Citation
In Search of Pathogenicity Factors of Xanthomonas citri pv. aurantifolii

Material Information

Title:
In Search of Pathogenicity Factors of Xanthomonas citri pv. aurantifolii
Creator:
BRUNINGS, ASHA MARCELLE ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Bacteria ( jstor )
DNA ( jstor )
Genes ( jstor )
Open reading frames ( jstor )
Plasmids ( jstor )
Proteins ( jstor )
Secretion ( jstor )
Signals ( jstor )
Xanthomonas ( jstor )
Xanthomonas campestris ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Asha Marcelle Brunings. 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
Resource Identifier:
436098672 ( OCLC )

Downloads

This item has the following downloads:

brunings_a ( .pdf )

brunings_a_Page_052.txt

brunings_a_Page_093.txt

brunings_a_Page_089.txt

brunings_a_Page_112.txt

brunings_a_Page_110.txt

brunings_a_Page_060.txt

brunings_a_pdf.txt

brunings_a_Page_118.txt

brunings_a_Page_072.txt

brunings_a_Page_037.txt

brunings_a_Page_066.txt

brunings_a_Page_115.txt

brunings_a_Page_049.txt

brunings_a_Page_020.txt

brunings_a_Page_091.txt

brunings_a_Page_013.txt

brunings_a_Page_117.txt

brunings_a_Page_002.txt

brunings_a_Page_056.txt

brunings_a_Page_009.txt

brunings_a_Page_006.txt

brunings_a_Page_116.txt

brunings_a_Page_036.txt

brunings_a_Page_029.txt

brunings_a_Page_113.txt

brunings_a_Page_073.txt

brunings_a_Page_031.txt

brunings_a_Page_059.txt

brunings_a_Page_035.txt

brunings_a_Page_004.txt

brunings_a_Page_012.txt

brunings_a_Page_083.txt

brunings_a_Page_099.txt

brunings_a_Page_087.txt

brunings_a_Page_008.txt

brunings_a_Page_054.txt

brunings_a_Page_022.txt

brunings_a_Page_061.txt

brunings_a_Page_055.txt

brunings_a_Page_001.txt

brunings_a_Page_071.txt

brunings_a_Page_106.txt

brunings_a_Page_030.txt

brunings_a_Page_108.txt

brunings_a_Page_076.txt

brunings_a_Page_107.txt

brunings_a_Page_033.txt

brunings_a_Page_075.txt

brunings_a_Page_005.txt

brunings_a_Page_050.txt

brunings_a_Page_081.txt

brunings_a_Page_094.txt

brunings_a_Page_090.txt

brunings_a_Page_058.txt

brunings_a_Page_053.txt

brunings_a_Page_111.txt

brunings_a_Page_095.txt

brunings_a_Page_092.txt

brunings_a_Page_026.txt

brunings_a_Page_057.txt

brunings_a_Page_114.txt

brunings_a_Page_102.txt

brunings_a_Page_070.txt

brunings_a_Page_046.txt

brunings_a_Page_039.txt

brunings_a_Page_043.txt

brunings_a_Page_088.txt

brunings_a_Page_100.txt

brunings_a_Page_028.txt

brunings_a_Page_007.txt

brunings_a_Page_048.txt

brunings_a_Page_065.txt

brunings_a_Page_014.txt

brunings_a_Page_086.txt

brunings_a_Page_080.txt

brunings_a_Page_010.txt

brunings_a_Page_041.txt

brunings_a_Page_062.txt

brunings_a_Page_019.txt

brunings_a_Page_079.txt

brunings_a_Page_063.txt

brunings_a_Page_069.txt

brunings_a_Page_098.txt

brunings_a_Page_084.txt

brunings_a_Page_016.txt

brunings_a_Page_109.txt

brunings_a_Page_120.txt

brunings_a_Page_038.txt

brunings_a_Page_101.txt

brunings_a_Page_011.txt

brunings_a_Page_024.txt

brunings_a_Page_021.txt

brunings_a_Page_018.txt

brunings_a_Page_067.txt

brunings_a_Page_045.txt

brunings_a_Page_040.txt

brunings_a_Page_097.txt

brunings_a_Page_044.txt

brunings_a_Page_003.txt

brunings_a_Page_034.txt

brunings_a_Page_119.txt

brunings_a_Page_027.txt

brunings_a_Page_025.txt

brunings_a_Page_042.txt

brunings_a_Page_078.txt

brunings_a_Page_105.txt

brunings_a_Page_051.txt

brunings_a_Page_032.txt

brunings_a_Page_015.txt

brunings_a_Page_064.txt

brunings_a_Page_068.txt

brunings_a_Page_077.txt

brunings_a_Page_023.txt

brunings_a_Page_096.txt

brunings_a_Page_104.txt

brunings_a_Page_085.txt

brunings_a_Page_103.txt

brunings_a_Page_082.txt

brunings_a_Page_074.txt

brunings_a_Page_047.txt

brunings_a_Page_017.txt


Full Text












IN SEARCH OF PATHOGENICITY FACTORS OF Xanthomonas citri pv. aurantifolii


By

ASHA MARCELLE BRUNINGS













A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

ASHA MARCELLE BRUNINGS















ACKNOWLEDGMENTS

Many debts have been incurred during the project described on the following

pages. The Organization of American States (OAS) provided funding during the first 2

academic years. I especially thank my supervisory committee chair Dean Gabriel, and all

of my current and former committee members Alice Harmon, Bill Gurley, Ken Cline,

John Davis, James Preston III, and Tom Mareci have all been very helpful.

I owe my colleagues in the lab Adriana Castafieda, Abdulwahid Al-Saadi, and

Basma El Yacoubi for their support and lively discussions. Many thanks go to Yong Ping

Duan, Anjaiah Vanamala, and Joseph Reddy, postdoctoral associates in the lab.

I owe more than I can say on these pages to Gary Marlow. He has gone above and

beyond what could reasonably be expected from a lab technician. Gary has been a

constant source of support over many years, both inside and outside the lab.

Pant Pathology Department chair Gail Wisler, has been an enormous source of

encouragement over the years; as were Lauretta Rahmes, Gail Harris, and Donna Perry of

the Pathology office. I thank the Plant Molecular and Cellular Biology Program, and

support staff, especially Melissa Webb, who has been very supportive.

My life has been enriched by the friendships of many people I met while pursuing

my degree. In this regard, a special note of thanks goes out to Jorge Vazquez, Erica

Gubrium, Doug Knudsen, Joyce Merritt, Shayna Sutherland, Karen Chamusco, Melanie

Cash, and Mukesh Jain.









Immeasurable is the support of my family. My parents Ernie and Mireille Brunings,

and the memory of my late brother, Raoul, have been an inspiration. My sons, Raoul and

Marlow; and my husband, James Meier, have been especially patient with me, both while

the research was ongoing and during the writing of this thesis.
















TABLE OF CONTENTS

page

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

LIST OF TABLES ................................................. ................... viii

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

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

CHAPTER

1 LITERATURE REVIEW ............................................................ ................ 1

Citrus Canker ..................................... .................................. ................
D disease Sym ptom s ............................................... ........... .............. .
H o st R a n g e ....................................................... ................ .. 2
Control Measures.............. .. ............... ..................
The Pathogen .................. ................... .................. .. .... ........... 3
Pathogenicity Factors..................................5.........5
Gene pthA and its Predicted Product .......................... ...............................5
Q uorum Sensing ....................................................... 8
N u tritio n ................................................................................................................ 8
Attachm ent .................................................9
Type II (General) Secretion (T2S) ............ ............................... ................11
Type III Secretion (T3 S) .............. .............................. ..................................... 12
T ype IV Secretion (T 4 S) ........................................................................ ... ... 15
Plasm ids pX A C33 and pX A C64 ....................................................................... 18
Plasm id Toxin-Antitoxin M odules .................................. .............................. ....... 19
TA M mechanism s .................. .................................... .............. ... 20
TA O peron O organization .......................................................... ............... 21
T testing Plasm id Stability .......................................................... ............... 22

2 A NOVEL PATHOGENICITY FACTOR REQUIRED FOR CITRUS CANKER
DISEASE IS CONTAINED ON A SELF-TRANSMISSIBLE PLASMID ...............23

M material and M methods ................... .......... ................................... ...................... 24
Bacterial Strains, Plasmids, and Recombinant DNA Techniques......................24
Plant Inoculations ...................... .... .............. ................... ..... .... 27
R e su lts ...........................................................................................2 9


v









B 69 H as M multiple Plasm ids and a T3SS.................................... .....................29
Gene pthB is Insufficient to Cause Canker..................................... ..................29
M apping of pBIM 2 and pBIM 6 ................................. ............... ....................29
Constructs ............ 3............ ........ ..........30
P athogenicity T ests............ .................................................................... ... ... 1
Self-transm issibility .................................................. ...... .. ...... .... 33
D discussion ..................................... .................. ............... ........... 33

3 COMPLETE PLASMID SEQUENCE .............................................. ...............35

M material and M methods ............... .............................................................. .......... 35
Bacterial Strains, Plasmids, and Culture Media...............................................35
General Bacteriological Techniques................................................................. 35
Recombinant DN A Techniques...................................... ......................... 36
D N A Sequence A analysis ............................................. ............................ 36
R e su lts .................. ................... ....................................................... 3 7
Features of Plasm id pX cB ........................................................ ............. 37
P ro m o te rs ....................................................................................................... 3 9
O pen R leading F ram es .............................................................. .....................47
T he pX cB virB C lu ster ................................................................................ ... 49
Similarity to Plasmids From X citri pv. citri and P. putida.............................58
D iscu ssion ............... .................................... ............................6 1

4 MARKER INTEGRATION MUTAGENESIS..................................64

M materials and M methods ....................................................................... .....................66
Bacterial Strains, Plasmids, and Culture Media............................................66
General Bacteriological Techniques................. ..............................................68
Recom binant DN A Techniques....................................... ......... ............... 68
Marker Interruption Mutagenesis .................................................................68
Phenotypic Tests .................. .............................. ...... ................. 69
Plant Inoculations ........................ ................ ................... ..... .... 73
P lasm id Stability in p lanta ....................................................... .....................73
R e su lts .................. ............ ....... ............................................................................ 7 3
Southern Blot Hybridization with orfs115-117 ............................................73
Marker Integrations and Mapping ..... .................... ...............75
P ath og enicity T ests............ ........................................................... .... .... .... .. 7 5
Plasm id Stability Tests in vitro........................................ ........................ 80
Stability T ests in p lanta ............................................................ ............... 82
D isc u ssio n ............................................................................................................. 8 3

5 C O N C L U SIO N S ..................... .... ............................ ........... ...... ... ...... 85


APPENDIX

A MAP OF POSITION 13301 TO 15700................................ ............. ........... 87









B CODONPREFERENCE FOR POSITION 13301 TO 16000....................92

L IST O F R E F E R E N C E S ......................................... ........... ................ .......................... 94

B IO G R A PH ICA L SK ETCH .................................... ............ ......................................108
















LIST OF TABLES


Table p

1-1 Relative pathogenicity of all known X citri strain groups on four citrus species......4

2-1 List of bacterial strains used in this study ..................................... ............... ..24

2-2 List of plasmids used in this study ............ ................................... ................26

3-1 Open reading frames identified in plasmid pXcB ..........................................40

3-2 Predicted promoters for the + strand of plasmid pXcB with a score of 0.85 or
h ig h er ............... ......................................... ..............................4 7

3-3 Predicted promoters for the strand of plasmid pXcB with a score of 0.85 or
h ig h er ............... ......................................... ..............................4 8

3-4 Transcriptional terminators identified in plasmid pXcB by GeSTer .....................49

3-5 Presence of N-terminal signal sequences, localization of putative proteins
(PSORT), and transmembrane helices (TmPred) ................................. .............. 50

3-6 Comparison of the pXcB virB cluster with several type IV secretion-systems. ......51

4-1 List of bacterial strains used in this study, in addition to those listed in Chapter 2.66

4-2 List of plasmids used in this study, in addition to those listed in Chapter 2............67

4-3 Summary of primers and resulting plasmids and Xanthomonas strains..................69

4-4 Stability tests for marker-integrated plasmids in vitro, grown in liquid culture with
and without antibiotic selection for the marker-integrated plasmid.......................81

4-5 Stability tests for mutant plasmids in vitro, alone, and together with vector
pUFR047 or complementation clone plasmidd pAB36.6). .....................................81

4-6 Stability tests for marker-integrated plasmids inplanta .......................................82
















LIST OF FIGURES


Figure page

1-1 Maps of plasmids from Xanthomonas citri pv. citri Type A strain 306.....................7

1-2 Pathway of the T-DNA strand through the Agrobacterium tumefaciens T4SS.......17

2-1 Members of the avr/pthA gene family in X citri pv. aurantifolii.............................28

2-2 Restriction map of plasmid pBIM2 and its subclones ................ ......... ..........30

2-3 Restriction map of plasmid pBIM6 ......................................................30

2-4 Restriction digests of plasmid pBIM2 for mapping purposes................................31

2-5 Plasm id pBIM 2 and subclones. ........................................ ........................... 32

3-1 M ap of plasm id pX cB ................................. ................. ....................................... 38

3-2 Codonpreference of plasmid pXcB reverse frames from position 20,000 to 21,000.46

3-3 G+C content of plasmid pXcB. ........ ....................................................................46

3-4 Potential promoters on plasmid pXcB .......... ................................................48

3-5 Comparison of the linear organization of several type IV secretion-systems ..........51

3-6 Phylogenetic trees compiled by CLUSTALW ......................................................52

3-7 Results of a BLASTX search limited to X axonopodis pv. citri............................59

3-8 Results from BLASTN and BLASTX searches of nucleotides 7000 to 8500 of
plasmid pXcB ...................................... ................................ .........59

3-9 Results of a BLASTX search with plasmid pXcB as query and the results limited to
genes from P. putida ......................... ....... .... .. ...... ............ 60

4-1 Diagram of marker integration mutagenesis. ............. .............. ................65

4-2 Primer AB47. .............................. ... ..... .. ................... 70

4-3 G arden B lot. ....................................................... ................. 7 1









4-4 Total DNA digested with EcoRI, Southern blot hybridized consecutively with orf
115, 116, and orfl 7 ..................... .. ........................... .. ...... .... ........... 72

4-5 Orientations for the integration of the suicide vector in pXcB. ............................74

4-6 Mapping of the lacZ promoter orientation in pB16.1, pB23.1-11, pB26.4.0, and
p B 3 1 .2 1 .......................................................................... 7 4

4-7 Construction of pA B 29 ......................... ......... ............. .... ......................76

4-8 Inoculation result of plasmid pAB29 mated into B69.4 and inoculated on Duncan
G rap efru it ......................................................................... 7 6

4-9 The lacZ promoter is functional in the B-strain. ................... ................... .......... 77

4-10 Results of initial inoculation of orfl 17 marker integration .............. .....................78

4-11 Clones used for attempts to complement maker-integrated mutants and the regions
of plasm id pX cB they represent ...................................................... ........ ...... 79

4-12 Attempts to complement marker-integrated mutants with plasmid pAB36.6..........79

4-13 Attempts to complement marker-integrated mutants with plasmid pJR8.2............80















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

IN SEARCH OF PATHOGENICITY FACTORS OF Xanthomonas citri pv. aurantifolii

By

Asha Brunings

May 2005

Chair: Dean W. Gabriel
Major Department: Plant Molecular and Cellular Biology

Xanthomonas citri pv. aurantifolii strain B69 is a causal agent of citrus canker

disease found in South America. GenepthB is located on a self-mobilizing plasmid,

pXcB, found in B69, and is necessary but insufficient for pathogenicity. The B69

derivatives that are cured of pXcB are non pathogenic. Subclones of pXcB were

introduced into a pXcB-cured derivative of B69, B69.4, together withpthB or an

isofunctional homologue, pthA, in an attempt to complement the mutant phenotype. None

of the subclones used in our study restored pathogenicity to B69.4/pthB, or B69.4/pthA.

The complete sequence of pXcB was determined, and 38 open reading frames

(ORFs) were identified. None of the identified putative genes appeared to be an obvious

pathogenicity factor. Almost one-third of pXcB appears to encode a large polycistronic

transcriptional unit, which includes a complete type IV secretion system, consisting of 12

genes. In addition, the transcriptional unit carries one gene upstream, and five additional

genes immediately downstream of the type IV secretion-system genes.









Marker integration, resulting in the interruption of two ORFs downstream of the

type IV secretion system, was performed. Both marker-integrated strains lost

pathogenicity on both lime and grapefruit. However, several clones carrying wild-type

copies of the putative genes encoded by the ORFs failed to complement the mutants.

Since plasmid instability could explain the inability to complement the mutants, and one

of the interrupted ORFs resembled a plasmid stability gene, the stability of the mutant

plasmids was tested in vitro and inplanta. No instability of the mutant plasmids was

observed. Stability of the mutant plasmids together with one of the complementation

plasmids was also tested in vitro. The mutant plasmids and the complementation plasmid

were stable together in liquid culture without selection pressure.














CHAPTER 1
LITERATURE REVIEW

Citrus Canker

Citrus canker is a bacterial disease of citrus that causes premature leaf and fruit

drop (Gottwald et al., 2001, 2002; Graham et al., 2004; Stall & Civerolo, 1991). Since

there is no cure or effective treatment available, citrus is subject to strict federal and state

quarantine regulations in the United States of America. These regulations require canker

eradication, including destruction of both infected citrus (e.g., showing symptoms of

citrus canker disease) and "exposed" citrus (e.g., all citrus trees within a 1900 ft radius of

an infected tree). Destruction of exposed trees is thought to eliminate 95% of subsequent

infections resulting from dispersal of inoculum from an infected tree (Gottwald et al.,

2001).

Disease Symptoms

Citrus canker disease symptoms first appear as oily circular lesions, 2-10 mm in

size, usually on the abaxial surface of the leaf (Brunings & Gabriel, 2003, Gottwald et al.,

2001, 2002; Graham et al., 2004; Stall & Civerolo, 1991). The lesions become raised and

blister-like, and eventually grow into white or yellow spongy pustules on leaves, stems,

thorns and fruit. The pustules then darken and thicken into light tan to brown rough,

corky cankers. The individual lesions look like craters, with a necrotic sunken center and

raised sides. The lesions on stems can sometimes fuse together and result in splitting of

the epidermis along the stem. Older lesions on leaves and fruit can have more elevated

margins and can be surrounded by a yellow chlorotic halo (which may disappear in time).






2


The sunken craters are especially noticeable on fruits, but do not penetrate far into the

rind. Defoliation and premature abscission of affected fruit occurs on heavily infected

trees. Damage to leaves, stems, and fruit results in a reduction of fruit quality and yield.

The actual cost of the disease is greater than simply reduced fruit quality and quantity.

Regulatory actions (including destruction of infected and exposed trees) plus restrictions

on shipping of fruit from areas that are affected by citrus canker cause additional losses

(Schubert et al., 2001).

Host Range

Citrus canker affects Rutaceous plants, primarily Citrus spp., Fortunella spp., and

Poncirus spp. (Gabriel, 2002), depending on the citrus canker strain involved. The

disease is most severe on grapefruit, some sweet oranges, Mexican (Key) limes, and

trifoliate orange (Gottwald et al., 2002). However the actual host range depends on the

strain of citrus canker (see below). In general, lush, young tissues are more susceptible to

citrus canker than older ones (Stall et al., 1982). Therefore every flush in the citrus

culture cycle (about three times a year) provides a period of vulnerability (Schubert et al.,

2001).

Control Measures

Quarantine measures are used by several countries to prevent introduction of citrus

canker, but outbreaks still occur (Gottwald et al., 2002). Breeding of citrus trees is

virtually impossible. Therefore no known citrus varieties are resistant against citrus

canker in a gene-for-gene manner (Gottwald et al.,1993). However, there are important

differences in host range among various citrus canker strains (see below) that may

indicate active host defences. In addition, there is some field resistance against the

disease, generally correlated with the size and number of stomatal openings and a lack of









aggressive growth by the citrus species (Goto, 1969; McLean & Lee, 1922; Stall et al.,

1982). More aggressive rootstocks increase susceptibility of the scion.

Citrus canker was first declared eradicated in Florida in 1933 (Schoulties et al.,

1987). In 1986, the disease reappeared (Stall & Civerolo, 1991) and was again declared

eradicated in 1994 (Gottwald et al., 2002). Two more outbreaks occurred: one in Miami

in 1995, and another on the Florida west coast in 1997 (Gottwald et al., 2002).

Eradication efforts are ongoing at the time of this writing; and include strict quarantine of

the pathogen, and removing all infected and exposed trees. It was reported that as of

April 2004 more than 2.9 million trees have been destroyed in Florida as a result of

regulatory action (Florida Citrus Mutual, 2004).

The Pathogen

Citrus canker is caused by two phylogenetically distinct groups of Xanthomonas:

one originating in Asia, and the other in South America (Gabriel, 2001). Each group

comprises subgroups that are pathogenic variants. The most widespread group is the

Asiatic group, or X citri pv. citri A (syn. X campestris pv. citri; X axonopodis pv. citri),

causing type A canker. This group is prevalent in citrus-growing regions throughout Asia,

and has been successfully eradicated in Northern Australia and South Africa. The second

group is the South American group (X. citri pv. aurantifolii B), causing B type canker.

This group was first identified in Argentina, and spread to Uruguay and Brazil.

Xanthomonas citri pv. aurantifolii B is more restricted in host range than that of type A. It

mainly affects Mexican (Key) lime and mandarin; and to a lesser extent, grapefruit.

Xanthomonas citri pv. aurantifolii C, causing C type canker, is closely related to type B

and also originated in South America. Its host range, however, is limited to Mexican

(Key) lime. Apparent derivatives ofX. citri pv. citri A have emerged (called types A* and









A"). Type A* was identified in Southwest Asia by Verniere et al. (1998). Type Aw was

more recently identified in Florida (Sun et al., 2000). Both A* and A" types have a host

range limited to Mexican lime. All citrus canker strains cause identical disease symptoms

on susceptible citrus, but host range differences and in particular the hypersensitive

response (HR) may play a role in restricting host range. The pathogenicity of different

strains of citrus canker is summarized in Table 1-1.

Table 1-1. Relative pathogenicity of all known X citri strain groups on four citrus
species
Canker group C. sinensis C. paradisi C. limon C. aurantifolii
(Sweet Orange) (Grapefruit) (Lemon) (Mexican lime)
X citri pv. citri A +++ ++++ +++ ++++
X citri pv. citri A* ++++
X citri pv. citri Aw HR ++++
X citri pv. aurantifolii B + + +++ ++++(white)
X citri pv. aurantifolii C HR HR HR ++++
+++ weak canker;++++ strong canker; no symptoms; HR Hypersensitive Response
From: Brunings & Gabriel (2003)

The pathogen is dispersed for short distances by rain splash. It enters the plant

mesophyll through the stomata, or through open wounds. After about 4 days,

watersoaking can be observed. The bacteria produce xanthan gum, which is highly

hygroscopic, and fills up the intercellular spaces by absorbing water and swelling. Within

7-10 days after inoculation, the typical canker craters rupture and cause bacteria to ooze

to the surface, from whence they are dispersed through rain splash. The citrus leaf miner

Phyllocnistis citrella Stainton, although not a long-distance vector for the pathogen,

greatly accelerates local spreading of citrus canker by extensively wounding the leaf and

allowing easy entry inside the leaf. Dispersal of xanthomonads over longer distances is

facilitated by man and machinery (e.g., pruning tools).









Pathogenicity Factors

GenepthA and its Predicted Product

Gene pthA, a member of the avrBs3/pthA family of genes, appears to encode a

protein required for pathogenicity of citrus-canker causing strains. Members of the

avrBs3/pthA gene family are widespread among xanthomonads. To date 27 members of

the family have been identified (Gabriel, 1999b; Leach & White, 1996) and they are

remarkably similar in sequence. The genes are all flanked by inverted terminal repeats, so

they could potentially transpose (De Feyter et al., 1993). In fact, as will be discussed

later, pthA is found on a plasmid which also has ISXc4, an insertional element shown

capable of transposition (Tu etal., 1989). The predicted proteins all have from 15 to 22

direct, leucine-rich repeats, which are nearly always 34-aa in length. In addition, the

predicted proteins all encode three nuclear localization sequences, and a C-terminal

eukaryotic transcriptional activation domain (Gabriel, 1999b; Zhu et al., 1998). Most

members of the gene family have been shown to be avirulence (avr) genes that act in a

gene-for-gene fashion to cause a hypersensitive response on plants that carry the cognate

resistance gene (Gabriel, 1999a). GenepthA was the first member of the gene family

shown to be required for pathogenicity (Swamp et al., 1991). Since then, many, but not

all, gene family members have been shown to contribute to pathogenicity (Yang et al.,

1994).

The 102-bp repeat regions of the members of the avrBs3/pthA gene family are

important for the host specificity of pathogenicity and gene-for-gene specificity of

avirulence (Yang et al., 1994). The 102-bp repeats differ slightly in sequence; if the

specific repeats of one member is swapped with those of another, the specificity of the

resulting chimeric gene changes to that of the donor of the repeat region (Yang et al.,









1994). The repeat region within the members of the gene family is a source for alternate

specificities due to intragenic and intergenic recombination (Yang & Gabriel, 1995a).

The nuclear localization sequences of PthA and AvrBs3 have been shown to be

functional and required for pathogenicity and avirulence, respectively (Szurek et al.,

2001; Yang & Gabriel, 1995b). In addition, the eukaryotic transcriptional activation

domain at the C-terminus is required for avirulence activity of avrXalO, a member of the

gene family that is required for avirulence ofX. oryzae pv. oryzae (Zhu et al., 1998).

X citri pv. citri A strain 306, as all other type A strains examined, has four

members of the avrBs3/pthA gene family, namedpthA1, A2, A3 and A4. The fourpthA

genes are located on two native plasmids, pXAC33 and pXAC64; each plasmid encodes

two members (Figure 1-1). GenepthA4 is the same size aspthA, (isolated from X citri

pv. citri strain 3212T), and is 99.7% identical topthA. Additionally, the second leucine-

rich tandem repeat of both genes is exceptional, encoding 33 amino acid residues instead

of the 34 that are typical of all other repeats in these genes and in most other repeats in

the gene family. GenepthA1 has one repeat region less thanpthA and pthA4; while pthA2

andpthA3 each have two repeats less. Although functional analyses have not been

reported, these observations indicate thatpthA4 of strain 306 is most likely a functional

equivalent ofpthA.

Additional pathogenicity genes belonging to the avrBs3/pthA gene family have

been identified in citrus canker-causing strains, including pthB, pthC (Yuan & Gabriel,

unpublished), andpthW(Al-Saadi & Gabriel, 2002). All are isofunctional withpthA

(capable of restoring pathogenicity to apthA mutant strain) and required for

pathogenicity.











tnpR ISXc4 resolvase


tnpA TN5045 transposase
tnpATN5045 transposese


Tn5044 transposase


PthA4 n Ptns ),
I Sxac3 trans pos ase


ISXc4


-Co-integrase resolution protein T


pXAC64
64920 bp


VirB2


TrwB (VirD4)


TrwC


VirB8


TnpA 5


TnpA


Peml PemK


Figure 1-1. Maps of plasmids from Xanthomonas citri pv. citri Type A strain 306. A)
Map of plasmid pXAC64. B) Map of plasmid pXAC33.
From: Brunings & Gabriel (2003)









Quorum Sensing

Many plant pathogenic bacteria utilize quorum sensing---the induction of genes in

response to an increase in population density---to turn on pathogenicity genes inside the

host environment. Pathogenicity genes, including the biosynthesis of extracellular

polysaccharide (xanthan gum), ofXanthomonas campestris pv. campestris, are regulated

by a cluster of genes, designated rpf (regulation ofpathogenicityfactors; Dow et al.,

2000). The complete genome sequence of the Asiatic strain of citrus canker, X citri pv.

citri strain 306 (da Silva et al., 2002) revealed that genes homologous for all of the X

campestris pv. campestris rpf cluster are present. In X campestris pv. campestris, rpfA

encodes an aconitase, implicated in iron homeostasis (Wilson et al., 1998), and rpfB and

rpfF are responsible for the synthesis of a small diffusible signal factor, DSF (Barber et

al., 1997). As the bacterial population in the host environment grows, the increase in the

concentration of DSF is thought to trigger transcription of pathogenicity factors when it

exceeds a certain minimum threshold. The genes rpfC, G, and H, are proposed to be

components of a sensory-transduction system (Tang et al., 1991). A transposon insertion

in X campestris pv. campestris rpfE caused reduced levels of exopolysaccharide and

some extracellular enzymes (endoglucanase and protease), while the level of another

extracellular enzyme, polygalacturonate lyase, increased (Dow et al., 2000).

Nutrition

One way plant pathogenic bacteria can derive nutrition from their host is by

secreting cell-wall degrading enzymes and using the breakdown product of the cell walls

as a source of nutrition. X citri produces enzymes that could assist in the breakdown of

the plant cell wall and provide nutrition. Its genome shows that it does not have pectin

esterases, but there are three pectate lyases, six cellulases, five xylanases and an









endoglucanase. The endoglucanase, BcsZ (gil22001634), belongs in family 8 of the

glycosyl-hydrolases which hydrolyze 1,4-P-D-glucosidic linkages in cellulose. In

addition there is a permease which imports degraded pectin products in a hydrogen-

transport coupled fashion into the bacterial cell. X citri pv. citri produces fewer cell-wall

degrading enzymes than X campestris pv. campestris, and da Silva et al. (2002) suggest

that this may be why the two pathogens cause different symptoms on their hosts. X

campestris pv. campestris causes black rot of crucifers, and sometimes displays blight

symptoms (Alvarez et al., 1994), and spreads systemically through the xylem. X citri pv.

citri, on the other hand, doesn't cause rotting or blight of citrus, never becomes systemic

and only causes local lesions. In addition, pthA, which is likely injected directly into the

host cell, is required for optimal growth of the pathogen in citrus (Swarup et al., 1991).

This implies that some citrus plant responses) may be necessary for nutrient release,

possibly endogenous loosening of the cell wall in preparation of cell division and cell

enlargement.

Attachment

Bacteria can attach to host cells with special proteins called adhesins, or specialized

organelles called pili (Lee & Schneewind, 2001). X citri has four gene clusters and two

separately located genes that are predicted to be involved in type IV pilus biosynthesis

and regulation. Two genes encoding proteins called fimbrillins, FimA gi|21243966 and

gi|21243967 (85% similar in predicted amino acid sequence), are located within one of

the clusters. A gene designatedpilA, located elsewhere in the genome, is similar to type II

pilin (PilE) from Neisseria meningitidis. The sequence of the twofimA genes is similar to

PilA from other bacteria and they are located in a cluster of genes containing other type

IV pilus genes pilB, pilC, pilD, pilR, and pilS. The gene products ofpilS and pilR are









homologous to two-component sensor/regulatory proteins and control expression ofpilA

(Hobbs et al., 1993; Wu & Kaiser, 1997). The major subunit of the type IV pilin is first

exported by the general secretary pathway (GSP). It has a short, basic N-terminal signal

sequence, unique to type IV pilus proteins. PilD is a specific leader sequence peptidase

that removes the signal sequence of PilA and other type IV pilus biosynthesis proteins,

and methylates the new N-terminus (Finlay & Falkow, 1997; Russel, 1998). Mature,

translocated pilin polymerizes at the plasma membrane, and the pilus is pushed through

the central cavity of the outer membrane secretin (Parge et al., 1995; Russel, 1998).

Interestingly, there are two genes encoding proteins similar to PilA in X citri.

Pseudomonas stutzeri has type IV pili that are required for DNA uptake and natural

transformation, and two genes encoding proteins similar to PilA (74% similar at the

amino acid level; Graupner & Wackernagel, 2001).

The pilus biogenesis machinery and assembly is highly conserved in bacteria

(Hultgren et al., 1993). Their assembly genes are similar to type II secretion genes, but

the N-terminal signal sequences are different (Russel, 1998). Type IV pili (also called

fimbriae) have been proposed to attach bacterial pathogens to the host cell wall (Farinha

et al., 1994; Kang et al., 2002) and retract (Skerker & Berg, 2001; Wall & Kaiser, 1999),

pulling the bacterium closer to the host cell (Wall & Kaiser, 1999). Type IV pili are

important for virulence ofRalstonia solanacearum (Kang et al., 2002) and Pseudomonas

aeruginosa (Hahn, 1997). Since most plant pathogenic bacteria, including X citri, can

provoke a nonhost HR, plant cell wall attachment may not be (very) host specific.

Just upstream of the xps cluster (see below), X citri has two genes with similarity

to xadA from X oryzae pv. oryzae. XadA is a non-fimbrial, adhesin-like outer membrane









protein required for virulence ofX. oryzae pv. oryzae (Ray et al., 2002). Proposed to be a

cell wall surface anchor protein, XadA belongs to a family of proteins that are more

similar at their C-terminus (which forms an outer membrane anchor domain) than at their

signal sequences (Marchler-Bauer et al., 2002). Some non-fimbrial adhesins are

autotransporters (type V secretion): they are exported across the bacterial inner

membrane by the general secretary pathway, and then secrete themselves across the outer

membrane (Henderson et al., 2000; Henderson & Nataro, 2001). One of the xadA genes

encodes a protein (gi|21244271) that is missing the unusually long, highly conserved N-

terminal signal sequence that is typical of autotransporters (Henderson et al., 2000) and is

therefore not likely to be a functional XadA homolog. The other XadA (gi|2124427) does

have a conserved autotransporter N-terminal sequence. Therefore, it may be involved in

tight adhesion to plant cell walls, and could potentially have a function in bacterial

virulence. Two yapH genes (similar to yapH from Yersiniapestis) are predicted to encode

proteins similar to XadA, but both lack the typical N-terminal signal sequence of

autotransporters.

Type II (General) Secretion (T2S)

Type II secretion systems, (reviewed by Sandkvist, 2001) are common, but not

ubiquitous in gram negative bacteria (Finlay & Falkow, 1997). Secretion occurs in two

steps (Finlay & Falkow, 1997; Lee & Schneewind, 2001;): the Sec (general secretion)

machinery exports substrates with a signal peptide across the inner membrane of gram

negative bacteria and the type II secretion genes secrete them across the outer membrane.

Bacterial T2S systems have been shown to secrete diverse molecules such as cellulases,

pectate lyases, toxins, proteases, and alkaline phosphatases (Russel, 1998).









There are two T2S clusters in the X citri genome associated with two

rearrangements of the otherwise highly syntenic genomes of X citri and X campestris pv

campestris (da Silva et al., 2002). The first one (the xcs cluster) consists of one

transcriptional unit of thirteen open reading frames, xcsC through xcsN, and one

(downstream of xcsN) is similar to a TonB-dependent receptor gene. Consistent with the

idea that this T2S cluster has been transferred horizontally, it has a higher (-68%) G+C

content than the surrounding DNA region (-65%). The second T2S system (the xps

cluster) contains two transcriptional units. One consists of the xpsE and xpsF genes and

the other of ten genes: xpsG through xpsN, xpsD and a conserved hypothetical gene,

similar to glycosyltransferase genes. Part of this cluster has a G+C content as low as

47.5%.

Type III Secretion (T3S)

Many plant and animal pathogenic bacteria have a T3S system (T3SS) consisting

of more than twenty proteins which together function to inject pathogenicity factors

directly into host cells (Buttner & Bonas, 2002; Hueck, 1998). The T3S genes of plant

pathogens are called hrp (hypersensitive response and pathogenicity) genes and are

required both for pathogenicity on hosts and elicitation of the HR on hosts and nonhosts

(Fenselau & Bonas, 1995; Roine et al., 1997). Nine of these genes are highly conserved

between the T3S systems of plant and animal pathogens and have been renamed hrc (hrp

conserved; Bogdanove et al. (1996). The X campestris pv. vesicatoria Hrp system can

secrete at least some heterologous type III secreted proteins from both plant and animal

pathogens (Rossier et al., 1999).

Expression of the T3SS can be contact-dependent (Ginocchio et al., 1994; Menard

et al., 1994; Pettersson et al., 1996; Rosqvist et al., 1994; Watarai et al., 1995), but in the









case of Salmonella, T3S may not be contact-dependent (Daefler, 1999). Close contact

with plant host cells appears necessary for X citri, since there is compelling evidence that

PthA is injected directly into host cells. First, it has recently been demonstrated directly

that a member of the AvrBs3/Pth family,AvrBs3, is secreted by T3S (Szurek et al.,

2001). Second, mutations in the T3SS ofX. citri pv. aurantifolii render the bacterium

nonpathogenic (El Yacoubi et al., 2004). Third, E. coli carrying a T3SS from

Pseudomonas syringae pv. syringae and apthA allele causes canker-like symptoms on

citrus (Kanamori & Tsuyumu, 1998). Fourth, if PthA is delivered directly into the plant

cell, it is capable of forming canker-like symptoms (Duan et al., 1999). Fifth, PthA

localizes to the nucleus (Yang & Gabriel, 1995b), and so does AvrBs3 (Szurek et al.,

2001). Finally, the transcriptional activation domain of AvrXalO, another member of the

AvrBs3/Pth family of proteins, is functional (Zhu et al., 1998).

The hrp genes are proposed to encode proteins that form a hrp pilus (Roine et al.,

1997). The X citri hrp cluster is part of a pathogenicityy island" in the main

chromosome, as indicated by the following features (Hacker & Kaper, 2000): 1) it spans

more than 23-kb; 2) it encodes a system that secretes pathogenicity factors into host cells;

3) it is always associated with pathogenic species ofXanthomonas; 4) it has regions with

a different G+C content than the rest of the X citri genome; 5) it carries mobile genetic

elements (transposases), and 6) it represents an unstable region of DNA since there are

differences between the X citri T3SS and the closely related T3SS cluster from X

campestris pv. vesicatoria, which is also part of a pathogenicity island (Buttner & Bonas,

2002; Noel et al., 2002). The genes of the hrp cluster are induced inplanta and controlled

by the hrp regulatory proteins HrpG and HrpX (Wengelnik & Bonas, 1996; Wengelnik et









al., 1996, 1999). For some HrpX-regulated genes involved in the T3S system, a PIP

(plant-inducible promoter) box has been identified (Fenselau & Bonas, 1995), although

there are genes with PIP boxes that are not regulated by HrpX, and there are genes whose

expression is under control of HrpG, and HrpX that do not have a PIP Box (Buttner &

Bonas, 2002).

Type III effectors are injected by the TTSS into the host cell. It is logical to assume

that expression of these effectors would be coordinately regulated with expression of the

T3SS genes. In a screen performed by Guttman et al. (2002) for type III secreted

Pseudomonas syringae effector proteins, 13 different hop (hrp/hrc outerprotein) genes

were identified, of which all but one had a hrp box (Innes et al., 1993) in their promoters.

The situation may be somewhat different in Xanthomonas. Although da Silva et al.

(2002) found twenty potential PIP boxes in the X citri genome, only a few of these

indicated potential T3 S effector proteins. Homologues of four avr genes (avrBs2,

avrXacEl, avrXacE2 and avrXacE3) and twopopC family effector genes were also

found. Interestingly, the avrXacE2 homologue, one ofthepopC homologues, and all four

members of the avrBs3/pthA gene family do not contain a PIP box (da Silva et al., 2002).

In fact, all avrBs3/pthA gene family members examined to date are constitutively

expressed and yet are known to be delivered by T3S (Knoop et al., 1991; Szurek et al.,

2001; Yang & Gabriel, 1995b).

Three mechanisms have been proposed to explain how the T3SS recognizes

effectors for secretion. The first proposes that N-terminal signal sequences in the secreted

protein are recognized by the T3SS and result in export of the effector (Miao & Miller,

2000; Mudgett et al., 2000). However, type III secreted proteins lack a clearly defined









signal sequence in their amino termini (Aldridge & Hughes, 2001). In the second

mechanism, molecular chaperones bind effector proteins transiently, preventing them

from folding incorrectly, and present them to the T3SS apparatus for subsequent

secretion (Wattiau et al., 1994, 1996) A third mechanism was proposed for the T3S of

Yop proteins by Yersinia (Anderson & Schneewind, 2001). They hypothesized that the

secretion signal was encoded in the messenger RNA (mRNA) instead of in the amino

acid sequence of the secreted protein. The three kinds of secretion signal are not mutually

exclusive. For example, YopE appears to have two separate secretion signals in its amino

acid sequence, one of which functions only in conjunction with the secretion chaperone,

SycE (Cheng et al., 1997). In Xanthomonas campestris pv. vesicatoria, a region of

AvrBs2 was determined to be required for T3S and translocation to the plant cell, but a

potential mRNA secretion signal was also found (Mudgett et al., 2000). It was reported

that 13 type III secreted proteins from Pseudomonas syringae have very similar N-

terminal regions, and most are predicted to localize to chloroplasts in the plant cell, which

could point to a common recognition mechanism as in chloroplasts, or to a common

origin of the signal sequences (Guttman et al., 2002). It has been proposed that a signal

sequence in the message of AvrB and AvrPto is responsible for their recognition by the

Pseudomonas syringae TTSS (Galan & Collmer, 1999).

Type IV Secretion (T4S)

One of the native plasmids of strain 306, pXAC64, appears to encode a type IV

secretion (T4S) or "adapted conjugation" system, and at least portions of a second

potential T4S cluster are located in the main chromosome (da Silva et al., 2002). T4S

systems mediate intercellular transfer of macromolecules (proteins or protein-DNA

complexes) from gram-negative bacteria to other bacteria or eukaryotic cells (Baron et









al., 2002; Christie, 2001). The prototype for T4S systems is the Agrobacterium

tumefaciens virB cluster (Christie, 1997), of which virB2 through virB11 have all been

shown to be essential for Agrobacterium virulence (Berger & Christie, 1994; Ward et al.,

1990). The adapted conjugation system is, as the name implies, required for the

conjugational transfer of plasmids, including self-mobilizing plasmids, which carry the

T4S transfer genes and an origin of transfer (oriTor mobilization site). Adapted

conjugation systems can also mobilize other plasmids in trans if the plasmid carries an

oriTsite (Christie and Vogel, 2000; Winans et al., 1996). A T4SS is also necessary for

the secretion of pertussis toxin by Bordetellapertussis (Christie & Vogel, 2000). T4S

systems also require a "coupling factor" (Cabez6n et al., 1997; Moncalian et al., 1999), a

homolog of the Agrobacterium VirD4 protein. The order in which different components

of the T4SS of Agrobacterium contact the T-DNA strand has been determined using a

transfer DNA immunoprecipitation (TrIP) assay (Cascales & Christie, 2004). The

proposed pathway is depicted in Figure 1-2 (Cascales & Christie, 2004).

The T4SS in the main chromosome of X. citri strain 306 is incomplete. There are

three partial copies of the virB6 gene, there is no virB5 homolog, and most virB

homologs that are present are incomplete. This partial virB cluster is surrounded by

several transposases, indicating that this area may have been subject to more than one

transposition event. This T4S cluster is not likely to encode a functional secretion system,

since it lacks essential genes. It has been suggested that the T4SS on plasmid pXAC64 is

also incomplete, since it appears to lack virB5, virB7, and virD4 homologs (da Silva et

al., 2002). This conclusion may be incorrect, and needs experimental verification.












