figure 2-10, compare A, B and D to C, a nd Figure 2-12). At later stages (14 dpi), increased raising of the epidermis and whitish coloration with soft or velvety appearance were observed at the macroscopic level. Th ese phenotypes coincided with a phase of increased cell expansion (dat a not shown and Figure 2-11, co mpare A, B to C and D). While areas of cell division were still visible, a significant su bset of cells became much larger and the leaf dramatically thickened (t wice that of the contro l leaf, see Figure 2-12). A critical preliminary conclusion from these an alyses indicated that the earliest visible canker phenotype was mainly due to cell divi sion, with a moderate cell expansion, while late onset phenotypes were due to scattered but dramatic increases in cell expansion. Fast-responding leaves Macroscopic analysis indicated that cell expansion was the primary phenotype with very little cell di vision occurring (Figure 2-13, compare B, C and D to A). Furthermore, several areas of ce ll lysis, were visible immediately under the abaxial epidermis. Bacterial growth in B69 and BIM2 infected leaves Canker visible symptoms (cell division, cell expansion and resulting cell death) appeared necessary for B69 growth as very few bacteria were vi sible in BIM2 infected tissue 14 dpi while numerous pockets

PAGE 50

38 of bacteria were seen in B 69 infected tissue (Figure 2-11, compare B to C and D and data not shown). Discussion In this study, we have used macroscopi c and microscopic phenotypic analysis in combination with targeted gene discovery techniques to understand how the pathogenicity factor pthB of X. c. aurantifolii belonging to the avrBs3/pthA gene family elicits host-specific citrus canker symptoms in a compatible plant microbe interaction. The nonpathogenic mutant BIM2, lacking pthB was used in combination with the wild type strain, Xca B69, to study th e specific effects of PthB on the plant cell transcriptome. A split-leaf inoculation experimental design wa s used to minimize leaf-to leaf variations in gene expression. In addition, bacterial cells were inoculated with high inoculum to: (1) ensure near saturation of infection sites, (2) maximize the synchronicity of the host response, and (3) artificially no rmalize the levels of bacter ial populations (wild type and mutant) present during early infec tion stages of the plant leaves (up to 2dpi). In order to obtain a collection of gene s potentially responsive to PthB, two complementary techniques, DD-PCR, and forward and revers e SSH, were used to enrich for: (1) transcript up-regulated when PthB is secreted in plant cells by X citri and (2) those upregulated in the absence of a functional PthB. Transcript analysis of a subset of 312 clones was conducted using reverse northern blot technique. Statistical analysis was used to identify a list of forty-ni ne PthB responsive genes and differential regulation for a subset of these was verified by northern blot analysis. Northern blot analysis was also conducted on several CCR that did not show differential regulation by reverse northern blot analysis (Appendix B). Several of th ese showed differential regulation when northern analysis was used suggesting a bett er sensitivity than with reverse northern

PAGE 51

39 analysis. This implies that the subtracti on libraries contain additional CCR that need identification. PthB Induces Cell Division and Cell Expansion in Citrus Leaves When inoculated on citrus leaves, Xanthomonas citri pv. aurantifolii was able to cause cell division and cell expansion, c onsistent with previous reports on pthAinduced phenotypes (Duan et al. 1999). Quantification of the three visible phenotypes of canker i.e cell division, cell expansion and the resulti ng thickening of the leaves was difficult due to (1) the heterogene ity of the cells in the spongy mesophyll and (2) the heterogeneity in distribution of the abundant air filled spaces in citrus leaf tissue. Therefore, as first approximation of the phenotype, quantification measurements were performed on areas where cellular activity wa s the most dramatic (areas of intense cell expansion, cell cycle activity and thicker leaf areas). Analysis of PthB induced symptoms over time revealed that the earliest visibl e phenotype associated with canker was cell division in the infected spongy mesophyll, whereas heterogeneous but massive cell expansion was observed at later stages of the infection. Interestingly, when canker developed rapidly, i.e advanced canker symptoms at 7 dpi versus 12 to 14 dpi, symptoms of cell division were found to be reduced co mpared to slower developing canker. In addition, cell expansion was the major phenot ype, primarily affecting mesophyll cells directly under the abax ial epidermis layer. This supports the hypothesis that the primary cellular mechanism affected by PthB alterati on of the plant cell tr anscriptome is the integrity of the cell wall and the induction of cell expansion. In turn, cell division could either be: (1) a consequence of modifi cation associated with cell expansion ( e.g. changes in cell volume) and (2) due to a second and dist inct effect of PthB. However, because cell expansion constituted a major phenotype in both rapid and slow developing canker,

PAGE 52

40 induction of cell expansion may be the primary consequence of pthB functions in the plant cell. Furthermore, a specific set of genes with homology to genes involved in cell growth were identified as responsive to PthB. PthB Induces the Expression of Cell Wall Remodeling Enzymes In order to understand PthB-induced phenotypes on citrus leaves, we have identified a set of forty-six genes (CCRs) sp ecifically regulated by the presence of this effector in the plant cell. Consistent w ith the PthB-induced morphological phenotypes, several CCRs were homologous to genes invol ved in plant cell wall modifications. Expansins Among these PthB-up-regulated plan t genes, two were homologous to -expansins. The role of the expansin gene fa mily in wall loosening (polymer creep) and cell expansion has been widely docum ented (Cosgrove, 2000). Expansins are extracellular proteins that facilitate ce ll wall expansion probably by altering hydrogen bonds between hemicellulosic wall components and cellulose microfibrils (Coscrove, 1998). These can act alone to induce cell wall extension in vitro however, in vivo they act with a suite of enzymes capable of rest ructuring the plant ce ll wall (Cosgrove, 1998). Consistent with this, several CCRs homol ogous to genes associated with cell wall remodeling were also identified. Pectate lyases Among CCRs associated with cell wall remodeling, CCR571 and CCR1453 were similar to pectate lyases (PLs). These enzymes are involved in hydrolysis of wall polymers, via cleavage of de-esterfi ed pectin, thereby fac ilitating cell expansion (Carpita and Gibeaut, 1993 and Domingo et al., 1998). Although the role of bacterial secreted PLs in cell wall degradation is we ll known (Collmer and Keen, 1998), the role of endogenous plant PLs in development has not been extensively examined. In pollen,

PAGE 53

41 plant PLs are thought to initiate the looseni ng of the cell wall enabling the emergence and growth of the pollen tube (Cosgrove et al ., 1997). PLs also medi ate cell wall breakdown in the styleÂ’s transmitting tissue, allowing pe netration of the pollen (Taniguchi et al., 1995, Wu et al., 1996). Thus, induction of plant PLs by PthB can help account for aspects of the disease phenotype. Cellulases Another PthB up-regulated CCR wa s homologous to the cellulase family, another class of ce ll wall remodeling enzymes. Cellulases catalyze the cleavage of internal 1,4 linkages of cellulose and are involved in seve ral aspects of plant development involving cell wall modifications including abscission, fruit softening and cell expansion (Lewis and Koehler, 1979, a nd Fisher and Bennet, 1991). Relevant to PthB induced phenotypes, it has been shown that constitutive expression of a poplar cellulase in A. thaliana led to a significant increase in cell size (Park et al., 2003). Beta-endo-mannanase In addition to CCRs homologous to expansins, PL, and cellulases, a fourth type of cell wall remodeling enzyme, a mannan endo-1,4D mannosidase (endo-beta-mannanase) was also identified as up-regulated by PthB. This enzyme catalyzes the hydrolysis of 1-4D mannosidic linkages in mannans, galactomannans, glucomannans and gala ctoglucoomannans (Matheson and McCleary, 1985 and Matheson, 1986) and has been implicat ed in cell wall weak ening during anther and pollen development (Filichkin et al., 2004) and in seed ripening where it is involved in mobilization of the mannan-containing cell walls of the tomato seed endosperm (Mo and Bewley, 2003). Caffeic acid methyl transferase Only one gene involved in cell wall metabolism was down regulated by PthB, CCR243. This clone was homologous a caffeic acid methyl

PAGE 54

42 transferase (COMT), belonging to the phenyl propanoid pathway that leads to lignin biosynthesis. Its expression has been shown to be regulated by biotic and abiotic elicitors including infection by avirulent and virulent bacteria (Toquin et al., 2003). It is possible that down-regulation of this enzyme relates to down-regulation of defense responses by down-regulation of lignin deposition. This ev ent could occur due to alterations in the lignin content or composition. COMT down -regulation is in acco rd with the cell expansion induced by PthB since mature wa lls lack acid-induced extension (Cosgrove, 1989). It is also interesting that fully expanded mature leaves are more resistant to canker, whereas young leaves (one half to two-third expanded) are the most sensitive ones (Graham et al., 2004). This is consistent w ith the hypothesis that PthB targets the cell wall, inducing cell expansion ultimately resulting in di sease progression. A synthesis of our results i ndicates that type III effe ctor PthB triggers the upregulation of an array of proteins whose combined activities induce cell wall loosening and cell expansion. The roles of expansins, PL s and cellulases in ce ll wall loosening have been shown to be complementary in othe r systems (Cosgrove et al.,1998; Carpita and Gibeault; 1993, Domingo et al ., 1998, Inouhe and Nevins, 1991). Enod8 and SAH7/LAT52 are a Link Betw een Canker Symptoms Development and Nodule Organogenesis and Pollen Tube Growth Respectively Two additional classes of CCRs (C CR575 and CCR109, 959 and 501) identified as up-regulated by PthB also s upport the theory that this effector targ ets cellular growth. The first one, CCR575, was homologous to Enod8 an early nodulin gene associated with the development of rhizobial nodule structures prior to nitrogen-fixat ion (Dickstein et al., 1988, 1993). Enod8 has sequence similarity to ex opolygalacturonase and lanatoside 15Â’O-acetylesterase (Pringle and Dickstein, 2003) Intriguingly, the up-regulation of Enod8

PAGE 55

43 in response to X. citri and Rhizobium suggests some common steps between nodule formation and canker pustule formation. This is also supported by the fact that both infections trigger cellular reprogramming events that lead to cellular growth. The function of Enod8 is unknown, but in-vitro characterization and se quence analysis predict that it is a cell wall localized esterase with acetylated ol igoor polysaccharides as substrates (Pringle and Dickstein, 2004). T hus the enzymatic activity of Enod8, its cell wall localization and involvement in both nodul e and canker pustule formation point to its involvement in modification of cell wall components during cellular growth. The second class of CCRs reinforcing the hypot hesis that the cell wall is the target of PthB, displayed homology to SAH7 and LAT52 genes encoding for members of the ole e I family of proteins. Originally identified as pollen allergens, members of this family have also been found expre ssed in other tissues (e.g. SAH7 in leaves). A recent study of one homologue LAT52 (tomato), indicates that these genes may be involved in controlling hydration and pollen tu be growth (Tang et al., 2002). LAT52 interaction with the pollen receptor kinase LePRK2 (LRR kinase ) led to the hypothesi s that binding of LAT52 initiates a signal transduction pathwa y required for pollen germination and pollen tube growth (Tang et al., 2002 and Johnson and Preuss, 2003). The up-regulation of a LAT52like gene in canker might, therefore, be part of a signali ng pathway leading to cell growth (the phenotype of both canker and pollen tube). Interestingly, pollen tube growth, which occurs by tip extension, i nvolves expansion and deposition of cell wall precursors at the growing tip and requires th e concerted action of endo-beta-mannanase, expansins and pectate lyases (Marin -Rodriguez et al., 2002, Cosgrove, 1998 and Filichkin et al., 2004), also found up-regulat ed during canker symptoms development.

