Title: Molecular characterization of bacteriocin-like activity in tomato race-three strains of Xanthomonas campestris pv. vesicatoria
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00100711/00001
 Material Information
Title: Molecular characterization of bacteriocin-like activity in tomato race-three strains of Xanthomonas campestris pv. vesicatoria
Physical Description: Book
Language: English
Creator: Tudor, Simone Michelle, 1969-
Publisher: University of Florida
Place of Publication: Gainesville Fla
Gainesville, Fla
Publication Date: 1999
Copyright Date: 1999
 Subjects
Subject: Plant Pathology thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Plant Pathology -- UF   ( lcsh )
Genre: government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )
 Notes
Summary: ABSTRACT: Tomato race 3 (T3) strains of Xanthomonas campestris pv. vesicatoria are inhibitory to other strains within the species. Ten cosmid clones were isolated from a T3 strain genomic library that were selected for their ability to inhibit a sensitive indicator strain in plate assays. Southern hybridization analysis grouped these into three subsets of overlapping activity- BCN-A+, BCN-B+, and BCN-C+. Groups could be differentiated by variations in inhibitory spectrum and levels of activity in plate assays. Transposon generated mutants deficient in each were used to mutate individual wild-type loci. Marker-exchange mutants all retained inhibitory activity. Two N-methyl-N'-nitro-N-nitrosoguanidine (NTG) mutants deficient in wild-type BCN+ activity were obtained. Restriction enzyme mapping, transposon mapping and subcloning were used to localize the genes responsible for expression of the BCN-A+ phenotype to a 8.0-kb KpnI/EcoRI fragment. Expression of this in a BCN-A- strain conferred both inhibitory activity and immunity to BCN-A+ (IMM-A+). Sequence analysis of the region revealed a cluster of seven open reading frames (ORF). The largest ORF (bcnA), approximately 3.6-kb, was required for BCN-A+ activity. IMM-A+ was encoded by an ORF downstream of this.
Summary: ABSTRACT (cont.): Homology searches with bcnA revealed no homology with any known genes involved in bacteriocin activity; however, it was homologous to wapA and Rhs elements. Both of these contain multiple copies of an almost identical ligand-binding motif. Seven copies of a similar motif were found in bcnA. BCN-A+ activity was found to be analogous to bacteriocin activity, and the name vesicacin A is proposed for the gene product encoded by bcnA. A bcnA-specific probe was homologous to genomic DNA of T3 strains and a bacteriocin-producing strain of X. c. pv. glycines. Tn3-gus::bcnA translational fusions were used to demonstrate expression of this gene in planta and in selected media types. T3 strains were inhibitory to a sensitive indicator strain in planta, but this effect was only observed when T3 strains were applied in advance of the sensitive strain. NTG-mutant analysis showed that production of these antimicrobial compounds was important in mediating the observed antagonism. Further analysis revealed vesicacin A to be an essential and dominant component in the suppression of the sensitive indicator strain in tomato leaf tissues.
Thesis: Thesis (Ph. D.)--University of Florida, 1999.
Bibliography: Includes bibliographical references (p. 108-120).
System Details: System requirements: World Wide Web browser and Adobe Acrobat Reader to view and print PDF files.
System Details: Mode of access: World Wide Web.
Statement of Responsibility: by Simone Michelle Tudor.
General Note: Title from first page of PDF file.
General Note: Document formatted into pages; contains vi, 121 p.; also contains graphics (some colored).
General Note: Vita.
 Record Information
Bibliographic ID: UF00100711
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 50751502
alephbibnum - 002531486
notis - AMP7408

Downloads

This item has the following downloads:

tudor ( PDF )


Full Text











MOLECULAR CHARACTERIZATION OF BACTERIOCIN-LIKE ACTIVITY IN
TOMATO RACE-THREE STRAINS OF XANTHOMONAS
CAMPESTRIS PV. VESICATORIA
















By

SIMONE MICHELLE TUDOR


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1999















ACKNOWLEDGMENTS

My sincere thanks go out to Dr. R. E. Stall, who served as the chair of my committee

for most of my time here. I have appreciated and enjoyed all of the time that he spent

sharing his ideas and passing on his wealth of knowledge. I also would like to thank the

new chair of my committee, Dr. J. B. Jones, for his interesting ideas, constant

enthusiasm, and the support which he gave to me. Appreciation also goes to Dr. D. R.

Pring, and J. F. Preston III who served on my supervisory committee. I also would like

to thank Dr. H. C. Kistler, who served on my committee for four years before leaving for

a new position.

I would like to specially thank Jerry Minsavage, for his willingness to share

generously of his time and ideas, and Ellen Dickstein, who gave me constant support.

The faculty and staff of the Department of Plant Pathology were very supportive

throughout my program, providing assistance whenever it was needed. I also would like

to thank the many friends that I have met during my stay here. I have enjoyed their

friendship and support.

Words cannot express the appreciation that I feel towards my family. My mother and

father have been a constant support throughout my life. I thank them for their love and

steadfast belief in me. Finally, to my beloved husband, Stephen, I owe my deepest

appreciation. I would like to thank him for his patience, love, and support throughout

my time here.
