VirB2
VVirB3
VirB5
VirB10
VirB8
VirB6

t4.- VirB4

VirB1l

1 -- VirB7

VirD4
Figure 1-2. Pathway of the T-DNA strand through the Agrobacterium tumefaciens T4SS.
The proteins in the vertical pathway make contact with the T-DNA strand in
the order indicated, the other proteins do not make direct contact with the
substrate, but are necessary at the steps indicated. Reprinted with permission
from Cascales & Christie (2004), Fig 4, page 1172

The virB cluster on pXAC64 does not contain any genes that are annotated as

virB7. It was found that VirB7 is required very early in the secretion process (Cascales &

Christie, 2004). It is proposed to stabilize the secretary apparatus of the A. tumefaciens

T4SS (Beaupre et al., 1997; Fernandez et al., 1996). The predicted protein gene product

of XACb0044 (gi|21110908) on pXAC64 is similar to VirB5 and is annotated in

GenBank as such. Second, a BLASTP (Altschul et al., 1990) search with the predicted

TrwB sequence from pXAC64 shows similarity with VirD4 and its homologs. TrwB

functions as a coupling factor for T4S systems and trwB is a known virD4 homolog

(Gomis-Ruth et al., 2002; Moncalian et al., 1999). Finally, the T4S virB cluster on

pXAC64 is similar in organization to that of the Ivh cluster from Legionella pneumophila,

Genbank accession Y19029 (Segal et al., 1999), where there also are two genes between

the homologs of virBS (lvhB5) and virB6 (lvhB6). Remarkably, the gene immediately









downstream of IvhB5 is a homolog of virB7, IvhB7. The gene immediately downstream

of virB5 in pXAC64 may therefore be a virB7 homolog, similar to the organization of the

T4SS in L. pneumophila. This gene organization is discussed in more detail in Chapter 3.

Plasmids pXAC33 and pXAC64

Xanthomonas citri pv. citri A strain 306 from Brazil has two native plasmids;

pXAC64 is 64.9 kb in size, while pXAC33 is 33.7 kb [(da Silva et al., 2002); refer Figure

1-1]. The restriction endonuclease maps and sizes of these two plasmids correspond

almost perfectly to the restriction endonuclease maps and estimated sizes of two

plasmids, pXW45J and pXW45N, respectively, from X citri pv. citri A strain XAS4501

from Japan that were characterized by Tu et al. (1989). Two transposable elements,

ISXc4 and ISXc5, were functionally characterized by transposition in E. coli and their

locations were mapped on plasmids pXW45N and pXW45J (Tu et al., 1989). The

corresponding locations of these IS elements on pXAC64 and pXAC33 are indicated in

Figure 1-1.

These IS elements are of potential interest because the TnpR resolvase of ISXc5

represents a new subfamily of recombinases responsible for resolution of cointegrates of

class II transposable elements, such as Tn3 (Liu et al., 1992, 1998). Intriguingly,

pXAC64, which presumably carries a functional allele ofpthA (pthA4), also has a tnpA

transposase gene that is 99.9% similar to tnpA of Tn5044, which is in the Tn3 subgroup

of the Tn3 family of transposases (Kholodii et al., 2000). Since all members of the

avrBs3/pthA family have terminal inverted repeats that are similar to Tn3 transposable

elements, it was hypothesized that these genes may transpose (De Feyter et al., 1993). It

should now be possible to experimentally test this hypothesis. If the T4SS on pXAC64 is

functional, the plasmid might be self-mobilizing, Self-mobilizing plasmids carry the









transfer functions necessary to transfer other plasmids with an origin of transfer (oriT). If

pthA4 is indeed a functional pthA homolog and pXAC64 is self-mobilizing, then

pXAC64 would be able to transfer itself into other xanthomonads resident on the same

host, some of which may lack ability to cause citrus canker. ISXc4 on the plasmid could

cause transposition of thepthA gene, and multiple copies would be created. The T4SS

would likely be instrumental in initiating such transfer.

Plasmid Toxin-Antitoxin Modules

Low-copy number plasmids constantly face the problem of stable maintenance in

the absence of selection pressure (Gerdes et al., 1986). The rate at which plasmid-free

cells accumulate in culture depends on the rate in which they arise, and the relative

growth of plasmid-containing and plasmid-free cells (Gerdes et al., 1986). The growth-

rate of plasmid-containing cells is depressed by plasmid metabolic load, which results

from the demands placed upon the host by replication, transcription and translation of the

plasmid genome. The load can become severe if the plasmid is used to express genes with

gene product that are toxic or interfere with the host's metabolism (Gerdes et al., 1986).

Low-copy number plasmids have developed multiple approaches to maintain their

segregational stability. Among these are (1) an active partition system ensures that each

daughter cell receives a copy of the newly replicated plasmid at cell-division; (2) site-

specific recombination of dimers and higher-order multimers which arise by homologous

recombination optimizes the number of plasmids available for segregation at the time of

cell division; (3) toxin-antitoxin (TA) systems. TA systems compromise the survival of

plasmid-free segregants (Lehnherr et al., 1993; Magnuson et al., 1996), mediate the

exclusion of competing plasmids (Cooper & Heinemann, 2000), and prevent cell-division

in plasmid-free daughter cells (Ruiz-Echevarria et al., 1995). A toxin-antitoxin locus









should stabilize any unstably inherited plasmid, no matter what causes the loss of the

plasmid (Gerdes et al., 1986).

TA Mechanisms

Toxin-antitoxin (TA) systems do not affect copy number of plasmids (Deane &

Rawlings, 2004; Radnedge et al., 1997; Rawlings, 1999), but make them more stable than

can be accounted for by replication control and active partition (Lehnherr et al., 1993).

The toxins are proteins encoded by TA systems are very potent and their artificial

overproduction leads to rapid and massive cell-killing, in most cases corresponding to

several orders of magnitude in reduction of viable-cell counts (Gerdes, 2000).

Presumably, as a result of this toxicity, attempts to clone the toxin gene by itself have

often been unsuccessful (Gerdes, 2000; Hayes, 1998; Smith and Rawlings, 1998), except

under tightly regulated non-leaky promoters (Gerdes, 2000; Roberts & Helinski, 1992),

or replacement of the start codon by one that reduces the level of toxin expression

(Gerdes, 2000).

There appear to be two mechanisms by which the antidote can neutralize the effect

of the toxin. In the hok (host killing)/sok (supressor of host killing) system, the antisense

sok RNA prevents synthesis of hok (Gerdes et al., 1990, 1992). In most other

mechanisms, the antitoxin is proposed to be a protein. The antitoxin forms a tight

complex with toxin (Gerdes, 2000) and inactivates it. The toxins are generally stable,

while the antitoxins are degraded by cellular proteases (Gerdes, 2000). Immediately after

cell division, plasmid-free cells have a pool of TA complexes plus a pool of free antitoxin

(Gerdes, 2000; Rawlings, 1999). While the antitoxin degrades, the toxin remains, and

kills the cells (Gerdes, 2000). There is a large amount of variability in the amino acid

sequence of TA encoding genes, especially between the sequence of the antitoxins, which









makes it difficult to identify TA complexes by homology searches (Deane & Rawlings,

2004; Gerdes, 2000; Rawlings, 1999).

TA Operon Organization

Stability operons appear to be organized similarly on different plasmids. In most

cases a gene encoding an antitoxin precedes a gene that encodes a toxin, and the genes

overlap (Hayes, 1998; Rawlings, 1999; Smith & Magnuson, 2004), suggesting

translational coupling (Roberts & Helinski, 1992). Both genes are predicted to encode

proteins of about 10 kDA. Putative promoter elements are typically present immediately

upstream of the antitoxin gene. The antitoxins autoregulate transcription of the TA

operons by binding to operator sites upstream or overlapping with the operon promoters

(Eberl et al., 1992; Gerdes, 2000; Magnuson et al., 1996; Magnuson & Yarmolinsky,

1998; Smith & Magnuson, 2004) and negatively regulating transcription (Magnuson et

al., 1996). ThepasABC promoter is autorepressed by PasA (Smith & Rawlings, 1998). In

the case of the Phd repressor/antitoxin protein, the N-terminus is required for repressor,

but not for antitoxin activity, while the C-terminus is required for antitoxin, but not for

repressor activities (Smith & Magnuson, 2004).

Autoregulation appears to be essential forpas-mediated plasmid stabilization

because when thepas genes were placed behind the IPTG-regulated tac promoter, they

were unable to stabilize a heterologous test plasmid (Smith & Rawlings, 1998). Under

conditions in which the killing function is artificially overexpressed from a strong

external promoter, killing is observed in plasmid-containing cells (Gerdes et al., 1986).

This indicates that the operon is tightly regulated under native conditions. The pas

stability system is unusual in that it consists of three genes rather than the two-gene

systems identified in other plasmids (Smith & Rawlings, 1998). ThepasA gene encodes









an antidote, pasB encodes a toxin, and pasC encodes a protein that appears to enhance the

neutralizing effect of the antidote (Smith & Rawlings, 1998). The ParD protein alone is

sufficient for autoregulation of theparDE operon (Gerdes, 2000). Expression ofparE

seems to rely on translational coupling with parD (Roberts & Helinski, 1992).

Testing Plasmid Stability

Roberts & Helinski (1992) tested plasmid stability by diluting mid-log phase

cultures under antibiotic selection into antibiotic-free medium and allowing them to grow

overnight to mid-log or stationary phase. Aliquots were plated on antibiotic-free medium

both before and after overnight growth. The stability of the plasmid was expressed as the

percentage of cells carrying plasmid at the initial and final time points.

The plasmid maintenance stability determinant of the large virulence plasmid

pMYSH6000 of .\/ge//t flexneri consists of two small open reading frames STBORF1

and STBORF2 and is likely to encode a postsegregational killing system (Radnedge et

al., 1997; Sayeed et al., 2000). This Stb system also exerts incompatibility against a co-

resident plasmid containing Stb.

Insertional mutagenesis showed that parD and parE were both required for plasmid

stabilization (Roberts & Helinski, 1992). The parDE operon is sufficient to confer vector-

independent, broad-host range stability under several conditions tested (Roberts &

Helinski, 1992). However, in rich-medium growth of E.coli carrying a wild-type recA

gene, this minimal region is insufficient for plasmid stabilization and requires the

presence of theparCBA operon (which functions to resolve plasmid multimers to

monomers) for efficient stabilization under certain growth conditions (Roberts &

Helinski, 1992).















CHAPTER 2
A NOVEL PATHOGENICITY FACTOR REQUIRED FOR CITRUS CANKER DISEASE IS
CONTAINED ON A SELF-TRANSMISSIBLE PLASMID

The only pathogenicity factors known to be required for citrus canker disease are a

functional T3SS and at least one member of the Xanthomonas avr/pth gene family (Gabriel,

1999b). Xanthomonas citri pv. aurantifolii B69 has been reported to have at least two native

plasmids (Civerolo, 1985), a T3SS (Leite, Jr. et al., 1994), and two DNA fragments that

hybridize with pthA, indicating the presence of two members of the Xanthomonas pthA gene

family (Yuan & Gabriel, unpublished results). One of these genes, pthB, is located on a self-

transmissible plasmid, pXcB, and DNA fragments carrying pthB complement a mutant Asiatic

strain that has pthA interrupted (Yuan & Gabriel, unpublished). Cloned DNA fragments carrying

the other potential pthA gene family member fail to complement the pthA mutation, therefore

pthB is apparently the only active member of thepthA gene family in B69 (Yuan & Gabriel,

unpublished results). Gene-swapping experiments showed thatpthA and pthB are isofunctional.

pthB is necessary for B69 to cause citrus canker disease (Yuan & Gabriel, unpublished

results). However, although pthB is required for pathogenicity, it is not the only gene on pXcB

that is required, since pthB cannot alone complement a B69 strain, B69.4, cured of pXcB (Yuan

& Gabriel, unpublished results). In this study, pXcB was mapped. Based on the map, several

subclones of pXcB were made, and then tested for self-transmissibility and pathogenicity by

mating into B69.4 carrying pthA.










Material and Methods

Bacterial Strains, Plasmids, and Recombinant DNA Techniques

The bacterial strains and plasmids used in this study are listed in Table 2-1 and Table 2-2.

Xanthomonas spp. were cultured in PYGM medium at 300C (De Feyter et al., 1990).

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

370C. For solid media, agar was added to 15 g/liter. Antibiotics were used at the following final

concentrations in [g/mL: ampicilin (Ap) 100 for high copy vectors, 50 for medium copy vectors;

chloramphenicol (Cm) 35, gentamycin (Gm) 3, kanamycin (Kn) 25, rifampicin (Rif) 75 or 50,

spectinomycin (Sp) 35, streptomycin (Sm 100), tetracyclin (Tc) 15.

Table 2-1. List of bacterial strains used in this study
Strain Relevant Characteristics Reference or source
E. coli
supE44A lacU169((p8OlacZAM15) hsd R17 recA1 gyrA96 Gibco-BRL,
DH5a thi-1 relAl Gaithersburg, MD
supE44 hsdS20(rB mB-) recA13 ara-14 proA2 lac Y1 gal K2 (Boyer and Roulland-
HB101 rpsL20 xyl-5 mtl-1 (Smr) Dussoix, 1969)
Xanthomonas
3213T X citri pv. citri, species type strain, ATCC 49118 (Gabriel et al., 1989)
3213Sp Sp' mutant of 3213T (Gabriel et al., 1989)
pthA::Tn5-gusA, marker-exchange mutant of 3213Sp
B21.2 (Sp'Knr) (Swamp et al., 1991)
B69 X citri pv. aurantifolii B strain B69 ATCC 51301
B69Sp Sp' derivative of B69 This study
pthB::pUFR004 marker interruption mutant of B69Sp Yuan & Gabriel,
BIM2 (Sp'Cmr); carries pBIM2. unpublished
pXcB::pUFR004 cointegrate derivative of B69Sp Yuan & Gabriel,
BIM6 (Sp'Cmr); carries pBIM6 unpublished
Yuan & Gabriel,
B69.4 Spontaneous Rif mutant of B69, cured of pXcB unpublished

Plasmids were transferred between bacteria using triparental matings as described (Swarup

et al., 1991), using pRK2013 as helper plasmid. In experiments to test self-transmissibility of a

particular plasmid, the helper strain was omitted.










For extraction of total DNA from Xanthomonas, 12 mL bacterial cultures were grown in

25 mL flasks at 300C with slow shaking. Cells were harvested at 8000 x g and washed twice

with 12 mL and 1.5 mL 50 mM Tris-HC1, 50 mM EDTA, 0.15 mM NaCl (pH 8), respectively.

The cells were resuspended in 627 pL of TES buffer (10 mM Tris-HC1, 10 mM EDTA, 0.5%

SDS, pH 7.8). Afterwards 33 [tL of protease stock solution was added, the tubes were inverted

several times to mix well, and the cell suspension was incubated at 370C for /2-3 hours. The

protease stock solution consists of 20 mg/mL protease in 10 mM Tris-HC1, 10 mM NaC1, pH7.5;

the protease was predigested at 370C for 1 hour and stored at -200C. The mixture was gently

mixed overnight by rotation with an equal volume of phenol:chloroform:isoamyl alcohol

(25:24:1) buffered with Tris-HCl pH 8. The layers were separated by centrifuging for 15 minutes

at 5,000 x g, and the top layer was gently transferred to a new tube with a pipet. A second

extraction with phenol:chloroform:isoamyl alcohol (25:24:1) was carried out, followed by an

extraction with chloroform:isoamyl alcohol (24:1). One-tenth volume of 3 M NaAc was added,

the tubes were inverted several times, and 0.9 volume of room temperature isopropanol was

added and mixed. Precipitated DNA was spooled out with a heat-sealed glass Pasteur pipet,

transferred to a tube containing 600 /L of 10 mM Tris-HC1, 1 mM EDTA, pH 8, and 200 /g/mL

RNaseA, and incubated for 1-2 hours at 370C. A phenol:chloroform:isoamyl alcohol and a

chloroform extraction were carried out, followed by precipitation with NaAc and 2 volumes of

room temperature 95% ethanol. After careful mixing, the total DNA was spooled out as before

and resuspended in 200 [tL sterile distilled water. To analyze the DNA, 1 tL was digested with a

restriction enzyme and run on a 0.7% agarose gel.











Plasmids were isolated from E. coli using alkaline lysis (Sambrook et al., 1989). For

extraction of plasmid DNA out of Xanthomonas, cells were washed twice in TEN (50mM Tris-

HC1, 50 mM EDTA, 150 mM NaC1, pH 8) buffer to remove excess xanthan gum prior to alkaline

lysis.

Table 2-2. List of plasmids used in this study
Plasmid Relevant characteristics Reference or source
pXcB Native, self-mobilizing 37-kb plasmid from B69 This study
pRK2013 ColE1, Knr, Tra helper plasmid (Figurski & Helinski, 1979)
pUFR047 IncW, Mob lacZa Par+ (Ap'Gmr) (De Feyter et al., 1993)
pUFR053 IncW, Mob+, lacZa Par+ (CmrGmr) Yuan & Gabriel, unpublished
pLAFR3 IncP1, cos+, Tra, Mob (Tc') (Staskawicz et al., 1987)
pBIM2 pthB::pUFR004 (Cmr) Yuan & Gabriel, unpublished
pBIM6 pXcB::pUFR004, cointegrated upstream of pthB (Cmr) Yuan & Gabriel, unpublished
pAB2.1 EcoRI-HindIII fragment of pZit45, containing pthA, in This study
pLAFR3
pAB 11.62 18-kb EcoRI fragment of pBIM2 in pUFR047 This study
pAB 12.2 14-kb EcoRI fragment of pBIM2 in pUFR047 This study
pAB 13.2 8.9-kb EcoRI fragment of pBIM2 in pUFR047 This study
pAB 14.1 7.2-kb EcoRI fragment of pBIM2 in pUFR047 This study
pAB 18.1 EcoRI-HindIII fragment of pYD9.3 in pUFR047 This study
pQY96 14-kb HindIII fragment containing pthB from pXcB, cloned Yuan & Gabriel, unpublished
in pUFR053
pYD9.3 pthA in pUC 118 (Ap') (Duan et al., 1999)
pUFY14.5 pthA in pGEM 7Zf(+) (Ap') (Yang & Gabriel, 1995b)
pYY40.9 StuI-HinclI fragment of pthA in pUFR004 Yang & Gabriel, unpublished
pZit45 pthA in pUFR047 (Swamp et al., 1992)
B69.4 Spontaneous Rif mutant of B69, cured of pXcB Yuan & Gabriel, unpublished

Electroporation was carried out using the Eppendorf Electroporator 2510, according to the

manufacturers' instructions. Electrocompetent cells were prepared by washing the cells from a

liquid culture with 1 x culture volume of distilled water. The cells were then washed with 0.1 x

volume of distilled water and resuspended in sterile 10% glycerol, followed by storage at -80C

until use. The cells were thawed out completely on ice (about 10 minutes). 1 tL of DNA solution

(diluted 1:10 to minimize the salt concentration) was added to 40 tL (for cuvettes with a gap

width of 1 mm) or 100 tL (for cuvettes with a gap width of 2 mm) of electrocompetent cells,










gently mixed and incubated on ice for 20-60 seconds. Electroporation was carried out in sterile

cuvettes with a voltage setting of 1800 V for cuvettes with a 1 mm gap width, and 2500 V for

cuvettes with a 2 mm gap width. To ensure that the electroporation was carried out with high

efficiency, the time constant was noted after each electroporation. The electroporation was

repeated with a greater dilution of DNA if the time constant was below 4.0 ms. Immediately after

electroporation, 1 mL rich medium (SOBG for E. coli cells) was added to the cuvette. Medium

and electroporated cells were then transferred to glass culture tubes and incubated for 1 hour at

370C while shaking. 200 pL of the mixture was then spread out on selective plates and grown

overnight at 37C.

Other standard recombinant DNA procedures were used essentially as described

(Sambrook et al., 1989). Restriction enzyme digestion, alkaline phosphatase treatment, DNA

ligation, and random priming reactions were performed as recommended by the suppliers.

Southern hybridization was performed using nylon membranes as described (Lazo & Gabriel,

1987).

Plant Inoculations

Citrus plants "Duncan Grapefruit" (Citrus paradise) and "Mexican Lime" or "Key Lime"

(Citrus aurantifolii) were grown under natural light in the exotic pathogen quarantine greenhouse

facilities at the Division of Plant Industry, Florida Department of Agriculture and Consumer

Services, Gainesville, Florida. Two-day old liquid bacterial cultures were rinsed and adjusted to

an OD600 of 0.4 with sterile CaCO3-saturated water, and pressure infiltrated into the abaxial leaf

surface of the plants using tuberculin syringes (Yang & Gabriel, 1995a). All inoculations were

repeated at least three times.









Total
DNA


m m


Plasmid
DNA
I ,I

m m


23.1 kb

9.4 kb

6.6 kb

4.4 kb


2.3 kb
2.0 kb


23.1 kb

9.4 kb

6.6 kb

4.4 kb


2.3 kb
2.0 kb


Figure 2-1. Members of the avr/pthA gene family in X citri pv. aurantifolii. A) Ethidium
bromide stained gel of strains B69 and B69.4. Strain B69 has multiple plasmids. B)
Southern hybridization of gel in A) probe: internal BamHI fragment ofpthA. Strain
B69 has 2 members of the avr/pthA gene family, both of which are on plasmids. DNA
in all lanes is restricted with EcoRI










Results

B69 Has Multiple Plasmids and a T3SS

Plasmid and total DNA were extracted from B69 and its derivative lacking pXcB, B69.4.

Comparing the lanes labeled plasmidd DNA B69" and plasmidd DNA B69.4" in Figure 2-1 part

A, it is clear that B69.4 (which lacks pXcB) has at least one additional plasmid.

Two plasmids in B69 have EcoRI fragments (-18 kb, and -4.5 kb) that hybridize to an

internal BamHI fragment ofpthA. The 18 kb fragment is derived from pXcB and is absent from

B69.4, which is cured of pXcB (Figure 2-1 part B). The 4.5 kb EcoRI fragment is located on a

plasmid other than pXcB and still present in B69.4.

Gene pthB is Insufficient to Cause Canker

Strain BIM2, in which pthB alone is mutated, was unable to cause canker. BIM2 was

complemented with pAB2.1, pZit45, pAB 18.1 (all with pthA), and pQY96 (withpthB); thus

confirming thatpthB is necessary for pathogenicity and thatpthA is isofunctional withpthB.

Shuttle vectors containing pthA orpthB were mated into B69.4 and inoculated on Duncan

Grapefruit and Mexican Lime leaves. None of the plasmids used to complement BIM2 restored

pathogenicity to B69.4; therefore pthB is necessary but insufficient to cause citrus canker, and

one or more additional pathogenicity factors must reside on pXcB.

Mapping of pBIM2 and pBIM6

Plasmid pBIM2 (pthB::pUFR004) does not restore pathogenicity when introduced into

B69.4, however, pBIM6 (pXcB::pUFR004) does. In pBIM2 the inserted vector interrupts pthB

(Figure 2-2), while in pBIM6 (Figure 2-3) the inserted vector appears to have no mutational

effect. Both pBIM2 and pBIM6 were mated into E. coli for restriction and Southern blot

analysis. Plasmid pBIM2 DNA was digested with a variety of restriction enzymes for mapping











purposes (Figure 2-4). Several restriction sites (for example EcoRI, HindIII, KpnI, PstI, and

Sstl) were easily mapped, since they occurred infrequently in pXcB. Additional sites for EcoRI,

SstI, and KpnI at the 3' end of the integrated vector (the start of subclone pAB 13.2, see Figure 2-

2) were also mapped. Additional sites were mapped based on restriction digests of pBIM2,

pBIM6, or their subclones (data not shown).

E H E E HP BP BPH ESK BKSa HE
ShB pBIM2
pAB14.1
SpAB12.2
I lr hB5 4 BKSa pAB11.62
pUFR004 oi pAB13.2
Figure 2-2. Restriction map of plasmid pBIM2 and its subclones. E=EcoRI, B=BamhHI,
H=HindIII, K=KpnI, P=Pstl, S=Sstl, Sa=SalI


E HE E H K H BB HE
I I pBIM6
pUFRM4
i kb duplicalmd regian
Figure 2-3. Restriction map of plasmid pBIM6. E=EcoRI, B=BamHI, H=HindIII, K=KpnI


Constructs

Based on the restriction map of pBIM2, the four EcoRI fragments of pBIM2 were

subcloned into the shuttle vector pUFR047 (De Feyter et al., 1993). The resulting subclones

(Figure 2-2) are pAB11.62 (16 kb), pAB12.2 (12.7 kb), pAB13.2 (8.9 kb), and pAB14.1 (6.3 kb).

Plasmid pAB11.62 contains the 5'-end ofpthB, while pAB13.2 contains the 3'-end. Both of

these clones hybridize topthA (Figure 2-5).











Gene pthA was recloned directionally into cosmid vector pLAFR3 (repP) from the

EcoRI-HindIII fragment from pYY14.5 (Yang & Gabriel, 1995a). The resulting clone, pAB2.1,

contains pthA driven by its native promoter.

Gene pthA was recloned directionally into cosmid vector pLAFR3 (repP) from the EcoRI-

HindIII fragment from pYY14.5 (Yang & Gabriel, 1995a). The resulting clone, pAB2.1, contains

pthA driven by its native promoter.

Pathogenicity Tests

Plasmid pAB2.1 complemented B21.2, apthA-interrupted mutant of an A strain of canker.

To test whether any of the subcloned fragments of pBIM2 would be able to confer pathogenicity

to B69.4 in conjunction with pAB2.1 (carryingpthA on a repP plasmid), pAB11.62, pAB12.2,

pAB13.2, and pAB14.1 (carrying BIM2 fragments on a repWplasmid) were introduced into

B69.4 together with pAB2.1. Plasmid pAB2.1 was mated into B69.4 as a negative control. None

of these combinations of plasmids restored B69.4 to pathogenicity.



i g
-


23.1 kb

9.4 kb

8.8 kb

4.4 kb



2.3 kb
2.0 kb


Figure 2-4. Restriction digests of plasmid pBIM2 for mapping purposes









Since none of the subcloned fragments of pXcB was able to restore B69.4 pathogenicity

in the presence ofpthA, it is possible that whatever other factors) were required for

pathogenicity were truncated in the subclones. Alternatively, it is possible that multiple factors

are required for pathogenicity in addition topthB, and that none of the subclones provided a full

complement of necessary factors.


EcoRI BamHI EcoRI+BamHI
SN II II N I-

i -

S< < < < ci < < c < <
0- M. M M. M 0- M. 0 M M


ii 0
% w -


Mi ::::::: II
a
as.


Figure 2-5. Plasmid pBIM2 and subclones. A) A 0.7% agarose gel of plasmid DNA extracted
from pBIM2 and its subclones. B) Southern hybridization of the gel in A; probe
internal BamHI fragment ofpthA










Self-transmissibility

Each of the pBIM2 subclones was tested for self-transmissibility. Of the recipient strain

(R; E. coli DH5a MCR, or HB101), donor stain (D; E. coli DH5a MCR/pAB1 1.62, pAB12.2,

pAB 13.2, or pAB14.1) and helper strain (H; E. coli DH5a MCR/pRK2013), every possible

combination (RxDxH, RxD, RxH, DxH) was allowed to mate overnight on plain LB plates and

streaked on selective plates afterwards, selecting for the donor plasmid in the recipient strain.

None of the donor strains used in these matings resulted in colonies for the R+D combination,

while all combinations of RxDxH resulted in viable colonies. Whatever other factor is required

for self-mobility was either truncated in the subclones or none of the subclones provided all of

the necessary mobility factors.

Discussion

Members of the AvrBs3/PthA family of proteins are probably transferred directly into host

cells by the hrp system (Casper-Lindley et al., 2002; Duan et al., 1999; Yang & Gabriel, 1995b).

Consistent with this theory, pthA can cause canker-like symptoms when introduced directly into

citrus cells using particle bombardment or an Agrobacterium-based delivery system (Duan et al.,

1999).

In X campestris pv. aurantifolii there is at least one and possibly two members of the

Xanthomonas avr/pth gene family, of which one, pthB, is functional and located on a self-

mobilizing plasmid, pXcB. When pthB by itself, without the rest of pXcB, is introduced inX.

campestris pv. aurantifolii cured of the plasmid (B69.4), B69.4 does not regain the ability to

cause citrus canker. These data imply that there is at least one other factor on pXcB required for

pathogenicity. Subcloned fragments of pBIM2 do not complementpthA upon introduction

together with pthA into B69.4. Possible explanations include that more than one additional factor










might be necessary, or that the required factors) are incomplete in the subclones. Since no

other genes are known to be critical for citrus canker disease, there might be part of the hrp

system on pXcB, or alternatively, a chaperone forpthB. pXcB was therefore sequenced

completely (see Chapter 3). However, there are no genes similar to hrp genes or chaperones were

found on the plasmid.

The fact thatpthB resides on a self-mobilizing plasmid raises the interesting point that a

Xanthomonas strain is able to mobilize pathogenicity genes by horizontal transfer. This might

explain how different Xanthomonas spp. are able to cause essentially the same disease, differing

only in their host range. Hypothetically, strains that already have a hrp system and factors

elsewhere in their genome that determine their host range could receive pXcB by horizontal

transfer. If pXcB were then stably maintained in the recipient strain, the result might be a strain

capable of causing citrus canker disease.















CHAPTER 3
COMPLETE PLASMID SEQUENCE

Chapter 2 showed thatpthB is necessary but insufficient for pathogenicity of Xanthomonas

campestris pv. aurantifolii strain B69. The fact that none of the pXcB subclones restored

pathogenicity together with pthA, means that the required factor or factors are incomplete on the

subcloned fragments. Plasmid pXcB was sequenced completely to identify all possible

pathogenicity factors.

Material and Methods

Bacterial Strains, Plasmids, and Culture Media

The bacterial strains and plasmids used in these experiments are listed in Table 2-1 and

Table 2-2 in Chapter 2. Xanthomonas strains were cultured as described in Chapter 2.

General Bacteriological Techniques

The self-mobilizing plasmid pXcB and its pthB-interrupted derivative pBIM2, were

transferred from X citri pv. aurantifolii strain B69 to E. coli DH5a or HB101 by biparental

matings. For all other matings a helper strain, pRK2013 (Figurski & Helinski, 1979), was used in

triparental matings (Swamp et al., 1991). The DNA sequence of pXcB was determined by

sequencing the EcoRI fragments of pBIM2 subcloned into pUFR047: pAB 11.62, pAB 12.2,

pAB 13.2, and pAB 14.1 (refer to Chapter 2). Regions overlapping the subclones were sequenced

directly from pBIM2.










Recombinant DNA Techniques

Plasmid DNA extractions, DNA analysis techniques, and electroporation were as described

in Chapter 2.

DNA Sequence Analysis

DNA sequencing was performed by the Interdisciplinary Center for Biotechnology

Research (ICBR) sequencing core. The GCG sequence analysis software package (Wisconsin

Package Version 10.3, Accelrys, Inc. San Diego, CA.) was used to assemble the sequence data.

Putative open reading frames (orfs) capable of encoding peptides of at least 50 amino acids

were identified using ORF Finder (National Center for Biotechnology Information, 2004) and

the Open Reading Frame Identification feature (Bayer College of Medicine, 2003). The program

Codonpreference (Wisconsin Package Version 10.3, Accelrys Inc., San Diego, CA), which

scores orfs based on the similarity of their codon usage compared to a codon usage table

(codonfrequency) or by their third position GC bias was used to eliminate orfs that were unlikely

to be coding sequences, and to determine sequencing errors, since it simplifies identification of

frame shifts. BLASTX (Bogdanove et al., 1998) was used to identify orfs for which homologues

exist in the GenBank non-redundant database. Whenever possible, putative Shine-Dalgarno

(Shine & Dalgarno, 1974) sequences (ribosome-binding sites) were identified.

Potential promoters were identified using the BCM Gene Feature Searches option for

prokaryotic promoter prediction by neural network (LBNL). Transcriptional terminators were

predicted using the program GESTER (Unniraman et al., 2001, 2002). Individual putative

protein sequences were then analyzed in more detail using the programs that are part of the

Expert Protein Analysis System (Swiss Institute of Bioinformatics, 2003). For predicted

molecular weight and pi, PSORT (Nakai & Horton, 1999), and MOTIF (Kyoto University










Bioinformatics Center, 2004) were used. Putative transmembrane regions were identified

using TmPred (Hofmann & Stoffel, 1993). For the identification of lipoprotein, the DOLOP (A

Database Of Bacterial LipOProteins) program was used (Madan & Sankaran, 2002).

Results

Features of Plasmid pXcB

pXcB is a circular double-stranded DNA molecule of 37,106 bp. Thirty eight orfs were

identified on pXcB (Table 3-1). There are a total of 30,729 base pairs of coding DNA, resulting

in an average gene density of 1.02 orfs/kb, and 828 coding base pairs/kb.

Open reading frames (orfs) were initially identified using ORF Finder, the BLASTX

program and the GCG package program Codonpreference. Thirty six orfs appeared to start with

an ATG initiation codon, and two (orfs 117 and 219) were predicted to start with a GTG.

Sequences 6-12 nucleotides upstream of predicted start sites were compared to the Shine-

Dalgarno consensus sequence TGGAGG (Shine & Dalgarno, 1974). Most orfs appeared to have

sequences similar to these ribosome binding sites. A map of the orfs found on pXcB is given in

Figure 3-1, they are listed in Table 3-1. The process used to find orfs is illustrated in Figure 3-2

for orf207 and orf208. In the figure the 5'-end of each orf is on the right hand side of the

horizontal box. The codon preference and third position codon bias of orf208 remain high

throughout the predicted orf indicating a real gene, while the third position codon bias of orf207

drops significantly in the latter half of the predicted orf, indicating that this orf might not

represent a real gene. Both orfs are similar to genes in the Genbank database. Orf207 is similar to

a hypothetical gene on pWWO and a hypothetical gene on pXAC64 (gi:21264228) plasmid (da

Silva et al., 2002), while orf208 is similar to the traA gene on pWWO from P. putida and

another hypothetical gene on pXAC64. Orf208 was determined to be more likely to encode a









gene, based on the better third position codon bias, and renamed traA. A similar process was

used to hand-annotate each orf on the plasmid.


parA2


orf201


pthB


orf202
orfl00


35k Ok


/


S 30k


secA
kfrA H
25k
tnpR 2
orf122 4/E
IS401-
orf120


pXcB

37106 bp


20k
/


stbB /
traD
traA


/ orf115
trwC yacB
yacA


Figure 3-1. Map of plasmid pXcB. E=EcoRI, H=HindIII. Orfs that are part of the T4SS are
indicated in red, genes homologous to other transfer genes in blue, genes similar to
plasmid maintenance genes in purple, the insertion element and resolvase in orange,
orfs similar to genes in the GenBank database in green, orfs with no similarity to
known genes in grey, and pthB in yellow.

The average G+C content was 60.9%, with a minimum of 46.25% and a maximum of

71.75%. The G+C percentage, calculated for windows of 300 bp, with a step size of 3 was

plotted against nucleotide position (Figure 3-3). The G+C content forms a trough in a region of


orf217


parA
orf215
orf214
orf213


orf113










relatively low G+C content, roughly from position 7200 to 7800. This region corresponds to

orfs 105 and 106. From 32560 to 34475 (corresponding to the 102 -nucleotide tandem, direct

repeat region of orfl30, pthB) the G+C content was remarkably high compared to the rest of

pXcB.

Promoters

Potential promoter regions were identified with the Bayer College of Medicine Gene

Feature Search for Promoter and Transcription Binding Site Predictions. Eleven promoters with

a score of 0.95 or higher were identified on the +strand, and fifteen on the -strand of pXcB.

Lowering the cutoff to 0.90 resulted in the prediction of eleven additional promoters on the

+strand and seven additional promoters on the -strand. An even lower cutoff of 0.85, resulted in

fifteen more predicted promoters on the +strand and twelve more on the -strand. A list of

potential promoters with a score of 0.85 or higher is given for the +strand in Table 3-2, and for

the -strand in Table 3-3; they are illustrated in Figure 3-4.

There do not appear to be separate promoters for orfs 101, through 103, orfl05 through

107, and for orfl08 through 114. Orfl01 through 104, orfs 105 through 108, and orfs 111 and

112 have overlapping coding sequences and are most likely transcriptionally and translationally

co-regulated. Orfs 113 and 114 are likely transcribed from the same promoter as orfs 111 and

112, since the closest promoter has a low cutoff (0.87) and is almost a thousand bases away from

the initiation codon of orfl 13. This could add five orfs to the virB cluster that are not related to

any known T4SS genes. Upstream of orfl 15 (at position 14256-14301) there is a potential

promoter (cutoff 0.86) for orfl 15 (which starts at position 14556). Orfs 115 through 117 could

be transcribed either separately or together with upstream orfs.













Table 3-1. Open reading frames identified in plasmid pXcB
Orf Position Gene #AA MW (kDa) pi Similarity
201 c(706..1527) orf201 273 30.4 10.94 pXAC33() gi:21264186, XF(2)
gi:10956774
202 c(2030..2359) orf202 109 11.4 7.82 nsh(3)
100 (2757..3143) mobD 128 14.8 9.27 pXAC'4-I4' gi:21264276, PP(5)
gi:32469966
101 (3172..3549) virB2 125 13.4 8.92 pXAC64 VirB2 gi:21264275, PP
gi:31745842


102 (3546..3845)



103 (3832..6501)








104 (6498..7163)



105" (7214..7429)/
(7229..7429)


virB3 99 11.0



virB4 889 101.5








virB5 221 23.9



orfl05/ 71/ 7.7/
virB7 66 9.51


8.55 pXAC64 VirB3 gi:21264274, PP
gi:31745843


6.41 pXAC64 VirB4 gi:21264273, PP
gi:31745844, 18150991, 32469963






8.80 pXAC64 VirB5 gi:21264272, PP
gi:31745845, 18150990, 32469962,
Pseudomonas syringae gi:28867760

9.67/ pXAC64 gi:21264271, PP gi:32469961,
7.06 31745846,18150989


Predicted function


CDD: pfam06921, VIRB2, VIRB2 type IV
secretion protein; COG3838, VirB2, Type IV
secretary pathway; cyclin T-pilin subunit (Lai
& Kado, 1998; Jones et al., 1996),
attachment, mating channel (Christie, 2001)
CDD: pfam05101, VirB3, Type IV secretary
pathway; COG3702, VirB3, Type IV
secretary pathway; attachment, stabilized by
VirB4, VirB6 (Christie, 2001)
CDD: pfam03135, CagETrbEVirB, CagE,
TrbE, VirB family, component of type IV
transporter system; COG3451, VirB4, Type
IV secretary pathway; translocation
energetic, ATP-ase (Christie, 2001);
ATP/GTP-binding site motif A (P-loop);
energy for pilus biogenesis (Cascales and
Christie, 2003)
attachment, minor pilin subunit, stabilized by
VirB6 (Christie, 2001)


mating channel, lipoprotein (Christie, 2001);
stabilizes VirB9, and thereby other VirB
proteins (Cascales & Christie, 2003)













Orf Position Gene #AA MW (kDa)
106 (7407..7829) orfl06 140 15.3


107 (7817..8683)


virB6 288


31.5


108 (8680..9354) virB8 224 25.0





109 (9379..10167) virB9 262 28.6



110 (10164..11384) virB1O 406 42.9


pi Similarity
6.72 PP gi:18150988, 32469960, 31745847,
pXAC64 gi:21264270
8.44 pXAC64 VirB6 gi:21264269, PP
gi:31745848, 32469959, 18150987,
Legionella pneumophila gi:19919314,
6249468




8.94 pXAC64 VirB8 gi:21264268, PP
gi:31745849, 32469958, 18150986,
Legionella pneumophila gi:19919315,
6249469


9.01 pXAC64 VirB9 gi:21264267, PP
gi:31745850, 18150985, 32469957


6.04 pXAC64 VirB10 gi:21264266, PP
gi:32469956, 31745851, 18150984


Predicted function


CDD: pfam04610, TrbL, TrbL/VirB6
plasmid conjugal transfer protein; COG3704,
VirB6, Type IV secretary pathway; mating
channel, required for the stability of VirB3
and VirB5, and formation of VirB7
homodimers (Hapfelmeier et al., 2000),
possibly a inner membrane pore component
(Christie, 2001)
CDD: pfam04335, VirB8, VirB8 protein;
COG3736, VirB8, Type IV secretary
pathway; mating channel, assembly factor for
positioning of VirB9 and VirB10 (Christie,
2001); bridge between subcomplexes
(Cascales & Christie, 2003)
CDD: CagX (6.9% aligned); COG3504,
VirB9, Type IV secretary pathway; stabilizes
T4SS; possibly outer membrane pore
(Cascales & Christie, 2003)
CDD: pfam03743, TrbI, Bacterial
conjugation TrbI-like protein; COG2948,
VirB 10, Type IV secretary pathway; mating
channel (Christie, 2001); bridge between
inner membrane and outer membrane
subcomplexes (Cascales & Christie, 2003)













Orf Position Gene #AA MW (kDa) pi Similarity
111 (11395..12414) virB1l 339 38.1 8.55 pXAC64 VirBll gi:21264265, PP
gi:31745852,32469955, 18150983


112 (12401..13312) virB1 303





113 (13358..13945) orfll3 195




114 (13958..14563) orfll4 201


31.9





20.8




21.7


115 (14556..14978) orfll5 140 15.9


9.19 pXAC64 VirB1 gi:21264264, pSB102
gi: 15919992, Brucella melitensis biovar
Abortus gi:8163884, B. melitensis
gi:17988369


6.80 E. coli hypothetical protein gi:21885930,
A. tumefaciens conserved hypothetical
protein gi: 17938747, Sinorhizobium
meliloti hypothetical protein gi: 16263173,
Rhizobium etli gi:21492820
9.62 Burkholderia cepacia gi:46320012, PP
putative nuclease gi:18150980, Proteus
vulgaris EDTA-resistant nuclease
gi:21233859, Yersinia enterocolitica
endonuclease gi:2208977, Salmonella
thyphimurium EDTA-resistant nuclease
gi:4903112
5.38 nsh


Predicted function
CDD: pfam00437, GSPII E, Type II/IV
secretion system protein; COG0630, VirB 11,
Type IV secretary pathway; COG4962,
CpaF, Flp pilus assembly protein, ATPase;
translocation energetic, ATP-ase (Lai &
Kado, 2000); energy for pilus biogenesis
(Cascales & Christie, 2003)
CDD: cd00254, SLT, Transglycosylase SLT
domain; pfam01464, SLT, Transglycosylase
SLT domain; mating channel,
transglycosylase (Christie, 2001; Lai & Kado,
2000); channel assembly (Cascales &
Christie, 2003)
ATP/GTP-binding site motif A (P-loop)




CDD: cd00138, PLDc, Phospholipase D;
ProDom: endonuclease plasmid EDTA-
resistant nuclease; Pfam: phospholipase D
active site motif













Orf Position Gene #AA MW (kDa)
116 (15110..15393) yacA 95 10.7


117 (15394..15681) yacB 95 10.8








205 c(15731..18706) trwC 991 109.6





206 c(18720..20294) trwB 524 58.2







208 c(20486..20881) traA 131 14.5


pi Similarity
9.98 Photorhabdus luminescens subsp.
laumondii gi:37525856, \i I.i.,ill, "i ,,i
europaea ATCC 19718 gi:30248566,
Bartonella henselae str. Houston-1
gi:49475484, \hg ii,/ sonnei YacA
gi:9507443
4.49 AiI ,lI 'i ,i,,i europaea ATCC 19718
30248567, Photorhabdus luminescens
subsp. laumondii TTO1 gi:37526182,
Azotobacter vinelandii gi:23102756,
Ralstonia metallidurans CH34
gi:48766888, Bartonella henselae str.
Houston-1 gi:49475483, S. sonnei YacB
gi:9507444
9.62 pXAC64 TrwC gi:21264259, PP
gi:31745859, 18150978,32469948,E.
coli gi: 19572639, 1084124



8.93 pXAC64 TrwB gi:21264258, PP TraB
gi:32469947, 18150977, 31745860, E.
coli TrwB gi: 1084123




6.31 pXAC64 hypothetical protein
gi:21264253, PP TraA gi:31745861,
32469946, 18150976


Predicted function


CDD: pfam05016, Plasmid stabilisation
system protein; COG3668, ParE, Plasmid
stabilization system protein





CDD: COG0507 RecD, ATP-dependent
exoDNAse (exonuclease V), alpha subunit -
helicase superfamily I member (DNA
replication, recombination, and repair);
Prosite pattern: ATP/GTP-binding site A
motif (P-loop), lipocalin,
CDD: pfam02534, TRAG, TraG/TraD family
member of the TraG/TraD (73.5%); Prosite
pattern: ATP/GTP-binding site motif A (P-
loop), walker-B site for nucleotide binding;
BLOCKS: TraG protein; ProDom: plasmid
TraD membrane inner ATP-binding; Pfam:
TraD/TraG













Orf Position Gene #AA MW (kDa)
118 (21269..21724) traD 151 17.0


119 (21795..22427) stbB


210 23.8


120 (22437..22827) orfl20 129
209 c(22978..23295) IS401 105




122 (23463..23894) orfl22 143


210 c(24077..25039) tnpR


211 c(25198..25884) kfrA



125 (26609..27384) secA


13.3
11.6




15.3


320 36.2


228 24.3



258 28.9


pi Similarity
9.55 pXAC64 hypothetical protein
gi:21264198, PP TraD gi:, 18150975,
32469945, E. coli gi:49065168,
25815153, yciA gi: 10955477
5.40 pXAC64 conserved hypothetical protein
gi:21264200, PP hypothetical protein
gi: 18150974, 32469944, X fastidiosa
conserved hypothetical protein
gi: 10956760, Salmonella typhimurium
stbB gi:2801371
8.06 PP hypothetical protein gi: 18150973
9.82 Azotobacter sp. FA8, gi:22077125,
Ralstonia metallidurans CH34
gi:48770250, Burkholderia cepacia IS401
gi:2497402, j\ i, i.i ii, ii,,p, europaea
ATCC 19718 gi:30248272
8.37 P. syringae pv. pisi Hrp effector candidate
gi:42475527, Ralstonia solanacearum
putative transmembrane protein
gi:17548482
9.39 Y. enterocolitica gi:37518401,A.
tumefaciens putative resolvase
gi:15891106, 17937554
4.86 pXAC64 KfrA gi:21264227,
Achromobacter denitrificans gi:45368551,
P. aeruginosa gi:37955784, plasmid RK2
KfrA gi:78651, P. alcaligenes gi:2429365
5.42 pXAC64 conserved hypothetical protein
gi:21264217, Burkholderia cepacia
R1808 gi:46320800


Predicted function


Insertion sequence (Wood et al., 2001)









CDD: pfam00239, Resolvase,COG1961,
PinR, Site-specific recombinases





CDD: pfam02810, SEC-C motif; COG3318,
Predicted metal-binding protein related to the
C-terminal domain of SecA













Orf Position Gene #AA MW (kDa) pi Similarity Predicted function
213 c(27823..29010) orf213 395 45.7 8.90 pXAC64 conserved hypothetical protein CDD: pfam06414, Zeta toxin; ATP/GTP-
gi:21264286, Burkholderia cepacia binding site A motif (P-loop)
R1808 gi:46322784, P. alcaligenes
unknown gi:49188515
214 c(29017..29301) orf214 94 10.4 4.56 pXAC64 hypothetical protein
gi:21264286
215 c(29368..29700) orf215 110 12.0 9.98 Y. enterocolitica orf78 gi:28373039
216 c(29693..30325)parAl 210 22.2 6.85 E. coli gi:46949068, Y. enterocolitica CDD: COG1192, Soj, ATPases involved in
orf79 gi:28373041, Y. pestis biovar chromosome partitioning; similar to orf219
Mediaevails str. 91001 ATPases involved
in chromosome partitioning gi:45476526
217 c(30438..31232) orf217 264 29.2 9.69 P. syringae pv. tomato str. DC3000
gi:28867389, A. tumefaciens hypothetical
protein gi: 10954854, Streptomyces
violaceoruber gi:32455690
130 (31719..35221) pthB 1168 123.1 6.10 X. citri pv. citri PthA gi:899439, pXAC64 CDD: pfam03377, Avirulence,Xanthomonas
PthA4 gi:21264293, 4163845 avirulence protein, Avr/PthA; pathogenicity
protein
219 c(35813..36442)parA2 209 22.1 5.59 pXAC33 partition protein A gi:21264226, CDD: pfam00991, ParA, ParA family
Bartonella henselae str. Houston-1 ATPase; COG 1192, Soj, ATPases involved
gi:49476005, C hil, I'i.,,,i limicola in chromosome partitioning; COG0455,
unknown gi: 10956076 ATPases involved in chromosome
partitioning
* CDD=Conserved Domain Database
** See text for explanation of two sets of data for Orfl05
(1) pXAC33= X citri pv. citri strain 306 plasmid pXAC33 (gi:38201775)
(2) XF= Xylellafastidiosa
(3) nsh = no significant hits
(4) pXAC64= X citri pv. citri strain 306 plasmid pXAC64 (gi:21264228)
(5) PP= P. putida plasmid pWWO
















10
o05


20
S 15


C 05
Soo0 0 ORF 207 c(20298 20717)


15



-oo ORF 208 c(20486 20881) h


i i iIi I III 1 I1 1 1 1 1 1 1 1 11 1 1,1 1 1 1
21,000 20,500 20,000
Figure 3-2. Codonpreference of plasmid pXcB reverse frames from position 20,000 to
21,000.