PAGE 56

44 PthB Induces Up-Regulation of a Tonoplast Aquaporin CCR313, identified as up-regulated by PthB, displayed sequence similarity to a tonoplast aquaporin of the TIPs family (Maurel, 1997 and 2002, and Hill et al., 2004). Besides cell wall loosening, expansion requires extensive solute and water uptake resulting in the formation of a prominent vacuolar compartment. This maintains the turgor pressure that drives cell expansion (Veytsman and Cosgrove, 1998). Expansion is thought to require high hydrolic permeability of the tonoplast in order to support water entry into the vacuole, and tonopl ast aquaporins (TIPs) play a cr itical role in this process (Ludevid et al., 1992; Chaumont et al., 1998) TIPs are enriched in zones of cell expansion (Tyerman et al. 2002) as well as in zones of active cell division where their upregulation is linked to vacuol e biogenesis (Marty, 1997). Whet her the identified tonoplast aquaporin is indeed a marker for cell divisi on and/or is actively i nvolved in driving the cellular expansion is unknown. PthB Induces Up-Regulation of Two Compon ents Involved in Vesicle Trafficking Cell expansion and cell division both re quire deposition of new wall components into the extending cell walls (Veytsman and Cosgrove, 1998) or into the cell plate of dividing cells (Staehlin and He pler, 1996 and Samuels et al 1995). This may be achieved via secretory processes involving vesicle traffi cking. However, most genes identified here suggest that in response to canker, cell walls are mainly extended w ithout the building of new cell wall components. This would imply th at walls become thinner as cells expand. This indeed has been observed at late stag es of canker (Figure 2-10 compare A, B to C and D). Although plant cell walls generally appear not to become thinner as they extend (Veytsman and Cosgrove, 1998), expansion with out new cell wall deposition could be at the origin of the cell lysis obser ved in advanced canker stages.

PAGE 57

45 Several CCRs identified as up-regulated by PthB are involved in vesicle trafficking. Among these, CCR673 and CCR12 58 have homology to RAB8B and beta COP respectively. RAB8B is a member of th e small GTPase gene family. The yeast homologue of Rab8 (also named RABE see Vernoud et al., 2003) regulates membrane trafficking to the daughter cell bud site (Salminen and Novick, 1987 and Goud et al., 1988). Interestingly, in tomato, members of this subfamily appear to be targeted by the Pseudomonas avirulence factor, AvrPto. This implie s that in susceptible plants, AvrPto may interfere with membrane trafficking pathways (Bogdanove and Martin, 2000). It has been suggested that RAB8B might be involve d in polarized secre tion of antimicrobial compounds (Bogdanove and Martin, 2000). In mammals, beta COP belongs to a larg e complex that coats COPI vesicles (Kreis et al. 1995). COPI vesicl es transport membrane proteins and soluble molecules in a retrograde, and possibly anterograde, direction through mammalian Golgi stacks (Nickel and Wieland, 1997 and Harter, 1999). In plants little is known about COPI vesicles. Recent evidence suggests that CO PI-like vesicles are functional in plant secretion and localize mainly to the Golgi appara tus as well as to the cell plate of dividing cells (Couchy et al, 2003). Hormone Pathways are Possibly Involved in Canker Symptoms Development Triggering of cell expansion as well as induction of expansins and pectate lyases constitutes a common point between the effect of pthB on citrus (this study) and that of the avirulence effector avrBs3 on susceptible pepper plants (Marois et al., 2002). Cell expansion induction by both effectors share similar features; however, several plant auxin-induced proteins of the SAUR family were found up-regulated by avrBs3 (Marois et al. 2002). Several clones identified as pthB responsive are regulated by auxin in other

PAGE 58

46 systems. These include the expansins (Cat ala et al., 2000, Civello et al., 1999, Hutchison et al., 1999) and the pectate ly ases (Domingo et al., 1998). In the pepper model, one of two identified -expansins was found up-regulated by exogenous application of auxin; whereas a second expansin as well as a p ectate lyase were not (Marois et al., 2002). These data suggest that an auxin-indepe ndent pathway might ope rate under certain conditions leading to cell expansion. In addi tion to a possible role of auxin in canker disease, there is evidence for the involvement of the gibberellic acid signaling pathway in the plant response to pthB CCRs with homology to entkaurenoic acid oxidase and possibly to ent-kaurene oxidase (KO) (of GA biosynthetic pathway) (Oszewski et al., 2002) and two clones with homology to the GAST1 family (GA induced genes) were identified. Interestingly the GAST1 homologues were up-regulated by pthB ; whereas the putative KO and KAO were down-regulated. This may be explained by feed-back regulation of KAO and KO expr ession by GA. However, feedback regulation of several enzymes of the GAs biosynthetic pathway has b een described in other systems, it has not been reported to occur in the case of KAO and KO (Olszewski et al., 2002). GA is known to regulate TIPs (Phill ips and Huttly, 1994, Ozga et al., 2002), expansins (Oka et al., 2001, Vogler et al., 2003, Lee and Kende, 2002, Chen and Bradford, 2000), GAST1-like genes (Kotilain en et al., 1999 and Aubert et al., 1998), endo-beta-mannanase (Dutta et al., 1997, Yamagu chi e al., 2001) and cellulases (Litts et al., 1990). Therefore, PthB may act on regulatory steps upstr eam of GA biosynthesis. The involvement of GA does not preclude that auxin is also involved since the latter is able to regulate the production of the bioactive GA1 in elongating shoots (Ross et al., 2000).

PAGE 59

47 Indeed, these two hormones are known to, in concert, promote cellular division and elongation (Cleland, 2001 and Davies, 1995). Conclusions and Future Prospects The tight relationship between cell divi sion and cell expansion makes it difficult to address the question of whether cell expansion or cel l division are the cellular pathways that are altered as a downstream c onsequence of PthB regulating the plant cell transcriptome. However, the following resu lts presented here s upport the hypothesis that cell wall loosening and expansion is the major plant cellular mechanism targeted by PthB: 1) cell expansion occurs whether canke r symptoms develop ra pidly or slowly, 2) genes involved in cell expansion have been identified as responsive to PthB, 3) cell expansion is triggered by one another member of the avrBs3/pthA gene family and 4) PthB responsive genes are involved in cell growth. Microscopic analysis of leaves show ing a slow canker symptom development indicated that cell division is the major visibl e phenotype in initial infection stages, while cell expansion remains at a moderate level. Du ring the late infection stage however, cells dramatically expanded leading to areas of cell lysis. It is possible that PthB induces cell expansion and cell division by targeting seve ral distinct cellular mechanisms. Another hypothesis is that PthB targets cell e xpansion by altering cell wall composition (loosening). This in-turn lead s to a cell autonomous respons e that mainly involves the triggering of cell division in the early st ages and massive cell wall loosening and expansion in later stages. The concentration of bacteria su rrounding infected cells and, therefore, the concentration of PthB protein secreted into the plant cells as well as the physiological state of the infected tissue (for example immature expanding leaves will readily expand) would modulate this response. When the conc entration of PthB is low,

PAGE 60

48 moderate expansion and the subsequent cha nge in cell volume would lead to cell division, while in later stages, elevated concentrations of Pt hB would lead to gross cell expansion and cell ly sis (Figure 2-14). The relationship between cell expansion a nd cell division in plant growth and development remains controversial. Whether gr owth starts by an increase in cell size, triggering division, or whether division occurs first followed by restorat ion of the original cell size (Foard, 1971 and Cleland, 2001) is mainly unknown. Studies on leaf primordial (LP) initiation may begin to resolve this issu e. Initially, since the first visible sign of a new LP is a periclinal division in the L1 or L2 layer of the shoot apical meristem, it was suggested that division occurs first (Steev es and Sussex, 1989). However, recent evidence indicates that cell enlargement is the first step in LP initiation since LPs can be induced by adding expansins either by microinjection of by up-regulation of e xpansin transcripts at the shoot apical meristem (Pien et al ., 2001, Fleming et al. 1997). Canker could follow a similar pattern where cells expand first a nd then divide in response to expansion. The canker phenotype is necessary for optimal growth and dispersal of X. citri (Swarup et al., 1991 and this study); therefore, induction of cell division and or expansion are key steps in canker disease development and, unlike AvrBs3 for Xcv PthA/B confer a benefit to X. citri strains carrying it The PTHA/B family of pa thogenicity effectors may prove to be a valuable tool in dissec ting the molecular events surrounding microbeinduced diseases since they are require d for pathogenesis and can induce canker symptoms alone. Finally, an understanding of the mechanisms by which PthB induces canker phenotypes could help unravel the intr icate relationship between cell division and cell expansion that occurs in plant development.