TABLE OF CONTENTS


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

A B S T R A C T .............................................................................................................. v

CHAPTERS

1 IN TR O D U C TIO N .. ...................................................................... . .... 1

2 LITER A TU R E REV IEW .................................................................... 7

3 PRODUCTION OF MULTIPLE BACTERIOCIN-LIKE COMPOUNDS ..... 27

M materials and M ethods .................................................................................. 29
Bacterial Strains, Plasmids and Culture Conditions ......................... 29
In Vitro A ntagonism A says ........................... ...................... .......... 29
In Planta A ntagonism A says .................................................. .......... 30
D N A M anipulations ........................................................ ............... 3 1
Screening of C osm id Library .......................... ...................... .......... 31
Southern H ybridization A analysis ............................................. .......... 32
Transposon M utagenesis ...................................................... ......... 33
M arker Exchange M utagenesis .......................................................... 33
Chemical Mutagenesis and Complementation Analysis ................... 34
R e su lts .............. .................................................................................. . . ..... 3 4
Screen of C osm id Library ............................. ........................ .......... 34
Differences Between BCN+ Groups ................................................... 35
M utagenesis ............................................................................. ......... 40
In P lanta Studies ......... .. .......................... .................................... 40
D iscu ssio n ..................................................................................................... 4 5

4 CHARACTERIZATION OF GENETIC DETERMINANTS AND
EVALUATION OF THEIR ROLE IN ANTIBIOSIS ........................... 49

M materials and M ethods .................................. ..................... ............. ........ 51
Bacterial Strains, Plasmids and Culture Conditions ........................... 51
In Vitro Antagonism and Immunity Assays ........................................ 52
In Planta A ntagonism A says .................................................. .......... 52
D N A Extractions ......... .. .......................... .................................... 53









D N A M anipulations ........................................................ ............... 53
B acterial C onjugations .. ........................................................ ......... 54
Southern Hybridization Analysis ............................ ............ 54
Physical Characterization of BCN+ Cosmid Clones ............................ 55
DNA Sequencing ................................... ..... . ........ ......... ........... 55
Construction of Promoter Fusions and Assays for 13-glucuronidase
A ctiv ity ................................................................................. . ........ 5 8
RN A Extraction and R T-PCR ..................................................... ......... 59
Ammonium Sulfate Fractionation and Ultrafiltration .......................... 61
R results ..... .............. ................................................. .. ........ 61
Physical Characterization of BCN-A+ Clones ................................... 61
Sequence Analysis of pXV12.1 (BCN-A+) .................................. 62
Evidence for Transcription and Translation of bcnA .......................... 71
Homology Search Results and bcnA Characteristics .......................... 72
E expression of bcnA ........................................... ... .......... ....... ......... 75
Distribution and Genomic Localization of bcnA ................................ 77
Physical Characterization of BCN-B+ Clones .................................... 77
Physical Characterization of BCN-C Clone ..................................... 83
Sequence Analysis of pXV5.1 (BCN-C ............................................. 83
Relative Contribution of Bacteriocin-like Activities to In Planta
C om petition ................................................................................. 87
D iscu ssio n ..................................................................................................... 9 0

5 C O N C L U SIO N S .......................................................................... 100

APPENDICES

A BACTERIAL STRAINS AND PLASMIDS USED IN STUDY ........................ 104

B CUSTOM DESIGNED OLIGONUCLEOTIDES .............................................. 107

R E F E R E N C E S ........................................................... .. ...................................... 10 8

B IO G R A PH ICAL SK ETCH ........................................................................ 121















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

MOLECULAR CHARACTERIZATION OF BACTERIOCIN-LIKE
ACTIVITY IN TOMATO RACE-THREE STRAINS OF XANTHOMONAS
CAMPESTRIS PV. VESICATORIA

By

Simone Michelle Tudor

December 1999

Chairman: Dr. Jeffrey B. Jones
Major Department: Plant Pathology


Tomato race 3 (T3) strains of Xanthomonas campestris pv. vesicatoria are inhibitory

to other strains within the species. Ten cosmid clones were isolated from a T3 strain

genomic library that were selected for their ability to inhibit a sensitive indicator strain in

plate assays. Southern hybridization analysis grouped these into three subsets of

overlapping activity- BCN-A BCN-B and BCN-C Groups could be differentiated

by variations in inhibitory spectrum and levels of activity in plate assays. Transposon

generated mutants deficient in each were used to mutate individual wild-type loci.

Marker-exchange mutants all retained inhibitory activity. Two N-methyl-N'-nitro-N-

nitrosoguanidine (NTG) mutants deficient in wild-type BCN+ activity were obtained.

Restriction enzyme mapping, transposon mapping and subcloning were used to

localize the genes responsible for expression of the BCN-A+ phenotype to a 8.0-kb









Kpnl/EcoRI fragment. Expression of this in a BCN-A- strain conferred both inhibitory

activity and immunity to BCN-A+ (IMM-A+). Sequence analysis of the region revealed a

cluster of seven open reading frames (ORF). The largest ORF (bcnA), approximately

3.6-kb, was required for BCN-A+ activity. IMM-A+ was encoded by an ORF

downstream of this. Homology searches with bcnA revealed no homology with any

known genes involved in bacteriocin activity; however, it was homologous to wapA and

Rhs elements. Both of these contain multiple copies of an almost identical ligand-binding

motif. Seven copies of a similar motif were found in bcnA. BCN-A+ activity was found

to be analogous to bacteriocin activity, and the name vesicacin A is proposed for the

gene product encoded by bcnA. A bcnA-specific probe was homologous to genomic

DNA of T3 strains and a bacteriocin-producing strain of X c. pv. glycines. Tn3-

gus::bcnA translational fusions were used to demonstrate expression of this gene in

plant and in selected media types.