Orfs 205 and 206 may share the same promoter, since there is no separate promoter


for orf205 and the coding sequences of the two orfs overlap, implying translational


coupling. Orf210 does not appear to have a promoter at the cutoff levels shown.


virB cluster trwB/C vanous genes pthB



G+C 60 V N V

50
45-----------------------------------------------
0 10,000 20,000 30,000
Position


Figure 3-3. G+C content of plasmid pXcB. Window size is 300, step size 3


Transcriptional terminators were identified with the GeSTer algorithm (Table 3-4).


Since there were two possible promoters found upstream of orfl00 and no transcriptional


terminators until after orfl 17, it is possible that this entire region, spanning almost 15 kb


and comprising of 18 orfs, is transcribed together. There were no strong promoters


(cutoff greater than 0.90) predicted immediately upstream of the trwB and trwC genes.


However, two promoters are predicted upstream of the traA gene (both with a score of










0.85). It is possible that traA, trwB, and trwC are transcribed from the same promoter.

There were two predicted transcriptional terminators downstream of trwC. Two potential

promoters were predicted to drive pthB transcription, but no transcriptional terminators

were found downstream ofpthB.

Table 3-2. Predicted promoters for the + strand of plasmid pXcB with a score of 0.85 or
higher
Start End Score Start End Score
17 62 0.96 21014 21059 0.86
1875 1920 1.00 21213 21258 0.92
2372 2417 0.90 22416 22461 0.92
2383 2428 1.00 22868 22913 0.95
3367 3412 0.89 25899 25944 1.00
3738 3783 0.94 26135 26180 0.85
6040 6085 0.90 26302 26347 0.94
6102 6147 0.85 27095 27140 0.91
7924 7969 0.98 27451 27496 0.99
7939 7984 0.92 27733 27778 0.92
11088 11133 0.88 28028 28073 0.97
12353 12398 0.89 28120 28165 0.86
14256 14301 0.86 28868 28913 0.87
15044 15089 0.94 30791 30836 0.89
15223 15268 0.87 31096 31141 1.00
15835 15880 0.90 31477 31522 0.94
19724 19769 0.85 31672 31717 0.88
19783 19828 0.89 35498 35543 0.99
20865 20910 0.96 35526 35571 0.87
36674 36719 0.95

Open Reading Frames

The largest orf found on pXcB was orfl30, which was predicted to encode PthB, a

member of the AvrBs3/PthA family of proteins. PthB was more than 90% identical to

PthA, and had thirteen 34-amino acid, tandem, direct repeats in the middle of the amino

sequence. The program TmPred was used to predict one weak transmembrane helix in

PthB, but the score was barely above the cutoff level (see Table 3-5). Since PthB is

proposed to be secreted by the T3SS, like other members of the AvrBs3/PthA family of

proteins, it most likely does not have a transmembrane helix.










Table 3-3. Predicted promoters for the
higher


Start
36510
35535
35047
34920
32395
31708
31368
30488
29413
29114
27504
26859
25946
22986
22195
20942
20900
18190


End
36465
35490
35002
34875
32350
31663
31323
30443
29368
29069
27459
26814
25901
22941
22150
20897
20855
18145


Score
0.85
0.99
0.91
0.95
0.86
0.99
0.96
0.86
0.98
0.88
0.96
0.98
1.00
0.88
0.85
0.85
0.85
0.93


strand of plasmid pXcB with a score of 0.85 or


orf
orf orf 115 orf
orf 1 of11 116
'100 i 113
orr virB cluster //
... .... .. .... .............. ....
-jc C; C-


orf
120 orf arf arf
IDSIb 122 Ir 213 rf 215 parA


.i I. : ;('" piB -pXcB


> > >>> >>>>> cutofftf


_> > > *> >> _> > >_ > >> > tff nma1-B

= <4 < < <- tennffatM


Figure 3-4. Potential promoters on plasmid pXcB. Purple arrows indicate predicted
promoters with a score of 0.95 or higher. Red arrows indicate additional
predicted promoters if the cutoff score is lowered to 0.90, while orange arrows
indicate additional promoters if the cutoff is set to 0.85. The direction of the
arrow indicated the direction of the promoter. The bottom line gives the
position and direction of transcriptional terminators

A large part (10.1 kb, 27%) of pXcB consisted of an apparently complete T4SS,

called the virB cluster, since the orfs closely resemble the organization and sequence of

the virB cluster of genes present in the crown gall agent Agrobacterium tumefaciens


O.9-0.94


Start
15287
15100
13369
12711
11960
11125
8301
8256
7795
6957
6206
5282
4739
3861
3684
2457
1987
1923
539


End
15242
15055
13324
12666
11915
11080
8256
8211
7750
6912
6161
5237
4694
3816
3639
2412
1942
1878
494


Score
0.85
0.97
0.98
0.93
0.92
0.87
0.94
0.88
0.87
0.98
0.96
0.86
0.85
0.94
0.95
0.88
0.97
0.96
0.89


> >>>>_ > > > > > >- cutoff 0.9&1.0
% % w.IC 4 w. %% w. -W. ^-% l % .510









(Christie, 1997). Orfl06 appeared to be part of the T4SS cluster, but it was not similar to

any known type IV secretion gene. Five orfs were found similar to mobilization or

transfer genes, two were similar to genes involved in plasmid partitioning, three orfs were

similar to other known genes, and 13 orfs were either similar to genes of unknown

function, or not similar to any known genes. There was one IS element and one resolvase

gene found.

Table 3-4. Transcriptional terminators identified in plasmid pXcB by GeSTer
Position Orf Strand
15893 yacB(orf117) +
22924 orfl20 +
15674 trwC (orf205)
22957 IS401 (orf209)
27618 orf213
27708 orf213
35633 parA2(orf219)
35729 parA2 (orf219)

The pXcB virB Cluster

The Agrobacterium virB cluster consists of eleven genes, virB 1 through virB 11 (Christie

& Vogel, 2000). The pXcB virB cluster seems to be a hybrid of a T4SS and unrelated

(orfl06, and 113 through 117) genes, which is not uncommon (Christie, 2001). For the

most part, the virB cluster is organized in the same way as the orfs in the Agrobacterium

virB cluster; the data comparing the pXcB virB cluster to several other T4SS is

summarized in Table 3-6 and depicted in Figure 3-5. There are significant differences

between the pXcB virB cluster and the prototype T4SS cluster from Agrobacterium. The

virB] gene is located downstream of virBO1 instead of upstream of virB2. There are two

orfs between virB5 and virB6 instead of virB6 being immediately downstream of virB5,

and virB7 is missing from in between virB6 and virB8.















Table 3-5. Presence of N-terminal signal sequences, localization of putative proteins

(PSORT), and transmembrane helices (TmPred)


Signal

Orf sequence


106 cleavable


107

108


109 cleavable


110

111


112 cleavable


113 cleavable


114 cleavable


cleavable


122 cleavable


uncleavable


Predicted

localization
cytoplasm

cytoplasm

cytoplasm

inner membrane

inner membrane

cytoplasm

outer membrane

inner membrane

periplasmic space

outer membrane

inner membrane

inner membrane

outer membrane

periplasmic space

inner membrane

cytoplasm

periplasmic space

outer membrane

periplasmic space

outer membrane

periplasmic space

outer membrane

cytoplasm

cytoplasm

cytoplasm

cytoplasm

inner membrane

cytoplasm

cytoplasm

cytoplasm

inner membrane

cytoplasm

periplasmic space

outer membrane

inner membrane

cytoplasm

cytoplasm

cytoplasm

cytoplasm

cytoplasm

cytoplasm

inner membrane

cytoplasm

cytoplasm


Transmemberane N-terminus Score

helices* INside or OUTside (cutoff 500)


Certainty
0528

0 133

0403

0232

0472

0 186

0600

0361

0929

0217

0495

0429

0939

0324

0297

0 193

0923

0 143

0915

0321

0278

0265

0344

0267

0453

0243

0 119

0314

0362

0318

0387

0378

0943

0363

0 134

0240

0 182

0480

0 164

0278

0212

0 115

0342

0 122


4949

4186


1602

2614


2499


12699

2642


1800


3678



948


1502


1924









4098






2644



1796


1119


846








1492

521










Table 3-6. Comparison of the pXcB virB cluster with several type IV secretion-systems.
Percent similarity is indicated in brackets
pXAC64 R46 (2) pWWO (3) R388 (4) R6K (5) T (7)
pXcB VirB () (IncN) (IncP-9) (IncW) (IncX) lvhB (6) VirB pXF51 (8)
VirB2 VirB2 (94) MpfA (42) EnhD (46) VirB2 (54) XFa0005
VirB3 VirB3 (100) TraA (ns) MpfB (67) TrwM (41) LvhB3 (39) VirB3 (53) XFa0006
VirB4 VirB4 (98) TraB (41) MpfC (74) TrwK (39) Pilx4 (37) LvhB4 (43) VirB4 (60) XFa0007 (40)
VirB5 VirB5 (98) TraC (40) MpfD (73) TrwJ (40) LvhB5 (39) VirB5 (57) XFa0008 (44)
VirB7 XACb0043 (89) Eex (55) orf200 (76) Eex(51) Eex (59) LvhB7 (45) VirB7 (41)
orfl06 XACb0042 (65) -orfl99 (66) LvrD (49)
VirB6 VirB6 (92) TraD (32) MpfE (69) TrwI (34) Pilx6 (32) LvhB6 (36) VirB6 (62) XFa0011 (35)
VirB8 VirB8 (98) MpfF (56) TrwG (37) Pilx8 (37) LvhB8 (40) VirB8 (58) XFa0012 (31)
VirB9 VirB9 (97) TraO (38) MpfG (73) TrwF (46) Pilx9 (31) LvhB9 (49) VirB9 (67) XFa0013 (41)
VirB10 VirB10 (94) TraF (42) MpfH (56) TrwE (45) Pilxl0 (41) LvhB10 (48) VirB10 (56) XFa0014 (43)
VirB11 VirB11 (89) TraG (42) Mpfl (56) TrwD (42) Pilxl1 (45) LvhB11 (48) VirB11 (64) XFa0015 (46)
VirB1 VirB1 (92) TraL (42) MpfJ (53) TrwN (38) Pilxl (35) -VirB1 (48) XFa0016 (38)
(1) Xanthomonas citri plasmid pXAC64 (gi:21264228)
(2) Salmonella typhimurium plasmid R46 (gi: 17530571)
(3) Pseudomonas putida plasmid pWWO (gi: 18150858)
(4) Escherichia coli plasmid R388 (gi:21885926)
(5) Escherichia coli plasmid R6K (gi: 12053564)
(6) Legionella pneumophila (gi:6249457)
(7) Agrobacterium tumefaciens Ti plasmid (gi: 16119780)
(8) Xylellafastidiosa plasmid pXF51 (gi:9112238)

pXcB (vir) ~N t p BO B n BI

pXAC64 (vlr) N2 L1 lo Bli B

pWWVO (mpf) LD k p H I J
Ivh N B5
pvh p p B1O BliR2 D4

TI plasmid (vlr) I B2- p B10 B1
Figure 3-5. Comparison of the linear organization of several type IV secretion-systems.
Arrows with the same color indicate homologous genes


A comparison of the pXcB virB orfs with other T4SS genes is given in Table 3-6.

The synteny of the genes in the pXcB virB cluster is identical to that of the virB cluster of

pXAC64, and interestingly also to that of the putative mating pair formation gene cluster

of the P. putida plasmid pWWO, and the IvhB cluster of Legionellapneumophila (Figure


3-9 and Table 3-6).














CC B4
-pXcB VirB2
p--pVWVO Mpf
pXAC64 VirB2
pXcB VirB4
-- Ti VirB2
pXAC64 VirB4
R6K PilX2
R6K Pilx4
R64TraM
Ti VirB4
R388TrwL R64
R64TraB
pXF51 XFa0005
R388TrwK
enhD
pXF51 XFa0007

(C)
R6K Pilx6

IvhB6

---pWWO MpfE

pXcB VirB6

pXAC64 VirB6

R46TraD

R388Trwl

pXF51 XFa0011

Ti VirB6

CC2420

Figure 3-6. Phylogenetic trees compiled by CLUSTALW and drawn with TreeView of
A) VirB2 homologs, B) VirB4 homologs, and C) VirB6 homologs

As in pXcB, the Legionellapneumophila IvhB cluster has two orfs between lvhB5

and lvhB6, and the first one has been identified as lvhB7 (Segal et al., 1999). The VirB7









protein is a lipoprotein (Fernandez et al., 1996) that interacts with and stabilizes VirB4,

VirB9, VirB10, and VirB11 (Baron et al., 1997; Krall et al., 2002). The VirB7 protein

can form both homodimers and heterodimers with VirB9 through a disulfide bridge

(Baron et al., 1997; Anderson et al., 1996) in the periplasmic space, and could possibly

serve as a chaperone (Christie, 2001). Orfs 105 and 106 of pXcB were examined more

closely. VirB7 is rather small (Agrobacterium VirB7 has 54 amino acid residues), and its

similarity to homologs is not immediately apparent. Orfl05 (the smallest of the two orfs

between lvhB5 and lvhB6, and in the same relative position as lvhB7), the most likely orf

to encode a VirB7 homolog, was examined with the DOLOP algorithm that identifies

probable lipoproteins.

If the start codon for orfl05 is at position 7214, DOLOP does not recognize it as a

lipoprotein, since one of the requirements is that the lipobox ([LV][ASTV][ASG][C]) is

within the first forty amino acid residues and the predicted orfl05 gene product has a

sequence LAGC that fits the lipobox consensus sequence, but ends at amino acid residue

forty one. If the start codon for orfl05 is the next available ATG initiation codon at

position 7229, resulting in a 66-amino acid residue gene product as opposed to a 71-

amino acid residue gene product, VirB7 indeed meets all the requirements for a

lipoprotein, and the DOLOP algorithm recognizes it as such. The lipobox consists of the

sequence LAGC, there are 4 positively charged amino acids within the first 22 residues

of the signal sequence, and the length of the hydropohobic stretch in the signal sequence

is 10 residues. Both versions of orfl05 have potential ribosome-binding sites. VirB7 has

two conserved cysteines, one of which is used for cleavage of the lipoprotein signal

sequence, while the other is for forming a disulfide bridge between VirB7 and VirB9









(Anderson et al., 1996; Spudich et al., 1996). Both conserved cysteines are present in the

predicted amino acid sequence of the orfl05 gene product at residue positions 36 and

position 48. The truncated version of orfl05 therefore likely encodes a VirB7 homolog,

and is annotated as such. In Table 3-1 data is given for both the longer gene and the

shorter version of orfl05. The longer version is called orfl05, since it does not meet the

requirements for a VirB7 homolog, while the shorter gene is called virB7.

The IncN plasmid R46, which has homologs for most of the virB cluster genes, and

appears to be organized similarly as pXcB, has only one orf between the virB5 and virB6

homolog. This gene is called eex for entry exclusion (Pohlman et al., 1994), and it

appears to be a virB7 homolog. The eex gene is 15% identical and 55% similar to pXcB's

virB7 gene, its gene product is a lipoprotein as assayed by DOLOP, and like the VirB7

from pXcB it has three cysteine residues, of which one is at the end of the signal

sequence and one of the others is likely used to form a disulfide bridge with the VirB9

homolog. The same argument is applicable to the eex genes from the IncW plasmid R388

and the IncX plasmid R6K, which are therefore also indicated as virB7 homologs in

Table 3-6. The Xylellafastidiosa conjugation cluster on plasmid pXF51 also has two orfs

(XFa0009 and XFa0010) between the virB5 and virB6 homologs. However, neither of

these orfs is a lipoprotein, and therefore neither is listed as a virB7 homolog.

The gene product of the orf between the virB7 and virB6 genes, orfl06, is similar to

hypothetical proteins of pWWO and pXAC64, in both cases the orf is the one

immediately upstream of the virB6 gene. The orf106 predicted protein has a cleavable N-

terminal signal sequence and is most likely to reside in the bacterial periplasmic space.

Alternatively, it could locate to the bacterial inner membrane.









The CLUSTALW program was used to for a phylogenetic tree of the VirB2 [major

pilus subunit (Lai & Kado, 1998)] and VirB4 [translocation energetic (Shirasu et al.,

1994)] homologs, and VirB6 (T4SS channel). VirB4 is the largest gene in the cluster and

highly conserved. The results are in Figure 3-6. For each putative protein analysis, the

ones from pXcB are closest to the ones from pXAC64, and these two are grouped

together with the Pseudomonasputida pWWO homologue. In each case they are

distinctly different from the Agrobacterium tumefaciens virB cluster.

Immediately downstream of the virB cluster on pXcB, there are five ORFs (113

through 117) that are not similar to any T4SS genes described to date. They might be

transcribed as one polycistronic message together with the virB cluster genes since there

is no transcriptional terminator between the virB cluster and ORFs 113 through 117.

Orf113 is a conserved hypothetical protein, similar to orf34 on the IncW plasmid R388.

Orfl 13 has a cleavable N-terminal signal sequence and is therefore predicted to localize

to the bacterial periplasm, in addition it has an ATP/GTP-binding site motif A (P-loop)

and is thus likely to hydrolyze ATP or GTP. Orfl 14 is most similar to several nucleases

including one from P. putida (57% identity, 70% similarity in predicted amino acid

sequence), an EDTA-resistant nuclease from Proteus vulgaris, nucleases from Yersinia

enterocolitica, Salmonella typhimurium, plasmid ColIB-P9 from \lN/gel// sonnei, and a

putative phospholipase D from Salmonella typhimurium. The orfl 15 gene product has no

significant similarity to any known proteins. Orfl 17 has a GTG start site. The probability

that orfl 17 is a true coding sequence was assessed with the Codonpreference algorithm.

The Codonpreference analysis of pXcB from position 13301 through 16000 (containing

orfs 113 through 117) is depicted in Appendix B. The codonpreference of orfl 15 was









quite low (hovering around the average for random sequence, but the third position GC

bias was significantly high throughout the orf, indicating that orfl 15 is translated. Both

the codonpreference and third position GC bias of orfs 116 and 117 were well above the

average for random sequences.

BLASTP analysis (Altschul et al., 1990) found multiple plasmid-encoded predicted

proteins that are similar to the toxin part of toxin-antitoxin plasmid stabilization systems.

Orfl 17 was 71% identical and 84% similar (Expect value le-31) to a conserved

hypothetical protein from Nitrosomonas europaea ATCC 19718 (gi:30248567). A

Conserved Domain Database search (Marchler-Bauer et al., 2002) with the predicted

protein sequence of orfl 17 resulted in 100% alignment with a 85-residue conserved

domain from plasmid stabilization proteins, pfam05016 (Expect value 4e-12).

The highest BLASTP hit, with experimental evidence for its function, was StaB

(gi:33517368) from Paracoccus imethylihieu, (Szymanik et al., 2004). Orfl 16 was only

weakly similar to an anti-toxin gene, but several similarities existed between orfl 16 and

117 and TA modules as described in Chapter 1.

The case for orfl 16 being an antitoxin gene was not strong, since there were no

significant BLASTP hits to anti-toxin genes. This does not rule out the possibility that it

is an antitoxin gene, since antitoxin genes are less well conserved than toxin genes

(Rawlings, 1999; Deane and Rawlings, 2004; Gerdes, 2000). Typically, TA systems are

organized in operons, and are driven off their own promoter, since the anti-toxin gene

negatively regulates expression of the operon. A potential promoter was predicted with a

score of 0.94, located from position 15044 through 15089 (see Table 3-2), immediately

upstream of orfl 16 (predicted Shine-Dalgarno sequence at 15100, and ATG start site at









15110). Both orfl 16 and orfl 17, like toxin-antitoxin proteins, are about 10 kDa, their

coding sequences overlap, suggesting transcriptional and translational coupling. It was

therefore possible that orfl 16 and orfl 17 encoded a plasmid stability system, with orfl 17

encoding a toxin and orfl 16 the antitoxin. This hypothesis was tested in Chapter 4.

Organizationally, pXcB is similar to the P. putida IncP-9 plasmid pWWO. The

origin of transfer, oriT, is predicted to be in a 431-nucleotide region on pWWO, between

the opposing genes traA and traD. pXcB has homologs of both traA and traD, and there

are 388 nucleotides in between the two genes. Analogous to pWWO, the origin of

transfer might lie in this area of pXcB. A pairwise BLASTN search of these two regions

shows that they are not similar to one another (results not shown). If pXcB's transfer

origin is in this region, it is not significantly similar to the oriT of pWWO, despite the

overall similarity of the two plasmids.

Several orfs are similar to mobilization genes of other plasmids. Two orfs are

similar to the P. putida and E. coli trwB (orf206) and trwC (orf205) genes. Interestingly,

TrwB is a homolog of the Agrobacterium VirD4, which is required for a functional T4SS

(Vergunst et al., 2000). Other than trwB, pXcB does not have any other virD4 homologs.

Possibly, TrwC could function as the VirD4 homolog for the T4SS on pXcB. Orf208 is

similar to traA, and orfl 18 is similar to traD, both sequential genes on pWWO.

Orf206 is similar to the traB from the P. putida pWWO plasmid, to trwB from the

IncW plasmid R388, and (weakly) to virD4 homologs in X citri pv. citri and X

campestris pv. campestris strain ATCC 33913 (da Silva et al., 2002). It has a cleavable

N-terminal signal sequence, three predicted transmembrane sequences and a P-loop. The

orf206 gene product is predicted to localize to the bacterial inner membrane and likely









capable of hydrolyzing ATP or GTP. The gene product of orf210 is similar to putative

resolvase and predicted to have one transmembrane helix.

Orf217 is similar only to unknown genes. Its gene product is predicted to have 2

transmembrane helices by TMPred and its product is the only one on pXcB that appears

to have an uncleavable N-terminal signal sequence. These results indicate that orf217

may be a cytoplasmic membrane protein. Orf122 is predicted to encode a protein most

likely localized in the periplasmic space (see Table 3-5) and is weakly similar to

gi:42475527, a novel Hrp effector candidate from Pseudomonas syringae pv. pisi.

However, at the time of writing, experimental evidence to support this classification of

holPpiXhas not been published. Orf122 may therefore be of some interest for future

research.

Similarity to Plasmids From X. citri pv. citri and P. putida

Plasmid pXcB is very similar both in DNA sequence and organization to plasmid

pXAC64 from X citri pv. citri (X. axonopodis pv. citri) strain 306 (da Silva et al., 2002).

An overview of the BLASTX results of pXcB against X citri pv. citri strain 306, using an

Expect (number of matches expected to be found merely by chance) threshold of 0.001, is

given in Figure 3-7. Twenty three genes on pXcB show similarity to pXAC64, and three

genes are similar to genes on pXAC33 (gi:38201775), another native plasmid in X citri

pv. citri. One orf (orf213) is similar to a gene from X campestris pv. campestris ATCC

33913 the causal agent of black rot (da Silva et al., 2002).









Color Key for Alignnent Scores
DB^^H 5--J-80


pXcB, I I
0 5K 1K 15K 20K 25K 30K 35K

m m-- m mm m a m




Figure 3-7. Results of a BLASTX search limited to X axonopodis pv. citri, with an
Expect threshold of 0.001


Color Ket for Alimnnent Scores
50-90


70 7250 7500 7750 800 8250
BLAST

HyOhllic)l Htothetical .irB
BLASTX VirB5 Paoei pro ___ei _____ _
Figure 3-8. Results from BLASTN and BLASTX searches of nucleotides 7000 to 8500
of plasmid pXcB, limited to results from X axonopodis pv. citri, with an
Expect threshold of 0.10

Some distinct differences are notable between pXcB and pXAC64. Plasmid

pXAC64 is 64.9 kb in length, and has two members of the avr/pth gene family, while

pXcB is 37.1 kb in length and has one. Both plasmids have a T4SS, with the same

organization (like several other T4SS, see Figure 3-5), including two orfs between the

genes that are similar to the virB5 and virB6 gene from the Agrobacterium virB cluster.

An exception to the similarity of the pXcB virB cluster to the pXAC64 virB cluster exists

in the region of the cluster that encodes orfl05 and orfl06, which are unrelated to any

T4SS genes. This coincides with a slight but consistent rise in the G+C content of the

DNA in this area (see Figure 3-3).

For the nucleotide sequence corresponding to orfs 105 through 107 (-650-bp long),

only a 40-bp region (of which 35-bp are identical) appears to be similar to the


1 [ <4.









corresponding region of pXAC64, whereas the similarity between pXcB and pXAC64 for

the rest of the virB cluster is 93%. The translated predicted proteins from these areas of

the plasmid are similar to each other, even though the nucleotide sequence is not. This is

demonstrated in Figure 3-8, where the upper part of the figure is the result of a BLASTN

(nucleotide query nucleotide database) search, and the lower part the result of a

BLASTX (translate nucleotide query protein database) search for nucleotides 7000 to

8500.

Color Key for Hlignnent Scores
50-80

1-49591 1 1 1 1 1 1 1
0 5K 10K 15K 20K 25K 30K 35K




Figure 3-9. Results of a BLASTX search with plasmid pXcB as query and the results
limited to genes from P. putida with an Expect value of 01.0

Even though BLASTN does not detect any significant similarity in the region

corresponding to orfs 105 and 106, BLASTX shows that the predicted proteins are

similar. This may mean that the region between orfs 104 and 108 in pXcB has evolved

separately from the rest of the virB cluster, resulting in significant changes in nucleotide

changes, but maintaining similarity at the protein level. This would suggest that this

region has evolved at a different rate than the rest of the virB cluster, and that the encoded

proteins have an important function, thus preserving the relative stability of the amino

acid sequence. The five orfs immediately downstream of the last gene in the virB cluster,

orfs 113 through 117, which may be part of the same transcriptional unit on pXcB, are

not similar to any genes on pXAC64 at either the nucleotide level or the amino acid level.









Several orfs on pXcB are similar to genes from pWWO, a native plasmid of P.

putida (Greated et al., 2001). A BLASTX (translated nucleotide query protein database)

search with results limited to P. putida illustrates this in Figure 3-9. It appears as if some

ancestral stretch of DNA was interspersed by horizontal transfer of genes, interrupting

one coding sequence with stretches of DNA from a different source (solid lines indicate

BLASTX hits to the same gene, connected by dashed lines indicating interrupting

regions).

Another striking resemblance between pXcB and pWWO is the similarity between

orfs 105 and 106 from pXcB and orfs 200 and 199 from pWWO. These are the same orfs

that have a different G+C content than the rest of the virB cluster and that show

significant similarity at the nucleotide level to the corresponding region of pXAC64, but

not at the amino acid level.

Discussion

Since X citri pv. citri, which harbors pXAC64, causes the same disease as X citri

pv. aurantifolii, which harbors pXcB, and both plasmids carry the only functional avr/pth

gene member, it would not have been surprising had pXcB been a simple deletion

derivative of pXAC64. However, while many genes on pXcB are similar to genes on

pXAC64 and pWWO from P. putida, there are differences between the three plasmids

significant enough to conclude that any one of them is not a simple deletion derivative of

another.

The predicted promoter sequences suggest that many if not all orfs of the virB

cluster are transcribed from the same promoter, including five orfs immediately

downstream from the virB cluster that are not similar to any known T4SS genes.









The overall organization of the virB cluster is like several other T4SS, with the

virB7 gene and one additional orf between virB5 and virB6. Although it was initially not

clear, closer inspection of orfl05 resulted in the conclusion that it encodes a VirB7

homolog. The predicted amino acid sequences of the virB7 gene and orfl06 from pXcB,

and XACb0043 and XACb0042 from pXAC64 are similar to each other, regardless of the

fact that their nucleotide sequences are not. This could mean that the orfs encode some

essential functions. The difference in nucleotide sequence seems to suggest differential

evolutionary rates, and possibly roles, for virB7 and orfl06. Alternatively, these two orfs

can have transferred horizontally from another source into pXcB. This later theory is

consistent with the fact that the G+C content for these two orfs is higher than that of the

rest of the virB cluster. It does not explain, however, why these interruptions would be in

the exact same location in pXcB and pXAC64, and why the orfs would have such

different nucleotides sequences while having similar amino acid sequences.

Compared to the rest of pXcB, pthB has a consistently high G+C content, also

implying that it transferred horizontally into pXcB. More evidence that pXcB is a hybrid

of stretches of DNA from many different sources, comes from the results of a BLAST

search on which it appears that one orf is interrupted by several different stretches of

DNA. Consistent with a hypothesis that it was transfer horizontally, pthB seems to have

resulted in the interruption of a partition region, since two orfs (216 and 219), both

similar to partition genes and each other (30% identical, 43% similar), appear on either

side ofpthB. Strangely, the predicted protein encoded by orf219 has a region that is

significantly similar to the conserved domains of the Par family of ATPases (see Table 3-

1), while the protein encoded by orf216 does not. One of the orfs (orf216), may have lost









its function, since it was redundant after the duplication event. It may no longer encode a

functional Par protein.

From the sequence of pXcB it is also clear that two subclones of pXcB discussed in

Chapter 2, plasmids pAB12.2 and pAB 14.1, contained separate parts of the virB cluster.

If indeed some or all of the orfs of the cluster do not have their own promoter, many orfs

would not be transcribed in the subclones. If any or all of the genes of the virB cluster are

required for pathogenicity, none of the subclones would have had the necessary genes to

cause citrus canker. Although the five orfs immediately downstream of the virB cluster,

are not part of known T4SS clusters, they might also be transcribed from a virB cluster

promoter, and even though they are all present in subclone pAB 12.2, they might not have

a promoter to transcribe the genes. To assay whether any of the orfs in the virB cluster are

required for pathogenicity, mutagenesis studies need to be performed.















CHAPTER 4
MARKER INTEGRATION MUTAGENESIS

As shown in Chapter 3, plasmid pXcB is about one-third comprised of what appears to be

a complete type IV secretion (T4S) system (see map in Figure 3-1). T4S systems are found in

gram-negative bacteria, transport a wide variety of substrates, and have been shown to be

required for pathogenicity in some cases (Cascales & Christie, 2003; Yeo & Waksman, 2004).

The prototype for T4S systems is the virB cluster from Agrobacterium tumefaciens which uses

the system to export a strand of DNA known as the T-strand to its plant host (Christie, 1997).

The T-strand integrates into the plant genome and contains coding sequences for genes that

affect plant cell division. Bordetellapertussis uses a T4SS to transport the multi-subunit

pertussis toxin, and some plamids use a T4SS for conjugal transfer (Christie and Vogel, 2000).

Since the initial definition of T4S systems based on the A. tumefaciens T-DNA transport system

(virB), the B. pertussis toxin exporter (ptl) and the transfer system of pKM101 (tra), many

different bacteria have been reported to have a T4SS. Some of these are required for intracellular

survival of pathogens in their host (Sieira et al., 2000). The T4SS encoded on pXcB appears to

be related to both conjugation systems and pathogenicity related T4S systems.










pXcB was integrated using a suicide vector in three different locations. Two orfs

(orfl 16, and orfl 17) were interrupted using marker integration (Figure 4-1) and the phenotypes

of the resulting mutant B-strains were assessed with respect to pathogenicity and self-mobilizing

ability of the mutant pXcB's. Attempts to complement the mutant phenotypes were carried out

using several different subclones of pXcB. The marker-integrated plasmids were also tested for

self-mobilizing ability and for plasmid stability.

Possible orientations
of the lacZ promoter internal fragment











5'-T



Possible orientations
of the lacZ promoter ,,,..,,s _


Figure 4-1. Diagram of marker integration mutagenesis. See text for the explanation











Materials and Methods

Bacterial Strains, Plasmids, and Culture Media

Strains used in this study are listed in Table 2-1 and Table 4-1. Plasmids are listed in Table

2-2 and Table 4-2. Xanthomonas strains were grown on PYGM media at 300C, and E. coli

strains on LB at 370C. Agar was added to a final concentration of 15 g/L for solid media.

Antibiotics were used at the following final concentrations in /g/mL: Ampicillin (Ap), 50 (for

low copy plasmid pUFR047), or 100 (for high copy plasmid pGEM-T Easy); Chloramphenicol

(Cm), 35; Gentamycin (Gm), 3; Kanamycin 10 (for initial selection after interruption of ORFs),

20 (for restreak of gene interruption mutants), and 50 (high copy helper plasmid pRK2013);

Tetracyclin (Tc), 15.

Table 4-1. List of bacterial strains used in this study, in addition to those listed in Chapter 2
Xanthomonas
Strains Relevant characteristics Reference or source
B 16.1 pXcB::pUFR004, marker-integrated immediately This study
upstream of orfl 15 (Sp'Cmr), result of mating
pAB16.1 into B69Sp2
B23.1-11 pXcB::pUFRO12, marker-integrated immediately This study
upstream of orfl 15 (Sp'Cm'Knr), result of mating
pAB23.1 into B69Sp2
B26.4.0 Marker-interrupted mutant of orfl 17 This study
(Sp'Cm'Knr), result of mating pAB26.4 into
B69Sp2
B26.4.1 Marker-interrupted mutant of orfl 17 This study
(Sp'Cm'Knr), result of mating pAB26.4 into
B69Sp2
B26.8.0 Marker-interrupted mutant of orfl 17 This study
(Sp'CmrKn)), result of mating pAB26.8 into
B69Sp2
B26.8.1 Marker-interrupted mutant of orfl 17 This study
(Sp'Cm'Knr), result of mating pAB26.8 into
B69Sp2
B31.2.1 Marker-interrupted mutant of orfl 16 This study
(Sp'Cm'Knr), result of mating pAB31.2 into
B69Sp2











Table 4-2. List of plasmids used in this study, in addition to those listed in Chapter 2


Plasmid


DUFRO12 suicide vector. CmrKnr


pAB 15.1

pAB 16.1

pAB23.1

pAB23.6

pAB25.1

pAB26.4

pAB26.8

pAB29

pAB30.1

pAB31.1

pAB31.2
pAB33.4

pAB34.1

pAB35.3

pAB36.6

pAV7.3
pB16.1
pB23.1-11
pB26.4.0
pB31.2.1

pJR7.1

pJR8.2
pUFR004


Relevant characteristics
PCR amplified (primers AB01 and AB02) internal
fragment of orfl 15 in pGEM-T Easy (Ap')
EcoRI orfl 15 fragment, subcloned from pAB 15.1 in
pUFR004 (Cmr)
EcoRI orfl 15 fragment, subcloned from pAB 15.1 in
pUFR12 (CmrKnr)
EcoRI orfl 15 fragment, subcloned from pAB 15.1 in
pUFR12 (CmrKnr)
PCR amplified (primers AB35 and AB36) internal
fragment of orfl 17 in pGEM-T Easy (Ap')
EcoRI orf 17 fragment, subcloned from pAB25.1 in
pUFR12 (CmrKnr)
EcoRI orfl 17 fragment, subcloned from pAB25.1 in
pUFR12 (CmrKnr)
BglII fragment of pB23.1 (containing pthB through
virB cluster) in pUFR047 (Ap'Gmr)
PCR amplified (primers AB43 and AB44) internal
fragment of orfl 16 in pGEM-T Easy (Ap')
EcoRI orfl 16 fragment, from pAB30.1 in pUFR12
(CmrKnr)
EcoRI orfl 16 fragment, from pAB30.1 in pUFR12
(CmrKnr)
orfs 115-117 cloned in pLAFR3 (Tc')
orfl 16 (PCR product from AB 46 and AB47) in
pGEM-T Easy (Ap')
EcoRI-HindIII orfl 16 fragment, subcloned from
pAB34.1 in pUFR047 (Ap', Gmr)
12.7-kb EcoRI fragment from pXcB cloned in
pUFR047 (Ap'Gmr)

orfl 16-117 cloned in pUFR047 (Ap'Gmr)
Marker-integrated plasmid from B16.1 (Cmr)
Marker-integrated plasmid from B23.1-11 (CmrKnr)
Marker-integrated plasmid from B26.4.0 (CmrKnr)
Marker-integrated plasmid from B31.2.1 (CmrKnr)
14.5-kb fragment from pXcB containing orfl00
through orfl 17 in pUFR047 (Ap'Gmr)
EcoRI subclone of pJR7.1 containing the 3'-end of
virB0O through orfl 17 in pUFR047 (Ap'Gmr)
suicide vector, Cmr


Reference or source

This study

This study

This study

This study

This study

This study

This study

This study

This study

This study

This study
This study

This study

This study

This study
Vanamala & Gabriel,
unpublished
This study
This study
This study
This study

Reddy & Gabriel, unpublished

Reddy & Gabriel, unpublished
De Feyter et al., 1990
De Feyter & Gabriel,
unpublished










General Bacteriological Techniques

Plasmids were transferred between X citri pv. aurantifolii and E. coli strains using

biparental and/or triparental matings. Plasmid pRK2013 was used as the helper strain in all

triparental matings.

Recombinant DNA Techniques

Xanthomonas total DNA was extracted and manipulated as described in Chapter 2.

Plasmid DNA was extracted from both E. coli and Xanthomonas as described in Chapter 2.

Other standard recombinant DNA procedures were used essentially as described in

(Sambrook et al., 1989). Polymerase chain reactions (PCR) were carried out using the Biometra

T-gradient thermocycler. A summary of the primers and the resulting clones and marker-

integrated mutants is given in Table 4-3.

Marker Interruption Mutagenesis

Two orfs on pXcB were interrupted by plasmid integration. For interruption of both ORFs,

an internal fragment was amplified by PCR and cloned into the pGEM-T Easy vector

(Promega). EcoRI fragments containing the internal fragments were then subcloned into either

suicide vector pUFR004 (De Feyter et al., 1990) or pUFR12. Plasmid pUFR12 has an extra 1.2-

kb PstI fragment that contains a gene encoding neomycin phosphotransferase (Knr). The

resulting plasmids were then mated into B69Sp using triparental matings, and selection was

carried out on selective PYGM agar plates. Each interruption resulted in vector integration into

pXcB and a duplication of the internal fragment of the target gene (see Figure 4-1). Primers used

for amplification of an internal fragment of orfl 16 were: AB43 TCGATGAGGCCCTGAAAA,

and AB44 AGCCGGCGACGTGTC, for orfl l7: AB35 ACAAGATCGCGACGACAT, and

AB36 GGCGGGACGTATGGAC.











Table 4-3. Summary of primers and resulting plasmids and Xanthomonas strains
Clone in Vector
pGEM-T subcloned Subclone Marker-integrated Marker-integrated
Primers Easy into name Xanthomonas strain plasmid
AB01+AB02 pAB15.1 pUFR004 pAB16.1 B16.1 pB16.1
pAB23.1, B23.1-11 pB23.1-11,
AB01+AB02 pAB15.1 pUFR012 pAB23.6 B23.6 pB23.6
B31.2.1, pB31.2.1,
AB43+AB44 pAB30.1 pUFR012 pAB31.2 B31.2.3 pB31.2.3
pAB26.4, B26.4.0, pB26.4.0,
AB35+AB36 pAB25.1 pUFR012 pAB26.8 B26.8.0 pB26.8.0
AB46+AB47 pAB34.1 pUFR047 pAB35.3 N/A N/A

A third plasmid integration was performed using primers AB01

TAGGAGTGGAAAGAATGG, and AB02 CAACCGCTCTCAGAG. The primers amplified

orfl 15 truncated at the 5'-end and the integration should therefore not result in orfl 15

interruption. Table 4-3 gives a summary of clones and Xanthomonas strains derived from

different primer combinations. The DNA sequence of orfs 115-117, predicted protein translation

and the position of the primers are given in Appendix A.