PAGE 61

49 Table 2-1: List of putative CCR identified by DD-PCR. CCR Homology e-value CCR23.2 Unknown protein [ A. thaliana] (NP_196103.1) 2e-37 CCR24.5 Putative protein [ A.thaliana ] (NP_195874.1) 3e-24 CCR1.1 Putative Transposase [ A. thaliana ] (NP_189803.1] 4e-33 CCR22.5 3-hydroxyisobutyryl-coA hydrolase [ A. thaliana ] (NP_193072.1) 1e-22 CCR27.1 Cytochrome P450 [soybean] (T05942) 5e-41 CCR11.4 Cytochrome f [ Nicotiana tabacum ] (NP_054512.1) 4e-53 CCR6.2 Phosphoribosyl pyrophosphate synthase [ Spinacia oleracea ] (CAB43599.1) 6e-25 CCR28.2 Putative mitochondri al carrier protein [ A. thaliana ] (NP_181124.1) 4e-34 CCR7.6 Copper Transport Protein [ A. thaliana ] (NP_200711.1) 4e-33 CCR8.2 Receptor-like protei n kinase-like (LRR) [ A.thaliana ] 6.8e-45 CCR6.4 Cellulase [ sweet orange ] (eC3.2.1.4) 1e-23 CCR28.4 Peroxidase [ A. thaliana ] (CAA66035.1) 2e-50 CCR12.1 Catalase [ Campylobacter jejuni ] (Q59296) 2e-10 CCR2.2 Bacterial-induced peroxidase [ Goss hirsutum ] (AF155124) 3e-26 CCR17.2 Peroxidase [ Nicotiana tabacum ] (BAA82306.1) 6e-63 CCR20.2 Pathogenicity-related protein 1a [barley] (AF245497) 2e-43 CCR15.1 Osmotin-like protein [ Fagus sylvatica ] (AJ298303) 2e-17 CCR9.5 Osmotin -like protein [ Fragaria x ananassa ] (AF1999508) 3e-61 CCR21.1 Auxin induced protein, putative [ A. thaliana ] (NP_176274.1) 1. 9e-3 CCR25.1 Ubiquitin-like protein [ A.thaliana ] (NP_194414.1) 4e-28

PAGE 62

50 Table 2-2: List of CCRs confirmed by reverse northern blot analysis. CCR Homology e value Ratio* ( 0.05) Ribosomal protein 1385 30S ribosomal protein S20 ( A. thaliana ) gi21592469 1e-28 -3.36 Unknown function 1065 EST ( O. sativa ) gi50919279 4e-11 4.79 1243 EST ( A. thaliana ) gi42569501 2e-40 -3.2 497 Putative protein ( O. sativa ) gi50919279 4e-27 4.89 767 Putative protein ( A. thaliana ) gi15241855 3e-24 -2.41 1111 No significant homology 2.95 171 No significant homology 3.60 137 No significant homology 3.09 809 No significant homology -3.14 1139 Ring Finger Protein ( A. thaliana ) gi26450511 5e-07 3.86 1061 Splicing factor RSZp22 ( A. thaliana ) gi21554419 2e-03 3.16 475 Zinc finger protein ( A. thaliana ) gi28416541 1e-07 -2.55 1312 LRR receptor kinase ( A. thaliana ) gi42562316 1e-43 3.25 Metabolism/energy 539 CytoF ( N. tabacum ) gi11465970 4e-53 -3.05 1415 RubisCO activase ( malus x domestica ) gi415852 9e-58 -2.60 1239 F1F0 ATPase inhibitor protein ( O. sativa ) gi 52077175 4e-10 -2.91 1057 Hydroxymet hyltransferase ( A. thaliana ) gi21593312 4e-73 3.53 901 UMP-kinase ( A. thaliana ) gi2497486 3e-38 2.64 1445 UMP-kinase ( A. thaliana ) gi2497486 1e-10 6.92 33 UDP-galactose epimerase ( A. thaliana ) gi9758701 4e-21 3.58 Transport 1339 Mitochondrial import inner membrane translocase S.U gi42568553 1e-19 -2.62 343 Copper T protein ( A. thaliana ) gi15237802 4e-33 4.92 1435 Monosaccharide-H+ symporter ( D. glomerata ) gi30349804 2e-13 -3.36 Protein modification/stability 915 Small ubiquitin-like modifier ( A. thaliana ) gi15236885 4e-28 -2.53 1262 Protease inhibitor/seed storage//LTP ( A. thaliana ) gi42567284 6e-04 5.35 1345 Aminopeptidase ( A. thaliana ) gi34098848 2e-21 3.09 409 Putative cysteine proteinase RD21A ( A. thaliana ) gi22136972 5e-32 -2.99 GA pathway 1535 GAST1-like protein ( A. thaliana ) gi25406361 3e-10 11.00 493 GAST1-like protein ( A. thaliana ) gi25406361 3e-34 6.10 237 Cyt. P450 ent-keuren oxydase ( Malus x domestica ) gi45551401 7e-44 -3.2 1051 Cyt P450 (possibly ent-kaurenoic acid oxidase) ( P. sativum ) gi27776451 1e-23 -2.46 Vesicle trafficking 673 RAB 8B ( Lotus corniculatus ) gi1370192 1e-28 6.68 1258 beta COP protein ( O. sativa ) gi50900798 7e-20 2.38 279 Phosphatase (put. membrane trafficking factor) ( A. thaliana ) gi21553471 3e-19 -2.51

PAGE 63

51 Table 2-2. Continued CCR Homology e value Ratio ( 0.05) Cell growth (cell wall metabolism and expansion) 889 Mannan endo 1,4 beta mannosidase ( O. sativa ) gi34912090 4e-15 3.53 113 Alpha expansin ( P. cerasus ) gi13898655 4e-49 4.06 1511 Alpha expansin ( P. cerasus ) gi13898655 4e-51 4.26 571 Pecate lyase ( malus x domestica ) gi 34980263 1e-54 7.89 1453 Pectate lyase ( A. thaliana ) gi21593312 12-15 4.69 243 Caffeic acid O-methyl transferase ( C. roseus ) gi 18025321 6e-59 -2.39 339 Cellulase (sweet orange) gi7488904 1e-23 34.53 575 Enod8 (early nodulin 8 like) ( A. thaliana ) gi26451820 3e-07 4.59 313 Tonoplast aquaporin gamma TIP (TIP3) ( A. thaliana ) gi3688799 5e-13 4.82 109 SAH7/LAT52 (ole e I allergen family) ( L. esculentum ) gi 295812 7e-17 3.78 959 SAH7/LAT52 (ole e I allergen family) ( L. esculentum) gi 295812 5e-11 3.73 501 SAH7/LAT52 (ole e I allergen family) ( L. esculentum ) gi 295812 9e-21 2.75 *: A positive ratio indicates up-regulation in B69 infected tissue compared to BIM2 infected tissue. A negative ration indicat es up-regulation in BI M2 infected tissue compared to B69 infected.tissue.

PAGE 64

52 Figure 2-1: Phenotype of B69 and BIM2 in fections on grapefruit leaves. BIM2 lacks PthB and induces formation of very sm all pustule like structures, reminiscent of canker pustules, at the edges of some inoculated areas. (A), (C) are BIM2 (pUFR004:: pthB ) inoculations on grapefruit leav es and (B), (D) are wt B69 inoculations. Pictures were taken 7 days post inoculation. AB C D AB C D AB C D

PAGE 65

53 Figure 2-2: Late B69 and BIM2 phenotypes. (A ) BIM2 inoculated l eaves 30 dpi and (B) B69 inoculated leaves 30 dpi. Note th e much attenuated phenotype of BIM2 infected leaves. B A B A

PAGE 66

54 Figure 2-3: Quantification of bacterial population two days post inoculation with B69 and BIM2. (cfu: colony forming unit), Exp1: experiment 1, Exp2: experiment 2). 1.E+00 1.E+02 1.E+04 1.E+060dpi2dpi0dpi2dpi0dpi2dpi0dpi2dpi B69B69BIM2BIM2 Exp1Exp2Exp1Exp2 cfu/cm2 1.E+00 1.E+02 1.E+04 1.E+060dpi2dpi0dpi2dpi0dpi2dpi0dpi2dpi B69B69BIM2BIM2 Exp1Exp2Exp1Exp2 cfu/cm2

PAGE 67

55 Figure 2-4: Diagram of PCR-Select cDNA subt raction. Type e molecules are formed only if the sequence is up-regulated in the tester cDNA. Solid lines represent the Rsa Idigested tester or driver cDNA. Solid boxes represent the outer part of the Adaptor 1 and 2R longer strands and corresponding PCR primer 1 sequence. Green boxes represent the inner part of Adaptor 1 and the corresponding Nested PCR primer 1 se quence; red boxes represent the inner part of Adaptor 2R and the corresponding Nested PCR primer 2R sequence. Poly A+RNA IsolationPoncirus trifoliata Cold Acclimated at 4 C for 2 days Poncirus trifoliata Non Acclimated ControlcDNA Synthesis by Reverse Transcriptase Double-stranded cDNA Synthesis Restriction Enzyme Digestion with Rsa I Adaptor Ligation to the Tester DNA Driver Tester Driver Adaptor 1Adaptor 2R AAAAAAA TTTTTTT TTTTTTTT AAAAAAA AAAAAAA TTTTTTTT TTTTTTTT AAAAAAA First Hybridization 68 C for 8 hrs a b c d a b c d a b c d e Second Hybridization 68 C for 16 hrs Driver a b c d e Fill in the ends a and d No amplification b bÂ’ No amplification c Linear amplification PCR Amplification using an Adaptor Primer 5Â’ 3Â’ 5Â’ 3Â’ e Exponential amplification Poly A+RNA IsolationPoncirus trifoliata Cold Acclimated at 4 C for 2 days Poncirus trifoliata Non Acclimated ControlcDNA Synthesis by Reverse Transcriptase Double-stranded cDNA Synthesis Restriction Enzyme Digestion with Rsa I Adaptor Ligation to the Tester DNA Driver Driver Tester Driver Adaptor 1Adaptor 2R AAAAAAA TTTTTTT TTTTTTTT AAAAAAA AAAAAAA TTTTTTT TTTTTTTT AAAAAAA AAAAAAA TTTTTTTT TTTTTTTT AAAAAAA First Hybridization 68 C for 8 hrs a b c d a b c d a b c d a b c d a b c d e a b c d e Second Hybridization 68 C for 16 hrs Driver Driver a b c d e a b c d e Fill in the ends a and d No amplification b bÂ’ No amplification b bÂ’ No amplification c Linear amplification PCR Amplification using an Adaptor Primer 5Â’ 3Â’ 5Â’ 3Â’ e Exponential amplification 5Â’ 3Â’ 5Â’ 3Â’ 5Â’ 3Â’ 5Â’ 3Â’ e Exponential amplification Citrus leaves infected with BIM2 Citrus leaves infected with B69 Poly A+RNA IsolationPoncirus trifoliata Cold Acclimated at 4 C for 2 days Poncirus trifoliata Non Acclimated ControlcDNA Synthesis by Reverse Transcriptase Double-stranded cDNA Synthesis Restriction Enzyme Digestion with Rsa I Adaptor Ligation to the Tester DNA Driver Tester Driver Adaptor 1Adaptor 2R AAAAAAA TTTTTTT TTTTTTTT AAAAAAA AAAAAAA TTTTTTTT TTTTTTTT AAAAAAA First Hybridization 68 C for 8 hrs a b c d a b c d a b c d e Second Hybridization 68 C for 16 hrs Driver a b c d e Fill in the ends a and d No amplification b bÂ’ No amplification c Linear amplification PCR Amplification using an Adaptor Primer 5Â’ 3Â’ 5Â’ 3Â’ e Exponential amplification Poly A+RNA IsolationPoncirus trifoliata Cold Acclimated at 4 C for 2 days Poncirus trifoliata Non Acclimated ControlcDNA Synthesis by Reverse Transcriptase Double-stranded cDNA Synthesis Restriction Enzyme Digestion with Rsa I Adaptor Ligation to the Tester DNA Driver Driver Tester Driver Adaptor 1Adaptor 2R AAAAAAA TTTTTTT TTTTTTTT AAAAAAA AAAAAAA TTTTTTT TTTTTTTT AAAAAAA AAAAAAA TTTTTTTT TTTTTTTT AAAAAAA First Hybridization 68 C for 8 hrs a b c d a b c d a b c d a b c d a b c d e a b c d e Second Hybridization 68 C for 16 hrs Driver Driver a b c d e a b c d e Fill in the ends a and d No amplification b bÂ’ No amplification b bÂ’ No amplification c Linear amplification PCR Amplification using an Adaptor Primer 5Â’ 3Â’ 5Â’ 3Â’ e Exponential amplification 5Â’ 3Â’ 5Â’ 3Â’ 5Â’ 3Â’ 5Â’ 3Â’ e Exponential amplification Citrus leaves infected with BIM2 Citrus leaves infected with B69