T3 strains were inhibitory to a sensitive indicator strain inplanta, but this effect was

only observed when T3 strains were applied in advance of the sensitive strain. NTG-

mutant analysis showed that production of these antimicrobial compounds was important

in mediating the observed antagonism. Further analysis revealed vesicacin A to be an

essential and dominant component in the suppression of the sensitive indicator strain in

tomato leaf tissues.















CHAPTER 1
INTRODUCTION

The importance of bacteria in plant diseases was not recognized until the late 1800s.

Common sentiment at this time was aptly described in a quote by Anton de Bary:

"according to the present state of our knowledge, parasitic bacteria are of little

importance as the contagion of plant diseases." Proponents who ascribed to this view

based their arguments on observations that bacteria were rarely isolated from diseased

tissue, that the nutritive poor conditions of the plant cell were insufficient to support

growth of bacteria, and that conditions within the plant cell were too acidic for bacterial

tolerance. Fischer and Smith hotly debated this issue in a series of exchanges published

in the late 1890s and early 1900s (Fischer and Smith, 1981). Bacteria are now known to

cause some of the most destructive of plant diseases (Kennedy and Lacy, 1982).

Bacterial spot disease caused by Xanthomonas campestris pv. vesicatoria (Doidge)

Dye was one of the earliest recorded bacterial diseases. It was first observed in 1914 in

South Africa and at around the same time in the U.S. (Jones et al., 1998c). Study of this

disease has continued since then, primarily because of the economic importance of the

crops it infects and the lack of effective control measures. Vegetable production in

Florida is a billion dollar industry, with bell peppers and fresh market tomatoes at the

forefront of the industry. Together they represent almost half the total earnings from

vegetable production in the state (USDA Statistical Service, 1998).









The etiological agent of this disease is a yellow-pigmented, Gram-negative, rod-

shaped bacterium. It affects all aerial portions of the plant, and causes water-soaking,

followed by characteristic necrotic leaf spots and in severe cases defoliation. There have

recently been major changes in the taxonomy of this organism that has resulted in the

renaming of the species. Extensive DNA-DNA hybridizations between strains allowed

the distinction of two DNA hybridization groups, A and B (Stall et al., 1994; Vauterin

et al., 1995). Group A strains were placed in X axonopodis and designated X

axonopodis pv. vesicatoria, while a new species name, X. vesicatoria, was created for

the group B strains (Vauterin et al., 1995). In some recent publications, the old

nomenclature is still the most widely used and will be used in this dissertation.

Despite the early observation of this disease and numerous studies focused on control

strategies, the disease still leads to major yield losses for these two crops (Pohronezny

and Volin, 1983). The most prevalent management practice is to spray plants with

copper compounds alone, or in combination with maneb or mancozeb (Conover and

Gerhold, 1981; Pohronezny et al., 1993). Effective control by this practice is limited by

the existence of high levels of copper resistance in field populations of X campestris pv.

vesicatoria. In one study, as many as 63% of strains sampled in field surveys were

copper resistant (Ritchie and Dittapongpitch, 1991). Additional limitations may occur as

the use of mancozeb in combination with copper reduces its fungicidal efficacy (Conover

and Gerhold, 1981; Pohronezny et al., 1986) and leads to increased incidences of late

blight and target spot on tomato. The same combination has also been shown to increase

the activity of aphids and armyworms (McCarter, 1992). Studies monitoring epiphytic

populations of X. campestris pv. vesicatoria have recently shown that bacteria may









escape lethal copper doses by residence in "protected sites" such as within buds

(Pernezny and Collins, 1997). High levels of copper resistance among natural X

campestris pv. vesicatoria strains, possible evasion of the bacteria to lethal copper doses,

increased incidences of other tomato diseases, and perhaps other factors may combine to

influence the lack of effective control by copper combinations, especially under

conditions of high disease pressure.

The use of resistant varieties containing single-gene resistances have been somewhat

successful at controlling this disease (Pohronezny et al., 1993); however, identification

of new populations of races able to overcome this type of resistance occurs frequently.

There are at least eight pepper and three tomato races that can be differentiated on host

genotypes (Jones et al., 1998c). One strategy proposed the use of mixed genotypes in

fields to minimize the effect of race variations (Kousik et al., 1996). In this study, plots

were planted with mixtures of susceptible and X campestris pv. vesicatoria resistant

(pepper races 1 and 2 resistant) genotypes and disease progress was followed While

disease attenuation was observed in these plots, exposure of bacterial strains to multiple

resistance genes may lead to the plausible development of a "super pathogen" capable of

overcoming multiple resistances. Efforts are now being focused on the integration of

quantitative resistance into commercially acceptable cultivars (Stall, personal

communication).

Cultural management practices, such as good field sanitation, are aimed at reducing

the spread of the pathogen. Pohronezny et al have shown that good field sanitation is

effective in controlling disease spread (Pohronezny et al., 1990); however, this strategy

is also ineffective when conditions are favorable for disease development.









There is a constant search for new ways to control this disease. Innovative

approaches such as the application of bacteriophage mixes have recently been shown to

reduce the severity of bacterial spot disease in greenhouses (Jones et al., 1998b).

One of the keys for designing an effective control strategy is a clear understanding of

pathogen processes, including; long-term survivability, dissemination and infection.

Numerous studies have been conducted on these processes of this pathosystem, and a

brief review of them follows.