Phenotypic Tests

For complementation tests, a 334 bp fragment containing the entire orfl 16, including 19

additional nucleotides upstream of the putative orfl 16 start codon, was PCR amplified from

B69Sp with primers AB46 AAGCTTTCCAGAACAACTCGATC and AB47

GAATTCTAACACATAAGGTGGCGC, and cloned into pGEM-T Easy (pAB34.1). Primers

AB46 and AB47 were designed so that the PCR product has an EcoRI site at the 5'-end and an

HindIII site at the 3'- end for directional cloning into pUFR047, and there are translational stop

codons in all three reading frames at the 5'-end before the start of the coding sequence of orfl 16

(see Figure 4-2). The EcoRI-HindIII fragment of pAB34.1 was subcloned into pUFR047











(pAB35.3). Plasmid pAB35.3 was mated into B23.1-11, B26.4.0, and B31.2.1, and inoculated

on both lime and grapefruit.


EcoRI
I I
5'-T G AA T T C TAAIC A C AT A AG G T G G C G C-3'
stopcodon stopcodon stopcodon
frame 1 frame 2 frame 3
*Is not actually part of the primer,
but exists in the vector

Figure 4-2. Primer AB47. The primer uses a "T" that occurs in the vector sequence to form a
stopcodon in one reading frame, while two "TAA" sequences in the vector supply
stopcodons in the other two reading frames

Plasmids pAV7.3, pJR7.1, and pJR8.2 were mated in B23.1-11, B26.4.0, and B31.2.1 to

attempt complementation. Plasmid pAB36.6 was constructed by reversing the insert of pAB12.2,

so that the lacZ promoter would run in the same direction as orfl 13-117. Both of these plasmids

were also mated into B23.1-11, B26.4.0, and B31.2.1. Plasmid pAB29 was constructed by

cloning the BglI fragment of pB16.1 containing all orfs from pthB through orfl 15 into the

BamHI site of pUFR047. Plasmid pAB29 was mated into B69.4 (cured of pXcB) and inoculated

to test for pathogenicity. To test whether the lacZ promoter is functional in B69 strains, pAB 18.1

containing pthA driven by the lacZ promoter, was mated into BIM2 (pthB::pUFR004) and

inoculated.

Self-transmissibility of pBIM2 and pBIM6 was tested. Since X citri pv. citri has a second

functional T4SS on another plasmid than pXcB (El-Yacoubi & Gabriel, unpublished results),

which can trans-complement a virB4 mutant of pXcB, it was necessary to first transfer the

mutant plasmids to E. coli DH5a and then demonstrate self-transmissibility to E. coli HB 101.

Matings were performed with and without helper to test for the ability of the marker-integrated

plasmids to transfer from one E. coli strain to another without the presence of the helper strain.












total DNA





1 1z 1I Is a
:22


s a s i ai !!jsjsS
1~ I i s^I
E
us
Id
C
id
cc M Zz
B M E E E m

.h E I ,U I~
9n u u u o CL CV
*Ug l E Eo E E I
E dAd
ri ri ri dj r ri Li Li dj r
0 0x x~~x


total DNA

i

E Z
S~ E E

L 1 z I I

> 11>11


Figure 4-3. Garden Blot. A) Ethidium bromide stained gel containing total DNA from different
xanthomonads, and plasmid DNA from B69 and B69.4. B-D) Southern blot from gel
in A, hybridized with a probe derived from B) orfl 15, C) orfl 16, and D) orfl 17

To test plasmid stability of marker-integrated plasmids integrated, integrated strains were

cultured by inoculating 1 kL of an overnight liquid starter culture (in liquid PYGM+Sp35+Kn20)

into 10 mL liquid PYGM with Sp35 (all strains used are Spr) and with Sp35Kn20 (selection for


klasmid
DNA



mm m

0 C M
X N Xt










the plasmid). After 4 days (approximately 20 generations) a dilution series of the cultures was

plated on PYGM plates with (Sp35+Cm35, Sp35Kn20) and without (Sp35) antibiotic selection.

orf116 orf117
orf115 mutants mutants
= mutant

2L CV7 C C3 C. k
(OrC (D Y (V O
D mmmmmm CM
23.1 kb
9.6 kb
6.6 kb--


4.4 kb-

2.3 kb-
2.0 kb-


Probe:
orf'115







Probe:
orf 16


Probe:
e orfl17


Figure 4-4. Total DNA digested with EcoRI, Southern blot hybridized consecutively with
probes derived from orfs 115, 116, and orfl 17










Plant Inoculations

Inoculations were done as described in Chapter 2. All inoculations were repeated at least

three times. Inoculated leaves were examined on 7 or 8 days post inoculation on Mexican Lime,

and on 9 or 10 days post inoculation on Duncan Grapefruit.

Plasmid Stability in plant

Two-day old liquid bacterial cultures were washed and then adjusted to an OD of 0.4 with

CaCO3-saturated water, and inoculated on Key Lime and Duncan Grapefruit. Inoculation patches

were removed and ground out immediately after inoculation (half of the leaves), and 8 days post

inoculation (dpi) for the second half of the leaves. For grinding out, three leaf punches (each

from a separate leaf inoculated with the same culture) were obtained with a number 3 cork borer

(0.84 cm2 total area), and ground in 1 mL CaCO3-saturated water with 1:5000 (v/v) Silwet).

Dilution series were spotted (10 ptL per spot) on non-selective (PYGM+Sp35) and non-selective

(PYGM+Sp35+Kn20) plates, and counted after 4 days incubation at 300C.

Results

Southern Blot Hybridization with orfs115-117

Total DNA from various xanthomonads was hybridized with probes made from orfl 15,

orfl 16, and orfl 17. In agreement with the fact that strain B69.4 is cured of pXcB, B69.4 DNA

does not hybridize to orfs 115-117. As shown in Figure 4-3, in addition to one copy of each of the

orfs 115-117 on plasmid pXcB, faint bands hybridizing to orfl 15 and orfl 17 are present in X

phaseoli var. fuscans. The hybridizing EcoRI fragments are of approximately the same size as

the fragment on pXcB, but are much weaker, suggesting they might be located on the

chromosome. A small EcoRI fragment in X.c. phaseoli also hybridizes to orfl 17.











EcoRI Hindlll EcoRI

dopUFR004
pUFRO12


downstream


Hindlll EcoRI
I I


Forward orientation:

pB23.1-11
pB31.2.1


Hindlll
26541


Reverse orientation:

pB16.1
pB26.4.0

Hindlll
26541


Figure 4-5. Orientations for the integration of the suicide vector in pXcB. Not drawn to scale








23.1 kb
9.4 kb
6.6 kb
4.4 kb

2.3 kb
2.0 kb





Figure 4-6. Mapping of the lacZ promoter orientation in pB16.1, pB23.1-11, pB26.4.0, and
pB31.2.1


Hindlll
8940


genes


EcoRI
I DlecZ


I pUFR0041
pUFRO12


Hindlll
8940


downstream
genes


I










Marker Integrations and Mapping

Orfl 16 and orfl 17 were interrupted using pUFR012. A third marker integration was

carried out, resulting in pUFR004 and pUFR012 integration downstream of orfl 15, followed by

a duplicate but incomplete copy of orfl 15 (lacking the 5'-end of orfl 15). Total DNA was

isolated from all integration derivatives, digested with EcoRI, and after electrophoresis

hybridized consecutively with probes derived from orfl 15, 116, and 117 (see Figure 4-4). In B69

Sp2, orfl 15 through 117 are all located on the same 12.6-kb EcoRI fragment. B69.4 is cured of

pXcB and does not hybridize to any one of the orfs used as a probe. In B 16.1 two fragments

(8811 bp and 4293 bp) hybridize to orfl 15, while orfl 16 and orfl 17 are located on the same

8811 bp EcoRI fragments. B31.2.1 and B31.2.3, both marker integration of orfl 16, each have

two fragments (8253 bp and 4679 bp) that hybridize to orfl 16, orfl 15 is located on the 4679-bp

fragment, while orfl 17 is located on the 8253-bp fragment. B26.4.0 and B26.8.0, marker

integration of orfl 17, have a 7952-bp fragment that hybridizes to orfl 17, and a 4953-bp

fragment that hybridizes to orfl 15, orfl 16, and orfl 17.

The introduction of a HindIII site with the suicide vector allowed the direction of the lacZ

promoter to be determined (Figure 4-5). The orientation in which the suicide vector integrated

into pXcB was determined, by transferring the integrated plasmids to E. coli DH5a (MCR), and

restriction digestion with HindIII (Figure 4-6).

Pathogenicity Tests

Plasmid pAB29 was constructed from pB16.1 (Figure 4-7). ABglII restriction site from

the integrated suicide vector pUFR004 and a BglII site just upstream of gene pthB were used to

clone a fragment of pXcB, including all orfs starting frompthB through orfl 15 into the BamHI











site of pUFR047. This clone was mated into the cured B-strain, B69.4, and the resulting strain

was used to inoculate lime and grapefruit leaves. No canker phenotype was seen (Figure 4-8).


EcoRI


EcoRI EcoRI
I.. I


Figure 4-7. Construction of pAB29, containingpthB through orfl 15 in pUFR047






















Figure 4-8. Inoculation result of plasmid pAB29 mated into B69.4 and inoculated on Duncan
Grapefruit


















B21.21
pAB8l







Figure 4-9. The lacZ promoter is functional in the B-strain. Strain BIM2, a marker integrated
mutant ofpthB, and B21.2, a marker exchange mutant ofpthA are complemented by
pAB 18.1, a clone ofpthA driven by the lacZ promoter in pUFR047

Strain B 16.1 and B23.1-11, which contain an intact and a partial copy of orfl 15, were both

non-pathogenic on lime and grapefruit. This result is surprising, since all genes are left intact. In

the case of B 16.1, the lacZ promoter runs in the reverse direction and there could be a polar

effect of the insertion on orfl 16 and orfl 17 if they are driven off the same promoter as orfl 15. If

orfl 16 and orfl 17 are driven off a separate promoter (as predicted, if orfl 16 and orfl 17 encode a

toxin-antitoxin system), the marker integration should have no effect on the transcription of

orfl 16 and orfl 17. In the case of B23.1-11, the lacZ promoter runs in the forward direction, and

should therefore have no effect on phenotype at all.

B31.2.1 and B31.2.3 resulted from marker integration of orfl 16. The orientation of

B31.2.1 was determined, and lacZ runs in the forward direction. Since the lacZ promoter is

functional in the B69 derivatives (pthA driven by the lacZ promoter complements BIM2, apthB-

marker integrated mutant; see Figure 4-9) If the lacZ promoter runs in the forward direction, the

insertions should not have a polar effect on orfl 17.










The marker-integrated mutants of orfl 17 gave inconsistent results upon inoculation.

Initially two colonies that resulted from marker integration of pAB26.4 (B26.4.0 and B26.4.1)

and two colonies that resulted from marker integration of pAB26.8 (B26.8.0 and B26.8.1) gave

canker on lime (Figure 4-10). Subsequent inoculations most often did not result in canker, but

occasionally a weak or fairly strong canker phenotype could be seen. This result could indicate

that orfl 17 is required under certain (at this time unknown) conditions.





B26.8.1 5 ~ B69 Sp2




B26. B26.4.0




B26]






Figure 4-10. Results of initial inoculation of orfl 17 marker integration

Clones used for attempts at complementing the mutant phenotypes are shown in Figure 4-

11. Plasmid pAB33.1 contains orfl 15 through orfl 17. Plasmid pAB35.3 was cloned using

primers AB46 and AB47 to place orfl 16 under control of the lacZ promoter in pUFR047,

pAV7.3 (Vanamala & Gabriel, unpublished) contains orfl 16 and orfl 17 under control of of the

lacZ promoter. Plasmid pAB36.6 was cloned by reversing the insert from pAB12.2 so that the

lacZ promoter runs in the same direction as orfsl 15 through 117. Plasmid pJR7.1 (Reddy &











Gabriel, unpublished) was cloned from pXcB to contain the entire virB cluster, orfsl00, and

113 through 117, and includes the putative promoter region in front of orfl00.


1 kt
115 116 117

orf00-orfl \ phB

pAB33.1
pAB35.3
pAV7.3
pAB36.6
pJR7.1
pJR8.2
Figure 4-11. Clones used for attempts to complement maker-integrated mutants and the regions
of plasmid pXcB they represent




116::pUF 69 Sp2 116::pUFRI 9 Sp2
pAB36.6 pAB36.6

16:: 6::
117::pUF FRi2 117::pUFR1 FR12
pAB36.6 pAB36.6
17:: 17::
UFR12 UFR12

115::pUF 115::pUFR4
pAB36.6 15:: pAB36.6 15::
pUFR4 pUFR4




Figure 4-12. Attempts to complement marker-integrated mutants with plasmid pAB36.6

However, repeated attempt to mate pJR7.1 into the marker-integrated B-strains, and hold

selection for both the integrated plasmid and pJR7.1 resulted in no or very few colonies. Clone

pJR8.2 was derived from pJR7.1 by deletion of the EcoRI fragment at the 5'- end of the clone, so

that pJR8.2 has the 3'-end ofvirB 10 through orfl 17. Contrary to pJR7.1, pJR8.2 could be mated










efficiently into all the marker-integrated strains. None of the above-mentioned clones

complemented any of the marker-integrated mutants tested. Results for attempts to complement

with clones pAB36.6 and pJR8.2 are shown in Figure 4-12 and Figure 4-13, respectively.




B69 Sp2
SB69 Sp2
B26.4.01
pJR8.2 B26.4.01
26.4.0 pJR8.2 B26.4.0


p8 31.2.1 B31.2.1 B31.2.1
pJR8.2

B23.1-1
B23.1-1 .1-11 pJR8.2 B23.1-11
pJR8.2




Figure 4-13. Attempts to complement marker-integrated mutants with plasmid pJR8.2

Plasmid Stability Tests in vitro

The mutant phenotype of the orfl 16 and orfl 17 marker-integrated mutants could be the

result of orfl 16 and orfl 17 being pathogenicity factors. An alternative hypothesis is that orfl 16

and orfl 17 form a toxin-antitoxin system and are responsible for pXcB stability. This was tested

in vitro by growing the marker-integrated bacterial strains in liquid medium with and without

selection for the marker-integrated plasmid for 4 days (approximately 20 generations). Dilution

series of each of the cultures were then plated on medium with (Sp35Kn20) and without (Sp35)

antibiotic selection for the marker-integrated plasmid. The data are given in Table 4-4. The

number of cfu's on plates with antibiotic selection for the marker-integrated plasmid (Cm35 or











Kn20) was not less than the number of cfu's without selection for the plasmids. Therefore the

marker-integrated plasmids appear to be stable in vitro.

Table 4-4. Stability tests for marker-integrated plasmids in vitro, grown in liquid culture with
and without antibiotic selection for the marker-integrated plasmid
Strain Plated on Grown on Sp35Kn20 Grown on Sp35
Sp35 6x10-8
B69Sp2 Sp35Cm35 0
Sp35Kn20 0
Sp35 1x10-9 2x10-9
B23.1-11 Sp35Cm35 2x10-9 2x10-9
Sp35Kn20 2x10-9 2x10-9
Sp35 3x10-9 1x10-9
B31.2.1 Sp35Cm35 2x10-9 2x10-9
Sp35Kn20 2x10-9 2x10-9
Sp35 6x10'8 6x10-8
B26.4.0 Sp35Cm35 2x10-9 2x10-9
Sp35Kn20 2x10-9 1x-10-9
Table 4-5. Stability tests for mutant plasmids in vitro, alone, and together with vector pUFR047
or complementation clone plasmidd pAB36.6). The marker-integrated plasmids ar
Cm35 and Kn20 resistant, while the empty vector and the complementation clone are
Gm3 resistant
Number of cfu's Number of cfu's
Strain Selection for: w/pUFR047 w/pAB36.6
2x10-7 6x10-7
Kn20 0 0
B69Sp2 Gm3 2x10- 4x10-
Kn20Gm3 0 0
3x10-7 3x10-6
Kn20 3x10-7 7x10-6
B23.1-11
2-11 Gm3 8x10-7 1x10-6
Kn20Gm3 7x 106 3x 107
3x10-6 3x10-7
Kn20 6x10-6 4x10-6
B31.2.1
B31.2.Gm3 1x10-7 5x10-7
Kn20Gm3 3x 106 5x10-6
6x10-6 1x10-7
Kn20 7x10-6 2x10-7
B26.4.0
B260 Gm3 4x 10-7 8x10-7
Kn20Gm3 6x 10-6 1x10-7

Stability of the marker-integrated plasmids and complementation plasmid when placed

together in the same strain was then tested. The possibility exists that either one of the plasmids

by itself will be stably maintained, but that an incompatibility between the two plasmids causes










one or both to become unstable. The experiment was carried out with strains B23.1-11,

B26.4.0, and B31.2.1. Complementation plasmid pAB36.6 was mated into each strain, and

restreaked on selective plates (PYGM+Kn20+Gm3). As a control, the empty vector pUFR047

was also mated into each strain. The results are given in Table 4-5. The number of cfu's on

selective plates was no less than the number of colonies on non-selective plates. This means that

virtually all of the strains maintained both plasmids separately, and together over -20

generations. No loss of either one of the plasmids, or the plasmids together was observed. This

demonstrated that in vitro, both the mutant plasmids and complementation plasmid pAB36.6

were stably maintained in B69.

Stability Tests in plant

To test plasmid stability in plant, bacterial cultures were inoculated and ground out

immediately and 8 dpi. Dilution series were then plated on selective and non-selective plates.

Results are summarized in Table 4-6. No significant loss of the marker-integrated plasmids was

observed. The marker-integrated plasmids appear to be stable inplanta.

Table 4-6. Stability tests for marker-integrated plasmids in plant
Strain Selection for: Number of cfu's 0 dpi Number ofcfu's 8 dpi
Sp35 1x10-6 5x 107
B69Sp2 Sp35Cm35 2x10-4 5x 10-3
Sp35Kn20 2x10-4 2x10-3
Sp35 5x 10-6 1x107
B23.1-11 Sp35Cm35 5x10-6 3x10-6
Sp35Kn20 4x10-6 1x10-6
Sp35 1x10-6 4x10-6
B31.2.1 Sp35Cm35 2x10-6 2x10-5
Sp35Kn20 2x10-6 1 x10-
Sp35 1x105 2x10-5
B26.4.0 Sp35Cm35 2x105 6x10-5
Sp35Kn20 1x105 2x10-5










Discussion

Three marker-interrupted mutants each were made of orfl 16 and orfl 17. The third marker-

integrated mutant was predicted to be fully pathogenic, since orfl 15, the intended target was left

intact. If the lacZ promoter is in the forward direction, it is possible that it can drive expression

of downstream genes, making the mutation non-polar. If the lacZ promoter is in the reverse

direction, the mutation would be a polar mutation. The suicide vector was integrated in two

different orientations in two independent marker-integrated mutants, and since the lacZ promoter

is functional in X citri pv. aurantifolii, it should be able to drive the orfs downstream of orfl 15 if

it runs in the forward orientation. There is the possibility that the lacZ is unable to drive orfl 16

and orfl 17, if those genes are expressed from there own (possibly tightly regulated) promoter.

The marker-integrated mutant of orfl 17 was fully pathogenic when first inoculated. During

subsequent inoculations the phenotype was inconsistent. It appears that orfl 17 is not required for

pathogenicity of B69. One explanation for the mutant phenotypes of each of the marker-

integrated mutants is that spontaneous mutations elsewhere on the plasmid or in the genome have

occurred which rendered the strain non-pathogenic. There were no obvious deviations from the

expected restriction patterns and hybridization patterns for either of the mutants.

All attempts to complement the mutant phenotypes were unsuccessful, which supports the

hypothesis that additional mutations have occurred. It seems unlikely that such spontaneous

mutations would have occurred in all of the independent marker-integrated mutants. It would

have been prudent to generate non-polar mutations from the start, instead of relying on possibly

polar mutations using marker integration. An entire suicide vector is integrated into the native

plasmid, and it is possible that unforeseen interactions take place.










Plasmid pAB29, contains pthB and the entire virB cluster with its own promoter,

through orfl 15. When this clone was mated into B69.4 (cured of pXcB), the resulting strain

transconjugant did not cause canker. It is therefore possible that the necessary factors) for

pathogenicity are located between orfl 15 and pthB.

Since orfl 16 and orfl 17 resemble reported plasmid stability systems, the hypothesis that

the marker-integrated plasmids were unstable in the X citri strains was tested. It was shown that

the marker-integrated plasmids are stable in the mutant strains in culture and in plant. The

possibility that the clones used for complementation were incompatible with the marker-

integrated plasmids was also tested in culture. No instability of either the mutant plasmid or the

complementation plasmids, or of the two plasmids together, was seen in culture. This experiment

was not done in plant, and might have been worthwhile, since the instability/incompatibility

may only be an issue under stress (low nutrition) conditions, as likely to be encountered in the

plant.














CHAPTER 5
CONCLUSIONS

Previous experiments have shown thatpthB, located on native plasmid pXcB in X

citri, is necessary but insufficient to cause canker on citrus. Subclones of pBIM2, a

marker-integrated mutant of pXcB, were used together with pthA in a different shuttle

vector in attempts to complement the mutant phenotype. The clones were mated into

B69.4/pAB2.1, the strain cured of pXcB withpthA added back into it. None of the

resulting strains caused canker on either Key Lime or Duncan Grapefruit. The factors)

necessary for pathogenicity are most likely interrupted in the subclones, or there is more

than one factor required, and the two were cloned into separate fragments.

The complete sequence of pXcB revealed an intact T4SS, similar to the T4SS

found on pXAC64, the native plasmid ofX. citri pv. citri (syn. X axonopodis pv. citri).

T4S systems are responsible for bacterial conjugation, and experiments have shown that

the T4SS of pXcB is required for self-mobilization (El Yacoubi and Gabriel, manuscript

in preparation). The T4SS on pXcB and pXAC64 differ slightly from that of A.

tumefaciens. On pXcB and pXAC64, the virB1 homologue is at the end of the virB

cluster instead of at the beginning, and the order of the virB6 and virB7 genes is not the

same. In addition, pXcB and pXAC64 have one additional orf between virB6 and virB7.

Remarkably, although the nucleotide for the rest of the virB cluster of pXcB shows very

high similarity to that of the pXAC64 virB cluster, the region between the virB5 and the

virB6 gene shows a much lower degree of similarity. This does not hold true for the









similarity at the translated level. This implies that this part of the sequence has evolved at

a different rate than the rest of the virB cluster.

Although many similarities exist between pXAC64 and pXcB, there are some

differences. pXAC has two members of the avrBs3/pthA gene family, including pthA,

whereas pXcB has only one, pthB. Plasmid pXcB has five additional orfs immediately

downstream of the virB cluster that could be transcribed from the same promoter, but

pXAC64 has no homologues of any of these genes. These orfs may therefore be suited

for identification purposes. A garden blot using different xanthomonads shows that

except for a weak band in X phaseoli var. fuscans, orfl 15 is unique to the B strain, and

orfl 16 seems to be completely unique.

Two marker-interruption mutants post pathogenicity on Key Lime and Duncan

Grapefruit, but this phenotype could not be complemented using the complementation

clones designed for this purpose. A third attempt at marker interruption, resulting in

integration but no gene disruption, also lost pathogenicity on citrus. No instability of the

marker-interrupted plasmids or the complementation plasmid was observed in liquid

culture and in plant. In addition, no incompatibility between the marker-interrupted

plasmids and the complementation plasmid was evident in liquid culture.




















APPENDIX A
MAP OF POSITION 13301 TO 15700


Orfs are hi hlighted as follows: Primers used for marker integration:
orfl 13: not done
orfl 14:yellow (frame a) not done
orfl 15:green (frame b) ABO1 and AB02
orfll6:blue frame a AB43 and AB44
orfl 17: AB35 and AB36


TTCGTGCATTGACCGGCTTTTCTCGCATGCGAGAAAGCTAACTCAGGAGAAGAAGGCATG
13301 ---------+---------+---------+---------+------------------+ 13360
a F V H *
b
c R A L T G F S R M R E S *

AAAAAACTGATAGTCGCGCTGTTTATCGGTGCTGCCT GGTTCCGCOGCACGCCTCCGAA
13361 ---------------------------+---------+ 13420
a
b
c

TATGGGTGCAAGGTGCTGCTGTGCCTTGCCAATCCCGCGTCCAATGGCGGCCCGAAGGGC
13421 ---------+---------+---------+---------+------------------+ 13480
a
b M G A R C C C A L P I P R P M A A R R A
c

GTGTCCGAATGCGTTCCCCCCATCGATCAGCTCTACCACGATCTCAGCAAGGGGCGGCCG
13481 ---------+---------+---------+---------+------------------+ 13540

a
b C P N A F P P S I S S T T I S A R G G R -
c M R S P H R S A L P R S Q Q G A A V-

TTCCCGACGTGCGATCTCGCGGACGGCAATGATGGTTCGAGCTACGCCCGGCAGGTCTAT
13541 ---------+---------+---------+---------+------------------+ 13600
a
b S R R A I S R T A M M V R A T P G R S M
c P D V R S R G R Q *

GACCCGTATGACCCTTGCCCGGCCCCGTTGCAACCTGCCGCACGCGGTTCCTATGTCGTG
13601 ---------------------------+---------+ 13660
a
b T R M T L A R P R C N L P H A V P M S C -
C














13661


13721


13781


13841


13901


CAAGGGCAGAAGAAAACGGGCGGCAACAAGCCGGGATGGTGGGGCGGCGACGGCTCGTAC
S------------------------------------------------ 13720


K G R R K R A A T S R D G G A A T A R T -
M V G R R R L V H-

ACGCTGAGCGGACAGCCGCAAGTGTCTCAGTCGCAAAGCGACTACGGCCATAGCTCGGGT
S---------------------------+---------+---------+ 13780

R *
A E R T A A S V S V A K R L R P *

GCGCGGGCCTGCGTCGGGAAGTCGGTCGGCTCATACACCGTAGGCAGCTACGACAGCAGC
S---------------------------+---------+---------+ 13840





GACACCGTGGACGTGTTCGACAAGGTGGTCTGGCAGCCGGCGCAGAATCCGCGAGCCATC
S---------------------------+---------+---------+ 13900





GACGTGTTCATCGACAACACGTGGCAGCAGCGCGTGCGCTGGTAAGCGGGGGTGGCAATG
S---------------------------+---------+---------+ 13960
M


CGTCGATCTATCGTCTGCGCGGCTCTGTTGGCCCTCGCCTCTCTGGCGGGGCTCAACAGC
13961 ------------------+--------+---- ----- ---- + 14020
R R S I V C A A L L A L A S L A G L N S




TTCACGGTCGGCCTGTTGGACAAGGTTCGCAACACGGTGGCCGCCGAGCCGGCCAGCGCC
14021 ---------+---------+---------+---------+------------------+ 14080
F T V G L L D K V R N T V A A E P A S A




CCGGACACGCAGACTGTCGAGGTCGCTTTCTCGCCGGACGGTCGGGCCGAAGCGTTGGTG
14081 ------------------+--------+---- ----- ---- + 14140
P D T Q T V E V A F S P D G R A E A L V




CTCAAGGTCATTCGTGCCGCGAAAACGTCAATCCGCTTGGCCGGCTACACCTTCACGTCG
14141 ------------------+--------+---- ----- ---- + 14200
L K V I R A A K T S I R L A G Y T F T S




CCGGCCGTCGTGCGCGCCCTGACCGATGCCAAGAAGCGCGGTGTCGATGTGGCTGTCGTG
14201 ---------+---------+---------+---------+------------------+ 14260
P A V V R A L T D A K K R G V D V A V V
M P R S A V S M W L S W -




Full Text

PAGE 1

IN SEARCH OF PATHOGENICITY FACTORS OF Xanthomonas citri pv. aurantifolii By ASHA MARCELLE BRUNINGS A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by ASHA MARCELLE BRUNINGS

PAGE 3

iii ACKNOWLEDGMENTS Many debts have been incurred during the project described on the following pages. The Organization of American Stat es (OAS) provided funding during the first 2 academic years. I especially thank my superv isory committee chair Dean Gabriel, and all of my current and former committee memb ers Alice Harmon, Bill Gurley, Ken Cline, John Davis, James Preston III, and Tom Mareci have all been very helpful. I owe my colleagues in the lab Adriana Casta eda, Abdulwahid Al-Saadi, and Basma El Yacoubi for their support and livel y discussions. Many thanks go to Yong Ping Duan, Anjaiah Vanamala, and Joseph Re ddy, postdoctoral associates in the lab. I owe more than I can say on these pages to Gary Marlow. He has gone above and beyond what could reasonably be expected from a lab technician. Gary has been a constant source of support over many year s, both inside and outside the lab. Pant Pathology Department chair Gail Wi sler, has been an enormous source of encouragement over the years; as were Laure tta Rahmes, Gail Harris, and Donna Perry of the Pathology office. I thank the Plant Mo lecular and Cellular Biology Program, and support staff, especially Melissa Webb, who has been very supportive. My life has been enriched by the friendships of many people I met while pursuing my degree. In this regard, a special note of thanks goes out to Jorge Vazquez, Erica Gubrium, Doug Knudsen, Joyce Merritt, Sha yna Sutherland, Karen Chamusco, Melanie Cash, and Mukesh Jain.

PAGE 4

iv Immeasurable is the support of my famil y. My parents Ernie and Mireille Brunings, and the memory of my late brother, Raoul, ha ve been an inspirati on. My sons, Raoul and Marlow; and my husband, James Meier, have be en especially patient with me, both while the research was ongoing and during the writing of this thesis.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES...........................................................................................................viii LIST OF FIGURES...........................................................................................................ix ABSTRACT....................................................................................................................... xi CHAPTER 1 LITERATURE REVIEW.............................................................................................1 Citrus Canker................................................................................................................1 Disease Symptoms.................................................................................................1 Host Range............................................................................................................2 Control Measures...................................................................................................2 The Pathogen.........................................................................................................3 Pathogenicity Factors....................................................................................................5 Gene pthA and its Predicted Product.....................................................................5 Quorum Se nsing....................................................................................................8 Nutrition................................................................................................................8 Attachment............................................................................................................9 Type II (General) Secretion (T2S)......................................................................11 Type III Secretion (T3S).....................................................................................12 Type IV Secretion (T4S).....................................................................................15 Plasmids pXAC33 and pXAC64.........................................................................18 Plasmid Toxin-Antitoxin Modules.............................................................................19 TA Mechanisms...................................................................................................20 TA Operon Organization.....................................................................................21 Testing Plasmid Stability.....................................................................................22 2 A NOVEL PATHOGENICITY FACTOR REQUIRED FOR CITRUS CANKER DISEASE IS CONTAINED ON A SELF-TRANSMISSIBLE PLASMID...............23 Material and Methods.................................................................................................24 Bacterial Strains, Plasmids, and Recombinant DNA Techniques.......................24 Plant Inoculations................................................................................................27 Results........................................................................................................................ .29

PAGE 6

vi B69 Has Multiple Plasmids and a T3SS..............................................................29 Gene pthB is Insufficient to Cause Canker..........................................................29 Mapping of pBIM2 and pBIM6..........................................................................29 Constructs............................................................................................................30 Pathogenicity Tests..............................................................................................31 Self-transmissibility.............................................................................................33 Discussion...................................................................................................................33 3 COMPLETE PLASMID SEQUENCE.......................................................................35 Material and Methods.................................................................................................35 Bacterial Strains, Plasmids, and Culture Media..................................................35 General Bacteriological Techniques....................................................................35 Recombinant DNA Techniques...........................................................................36 DNA Sequence Analysis.....................................................................................36 Results........................................................................................................................ .37 Features of Plasmid pXcB...................................................................................37 Promoters.............................................................................................................39 Open Reading Frames.........................................................................................47 The pXcB virB Cluster........................................................................................49 Similarity to Plasmids From X. citri pv. citri and P. putida ................................58 Discussion...................................................................................................................61 4 MARKER INTEGRATION MUTAGENESIS..........................................................64 Materials and Methods...............................................................................................66 Bacterial Strains, Plasmids, and Culture Media..................................................66 General Bacteriological Techniques....................................................................68 Recombinant DNA Techniques...........................................................................68 Marker Interruption Mutagenesis........................................................................68 Phenotypic Tests..................................................................................................69 Plant Inoculations................................................................................................73 Plasmid Stability in planta ..................................................................................73 Results........................................................................................................................ .73 Southern Blot Hybridization with orfs115-117...................................................73 Marker Integrations and Mapping.......................................................................75 Pathogenicity Tests..............................................................................................75 Plasmid Stability Tests in vitro ............................................................................80 Stability Tests in planta .......................................................................................82 Discussion...................................................................................................................83 5 CONCLUSIONS........................................................................................................85 APPENDIX A MAP OF POSITION 13301 TO 15700......................................................................87

PAGE 7

vii B CODONPREFERENCE FOR POSITION 13301 TO 16000.....................................92 LIST OF REFERENCES...................................................................................................94 BIOGRAPHICAL SKETCH...........................................................................................108

PAGE 8

viii LIST OF TABLES Table page 1-1 Relative pathogenicity of all known X. citri strain groups on four citrus species......4 2-1 List of bacterial strains used in this study................................................................24 2-2 List of plasmids used in this study...........................................................................26 3-1 Open reading frames identified in plasmid pXcB....................................................40 3-2 Predicted promoters for the + strand of plasmid pXcB with a score of 0.85 or higher........................................................................................................................4 7 3-3 Predicted promoters for the strand of plasmid pXcB with a score of 0.85 or higher........................................................................................................................4 8 3-4 Transcriptional terminators iden tified in plasmid pXcB by GeSTer.......................49 3-5 Presence of N-terminal signal sequen ces, localization of putative proteins (PSORT), and transmembrane helices (TmPred).....................................................50 3-6 Comparison of the pXcB virB cluster with several type IV secretion-systems.......51 4-1 List of bacterial strains used in this study, in addition to those listed in Chapter 2.66 4-2 List of plasmids used in this study, in addition to those listed in Chapter 2............67 4-3 Summary of primers and resulting plasmids and Xanthomonas strains...................69 4-4 Stability tests for ma rker-integrated plasmids in vitro grown in liquid culture with and without antibiotic se lection for the marker-integrated plasmid.........................81 4-5 Stability tests for mutant plasmids in vitro alone, and together with vector pUFR047 or complementation clone (plasmid pAB36.6).......................................81 4-6 Stability tests for ma rker-integrated plasmids in planta ..........................................82

PAGE 9

ix LIST OF FIGURES Figure page 1-1 Maps of plasmids from Xanthomonas citri pv. citri Type A strain 306.....................7 1-2 Pathway of the T-DNA strand through the Agrobacterium tumefaciens T4SS.......17 2-1 Members of the avr/pthA gene family in X. citri pv. aurantifolii.............................28 2-2 Restriction map of plasmid pBIM2 and its subclones..............................................30 2-3 Restriction map of plasmid pBIM6..........................................................................30 2-4 Restriction digests of pl asmid pBIM2 for mapping purposes..................................31 2-5 Plasmid pBIM2 and subclones.................................................................................32 3-1 Map of plasmid pXcB..............................................................................................38 3-2 Codonpreference of plasmid pXcB reverse frames from position 20,000 to 21,000.46 3-3 G+C content of plasmid pXcB.................................................................................46 3-4 Potential promoters on plasmid pXcB.....................................................................48 3-5 Comparison of the linear organization of several type IV secretion-systems..........51 3-6 Phylogenetic trees compiled by CLUSTALW.........................................................52 3-7 Results of a BLASTX search limited to X. axonopodis pv. citri..............................59 3-8 Results from BLASTN and BLASTX s earches of nucleotides 7000 to 8500 of plasmid pXcB...........................................................................................................59 3-9 Results of a BLASTX search with plasmi d pXcB as query and the results limited to genes from P. putida ................................................................................................60 4-1 Diagram of marker integration mutagenesis............................................................65 4-2 Primer AB47............................................................................................................70 4-3 Garden Blot..............................................................................................................71

PAGE 10

x 4-4 Total DNA digested with Eco RI, Southern blot hybridiz ed consecutively with orf 115, 116, and orf117................................................................................................72 4-5 Orientations for the integration of the suicide vector in pXcB................................74 4-6 Mapping of the lacZ promoter orientation in pB16.1, pB23.1-11, pB26.4.0, and pB31.2.1...................................................................................................................74 4-7 Construction of pAB29............................................................................................76 4-8 Inoculation result of plasmid pAB29 ma ted into B69.4 and inoculated on Duncan Grapefruit.................................................................................................................76 4-9 The lacZ promoter is functiona l in the B-strain.......................................................77 4-10 Results of initial inoculati on of orf117 marker integrations....................................78 4-11 Clones used for attempts to complement maker-integrated mutants and the regions of plasmid pXcB they represent...............................................................................79 4-12 Attempts to complement marker-int egrated mutants with plasmid pAB36.6..........79 4-13 Attempts to complement marker-integrated mutants with plasmid pJR8.2.............80

PAGE 11

xi Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science IN SEARCH OF PATHOGENICITY FACTORS OF Xanthomonas citri pv. aurantifolii By Asha Brunings May 2005 Chair: Dean W. Gabriel Major Department: Plant Molecular and Cellular Biology Xanthomonas citri pv. aurantifolii strain B69 is a causal agent of citrus canker disease found in South America. Gene pthB is located on a self-mobilizing plasmid, pXcB, found in B69, and is necessary but insufficient for pathogenicity. The B69 derivatives that are cured of pXcB are non pathogenic. Subclones of pXcB were introduced into a pXcB-cured deri vative of B69, B69.4, together with pthB or an isofunctional homologue, pthA in an attempt to complement the mutant phenotype. None of the subclones used in our st udy restored pathogenicity to B69.4/ pthB or B69.4/ pthA The complete sequence of pXcB was determined, and 38 open reading frames (ORFs) were identified. None of the identifi ed putative genes app eared to be an obvious pathogenicity factor. Almost one-third of pX cB appears to encode a large polycistronic transcriptional unit, which includes a complete type IV secretion system, consisting of 12 genes. In addition, the transcri ptional unit carries one gene upstream, and five additional genes immediately downstream of the type IV secretion-system genes.

PAGE 12

xii Marker integration, resulting in the inte rruption of two ORFs downstream of the type IV secretion system, was performe d. Both marker-integrated strains lost pathogenicity on both lime and grapefruit. However, several clone s carrying wild-type copies of the putative genes encoded by the ORFs failed to complement the mutants. Since plasmid instability could explain the in ability to complement the mutants, and one of the interrupted ORFs resembled a plasmid stability gene, the stability of the mutant plasmids was tested in vitro and in planta No instability of the mutant plasmids was observed. Stability of the mutant plasmids together with one of the complementation plasmids was also tested in vitro The mutant plasmids and the complementation plasmid were stable together in liquid cu lture without selection pressure.

PAGE 13

1 CHAPTER 1 LITERATURE REVIEW Citrus Canker Citrus canker is a bacterial disease of citr us that causes premature leaf and fruit drop (Gottwald et al., 2001, 2002; Graham et al., 2004; Stall & Civerolo, 1991). Since there is no cure or effective treatment available, citrus is subject to strict federal and state quarantine regulations in the United States of America. These regulations require canker eradication, including destruc tion of both infected citrus (e.g., showing symptoms of citrus canker disease) and “expos ed” citrus (e.g., all citrus tree s within a 1900 ft radius of an infected tree). Destruction of exposed trees is thought to eliminate 95% of subsequent infections resulting from dispersal of inoc ulum from an infected tree (Gottwald et al., 2001). Disease Symptoms Citrus canker disease symptoms first app ear as oily circular lesions, 2-10 mm in size, usually on the abaxial surface of the leaf (Brunings & Gabriel, 2003, Gottwald et al., 2001, 2002; Graham et al., 2004; Stall & Civerolo, 1991). The lesions become raised and blister-like, and eventually grow into white or yellow spongy pustules on leaves, stems, thorns and fruit. The pustules then darken and thicken into light tan to brown rough, corky cankers. The individual lesions look like cr aters, with a necrotic sunken center and raised sides. The lesions on stems can someti mes fuse together and result in splitting of the epidermis along the stem. Older lesions on leaves and fruit can have more elevated margins and can be surrounded by a yellow chloro tic halo (which may disappear in time).

PAGE 14

2 The sunken craters are especially noticeable on fruits, but do not penetrate far into the rind. Defoliation and premature abscission of a ffected fruit occurs on heavily infected trees. Damage to leaves, stems, and fruit resu lts in a reduction of fruit quality and yield. The actual cost of the disease is greater than simply reduced fruit quality and quantity. Regulatory actions (including dest ruction of infected and expose d trees) plus restrictions on shipping of fruit from areas that are aff ected by citrus canker cause additional losses (Schubert et al., 2001). Host Range Citrus canker affects Rutaceous plants, primarily Citrus spp., Fortunella spp. and Poncirus spp. (Gabriel, 2002), depending on the ci trus canker strain involved. The disease is most severe on grapefruit, so me sweet oranges, Mexican (Key) limes, and trifoliate orange (Gottwald et al., 2002). However the actual host range depends on the strain of citrus canker (see be low). In general, lush, young tissu es are more susceptible to citrus canker than older ones (Stall et al., 1982). Therefore every flush in the citrus culture cycle (about three times a year) pr ovides a period of vu lnerability (Schubert et al., 2001). Control Measures Quarantine measures are used by several countries to prev ent introduction of citrus canker, but outbreaks s till occur (Gottwald et al., 2002). Breeding of citrus trees is virtually impossible. Therefor e no known citrus varieties ar e resistant against citrus canker in a gene-for-gene manner (Gottwald et al. ,1993). However, there are important differences in host range among various citr us canker strains (see below) that may indicate active host defences. In addition, there is some field resistance against the disease, generally correlated with the size a nd number of stomatal openings and a lack of

PAGE 15

3 aggressive growth by the citrus specie s (Goto, 1969; McLean & Lee, 1922; Stall et al., 1982). More aggressive rootstocks in crease susceptibility of the scion. Citrus canker was first declared erad icated in Florida in 1933 (Schoulties et al., 1987). In 1986, the disease reappeared (Stall & Civerolo, 1991) and was again declared eradicated in 1994 (Gottwald et al., 2002). Two more outbreaks occurred: one in Miami in 1995, and another on the Florid a west coast in 1997 (Gottwald et al., 2002). Eradication efforts are ongoing at the time of th is writing; and include strict quarantine of the pathogen, and removing all infected and ex posed trees. It was reported that as of April 2004 more than 2.9 million trees have b een destroyed in Florida as a result of regulatory action (Florida Citrus Mutual, 2004). The Pathogen Citrus canker is caused by two phyl ogenetically distin ct groups of Xanthomonas : one originating in Asia, a nd the other in South Americ a (Gabriel, 2001). Each group comprises subgroups that are pathogenic va riants. The most widespread group is the Asiatic group, or X. citri pv. citri A (syn. X. campestris pv. citri; X. axonopodis pv. citri), causing type A canker. This group is prevalen t in citrus-growing regions throughout Asia, and has been successfully eradicated in Nort hern Australia and South Africa. The second group is the South American group ( X. citri pv. aurantifolii B), causing B type canker. This group was first identified in Argen tina, and spread to Uruguay and Brazil. Xanthomonas citri pv. aurantifolii B is more restricted in host range than that of type A. It mainly affects Mexican (Key) lime and mandari n; and to a lesser extent, grapefruit. Xanthomonas citri pv. aurantifolii C, causing C type canke r, is closely related to type B and also originated in South America. It s host range, however, is limited to Mexican (Key) lime. Apparent derivatives of X. citri pv. citri A have emerged (called types A* and

PAGE 16

4 Aw). Type A* was identified in Southwest Asia by Vernire et al. (1998). Type AW was more recently identified in Florida (Sun et al., 2000). Both A* and Aw types have a host range limited to Mexican lime. All citrus ca nker strains cause identical disease symptoms on susceptible citrus, but host range differen ces and in particular the hypersensitive response (HR) may play a role in restricting host range. Th e pathogenicity of different strains of citrus canker is summarized in Table 1-1. Table 1-1. Relative pat hogenicity of all known X. citri strain groups on four citrus species Canker group C. sinensis (Sweet Orange) C. paradisi (Grapefruit) C. limon (Lemon) C. aurantifolii (Mexican lime) X. citri pv. citri A X. citri pv. citri A* X. citri pv. citri AW HR X. citri pv. aurantifolii B (white) X. citri pv. aurantifolii C HR HR HR +++ weak canker;++++ strong canker; no symptoms; HR Hypersensitive Response From: Brunings & Gabriel (2003) The pathogen is dispersed for short distan ces by rain splash. It enters the plant mesophyll through the stomata, or through open wounds. After about 4 days, watersoaking can be observed. The bacteria produce xanthan gum, which is highly hygroscopic, and fills up the intercellular sp aces by absorbing water and swelling. Within 7-10 days after inoculation, the typical canker craters rupture and cause bacteria to ooze to the surface, from whence they are disperse d through rain splash. The citrus leaf miner Phyllocnistis citrella Stainton, although not a long-dis tance vector for the pathogen, greatly accelerates local spreading of citr us canker by extensively wounding the leaf and allowing easy entry inside the leaf. Disper sal of xanthomonads over longer distances is facilitated by man and machinery (e.g., pruning tools).