PAGE 68

56 Figure 2-5: Distribution of potential citrus ca nker responsive genes. ribosomal protein 20% no homology 24% abiotic and biotic stress response 6% unknown function 8% signaling 5% hormone metabolism and signaling 1% transport 4% protein stability and degradation 3% secondary metabolism 4% metabolism/energy 9% transcription and translation 6% cell growth and division 10% signaling 3% hormone metabolism and signaling 2% ribosomal protein 28% no homology 14% abiotic and biotic stress response 15% unknown function 1%metabolism/energy 14% transcription and translation 6% cell growth and division 2% secondary metabolism 6% protein stability/degradation 4% transport 5%A) FS B) RS ribosomal protein 20% no homology 24% abiotic and biotic stress response 6% unknown function 8% signaling 5% hormone metabolism and signaling 1% transport 4% protein stability and degradation 3% secondary metabolism 4% metabolism/energy 9% transcription and translation 6% cell growth and division 10% signaling 3% hormone metabolism and signaling 2% ribosomal protein 28% no homology 14% abiotic and biotic stress response 15% unknown function 1%metabolism/energy 14% transcription and translation 6% cell growth and division 2% secondary metabolism 6% protein stability/degradation 4% transport 5%A) FS B) RS signaling 3% hormone metabolism and signaling 2% ribosomal protein 28% no homology 14% abiotic and biotic stress response 15% unknown function 1%metabolism/energy 14% transcription and translation 6% cell growth and division 2% secondary metabolism 6% protein stability/degradation 4% transport 5%A) FS B) RS

PAGE 69

57 Figure 2-6: Distribution and origin of the clones stam ped on the nitrocellulose membranes used in reverse northern blot analysis. 131 16 161 4 0 50 100 150 200 DD SSH

PAGE 70

58 Figure 2-7: Cluster analysis of genes differentially regulate d by PthB. In green are genes down-regulated by PthB, and in red ar e genes up-regulated by PthB -level

PAGE 71

59 Figure 2-8: Northern blot an alysis of CCR genes found diffe rentially regulated by reverse northern blot analysis rRNA was used as control for loading. Enod8 B69 BIM2 rRNA 2 dpi 18 S TIP LAT52 Exp PL GAST1 CCR137 B69 BIM2 2 dpi cellulase B69 BIM2 rRNA 2 dpi B69 BIM2 rRNA 2 dpi Enod8 B69 BIM2 rRNA 2 dpi 18 S TIP LAT52 Exp PL GAST1 CCR137 B69 BIM2 2 dpi cellulase B69 BIM2 rRNA 2 dpi B69 BIM2 rRNA 2 dpi B69 BIM2 rRNA 2 dpi B69 BIM2 rRNA 2 dpi

PAGE 72

60 100 m 100 m 100 m 100 m 100 m 100 m 100 m 100 m 100 m A B C 100 m 100 m 100 m 100 m 100 m 100 m 100 m 100 m 100 m A B C Figure 2-9: Microscopi c phenotype of leaves inoculat ed with B69 (wt) and BIM2 (nonpathogenic mutant lacking PthB). 7dpi BIM2 infected leaves, A; 7dpi B69 infected leaves, B, C. By In canke r-infected tissue, by 7 dpi, air spaces of the spongy mesophyll are almost inexistent These spaces are replaced by new divided cells as well as by cell of larg er size, resulting in thickening of the leaves.

PAGE 73

61 Figure 2-10: Microscopic phenot ype of leaves inoculated with B69 (wt) and BIM2 (nonpathogenic mutant lacking PthB). 7dpi BIM2 infected leaves, C; 7dpi B69 infected leaves, A, B, D. At 40X magnification, pockets of bacterial cells are visible surrounding mesophyll cells of B69 infected tissue while almost no bacteria is present in BIM2 infected ti ssue. Also not the areas of cell lysis in B69 infected tissue. 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 mA B C D 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 m 50 mA B C D

PAGE 74

62 Figure 2-11: Microscopic phenot ype of leaves inoculated with B69 (wt) and BIM2 (nonpathogenic mutant lacking PthB) at 14 dpi. A, B: BIM2, and C, D:B69 infected leaves. Note High levels of bact eria in B69 infected leaves compared to BIM2 infected leaves, as well as possible wall thinning of cells in B69 infected tissue. A C B D 10 m 10 m 10 m 10 m A C B D 10 m 10 m 10 m 10 m 10 m 10 m 10 m 10 m 10 m 10 m 10 m 10 m

PAGE 75

63 Figure 2-13: Quantification of leaf thicke ning and cell division during B69 and BIM2 infection on Duncan grapefruit leaves These measurements where taken on “slow canker-developing” leaves, i.e leaves showing high rate of cell division when inoculated with B69. The number of cells from abaxial epidermis to adaxial epidermis was calculated by counti ng the number of ce lls that a virtual line perpendicular to the epidermal layers would cross. Ten lanes were used n the analysis and the number shown are averages. 0 5 10 15 20 25 30 02714 0 200 400 600 800 1000 02714 B69 BIM2 0 5 10 15 20 25 30 02714 0 200 400 600 800 1000 02714 B69 BIM2 0 5 10 15 20 25 30 02714 0 200 400 600 800 1000 02714 0 5 10 15 20 25 30 02714 0 200 400 600 800 1000 02714 B69 BIM2Days post inoculationLeaf thickness ( m) Number of cells from abaxial epidermis to adaxial epidermis (in cross section) 0 5 10 15 20 25 30 02714 0 200 400 600 800 1000 02714 B69 BIM2 0 5 10 15 20 25 30 02714 0 200 400 600 800 1000 02714 B69 BIM2 0 5 10 15 20 25 30 02714 0 200 400 600 800 1000 02714 0 5 10 15 20 25 30 02714 0 200 400 600 800 1000 02714 B69 BIM2Days post inoculationLeaf thickness ( m) Number of cells from abaxial epidermis to adaxial epidermis (in cross section)

PAGE 76

64 Figure 2-13: Microscopic symp toms of rapidly developing canker. 14dpi BIM2 infected leaves, A; 14dpi B69 infected leaves, B, C, D. Note the highly enlarged cells the large areas of cell lysis and the absen ce of high rate of cell division in B69 infected tissue. 50 m A B C D 50 m 50 m 50 m A B C D

PAGE 77

65 Figure 2-14: Possible model for PthB eff ects on susceptible citrus cell showing parallel pathways activati ng cell division and expansion. PthBSignaling molecules: LAT52, Auxin /GA Exps, PLs, cellulase, Mannanase, TIP3, Enod8, GAST1, RAB8B, BetaCOP?Cell wall loosening Cell Expansion Cell Division? Rapid multiplication of X. citri and Canker phenotypes PthBSignaling molecules: LAT52, Auxin /GA Exps, PLs, cellulase, Mannanase, TIP3, Enod8, GAST1, RAB8B, BetaCOPCell division Effector molecules:?Cell wall loosening Cell Expansion Cell Division? Rapid multiplication of X. citri and Canker phenotypes Cell expansion Effector molecules: PthBSignaling molecules: LAT52, Auxin /GA Exps, PLs, cellulase, Mannanase, TIP3, Enod8, GAST1, RAB8B, BetaCOP?Cell wall loosening Cell Expansion Cell Division? Rapid multiplication of X. citri and Canker phenotypes PthBSignaling molecules: LAT52, Auxin /GA Exps, PLs, cellulase, Mannanase, TIP3, Enod8, GAST1, RAB8B, BetaCOPCell division Effector molecules:?Cell wall loosening Cell Expansion Cell Division? Rapid multiplication of X. citri and Canker phenotypes Cell expansion Effector molecules:

PAGE 78

66 CHAPTER 3 CHANGES IN SUMO CONJUGATION ARE ASSOCIATED WITH CITRUS CANKER DISEASE Introduction Citrus canker is an important disease of citrus worldwid e (Civerolo, E., 1984). It is caused by several pathovars of Xanthomonas citri which differ mainly in their host range (Shubert et al, 2001, Verniere et al, 1998). Canker infecti ons cause defoliation, fruit blemishes, premature fruit drop and tree declin e, resulting in severe economical losses (Shubert et al, 2001). Considerable internati onal regulatory efforts are implemented to prevent the spreading of the already quar antined pathogen, with negative effects on national and international trad e of citrus (Timmer et al 1996; Shubert et al, 2002). Canker symptoms are characterized by erumpe nt corky lesions that can affect all aerial parts of citrus trees (Shubert et al, 2002). Microscopy studies showed that canker lesions result from hyperplasia (cell divisi on) and hypertrophy (cel l expansion) in the spongy mesophyll tissue, where the bacteria cont act plant cells (Swarup et al, 1991; Duan et al, 1999 and Chapter 2). Ultimately, this in tense increase in cellular growth ruptures the epidermis and causes necrosis. The rupture of the epidermis is thought to be crucial for bacterial dissemination a nd spread of the disease (Graham and Gottwald, 1991; Duan et al, 1999). A crucial step towards understanding citrus canker disease was the identification of a pathogenicity gene, pthA required by X. citri pv. citri to cause canker on citrus (Swarup et al., 1991). Since then, all canker-causing stra ins have been shown to carry at least two

PAGE 79

67 members of the pthA gene family, with one copy suffici ent for most or all pathogenicity (Yang and Gabriel, 1995; Al -Saadi and Gabriel unpublished). pthA found in X. citri pv. citri ( Xcc ) of the Asiatic group of strains, and pthB found in X. citri pv. aurantifolii B69 ( Xca ) of the South American group, have been shown to be interchangeable in their ability to elicit canker (Yuan a nd Gabriel, unpublished). As for pthA of Xcc pthB of Xca was also shown to be required for pat hogenicity on citrus (Yuan and Gabriel, unpublished, and Chapter 2), and therefore, the B69 derivative mutant strain BIM2 lacking pthB does not elicit the typical macroscopic symptoms associated with canker disease (Chapter 2). When transferred to other xanthomonads carrying a functional type III secretion system (TTSS), or tran siently expressed in leaf cells, pthA was found to induce cell division, cell expansion, and rupt ure of the epidermis the three most prevalent canker symptoms (Swarup et al 1991 and 1992; Duan et al, 1999). It was therefore concluded that pthA alone was able to cause canke r-like symptoms and that its delivery into the plant cell relies on a functional TTSS. Members of the pthA gene family are also found in non-canker causing strains of Xanthomonas Examples of genes belonging to this gene family include avrBs3 and avrBs3-2 of Xanthomonas campestris pv. vesicatoria (Bonas et al, 1989, and Bonas et al, 1993), avrXa10 and avrXa7 of Xanthomonas oryzae pv. oryzae (Hopkins et al, 1992); along with avrB4 avrb6 and avrb7 of Xanthomonas campestris pv. malvacearum (De Feyter and Gabriel, 1991 and 1993). Proteins encoded by memb ers of this gene family are 90 to 97% similar and are characterized by several structural features essential for their function in avirulence and/ or pathogenicity. Such features include 1) nearly identical 102-bp tandem repeats in their center, 2) C-te rminal nuclear localization signals (NLS),

PAGE 80

68 and 3) C-terminal eukaryotic acidic transcri ptional activator (Herbers et al, 1992; Yang et al, 1994; Zhu et al, 1998; Yang et al, 2000; Yang and Gabriel, 1995; Van den Ackerveken et al, 1996, Szurek et al, 2001). Little is known about how canker disease is initiated in planta In order to understand the molecular mechanism underlyi ng canker, a differential display PCR experiment was conducted to identify plan t genes potentially responsive to canker (Chapter 2). At two days pos t inoculation (dpi), transcri pts were compared between leaves inoculated with B69 and leaves inoc ulated with BIM2 (B69 derivative carrying a non-functional pthB ). One clone was related to AtSUMO1 from Arabidopsis SUMO belongs to the ubiquitin family of proteins that are conjugated to target proteins; however;its functions are distin ct from those of ubiquitin. SUMO conjugation has been shown to be an important regulatory step in processes such as protein stability, s ubcellular localizati on, and response to various stresses. SUMOylation is carried out in a ATP-dependa nt reaction cascade si milar to the E1-E2E3 reactions responsible for ubiquitin conj ugation (Melchior F., 2000; Kim et al, 2002; Kurepa et al, 2003). In addition, SUMO modifi cation has been shown to be important for cell cycle progression in yeast. Specifically, temperaturesensitive mutants lacking a functional SUMO conjugation pathway have been shown to arrest at the G2/M transition (Seufert et al, 1995; Johnson and Gupta, 2001). Such work is of interest, as Xanthomonas citri infection triggers division of mesophy ll cells contacted by the bacteria. Recent work has shown that strain s of the phytopathogenic bacterium Xanthomonas campestris pv. vesicatoria encode at leas t two type III effectors with demonstrated SUMO protease activity (Hot son et al, 2003; Roden et al, 2004). Though

PAGE 81

69 loss of these SUMO protease-like effectors di d not affect pathogenicity on susceptible plants, it raises the possibility that th e plant SUMO conjugation pathway could be targeted during infection by X. c vesicatoria (Hotson et al 2003; Roden et al, 2004). This study indicates that: 1) changes in plant protein SUMO ylation profiles occurred after host infection by Xanthomonas citri pv. aurantifolii, 2) these changes in SUMOylation profiles were of two types, gene pthB -dependent and independent, and 3) these changes in SUMOylation profiles di d not occur following challenge with a nonpathogenic mutant strain lacking a TTSS. Toge ther, these data indicate that the TTSS of Xca delivers one or more effectors that direct ly, or indirectly, de -conjugate SUMO from host proteins in vivo Materials and Methods Plant Inoculations All inoculations were done with needle-l ess syringes on the abaxial surface of the leaf. Plants ( Citrus paradisi ‘Duncan’ grapefruit) were grown under greenhouse conditions. Inoculations involvi ng strains B69 and its derivati ves were carried out in BL3P level containment (refer to Federal Re gister Vol.52 no 154, 1987) at the Division of Plant industry, Florida Department of Agri culture, Gainesville, FL. For inoculation, bacterial suspensions were standardized in sterile 10mM CaCO3 (mock) to an optical density of 0.5 and pressure -infiltrated. For phenotypic obs ervation, inoculations were repeated at least three times. For protein ex traction, a split leaf i noculation scheme was followed to normalize differences due to physiol ogical state of inoculat ed tissue. For each combination of treatments (i.e. mock/B69 and mock/BIM2), one treatment was inoculated on the right side of the mid-vein and the other strain on the left side of the

PAGE 82

70 mid-vein. For each split-leaf experiment thr ee trees were used, with an average of 10 leaves inoculated per tree (approximatel y 5 leaves per treatment combination). Bacterial Strains and Culture Media Bacterial strains and plasmids used in th is study are listed in Table 1 Appendix A. All Xanthomonas strains were cultured in PYGM medium at 30C (De Feyter et al. 1990). Escherichia coli were grown on Luria-Bertani (LB) medium (Sambrook et al., 1990). For culture on solid media, agar was adde d at 15 g/L. Antibiotics were used at the following concentrations: Spectinomycin (S p), 35 mg/L; Kanamycin (Kn), 12.5 mg/L; Chloramphenicol (Cm), 35 mg/L; Gentomycin (Gt), 3 mg/L. Marker Integration Mutagenesis hrpG gene knock-out mutation was generated by triparental matings (as described in Chapter 1). Briefly, a 550 bp internal fragment of hrpG was cloned in the suicide vector pUFR012 [derivative of pUFR004 ca rrying kanamycin resistance (Gabriel laboratory, unpublished)] creating pB Y23. Transconjugants resulting from E. coli DH5/pBY23, DH5/pRK2013 (helper plasmid) and B69 matings were selected on spectinomycin to select against E. coli and chloramphenicol and ka namycin to select for plasmid insertion events. Putative transconj ugants were purified to a single colony, and Southern hybridization was used to confirm the integration of suicide vector pBY23 in hrpG For complementation purposes, a Hin dIII to Kpn I fragment was cloned out of plasmid pXG8 (REF) and recloned in pU FR053 (Yuan and Gabriel, unpublished) creating pBY24. DH5/pBY24 was used in triparental matings to create B23.5/pBY24 (B23.5c and B23.5c1). Putative exconjugant s were purified to a single colony, and

PAGE 83

71 Southern hybridization was used to conf irm the presence of the complementation plasmid. Total DNA extractions were performed as described in Gabriel and De Feyter (1992). Southern hybridizations were perf ormed as described by Lazo and Gabriel (1987). Bioinformatics Alignments and box shading were carried out using Clustal W (http://clustalw.genome.jp). Protein Extraction and Western Blotting Citrus leaf tissue was harvested at 0, 2 or 7 days post inocula tion (dpi), depending on the experiment, and ground to a fine powder in liquid nitrogen. Soluble proteins were extracted in two volumes of extraction buffer (50mM Tris, pH = 8.0, 300mM sucrose, 2mM EDTA, 0.3% DIECA, 10mM N-ethylmaleimide, 1 g/ l pepstatin, 1 g/ l leupeptin, and 7.5% w/v PVPP). Extracts were vortexed and briefly sonicated, then clarified by two rounds of centrifugation at 16,000 x g for 10 min at 4C. Soluble proteins were quantified by the BCA assa y (Pierce Biotechnology, Rockford, IL). Proteins were separated by polyacryalmid e electrophoresis on a 15% Tris-Tricine gel, and transferred to PVDF membrane (Millipore, Bedford, MA). For immunoblot analysis, membranes were probed with 1:2,500 immunopurified polyclonal PopSUMO1 (gi:23997054) antiserum (Cocalico, Reamstown, PA) diluted in phosphate buffered saline (137 mM NaCl, 2.7 mM KCl, 1.4 mM K2HP O4, 10.1 mM Na2HPO4, pH 7.4) containing 0.1% Tween 20 (T-PBS) with 1% v/v goat seru m (Sigma, St. Louis, MO). The antibodies were raised against purified PopSUMO1, whic h also contained an additional N-terminal hexahistidine tag generated by PCR (Reed, J., MasterÂ’s Thesis Univ ersity of Florida, 2005). For secondary antibody, the membranes were probed with 1:25,000 horseradish