With the exception of selected species, phytopathogenic bacteria survive poorly in the

environment when not in association with living hosts. Disease control efforts are often

focused on the elimination of survival reservoirs. Long-term survival ofX. campestris

pv. vesicatoria has been shown to occur primarily in crop residues, and volunteer plants

(Jones et al., 1986a). Survival on infested pepper seeds for up to 10 months has been

reported (Bashan et al., 1982). In soils artificially infested with X campestris pv.

vesicatoria soil survival is poor; detection of the bacterium only occurred for 16 days

(Bashan et al., 1982).

Dissemination ofX. campestris pv. vesicatoria is usually facilitated by wind-blown

rain (Volcani, 1969). Once in contact with the plant surface, mechanisms of attachment

are thought to facilitate colonization. Production of fimbriae or pili by X campestris pv.

vesicatoria is thought to mediate this attachment (Romantschuck, 1994). Colonization

of the host does not necessarily lead immediately to the diseased state. Leben introduced

the concept of a resident phase of bacterial colonization, in which pathogens are able to

replicate on the phyllosphere without causing visible symptoms (Leben, 1974).

Epiphytic colonization by X campestris pv. vesicatoria of pepper and tomato surfaces









has been well documented (Sharon et al., 1982; McGuire and Jones, 1991). Epiphytic

survival is dependent on environmental factors such as relative humidity, with X

campestris pv. vesicatoria increasing 10-100-fold on tomato leaves at high relative

humidity compared to numbers at low humidity (Timmer et al., 1987). The transition

between epiphyte and pathogen is not clear, and may suggest that an apparent infection

threshold in which epiphytic populations are high enough, must be reached before

infection will occur (McGuire and Jones, 1991). Others believe that successful invasion

of the plant tissue is enabled when conditions favorable for disease development occur

(Goodman, 1982).

Entry ofX. campestris pv. vesicatoria into the leaf occurs through wounds (Vakili,

1967) or natural openings such as stomata, hydathodes and lenticels. The importance of

stomata in infection was shown by Ramos and Volin (1987) who illustrated that disease

severity was significantly reduced by the physiological or chemical suppression of

stomatal opening. Entry may not lead to visible symptoms. Populations of X. campestris

pv. vesicatoria as high as 1. lx 107 CFU/g leaf tissue were recovered from symptomless

plants, raising the possibility of endophytic survival (Bashan et al., 1982).

The race compositions of bacterial populations are constantly changing. Distribution

of X. campestris pv. vesicatoria races vary over time. Factors which have been

implicated in such shifts include change in plant genotypes (Jones et al., 1998a),

introduction of new races on seeds or transplants (Pohronezny et al., 1992) and

differences in strain competitiveness (Stall et al., 1986). Antibiosis is one of the factors

commonly involved in microbial competition. In 1994, it was first reported that tomato

race three (T3) strains of X campestris pv. vesicatoria were inhibitory to T strains of the






6


same bacterium (El-Morsy et al., 1994). A subsequent study on this revealed that

production of bacteriocin-like compounds was responsible for inhibition (Tudor, 1995).

The objectives of the present study are aimed at further characterization of these

compounds including identification of the genes involved, and an examination of the

possible role that they play in mediating dynamics of plant populations.















CHAPTER 2
REVIEW OF LITERATURE

Substances involved in microbial warfare have been fascinating targets of study for

microbiologists. They encompass a wide variety of compounds with very different

structures and modes of action. They include low-molecular weight "classical"

antibiotics, which are usually synthesized by multienzyme complexes and are secondary

metabolites (Birch and Patil, 1985); bacteriolytic enzymes such as the endo-beta-N-

acetylglucosaminidases secreted by staphylococcal species (Valisena et al., 1982);

heamolysins (Fath and Kolter, 1993); bacteriophages (Vidaver, 1976); and a special

class of compounds called bacteriocins that are produced directly as ribosomally

synthesized polypeptides (Jack et al., 1995).

Studies on the observed antagonism between X campestris pv. vesicatoria T3 and

other strains of the same bacterium indicated the characteristics of inhibition most closely

resembled those of bacteriocins (Tudor, 1995). Several "bacteriocin-like" compounds

have been identified and so named because of the lack of a precise definition of a

bacteriocin. The first bacteriocin to be identified was produced by Escherichia coli and

was given the name colicin (Reeves, 1965). It was distinguished from previously

identified antibiotics by the narrow range of organisms it inhibited, relative to broad

spectrum antibiotic compounds and by its chemical composition. Jacob, in recognizing

that colicin-type compounds were produced by non-coliform bacteria, coined the more

general term "bacteriocin" (Pattus et al., 1990). Thus, colicins became the prototype









bacteriocins, and their properties were used to define the class of compounds now known

as bacteriocins. The characteristics of bacteriocins were as follows: narrow inhibitory

spectrum; presence of a biologically active protein moiety; bactericidal mode of action;

attachment to specific cell surface receptors; plasmid borne determinants; and

inducibility by DNA damaging agents such as ultra-violet radiation and mitomycin C

(Reeves, 1972). Although these provided the initial definition, several compounds with

comparable but not identical characteristics have been classified as bacteriocins. For

example, agrocin, a bacteriocin produced by Agrobacterium radiobacter strain K84 is a

disubstituted, fraudulent adenine nucleoside analog with no protein component (Tate et

al., 1979), while other bacteriocins do not share the adsorption specificity of colicins

(Bhunia et al., 1991). Despite these inconsistencies, bacteriocins with varied

characteristics have been reported for all major groups of Eubacteria and Archaebacteria

(Riley, 1998). The heterogeneity between this group of compounds can be somewhat

refined into two major groups: bacteriocins produced by Gram positive organisms versus

those of Gram negative bacteria. Properties of bacteriocins of the Archaebacteria align

most closely with those of the first group (Riley, 1998). The major differences between

these two groups will be highlighted within the discussion to follow.