PAGE 17

5 Pathogenicity Factors Gene pthA and its Predicted Product Gene pthA a member of the avrBs3/pthA family of genes, appears to encode a protein required for pathogenicity of citr us-canker causing strains. Members of the avrBs3/pthA gene family are widespread among xanthomonads. To date 27 members of the family have been identified (Gabri el, 1999b; Leach & White, 1996) and they are remarkably similar in sequence. The genes are all flanked by inverted terminal repeats, so they could potentially transpose (De Feyter et al., 1993). In fact, as will be discussed later, pthA is found on a plasmid which also has IS Xc 4, an insertional element shown capable of transposition (Tu et al., 1989). The predicted proteins all have from 15 to 22 direct, leucine-rich repeats, which are n early always 34-aa in length. In addition, the predicted proteins all encode three nuclea r localization sequences and a C-terminal eukaryotic transcriptional activ ation domain (Gabriel, 1999b; Zhu et al., 1998). Most members of the gene family have been shown to be avirulence (avr) genes that act in a gene-for-gene fashion to cause a hypersensiti ve response on plants that carry the cognate resistance gene (Gabriel, 1999a). Gene pthA was the first member of the gene family shown to be required for pathogenicity (Swarup et al., 1991). Since then, many, but not all, gene family members have been show n to contribute to pathogenicity (Yang et al. 1994). The 102-bp repeat regions of the members of the avrBs3/pthA gene family are important for the host specificity of pathoge nicity and gene-for-g ene specificity of avirulence (Yang et al., 1994). The 102-bp repeats differ slightly in sequence; if the specific repeats of one member is swapped with those of another, th e specificity of the resulting chimeric gene changes to that of the donor of the repeat region (Yang et al.,

PAGE 18

6 1994). The repeat region within the members of the gene fam ily is a source for alternate specificities due to intragenic and interg enic recombination (Yang & Gabriel, 1995a). The nuclear localization sequences of PthA and AvrBs3 have been shown to be functional and required for pathogenicity and avirulence, respectively (Szurek et al., 2001; Yang & Gabriel, 1995b). In addition, th e eukaryotic transcriptional activation domain at the C-terminus is re quired for avirulence activity of avrXa10 a member of the gene family that is re quired for avirulence of X. oryzae pv. oryzae (Zhu et al., 1998). X. citri pv. citri A strain 306, as all other type A strains examined, has four members of the avrBs3/pthA gene family, named pthA1 A2 A3 and A4 The four pthA genes are located on two nativ e plasmids, pXAC33 and pXAC64; each plasmid encodes two members (Figure 1-1). Gene pthA4 is the same size as pthA, (isolated from X. citri pv. citri strain 3212T), and is 99.7% identical to pthA Additionally, the second leucinerich tandem repeat of both genes is exceptio nal, encoding 33 amino acid residues instead of the 34 that are typical of a ll other repeats in these genes and in most other repeats in the gene family. Gene pthA1 has one repeat region less than pthA and pthA4 ; while pthA 2 and pthA3 each have two repeats less. Although functional analyses have not been reported, these observations indicate that pthA4 of strain 306 is most likely a functional equivalent of pthA Additional pathogenicity genes belonging to the avrBs3/pthA gene family have been identified in citrus ca nker-causing strains, including pthB pthC (Yuan & Gabriel, unpublished), and pthW (Al-Saadi & Gabriel, 2002) All are isofunctional with pthA (capable of restoring pathogenicity to a pthA mutant strain) and required for pathogenicity.

PAGE 19

7 Figure 1-1. Maps of plasmids from Xanthomonas citri pv. citri Type A strain 306. A) Map of plasmid pXAC64. B) Map of plasmid pXAC33. From : Brunings & Gabriel (2003)

PAGE 20

8 Quorum Sensing Many plant pathogenic bacteria utilize quorum sensing---t he induction of genes in response to an increase in population density--to turn on pathogenici ty genes inside the host environment. Pathogenicity genes, incl uding the biosynthesis of extracellular polysaccharide (xanthan gum), of Xanthomonas campestris pv. campestris, are regulated by a cluster of genes, designated rpf ( r egulation of p athogenicity f actors; Dow et al., 2000). The complete genome sequence of th e Asiatic strain of citrus canker, X. citri pv. citri strain 306 (da Silva et al., 2002) revealed that genes homologous for all of the X. campestris pv. campestris rpf cluster are present. In X. campestris pv. campestris, rpfA encodes an aconitase, implicated in iron homeostasis (Wilson et al., 1998), and rpfB and rpfF are responsible for the synthesis of a small diffusible signal factor, DSF (Barber et al., 1997). As the bacterial population in the host environment grows, the increase in the concentration of DSF is thought to trigger tran scription of pathogenici ty factors when it exceeds a certain minimum threshold. The genes rpfC G and H are proposed to be components of a sensory-transduction system (Tang et al., 1991). A transposon insertion in X. campestris pv. campestris rpfE caused reduced levels of exopolysaccharide and some extracellular enzymes (endoglucanase a nd protease), while the level of another extracellular enzyme, polygalact uronate lyase, increased (Dow et al., 2000). Nutrition One way plant pathogenic bacteria can de rive nutrition from their host is by secreting cell-wall degrading enzymes and us ing the breakdown product of the cell walls as a source of nutrition. X. citri produces enzymes that could assist in the breakdown of the plant cell wall and provide nutrition. Its ge nome shows that it does not have pectin esterases, but there are three pectate lyas es, six cellulases, five xylanases and an

PAGE 21

9 endoglucanase. The endoglucanase, BcsZ (gi|22001634), belongs in family 8 of the glycosyl-hydrolases which hydrolyze 1,4-D-glucosidic linkages in cellulose. In addition there is a permease which imports degraded pectin products in a hydrogentransport coupled fashion into the bacterial cell. X. citri pv. citri produces fewer cell-wall degrading enzymes than X. campestris pv. campestris, and da Silva et al. (2002) suggest that this may be why the two pathogens cause different symptoms on their hosts. X. campestris pv. campestris causes black rot of cruc ifers, and sometimes displays blight symptoms (Alvarez et al., 1994), and spreads systemically through the xylem. X. citri pv. citri, on the other hand, doesnÂ’t cause rotting or blight of ci trus, never becomes systemic and only causes local lesions. In addition, pthA which is likely injected directly into the host cell, is required for optimal growth of the pathogen in citrus (Swarup et al. 1991). This implies that some citrus plant respons e(s) may be necessary for nutrient release, possibly endogenous loosening of the cell wall in preparation of ce ll division and cell enlargement. Attachment Bacteria can attach to host cells with special proteins ca lled adhesins, or specialized organelles called pili (Lee & Schneewind, 2001). X. citri has four gene clusters and two separately located genes that are predicted to be involved in type IV pilus biosynthesis and regulation. Two genes encoding proteins called fimbrillins, FimA gi|21243966 and gi|21243967 (85% similar in pr edicted amino acid sequence), are located within one of the clusters. A gene designated pilA located elsewhere in the genome, is similar to type II pilin (PilE) from Neisseria meningitidis The sequence of the two fimA genes is similar to PilA from other bacteria and they are located in a cluster of genes containing other type IV pilus genes pilB pilC pilD pilR and pilS The gene products of pilS and pilR are

PAGE 22

10 homologous to two-component sensor/regulat ory proteins and control expression of pilA (Hobbs et al., 1993; Wu & Kaiser, 1997). The major subunit of the type IV pilin is first exported by the general secretory pathway (GSP). It has a short, basic N-terminal signal sequence, unique to type IV pilus proteins. PilD is a specific l eader sequence peptidase that removes the signal sequen ce of PilA and other type IV pilus biosynthesis proteins, and methylates the new N-terminus (Fin lay & Falkow, 1997; Russel, 1998). Mature, translocated pilin polymerizes at the plasma membrane, and the pilus is pushed through the central cavity of the out er membrane secretin (Parge et al., 1995; Russel, 1998). Interestingly, there are two genes en coding proteins similar to PilA in X. citri Pseudomonas stutzeri has type IV pili that are re quired for DNA uptake and natural transformation, and two genes encoding proteins similar to PilA (74% similar at the amino acid level; Graupner & Wackernagel, 2001). The pilus biogenesis machinery and assemb ly is highly conserved in bacteria (Hultgren et al., 1993). Their assembly genes are simila r to type II secretion genes, but the N-terminal signal sequences are different (Russel, 1998). Type IV pili (also called fimbriae) have been proposed to attach b acterial pathogens to th e host cell wall (Farinha et al., 1994; Kang et al., 2002) and retract (Skerker & Be rg, 2001; Wall & Kaiser, 1999), pulling the bacterium closer to the host ce ll (Wall & Kaiser, 1999). Type IV pili are important for virulence of Ralstonia solanacearum (Kang et al., 2002) and Pseudomonas aeruginosa (Hahn, 1997). Since most plant pathogenic bacteria, including X. citri can provoke a nonhost HR, plant cell wall attachment may not be (very) host specific. Just upstream of the xps cluster (see below), X. citri has two genes with similarity to xadA from X. oryzae pv. oryzae. XadA is a non-fimbri al, adhesin-like outer membrane

PAGE 23

11 protein required for virulence of X. oryzae pv. oryzae (Ray et al., 2002). Proposed to be a cell wall surface anchor protein, XadA belongs to a family of proteins that are more similar at their C-terminus (which forms an outer membrane anchor domain) than at their signal sequences (Marchler-Bauer et al., 2002). Some non-fimbrial adhesins are autotransporters (type V secretion): they are exported across the bacterial inner membrane by the general secretory pathway, a nd then secrete themselves across the outer membrane (Henderson et al., 2000; Henderson & Nataro, 2001). One of the xadA genes encodes a protein (gi|21244271) that is mi ssing the unusually long, highly conserved Nterminal signal sequence that is ty pical of autotransporters (Henderson et al., 2000) and is therefore not likely to be a functional Xa dA homolog. The other XadA (gi|2124427) does have a conserved autotransporter N-terminal sequence. Therefore, it may be involved in tight adhesion to plant cell walls, and coul d potentially have a f unction in bacterial virulence. Two yapH genes (similar to yapH from Yersinia pestis ) are predicted to encode proteins similar to XadA, but both lack th e typical N-terminal signal sequence of autotransporters. Type II (General) Secretion (T2S) Type II secretion systems, (reviewed by Sandkvist, 2001) are common, but not ubiquitous in gram negative bacteria (Finla y & Falkow, 1997). Secretion occurs in two steps (Finlay & Falkow, 1997; Lee & Schneew ind, 2001;): the Sec (general secretion) machinery exports substrates with a signal peptide across the inner membrane of gram negative bacteria and the type II secretion genes secrete them across the outer membrane. Bacterial T2S systems have been shown to s ecrete diverse molecules such as cellulases, pectate lyases, toxins, proteases, and alkaline phosphatases (Russel, 1998).

PAGE 24

12 There are two T2S clusters in the X. citri genome associated with two rearrangements of the otherwise highly syntenic genomes of X. citri and X. campestris pv campestris (da Silva et al., 2002). The first one (the xcs cluster) consists of one transcriptional unit of thir teen open reading frames, xcsC through xcsN and one (downstream of xcsN ) is similar to a TonB-dependent receptor gene. Consistent with the idea that this T2S cluster has been transf erred horizontally, it ha s a higher (~68%) G+C content than the surrounding DNA region (~65%). The second T2S system (the xps cluster) contains two transcripti onal units. One c onsists of the xpsE and xpsF genes and the other of ten genes: xpsG through xpsN xpsD and a conserved hypothetical gene, similar to glycosyltransferase genes. Part of this cluster has a G+C content as low as 47.5%. Type III Secretion (T3S) Many plant and animal pathogenic bacteria have a T3S system (T3SS) consisting of more than twenty proteins which togeth er function to inject pathogenicity factors directly into host cells (B uttner & Bonas, 2002; Hueck, 1998). The T3S genes of plant pathogens are called hrp ( h ypersensitive r esponse and p athogenicity) genes and are required both for pathogenicity on hosts and elicitation of the HR on hosts and nonhosts (Fenselau & Bonas, 1995; Roine et al., 1997). Nine of these genes are highly conserved between the T3S systems of plant and an imal pathogens and have been renamed hrc ( hr p c onserved; Bogdanove et al. (1996). The X. campestris pv. vesicatoria Hrp system can secrete at least some heterologous type III s ecreted proteins from both plant and animal pathogens (Rossier et al., 1999). Expression of the T3SS can be contact-dependent (Ginocchio et al., 1994; Menard et al., 1994; Pettersson et al., 1996; Rosqvist et al., 1994; Watarai et al., 1995), but in the

PAGE 25

13 case of Salmonella T3S may not be contact-depende nt (Daefler, 1999). Close contact with plant host cells appears necessary for X. citri since there is compelling evidence that PthA is injected directly into host cells. Firs t, it has recently been demonstrated directly that a member of the AvrBs3/Pth family,AvrBs3, is secreted by T3S (Szurek et al., 2001). Second, mutations in the T3SS of X. citri pv. aurantifolii render the bacterium nonpathogenic (El Yacoubi et al., 2004). Third, E. coli carrying a T3SS from Pseudomonas syringae pv. syringae and a pthA allele causes canker-like symptoms on citrus (Kanamori & Tsuyumu, 1998). Fourth, if Pt hA is delivered directly into the plant cell, it is capable of formi ng canker-like symptoms (Duan et al., 1999). Fifth, PthA localizes to the nucleus (Yang & Gabr iel, 1995b), and so does AvrBs3 (Szurek et al., 2001). Finally, the transcripti onal activation domain of AvrX a10, another member of the AvrBs3/Pth family of proteins, is functional (Zhu et al., 1998). The hrp genes are proposed to encode proteins that form a hrp pilus (Roine et al., 1997). The X. citri hrp cluster is part of a “patho genicity island” in the main chromosome, as indicated by the following f eatures (Hacker & Kaper, 2000): 1) it spans more than 23-kb; 2) it encodes a system that secretes pathogenicity factors into host cells; 3) it is always associated with pathogenic species of Xanthomonas ; 4) it has regions with a different G+C content than the rest of the X. citri genome; 5) it carries mobile genetic elements (transposases), and 6) it represents an unstable region of DNA since there are differences between the X. citri T3SS and the closely related T3SS cluster from X. campestris pv. vesicatoria, which is also part of a pathogenicity island (Buttner & Bonas, 2002; Nol et al., 2002). The genes of the hrp cluster are induced in planta and controlled by the hrp regulatory proteins HrpG and HrpX (Wengelnik & Bonas, 1996; Wengelnik et

PAGE 26

14 al., 1996, 1999). For some HrpX-regulated genes involved in the T3S system, a PIP ( p lanti nducible p romoter) box has been identified (F enselau & Bonas, 1995), although there are genes with PIP boxes that are not regulated by HrpX, and there are genes whose expression is under control of HrpG, and HrpX that do not have a PIP Box (Buttner & Bonas, 2002). Type III effectors are injected by the TTSS into the host cell. It is logical to assume that expression of these effectors would be c oordinately regulated w ith expression of the T3SS genes. In a screen performed by Guttman et al. (2002) for type III secreted Pseudomonas syringae effector proteins, 13 different hop ( h rp/hrc o uter p rotein) genes were identified, of wh ich all but one had a hrp box (Innes et al., 1993) in their promoters. The situation may be somewhat different in Xanthomonas. Although da Silva et al. (2002) found twenty poten tial PIP boxes in the X. citri genome, only a few of these indicated potential T3S effector proteins. Homologues of four avr genes ( avrBs2, avrXacE1, avrXacE2 and avrXacE3 ) and two popC family effector genes were also found. Interestingly, the avrXacE2 homologue, one of the popC homologues, and all four members of the avrBs3/pthA gene family do not contain a PIP box (da Silva et al., 2002). In fact, all avrBs3/pthA gene family members examined to date are constitutively expressed and yet are known to be delivered by T3S (Knoop et al., 1991; Szurek et al., 2001; Yang & Gabriel, 1995b). Three mechanisms have been proposed to explain how the T3SS recognizes effectors for secretion. The first proposes that N-terminal signal seque nces in the secreted protein are recognized by the T3SS and result in export of the effector (Miao & Miller, 2000; Mudgett et al., 2000). However, type III secreted proteins lack a clearly defined

PAGE 27

15 signal sequence in their amino termini (Aldridge & Hughes, 2001). In the second mechanism, molecular chaperones bind effect or proteins transiently, preventing them from folding incorrectly, and present them to the T3SS apparatus for subsequent secretion (Wattiau et al., 1994, 1996) A third mechanism was proposed for the T3S of Yop proteins by Yersinia (Anderson & Schneewind, 2001). They hypothesized that the secretion signal was encoded in the messe nger RNA (mRNA) instead of in the amino acid sequence of the secreted protein. The thr ee kinds of secretion signal are not mutually exclusive. For example, YopE appears to have two separate secreti on signals in its amino acid sequence, one of which f unctions only in conjunction w ith the secretion chaperone, SycE (Cheng et al., 1997). In Xanthomonas campestris pv. vesicatoria, a region of AvrBs2 was determined to be required for T3S and translocation to the plant cell, but a potential mRNA secretion signal was also found (Mudgett et al., 2000). It was reported that 13 type III secreted proteins from Pseudomonas syringae have very similar Nterminal regions, and most are predicted to loca lize to chloroplasts in the plant cell, which could point to a common recognition mechanis m as in chloroplasts, or to a common origin of the signal sequences (Guttman et al. 2002). It has been proposed that a signal sequence in the message of AvrB and AvrPto is responsible for their recognition by the Pseudomonas syringae TTSS (Galn & Collmer, 1999). Type IV Secretion (T4S) One of the native plasmids of strain 306, pXAC64, appears to encode a type IV secretion (T4S) or “adapted conjugation” system, and at least portions of a second potential T4S cluster are located in the main chromosome (da Silva et al., 2002). T4S systems mediate intercellular transfer of macromolecules (pro teins or protein-DNA complexes) from gram-negative bacteria to other bacteria or e ukaryotic cells (Baron et

PAGE 28

16 al., 2002; Christie, 2001). The prototype for T4S systems is the Agrobacterium tumefaciens virB cluster (Christie, 1997), of which virB2 through virB11 have all been shown to be essential for Agrobacterium virulence (Berger & Christie, 1994; Ward et al., 1990). The adapted conjugation system is, as the name implies, required for the conjugational transfer of plasmids, including self-mobilizing plasmids, which carry the T4S transfer genes and an origin of transfer ( oriT or mobilization site). Adapted conjugation systems can also mobilize other plasmids in trans if the plasmid carries an oriT site (Christie and Vogel, 2000; Winans et al., 1996). A T4SS is also necessary for the secretion of pertussis toxin by Bordetella pertussis (Christie & Vogel, 2000). T4S systems also require a “coupling factor” (Cabezn et al., 1997; Moncalin et al., 1999), a homolog of the Agrobacterium VirD4 protein. The order in which different components of the T4SS of Agrobacterium contact the T-DNA strand has been determined using a transfer DNA immunoprecipitation (TrIP) a ssay (Cascales & Christie, 2004). The proposed pathway is depicted in Figu re 1-2 (Cascales & Christie, 2004). The T4SS in the main chromosome of X. citri strain 306 is incomplete. There are three partial copies of the virB6 gene, there is no virB5 homolog, and most virB homologs that are present are incomplete. This partial virB cluster is surrounded by several transposases, indicating that this area may have been subject to more than one transposition event. This T4S cluster is not likely to encode a func tional secretion system, since it lacks esse ntial genes. It has been suggested that the T4SS on plasmid pXAC64 is also incomplete, since it appears to lack virB5, virB7 and virD4 homologs (da Silva et al., 2002). This conclusion may be incorrect, and needs experimental verification.

PAGE 29

17 Figure 1-2. Pathway of th e T-DNA strand through the Agrobacterium tumefaciens T4SS. The proteins in the vertical pathway make contact with the T-DNA strand in the order indicated, the other proteins do not make direct contact with the substrate, but are necessary at the step s indicated. Reprinted with permission from Cascales & Christie (2004), Fig 4, page 1172 The virB cluster on pXAC64 does not contain any genes that are annotated as virB7 It was found that VirB7 is required very early in the secreti on process (Cascales & Christie, 2004). It is proposed to st abilize the secretory apparatus of the A. tumefaciens T4SS (Beaupre et al., 1997; Fernandez et al., 1996). The predicted protein gene product of XACb0044 (gi|21110908) on pXAC64 is sim ilar to VirB5 and is annotated in GenBank as such. Second, a BLASTP (Altschul et al., 1990) search with the predicted TrwB sequence from pXAC64 shows similari ty with VirD4 and its homologs. TrwB functions as a coupling factor for T4S systems and trwB is a known virD4 homolog (Gomis-Ruth et al., 2002; Moncalin et al., 1999). Finally, the T4S virB cluster on pXAC64 is similar in orga nization to that of the lvh cluster from Legionella pneumophila Genbank accession Y19029 (Segal et al., 1999), where there also are two genes between the homologs of virB5 ( lvhB5 ) and virB6 ( lvhB6 ). Remarkably, the gene immediately

PAGE 30

18 downstream of lvhB5 is a homolog of virB7 lvhB7 The gene immediately downstream of virB5 in pXAC64 may therefore be a virB7 homolog, similar to the organization of the T4SS in L. pneumophila This gene organizati on is discussed in more detail in Chapter 3. Plasmids pXAC33 and pXAC64 Xanthomonas citri pv. citri A strain 306 from Brazil has two native plasmids; pXAC64 is 64.9 kb in size, while pXAC33 is 33.7 kb [(da Silva et al., 2002); refer Figure 1-1]. The restriction endonuc lease maps and sizes of these two plasmids correspond almost perfectly to the re striction endonuclease maps and estimated sizes of two plasmids, pXW45J and pXW45N, respectively, from X. citri pv. citri A strain XAS4501 from Japan that were characterized by Tu et al (1989). Two transposable elements, ISXc4 and ISXc5, were functionally characterized by transposition in E. coli and their locations were mapped on plasmids pXW45N and pXW45J (Tu et al., 1989). The corresponding locations of th ese IS elements on pXAC64 and pXAC33 are indicated in Figure 1-1. These IS elements are of potential intere st because the TnpR resolvase of ISXc5 represents a new subfamily of recombinases responsible for re solution of cointegrates of class II transposable elements, such as Tn 3 (Liu et al., 1992, 1998). Intriguingly, pXAC64, which presumably carri es a functional allele of pthA ( pthA4 ), also has a tnpA transposase gene that is 99.9% similar to tnpA of Tn5044, which is in the Tn 3 subgroup of the Tn 3 family of transposases (Kholodii et al., 2000). Since all members of the avrBs3/pthA family have terminal inverted repeats that are similar to Tn 3 transposable elements, it was hypothesized that th ese genes may transpose (De Feyter et al., 1993). It should now be possible to experimentally te st this hypothesis. If the T4SS on pXAC64 is functional, the plasmid might be self-mobilizing, Self-mobilizing plasmids carry the

PAGE 31

19 transfer functions necessary to tr ansfer other plasmids with an origin of transfer (oriT). If pthA4 is indeed a functional pthA homolog and pXAC64 is self-mobilizing, then pXAC64 would be able to tran sfer itself into other xanthom onads resident on the same host, some of which may lack ability to cau se citrus canker. ISXc 4 on the plasmid could cause transposition of the pthA gene, and multiple copies would be created. The T4SS would likely be instrumental in initiating such transfer. Plasmid Toxin-Antitoxin Modules Low-copy number plasmids constantly face the problem of stable maintenance in the absence of select ion pressure (Gerdes et al., 1986). The rate at which plasmid-free cells accumulate in culture depends on the ra te in which they arise, and the relative growth of plasmid-containing and plasmid-free cells (Gerdes et al., 1986). The growthrate of plasmid-containing cells is depre ssed by plasmid metabolic load, which results from the demands placed upon the host by replic ation, transcription and translation of the plasmid genome. The load can become severe if the plasmid is used to express genes with gene product that are toxic or interfer e with the host's metabolism (Gerdes et al., 1986). Low-copy number plasmids have developed multiple approaches to maintain their segregational stability. Among these are (1) an active partition system ensures that each daughter cell receives a copy of the newly replicated plasmi d at cell-division; (2) sitespecific recombination of dimers and highe r-order multimers which arise by homologous recombination optimizes the number of plasmids available for segreg ation at the time of cell division; (3) toxin-antitoxin (TA) systems. TA systems compromise the survival of plasmid-free segregants (Lehnherr et al., 1993; Magnuson et al., 1996), mediate the exclusion of competing plasmids (Cooper & Heinemann, 2000), and prevent cell-division in plasmid-free daughter cells (Ruiz-Echevarria et al., 1995). A toxin-antitoxin locus

PAGE 32

20 should stabilize any unstably inherited plasmid, no matter wh at causes the loss of the plasmid (Gerdes et al., 1986). TA Mechanisms Toxin-antitoxin (TA) systems do not aff ect copy number of plasmids (Deane & Rawlings, 2004; Radnedge et al., 1997; Rawlings, 1999), but make them more stable than can be accounted for by replication c ontrol and active partition (Lehnherr et al., 1993). The toxins are proteins encoded by TA syst ems are very potent and their artificial overproduction leads to rapid and massive ce ll-killing, in most cases corresponding to several orders of magnitude in reduction of viable-c ell counts (G erdes, 2000). Presumably, as a result of this toxicity, atte mpts to clone the toxin gene by itself have often been unsuccessful (Gerdes, 2000; Haye s, 1998; Smith and Rawlings, 1998), except under tightly regulated non-leaky promoters (Gerdes, 2000; Roberts & Helinski, 1992), or replacement of the start codon by one th at reduces the level of toxin expression (Gerdes, 2000). There appear to be two mechanisms by wh ich the antidote can neutralize the effect of the toxin. In the hok (ho st k illing)/sok (s upressor o f host k illing) system, the antisense sok RNA prevents synthesis of hok (Gerdes et al., 1990, 1992). In most other mechanisms, the antitoxin is proposed to be a protein. The antitoxin forms a tight complex with toxin (Gerdes, 2000) and inactiv ates it. The toxins are generally stable, while the antitoxins are degraded by cellula r proteases (Gerdes, 2000). Immediately after cell division, plasmid-free cells have a pool of TA complexe s plus a pool of free antitoxin (Gerdes, 2000; Rawlings, 1999). While the an titoxin degrades, the toxin remains, and kills the cells (Gerdes, 2000). There is a la rge amount of variability in the amino acid sequence of TA encoding genes, especially between the sequence of the antitoxins, which

PAGE 33

21 makes it difficult to identify TA comple xes by homology searches (Deane & Rawlings, 2004; Gerdes, 2000; Rawlings, 1999). TA Operon Organization Stability operons appear to be organized similarly on different plasmids. In most cases a gene encoding an antitoxin precedes a gene that encodes a toxin, and the genes overlap (Hayes, 1998; Rawlings, 199 9; Smith & Magnuson, 2004), suggesting translational coupling (Roberts & Helinski, 19 92). Both genes are predicted to encode proteins of about 10 kDA. Putative promoter elements are typically present immediately upstream of the antitoxin gene. The antitoxins autoregulate transcription of the TA operons by binding to operator sites upstream or overlapping with the operon promoters (Eberl et al., 1992; Gerdes, 2000; Magnuson et al., 1996; Magnuson & Yarmolinsky, 1998; Smith & Magnuson, 2004) and negativel y regulating transcription (Magnuson et al., 1996). The pasABC promoter is autorepressed by Pa sA (Smith & Rawlings, 1998). In the case of the Phd repressor/an titoxin protein, the N-terminus is required for repressor, but not for antitoxin activity, while the C-terminus is required for antitoxin, but not for repressor activities (Smith & Magnuson, 2004). Autoregulation appears to be essential for pas -mediated plasmid stabilization because when the pas genes were placed behind the IPTG-regulated tac promoter, they were unable to stabilize a heterologous test plasmid (Smith & Rawlings, 1998). Under conditions in which the killing function is artificially overexpressed from a strong external promoter, killing is observe d in plasmid-containing cells (Gerdes et al., 1986). This indicates that the operon is tigh tly regulated under native conditions. The pas stability system is unusual in that it consists of three genes rather than the two-gene systems identified in other plasmi ds (Smith & Rawlings, 1998). The pasA gene encodes

PAGE 34

22 an antidote, pasB encodes a toxin, and pasC encodes a protein that appears to enhance the neutralizing effect of the an tidote (Smith & Rawlings, 1998). The ParD protein alone is sufficient for autoregulation of the parDE operon (Gerdes, 2000). Expression of parE seems to rely on translational coupling with parD (Roberts & Helinski, 1992). Testing Plasmid Stability Roberts & Helinski (1992) tested plasmi d stability by diluting mid-log phase cultures under antibiotic selection into antibio tic-free medium and allowing them to grow overnight to mid-log or stationary phase. A liquots were plated on antibiotic-free medium both before and after overnight growth. The stability of the plasmid was expressed as the percentage of cells carry ing plasmid at the initia l and final time points. The plasmid maintenance stability determinant of the large virulence plasmid pMYSH6000 of Shigella flexneri consists of two small open reading frames STBORF1 and STBORF2 and is likely to encode a postsegregational killing system (Radnedge et al., 1997; Sayeed et al., 2000). This Stb system also exerts incompatibility against a coresident plasmid containing Stb. Insertional mutagenesis showed that parD and parE were both required for plasmid stabilization (Roberts & Helinski, 1992). The parDE operon is sufficient to confer vectorindependent, broad-host range stability under several c onditions tested (Roberts & Helinski, 1992). However, in rich-medium growth of E.coli carrying a wild-type recA gene, this minimal region is insufficient for plasmid stabilization and requires the presence of the parCBA operon (which functions to resolve plasmid multimers to monomers) for efficient stabilization unde r certain growth co nditions (Roberts & Helinski, 1992).

PAGE 35

23 CHAPTER 2 A NOVEL PATHOGENICITY FA CTOR REQUIRED FOR CITR US CANKER DISEASE IS CONTAINED ON A SELF-TRANSMISSIBLE PLASMID The only pathogenicity factors known to be required for citrus canker disease are a functional T3SS and at least one member of the Xanthomonas avr/pth gene family (Gabriel, 1999b). Xanthomonas citri pv. aurantifolii B69 has been report ed to have at least two native plasmids (Civerolo, 1985), a T3SS (Leite, Jr. et al., 1994), and two DNA fragments that hybridize with pthA indicating the presence of two members of the Xanthomonas pthA gene family (Yuan & Gabriel, unpublished results). One of these genes, pthB is located on a selftransmissible plasmid, pXcB, and DNA fragments carrying pthB complement a mutant Asiatic strain that has pthA interrupted (Yuan & Ga briel, unpublished). Clone d DNA fragments carrying the other potential pthA gene family member fail to complement the pthA mutation, therefore pthB is apparently the only active member of the pthA gene family in B69 (Yuan & Gabriel, unpublished results). Gene-swapping experiments showed that pthA and pthB are isofunctional. pthB is necessary for B69 to cause citrus canker disease (Yuan & Gabriel, unpublished results). However, although pthB is required for pathogenicity, it is not the only gene on pXcB that is required, since pthB cannot alone complement a B69 strain, B69.4, cured of pXcB (Yuan & Gabriel, unpublished results). In this study, pXcB was mapp ed. Based on the map, several subclones of pXcB were made, and then tested for self-transmissibility and pathogenicity by mating into B69.4 carrying pthA

PAGE 36

24 Material and Methods Bacterial Strains, Plasmids, and Recombinant DNA Techniques The bacterial strains and plasmids used in this study are listed in Table 2-1 and Table 2-2. Xanthomonas spp. were cultured in PYGM medium at 30 C (De Feyter et al., 1990). Escherichia coli strains were grown in LuriaBertani (LB) medium (Sambrook et al., 1989) at 37 C. For solid media, agar was added to 15 g/lite r. Antibiotics were used at the following final concentrations in g/mL: ampicilin (Ap) 100 fo r high copy vectors, 50 for medium copy vectors; chloramphenicol (Cm) 35, gentamycin (Gm) 3, kanamycin (Kn) 25, rifampicin (Rif) 75 or 50, spectinomycin (Sp) 35, streptomycin (Sm 100), tetracyclin (Tc) 15. Table 2-1. List of bacteria l strains used in this study Strain Relevant Characteristics Reference or source E. coli DH5 supE44 lacU169( 80 lacZ M15) hsd R17 recA1 gyrA96 thi-1 relA1 Gibco-BRL, Gaithersburg, MD HB101 supE44 hsdS20 (rB -mB -) recA13 ara-14 proA2 lac Y1 gal K2 rpsL20 xyl-5 mtl-1 (Smr) (Boyer and RoullandDussoix, 1969) Xanthomonas 3213T X. citri pv. citri, species type strain, ATCC 49118 (Gabriel et al., 1989) 3213Sp Spr mutant of 3213T (Gabriel et al., 1989) B21.2 pthA ::Tn5gus A, marker-exchange mutant of 3213Sp (SprKnr) (Swarup et al., 1991) B69 X. citri pv. aurantifolii B strain B69 ATCC 51301 B69Sp Spr derivative of B69 This study BIM2 pthB ::pUFR004 marker interruption mutant of B69Sp (SprCmr); carries pBIM2. Yuan & Gabriel, unpublished BIM6 pXcB::pUFR004 cointegrate derivative of B69Sp (SprCmr); carries pBIM6 Yuan & Gabriel, unpublished B69.4 Spontaneous Rifr mutant of B69, cured of pXcB Yuan & Gabriel, unpublished Plasmids were transferred between bacteria us ing triparental matings as described (Swarup et al., 1991), using pRK2013 as helper plasmid. In expe riments to test self-transmissibility of a particular plasmid, the helper strain was omitted.

PAGE 37

25 For extraction of total DNA from Xanthomonas 12 mL bacterial cultures were grown in 25 mL flasks at 30 C with slow shaking. Cells were harvested at 8000 g and washed twice with 12 mL and 1.5 mL 50 mM Tris-HCl, 50 mM EDTA, 0.15 mM NaCl (pH 8), respectively. The cells were resuspended in 627 L of TES buffer (10 mM Tris-H Cl, 10 mM EDTA, 0.5% SDS, pH 7.8). Afterwards 33 L of protease st ock solution was added, the tubes were inverted several times to mix well, and the cell suspension was incubated at 37 C for -3 hours. The protease stock solution consists of 20 mg/mL protease in 10 mM Tris-HCl, 10 mM NaCl, pH7.5; the protease was predigested at 37 C for 1 hour and stored at -20 C. The mixture was gently mixed overnight by rotation with an equal volume of phenol:chloroform:isoamyl alcohol (25:24:1) buffered with Tris-HCl pH 8. The laye rs were separated by centrifuging for 15 minutes at 5,000 g and the top layer was gently transferre d to a new tube with a pipet. A second extraction with phenol:chloroform:isoamyl alc ohol (25:24:1) was carri ed out, followed by an extraction with chloroform:isoamyl alcohol (24: 1). One-tenth volume of 3 M NaAc was added, the tubes were inverted several times, and 0.9 volume of room temperature isopropanol was added and mixed. Precipitated DNA was spooled out with a heat-sealed glass Pasteur pipet, transferred to a tube containing 600 L of 10 mM Tris-HCl, 1 mM EDTA, pH 8, and 200 g/mL RNaseA, and incubated for 1-2 hours at 37 C. A phenol:chloroform:isoamyl alcohol and a chloroform extraction were carried out, followe d by precipitation with Na Ac and 2 volumes of room temperature 95% ethanol. After careful mi xing, the total DNA was s pooled out as before and resuspended in 200 L sterile distilled water. To analyze the DNA, 1L was digested with a restriction enzyme and run on a 0.7% agarose gel.

PAGE 38

26 Plasmids were isolated from E. coli using alkaline lysis (Sambrook et al., 1989). For extraction of plasmid DNA out of Xanthomonas cells were washed twice in TEN (50mM TrisHCl, 50 mM EDTA, 150 mM NaCl, pH 8) buffer to remove excess xanthan gum prior to alkaline lysis. Table 2-2. List of plasmids used in this study Plasmid Relevant characteristics Reference or source pXcB Native, self-mobilizing 37-kb plasmid from B69 This study pRK2013 ColE1, Knr, Tra+, helper plasmid (Figurski & Helinski, 1979) pUFR047 IncW, Mob+, lacZ +, Par+ (AprGmr) (De Feyter et al., 1993) pUFR053 IncW, Mob+, lacZ +, Par+ (CmrGmr) Yuan & Gabriel, unpublished pLAFR3 IncP1, cos+, Tra-, Mob+ (Tcr) (Staskawicz et al., 1987) pBIM2 pthB::pUFR004 (Cmr ) Yuan & Gabriel, unpublished pBIM6 pXcB::pUFR004, cointegrated upstream of pthB (Cmr ) Yuan & Gabriel, unpublished pAB2.1 EcoRI-HindIII fragment of pZit45, containing pthA, in pLAFR3 This study pAB11.62 18-kb EcoRI fragment of pBIM2 in pUFR047 This study pAB12.2 14-kb EcoRI fragment of pBIM2 in pUFR047 This study pAB13.2 8.9-kb EcoRI fragment of pBIM2 in pUFR047 This study pAB14.1 7.2-kb EcoRI fragment of pBIM2 in pUFR047 This study pAB18.1 EcoRI-HindIII fragment of pYD9.3 in pUFR047 This study pQY96 14-kb HindIII fragment containing pthB from pXcB, cloned in pUFR053 Yuan & Gabriel, unpublished pYD9.3 pthA in pUC118 (Apr) (Duan et al., 1999) pUFY14.5 pthA in pGEM 7Zf(+) (Apr) (Yang & Gabriel, 1995b) pYY40.9 StuI-HincII fragment of pthA in pUFR004 Yang & Gabriel, unpublished pZit45 pthA in pUFR047 (Swarup et al., 1992) B69.4 Spontaneous Rifr mutant of B69, cured of pXcB Yuan & Gabriel, unpublished Electroporation was carried out using the E ppendorf Electroporator 2510, according to the manufacturersÂ’ instructions. Electrocompetent ce lls were prepared by washing the cells from a liquid culture with 1 culture vol ume of distilled water. The cells were then washed with 0.1 volume of distilled water and resuspended in st erile 10% glycerol, followed by storage at -80 C until use. The cells were thawed out completely on ice (about 10 minutes). 1 L of DNA solution (diluted 1:10 to minimize the salt concentration) was added to 40 L (for cuvettes with a gap width of 1 mm) or 100 L (for cuvettes with a gap width of 2 mm) of electrocompetent cells,

PAGE 39

27 gently mixed and incubated on ice for 20-60 seco nds. Electroporation was carried out in sterile cuvettes with a voltage setting of 1800 V for cuvettes with a 1 mm gap width, and 2500 V for cuvettes with a 2 mm gap width. To ensure that the electroporation was carried out with high efficiency, the time constant was noted afte r each electroporation. The electroporation was repeated with a greater dilution of DNA if the ti me constant was below 4.0 ms. Immediately after electroporation, 1 mL ri ch medium (SOBG for E. coli cells) was added to the cuvette. Medium and electroporated cells were th en transferred to gla ss culture tubes and in cubated for 1 hour at 37 C while shaking. 200 L of the mixture was th en spread out on selective plates and grown overnight at 37 C. Other standard recombinant DNA procedures were used essentially as described (Sambrook et al., 1989). Restriction enzyme digesti on, alkaline phosphatase treatment, DNA ligation, and random priming reactions were performed as recommended by the suppliers. Southern hybridization was performed using nyl on membranes as described (Lazo & Gabriel, 1987). Plant Inoculations Citrus plants “Duncan Grapefruit” ( Citrus paradisi ) and “Mexican Lime” or “Key Lime” ( Citrus aurantifolii ) were grown under natural light in th e exotic pathogen quarantine greenhouse facilities at the Division of Pl ant Industry, Florida Department of Agriculture and Consumer Services, Gainesville, Florida. Two-day old liquid bacterial cultures were rinsed and adjusted to an OD600 of 0.4 with sterile CaCO3-saturated water, and pressure infiltrated into the abaxial leaf surface of the plants using tuberculin syringes (Yang & Gabriel, 1995a). All inoculations were repeated at least three times.

PAGE 40

28 Figure 2-1. Members of the avr/pthA gene family in X. citri pv. aurantifolii. A) Ethidium bromide stained gel of strains B69 and B 69.4. Strain B69 has multiple plasmids. B) Southern hybridization of gel in A) probe: internal Bam HI fragment of pthA Strain B69 has 2 members of the avr/pthA gene family, both of which are on plasmids. DNA in all lanes is restricted with Eco RI

PAGE 41

29 Results B69 Has Multiple Plasmids and a T3SS Plasmid and total DNA were extracted from B69 and its derivative lacking pXcB, B69.4. Comparing the lanes labeled “plasmid DNA B69” and “plasmid DNA B69.4” in Figure 2-1 part A, it is clear that B69.4 (which lacks pX cB) has at least one additional plasmid. Two plasmids in B69 have Eco RI fragments (~18 kb, and ~4.5 kb) that hybridize to an internal Bam HI fragment of pthA The 18 kb fragment is derived from pXcB and is absent from B69.4, which is cured of pXcB (Figure 2-1 part B). The 4.5 kb Eco RI fragment is located on a plasmid other than pXcB and still present in B69.4. Gene pthB is Insufficient to Cause Canker Strain BIM2, in which pthB alone is mutated, was unable to cause canker. BIM2 was complemented with pAB 2.1, pZit45, pAB18.1 (all with pthA ), and pQY96 (with pthB ); thus confirming that pthB is necessary for pathogenicity and that pthA is isofunctional with pthB Shuttle vectors containing pthA or pthB were mated into B69.4 and inoculated on Duncan Grapefruit and Mexican Lime leaves. None of the plasmids used to complement BIM2 restored pathogenicity to B69.4; therefore pthB is necessary but insufficient to cause citrus canker, and one or more additional pathogenic ity factors must reside on pXcB. Mapping of pBIM2 and pBIM6 Plasmid pBIM2 ( pthB ::pUFR004) does not restore pat hogenicity when introduced into B69.4, however, pBIM6 (pXcB::pUFR004) does. In pBIM2 the inserted vector interrupts pthB (Figure 2-2), while in pBIM6 (Figure 2-3) the inserted vector appears to have no mutational effect. Both pBIM2 and pBIM6 were mated into E. coli for restriction and Southern blot analysis. Plasmid pBIM2 DNA was digested with a variety of restriction enzymes for mapping

PAGE 42

30 purposes (Figure 2-4). Several re striction sites (for example Eco RI, Hind III, Kpn I, Pst I, and SstI ) were easily mapped, since they occurred infrequently in pXcB Additional sites for Eco RI, Sst I, and Kpn I at the 3Â’ end of the integrated vector (the start of subcl one pAB13.2, see Figure 22) were also mapped. Additional sites were mapped based on restriction digests of pBIM2, pBIM6, or their subclones (data not shown). Figure 2-2. Restriction map of pl asmid pBIM2 and its subclones. E= Eco RI, B= Bamh HI, H= Hin dIII, K= Kpn I, P= Pst I, S= Sst I, Sa= Sal I Figure 2-3. Restriction map of plasmid pBIM6. E= Eco RI, B= Bam HI, H= Hin dIII, K= Kpn I Constructs Based on the restriction map of pBIM2, the four Eco RI fragments of pBIM2 were subcloned into the shuttle vector pUFR047 (De Feyter et al., 1993). The resulting subclones (Figure 2-2) are pAB11.62 (16 kb), pAB12.2 ( 12.7 kb), pAB13.2 (8.9 kb), and pAB14.1 (6.3 kb). Plasmid pAB11.62 contains the 5Â’-end of pthB while pAB13.2 contains the 3Â’-end. Both of these clones hybridize to pthA (Figure 2-5).