PAGE 84

72 peroxidase conjugated donkey anti-ra bbit secondary antibodies (Amersham, Buckinghamshire, England) diluted in 1X T-PBS. Chemilluminescence was carried out according to the manufacturer’s instructions using the ECL plus (+) kit (Amersham). Following chemilluminescence, each membrane was rinsed in 1X T-PBS and stained with Coomassie R250 as a loading control. Results SUMO Conjugation Profiles are Altered in X. citri-Infected Leaves The grapefruit partial cDNA, CCR915 was identified by differential display as being canker responsive. Following reverse no rthern blot analysis, CCR915 which shows homology to SUMO, was found up-regulated in leaves inoculated with BIM2 (lacking pthB ) compared to leaves inoculated with B69 (w t) (Chapter 2). To determine if shifts in SUMO transcript abundance reflected regula tion at the protein levels, a split-leaf experiment was conducted in which Duncan gr apefruit leaves were mock inoculated on one side of the mid-vein, and Xanthomonas citri pv. aurantifolii strain B69 was inoculated on the other side Soluble extracts taken from canker or mock –inoculated leaves were probed for CitSUMO and CitS UMO-conjugated protei ns using PopSUMO1 antibodies. The grapefruit sequence was highly similar to poplar SUMO isoform PopSUMO1 (gi:23997054) (Figure 3-1) and as expected, the grapefruit SUMO and its protein conjugates cross-reacted with anti bodies raised against PopSUMO1. Using anti PopSUMO1, it was found that at two days post inoculation, the profile of SUMO conjugation is noticeably altered (Figure 3-2). The amounts of free CitSUMO and high molecular weight CitSUMO conjugated proteins were higher in B69-in filtrated leaves as compared to mock-infiltrated leaves.

PAGE 85

73 SUMO Conjugation Profiles in Infected Leaves are Partially PthB Dependent To determine if SUMOylation patterns were associated with disease symptom development, a split leaf inoculation experi ment was conducted and the effects of three separate treatments examined over time. Split-le aves were mock infiltrated, or inoculated with wild type strain B6 9, or the non-pathogenic mutant strain BIM2, which lacks pthB At 0, 2, and 7 dpi, half-leaves were harveste d and soluble proteins examined by western blot analysis. SUMO profiles of leaves inoculated with B69 were compared to those of leaves inoculated with mutant BIM2 at two dpi. There were no cha nges in the abundance of free CitSUMO or SUMOylated proteins in BIM2 i noculated leaves (Figur e 3-3, lane 4 and 5). The expected changes were seen in leaves inoculated with B69, i.e an increase in the amount of free SUMO and SUMO-conjugated pr oteins (Figure 3-3, lane 7 and 8). SUMO profiles at 7 days post inoculation revealed that the majority of the high molecular weight conjugates seen at 2 dpi in canker infected leaves were lost (Figure 3-3, lane 8 and 9). Interestingly, this loss of high molecular weight conjugates was also observed in leaves inoculated with non-pathogenic mutant strain BIM2. Whether the identities of SUMOylated prot eins in canker infected leaves are similar to the ones in BIM2 infected leaves is unknown; howev er, in both cases, SUMO de-conjugation occurred 7 dpi. These findings suggest that the SUMO de-conjugation observed at 7 dpi, in both BIM2and B69-inoculated leaves is PthB-independent and is also independent of the development of the macromolecular di sease symptom of canker (Figure 3-4). Conversely, the increase in the amounts of free SUMO and SUMO-conjugated proteins seen at 2 dpi with B69 are PthB-dependent.

PAGE 86

74 SUMO De-Conjugation Observed at 7 days Following Infection with B69 and BIM2 is Dependent on a Functional Type III Secretion System To determine if the SUMO de-conjugation observed at day 7 post inoculation in both B69and BIM2-inoculated leaves is dependent on a functional type III secretion system, a hrpG integrative mutant, B23.5, was generated. B23.5 was no longer pathogenic on citrus, and the hrpGphenotype was complemented after transformation of B23.5 with pUFR057::Xcv hrpG (Figure 3-5). There was no SUMO de-conjugation at da y 7 following inoculation with B23.5 (Figure 3-6), indicating that SUMO de-conjugation relies on a functional TTSS. In addition, B23.5 inoculation stimulated accu mulation of a 45kDa SUMO conjugate. A SUMOylated product of similar size was obser ved in leaves inoculated with B69 and BIM2, but did not accumulate (Figure 3-3). Discussion A great deal of effort has been directed towards investigating the mechanisms by which plants mount defense responses toward s pathogenic bacteria. Most studied cases involve incompatible plant microbe interactio ns that lead to the classical hypersensitive response or HR (Malek et al., 2000; Kazan et al, 2001). However, far less effort has been invested in trying to elucidate the mechanis ms by which a specific pathogen, or a group of pathogens elicit a particular disease with specific sets of morphological and molecular symptoms. In an effort to understand the processes by which different pathovars of Xanthomonas citri trigger canker symptoms, a canker responsive gene with sequence similarity to the SUMO gene family was identified by differential display PCR. The SUMO conjugation pathway in canker di sease was investigated using a splitleaf inoculation experiment to normalize for leaf-to-leaf vari ations. It was found that at 2

PAGE 87

75 dpi, X. citri pv. aurantifolii infection induces an increase in free CitSUMO and an increase in the number of high molecular we ight SUMOylated proteins. These changes were not observed in mock-inoculated leav es. SUMO conjugation in plants and other systems has been shown to be up-regulated by various instances of biotic and abiotic stresses (Kurepa et. al., 2003, Lois et al., 2003 and OÂ’Donnell et. al., 2003). In order to test if changes in SUMO conjugation observed were specific to X. citri pv. aurantifolii infection, two mu tant strains unable to cause ca nker on citrus were used in this study, BIM2 (interrupted in pathogenicity gene pthB ) and B23.5 (interrupted in the TTSS regulatory gene, hrpG ). Disruption of hrpG was previously shown to disable the type III secretion system in Xanthomonas (Wengelnik et al. 1996). Using split leaf inoculations, it was shown that in BIM2 i noculated leaves, at 2 dpi, there were no changes in the amount of free SUMO and SUMO ylated high molecula r weight proteins. Thus, the increase in free SUMO and in the number of SUMOylated proteins is likely to be a PthB-specific plant respons e rather than a general stre ss response. A large number of SUMO targets identified in ot her organisms are cell-cycle re lated (Melchior, 2000). It has been shown in yeast ( Saccharomyces cerevisiae ) that temperature-sensitive mutants corresponding to SUMO and the enzymes invol ved in its conjugation pathway arrest the cell cycle at the G2/M transition, therefore, showing a critical role for SUMO in cell cycle progression (Johnson and Gupta, 2001). It is possible that the observed upregulation of SUMOylated proteins and free SUMO reflects activation of the plant cell cycle by X. c. pv. aurantifolii in the early stages of infection. Remarkably, this increase in free SUMO and in the amount of high molecular weight SUMOyl ated proteins is lost 7 dpi, potentially indicating a transition to a second disease phase. The deconjugation

PAGE 88

76 phenotype observed at 7 dpi with B69 is al so observable at 7 dpi with BIM2, and therefore, the triggering f actor of de-conjugation is pr obably independent of PthB. The possibility that another effector c ould be the trigger of the de-conjugation observed at 7 dpi came from the finding that the TTSS mutant B23.5, did not induce deconjugation. Therefore it is po ssible that another type thre e effector, beside PthB is responsible for the de-conjugation observed at day 7. Alternatively, it is possible that the second PthA homologue, PthB0 (not required for canker, Ch apter 2), found in B69 and BIM2 is also able to trigger the de-conjugation observed 7dpi. It has been proposed that the abundance of SUMO proteases in X. campestris pv. vesicatoria could reflect an importa nt role of theses effectors in Xcv pathogenesis (Hotson et al. 2003 and Roden et al 2004) However, none of the iden tified proteases have been implicated in disease and are, in fact, disp ensable. Given the cri tical role of SUMO conjugation in cell cycle proce sses (Melchior, 2000), and th e lack of apparent SUMO proteases encoded by another canker causing strain X. citri pv. citri it is possible that the late de-conjugation phenotype is not directly triggere d by a type III effector of a protease nature, but rather that a type III effecto r(s) acts to induce endogenous citrus SUMO protease(s) leading to the de -conjugation observed in late stages of canker infection. Both hypotheses are not mutually exclus ive and characteriza tion of additional X. citri effectors as well as citrus pr oteins SUMOylated in respon se to canker are required to better characterize the involvement of SUMOyl ation in the infection process of canker causing xanthomonads.

PAGE 89

77 Figure 3-1: Alignment of grapefruit SUMO (partial sequence) with (PopSUMO1, gi:23997054, and AtSUMO1, At4g26840).

PAGE 90

78 B 6 9r S U M O Molecular Mass (kDa)20.4 29.6 37.4 54.6 7.0 98.0 206.7 115.8 A B M o c k B 6 9r S U M O Molecular Mass (kDa)20.4 29.6 37.4 54.6 7.0 7.0 98.0 206.7 115.8 A B M o c k Figure 3-2: SUMO profiles of B69a nd mock-challenged gr apefruit leaves. 10 g of crude protein from day 2 of the spl it leaf experiment was separated by electrophoresis, blotted to PVDF and (A) probed with purified Pop SUMO1 antisera. Lane 1, Mock treated leaf; la ne 2, B69 inoculated leaf; lane 3, 2 ng purified recombinant Pop SUMO1. ([]): high molecula r weight SUMOylated proteins. (): un-conjugated SUMO. (B) The membrane was stained with Coomassie R250 as a loading control (S hown is the small subunit of Rubisco).

PAGE 91

79 Molecular Mass (kDa)20.4 29.6 37.4 54.6 98.0 206.7 115.8 027027027 MockBIM2B69 DPI Treatment AMolecular Mass (kDa)20.4 29.6 37.4 54.6 98.0 206.7 115.8 027027027 MockBIM2B69 DPI Treatment A Figure 3-3: SUMO de-conjugation occurs 7 days after infecti on. Leaves were inoculated with Mock, BIM2, and B69 strains. 7.5 g of crude protein from 0, 2, and 7 dpi from each treatment of the spli t leaf experiment was separated by electrophoresis, blotted to PVDF and (Upper panel) probed with purified PopSUMO1 antisera. ([]): high molecu lar weight SUMOylated proteins. (): un-conjugated SUMO. (Lower panel) The membrane was stained with Coomassie R250 as a loading control (Show n is the small subunit of Rubisco).