The restricted specificity of bacteriocins points to the supposition that they are

primarily involved in mediating population dynamics at the intraspecific level. This type

of inhibition is highly advantageous, as organisms within the same species are likely to

be competing for the same niche resources. In order to examine the role that bacteriocins

play in shaping population structures at this level, a number of in vitro studies

investigating competition between bacteriocin-producer and -sensitive strains have been









conducted. Most of these have centered on colicin producing strains (Riley and Gordon,

1999). Observations of the flux between these two phenotypes under different conditions

have led to the development of models to define the co-existence of these two phenotypes

in a given habitat. The models are based on the assumption that bacteriocin production

has an associated fitness cost that is reflected in their lower growth rate (Chao and Levin,

1981). Colicin production is lethal to the cells that produce it. Not all of the cells within

a given population are actively producing colicin, but those that do are lysed upon

colicin release (Spangler et al., 1985). This, associated with the additional energetic

cost of maintaining plasmid-borne determinants for both production of, and immunity to

the bacteriocin, are thought to account for the lowered growth rate of colicin producing

strains. Problems arise however, when using this assumption to describe non-colicin

systems. For most bacteriocins produced by Gram positive bacteria, synthesis is not

lethal to the cells that produce it (Jack et al, 1995), and, in some cases the genes

involved in bacteriocin activities are chromosomal (Gaggero et al., 1993). Another way

of off-setting the fitness cost of bacteriocin production is based on the finding that some

bacteriocins have multiple functions. For example, bacteriocin small, produced by

Rhizobium leguminosarum, is also an autoinducer that is involved in the symbiotic

interaction between the bacteria and plants (Schripsema et al., 1996). The possibility

still exists that many of the bacteriocins that have been characterized to-date may have

additional, undiscovered functions.

Based on the first assumption, models predict that bacteriocin-susceptible strains are

generally favored in poor habitats where the rate of resource competition is high, and

that bacteriocin producers would be favored in nutrient-rich environments. Experimental









results show this to be an accurate assessment (Frank, 1994). In addition, bacteriocin

production has been shown to be induced in nutrient-poor environments, perhaps a

compensatory adjustment for their poor competitiveness in these types of environments

(Riley and Gordon, 1999). The high frequencies of bacteriocin-producer strains found in

natural populations, roughly 30% (Riley and Gordon, 1992), suggests a selective

advantage over the cost of their production.

Surprisingly, invasion of a bacteriocin producing strain into a sensitive population

does not necessarily dictate its immediate take-over. In fact, studies have shown that the

strain with the highest frequency in the population is rarely displaced in this situation,

unless the frequency of the invader is greater. Exceptions to this do however occur in

structured habitats, where there is a partitioning of resources and rare phenotypes have

more of a selective advantage (Chao and Levin, 1981).

There has been a resurgence on the study of bacteriocins in recognition of their

potential practical applications. Most of this work has focused on characterization of the

genetic elements used to effect bacteriocin production. Colicin biosynthesis genes, were

the first to be isolated from either small, non-self-transmissible plasmids (Riley, 1998)

or large low copy plasmids which are transferred by conjugation (Harkness and

Oelschlaeger, 1991). Bacteriocinogenic determinants of most Gram-negative and

archaebacteria have subsequently been found on plasmids (Cheung et al., 1997;

Barefoot et al., 1992). While this is also true for most Gram-positive bacteriocins (Jack

et al., 1995), a number of these have also been found on chromosomes. Helveticin J, a

bacteriocin produced by Lactobacillus helveticus 481 was localized to the chromosome









when plasmid-cured strains retained bacteriocin activity (Joerger and Klaenhammer,

1986).

Several genes are involved in bacteriocin expression. Typically a group of genes,

each with a specific role in bacteriocin production are clustered. Colicin gene clusters are

usually transcribed as an operon that consists of a structural or activity gene, a gene

involved in specific immunity to the bacteriocin and a third lysis gene that is involved in

bacteriocin release from the cell (Riley and Gordon, 1999). These genes are usually

transcribed in the same direction, although in some cases, as is the case with colicin B

(Braun et al., 1994) the immunity gene is transcribed in the opposite orientation. The

presence of all these genes is not ubiquitous to colicinogenic strains, colicins B and M

both lack a lysis gene (Braun et al., 1994). The gene clusters involved in bacteriocin

synthesis by Gram-positive bacteria contain many more genes and have a more complex

transcriptional organization. A polycistronic operon containing eight genes involved in

nisin biosynthesis have been isolated from Lactococcus lactis (Steen et al., 1991). The

structural gene, nisA is the first in the operon and is directly followed by three genes,

nisB, nisT, and nisC thought to be involved in bacteriocin export. There are regions of

overlap between nisT, and nisC. Four additional genes lie directly downsteam of these,

nisI, nisP, nisR, and nisK. Immunity to nasin is thought to be mediated by the

lipoprotein encoded by nisI, while, nisR, and nisK are involved in the regulation of

nisin biosynthesis genes (Kuipers, et al., 1993). The structure and content of this operon

is similar to many of the bacteriocin operons in Gram-positive bacteria (Jack et al.,

1995).