PAGE 43

31 Gene pthA was recloned directionally in to cosmid vector pLAFR3 ( repP ) from the Eco RIHin dIII fragment from pYY14.5 (Yang & Gabrie l, 1995a). The resulting clone, pAB2.1, contains pthA driven by its native promoter. Gene pthA was recloned directionally in to cosmid vector pLAFR3 ( repP ) from the Eco RIHin dIII fragment from pYY14.5 (Yang & Gabriel, 1995a). The resulting clone, pAB2.1, contains pthA driven by its native promoter. Pathogenicity Tests Plasmid pAB2.1 complemented B21.2, a pthA -interrupted mutant of an A strain of canker. To test whether any of the subc loned fragments of pBIM2 would be able to confer pathogenicity to B69.4 in conjunction with pAB2.1 (carrying pthA on a repP plasmid), pAB11.62, pAB12.2, pAB13.2, and pAB14.1 (carrying BIM2 fragments on a repW plasmid) were introduced into B69.4 together with pAB2.1. Plasmid pAB2.1 was ma ted into B69.4 as a negative control. None of these combinations of plasmids restored B69.4 to pathogenicity. Figure 2-4. Restriction digests of plasmid pBIM2 for mapping purposes

PAGE 44

32 Since none of the subcloned fragments of pXcB was able to restore B69.4 pathogenicity in the presence of pthA it is possible that whatever ot her factor(s) were required for pathogenicity were truncated in the subclones. Al ternatively, it is possibl e that multiple factors are required for pathogenicity in addition to pthB and that none of the s ubclones provided a full complement of necessary factors. Figure 2-5. Plasmid pBIM2 and subclones. A) A 0.7% agarose gel of plasmid DNA extracted from pBIM2 and its subclones. B) Southe rn hybridization of the gel in A; probe internal BamHI fragment of pthA

PAGE 45

33 Self-transmissibility Each of the pBIM2 subclones was tested for se lf-transmissibility. Of the recipient strain ( R; E. coli DH5 MCR, or HB101), donor stain ( D; E. coli DH5 MCR/pAB11.62, pAB12.2, pAB13.2, or pAB14.1) and helper strain ( H; E. coli DH5 MCR/pRK2013), every possible combination (RDH, RD, RH, DH) was allowe d to mate overnight on plain LB plates and streaked on selective plates afte rwards, selecting for the donor plasmid in the recipient strain. None of the donor strains used in these matings resulted in colonies for the R+D combination, while all combinations of RDH resulted in viab le colonies. Whatever ot her factor is required for self-mobility was either truncated in the subc lones or none of the subc lones provided all of the necessary mobility factors. Discussion Members of the AvrBs3/PthA family of proteins are probably transferred directly into host cells by the hrp system (Casper-Lindley et al., 2002; Duan et al., 1999; Yang & Gabriel, 1995b). Consistent with this theory, pthA can cause canker-like symptoms when introduced directly into citrus cells using particle bombardment or an Agrobacterium -based delivery system (Duan et al., 1999). In X. campestris pv. aurantifolii there is at least one and possibly two members of the Xanthomonas avr/pth gene family, of which one, pthB is functional and located on a selfmobilizing plasmid, pXcB. When pthB by itself, without the rest of pXcB, is introduced in X. campestris pv. aurantifolii cured of the plasmid (B 69.4), B69.4 does not regain the ability to cause citrus canker. These data imply that there is at least one other f actor on pXcB required for pathogenicity. Subcloned fragment s of pBIM2 do not complement pthA upon introduction together with pthA into B69.4. Possible explanations include that more than one additional factor

PAGE 46

34 might be necessary, or that the required factor (s) are incomplete in the subclones. Since no other genes are known to be cri tical for citrus canker disease, there might be part of the hrp system on pXcB, or altern atively, a chaperone for pthB pXcB was therefore sequenced completely (see Chapter 3). However, there are no genes similar to hrp genes or chaperones were found on the plasmid. The fact that pthB resides on a self-mobilizing plasmid raises the interesting point that a Xanthomonas strain is able to mobilize pathogenicity genes by horizontal transfer. This might explain how different Xanthomonas spp. are able to cause essentia lly the same disease, differing only in their host range. Hypothetica lly, strains that already have a hrp system and factors elsewhere in their genome that determine their host range could receiv e pXcB by horizontal transfer. If pXcB were then stably maintained in the recipient strain, the result might be a strain capable of causing citrus canker disease.

PAGE 47

35 CHAPTER 3 COMPLETE PLASMID SEQUENCE Chapter 2 showed that pthB is necessary but insufficient for pathogenicity of Xanthomonas campestris pv. aurantifolii strain B 69. The fact that none of th e pXcB subclones restored pathogenicity together with pthA means that the required factor or factors are incomplete on the subcloned fragments. Plasmid pXcB was seque nced completely to identify all possible pathogenicity factors. Material and Methods Bacterial Strains, Plasmids, and Culture Media The bacterial strains and plasmids used in th ese experiments are listed in Table 2-1 and Table 2-2 in Chapter 2. Xanthomonas strains were cultured as described in Chapter 2. General Bacteriological Techniques The self-mobilizing pl asmid pXcB and its pthB -interrupted derivative pBIM2, were transferred from X. citri pv. aurantifolii strain B69 to E. coli DH5 or HB101 by biparental matings. For all other matings a helper strain, pRK2013 (Figurski & Helinski, 1979), was used in triparental matings (Swarup et al., 1991). The DNA sequence of pXcB was determined by sequencing the EcoRI fragments of pBIM2 subclone d into pUFR047: pAB11.62, pAB12.2, pAB13.2, and pAB14.1 (refer to Chapter 2). Regions overlapping the subclones were sequenced directly from pBIM2.

PAGE 48

36 Recombinant DNA Techniques Plasmid DNA extractions, DNA analysis techniqu es, and electroporation were as described in Chapter 2. DNA Sequence Analysis DNA sequencing was performed by the Inte rdisciplinary Center for Biotechnology Research (ICBR) sequencing core The GCG sequence analysis software package (Wisconsin Package Version 10.3, Accelrys, Inc. San Diego, CA.) was used to assemble the sequence data. Putative open reading frames (orfs) capable of encoding peptides of at least 50 amino acids were identified using ORF Finder (National Center for Biotechnology Information, 2004) and the Open Reading Frame Identification feature (Bayer College of Medicine, 2003). The program Codonpreference (Wisconsin Package Version 10.3, Accelrys Inc., San Diego, CA), which scores orfs based on the similarity of thei r codon usage compared to a codon usage table (codonfrequency) or by their third position GC bias was used to eliminate orfs that were unlikely to be coding sequences, and to determine sequenc ing errors, since it simplifies identification of frame shifts. BLASTX (Bogdanove et al., 1998) was used to identify orfs for which homologues exist in the GenBank non-redundant database. Whenever possible, putative Shine-Dalgarno (Shine & Dalgarno, 1974) sequences (ribos ome-binding sites) were identified. Potential promoters were identified using the BCM Gene Feature Searches option for prokaryotic promoter prediction by neural network (LBNL). Tran scriptional terminators were predicted using the prog ram GESTER (Unniraman et al ., 2001, 2002). Individual putative protein sequences were then analyzed in more detail using the programs that are part of the Expert Protein Analysis System (Swiss Inst itute of Bioinformatic s, 2003). For predicted molecular weight and pI, PSORT (Nakai & Horton, 1999), and MOTIF (Kyoto University

PAGE 49

37 Bioinformatics Center, 2004) were used. Puta tive transmembrane regions were identified using TmPred (Hofmann & Stoffel, 1993). For th e identification of lipop rotein, the DOLOP (A Database Of Bacterial LipOProteins) pr ogram was used (Madan & Sankaran, 2002). Results Features of Plasmid pXcB pXcB is a circular double-stranded DNA mol ecule of 37,106 bp. Thirty eight orfs were identified on pXcB (Table 3-1). There are a total of 30,729 base pairs of coding DNA, resulting in an average gene density of 1.02 orfs/kb, and 828 coding base pairs/kb. Open reading frames (orfs) were initially identified using ORF Finder, the BLASTX program and the GCG package program Codonpreference. Thirty six orfs app eared to start with an ATG initiation codon, and two (orfs 117 and 219) were predicted to start with a GTG. Sequences 6-12 nucleotides upstream of predicte d start sites were compared to the ShineDalgarno consensus sequence TGGAGG (Shine & Dalg arno, 1974). Most orfs appeared to have sequences similar to these ribosome binding sites. A map of the orfs found on pXcB is given in Figure 3-1, they are listed in Tabl e 3-1. The process used to find orfs is illustrated in Figure 3-2 for orf207 and orf208. In the figure the 5'-end of each orf is on the right hand side of the horizontal box. The codon preference and third position codon bias of orf208 remain high throughout the predicted orf indica ting a real gene, while the thir d position codon bias of orf207 drops significantly in the latter half of the predicted orf, i ndicating that this orf might not represent a real gene. Both orfs are similar to genes in the Genbank database. Orf207 is similar to a hypothetical gene on pWWO and a hypotheti cal gene on pXAC64 (gi:21264228) plasmid (da Silva et al., 2002), while orf208 is similar to the traA gene on pWWO from P. putida and another hypothetical gene on pXAC 64. Orf208 was determined to be more likely to encode a

PAGE 50

38 gene, based on the better third position codon bias, and renamed traA A similar process was used to hand-annotate each orf on the plasmid. Figure 3-1. Map of plasmid pXcB. E= Eco RI, H= Hin dIII. Orfs that are part of the T4SS are indicated in red, genes homol ogous to other transfer genes in blue, genes similar to plasmid maintenance genes in purple, the in sertion element and resolvase in orange, orfs similar to genes in the GenBank databa se in green, orfs with no similarity to known genes in grey, and pthB in yellow. The average G+C content was 60.9%, with a minimum of 46.25% and a maximum of 71.75%. The G+C percentage, calculated for win dows of 300 bp, with a step size of 3 was plotted against nucleotide position (Figure 3-3) The G+C content forms a trough in a region of

PAGE 51

39 relatively low G+C content, roughly from position 7200 to 7800. This region corresponds to orfs 105 and 106. From 32560 to 34475 (corresponding to the 102 -nucleotide tandem, direct repeat region of orf130, pthB ) the G+C content was remarkably high compared to the rest of pXcB. Promoters Potential promoter regions were identified with the Bayer College of Medicine Gene Feature Search for Promoter and Transcription Binding Site Predictions. Eleven promoters with a score of 0.95 or higher were identif ied on the +strand, and fifteen on the strand of pXcB. Lowering the cutoff to 0.90 resulted in the pred iction of eleven additional promoters on the +strand and seven additional prom oters on the -strand. An even lo wer cutoff of 0.85, resulted in fifteen more predicted promoters on the +stra nd and twelve more on th e -strand. A list of potential promoters with a score of 0.85 or higher is given for th e +strand in Table 3-2, and for the strand in Table 3-3; they ar e illustrated in Figure 3-4. There do not appear to be separate pr omoters for orfs 101, through 103, orf105 through 107, and for orf108 through 114. Orf101 through 104, orfs 105 through 108, and orfs 111 and 112 have overlapping coding sequences and are most likely transcriptionally and translationally co-regulated. Orfs 113 and 114 are likely transcri bed from the same promoter as orfs 111 and 112, since the closest promoter ha s a low cutoff (0.87) and is almo st a thousand bases away from the initiation codon of orf113. This could add five orfs to the virB cluster that are not related to any known T4SS genes. Upstream of orf115 (at position 14256-14301) there is a potential promoter (cutoff 0.86) for orf115 (which star ts at position 14556). Orfs 115 through 117 could be transcribed either separately or together with upstream orfs.

PAGE 52

40Table 3-1. Open reading frames identified in plasmid pXcB Orf Position Gene #AA MW (kDa)pI Similarity Predicted function 201 c(706..1527) orf201 273 30.4 10.94pXAC33(1) gi:21264186, XF(2) gi:10956774 202 c(2030..2359) orf202 109 11.4 7.82 nsh(3) 100 (2757..3143) mobD 128 14.8 9.27 pXAC64(4) gi:21264276, PP(5) gi:32469966 101 (3172..3549) virB2 125 13.4 8.92 pXAC64 VirB2 gi:21264275, PP gi:31745842 CDD: pfam06921, VIRB2, VIRB2 type IV secretion protein; COG3838, VirB2, Type IV secretory pathway; cyclin T-pilin subunit (Lai & Kado, 1998; Jones et al. 1996), attachment, mating channel (Christie, 2001) 102 (3546..3845) virB3 99 11.0 8.55 pXAC64 VirB3 gi:21264274, PP gi:31745843 CDD: pfam05101, VirB3, Type IV secretory pathway; COG3702, VirB3, Type IV secretory pathway; attachment, stabilized by VirB4, VirB6 (Christie, 2001) 103 (3832..6501) virB4 889 101.5 6.41 pXAC64 VirB4 gi:21264273, PP gi:31745844, 18150991, 32469963 CDD: pfam03135, CagE_TrbE_VirB, CagE, TrbE, VirB family, component of type IV transporter system; COG3451, VirB4, Type IV secretory pathway; translocation energetics, ATP-ase (Christie, 2001); ATP/GTP-binding site motif A (P-loop); energy for pilus biogenesis (Cascales and Christie, 2003) 104 (6498..7163) virB5 221 23.9 8.80 pXAC64 VirB5 gi:21264272, PP gi:31745845, 18150990, 32469962, Pseudomonas syringae gi:28867760 attachment, minor pilin subunit, stabilized by VirB6 (Christie, 2001) 105** (7214..7429)/ (7229..7429) orf105/ virB7 71/ 66 7.7/ 9.51 9.67/ 7.06 pXAC64 gi:21264271, PP gi:32469961, 31745846, 18150989 mating channel, lipoprotein (Christie, 2001); stabilizes VirB9, and thereby other VirB proteins (Cascales & Christie, 2003)

PAGE 53

41Orf Position Gene #AA MW (kDa)pI Similarity Predicted function 106 (7407..7829) orf106 140 15.3 6.72 PP gi:18150988, 32469960, 31745847, pXAC64 gi:21264270 107 (7817..8683) virB6 288 31.5 8.44 pXAC64 VirB6 gi:21264269, PP gi:31745848, 32469959, 18150987, Legionella pneumophila gi:19919314, 6249468 CDD: pfam04610, TrbL, TrbL/VirB6 plasmid conjugal transfer protein; COG3704, VirB6, Type IV secretory pathway; mating channel, required for the stability of VirB3 and VirB5, and formation of VirB7 homodimers (Hapfelmeier et al. 2000), possibly a inner membrane pore component (Christie, 2001) 108 (8680..9354) virB8 224 25.0 8.94 pXAC64 VirB8 gi:21264268, PP gi:31745849, 32469958, 18150986, Legionella pneumophila gi:19919315, 6249469 CDD: pfam04335, VirB8, VirB8 protein; COG3736, VirB8, Type IV secretory pathway; mating channel, assembly factor for positioning of VirB9 and VirB10 (Christie, 2001); bridge between subcomplexes (Cascales & Christie, 2003) 109 (9379..10167) virB9 262 28.6 9.01 pXAC64 VirB9 gi:21264267, PP gi:31745850, 18150985, 32469957 CDD: CagX (6.9% aligned); COG3504, VirB9, Type IV secretory pathway; stabilizes T4SS; possibly outer membrane pore (Cascales & Christie, 2003) 110 (10164..11384) virB10 406 42.9 6.04 pXAC64 VirB10 gi:21264266, PP gi:32469956, 31745851, 18150984 CDD: pfam03743, TrbI, Bacterial conjugation TrbI-like protein; COG2948, VirB10, Type IV secretory pathway; mating channel (Christie, 2001); bridge between inner membrane and outer membrane subcomplexes (Cascales & Christie, 2003)

PAGE 54

42Orf Position Gene #AA MW (kDa)pI Similarity Predicted function 111 (11395..12414) virB11 339 38.1 8.55 pXAC64 VirB11 gi:21264265, PP gi:31745852, 32469955, 18150983 CDD: pfam00437, GSPII_E, Type II/IV secretion system protein; COG0630, VirB11, Type IV secretory pathway; COG4962, CpaF, Flp pilus assembly protein, ATPase; translocation energetics, ATP-ase (Lai & Kado, 2000); energy for pilus biogenesis (Cascales & Christie, 2003) 112 (12401..13312) virB1 303 31.9 9.19 pXAC64 VirB1 gi:21264264, pSB102 gi:15919992, Brucella melitensis biovar Abortus gi:8163884, B. melitensis gi:17988369 CDD: cd00254, SLT, Transglycosylase SLT domain; pfam01464, SLT, Transglycosylase SLT domain; mating channel, transglycosylase (Christie, 2001; Lai & Kado, 2000); channel assembly (Cascales & Christie, 2003) 113 (13358..13945) orf113 195 20.8 6.80 E. coli hypothetical protein gi:21885930, A. tumefaciens conserved hypothetical protein gi:17938747, Sinorhizobium meliloti hypothetical protein gi:16263173, Rhizobium etli gi:21492820 ATP/GTP-binding site motif A (P-loop) 114 (13958..14563) orf114 201 21.7 9.62 Burkholderia cepacia gi:46320012, PP putative nuclease gi:18150980, Proteus vulgaris EDTA-resistant nuclease gi:21233859, Yersinia enterocolitica endonuclease gi:2208977, Salmonella thyphimurium EDTA-resistant nuclease gi:4903112 CDD: cd00138, PLDc, Phospholipase D; ProDom: endonuclease plasmid EDTAresistant nuclease; Pfam: phospholipase D active site motif 115 (14556..14978) orf115 140 15.9 5.38 nsh

PAGE 55

43Orf Position Gene #AA MW (kDa)pI Similarity Predicted function 116 (15110..15393) yacA 95 10.7 9.98 Photorhabdus luminescens subsp. laumondii gi:37525856, Nitrosomonas europaea ATCC 19718 gi:30248566, Bartonella henselae str. Houston-1 gi:49475484, Shigella sonnei YacA gi:9507443 117 (15394..15681) yacB 95 10.8 4.49 Nitrosomonas europaea ATCC 19718 30248567, Photorhabdus luminescens subsp. laumondii TTO1 gi:37526182, Azotobacter vinelandii gi:23102756, Ralstonia metallidurans CH34 gi:48766888, Bartonella henselae str. Houston-1 gi:49475483, S. sonnei YacB gi:9507444 CDD: pfam05016, Plasmid stabilisation system protein; COG3668, ParE, Plasmid stabilization system protein 205 c(15731..18706) trwC 991 109.6 9.62 pXAC64 TrwC gi:21264259, PP gi:31745859, 18150978, 32469948, E. coli gi:19572639, 1084124 CDD: COG0507 RecD, ATP-dependent exoDNAse (exonuclease V), alpha subunit helicase superfamily I member (DNA replication, recombination, and repair); Prosite pattern: ATP/GTP-binding site A motif (P-loop), lipocalin, 206 c(18720..20294) trwB 524 58.2 8.93 pXAC64 TrwB gi:21264258, PP TraB gi:32469947, 18150977, 31745860, E. coli TrwB gi:1084123 CDD: pfam02534, TRAG, TraG/TraD family member of the TraG/TraD (73.5%); Prosite pattern: ATP/GTP-binding site motif A (Ploop), walker-B site for nucleotide binding; BLOCKS: TraG protein; ProDom: plasmid TraD membrane inner ATP-binding; Pfam: TraD/TraG 208 c(20486..20881) traA 131 14.5 6.31 pXAC64 hypothetical protein gi:21264253, PP TraA gi:31745861, 32469946, 18150976

PAGE 56

44Orf Position Gene #AA MW (kDa)pI Similarity Predicted function 118 (21269..21724) traD 151 17.0 9.55 pXAC64 hypothetical protein gi:21264198, PP TraD gi:, 18150975, 32469945, E. coli gi:49065168, 25815153, yciA gi:10955477 119 (21795..22427) stbB 210 23.8 5.40 pXAC64 conserved hypothetical protein gi:21264200, PP hypothetical protein gi:18150974, 32469944, X. fastidiosa conserved hypothetical protein gi:10956760, Salmonella typhimurium stbB gi:2801371 120 (22437..22827) orf120 129 13.3 8.06 PP hypothetical protein gi:18150973 209 c(22978..23295) IS401 105 11.6 9.82 Azotobacter sp. FA8, gi:22077125, Ralstonia metallidurans CH34 gi:48770250, Burkholderia cepacia IS401 gi:2497402, Nitrosomonas europaea ATCC 19718 gi:30248272 Insertion sequence (Wood et al. 2001) 122 (23463..23894) orf122 143 15.3 8.37 P. syringae pv. pisi Hrp effector candidate gi:42475527, Ralstonia solanacearum putative transmembrane protein gi:17548482 210 c(24077..25039) tnpR 320 36.2 9.39 Y. enterocolitica gi:37518401, A. tumefaciens putative resolvase gi:15891106, 17937554 CDD: pfam00239, Resolvase,COG1961, PinR, Site-specific recombinases 211 c(25198..25884) kfrA 228 24.3 4.86 pXAC64 KfrA gi:21264227, Achromobacter denitrificans gi:45368551, P. aeruginosa gi:37955784, plasmid RK2 KfrA gi:78651, P. alcaligenes gi:2429365 125 (26609..27384) secA 258 28.9 5.42 pXAC64 conserved hypothetical protein gi:21264217, Burkholderia cepacia R1808 gi:46320800 CDD: pfam02810, SEC-C motif; COG3318, Predicted metal-binding protein related to the C-terminal domain of SecA

PAGE 57

45Orf Position Gene #AA MW (kDa)pI Similarity Predicted function 213 c(27823..29010) orf213 395 45.7 8.90 p XAC64 conserved hypothetical protein gi:21264286, Burkholderia cepacia R1808 gi:46322784, P. alcaligenes unknown gi:49188515 CDD: pfam06414, Zeta_toxin; ATP/GTPbinding site A motif (P-loop) 214 c(29017..29301) orf214 94 10.4 4.56 pXAC64 hypothetical protein gi:21264286 215 c(29368..29700) orf215 110 12.0 9.98 Y. enterocolitica orf78 gi:28373039 216 c(29693..30325) parA1 210 22.2 6.85 E. coli gi:46949068, Y. enterocolitica orf79 gi:28373041, Y. pestis biovar Mediaevails str. 91001 ATPases involved in chromosome partitioning gi:45476526 CDD: COG1192, Soj, ATPases involved in chromosome partitioning; similar to orf219 217 c(30438..31232) orf217 264 29.2 9.69 P. syringae pv. tomato str. DC3000 gi:28867389, A. tumefaciens hypothetical protein gi:10954854, Streptomyces violaceoruber gi:32455690 130 (31719..35221) pthB 1168 123.1 6.10 X. citri pv. citri PthA gi:899439, pXAC64 PthA4 gi:21264293, 4163845 CDD: pfam03377, Avirulence, Xanthomonas avirulence protein, Avr/PthA; pathogenicity protein 219 c(35813..36442) parA2 209 22.1 5.59 pXAC33 partition protein A gi:21264226, Bartonella henselae str. Houston-1 gi:49476005, Chlorobium limicola unknown gi:10956076 CDD: pfam00991, ParA, ParA family ATPase; COG1192, Soj, ATPases involved in chromosome partitioning; COG0455, ATPases involved in chromosome partitioning CDD=Conserved Domain Database ** See text for explanation of two sets of data for Orf105 (1) pXAC33= X. citri pv. citri strain 306 plasmid pXAC33 (gi:38201775) (2) XF= Xylella fastidiosa (3) nsh = no significant hits (4) pXAC64= X. citri pv. citri strain 306 plasmid pXAC64 (gi:21264228) (5) PP= P. putida plasmid pWWO

PAGE 58

46 Figure 3-2. Codonpreference of plasmid pXcB reverse frames from position 20,000 to 21,000. Orfs 205 and 206 may share the same promoter since there is no separate promoter for orf205 and the coding sequences of th e two orfs overlap, implying translational coupling. Orf210 does not appear to have a promoter at the cutoff levels shown. Figure 3-3. G+C content of plasmid pX cB. Window size is 300, step size 3 Transcriptional terminators were identifie d with the GeSTer algorithm (Table 3-4). Since there were two possible promoters f ound upstream of orf100 and no transcriptional terminators until after orf117, it is possible that this entire region, spanning almost 15 kb and comprising of 18 orfs, is transcribed together. There were no strong promoters (cutoff greater than 0.90) predic ted immediately upstream of the trwB and trwC genes. However, two promoters are predicted upstream of the traA gene (both with a score of 21,000 20,500 20,000 20,000 20,500 21,000 2.0 1.5 1.0 0.5 -0.0 2.0 1.5 1.0 0.5 -0.0 2.0 1.5 1.0 0.5 -0.0 0.0 0.5 1.0 0.0 0.5 1.0 0.0 0.5 1.0 ORF 207 c(20298..20717) ORF 208 c(20486..20881) virBcluster Orfs 113-117 trwB/C pthB various genes 0 10,000 20,000 30,000 45 50 55 60 65 70 G+C (%) Position

PAGE 59

47 0.85). It is possible that traA trwB and trwC are transcribed from the same promoter. There were two predicted transcri ptional terminators downstream of trwC Two potential promoters were predicted to drive pthB transcription, but no tran scriptional terminators were found downstream of pthB Table 3-2. Predicted promoters for the + stra nd of plasmid pXcB with a score of 0.85 or higher Start End Score Start End Score 17 62 0.96 21014 21059 0.86 1875 1920 1.00 21213 21258 0.92 2372 2417 0.90 22416 22461 0.92 2383 2428 1.00 22868 22913 0.95 3367 3412 0.89 25899 25944 1.00 3738 3783 0.94 26135 26180 0.85 6040 6085 0.90 26302 26347 0.94 6102 6147 0.85 27095 27140 0.91 7924 7969 0.98 27451 27496 0.99 7939 7984 0.92 27733 27778 0.92 11088 11133 0.88 28028 28073 0.97 12353 12398 0.89 28120 28165 0.86 14256 14301 0.86 28868 28913 0.87 15044 15089 0.94 30791 30836 0.89 15223 15268 0.87 31096 31141 1.00 15835 15880 0.90 31477 31522 0.94 19724 19769 0.85 31672 31717 0.88 19783 19828 0.89 35498 35543 0.99 20865 20910 0.96 35526 35571 0.87 36674 36719 0.95 Open Reading Frames The largest orf found on pXcB was orf130, which was predicted to encode PthB, a member of the AvrBs3/PthA family of protei ns. PthB was more than 90% identical to PthA, and had thirteen 34-amino acid, tandem, direct repeats in the middle of the amino sequence. The program TmPred was used to predict one weak transmembrane helix in PthB, but the score was barely above the cutoff level (see Table 3-5). Since PthB is proposed to be secreted by the T3SS, like ot her members of the AvrBs3/PthA family of proteins, it most lik ely does not have a transmembrane helix.

PAGE 60

48 Table 3-3. Predicted promoters for the strand of plasmid pXcB with a score of 0.85 or higher Start End Score Start End Score 36510 36465 0.85 15287 15242 0.85 35535 35490 0.99 15100 15055 0.97 35047 35002 0.91 13369 13324 0.98 34920 34875 0.95 12711 12666 0.93 32395 32350 0.86 11960 11915 0.92 31708 31663 0.99 11125 11080 0.87 31368 31323 0.96 8301 8256 0.94 30488 30443 0.86 8256 8211 0.88 29413 29368 0.98 7795 7750 0.87 29114 29069 0.88 6957 6912 0.98 27504 27459 0.96 6206 6161 0.96 26859 26814 0.98 5282 5237 0.86 25946 25901 1.00 4739 4694 0.85 22986 22941 0.88 3861 3816 0.94 22195 22150 0.85 3684 3639 0.95 20942 20897 0.85 2457 2412 0.88 20900 20855 0.85 1987 1942 0.97 18190 18145 0.93 1923 1878 0.96 539 494 0.89 Figure 3-4. Potential promoters on plasmid pXcB. Purple arrows indicate predicted promoters with a score of 0.95 or highe r. Red arrows indicate additional predicted promoters if the cutoff score is lowered to 0.90, while orange arrows indicate additional promoters if the cu toff is set to 0.85. The direction of the arrow indicated the direction of th e promoter. The bottom line gives the position and direction of tr anscriptional terminators A large part (10.1 kb, 27%) of pXcB consisted of an apparently complete T4SS, called the virB cluster, since the orfs closely rese mble the organization and sequence of the virB cluster of genes presen t in the crown gall agent Agrobacterium tumefaciens

PAGE 61

49 (Christie, 1997). Orf106 appeared to be part of the T4SS clus ter, but it was not similar to any known type IV secretion gene. Five or fs were found similar to mobilization or transfer genes, two were similar to genes involved in plasmid partitioning, three orfs were similar to other known genes, and 13 orfs were either similar to genes of unknown function, or not similar to any known genes. There was one IS element and one resolvase gene found. Table 3-4. Transcriptional terminators identified in plasmid pXcB by GeSTer Position Orf Strand 15893 yacB (orf117) + 22924 orf120 + 15674 trwC (orf205) 22957 IS401 (orf209) 27618 orf213 27708 orf213 35633 parA2 (orf219) 35729 parA2 (orf219) The pXcB virB Cluster The Agrobacterium virB cluster consists of eleven ge nes, virB1 through virB11 (Christie & Vogel, 2000). The pXcB virB cluster seems to be a hybr id of a T4SS and unrelated (orf106, and 113 through 117) genes, which is not uncommon (Christie, 2001). For the most part, the virB cluster is organized in the same way as the orfs in the Agrobacterium virB cluster; the data comparing the pXcB virB cluster to several other T4SS is summarized in Table 3-6 and depicted in Fi gure 3-5. There are sign ificant differences between the pXcB virB cluster and the prototype T4SS cluster from Agrobacterium The virB1 gene is located downstream of virB10 instead of upstream of virB2 There are two orfs between virB5 and virB6 instead of virB6 being immediately downstream of virB5 and virB7 is missing from in between virB6 and virB8

PAGE 62

50 Table 3-5. Presence of N-term inal signal sequences, localization of putative proteins (PSORT), and transmembrane helices (TmPred) Orf Signal sequence Predicted localization Certainty Transmemberane helices* N-terminus INside or OUTside Score (cutoff=500) 201 cytoplasm 0.528 202 cytoplasm 0.133 100 cytoplasm 0.403 101 inner membrane 0.232 3 in 4949 102 inner membrane 0.472 2 in 4186 103 cytoplasm 0.186 104 outer membrane 0.600 1 in 1602 105 inner membrane 0.361 2 in 2614 106 cleavable periplasmic space outer membrane 0.929 0.217 1 out 2499 107 inner membrane 0.495 7 out 12699 108 inner membrane 0.429 1 in 2642 109 cleavable outer membrane periplasmic space 0.939 0.324 1 out 1800 110 inner membrane 0.297 2 out 3678 111 cytoplasm 0.193 112 cleavable periplasmic space outer membrane 0.923 0.143 1 in 948 113 cleavable periplasmic space outer membrane 0.915 0.321 1 in 1502 114 cleavable periplasmic space outer membrane 0.278 0.265 2 out 1924 115 cytoplasm 0.344 116 cytoplasm 0.267 117 cytoplasm 0.453 205 cytoplasm 0.243 206 cleavable inner membrane 0.119 3 in 4098 208 cytoplasm 0.314 118 cytoplasm 0.362 119 cytoplasm 0.318 120 inner membrane 0.387 1 in 2644 209 cytoplasm 0.378 122 cleavable periplasmic space outer membrane 0.943 0.363 1 in 1796 210 inner membrane 0.134 1 out 1119 211 cytoplasm 0.240 125 cytoplasm 0.182 1 in 846 213 cytoplasm 0.480 214 cytoplasm 0.164 215 cytoplasm 0.278 216 cytoplasm 0.212 217 uncleavable inner membrane 0.115 2 in 1492 130 cytoplasm 0.342 1 in 521 219 cytoplasm 0.122

PAGE 63

51 Table 3-6. Comparison of the pXcB virB cluster with several type IV secretion-systems. Percent similarity is indicated in brackets pXcB pXAC64 VirB (1) R46 (2) (IncN) pWWO (3) (IncP-9) R388 (4) (IncW) R6K (5) (IncX) lvhB (6) Ti (7) VirB pXF51 (8) VirB2 VirB2 (94) MpfA (42) EnhD (46) VirB2 (54) XFa0005 VirB3 VirB3 (100) TraA (ns) MpfB (67) TrwM (41) LvhB3 (39) VirB3 (53) XFa0006 VirB4 VirB4 (98) TraB (41) MpfC (74) TrwK (39) Pilx4 (37) LvhB4 (43) VirB4 (60) XFa0007 (40) VirB5 VirB5 (98) TraC (40) MpfD (73) TrwJ (40) LvhB5 (39) VirB5 (57) XFa0008 (44) VirB7 XACb0043 (89) Eex (55) orf200 (76) Eex (51) Eex (59) LvhB7 (45) VirB7 (41) orf106 XACb0042 (65) orf199 (66) LvrD (49) VirB6 VirB6 (92) TraD (32) MpfE (69) TrwI (34) Pilx6 (32) LvhB6 (36) VirB6 (62) XFa0011 (35) VirB8 VirB8 (98) MpfF (56) TrwG (37) Pilx8 (37) LvhB8 (40) VirB8 (58) XFa0012 (31) VirB9 VirB9 (97) TraO (38) MpfG (73) TrwF (46) Pilx9 (31) LvhB9 (49) VirB9 (67) XFa0013 (41) VirB10 VirB10 (94) TraF (42) MpfH (56) TrwE (45) Pilx10 (41) LvhB10 (48) VirB10 (56) XFa0014 (43) VirB11 VirB11 (89) TraG (42) MpfI (56) TrwD (42) Pilx11 (45) LvhB11 (48) VirB11 (64) XFa0015 (46) VirB1 VirB1 (92) TraL (42) MpfJ (53) TrwN (38) Pilx1 (35) VirB1 (48) XFa0016 (38) (1) Xanthomonas citri plasmid pXAC64 (gi:21264228) (2) Salmonella typhimurium plasmid R46 (gi:17530571) (3) Pseudomonas putida plasmid pWWO (gi:18150858) (4) Escherichia coli plasmid R388 (gi:21885926) (5) Escherichia coli plasmid R6K (gi:12053564) (6) Legionella pneumophila (gi:6249457) (7) Agrobacterium tumefaciens Ti plasmid (gi:16119780) (8) Xylella fastidiosa plasmid pXF51 (gi:9112238) Figure 3-5. Comparison of the linear organization of several type IV secretion-systems. Arrows with the same colo r indicate homologous genes A comparison of the pXcB virB orfs with other T4SS genes is given in Table 3-6. The synteny of the genes in the pXcB virB cluster is identical to that of the virB cluster of pXAC64, and interestingly also to that of the putative mating pair formation gene cluster of the P. putida plasmid pWWO, and the lvhB cluster of Legionella pneumophila (Figure 3-9 and Table 3-6). pXAC64 (vir) Ti plasmid (vir) B2B3B4B5B7 B6B8B9B10B11 B1 200 199 ABCDEFGHIJpWWO(mpf) pXcB(vir) B2B3B4B5B6B8B9B10B11B1 B2B3B4B5B7B6B8B9B10B11B1lvh B2B3B4B5B7B6B8B9B10B11D4 lvrD

PAGE 64

52 Figure 3-6. Phylogenetic trees compiled by CLUSTALW and drawn with TreeView of A) VirB2 homologs, B) VirB4 homo logs, and C) VirB6 homologs As in pXcB, the Legionella pneumophila lvhB cluster has two orfs between lvhB5 and lvhB6 and the first one has been identified as lvhB7 (Segal et al., 1999). The VirB7 CC2417 pWWOMpfA pXcBVirB2 pXAC64 VirB2 Ti VirB2 R6K PilX2 R64 TraM R388 TrwL pXF51 XFa0005 enhD (A) LvhB4 CC B4 pWWOMpf pXcBVirB4 pXAC64 VirB4 R6K Pilx4 Ti VirB4 R64 TraB R388 TrwK pXF51 XFa0007 (B) R6K Pilx6 lvhB6 pWWOMpfE pXcBVirB6 pXAC64 VirB6 R46 TraD R388 TrwI pXF51 XFa0011 Ti VirB6 CC2420 (C)

PAGE 65

53 protein is a lipoprotein (Fernandez et al., 1996) that interacts with and stabilizes VirB4, VirB9, VirB10, and VirB11 (Baron et al., 1997; Krall et al., 2002). The VirB7 protein can form both homodimers and heterodimers with VirB9 through a disulfide bridge (Baron et al., 1997; Anderson et al., 1996) in the periplasmic space, and could possibly serve as a chaperone (Christi e, 2001). Orfs 105 and 106 of pXcB were examined more closely. VirB7 is rather small ( Agrobacterium VirB7 has 54 amino acid residues), and its similarity to homologs is not immediately apparent. Orf105 (t he smallest of the two orfs between lvhB5 and lvhB6 and in the same relative position as lvhB7 ), the most likely orf to encode a VirB7 homolog, was examined with the DOLOP algor ithm that identifies probable lipoproteins. If the start codon for orf105 is at positi on 7214, DOLOP does not recognize it as a lipoprotein, since one of the requirements is that the lipobox ([LV] [ASTV][ASG][C]) is within the first forty amino acid residues and the predicted orf105 gene product has a sequence LAGC that fits the lipobox consensu s sequence, but ends at amino acid residue forty one. If the start codon for orf105 is the next available ATG initiation codon at position 7229, resulting in a 66-amino acid residue gene product as opposed to a 71amino acid residue gene product, VirB7 i ndeed meets all the requirements for a lipoprotein, and the DOLOP algorithm recognize s it as such. The lipobox consists of the sequence LAGC, there are 4 positively charged amino acids within the first 22 residues of the signal sequence, and the length of the hydropohobic stretch in the signal sequence is 10 residues. Both versions of orf105 ha ve potential ribosome-binding sites. VirB7 has two conserved cysteines, one of which is used for cleavage of the lipoprotein signal sequence, while the other is for forming a disulfide bridge between VirB7 and VirB9

PAGE 66

54 (Anderson et al., 1996; Spudich et al., 1996). Both conserved cysteines are present in the predicted amino acid sequence of the orf105 gene product at resi due positions 36 and position 48. The truncated vers ion of orf105 therefore likel y encodes a VirB7 homolog, and is annotated as such. In Table 3-1 data is given for both the longer gene and the shorter version of orf105. Th e longer version is called orf 105, since it does not meet the requirements for a VirB7 homolog, wh ile the shorter gene is called virB7 The IncN plasmid R46, which has homologs for most of the virB cluster genes, and appears to be organized similarly as pXcB, has only one orf between the virB5 and virB6 homolog. This gene is called eex for entry exclusion (Pohlman et al., 1994), and it appears to be a virB7 homolog. The eex gene is 15% identical a nd 55% similar to pXcBÂ’s virB7 gene, its gene product is a lipoprotein as assayed by DOLOP, and like the VirB7 from pXcB it has three cysteine residues, of which one is at the end of the signal sequence and one of the others is likely used to form a disulfide bridge with the VirB9 homolog. The same argument is applicable to the eex genes from the IncW plasmid R388 and the IncX plasmid R6K, which are therefore also indicated as virB7 homologs in Table 3-6. The Xylella fastidiosa conjugation cluster on plasmid pXF51 also has two orfs (XFa0009 and XFa0010) between the virB5 and virB6 homologs. However, neither of these orfs is a lipoprotein, and therefore neither is listed as a virB7 homolog. The gene product of the orf between the virB7 and virB6 genes, orf106, is similar to hypothetical proteins of p WWO and pXAC64, in both cas es the orf is the one immediately upstream of the virB6 gene. The orf106 predicted protein has a cleavable Nterminal signal sequence and is most likely to reside in the bacter ial periplasmic space. Alternatively, it could locate to the bacterial inner membrane.