PAGE 92

80 B B A A C C B B A A C C Figure 3-4: Split leaf inoculation of Xanthomonas citri pv. aurantifolii (B69) and derivative BIM2 mutant. Duncan grapef ruit leaf 7 dpi with B69 (shown on the left side of the mid-vein and BIM2 (shown on the right side of the mid-vein). (A) adaxial side and (B) abaxial side of the leaf. Note the whitish canker characteristic of the Xca strain and yellowing associated with the day 7 post inoculation canker phenotype. (C ) Advanced B69 canker phenotype.

PAGE 93

81 B69 B23.5c B23.5c1 B69 B23.5c B23.5c1AB Hi ndIII B69 B23.5 B23.5c Hi ndIII B69 B23.5 B23.5c B69 B23.5c B23.5c1 B69 B23.5c B23.5c1AB Hi ndIII B69 B23.5 B23.5c Hi ndIII B69 B23.5 B23.5c Figure 3-5: B69 mutant derivative B23.5 lacks a functional Type III s ecretion system. (A) Southern blot hybridization profile s contrast B69, B23.5 and B23.5c (B23.5/ hrpG ). DNA was digested with Hin dIII and probed with the same internal fragment of hrpG used as homology region for marker interruption. (B) B69 and B23.5c inoculation on Duncan grapefruit. hrpG complemented the hrpphenotype of B23.5

PAGE 94

82 20.4 29.6 37.4 54.6 7.0 98.0 206.7 115.8 Molecular Mass (kDa)027027 B23.5B69* 20.4 29.6 37.4 54.6 7.0 98.0 206.7 115.8 Molecular Mass (kDa)027027 B23.5B69 20.4 29.6 37.4 54.6 7.0 98.0 206.7 115.8 20.4 29.6 37.4 54.6 7.0 98.0 206.7 115.8 20.4 20.4 29.6 29.6 37.4 37.4 54.6 54.6 7.0 7.0 98.0 98.0 206.7 206.7 115.8 115.8 Molecular Mass (kDa)027027 B23.5B69027027 027027 B23.5B69* Figure 3-6: SUMO de-conjugati on at 7 dpi requires a functional TTSS. Leaves were inoculated with B23.5 and B69 strains. 7.5 g of crude protein from 0, 2, and 7 dpi from each split leaf treatment was separated by electrophoresis, blotted to PVDF and (A) probed with purif ied PopSUMO1 antisera. ([]): high molecular weight SUMOylated proteins. (): un-conjugated SUMO. (*) novel 70kDa protein unique to B23.5 7 dpi leaves. Equal amounts of protein was loaded in each lane.

PAGE 95

83 APPENDIX A LIST OF PLASMIDS AND STRAINS Table A-1: List of strains a nd plasmids used in this study. Strains or plasmids Relevant ch aracteristics Reference or source Escherichia coli DH5 F-, end A1, hsd R17(rk -mk -), sup E44, thi -1, rec A1, gyr A, rel A1, 80d lac ZM15, ( lac ZYAarg F)U169 Gibco BRL, Gaithesburg, MD HB101 sup E44, hsd S20(rk -mk -), rec A13, ara -14, pro A2, lac Y1, gal K2, rps L20, xyl -5, mtl -1, SmR Boyer and RoullandDussoix ED8767 sup E44, sup F58, hsd S3(rk -rk r), rec A56, gal K2, gal T22, met B1 Murray et al. 1977 Xanthomonas 3213T X. citri pv. citri A Gabriel et al, 1989 3213Sp X. citri pv. citri A, SpR derivative of 3213 Swarup et al., 1991 B21.1 pthA ::Tn5-gusA, marker exchanged mutant of 3213Sp, SpRKnR Swarup et al., 1991 B69 X. axonopodis pv. aurantifolii 69, ATCC, B form citrus canker type strain B69Sp Spntaneous SpR derivative of 69, SpR Unpublished BIM2 pthB ::CmR, marker integration mutant of B69Sp, SpRCmR Unpublished BIM6 Marker integration mutant of B69Sp, CmR integrated upstream of pthB SpRCmR Unpublished B13.2 VirB4 ::CmR, marker integration mutant of B69Sp, SpRCmR This study

PAGE 96

84 Table A-1. Continued Strains or plasmids Relevant ch aracteristics Reference or source B13.1 VirB40::CmR, marker integration mutant of B69Sp, SpRCmR This study B69.4 Unpublished pRK2013 ColE1, KmR,Tra+, helper plasmid Figurski and Helinski, 1979 pUFR004 ColE1, Mob+, Cmr, lacZ+ De Feyter et al, 1990 pUFR012 Derivative of pUFR004 with Kn resistence. ColE1, Mob+, KnRCmR, lacZ+ Unpublished pBY13 270 bp fragment of virB4 cloned in pUFR004, CmR This study pB13.1 virB4 ::pBY13 of pXcB0, CmR This study pB13.2, pB13.4, pB13.5 virB4 ::pBY13 of pXcB, CmR This study PXcB Natural plasmid of B69 carrying pthB Unpublished pXcB0 Natural plasmid of B69 carrying pthB0 Unpublished pBIM2 pthB ::Cm R (pYY40.10) of pXcB, CmR Unpublished pBIM6 pXcB::CmR(pQY92.1), pthB still functional, CmR Unpublished pBY23 550 bp fragment of hrpG cloned in pUFR012, KnR, CmR This study pBY23c HrpG from pXG8 (REF) cloned in pUFR53 This study B23.5 hrpG ::pBY23 of B69, KnR CmR This study B23.5c B23.5/pBY23c This study

PAGE 97

85 APPENDIX B NORTHERN BLOT ANALYSIS OF CCRS Figure B-1: Northern blot an alysis of CCR genes not found differentially regulated by reverse northern blot PR2 PR1 Mock BIM2 B69 Mock BIM2 B69 2 dpi7 dpi 18S PR2 PR1 Mock BIM2 B69 Mock BIM2 B69 2 dpi7 dpi Mock BIM2 B69 Mock BIM2 B69 2 dpi7 dpi 18S rRNA PDF1 Mock BIM2 B69 BIM2 B69 2 dpi7 dpi rRNA PDF1 Mock BIM2 B69 BIM2 B69 2 dpi7 dpi PDF1 Mock BIM2 B69 BIM2 B69 2 dpi7 dpi PR5 B69 BIM2 2 dpi PR5 B69 BIM2 2 dpiRD22 GST CHI rRNA rRNA rRNA B69 BIM2 2 dpi Exp rRNA B69 BIM2 2 dpi Exp rRNA B69 BIM2 2 dpi Exp rRNA B69 BIM2 2 dpi CaCO3 BIM2 B69 BIM2 B69 2dpi7dpi TIP Exp CaCO3 BIM2 B69 BIM2 B69 2dpi7dpi TIP Exp CaCO3 BIM2 B69 BIM2 B69 2dpi7dpi TIP Exp Frap/tor rRNA B69 BIM2 2 dpi Frap/tor rRNA B69 BIM2 2 dpi pip3 B69 BIM218 S2 dpiPR2 PR1 Mock BIM2 B69 Mock BIM2 B69 2 dpi7 dpi 18S PR2 PR1 Mock BIM2 B69 Mock BIM2 B69 2 dpi7 dpi Mock BIM2 B69 Mock BIM2 B69 2 dpi7 dpi 18S rRNA PDF1 Mock BIM2 B69 BIM2 B69 2 dpi7 dpi rRNA PDF1 Mock BIM2 B69 BIM2 B69 2 dpi7 dpi PDF1 Mock BIM2 B69 BIM2 B69 2 dpi7 dpi PR5 B69 BIM2 2 dpi PR5 B69 BIM2 2 dpiRD22 GST CHI rRNA rRNA rRNA B69 BIM2 2 dpi Exp rRNA B69 BIM2 2 dpi Exp rRNA B69 BIM2 2 dpi Exp rRNA B69 BIM2 2 dpi CaCO3 BIM2 B69 BIM2 B69 2dpi7dpi TIP Exp CaCO3 BIM2 B69 BIM2 B69 2dpi7dpi TIP Exp CaCO3 BIM2 B69 BIM2 B69 2dpi7dpi TIP Exp Frap/tor rRNA B69 BIM2 2 dpi Frap/tor rRNA B69 BIM2 2 dpi pip3 B69 BIM218 S2 dpi

PAGE 98

86 LIST OF REFERENCES Alfano, J.R. and Collmer, A. 1996. Bacterial path ogens in plants: life up against the wall. Plant Cell. 8: 1683-1698. Alfano, J.R. and Collmer, A. 1997. The type III (Hrp) secretion pathway of plant pathogenic bacteria: traffick ing harpins, Avr proteins and death. J. Bacteriol. 179:5655-5662. Anderson, D.M., Fouts, D.E., Collmer, A. and Schneewind, O. 1999. Reciprocal secretion of proteins by th e bacterial type III machines of plant and animal pathogens suggests recognition of mRNA targ eting signals. Proc. Nat. Acad. Sci. USA. 96:12839-12843. BergeyÂ’s Mannual of Determinative Bacteriology, 9th Eddition, JG Holt (ed.), Williams and Wilkins, Baltimore, MD, USA. Bonas, U., Stall, R.E. and Staskawicz, B. 1989. Genetic and structural characterization of the avirulence gene, avrBs3 from Xanthomonas campestris pv. vesicatoria. Mol. & Gen. Genet. 218:127-136. Boyer, H. W., and Roulland-Dussoix, D. 1969. A complementation analysis of the restriction and modi fication of DNA in Escherichia coli J. Mol. Biol. 41:459-465. Brunings A.M. and Gabriel D.W. 2003. Xant homonas citri: breaking the surface. Molec. Plant Pathol. 4(3):141-157. Burns, D.L., 1999. Biochemistry of type IV secretion. Curr. Opin. Microbiol. 1999. 2(1):25-29. Christie, P.J. 1997. Agrobacterium T-Complex tr ansport apparatus: a paradigm for a new family of multifunctional tran sporters in Eubacteria. J. Bacteriol. 179:3085-3094. Christie, P.J. 2001. Type IV secretion: in tracellular tran sfer of macromolecules by systems ancestrally related to conjuga tion machines. Mol. Microbiol. 40:294-305. Christie, P.J. and Vogel, J.P. 2000. Bacteria l type IV secretion: conjugation systems adapted to deliver effector molecules to host cells. Trends Microbiol. 8:354-360. Cornelis, G.R. and VanGijsegem, F. 2000. Asse mbly and function of type III secretory systems. Annu. Rev. Microbiol. 54:735-774.