Most of the bacteriocins isolated from E. coli and other Gram-nagetive bacteria are

high molecular weight compounds, ranging in size from 29,000 to 81,000 (Braun et al.,

1995). Pyocins S1 and S2, isolated from Pseudomonas areuginosa, have molecular

weights of 65,600 and 74,000, respectively (Sano et al., 1993b). Proteolysis, deletion,

and chimera analysis have been used to define three distinct functional domains within

these molecules (Sano et al., 1993a; Baty et al., 1988). The N-terminal domain is

involved in translocation of the bacteriocin across the membrane of the sensitive strain,

while the central domain is involved in its specific binding to receptors expressed on the

surface of the sensitive cell. The C-terminal domain is the activity center of the molecule

as it contains the killing function. The order of these domains is not always the same. In

pyocins S1 and S2, the receptor-binding domain and translocation domains are swapped

in position (Sano et al., 1993a). Sequence analysis and subsequent alignment of colicin

and colicin-like bacteriocins reveal that certain pairs resemble each other only in one or

two domains. For example, Colicins E2 and E3 share almost identical N-terminal and

central domains; however, they differ significantly at the C-terminus of the molecule

(Konisky, 1982). In contrast, Pyocins S1 and S1 share an almost identical C-terminal

domain with colicin E2, but differ substantially in the other portion of the molecule

(Sano et al., 1993a).

Evidence that the C-terminal domain was the activity center of the molecule was

obtained when deletion constructs lacking the first two domains were able to kill when

translocated into the cell by osmotic shock treatment, which rendered the cell

temporarily permeable to proteins (Braun et al., 1994). The most common type of

activity associated with the C-terminal domain is channel-formation in the membranes of









sensitive cells. X-ray crystallographic structural analysis shows that it is composed of

eight hydrophilic and two hydrophobic helices. Insertion of the hydrophobic helices into

the membrane results in depletion of phosphate required for ATP synthesis and inhibition

of cellular respiration (Cramer et al., 1995). This membrane depolarization leads to

immediate cell lysis. The C-terminal domains of other bacteriocins have endonuclease

activity. Colicins E2, E7, and pyocins S and S2 have endodeoxyribonuclease activity

and exert their effect by cellular degradation of DNA (Pattus et al., 1990; Sano et al.,

1998b). Colicins E3, E4 and cloacin DF13 are endoribonucleases which cleave 16S

rRNA from the 30S subunit of ribosomes, thereby inhibiting protein biosynthesis (Pattus

et al., 1990). The C-terminal domain of colicin M blocks the biosynthesis of both

peptidoglycan and the O-antigen leading to cell death (Harkness and Oelschlaeger,

1991).

Bacteriocins of Gram-positive bacteria are usually much smaller peptides, typically

containing less than 60 amino acid residues (Jack et al., 1995). No functional domains

have been identified in these molecules and they show great variation with respect to

distribution of polar and non-polar amino acids, and acid and basic residues (Bruno and

Montville, 1993). A class of these bacteriocins, called lantibiotics, are known for their

content of the novel, posttranslationally modified amino acids, lanthionine and

methyllanthionine (Diep et al., 1994). Voltage-dependent or independent

permeabilization of cytoplasmic membranes is their primary mode of action (Bruno and

Montville, 1993).

Bacteriocin genes are generally regulated. Up-regulation of several of these genes

has been induced, in vitro, by DNA-damaging agents such as ultraviolet radiation and









mitomycin-C (Spangler et al., 1985). They are thought to induce the SOS response

regulation system (Rakin et al., 1996). This response leads to the activation of the RecA

protease which specifically degrades the LexA protein which is a repressor of bacteriocin

synthesis (Spangler et al., 1985). The genes involved in bacteriocin immunity are

constitutively expressed at low levels (Lakey et al., 1994). While continual production

of active bacteriocin may not be needed for cellular survival, expression of the immunity

protein in an environment of bacteriocin producing cells is a necessary trait.

A two-component signal-transducing system has recently been found to be involved

in expression of many of the bacteriocins secreted by Gram-positive bacteria (Jack et al.,

1995). The gene cluster of sakacin A, a bacteriocin secreted by Lactobacillus sake,

contains two genes, sapK and sapR which encode putative histidine kinase and response

regulator proteins, respectively. Both gene products have significant similarity with

corresponding proteins involved in the accessory gene regulatory (Agr) two-component

signal transduction system in Staphylococcus aureus, where it controls expression of

exoproteins during post-exponential growth (Axelsson and Holk, 1995). Frameshift

mutation and deletion analysis showed that they are both required for bacteriocin

expression and immunity. The signals detected in the response are unknown at this point.

However, in another type of regulation, a secreted peptide which was initially identified

as the bacteriocin itself, is shown to induce synthesis of the bacteriocin. The mechanism

of induction in this system is not known (Diep et al., 1995).

In order for bacteriocins to express activity they must be exported from the cell. For

Gram-negative bacteriocins, none of the typical N-terminal cleavable signal sequences

which direct exoproteins to the cytoplasmic membrane were identified (Lakey et al.,









1994). This was the first indication that they were not secreted via the general secretary

(sec) pathway commonly used for export of exoproteins in Gram-negative bacteria.

Instead, bacteriocin release is usually mediated by the activity of the lysis gene product

which is associated with the bacteriocin gene cluster. Evidence for this was obtained

when removal of the lysis gene (cal) from the colicin A plasmid lead to an accumulation

of colicin A in the cell wall (Lakey et al., 1994). The colicin A lysis protein contains an

N-terminal signal sequence, cleavage of which is necessary for its activity. Acylation of

the lysis protein is also necessary for its activity. Mature lysis proteins were localized to

both the cytoplasmic membrane and the outer membrane, so they may permeabilize both

membranes. Co-incident with the activity if the lysis protein is the dimerization and

activation of phospholipase A in the outer membrane which generates membrane-

damaging phospholipids and leads to the simultaneous release of colicin along with many

other molecules from the cell (Dekker et al., 1999). The result is death for the producing

cell. The process that enables the release of colicins lacking associated lysis genes is

unknown at this time.