PAGE 67

55 The CLUSTALW program was used to for a phylogenetic tree of the VirB2 [major pilus subunit (Lai & Kado, 1998)] and Vi rB4 [translocation energetics (Shirasu et al., 1994)] homologs, and VirB6 (T4SS channel). VirB 4 is the largest gene in the cluster and highly conserved. The results are in Figure 3-6. For each putative protein analysis, the ones from pXcB are closest to the ones from pXAC64, and these two are grouped together with the Pseudomonas putida pWWO homologue. In each case they are distinctly different from the Agrobacterium tumefaciens virB cluster. Immediately downstream of the virB cluster on pXcB, there are five ORFs (113 through 117) that are not similar to any T4SS genes described to date. They might be transcribed as one polycistronic message together with the virB cluster genes since there is no transcriptional terminator between the virB cluster and ORFs 113 through 117. Orf113 is a conserved hypotheti cal protein, similar to orf 34 on the IncW plasmid R388. Orf113 has a cleavable N-terminal signal seque nce and is therefore predicted to localize to the bacterial periplasm, in addition it has an ATP/GTP-binding site motif A (P-loop) and is thus likely to hydrolyze ATP or GTP. Orf114 is most similar to several nucleases including one from P. putida (57% identity, 70% similarity in predicted amino acid sequence), an EDTA-resistant nuclease from Proteus vulgaris nucleases from Yersinia enterocolitica Salmonella typhimurium plasmid ColIB-P9 from Shigella sonnei, and a putative phospholipase D from Salmonella typhimurium The orf115 gene product has no significant similarity to any known proteins. Orf117 has a GTG start site. The probability that orf117 is a true coding sequence was assessed with the Codonpr eference algorithm. The Codonpreference analysis of pXcB from position 13301 through 16000 (containing orfs113 through 117) is depicted in Appendi x B. The codonpreference of orf115 was

PAGE 68

56 quite low (hovering around the average for random sequence, but the third position GC bias was significantly high throughout the orf, indicating that orf115 is translated. Both the codonpreference and third position GC bias of orfs 116 and 117 were well above the average for random sequences. BLASTP analysis (Altschul et al. 1990) found multiple plasmid-encoded predicted proteins that are similar to the toxin part of toxin-antitoxin plasmid stabilization systems. Orf117 was 71% identical and 84% similar (Expect value 1e-31) to a conserved hypothetical protein from Nitrosomonas europaea ATCC 19718 (gi:30248567). A Conserved Domain Database search (Marchler-Bauer et al., 2002) with the predicted protein sequence of orf117 resulted in 100% alignment with a 85-residue conserved domain from plasmid stabilization pr oteins, pfam05016 (Expect value 4e-12). The highest BLASTP hit, with experime ntal evidence for its function, was StaB (gi:33517368) from Paracoccus methylutens (Szymanik et al., 2004). Orf116 was only weakly similar to an anti-toxin gene, but se veral similarities existed between orf116 and 117 and TA modules as described in Chapter 1. The case for orf116 being an antitoxin ge ne was not strong, since there were no significant BLASTP hits to anti-toxin genes. Th is does not rule out th e possibility that it is an antitoxin gene, since antitoxin genes are less well conserved than toxin genes (Rawlings, 1999; Deane and Rawlings, 2004; Gerdes, 2000). Typically, TA systems are organized in operons, and are driven off thei r own promoter, since the anti-toxin gene negatively regulates expression of the operon. A potential promoter was predicted with a score of 0.94, located from position 15044 through 15089 (see Table 3-2), immediately upstream of orf116 (predicted Shine-Dalgar no sequence at 15100, and ATG start site at

PAGE 69

57 15110). Both orf116 and orf117, like toxin-antito xin proteins, are about 10 kDa, their coding sequences overlap, sugges ting transcriptional and tran slational coupling. It was therefore possible that orf116 and orf117 enc oded a plasmid stability system, with orf117 encoding a toxin and orf116 the antitoxin. Th is hypothesis was tested in Chapter 4. Organizationally, pXcB is similar to the P. putida IncP-9 plasmid pWWO. The origin of transfer, oriT, is predicted to be in a 431-nucleotide region on pWWO, between the opposing genes traA and traD. pXcB has homologs of both traA and traD and there are 388 nucleotides in between the two ge nes. Analogous to pWWO, the origin of transfer might lie in this ar ea of pXcB. A pairwise BLASTN search of these two regions shows that they are not similar to one anot her (results not shown). If pXcBÂ’s transfer origin is in this region, it is not significantly similar to th e oriT of pWWO, despite the overall similarity of the two plasmids. Several orfs are similar to mobilization genes of other plasmids. Two orfs are similar to the P. putida and E. coli trwB (orf206) and trwC (orf205) genes. Interestingly, TrwB is a homolog of the Agrobacterium VirD4, which is required for a functional T4SS (Vergunst et al., 2000). Other than trwB pXcB does not have any other virD4 homologs. Possibly, TrwC could function as the VirD 4 homolog for the T4SS on pXcB. Orf208 is similar to traA and orf118 is similar to traD both sequential genes on pWWO. Orf206 is similar to the traB from the P. putida pWWO plasmid, to trwB from the IncW plasmid R388, and (weakly) to virD4 homologs in X. citri pv. citri and X. campestris pv. campestris strain ATCC 33913 (da Silva et al., 2002). It has a cleavable N-terminal signal sequence, three predicte d transmembrane sequences and a P-loop. The orf206 gene product is predicte d to localize to the bacteria l inner membrane and likely

PAGE 70

58 capable of hydrolyzing ATP or GTP. The ge ne product of orf210 is similar to putative resolvase and predicted to have one transmembrane helix. Orf217 is similar only to unknown genes. It s gene product is predicted to have 2 transmembrane helices by TMPred and its prod uct is the only one on pXcB that appears to have an uncleavable N-terminal signal se quence. These results indicate that orf217 may be a cytoplasmic membrane protein. Orf 122 is predicted to encode a protein most likely localized in the periplasmic space (s ee Table 3-5) and is weakly similar to gi:42475527, a novel Hrp effector candidate from Pseudomonas syringae pv. pisi. However, at the time of writing, experimental evidence to support th is classification of holPpiX has not been published. Orf122 may theref ore be of some interest for future research. Similarity to Plasmids From X. citri pv. citri and P. putida Plasmid pXcB is very similar both in DNA sequence and organization to plasmid pXAC64 from X. citri pv. citri ( X. axonopodis pv. citri) strain 306 (da Silva et al., 2002). An overview of the BLASTX results of pXcB against X. citri pv. citri strain 306, using an Expect (number of matches expected to be found merely by chance) threshold of 0.001, is given in Figure 3-7. Twenty three genes on pXcB show similarity to pXAC64, and three genes are similar to genes on pXAC33 (g i:38201775), another native plasmid in X. citri pv. citri. One orf (orf213) is similar to a gene from X. campestris pv. campestris ATCC 33913 the causal agent of black rot (da Silva et al., 2002).

PAGE 71

59 Figure 3-7. Results of a BLASTX search limited to X. axonopodis pv. citri, with an Expect threshold of 0.001 Figure 3-8. Results from BL ASTN and BLASTX searches of nucleotides 7000 to 8500 of plasmid pXcB, limited to results from X. axonopodis pv. citri, with an Expect threshold of 0.10 Some distinct differences are notab le between pXcB and pXAC64. Plasmid pXAC64 is 64.9 kb in length, a nd has two members of the avr/pth gene family, while pXcB is 37.1 kb in length and has one. Both plasmids have a T4SS, with the same organization (like several other T4SS, see Figure 3-5), including two orfs between the genes that are similar to the virB5 and virB6 gene from the Agrobacterium virB cluster. An exception to the similarity of the pXcB virB cluster to the pXAC64 virB cluster exists in the region of the cluste r that encodes orf105 and orf 106, which are unrelated to any T4SS genes. This coincides with a slight but consistent rise in the G+C content of the DNA in this area (see Figure 3-3). For the nucleotide sequence correspondi ng to orfs 105 through 107 (~650-bp long), only a 40-bp region (of which 35-bp are iden tical) appears to be similar to the

PAGE 72

60 corresponding region of pXAC64, whereas the si milarity between pXcB and pXAC64 for the rest of the virB cluster is 93%. The translated pred icted proteins from these areas of the plasmid are similar to each other, even though the nucleotide sequence is not. This is demonstrated in Figure 3-8, where the upper part of the figure is the result of a BLASTN (nucleotide query nucleotide database) se arch, and the lower part the result of a BLASTX (translate nucleotide query protei n database) search fo r nucleotides 7000 to 8500. Figure 3-9. Results of a BLASTX search w ith plasmid pXcB as query and the results limited to genes from P. putida with an Expect value of 01.0 Even though BLASTN does not detect a ny significant similarity in the region corresponding to orfs 105 and 106, BLASTX sh ows that the predicted proteins are similar. This may mean that the region between orfs 104 and 108 in pXcB has evolved separately from the rest of the virB cluster, resulting in sign ificant changes in nucleotide changes, but maintaining similarity at the pr otein level. This would suggest that this region has evolved at a different rate than the rest of the virB cluster, and that the encoded proteins have an important f unction, thus preserving the rela tive stability of the amino acid sequence. The five orfs immediatel y downstream of the last gene in the virB cluster, orfs 113 through 117, which may be part of th e same transcriptional unit on pXcB, are not similar to any genes on pXAC64 at either the nucleotide level or the amino acid level.

PAGE 73

61 Several orfs on pXcB are similar to genes from pWWO, a native plasmid of P. putida (Greated et al., 2001). A BLASTX (translated nucle otide query protein database) search with results limited to P. putida illustrates this in Figure 3-9. It appears as if some ancestral stretch of DNA was interspersed by horizontal transfer of genes, interrupting one coding sequence with stretches of DNA from a different source (s olid lines indicate BLASTX hits to the same gene, connected by dashed lines indicating interrupting regions). Another striking resemblance between pXcB and pWWO is the similarity between orfs 105 and 106 from pXcB and orfs 200 and 199 from pWWO. These are the same orfs that have a different G+C content than the rest of the virB cluster and that show significant similarity at the nucleotide leve l to the corresponding region of pXAC64, but not at the amino acid level. Discussion Since X. citri pv. citri, which harbors pXAC64, causes the same disease as X. citri pv. aurantifolii, which harbors pXcB, and both plasmids carry the only functional avr/pth gene member, it would not have been surprising had pXcB been a simple deletion derivative of pXAC64. However, while many genes on pXcB are similar to genes on pXAC64 and pWWO from P. putida there are differences between the three plasmids significant enough to conclude that any one of them is not a simple deletion derivative of another. The predicted promoter sequences suggest that many if not all orfs of the virB cluster are transcribed from the same pr omoter, including five orfs immediately downstream from the virB cluster that are not simila r to any known T4SS genes.

PAGE 74

62 The overall organization of the virB cluster is like several other T4SS, with the virB7 gene and one additional orf between virB5 and virB6 Although it was initially not clear, closer inspection of orf105 resulted in the conclusi on that it encodes a VirB7 homolog. The predicted amino acid sequences of the virB7 gene and orf106 from pXcB, and XACb0043 and XACb0042 from pXAC64 are sim ilar to each other, regardless of the fact that their nucleotide seque nces are not. This could mean that the orfs encode some essential functions. The difference in nucleot ide sequence seems to suggest differential evolutionary rates, and possibly roles, for virB7 and orf106. Alternativ ely, these two orfs can have transferred horizontally from anothe r source into pXcB. This later theory is consistent with the fact that the G+C content for these two orfs is higher than that of the rest of the virB cluster. It does not explain, however, why these interruptions would be in the exact same location in pXcB and p XAC64, and why the orfs would have such different nucleotides sequences while having similar amino acid sequences. Compared to the rest of pXcB, pthB has a consistently high G+C content, also implying that it transferred hor izontally into pXcB. More ev idence that pXcB is a hybrid of stretches of DNA from many different so urces, comes from the results of a BLAST search on which it appears that one orf is interrupted by several different stretches of DNA. Consistent with a hypothesis th at it was transfer horizontally, pthB seems to have resulted in the interruption of a parti tion region, since two or fs (216 and 219), both similar to partition genes and each other (30% identical, 43% similar), appear on either side of pthB Strangely, the predicted protein en coded by orf219 has a region that is significantly similar to the c onserved domains of the Par fa mily of ATPases (see Table 31), while the protein encoded by orf216 does not One of the orfs (orf216), may have lost

PAGE 75

63 its function, since it was redundant after the dup lication event. It may no longer encode a functional Par protein. From the sequence of pXcB it is also clear that two subclones of pXcB discussed in Chapter 2, plasmids pAB12.2 and pAB14.1, contained separate parts of the virB cluster. If indeed some or all of the orfs of the clus ter do not have their own promoter, many orfs would not be transcribed in the subclone s. If any or all of the genes of the virB cluster are required for pathogenicity, none of the subcl ones would have had the necessary genes to cause citrus canker. Although the five orfs immediately downstream of the virB cluster, are not part of known T4SS clusters, they might also be transcribed from a virB cluster promoter, and even though they are all presen t in subclone pAB12.2, they might not have a promoter to transcribe the genes. To assay whether any of the orfs in the virB cluster are required for pathogenicity, mutagenesi s studies need to be performed.

PAGE 76

64 CHAPTER 4 MARKER INTEGRATION MUTAGENESIS As shown in Chapter 3, plasmid pXcB is about one-third comprised of what appears to be a complete type IV secretion (T4S) system (s ee map in Figure 3-1). T4S systems are found in gram-negative bacteria, transport a wide variet y of substrates, and have been shown to be required for pathogenicity in some cases (C ascales & Christie, 2003; Yeo & Waksman, 2004). The prototype for T4S systems is the virB cluster from Agrobacterium tumefaciens which uses the system to export a strand of DNA known as th e T-strand to its plant host (Christie, 1997). The T-strand integrates into the plant genome and contains coding sequences for genes that affect plant cell division. Bordetella pertussis uses a T4SS to transport the multi-subunit pertussis toxin, and some plamids use a T4SS fo r conjugal transfer (Christie and Vogel, 2000). Since the initial definition of T4S systems based on the A. tumefaciens T-DNA transport system ( virB ), the B. pertussis toxin exporter ( ptl ) and the transfer system of pKM101 ( tra ), many different bacteria have been reported to have a T4SS. Some of these are required for intracellular survival of pathogens in their host (Sieira et al., 2000). The T4SS encoded on pXcB appears to be related to both conjugation systems and pathogenicity related T4S systems.

PAGE 77

65 pXcB was integrated using a suicide vector in three different locations. Two orfs (orf116, and orf117) were interrupted using marker integration (Figure 4-1) and the phenotypes of the resulting mutant B-strains were assessed with respect to pathogenicity and self-mobilizing ability of the mutant pXcBÂ’s. Attempts to comp lement the mutant pheno types were carried out using several different subclones of pXcB. The marker-integrated pl asmids were also tested for self-mobilizing ability and for plasmid stability. Figure 4-1. Diagram of marker integration mutagenesis. See text for the explanation

PAGE 78

66 Materials and Methods Bacterial Strains, Plasmids, and Culture Media Strains used in this study are listed in Table 2-1 and Table 4-1. Plasmi ds are listed in Table 2-2 and Table 4-2. Xanthomonas strains were grown on PYGM media at 30 C, and E. coli strains on LB at 37 C. Agar was added to a final concentration of 15 g/L for solid media. Antibiotics were used at the following final concentrations in g/mL: Ampicillin (Ap), 50 (for low copy plasmid pUFR047), or 100 (for high co py plasmid pGEM-T Easy); Chloramphenicol (Cm), 35; Gentamycin (Gm), 3; Kanamycin 10 (for initial selection after interruption of ORFs), 20 (for restreak of gene in terruption mutants), and 50 (hi gh copy helper plasmid pRK2013); Tetracyclin (Tc), 15. Table 4-1. List of bacterial strains used in th is study, in addition to those listed in Chapter 2 Xanthomonas Strains Relevant characteristics Reference or source B16.1 pXcB::pUFR004, marker-integrated immediately upstream of orf115 (SprCmr), result of mating pAB16.1 into B69Sp2 This study B23.1-11 pXcB::pUFR012, marker-integrated immediately upstream of orf115 (SprCmrKnr), result of mating pAB23.1 into B69Sp2 This study B26.4.0 Marker-interrupted mutant of orf117 (SprCmrKnr), result of mating pAB26.4 into B69Sp2 This study B26.4.1 Marker-interrupted mutant of orf117 (SprCmrKnr), result of mating pAB26.4 into B69Sp2 This study B26.8.0 Marker-interrupted mutant of orf117 (SprCmrKnr), result of mating pAB26.8 into B69Sp2 This study B26.8.1 Marker-interrupted mutant of orf117 (SprCmrKnr), result of mating pAB26.8 into B69Sp2 This study B31.2.1 Marker-interrupted mutant of orf116 (SprCmrKnr), result of mating pAB31.2 into B69Sp2 This study

PAGE 79

67 Table 4-2. List of plasmids used in this study, in addition to those listed in Chapter 2 Plasmid Relevant characteristics Reference or source pAB15.1 PCR amplified (primers AB01 and AB02) internal fragment of orf115 in pGEM-T Easy (Apr) This study pAB16.1 Eco RI orf115 fragment, subcloned from pAB15.1 in pUFR004 (Cmr) This study pAB23.1 Eco RI orf115 fragment, subcloned from pAB15.1 in pUFR12 (CmrKnr) This study pAB23.6 Eco RI orf115 fragment, subcloned from pAB15.1 in pUFR12 (CmrKnr) This study pAB25.1 PCR amplified (primers AB35 and AB36) internal fragment of orf117 in pGEM-T Easy (Apr) This study pAB26.4 EcoRI orf117 fragment, subcloned from pAB25.1 in pUFR12 (CmrKnr) This study pAB26.8 EcoR I orf117 fragment, subcloned from pAB25.1 in pUFR12 (CmrKnr) This study pAB29 Bgl II fragment of pB23.1 (containing pthB through virB cluster ) in pUFR047 (AprGmr) This study pAB30.1 PCR amplified (primers AB43 and AB44) internal fragment of orf116 in pGEM-T Easy (Apr) This study pAB31.1 Eco RI orf116 fragment, from pAB30.1 in pUFR12 (CmrKnr) This study pAB31.2 Eco RI orf116 fragment, from pAB30.1 in pUFR12 (CmrKnr) This study pAB33.4 orfs 115-117 cloned in pLAFR3 (Tcr) This study pAB34.1 orf116 (PCR product from AB 46 and AB47) in pGEM-T Easy (Apr) This study pAB35.3 Eco RIHin dIII orf116 fragment, subcloned from pAB34.1 in pUFR047 (Apr, Gmr) This study pAB36.6 12.7-kb Eco RI fragment from pXcB cloned in pUFR047 (AprGmr) This study pAV7.3 orf116-117 cloned in pUFR047 (AprGmr) Vanamala & Gabriel, unpublished pB16.1 Marker-integrated plasmid from B16.1 (Cmr) This study pB23.1-11 Marker-integrated plasmid from B23.1-11 (CmrKnr) This study pB26.4.0 Marker-integrated plasmid from B26.4.0 (CmrKnr) This study pB31.2.1 Marker-integrated plasmid from B31.2.1 (CmrKnr) This study pJR7.1 14.5-kb fragment from pXcB containing orf100 through orf117 in pUFR047 (AprGmr) Reddy & Gabriel, unpublished pJR8.2 Eco RI subclone of pJR7.1 containing the 3'-end of virB10 through orf117 in pUFR047 (AprGmr) Reddy & Gabriel, unpublished pUFR004 suicide vector, Cmr De Feyter et al., 1990 pUFR012 suicide vector, CmrKnr De Feyter & Gabriel, unpublished

PAGE 80

68 General Bacteriological Techniques Plasmids were transferred between X. citri pv. aurantifolii and E. coli strains using biparental and/or triparental matings. Plasmid pRK2013 was used as the helper strain in all triparental matings. Recombinant DNA Techniques Xanthomonas total DNA was extracted and manipu lated as described in Chapter 2. Plasmid DNA was extracted from both E. coli and Xanthomonas as described in Chapter 2. Other standard recombinant DNA procedures were used essentially as described in (Sambrook et al., 1989). Polymerase chain reactions (PCR) were carried out using the Biometra T-gradient thermocycler. A summary of the primers and the resulting clones and markerintegrated mutants is given in Table 4-3. Marker Interruption Mutagenesis Two orfs on pXcB were interrupted by plasmi d integration. For interruption of both ORFs, an internal fragment was amplif ied by PCR and cloned into the pGEM-T Easy vector (Promega). Eco RI fragments containing the internal frag ments were then subcloned into either suicide vector pUFR004 (De Feyter et al., 1990) or pUFR12. Plasmid pU FR12 has an extra 1.2kb Pst I fragment that contains a gene en coding neomycin phosphotransferase (Knr). The resulting plasmids were then mated into B69S p using triparental mati ngs, and selection was carried out on selective PYGM agar plates. Each interruption resulted in vector integration into pXcB and a duplication of the internal fragment of the target gene (see Figure 4-1). Primers used for amplification of an internal fragme nt of orf116 were: AB43 TCGATGAGGCCCTGAAAA, and AB44 AGCCGGCGACGTGTC, for orf 117: AB35 ACAAGATCGCGACGACAT, and AB36 GGCGGGACGTATGGAC.

PAGE 81

69 Table 4-3. Summary of primer s and resulting plasmids and Xanthomonas strains Primers Clone in pGEM-T Easy Vector subcloned into Subclone name Marker-integrated Xanthomonas strain Marker-integrated plasmid AB01+AB02 pAB15.1 pUFR004 pAB16.1 B16.1 pB16.1 AB01+AB02 pAB15.1 pUFR012 pAB23.1, pAB23.6 B23.1-11 B23.6 pB23.1-11, pB23.6 AB43+AB44 pAB30.1 pUFR012 pAB31.2 B31.2.1, B31.2.3 pB31.2.1, pB31.2.3 AB35+AB36 pAB25.1 pUFR012 pAB26.4, pAB26.8 B26.4.0, B26.8.0 pB26.4.0, pB26.8.0 AB46+AB47 pAB34.1 pUFR047 pAB35.3 N/A N/A A third plasmid integration was performed using primers AB01 TAGGAGTGGAAAGAATGG, and AB02 CAACCGCTCTCAGAG. The primers amplified orf115 truncated at the 5'-end and the integr ation should therefore not result in orf115 interruption. Table 4-3 gives a summary of clones and Xanthomonas strains derived from different primer combinations. The DNA sequen ce of orfs115-117, predicte d protein translation and the position of the primer s are given in Appendix A. Phenotypic Tests For complementation tests, a 334 bp fragment containing the entire orf116, including 19 additional nucleotides upstream of the putat ive orf116 start codon, was PCR amplified from B69Sp with primers AB46 AAGCT TTCCAGAACAACTCGATC and AB47 GAATTCTAACACATAAGGTGGCGC, and cloned into pGEM-T Easy (pAB34.1). Primers AB46 and AB47 were designed so that the PCR product has an EcoR I site at the 5Â’-end and an Hin dIII site at the 3Â’end for directional clonin g into pUFR047, and there are translational stop codons in all three reading frames at the 5Â’-end be fore the start of the coding sequence of orf116 (see Figure 4-2). The Eco RIHin dIII fragment of pAB34.1 was subcloned into pUFR047

PAGE 82

70 (pAB35.3). Plasmid pAB35.3 was mated into B23.1-11, B26.4.0, and B31.2.1, and inoculated on both lime and grapefruit. Figure 4-2. Primer AB47. The prim er uses a “T” that occurs in the vector sequence to form a stopcodon in one reading frame, while two “TAA” sequences in the vector supply stopcodons in the other two reading frames Plasmids pAV7.3, pJR7.1, and pJR8.2 were mated in B23.1-11, B26.4.0, and B31.2.1 to attempt complementation. Plasmid pAB36.6 was c onstructed by reversing the insert of pAB12.2, so that the lacZ promoter would run in the same direc tion as orf113-117. Both of these plasmids were also mated into B23.1-11, B26.4.0, and B31.2.1. Plasmid pAB29 was constructed by cloning the Bgl II fragment of pB16.1 containing all orfs from pthB through orf115 into the Bam HI site of pUFR047. Plasmid pAB29 was mated into B69.4 (cured of pXcB) and inoculated to test for pathogenicity. To test whether the lacZ promoter is functional in B69 strains, pAB18.1 containing pthA driven by the lacZ promoter, was mated into BIM2 ( pthB ::pUFR004) and inoculated. Self-transmissibility of pBIM 2 and pBIM6 was tested. Since X. citri pv. citri has a second functional T4SS on another plasmid than pXcB (El-Yacoubi & Gabriel, unpublished results), which can trans-complement a virB4 mutant of pXcB, it was nece ssary to first transfer the mutant plasmids to E. coli DH5 and then demonstrate self-transmissibility to E. coli HB101. Matings were performed with and without helper to test for the ability of the marker-integrated plasmids to transfer from one E. coli strain to another without the presence of the helper strain. T G A A T T C T A A C A C A T A A G G T G G C G C* EcoRI stopcodonframe 1stopcodonframe 2 stopcodonframe 35’-3’•Is not actually part of the primer, but exists in the vector

PAGE 83

71 Figure 4-3. Garden Blot. A) Et hidium bromide stained gel cont aining total DNA from different xanthomonads, and plasmid DNA from B69 an d B69.4. B-D) Southern blot from gel in A, hybridized with a probe derived from B) orf115, C) orf116, and D) orf117 To test plasmid stability of marker-integrated plasmids integrated, integrated strains were cultured by inoculating 1 L of an overnight liquid starter culture (i n liquid PYGM+Sp35+Kn20) into 10 mL liquid PYGM with Sp35 (all strains used are Spr) and with Sp35Kn20 (selection for

PAGE 84

72 the plasmid). After 4 days (approximately 20 gene rations) a dilution series of the cultures was plated on PYGM plates with (Sp35+Cm35, Sp35K n20) and without (Sp35) antibiotic selection. Figure 4-4. Total DNA digested with Eco RI, Southern blot hybridi zed consecutively with probes derived from orfs 115, 116, and orf117

PAGE 85

73 Plant Inoculations Inoculations were done as described in Chapter 2. All inoculations were repeated at least three times. Inoculated leaves were examined on 7 or 8 days post inoc ulation on Mexican Lime, and on 9 or 10 days post inocul ation on Duncan Grapefruit. Plasmid Stability in planta Two-day old liquid bacterial cultures were wash ed and then adjusted to an OD of 0.4 with CaCO3-saturated water, and inoculated on Key Lime and Duncan Grapefruit. Inoculation patches were removed and ground out immediately after inoc ulation (half of the leaves), and 8 days post inoculation (dpi) for the second half of the leaves. For grindi ng out, three leaf punches (each from a separate leaf inoculated with the same culture) were obtained with a number 3 cork borer (0.84 cm2 total area), and ground in 1 mL CaCO3-saturated water with 1:5000 (v/v) Silwet). Dilution series were spotted (10 L per spot ) on non-selective (PYGM+S p35) and non-selective (PYGM+Sp35+Kn20) plates, and counted after 4 days incubation at 30 C. Results Southern Blot Hybridiz ation with orfs115-117 Total DNA from various xanthomonads was hybrid ized with probes made from orf115, orf116, and orf117. In agreement with the fact th at strain B69.4 is cured of pXcB, B69.4 DNA does not hybridize to orfs115-117. As shown in Figur e 4-3, in addition to one copy of each of the orfs 115-117 on plasmid pXcB, faint bands hybri dizing to orf115 and orf117 are present in X. phaseoli var. fuscans. The hybridizing Eco RI fragments are of approximately the same size as the fragment on pXcB, but are much weaker suggesting they might be located on the chromosome. A small Eco RI fragment in X.c. phaseoli also hybridizes to orf117.

PAGE 86

74 Figure 4-5. Orientations for the integration of the suicide vector in pXcB. Not drawn to scale Figure 4-6. Mapping of the lacZ promoter orientation in pB16.1, pB23.1-11, pB26.4.0, and pB31.2.1

PAGE 87

75 Marker Integrations and Mapping Orf116 and orf117 were interrupted using pUFR012. A third marker integration was carried out, resulting in pUFR 004 and pUFR012 integrations dow nstream of orf115, followed by a duplicate but incomplete copy of orf115 (l acking the 5'-end of orf115). Total DNA was isolated from all integration derivatives, digested with Eco RI, and after electrophoresis hybridized consecutively with probes derived from orf115, 116, and 117 (see Figure 4-4). In B69 Sp2, orf115 through 117 are all located on the same 12.6-kb Eco RI fragment. B69.4 is cured of pXcB and does not hybridize to a ny one of the orfs used as a probe. In B16.1 two fragments (8811 bp and 4293 bp) hybridize to orf115, while orf116 and orf117 are located on the same 8811 bp Eco RI fragments. B31.2.1 and B31.2.3, both marker integrations of orf116, each have two fragments (8253 bp and 4679 bp) that hybridi ze to orf116, orf115 is located on the 4679-bp fragment, while orf117 is located on th e 8253-bp fragment. B26.4.0 and B26.8.0, marker integrations of orf117, have a 7952-bp fragme nt that hybridizes to orf117, and a 4953-bp fragment that hybridizes to orf115, orf116, and orf117. The introduction of a Hin dIII site with the suicide vector allowed the direction of the lacZ promoter to be determined (Figure 4-5). The orie ntation in which the suicide vector integrated into pXcB was determined, by transf erring the integrat ed plasmids to E. coli DH5 (MCR), and restriction digestion with Hin dIII (Figure 4-6). Pathogenicity Tests Plasmid pAB29 was constructe d from pB16.1 (Figure 4-7). A Bgl II restriction site from the integrated suicide vector pUFR004 and a Bgl II site just upstream of gene pthB were used to clone a fragment of pXcB, including all orfs starting from pthB through orf115 into the Bam HI

PAGE 88

76 site of pUFR047. This clone was mated into the cured B-strain, B69.4, and the resulting strain was used to inoculate lime and grapefruit leav es. No canker phenotype was seen (Figure 4-8). Figure 4-7. Construction of pAB29, containing pthB through orf115 in pUFR047 Figure 4-8. Inoculation result of plasmid pAB29 mated into B69.4 and inoculated on Duncan Grapefruit

PAGE 89

77 Figure 4-9. The lacZ promoter is functional in the B-stra in. Strain BIM2, a marker integrated mutant of pthB and B21.2, a marker exchange mutant of pthA are complemented by pAB18.1, a clone of pthA driven by the lacZ promoter in pUFR047 Strain B16.1 and B23.1-11, which contain an in tact and a partial copy of orf115, were both non-pathogenic on lime and grapefruit. This result is surpri sing, since all genes are left intact. In the case of B16.1, the lacZ promoter runs in the reverse di rection and there could be a polar effect of the insertion on orf116 a nd orf117 if they are driven off the same promoter as orf115. If orf116 and orf117 are driven off a separate promot er (as predicted, if or f116 and orf117 encode a toxin-antitoxin system), the marker integrati on should have no effect on the transcription of orf116 and orf117. In the case of B23.1-11, the lacZ promoter runs in the forward direction, and should therefore have no effect on phenotype at all. B31.2.1 and B31.2.3 resulted from marker inte gration of orf116. The orientation of B31.2.1 was determined, and lacZ runs in the forward direction. Since the lacZ promoter is functional in the B69 derivatives ( pthA driven by the lacZ promoter complements BIM2, a pthB marker integrated mutant; see Figure 4-9) If the lacZ promoter runs in the forward direction, the insertions should not have a polar effect on orf117.

PAGE 90

78 The marker-integrated mutants of orf117 ga ve inconsistent results upon inoculation. Initially two colonies that resulted from marker integration of pAB26.4 (B26.4.0 and B26.4.1) and two colonies that resulted from marker integration of pAB26.8 (B26.8.0 and B26.8.1) gave canker on lime (Figure 4-10). Subsequent inoculati ons most often did not result in canker, but occasionally a weak or fairly strong canker phenot ype could be seen. This result could indicate that orf117 is required under cert ain (at this time unknown) conditions. Figure 4-10. Results of initial inocul ation of orf117 marker integrations Clones used for attempts at complementing the mutant phenotypes are shown in Figure 411. Plasmid pAB33.1 contains orf115 through orf117. Plasmid pAB35.3 was cloned using primers AB46 and AB47 to place orf116 under control of the lacZ promoter in pUFR047, pAV7.3 (Vanamala & Gabriel, unpublished) contai ns orf116 and orf117 under control of of the lacZ promoter. Plasmid pAB36.6 was cloned by revers ing the insert from pAB12.2 so that the lacZ promoter runs in the same direction as orfs115 through 117. Plasmid pJR7.1 (Reddy &

PAGE 91

79 Gabriel, unpublished) was cloned fro m pXcB to contain the entire virB cluster, orfs100, and 113 through 117, and includes the putative pr omoter region in front of orf100. Figure 4-11. Clones used for attempts to comp lement maker-integrated mutants and the regions of plasmid pXcB they represent Figure 4-12. Attempts to complement mark er-integrated mutants with plasmid pAB36.6 However, repeated attempt to mate pJR7.1 in to the marker-integrate d B-strains, and hold selection for both the integrated plasmid and pJR 7.1 resulted in no or very few colonies. Clone pJR8.2 was derived from pJR7.1 by deletion of the Eco RI fragment at the 5'end of the clone, so that pJR8.2 has the 3'-end of virB10 through or f117. Contrary to pJR7.1, pJR8.2 could be mated

PAGE 92

80 efficiently into all the marker-integrated strains. None of the above-mentioned clones complemented any of the marker-integrated mutant s tested. Results for attempts to complement with clones pAB36.6 and pJR8.2 are shown in Figure 4-12 and Figure 4-13, respectively. Figure 4-13. Attempts to complement mark er-integrated mutants with plasmid pJR8.2 Plasmid Stability Tests in vitro The mutant phenotype of the orf116 and orf 117 marker-integrated mutants could be the result of orf116 and orf117 being pathogenicity f actors. An alternative hy pothesis is that orf116 and orf117 form a toxin-antitoxin system and are responsible for pXcB stab ility. This was tested in vitro by growing the marker-integrated bacterial strains in liquid medi um with and without selection for the marker-integrated plasmid for 4 days (approximately 20 generations). Dilution series of each of the cultures were then plated on medium with (Sp35Kn20) and without (Sp35) antibiotic selection for the mark er-integrated plasmid. The data are given in Table 4-4. The number of cfuÂ’s on plates with antibiotic sele ction for the marker-integrated plasmid (Cm35 or

PAGE 93

81 Kn20) was not less than the numbe r of cfuÂ’s without selection fo r the plasmids. Therefore the marker-integrated plasmids appear to be stable in vitro Table 4-4. Stability tests fo r marker-integrated plasmids in vitro grown in liquid culture with and without antibiotic selection for the marker-integrated plasmid Strain Plated on Grown on Sp35Kn20 Grown on Sp35 Sp35 610-8 Sp35Cm35 0 B69Sp2 Sp35Kn20 0 Sp35 110-9 210-9 Sp35Cm35 210-9 210-9 B23.1-11 Sp35Kn20 210-9 210-9 Sp35 310-9 110-9 Sp35Cm35 210-9 210-9 B31.2.1 Sp35Kn20 210-9 210-9 Sp35 610-8 610-8 Sp35Cm35 210-9 210-9 B26.4.0 Sp35Kn20 210-9 110-9 Table 4-5. Stability test s for mutant plasmids in vitro alone, and together with vector pUFR047 or complementation clone (plasmid pAB 36.6). The marker-integrated plasmids ar Cm35 and Kn20 resistant, while the empty v ector and the complementation clone are Gm3 resistant Strain Selection for: Number of cfuÂ’s w/pUFR047 Number of cfuÂ’s w/pAB36.6 210-7 610-7 Kn20 0 0 Gm3 210-7 410-7 B69Sp2 Kn20Gm3 0 0 310-7 310-6 Kn20 310-7 710-6 Gm3 810-7 110-6 B23.1-11 Kn20Gm3 710-6 310-7 310-6 310-7 Kn20 610-6 410-6 Gm3 110-7 510-7 B31.2.1 Kn20Gm3 310-6 510-6 610-6 110-7 Kn20 710-6 210-7 Gm3 410-7 810-7 B26.4.0 Kn20Gm3 610-6 110-7 Stability of the marker-integrated plasmids and complementation plasmid when placed together in the same strain was then tested. The po ssibility exists that eith er one of the plasmids by itself will be stably maintained, but that an incompatibility between the two plasmids causes

PAGE 94

82 one or both to become unstable. The experi ment was carried out with strains B23.1-11, B26.4.0, and B31.2.1. Complementation plasmid pA B36.6 was mated into each strain, and restreaked on selective plates (PYGM+Kn20+Gm 3). As a control, the empty vector pUFR047 was also mated into each strain. The results are given in Table 4-5. The number of cfuÂ’s on selective plates was no less than the number of co lonies on non-selective pl ates. This means that virtually all of the strains maintained both plasmids separately, and together over ~20 generations. No loss of either one of the plasmi ds, or the plasmids together was observed. This demonstrated that in vitro both the mutant plasmids and complementation plasmid pAB36.6 were stably maintained in B69. Stability Tests in planta To test plasmid stability in planta bacterial cultures were inoculated and ground out immediately and 8 dpi. Dilution series were then plated on selective a nd non-selective plates. Results are summarized in Table 4-6. No significant loss of the marker-integrated plasmids was observed. The marker-integrated plasmids appear to be stable in planta Table 4-6. Stability tests fo r marker-integrated plasmids in planta Strain Selection for: Number of cfuÂ’s 0 dpi Number of cfuÂ’s 8 dpi Sp35 110-6 510-7 Sp35Cm35 210-4 510-3 B69Sp2 Sp35Kn20 210-4 210-3 Sp35 510-6 110-7 Sp35Cm35 510-6 310-6 B23.1-11 Sp35Kn20 410-6 110-6 Sp35 110-6 410-6 Sp35Cm35 210-6 210-5 B31.2.1 Sp35Kn20 210-6 110-5 Sp35 110-5 210-5 Sp35Cm35 210-5 610-5 B26.4.0 Sp35Kn20 110-5 210-5

PAGE 95

83 Discussion Three marker-interrupted mutants each were made of orf116 and orf117. The third markerintegrated mutant was predicted to be fully pat hogenic, since orf115, the intended target was left intact. If the lacZ promoter is in the forward direction, it is possible that it can drive expression of downstream genes, making the mutation non-polar. If the lacZ promoter is in the reverse direction, the mutation would be a polar mutati on. The suicide vector was integrated in two different orientations in tw o independent marker-integrated mutants, and since the lacZ promoter is functional in X. citri pv. aurantifolii, it shoul d be able to drive the or fs downstream of orf115 if it runs in the forward orientation. There is the possi bility that the lacZ is unable to drive orf116 and orf117, if those genes are expressed from th ere own (possibly tightly regulated) promoter. The marker-integrated mutant of orf117 was fu lly pathogenic when firs t inoculated. During subsequent inoculations the phenot ype was inconsistent. It appears that orf117 is not required for pathogenicity of B69. One explanation for th e mutant phenotypes of each of the markerintegrated mutants is that spontaneous mutations elsewhere on the plasmid or in the genome have occurred which rendered the st rain non-pathogenic. There were no obvious deviations from the expected restriction patterns and hybridiza tion patterns for either of the mutants. All attempts to complement the mutant phe notypes were unsuccessful, which supports the hypothesis that additional mutations have occurre d. It seems unlikely that such spontaneous mutations would have occurred in all of the independent marke r-integrated mutants. It would have been prudent to generate non-polar mutations from the start, instead of relying on possibly polar mutations using marker integrations. An entir e suicide vector is integrated into the native plasmid, and it is possible that unfor eseen interactions take place.

PAGE 96

84 Plasmid pAB29, contains pthB and the entire virB cluster with its own promoter, through orf115. When this clone was mated into B69.4 (cured of pXcB), the resulting strain transconjugant did not cause canke r. It is therefore possible th at the necessary factor(s) for pathogenicity are located between orf115 and pthB Since orf116 and orf117 resemble reported plasmid stability systems, the hypothesis that the marker-integrated plasmi ds were unstable in the X. citri strains was tested. It was shown that the marker-integrated plasmids are stable in the mutant strains in culture and in planta The possibility that the clones used for complementation were incompatible with the markerintegrated plasmids was also tested in culture. No instability of either the mutant plasmid or the complementation plasmids, or of the two plasmids together, was seen in cu lture. This experiment was not done in planta and might have been worthwhile, since the instability/incompatibility may only be an issue under stress (low nutrition) conditions, as likely to be encountered in the plant.

PAGE 97

85 CHAPTER 5 CONCLUSIONS Previous experiments have shown that pthB located on native plasmid pXcB in X. citri is necessary but insufficient to caus e canker on citrus. Subclones of pBIM2, a marker-integrated mutant of pX cB, were used together with pthA in a different shuttle vector in attempts to complement the mu tant phenotype. The clones were mated into B69.4/pAB2.1, the strain cured of pXcB with pthA added back into it. None of the resulting strains caused canker on either Key Lime or Duncan Grapefruit. The factor(s) necessary for pathogenicity are most likely inte rrupted in the subclones, or there is more than one factor required, and the two were cloned into separate fragments. The complete sequence of pXcB revealed an intact T4SS, similar to the T4SS found on pXAC64, the native plasmid of X. citri pv. citri (syn. X. axonopodis pv. citri). T4S systems are responsible for bacterial c onjugation, and experiments have shown that the T4SS of pXcB is required for self-mob ilization (El Yacoubi and Gabriel, manuscript in preparation). The T4SS on pXcB and pXAC64 differ slightly from that of A. tumefaciens On pXcB and pXAC64, the virB1 homologue is at the end of the virB cluster instead of at the be ginning, and the order of the virB6 and virB7 genes is not the same. In addition, pXcB and pXAC64 have one additional orf between virB6 and virB7 Remarkably, although the nucleo tide for the rest of the virB cluster of pXcB shows very high similarity to that of the pXAC64 virB cluster, the region between the virB5 and the virB6 gene shows a much lower degree of sim ilarity. This does not hold true for the

PAGE 98

86 similarity at the translated level. This implie s that this part of th e sequence has evolved at a different rate than the rest of the virB cluster. Although many similarities exist between pXAC64 and pXcB, there are some differences. pXAC has two members of the avrBs3/pthA gene family, including pthA whereas pXcB has only one, pthB Plasmid pXcB has five additional orfs immediately downstream of the virB cluster that could be transcri bed from the same promoter, but pXAC64 has no homologues of any of these gene s. These orfs may therefore be suited for identification purposes. A garden blot using different xanthomonads shows that except for a weak band in X. phaseoli var. fuscans, orf115 is unique to the B strain, and orf116 seems to be completely unique. Two marker-interruption mu tants post pathogenicity on Key Lime and Duncan Grapefruit, but this phenotype could not be complemented using the complementation clones designed for this purpose. A third a ttempt at marker interruption, resulting in integration but no gene disruption, also lost pa thogenicity on citrus. No instability of the marker-interrupted plasmids or the comp lementation plasmid was observed in liquid culture and in planta In addition, no incompatibility between the marker-interrupted plasmids and the complementation plasmid was evident in liquid culture.