PAGE 99

87 Cubero, J. and Graham, J.H. 2002. Genetic relationship among worldwide strains of Xanthomonas causing canker in citrus species and design of new primers for their identification by PCR. Appl. E nviron. Microbiol. 68:1257-1264. da Silva, A. C., J. A. Ferro, et al. (2 002). “Comparison of the genomes of two Xanthomonas pathogens with differing host sp ecificities.” Na ture 417(6887): 45963. De Feyter, R., Kado, C. I., and Gabriel, D. W. 1990. Small st able shuttle vectors for use in Xanthomonas Gene 88:65-72. De Feyter, R., and Gabriel, D. W. 1991. At least six avirul ence genes are clustered on a 90-kilobase plasmid in Xanthomonas campestris pv. malvacearum. Mol. PlantMicrobe Interact. 4:423-432. De Feyter, R., Yang, Y., and Ga briel, D. W. 1993. Gene-for-g enes interactions between cotton R genes and Xanthomonas campestris pv. malvacearum avr genes. Mol. Plant-Microbe Interact. 6:225-237. Duan, Y.P., Castaneda, A.L., Zaho, G., Erdos, G. and Gabriel, D.W. 1999. Expression of a single, host-specific, bacter ial pathogenicity gene in plant cells elicits division, enlargement and cell death. Mol. Plant-Microbe Interact. 12:556-560. Egel, D.S., Graham, J.H. and Stal l, R.E. 1991. Genomic relatedness of Xanthomonas campestris strains causing diseases of ci trus. Appl. Environ. Microbiol. 57:2724-2730. El Yacoubi, B., Brunings,A., Yuan, Q. and Gabriel, D.W. 2001. A self-transmissible plasmid isolated from Xanthomonas campestris carries a member of the avr/pth gene family and additional factor(s) requi red for pathogenicity. Abstract of the 10th International Congress of Mo lecular Plant-Microbe Inte ractions, Madison, WI, 1014 July 2000, #650. Falcow, S. 1996. The evolution of pathogenicity in Escherichia, Shigella and Salmonela ; in Cellular and Molecular biology (ed.) F.C. Neidhadz (Washington DC: American Society for Microbiology). 2723-2729. Figurski, D. H., and Helinski, D. R. 1979. Re plication of an origin-containing derivatives of plasmid RK2 dependent on a plasmid function provided in trans Proc. Natl. Acad. Sci. USA 76:1648-1652. Gabriel, D. W., Kingsley, M., Hunter, J. E ., and Gottwald, T. R. 1989. Reinstatement of Xanthomonas citri (ex Hasse) and X. phaseoli (ex Smith) and reclassification of all X. campestris pv. citri strains. Int. J. Syst. Bacteriol. 39:14-22.

PAGE 100

88 Gabriel, D. W., and De Feyter, R. 1992. RFLP analyses and gene tagging for bacterial identification and taxonomy. Pages 5166 in: Molecular Plant Pathology: A Practical Approach. Vol. 1. S. J. Gurr, M. J. McPherson, and D. J. Bowles, eds. IRL Press, Oxford. Gabriel, D.W. 1999. Why do plant pathogens carry avirulence genes? Physiol. Mol. Plant Pathol. 55: 205-214. Gottwald, T.R., Graham, J.H., Schubert, T.S. 2002. Citris canker: the pathogen and its inpact. Online. Plant health Progress. doi:10.1094/PHP-2002-0812-01RV.http://plant managementnetwork.org/pub/php/review/citruscanker/. Graham, J.H., Gottwald, T.R., Cubero, J., Achor, D.S. 2004. Xanthomonas axonopodis pv citri factors affecting successful eradic ation of citrus canker. Molec. Plant Pathol. 5:1-15. He, S.Y. 1998. Type III protein secretion system in plant and animal pathogenic bacteria. Annu. Rev. Phytopathol. 36: 363-392. Hildebrand, D.C., Palleroni, N.J. and Schroth, M.N. 1990. Deoxyribonucleic acid relatedness of 24 xanthomona d strains representing 23 Xanthomonas campestris pathovars and Xanthomonas fragariae J. Appl. Bacteriol. 68: 263-269. Jin, Q. and S.Y. He (2001). “Role of the Hr p pilus in type III protein secretion in Pseudomonas syringae .” Science 294(5551): 2556-2558. Jones, J.B., Bouzar, H., Stall, R.E., Almira E.C., Roberts, P.D., Bowen, B.W., Subderry, J., Strickler, P.M., and Chun, J. 2000. Sy stematic analysis of Xanthomonads (Xanthomonas spp.) associated with pepper and tomato lesi ons. Int. J. Syst. Evol. Microbiol. 50:1211-1219. Keen N.T. 1990. Gene for gene complementr ity in plant-pathogen interactions. Annu. Rev. Genet. 24:447-63. Kingsley, M.T., Gabriel, D.W., Marlow G.C. and Roberts, P.D. 1993. The opsX locus of Xanthomonas campestris affects host range and bios ynthesis of lipopolysaccharide and extracellular polysaccharide. J. Bacteriol. 175:5839-5850. Kubori, T. Matsushima, Y., Nakamura, D., Uralil, J., Lara-Tajero, M., Sukhan, A., Galan, J.E., and Aizawa, S. 1998. Supramolecular structure of the Salmonella typhimurium type III pretein secretion system. Science. 280:602-605. Lawrance, J.G. and Roth, J.R. 1996. Selfish operons: horizontal transfer may drive the evolution of gene cluste rs. Genetics. 143: 1843-9417. Lazo, G. R., and Gabriel, D. W. 1987. Conservation of plasmid DNA sequences and pathovar identification of strains of Xanthomonas campestris Phytopathology 77: 448-453.

PAGE 101

89 Lazo, G. R., Roffey, R., and Ga briel, D. W. 1987. Pathovars of Xanthomonas campestris are distinguishable by restri ction fragment length polymorphisms. Int. J. Syst. Bacteriol. 37:214-221. Leach, J. E. and White F. F. 1996. Bacteria l avirulence genes. Annu. Rev. Phytopathol. 34:153-179. Leong, S. A., Ditta, G. S., and Helinski, D. R. 1982. Heme biosynthesis in Rhizobium: Identification of a cloned gene coding fo r aminolevulinic acid synthetase from Rhizobium meliloti J. Bio. Chem. 257:8724-8730. Lorian, V. (ed.) 1986. Antibiotics in Laborat ory Medicine, Second edition, Williams & Wilkins, Baltimore. 683-721. Marenda, M., Brito, B., Callard, D., Genin, S., Barberis, P., Boucher, C. and Arlat, M. 1998. Prha controls a novel regulatory pathway require d for the specific induction of Ralstonia solanacearum hrp genes in the presence of palnt cells. Mol. Microbiol. 27:437-453. Marois, E., Van den Ackerveken, G. and Bonas, U. (2002) The Xanthomonas type III effector protein AvrBs3 modulates pl ant gene expression and induces cell hypertrophy in the susceptible host. Mo l. Plant-Microbe In teract. 15(7), 637-46. Murray, N. E., Brammar, W. J., and Murra y, K. 1977. Lambdoid phages that simplify the recovery of in vitro recombinants. Mol. Gen. Genet. 150:53-61. Sambrook, J., Fritsch, E. F., and Maniatis, T. A. 1989. Molecular Cl oning: A Laboratory Manual. 2nd ed. Cold Spring Habor Laboratory, Cold Spring Habor, NY. Silhavy, T.J. 1997. Death by leatha l injection. Scie nce. 278:1085-1086. 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 Civerolo, E.L. 1991. Research relating to the recent outbreak of citrus canker. Annu. Rev. Phytopathol. 29:399-420. Swarup, S., De Feyter, R., Brlansky, R. H., and Gabriel, D. W. 1991. A pathogenicity locus from Xanthomonas citri enables strains from several pathovars of X. campestris to elicit canker-like lesion s on citrus. Phytopathology 81:802-809. Swarup,S., Yang, Y., Kingsley, M. K., and Gabriel, D. W. 1992. An Xanthomonas citri pathogenicity gene, pthA, pleiotropically encodes gratuitous avirulence on nonhost. Mol. Plant-Microbe Interact. 5:204-213. Syvanen, M. and Kado, C.I. 1998. Horizontal gene transfer. London: Chapman and Hall.

PAGE 102

90 Vernoud, V., Horton, A.C., Yang, Z. and Nielse n, E. (2003) Analysis of the Small GTPase Gene Superfamily of Arabi dopsis. Plant Physio l. 131(3), 1191-1208. Wayne, L.G., Brenner, D.J., Colwell, R.R ., Grimont, P.A.D., Kandler, O., Krichevsky, M.I., Moore, L.H., Moore, W.E.C., Murra y, R.G.E., Stackebrandt, E., Starr, M.P. and Truper, H.G. 1987. Report of the Ad Hoc Committee on reconciliation of approaches to bacterial systematics. Int. J. Winans, S.C., Burns, D.L. and Christie, P. J. 1996. Adaptation of a conjugal transfer system for the export of pathogenic macr omolecules. Trends Microbiol. 4:64-68. Yang, Y., Yuan, Q. and Gabriel, D.W. 1996. Watersoaking function( s) of XcmH1005 are redundantly encoded by members of the Xanthomonas avr/pth gene family. Mol. Plant-Microbe Interact. 9:105-113.

PAGE 103

91 BIOGRAPHICAL SKETCH Basma El Yacoubi was born on Octobe r 12 1973, in Mekns, Morocco. She obtained her D.E.U.G and Licen se in cell biology and physio logy from University Joseph Fourier in Grenoble, France; and her Maitri se cell biology and physiology from Paris 7 University in Paris, France. In the summer of 1996, she joined the University of Florida, and attended the English Language Institute during fall 1996. In spring 1997, she began her graduate studies, and obtai ned a Master of Science degr ee from the Department of Environmental Horticulture. In fall 1999, sh e joined the Plant Molecular and Cellular Biology program, working on her Ph.D. in the department of Plant Pathology.