Cell death is not associated with release of bacteriocins from Gram-positive bacteria;

instead a dedicated system of export, similar to the ABC transport system found in many

bacteria, is thought to facilitate their release (Fath and Kolter, 1993). Comparisons of

actual N-terminal amino acid sequences of purified bacteriocin preparations with

predicted products led to the identification of N-terminal leader sequences of

approximately 18-30 residues which are involved in targeting of bacteriocins to the

cytoplasmic membrane (Jack et al., 1995). ABC export systems consist of a core set of

required and sometimes accessory, membrane associated proteins. These proteins are









able to recognize the substrate, and in an ATP-dependent manner, translocate it through

the membrane. Several bacteriocin operons of Gram-positive bacteria are known to

encode proteins with significant homologies to proteins with known involvement in ABC

export systems. For example, nisT, a gene in the nisin operon, encodes a 600 amino

acid protein which is highly homologous to the hemolysin B (hlyB) gene (Engelke et al.,

1992). The HlyB protein is an inner membrane protein that is involved in the export of

alpha-hemolysin from E. coli (Fath and Kolter, 1993). Although unusual, sec-

dependent export has been reported for at least one bacteriocin. The gene encoding

enteriocin P, a bacteriocin produced by Enterococcusfaecum, is proceeded by a typical,

sec-dependent leader sequence. Results show that Enterocin P is produced and secreted

in a sec-dependent manner (Cintas et al., 1997).

The first step involved in cell lysis by bacteriocins is their interaction with the cell

membranes of sensitive cells. In all of the Colicins examined thusfar, direct binding to

membrane receptors has been reported. Colicins are known to parasitize outer membrane

receptors which are normally used for other physiological purposes. For example,

Colicin El binds to the BtuB receptor, which is normally involved in vitamin B12 uptake

(Braun et al., 1994). For colicins, uptake through the outer and inner membranes is via

two distinct routes, the Ton system or the Tol system. Both systems consist of a series of

polypeptides that interact to span both membranes and facilitate the uptake of molecules

into cells. The Ton system is energy dependent and is utilized in the uptake of iron and

vitamin B12 into cells. The Tol system is not energy dependent.

Binding of Gram-positive bacteriocins to sensitive cells is mainly thought to be a

non-specific process, which is affected by the pH of the cell wall (Bhunia et al., 1991).









The non-specific binding of these compounds may account for their generally broader

spectrum of activity compared to their Gram-negative counterparts. The single-cell wall

of these bacteria allows free passage of relatively large molecules, perhaps reducing the

need for specific binding and import of these compounds. Specific binding has however

been reported for lyostaphin, a bacteriocin produced by Staphylococcus aureus. The 92

C-terminal residues of this molecule mediate its specific binding to S. aureus strains

(Baba and Schneewind, 1996)

Even more striking than the specificity and lethality of bacteriocins is the efficiency

at which their activity is neutralized by the products of immunity genes. This is crucial

as a mechanism of self-protection for bacteriocin producing strains. Although different

mechanisms of immunity have been reported, the one characteristic they share is their

specificity to the bacteriocins with which they are associated. In some cases the

mechanism of immunity is by direct binding of the immunity protein to the bacteriocin.

In the case of colicins with nuclease activity, studies using hybrid and truncated colicin-

constructs showed that the immunity protein bound directly to the C-terminal, nuclease

domain of the molecule and prevented its lethal activity. The mechanism whereby the

immunity protein is removed from the complex to regenerate active bacteriocin is

unknown (Konisky, 1982). In other cases there is no direct evidence that the

bacteriocins are released from cells in association with the immunity protein. In these

cases, immunity has been shown to operate at the cytoplasmic membrane (Song et al.,

1991). It is proposed that there is an interaction between the transmembrane helicies of

the bacteriocin and those of the immunity protein (Zhang and Cramer, 1993).









Bacteriocin-resistant phenotypes, without specific immunity genes, can also be

found in natural populations. Bacteriocins generally require specific receptors and

translocation machinery for binding to, and entry into sensitive cells. Resistant

phenotypes show mutations in one or both of these components (Riley and Gordon,

1999). An additional mechanism of resistance was suggested to involve synthesis of long

O-antigen chains of LPS, the first barrier for most microbial toxins (Lakey et al., 1994).

Immunity genes have also been found in operons encoding bacteriocins of many

Gram positive bacteria; however, the mechanisms of immunity is poorly understood. In

some cases the presence of the immunity gene is insufficient to confer complete

protection as is the case with enterocin A, a bacteriocin produced by Enterococcus

faecium (O'Keeffe et al., 1999). Several mechanisms have been proposed for immunity

to these types of bacteriocins. Direct competition for binding sites on the surfaces of

sensitive cells, between the bacteriocin and immunity proteins has been suggested as a

mechanism of immunity (Quadri et al., 1995), while evidence for active proteolytic

degradation or extrusion from the cell has been obtained in other cases (O'Keeffe et al.,

1999).

The in-depth study of bacteriocins provides valuable insights towards understanding

the forces that shape natural bacterial populations, protein-protein interactions, transport

across bacterial membranes and protein conformational changes. Besides these elements

of basic research, can we use our understanding to manipulate this natural force used in

competition among bacteria in an attempt to facilitate more effective plant disease control

strategies? Limited attempts at this have been made over the years with varying degrees

of success (Vidaver, 1983).