PAGE 99

87 APPENDIX A MAP OF POSITION 13301 TO 15700 Orfs are highlighted as follows: Prim ers used for marker integration: orf113: red (frame a) not done orf114: yellow (frame a) not done orf115: green (frame b) AB01 and AB02 orf116: blue (frame a) AB43 and AB44 orf117: violet (frame c) AB35 and AB36 TTCGTGCATTGACCGGCTTTTCTCGCATGCGAGAAAGCTAACTCAGGAGAAGAAGGCATG 13301 ---------+---------+---------+---------+---------+---------+ 13360 a F V H M b c R A L T G F S R M R E S AAAAAACTGATAGTCGCGCTGTTTATCGGTGCTGCTGCTGTGTCCGCGCACGCCTCCGAA 13361 ---------+---------+---------+---------+---------+---------+ 13420 a K K L I V A L F I G A A A V S A H A S E b c TATGGGTGCAAGGTGCTGCTGTGCCTTGCCAATCCCGCGTCCAATGGCGGCCCGAAGGGC 13421 ---------+---------+---------+---------+---------+---------+ 13480 a Y G C K V L L C L A N P A S N G G P K G b M G A R C C C A L P I P R P M A A R R A c GTGTCCGAATGCGTTCCCCCCATCGATCAGCTCTACCACGATCTCAGCAAGGGGCGGCCG 13481 ---------+---------+---------+---------+---------+---------+ 13540 a V S E C V P P I D Q L Y H D L S K G R P b C P N A F P P S I S S T T I S A R G G R c M R S P H R S A L P R S Q Q G A A V TTCCCGACGTGCGATCTCGCGGACGGCAATGATGGTTCGAGCTACGCCCGGCAGGTCTAT 13541 ---------+---------+---------+---------+---------+---------+ 13600 a F P T C D L A D G N D G S S Y A R Q V Y b S R R A I S R T A M M V R A T P G R S M c P D V R S R G R Q GACCCGTATGACCCTTGCCCGGCCCCGTTGCAACCTGCCGCACGCGGTTCCTATGTCGTG 13601 ---------+---------+---------+---------+---------+---------+ 13660 a D P Y D P C P A P L Q P A A R G S Y V V b T R M T L A R P R C N L P H A V P M S C c

PAGE 100

88 CAAGGGCAGAAGAAAACGGGCGGCAACAAGCCGGGATGGTGGGGCGGCGACGGCTCGTAC 13661 ---------+---------+---------+---------+---------+---------+ 13720 a Q G Q K K T G G N K P G W W G G D G S Y b K G R R K R A A T S R D G G A A T A R T c M V G R R R L V H ACGCTGAGCGGACAGCCGCAAGTGTCTCAGTCGCAAAGCGACTACGGCCATAGCTCGGGT 13721 ---------+---------+---------+---------+---------+---------+ 13780 a T L S G Q P Q V S Q S Q S D Y G H S S G b R c A E R T A A S V S V A K R L R P GCGCGGGCCTGCGTCGGGAAGTCGGTCGGCTCATACACCGTAGGCAGCTACGACAGCAGC 13781 ---------+---------+---------+---------+---------+---------+ 13840 a A R A C V G K S V G S Y T V G S Y D S S b c GACACCGTGGACGTGTTCGACAAGGTGGTCTGGCAGCCGGCGCAGAATCCGCGAGCCATC 13841 ---------+---------+---------+---------+---------+---------+ 13900 a D T V D V F D K V V W Q P A Q N P R A I b c GACGTGTTCATCGACAACACGTGGCAGCAGCGCGTGCGCTGGTAAGCGGGGGTGGCAATG 13901 ---------+---------+---------+---------+---------+---------+ 13960 a D V F I D N T W Q Q R V R W M b c CGTCGATCTATCGTCTGCGCGGCTCTGTTGGCCCTCGCCTCTCTGGCGGGGCTCAACAGC 13961 ---------+---------+---------+---------+---------+---------+ 14020 a R R S I V C A A L L A L A S L A G L N S b c TTCACGGTCGGCCTGTTGGACAAGGTTCGCAACACGGTGGCCGCCGAGCCGGCCAGCGCC 14021 ---------+---------+---------+---------+---------+---------+ 14080 a F T V G L L D K V R N T V A A E P A S A b c CCGGACACGCAGACTGTCGAGGTCGCTTTCTCGCCGGACGGTCGGGCCGAAGCGTTGGTG 14081 ---------+---------+---------+---------+---------+---------+ 14140 a P D T Q T V E V A F S P D G R A E A L V b c CTCAAGGTCATTCGTGCCGCGAAAACGTCAATCCGCTTGGCCGGCTACACCTTCACGTCG 14141 ---------+---------+---------+---------+---------+---------+ 14200 a L K V I R A A K T S I R L A G Y T F T S b c CCGGCCGTCGTGCGCGCCCTGACCGATGCCAAGAAGCGCGGTGTCGATGTGGCTGTCGTG 14201 ---------+---------+---------+---------+---------+---------+ 14260 a P A V V R A L T D A K K R G V D V A V V b M P R S A V S M W L S W c

PAGE 101

89 GTGGACTACAAGAACAATCTCGTGGAAAGTCGCTCGAACACATCGCGGCAAGCGCTCAAC 14261 ---------+---------+---------+---------+---------+---------+ 14320 a V D Y K N N L V E S R S N T S R Q A L N b W T T R T I S W K V A R T H R G K R S T c CTGCTGGTGCATGCCGGCATCCCGACTCGTACGGTCAGCGTCTACCCGATCCACCACGAC 14321 ---------+---------+---------+---------+---------+---------+ 14380 a L L V H A G I P T R T V S V Y P I H H D b C W C M P A S R L V R S A S T R S T T T c AAATACATCGTGGCCGACGGCCTGCACGTCGAAACCGGGAGCTTCAATTACACGATGGCC 14381 ---------+---------+---------+---------+---------+---------+ 14440 a K Y I V A D G L H V E T G S F N Y T M A b N T S W P T A C T S K P G A S I T R W P c GCCGCTCGCCGTAACAGCGAGAACGTGCTTGTGGTGTGGAACAACGCGGCGGTAGCCGGC 14441 ---------+---------+---------+---------+---------+---------+ 14500 a A A R R N S E N V L V V W N N A A V A G b P L A V T A R T C L W C G T T R R c CAGTACATTGCGCACTGGCAAAGCCGTTGGGGGCAGGGCCAGCCGTATCAATCCAATGAC 14501 ---------+---------+---------+---------+---------+---------+ 14560 a Q Y I A H W Q S R W G Q G Q P Y Q S N D b M T c AB01 -----------------> TAGCGTAGGAGTGGAAAGAATGGAAGTGACAAACAACGAAATCGCAGAACGACGCGCTAT 14561 ---------+---------+---------+---------+---------+---------+ 14620 a b S V G V E R M E V T N N E I A E R R A M c GTTCGAGGCTCTGAATCGTGAAACGTCGGAAGCCGCCGTTGCTTTTCGGGACTACTGGCG 14621 ---------+---------+---------+---------+---------+---------+ 14680 a b F E A L N R E T S E A A V A F R D Y W R c CAAGGAGGGTGAGCGGTGGAGGGAGGCGGACAAGCTCGTGAAGTCGGTTGTGCGCGGTGA 14681 ---------+---------+---------+---------+---------+---------+ 14740 a b K E G E R W R E A D K L V K S V V R G E c AACACCGGTGACGGATGAGGTAATTGAAAACTGCCGGCTCGCGGTCATGCGGCTGCATCA 14741 ---------+---------+---------+---------+---------+---------+ 14800 a b T P V T D E V I E N C R L A V M R L H Q c M R

PAGE 102

90 GATTCGCTACGTGCTCGGACAATTGATGCTGCCTGTCGGATCGCCTTCGCCGGGCGTCCT 14801 ---------+---------+---------+---------+---------+---------+ 14860 a b I R Y V L G Q L M L P V G S P S P G V L c TGATGCCGTGATGGACGTTATCCGGCGCATCGGGAATGACACCACGGAAGCGACGAAGAA 14861 ---------+---------+---------+---------+---------+---------+ 14920 a b D A V M D V I R R I G N D T T E A T K N c M P M T P R K R R R I TCTCAAGTTGATCCTCGGATGGCAGGCTGCCGCGCGAATTGAGGTCGAGCCGGAGTAAGC 14921 ---------+---------+---------+---------+---------+---------+ 14980 a M A G C R A N b L K L I L G W Q A A A R I E V E P E c S S AB02 <-------------CCCTCTGAGAGCGGTTGTCGCACTACCCCAAAGGCGCTTCATGCGCCTTTGCCGTTTTTG 14981 ---------+---------+---------+---------+---------+---------+ 15040 GAGACTCTCGCCAAC a b M R L C R F W c AB47 ---------GCGAGATAGTTTCAACTCATTCGGCAATGTACTACGATGTACTCTCGATGTAACACATAA 15041 ---------+---------+---------+---------+---------+---------+ 15100 a M Y S R C N T b R D S F N S F G N V L R C T L D V T H K c M Y Y D V L S M AB43 -------> -----------------> GGTGGCGCCATGAGCGAAGCAACCTTTACATTCCGAGTCGATGAGGCCCTGAAAAGTGAG 15101 ---------+---------+---------+---------+---------+---------+ 15160 a M S E A T F T F R V D E A L K S E b V A P M R P c TTTTCCACGGCCGCGAAGGCCCACGACCGCACGGGCGCGCAGCTCTTGCGCGACTTCATG 15161 ---------+---------+---------+---------+---------+---------+ 15220 a F S T A A K A H D R T G A Q L L R D F M b c CGGGACTATGTGAAGCAGCAACAGGAAGCGGCAGAATACAACGCATGGCTACGCGCCAAG 15221 ---------+---------+---------+---------+---------+---------+ 15280 a R D Y V K Q Q Q E A A E Y N A W L R A K b M c M A T R Q G GTCGAGCGCTCACGGGCCTCAGCCAACGCCGGTCATCTGGTGCCGACCGCAGAGGTCGAA 15281 ---------+---------+---------+---------+---------+---------+ 15340 a V E R S R A S A N A G H L V P T A E V E b c R A L T G L S Q R R S S G A D R R G R S

PAGE 103

91 AB44 <--------------<---GCCAAATTCGCAGCGCGTCGAGCTGCGACACGTCGCCGGCTTGAAGCGTCGAAGTGATCG 15341 ---------+---------+---------+---------+---------+---------+ 15400 CTGTGCAGCGGCCGA a A K F A A R R A A T R R R L E A S K b c Q I R S A S S C D T S P A M I E AB46 AB35 <---------------------------> AGTTGTTCTGGACGCCGGAGGCGATACAAGATCGCGACGACATTTACGACTACATTGAGG 15401 ---------+---------+---------+---------+---------+---------+ 15460 a b c L F W T P E A I Q D R D D I Y D Y I E A CTGACAACCCGGTTGCCGCGCTGGACCTTGATGAGCTGTTCGAGGAGAAAGCGGCGCTGC 15461 ---------+---------+---------+---------+---------+---------+ 15520 a M S C S R R K R R C b c D N P V A A L D L D E L F E E K A A L L TGGTCGATCATCCGAGTCTAGGCCGGGTTGGCCGCGTAGCGGGCACGTGCGAACTGGTCG 15521 ---------+---------+---------+---------+---------+---------+ 15580 a W S I I R V b c V D H P S L G R V G R V A G T C E L V A CGCACCGTAGTTACTTACTGATCTATGACGTGGCCGGTGACTTGGTGCGTGTGTTGAACG 15581 ---------+---------+---------+---------+---------+---------+ 15640 a M T W P V T W C V C b c H R S Y L L I Y D V A G D L V R V L N V AB36 <--------------TGGTCCATACGTCCCGCCAATGGCCGCCCGTCCGAGAGTAGTCCAGGCGGCCACGACTGC 15641 ---------+---------+---------+---------+---------+---------+ 15700 CAGGTATGCAGGGCGG a b M A A R P R V V Q A A T T A c V H T S R Q W P P V R E

PAGE 104

APPENDIX B CODONPREFERENCE FOR POSITION 13301 TO 16000

PAGE 105

93 13,500 14,000 14,500 15,000 15,500 16,000 16,000 15,500 15,000 14,500 14,000 13,500 2.0 1.5 1.0 0.5 -0.0 2.0 1.5 1.0 0.5 -0.0 2.0 1.5 1.0 0.5 -0.0 0.0 0.5 1.0 0.0 0.5 1.0 0.0 0.5 1.0 ---------orf113 -----------------orf114-----------orf116-------orf115-------orf117-

PAGE 106

94 LIST OF REFERENCES Aldon, D, Brito, B, Boucher, C, & Genin, S. (2000). A bacterial sensor of plant cell contact controls the tran scriptional induction of Ralstonia solanacearum pathogenicity genes. EMBO J 19 2304-2314. Aldridge, P & Hughes, KT (2001) How and when are substrat es selected for type III secretion? Trends Microbiol 9 209-213. Al-Saadi, A & Gabriel, DW (2002). Mo lecular characterization of a new Xanthomonas citri strain isolated in Florida. Phytopathology 92 S3. Altschul, SF, Gish, W, Miller, W, Myer s, EW, & Lipman, DJ (1990). Basic local alignment search tool. J Mol Biol 215 403-410. Alvarez, AM, Benedict, AA, Mizumoto, CY, Hunter, JE, & Gabriel, DW (1994). Serological, pathological and ge netic diversity among strains of Xanthomonas campestris infecting crucifers. Phytopathology 84 1449-1457. Anderson, DM & Schneewind, O (2001). A mRNA Signal for the Type III Secretion of Yop Proteins by Yersinia enterocolitica Science 278 1140-1143. Anderson, LB, Hertzel, AV, & Das, A (1996). Agrobacterium tumefaciens VirB7 and VirB9 form a disulfide-linked protein complex. Proc Natl Acad Sci 93 8889-8894. Barber, CE, Tang, J-L, Feng, J-X, Pan, MQ Wilson, T.J., Slater, H., Dow, J.M., Williams, P., & Daniels, M.J. (1997). A novel regulatory system required for pathogenicity of Xanthomonas campestris is mediated by a small diffusible signal molecule. Mol Microbiol 24 555-566. Baron, C, O'Callaghan, D, & Lanka, E (2002) Bacterial secrets of secretion: EuroConference on the biology of t ype IV secretion processes. Mol Microbiol 43 1359-1365. Baron, C, Thorstenson, YR, & Zambryski, PC (1997). The lipoprotein VirB7 interacts with VirB9 in the membranes of Agrobacterium tumefaciens. J Bacteriol 179 1211-1218. Bayer College of Medicine HGSC (2003). BCM Search Launcher. http://searchlauncher.bcm.tmc.e du/seq-search/gene-search.html last accessed December 25, 2004.

PAGE 107

95 Berger, BR & Christie, PJ (1994). Geneti c complementation analysis of the Agrobacterium tumefaciens virB operon: virB2 through virB11 are essential virulence genes. J Bacteriol 176 3646-3660. Bogdanove, AJ, Bauer, DW, & Beer, SV (1998). Erwinia amylovora Secretes DspE, a Pathogenicity Factor and Functional Av rE Homolog, through the Hrp (Type III Secretion) Pathway. J Bacteriol 180 2244-2247. Bogdanove, AJ, Wei, JM, Zhao, L, & Beer, SV (1996). Erwinia amylovora secretes harpin via a type III pathwa y and contains a homolog of yopN of Yersinia spp. J Bacteriol 178 1720-1730. Bonas, U, Fenselau, S, Minsavage, GV, Stas kawicz, BJ, & Stall, RE (1991). Isolation of a Gene Cluster from Xanthomonas campestris pv. vesicatoria that Determines Pathogenicity and the Hypersensiti ve Response on Pepper and Tomato. Mol PlantMicrobe Interact 4 81-88. Boyer, HW & 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, AM & Gabriel, DW (2003) Xanthomonas citri : breaking the surface. Mol. Plant Pathol. 4 141-157. Buttner, D & Bonas, U (2002). Getting across--b acterial type III effector proteins on their way to the plant cell. EMBO J 21 5313-5322. Cabezn, E, Sastre, JI, & De la Cruz, F ( 1997). Genetic evidence of a coupling role for the TraG protein family in bacterial conjugation. Mol Gen Genet 254 400-406. Cascales, E & Christie, PJ (2003). The versat ile bacterial type IV secretion systems. Nat Rev Microbiol 1 137-149. Cascales, E & Christie, PJ (2004). Definition of a Bacterial Type IV Secretion Pathway for a DNA Substrate. Science 304 1170-1173. Casper-Lindley, C, Dahlbeck, D, Clar k, ET, & Staskawicz, BJ (2002). Direct biochemical evidence for type III secreti on-dependent translocation of the AvrBs2 effector protein into plant cells. Proc Natl Acad Sci 99 8336-8341. Chakrabarty, PK, Duan, YP, & Gabriel, DW (1997). Cloning and Ch aracterization of a Member of the Xanthomonas avr/pth Gene Family That Evades All Commercially Utilized Cotton R Genes in the United States. Phytopathology 87 1160-1167. Cheng, LW, Anderson, DM, & Schneewind, O (1997). Two independent type III secretion mechanisms for YopE in Yersinia enterocolitica Mol Microbiol 24 757765.

PAGE 108

96 Christie, PJ (1997). Agrobacterium T-Complex Transport Appa ratus: a Paradigm for a New Family of Multifunctional Transporters in Eubacteria. J Bacteriol 179 30853094. Christie, PJ (2001). Type IV secretion: in tercellular transfer of macromolecules by systems ancestrally related to conjugation machines. Mol Microbiol 40 294-305. Christie, PJ & Vogel, JP (2000) Bacterial type IV secreti on: conjugation systems adapted to deliver effector molecules to host cells. Trends Microbiol 8 354-360. Civerolo, EL (1985). Indigenous Plasmids in Xanthomonas campestris pv. citri. Phytopathology 75 524-528. Cooper, TF & Heinemann, JA (2000). Postsegr egational killing does not increase plasmid stability but acts to mediate the exclusion of competing plasmids. Proc Natl Acad Sci 97 12643-12648. da Silva, AC, Ferro, JA, Reinach, FC, Farah, CS, Furlan, LR, Quaggio, RB, MonteiroVitorello, CB, Van Sluys, MA, Almeida, NF, Alves, LM, do Amaral, AM, & Bertolini, MC (2002). Comparison of the genomes of two Xanthomonas pathogens with differing host specificities. Nature 417 459-463. Daefler, S (1999). Type III secretion by Salmonella typhimurium does not require contact with a eukaryotic host. Mol Microbiol 31 45-52. Deane, SM & Rawlings, DE (2004). Plasmi d evolution and interaction between the plasmid addiction stability systems of two related broad-hos t-range IncQ-like plasmids. J Bacteriol 186 2123-2133. De Feyter, R, Kado, CI, & Gabriel, DW (1990) Small stable shuttle vectors for use in Xanthomonas Gene 88 65-72. De Feyter, R, Yang, Y, & Gabriel, DW ( 1993). Gene-for-genes interactions between cotton R genes and Xanthomonas campestris pv. malvacearum avr genes. Mol Plant-Microbe Interact 6 225-237. Dow, JM, Feng, J-X, Barber, CE, Tang, J-L, & Daniels, MJ (2000). Novel genes involved in the regulation of pathogenicity factor production w ithin the rpf gene cluster of Xanthomonas campestris Microbiol 146 885-891. Duan, YP, Castaneda, AL, Zhao, G, Erdos, G, & Gabriel, DW (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. Eberl, L, Givskov, M, & Schwab, H (1992) The divergent promoters mediating transcription of the par locus of plasmid RP4 are subject to autoregulation. Mol Microbiol 6 1969-1979.

PAGE 109

97 El Yacoubi, B, Brunings, AM, Yuan, Q, & Ga briel, DW (2004). A self-transmissibile plasmid from Xanthomonas campestris carries a member of the avr/pth gene family and additional factor(s) require d for pathogenicity. (Abstract) Swiss Institute of Bioinformatics (2 003). ExPASy Proteomics Server. http://us.expasy.org/ last accessed December 25, 2004. Farinha, MA, Conway, BD, Glasier, LM, Ellert, NW, Irvin, RT, Sherburne, R, & Paranchych, W (1994). Alteration of the pilin adhesin of Pseudomonas aeruginosa PAO results in normal pilus biogenesi s but a loss of adherence to human pneumocyte cells and decreased virulence in mice. Infect Immun 62 4118-4123. Federal Register (1990). Citrus Ca nker Regulations. Proposed Rules. Federal Register 55 11209-11220. Fenselau, S, Balbo, I, & Bonas, U ( 1992). Determinants of Pathogenicity in Xanthomonas campestris pv. vesicatoria are Related to Proteins Involved in Secretion in Bacterial Pathogens of Animals. Mol Plant-Microbe Interact 5 390-396. Fenselau, S & Bonas, U (1995). Sequence and expression analys is of the hrpB pathogenicity operon of Xanthomonas campestris pv. vesicatoria which encodes eight proteins with similarity to component s of the Hrp, Ysc, Spa, and Fli secretion systems. Mol Plant-Microbe Interact 8 845-854. Fernandez, D, Dang, TA, Spudich, GM, Zhou, X, Berger, BR, & Christie, PJ (1996). The Agrobacterium tumefaciens virB7 gene product, a proposed component of the Tcomplex transport apparatus, is a membra ne-associated lipoprote in exposed at the periplasmic surface. J Bacteriol 178 3156-3167. Figurski, DH & Helinski, DR (1979). Replica tion of an origin-containing derivative of plasmid RK2 dependent on a plasmid function provided in trans Proc Natl Acad Sci 76 1648-1652. Finlay, BB & Falkow, S (1997). Common themes in microbial pathogenicity revisited. Microbiol Mol Biol Rev 61 139-169. Florida Citrus Mutual (2004). Citrus Canker Bulletin, April 2004 update. http://www.flcitrusmutual.com/content/docs/issues/april%202004.pdf last accessed December 25, 2004. Gabriel, DW (1999a). The Xanthomonas avr/pth Gene Family. In: G. Stacey & N.T. Keen (Eds.), Plant-Microbe Interactions (pp. 39-55). St. Paul, Minnesota: APS Press. Gabriel, DW (1999b). Why do pathoge ns carry avirulence genes? Phys Mol Plant Path 55 205-214.

PAGE 110

98 Gabriel, DW (2001). Citrus Canker. In : O.C. Maloy & T.D. Murray (Eds.), Encyclopedia if Plant Pahtology (pp. 215-217). New York: John Wiley and Sons. Gabriel, DW (2002). Citrus Canker Disease. http://www.biotech.ufl.edu/Pl antContainment/canker.htm last accessed December 25, 2004. Gabriel, DW, Kingsley, M, Hunter, JE, & Gottwald, TR (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. Galn, JE & Collmer, A (1999). Type III Secr etion Machines: Bacterial Devices for Protein Delivery into Host Cells. Science 284 1322-1328. Gerdes, K (2000). Toxin-antitoxin modules ma y regulate synthesis of macromolecules during nutritional stress. J Bacteriol 182 561-572. Gerdes, K, Nielsen, A, Thorsted, P, & Wa gner, EG (1992). Mechanism of killer gene activation. Antisense RNA-dependent RNas e III cleavage ensures rapid turn-over of the stable hok, srnB and pndA effector messenger RNAs. J Mol Biol 226 637649. Gerdes, K, Rasmussen, PB, & Molin, S (1986) Unique type of plasmid maintenance function: postsegregational killing of plasmid-free cells. Proc Natl Acad Sci 83 3116-3120. Gerdes, K, Thisted, T, & Martinussen, J (1990) Mechanism of post-segregational killing by the hok/sok system of plasmid R1: sok antisense RNA regulates formation of a hok mRNA species correlated with killing of plasmid-free cells. Mol Microbiol 4 1807-1818. Ginocchio, CC, Olmsted, SB, Wells, CL, & Ga ln, JE (1994). Cont act with epithelial cells induces the formation of surface appendages on Salmonella typhimurium Cell 76 717-724. Gomis-Ruth, FX, Moncalin, G, & De la Cruz, F (2002). Conjugative plasmid protein TrwB, an integral membrane type IV s ecretion system coupling protein. Detailed structural features and mappi ng of the active site cleft. J Biol Chem 277 75567566. Goto, M (1969). Studies on citrus canker in Japan. Proc First Int Citrus Symp 3 12511252. Gottwald, TR, Graham, JH, Civerolo, EL, Barre tt, HC, & Hearn, CJ (1993). Differential host reaction of citrus and citrus relatives to citrus canker and ci trus bacterial spot determined by mesophyll susceptibility. Plant Disease 77 1004-1009.

PAGE 111

99 Gottwald, TR, Graham, JH, & Schubert, TS (20 02). Citrus Canker: The pathogen and its impact. Online. Plant Health Progress doi 10.1094/PHP-2002-0812-01-RV. Gottwald, TR, Hughes, G, Graham, JH, Sun, X, & Riley, T (2001). The Citrus Canker Epidemic in Florida: The Scientific Basi s of Regulatory Eradication Policy for an Invasive Species. Phytopathology 91 30-34. Graham, JH, Gottwald, TR, Cubero, J, and Achor, DS (2004) Xanthomonas axonopodis pv. citri : factors affecting successful er adication of citrus canker. Mol. Plant Pathol. 5 1-15. Graupner, S & Wackernagel, W (2001). Pseudomonas stutzeri has two closely related pilA genes (type IV pilus sdtr uctural protein) with opposite influences on natural genetic transformation. J Bacteriol 183 2359-2366. Greated, A, Lambertson, L, Williams, PA, & Thomas, CM (2001). Complete nucleotide sequence of IncP-9 plasmid pWW0 (UnPub) Guttman, DS, Vinatzer, BA, Sarkar, SF, Rana ll, MV, Kettler, G, & Greenberg, JT (2002). A functional screen for the type III (H rp) secretome of the plant pathogen Pseudomonas syringae Science 295 1722-1726. Hacker, J & Kaper, JB (2000). Pathogenicit y islands and the evolution of microbes. Ann Rev Microbiol 54 641-679. Hahn, HP (1997). The type-4 pilus is the major virulence-asso ciated adhesin of Pseudomonas aeruginosa--a review. Gene 192 99-108. Hapfelmeier, S, Domke, N, Zambryski, PC & Baron, C (2000). VirB6 Is Required for Stabilization of VirB5 and VirB3 and Formation of VirB7 Homodimers in Agrobacterium tumefaciens J Bacteriol 182 4505-4511. Hayes, F (1998). A family of stability determinants in pathogenic bacteria. J Bacteriol 180 6415-6418. Henderson, IR, Cappello, R, & Nataro, JP (2000) Autotransporter prot eins, evolution and redefining protein secretion. Trends Microbiol 8 529-532. Henderson, IR & Nataro, JP ( 2001). Virulence functions of autotransporter proteins. Infect Immun 69 1231-1243. Hobbs, M, Collie, ES, Free, PD, Livingston, SP & Mattick, JS (1993). PilS and PilR, a two-component transcriptional regulatory system controlling expression of type 4 fimbriae in Pseudomonas aeruginosa Mol Microbiol 7 669-682.

PAGE 112

100 K. Hofmann & W. Stoffel (1993). TMbase A database of membrane spanning proteins segments. Biol Chem. Hoppe-Seyler 374 ,166. http://www.ch.embnet.org/software/TMPRED_form.html last accessed December 25, 2004. Hopkins, CM, White, FF, Choi, S-H, and Leach, JE (1992). Identification of a Family of Avirulence Genes from Xanthomonas oryzae pv. oryzae Mol Plant-Microbe Interact 5 451-459. Hueck, CJ (1998). Type III Protein Secretion Sy stems in Bacterial Pathogens of Animals and Plants. Microbiol Mol Biol Rev 62 379-433. Hultgren, SJ, Abraham, S, Caparon, M, Falk, P, St Geme, J, & Normak, S (1993). Pilus and nonpilus bacterial adhesins: assemb ly and function in cell recognition. Cell 73 887-901. Innes, RW, Bent, AW, Kunkel, BN, Bisgrove SR, & Staskawicz, BJ (1993). Molecular analysis of avirulence gene avrRpt2 and identification of a putative regulatory sequence common to all known Pseudomonas syringae avirulence genes. J Bacteriol 175 4859-4869. Jones, AL, Lai, E, Shirasu, K, & Kado, CI (1996). VirB2 is a processed pilin-like protein encoded by the Agrobacterium tumefaciens Ti plasmid. J Bacteriol 178 57065711. Kanamori, H & Tsuyumu, S (1998). Comparison of Nucleotide Sequences of Cankerforming and Non-canker-forming pthA Homologues in Xanthomonas campestris pv. citri. Ann Phytopathol Soc Jpn 64 462-470. Kang, Y, Liu, H, Genin, S, Schell, MA, & Denny, TP (2002). Ralstonia solanacearum requires type 4 pili to adhere to multipl e surfaces and for natu ral transformation and virulence. Mol Microbiol 46 427-437. Kholodii, G, Yurieva, O, Mindlin, S, Go rlenko, Z, Rybochkin, V, & Nikiforov, V (2000). Tn5044, a novel Tn3 family transposon codi ng for temperature-sensitive mercury resistance. Res Microbiol 151 291-302. Knoop, V, Staskawicz, BJ, & Bonas, U (1991). Expression of the avirulence gene avrBs3 from Xanthomonas campestris pv. vesicatoria is not under the control of hrp genes and is independent of plant factors. J Bacteriol 173 7142-7150. Krall, L, Wiedemann, U, Unsin, G, Weiss, S, Domke, N, & Baron, C (2002). Detergent extraction identifies different VirB protein subassemblies of the type IV secretion machinery in the membranes of Agrobacterium tumefaciens Proc Natl Acad Sci 99 11405-11410.

PAGE 113

101 Kyoto University Bioinformatics Center (2004 ). MOTIF: Searching Protein and Nucleic Acid Sequence Motifs. http://motif.genome.ad.jp/ last accessed December 25, 2004. Lai, E & Kado, CI (1998). Processed VirB2 Is the Major Subunit of the Promiscuous Pilus of Agrobacterium tumefaciens J Bacteriol 180 2711-2717. Lai, E & Kado, CI (2000). The T-pilus of Agrobacterium tumefaciens Trends Microbiol 8 361-369. Lazo, GR & Gabriel, DW (1987). Cons ervation of Plasmid DNA Sequences and Pathovar Identification of Strains of Xanthomonas campestris Phytopathology 77 448-453. Leach, JE & White, FF (1996). Bacterial avirulence genes. Annu Rev Phytopathol 34 179. Lee, VT & Schneewind, O (2001). Protein secr etion and the pathogenesis of bacterial infections. Genes & Dev 15 1725-1752. Lehnherr, H, Maguin, E, Jafri, S, & Yarmolin sky, MB (1993). Plasmid addiction genes of bacteriophage P1: doc, which causes cel l death on curing of prophage, and phd, which prevents host death when prophage is retained. J Mol Biol 233 414-428. Leite, RP, Jr, Egel, DS, & Stall, RE (1994). Genetic analysis of hrp -related DNA sequences of Xanthomonas campestris strains causing di seases of citrus. Appl Environ Microbiol 60 1078-1086. Liu, CC, Huhne, R, Tu, J, Lorbach, E, & Droge, P (1998). The resolvase encoded by Xanthomonas campestris transposable element ISXc5 constitutes a new subfamily closely related to DNA invertases. Genes Cells 3 221-233. Liu, CC, Wang, H, Chou, HC, Chang, WT, & Tu, J (1992). Analysis of the genes and gene products of Xanthomonas transposable elements ISXc5 and ISXc4. Gene 120 99-103. Madan BM & Sankaran, K (2002) DOLOP database of bacterial lipoproteins, Bioinformatics 18 : 641-643. http://www.mrc-lmb.cam.ac.uk/genomes/dolop/dolop.htm last accessed December 25, 2004. Magnuson, R, Lehnherr, H, Mukhopadhyay, G, & Yarmolinsky, MB (1996). Autoregulation of the plasmid addi ction operon of bacteriophage P1. J Biol Chem 271 18705-18710. Magnuson, R & Yarmolinsky, MB (1998). Corepr ession of the P1 addiction operon by Phd and Doc. J Bacteriol 180 6342-6351.

PAGE 114

102 Marchler-Bauer, A, Panchenko, AR, Shoemake r, BA, Thiessen, PA, Geer, LY, & Bryant, SH (2002). CDD: a database of conserved domain alignments with links to domain three-dimensional structure. Nucl Acids Res 30 281-283. McLean, F & Lee, A (1922). The re sistance of citr us canker to Citrus nobilis and a suggestion as to the production of re sistant varieties on other species. Phytopathol 11 109-115. Menard, R, Sansonetti, PJ, & Pa rsot, C (1994). The secretion of the Shigella flexneri Ipa invasins is activated by epithelial ce lls and controlled by IpaB and IpaD. EMBO J 13 5293-5302. Miao, EA & Miller, SI (2000). A conserved amino acid sequence dir ecting intracellular type III secretion by Salmonella typhimurium Proc Natl Acad Sci 97 7539-7544. Moncalin, G, Cabezn, E, Alkorta, I, Valle, M, Moro, F, Valpuesta, JM, Go i, FM, & De la Cruz, F (1999). Characterizati on of ATP and DNA Binding Activities of TrwB, the Coupling Protein Essen tial in Plasmid R388 Conjugation. J Biol Chem 274 36117-36124. Mudgett, MB, Chesnokova, O, Dahlbeck, D, Clark, ET, Rossier, O, Bonas, U, & Staskawicz, BJ (2000). Molecular signals required for type III secretion and translocation of the Xanthomonas campestris AvrBs2 protein to pepper plants. Proc Natl Acad Sci 97 13324-13329. Nakai, K & Horton, P (1999). PSORT: a program for detecting the sorting signals of proteins and predicting thei r subcellular localization, Trends Biochem Sci, 24 (1) 34-35 (1999). http://psort.nibb.ac.jp/ last accessed December 25, 2004. National Center for Biotechnology Inform ation (2004). ORF Finder (Open Reading Frame Finder). http://www.ncbi.nlm.nih.gov/gorf/gorf.html last accessed December 15, 2004. Nol, L, Thieme, F, Nennstiel, D, & B onas, U (2002). Two novel type III-secreted proteins of Xanthomonas campestris pv. vesicatoria are en coded within the hrp pathogenicity island. J Bacteriol 184 1340-1348. Parge, HE, Forest, KT, Hickey, MJ, Christen sen, DA, Gertzoff, ED, & Tainer, JA (1995). Structure of the fibre-forming protein pilin at 2.6 A resolution. Nature 378 32-38. Peet, RC, Lindgren, PB, Willis, DK, and Pa napoulos, NJ (1986) Identification and cloning of genes involved in phaseolotoxin production by Pseudomonas syringae pv. phaseolicola. J Bacteriol 166 1096-1105. Pettersson, J, Nordfelth, R, Dubinina, E, Bergman, T, Gustafsson, M, Magnusson, KE, & Wolf-Watz, H (1996). Modulation of Vi rulence Factor Expression by Pathogen Target Cell Contact. Science 273 1231-1233.

PAGE 115

103 Pohlman, RF, Genetti, HD, & Winans, SC (1994) Entry exclusion of the IncN plasmid pKM101 is mediated by a single hydrophilic protein containing a lipid attachment motif. Plasmid 31 158-165. Radnedge, L, Davis, MA, Youngren, B, & Austin, SJ (1997). Plasmid maintenance functions of the large virulence plasmid of Shigella flexneri. J Bacteriol 179 3670-3675. Rawlings, DE (1999). Proteic toxin-antitoxin, bacterial plasmid addiction systems and their evolution with special refere nce to the pas system of pTF-FC2. FEMS Microbiol Lett 176 269-277. Ray, SK, Pajeshwari, R, Sharma, Y, & Son ti, RV (2002). A high-molecular-weight outer membrane protein of Xanthomonas oryzae pv. oryzae exhibits similarity to nonfimbrial adhesins of animal pathogenic bacteria and is required for optimum virulence. Mol Microbiol 46 637-64. Roberts, RC & Helinski, DR (1992). Definition of a minimal plasmid stabilization system from the broad-host-range plasmid RK2. J Bacteriol 174 8119-8132. Roine, E, Wei, W, Yuan, J, Nurmiaho-La ssila, E, Kalkkinen, N, Romantschuk, M, & He, SY (1997). Hrp pilus: An hrp -dependent bacterial surface appendage produced by Pseudomonas syringae pv. tomato DC3000. Proc Natl Acad Sci 94 3459-3464. Rosqvist, R, Magnusson, KE, & Wolf-Watz, H (1994). Target cell contact triggers expression and polarized tran sfer of Yersinia YopE cytotoxin into mammalian cells. EMBO J 13 964-972. Rossier, O, Wengelnik, K, Ha hn, K, & Bonas, U (1999). The Xanthomonas Hrp type III system secretes proteins from plan t and mammalian bacterial pathogens. Proc Natl Acad Sci 96 9368-9373. Ruiz-Echevarria, MJ, Gimenez-Gallego, G, Sa bariegos-Jareno, R, & Diaz-Orejas, R (1995). Kid, a small protein of the parD stability system of plasmid R1, is an inhibitor of DNA replication acting at the initiation of DNA synthesis. J Mol Biol 247 568-577. Russel, M (1998). Macromolecular assembly and secretion across the bacterial cell envelope: type II protein secretion systems. J Mol Biol 279 485-499. Sambrook, J, Fritsch, EF, & Maniatis, TA (1989). Molecular cloning: A Laboratory Manual New York: Cold Spring Harbor. Sandkvist, M (2001). Biology of type II secretion. Mol Microbiol 40 271-283. Sayeed, S, Reaves, L, Radnedge, L, & Austin, SJ (2000). The stability region of the large virulence plasmid of Shigella flexneri en codes an efficient po stsegregational killing system. J Bacteriol 182 2416-2421.

PAGE 116

104 Schoulties, CL, Civerolo, EL, Miller, JW, Stall, RE, Krass, CJ, Poe, SR, & Ducharme, EP (1987). Citrus Canker in Florida. Plant Disease 71 388-393. Schubert, TS, Rizvi, SA, Sun, X, Gottw ald, TR, Graham, JH, & Dixon, WN (2001). Meeting the Challenge of Eradicating Citrus Canker in Florida Again. Plant Disease 85 340-356. Segal, G, Russo, JJ, & Shuman, H (1999). Rela tionships between a new type IV secretion system and the icm/dot virulence system of Legionella pneumophila Mol Microbiol 34 799-809. Shine, J & Dalgarno, L (1974). The 3'-ter minal sequence of Escherichia coli 16S ribosomal RNA: complementarity to nonsense triplets and ribosome binding sites. Proc Natl Acad Sci 71 1342-1346. Shirasu, K, Koukolikova-Nicola, Z, Hohn, B, & Kado, CI (1994). An inner-membraneassociated virulence protein essential for T-DNA transfer from Agrobacterium tumefaciens to plants exhibits ATPase activity and similarities to conjugative transfer genes. Mol Microbiol 11 581-588. Sieira, R, Comerci, DJ, Snchez, DO, & Ugalde, R (2000). A Homo logue of an Operon Required for DNA transfer in Agrobacterium Is Required in Brucella abortus for Virulence and Intracel lular Multiplication. J Bacteriol 182 4849-4855. Skerker, JM & Berg, HC (2001). Direct obser vation of extension and retraction of type IV pili. Proc Natl Acad Sci 98 6901-604. Smith, AS & Rawlings, DE (1998). Autore gulation of the pTF-FC2 proteic poisonantidote plasmid addiction system (pas) is essential for plasmid stabilization. J Bacteriol 180 5463-5465. Smith, JA & Magnuson, R (2004). Modular organi zation of the Phd re pressor/antitoxin protein. J Bacteriol 186 2692-2698. Spudich, GM, Fernandez, D, Zhou, X, & Chri stie, PJ (1996). Intermolecular disulfide bonds stabilize VirB7 homodimers and VirB7/VirB9 heterodimers during biogenesis of the Agrobacterium tumefaciens T-complex transport apparatus. Proc Natl Acad Sci 93 7512-7517. Stall, RE & Civerolo, EL (1991). Research Relating to the Recent Outbreak of Citrus Canker in Florida. Ann Rev Phytopathol 29 399-420. Stall, RE, Marco, GM, & Canteros de Eche nique, BI (1982). Importance of mesophyll in mature-leaf resistance to cancrosis of citrus. Phytopathology 72 1097-1100. Staskawicz, BJ, Dahlbeck, D, Keen, N, & Napoli, C (1987). Molecular characterization of cloned avirulence genes from race 0 and race 1 of Pseudomonas syringae pv. glycinea J Bacteriol 169 5789-5794.

PAGE 117

105 Sun, X, Stall, RE, Cubero, J, Gottwald, TR, Graham, JH, Dixon, WN, Schubert, TS, Peacock, ME, Dickstein, ER, & Chaloux, PH (2000). Detection of a unique isolate of citrus canker bacterium from Key Lime in Wellington and Lake Worth, Florida. (Abstract) Swarup, S, De Feyter, R, Brlansky, RH, & Ga briel, DW (1991). A Pathogenicity Locus from Xanthomonas citri Enables Strains from Several Pathovars of X. campestris to Elicit Cankerlike Lesions on Citrus. Phytopathology 81 802-809. Swarup, S, Yang, Y, Kingsley, MT & Gabriel, DW (1992). A Xanthomonas citri pathogenicity gene, pthA pleiotropically encodes gr atuitous avirulence on nonhosts. Mol Plant-Microbe Interact 5 204-213. Szurek, B, Marois, E, Bonas, U, & Van De r Ackerveken, G (2001). Eukaryotic features of the Xanthomonas type III effector AvrBs3: protein domains involved in transcriptional activation and the interact ion with nuclear import receptors from pepper. The Plant J 26 523-534. Szymanik, M, Bartosik, D, & Wlodarczyk, M (2004). Genetic organization of the basic replicon of plasmid pMTH4 of a facu ltatively methylotrophic bacterium Paracoccus methylutens DM12. Curr Microbiol 48 291-294. Tang, J-L, Lin, YN, Barber, CE, Dow, JM, Wootton, JC, & Daniel s, MJ (1991). Genetic and molecular analysis of a cluster of rpf genes involved in positive regulation of synthesis of extracellular enzymes a nd polysaccharide in Xanthomonas campestris pathovar campestris. Mol Gen Genet 226 409-417. Tu, J, Wang, H, Chang, S, Charng, Y, Lu rz, R, Dobrinski, B, & Wu, W (1989). Transposable elements of Xanthomonas campestris pv. citri originating from indigenous plasmids. Mol Gen Genet 217 505-510. Unniraman, S, Prakash, R, & Nagaraja, V (2001). Alternate Paradigm for Intrinsic Transcription Termination in Eubacteria. J Biol Chem 276 41850-41855. Unniraman, S, Prakash, R, & Nagaraja, V ( 2002). Conserved economics of transcription termination in eubacteria. Nucl Acids Res 30 675-684. Vergunst, AC, Schrammeijer, B, den Bulk -Ras, A, de Vlaam, CMT, Regensburg-Tu nk, TJG, & Hooykaas, PJJ (2000). VirB/D4-De pendent Protein Translocation from Agrobacterium into Plant Cells. Science 290 979-982. Verniere, C, Hartung, JS, Pruvost, OP, Cive rolo, EL, Alvarez, AM, Maestri, P, & Luisetti, J (1998). Characterization of phenotypically distinct strains of Xanthomonas axonopodis pv. citri from Southwest Asia. Europ J Plant Pathol 104 477-487. Wall, D & Kaiser, D (1999). Type IV pili and cell motility. Mol Microbiol 32 1-10.

PAGE 118

106 Ward, JE, Dale, EM, Christie, PJ, Nester EW, & Binns, AN (1990). Complementation analysis of Agrobacterium tumefaciens Ti Plasmid virB Genes by Use of a vir promoter Expression Vector: virB9 virB10 and virB11 Are Essential Virulence Genes. J Bacteriol 172 5187-5199. Watarai, M, Tobe, T, Yoshikawa, M, & Sa sakawa, C (1995). Contact of Shigella with host cells triggers release of Ipa invasins and is an essential function of invasiveness. EMBO J 14 2461-2470. Wattiau, P, Bernier, B, Deslee, P, Michie ls, T, & Cornelis, GR (1994). Individual chaperones required for Yop secretion by Yersinia. Proc Natl Acad Sci 91 1049110497. Wattiau, P, Woestyn, S, & Cornelis, GR ( 1996). Customized secretion chaperones in pathogenic bacteria. Mol Microbiol 20 255-262. Wengelnik, K & Bonas, U (1996). HrpXv, an Ar aC-type regulator, activates expression of five of the six loci in the hrp cluster of Xanthomonas campestris pv. vesicatoria. J Bacteriol 178 3462-3469. Wengelnik, K, Rossier, O, & Bonas, U (1999). Mu tations in the regula tory gene hrpG of Xanthomonas campestris pv. vesicatoria result in cons titutive expression of all hrp genes. J Bacteriol 181 6828-6831. Wengelnik, K, Van Der Ackerveken, G, & B onas, U (1996). HrpG, a key hrp regulatory protein of Xanthomonas campestris pv. vesicatoria is homologous to twocomponent response regulators. Mol Plant-Microbe Interact 9 704-712. Wilson, TJG, Bertrand, N, Tang, J-L, Feng, J-X, Pan, MQ, Barber, CE, Dow, JM, & Daniels, MJ (1998). The rpfA gene of Xanthomonas campestris pv. campestris which is involved in regulation of pa thogenicity factor production encodes an aconitase. Mol Microbiol 28 961-970. Winans, SC, Burns, DL, & Christie, PJ (1996) Adaptation of a conjugal transfer system for the export of pathogenic macromolecules. Trends Microbiol 4 64-68. Wood, DW, Setubal, JC, Kaul, R, Monks, DE, Kitajima, JP, Okura, VK, Zhou, Y, Chen, L, Wood, GE, Almeida, NFJ, Woo, L, Che n, Y, Paulsen, IT, Eisen, JA, Karp, PD, Bovee, DS, Chapman, P, & Clendenning, J (2001). The genome of the natural genetic engineer Agrobacterium tumefaciens C58. Science 294 2317-2323. Wu, SS & Kaiser, D (1997). Regulation of expression of the pilA gene in Myxococcus xanthus J Bacteriol 179 7748-7758. Yang, Y, De Feyter, R, & Gabriel, DW ( 1994). Host-specific Symptoms and Increased Release of Xanthomonas citri and X. campestris pv. malvacearum from Leaves Are Determined by the 102-bp Tandem Repeats of pthA and avrb6 Respectively. Mol Plant-Microbe Interact 7 345-355.

PAGE 119

107 Yang, Y & Gabriel, DW (1995a). Intragenic recombination of a si ngle plant pathogen gene provides a mechanism for the e volution of new host specificities. J Bacteriol 177 4963-4968. Yang, Y & Gabriel, DW (1995b). Xanthomonas avirulence/pathogenicity gene family encodes functional plant nuc lear targeting signals. Mol Plant-Microbe Interact 8 627-631. Yeo, H & Waksman, G (2004). Unveiling Molecula r Scaffolds of the Type IV Secretion System. J Bacteriol 186 1919-1926. Zhu, W, Yang, B, Chittoor, JM, Johnson, LB, & White, FF (1998). AvrXa10 Contains an Acidic Transcriptional Activation Doma in in the Functionally Conserved C Terminus. Mol Plant-Microbe Interact 11 824-832.

PAGE 120

108 BIOGRAPHICAL SKETCH Asha Marcelle Brunings completed her B. S. at the Anton de Kom University of Suriname in 1994, with a major in agronomy. She then spent 2 years on the Oil Palm Plantation Victoria (in Suriname’s interior) as head of the Research Department, breeding interspecific oil palm hybrids for resi stance against Lethal Yellowing. In 1998, she started the Plant Molecular Biology graduate pr ogram with Dr. Dean Gabriel as adviser. From August 1996 through August 1998, she was funded by the Organization of American States (OAS). In December 2004, she defended her master’s thesis “In Search of Pathogenicity Factors of Xanthomonas citri pv. aurantifolii.”