Although early attempts at biocontrol of plant diseases using antagonistic strains were

made, there has been a recent resurgence in attempts to develop biocontrol agents which

are perceived as being more environmentally safe than conventional chemical methods.

Research in this direction has led to the identification of a critical characteristic essential

for successful use of organisms in this manner and that is, the organisms ability to

efficiently colonize and compete on the host plant. The importance of this could be seen

in studies on the suppression of Fusarium wilt with Pseudomonasputida (Tari and

Anderson, 1988). The agglutination (agg ) phenotype of this bacterium is associated

with its ability to colonize roots. Tn5 generated agg- mutants not only had reduced ability

to colonize roots, but also provided less protection against the pathogen, indicating that

colonization is necessary for control to occur. Further evidence for this was obtained

from the analysis ofAgrobacterium radiobacter strains K84 and J73 (Thompson, 1993).

Both strains produce agrocins active against A. tumefaciens, but only strain K84

protected tomato plants against crown gall. Strain J73 is a very poor colonizer of tomato

compared to strain 84, a fact believed to be responsible for its inability to control the

disease. Direct transfer of the agrocin 84 plasmid to strain C58 a poor colonizer,

yielded transconjugants that continued to be less effective at reducing tumor formation

than strain K84 (Thompson, 1993). Once initial colonization occurs, organisms must be

able to continually compete for niche resources. Theoretically, biocontrol strains should

have very similar patterns of resource utilization to be effective competitors. Biological

control of ice-nucleating (Ice ) Pseudomonas syringae strains by (Ice-) strains was

investigated using a niche-overlapping index. This index is reflective of the level of

shared carbon resources between two organisms (Lindow, 1987). Control of Ice









bacteria is thought to be mediated by their inability to compete for the identical C

resources as their Ice- counterpart. Because of the importance of colonization, some

studies have specifically addressed the colonization potential of a bacterial strain prior to

testing its biocontrol potential. Non-pathogenic mutants of Pseudomonas solanacearum

were tested for their ability to control bacterial wilt caused by the same pathogen.

However, before this was done, the non-pathogenic Hrp- mutants, generated by Tn3-

mutagenesis were first screened and selected for their ability to survive on tomato plants.

Selected mutants were able to survive endophytically for up to three months and confer a

protective effect to the pathogen (Frey et al., 1994).

Despite the identification of strains that are efficient host colonizers and produce

antagonistic compounds, biological control using these strains is still not completely

effective. Perhaps the biggest hindrance to the implementation of these strategies is the

lack of predictability of the control it renders. Sakthivel and Mew (1991) used N-methyl-

N-nitro-N-nitrosoguanidine (NTG)-generated, non-pathogenic, bacteriocin-producing

mutants ofXanthomonas oryzae pv. oryzae to control bacterial blight of rice. Although

the strains used were consistently able to survive as epiphytes up to four-weeks post

inoculation, the reduction in disease symptoms was highly variable, ranging from 31-

99% in greenhouse tests. The variability was reflective of differences between the strains

used. Similar inconsistencies were observed when using avirulent Erwinia amylovora,

E. herbicola or Pseudomonas tabaci to control fire blight (McIntyre et al., 1973).

Control was variable with respect to timing of treatment application relative to

inoculation with the pathogen. On Bartlett pear, delay of symptom expression only

occurred when treatments were applied 24 hours but not 0.5 hours prior to pathogen









inoculation, while on Jonathan apple, similar results were achieved for both inoculation

times. Protection of tobacco plants from bacterial wilt by avirulent bacteriocin-producing

strains of the pathogen (Pseudomonas solanacearum) showed yet another type of

variability (Chen et al., 1981). Reductions in disease severity by these strains were

variable according to the population level of the virulent strain in the soil. Control was

negligible when concentrations were above 106 CFU/g dry soil.

Variability in these and other instances is indicative of the poor understanding of

underlying processes involved in biocontrol by these strains. In most cases, there is little

direct evidence for the role of antibiosis in the process. In the first example, there is no

experimental evidence for the production of bacteriocins on rice plants. Indirect evidence

for this was obtained for biocontrol using avirulent bacteriocin-producing strains of P.

solanacearum, when avirulent, non- bacteriocin-producing strains were found to be less

effective at reducing wilt symptoms (Chen and Echandi, 1984). More direct evidence

for this does occur for the use of syringacin 4-A, a bacteriocin produced by

Pseudomonas syringae. When purified preparations of the bacteriocin were applied to

bean leaves prior to inoculation with P. phaseolicola, there was a significant reduction in

lesion counts relative to the control (Vidaver, 1976).

A clear understanding of the factors involved in biocontrol will be required before

successful implementation of this strategy is to occur. Currently, the only successful

commercial use of a bacteriocin-producing strain to control a plant disease is the

biological control of crown gall disease. In 1974, Kerr and Htay showed rhizosphere

control of Agrobacterium tumefaciens, by the closely related non-pathogenic soil

organism A. radiobacter strain K84 (Kerr and Htay, 1974). Control is mediated by the









production of a bacteriocin, agrocin 84. Sensitivity to the bacteriocin is linked to

virulence, with both traits being encoded on the tumor inducing (Ti) plasmid. The

reason for the linkage is based on the fact that agrocin 84 uptake is facilitated by binding

to receptors which are normally involved in uptake of agropines, which are sugars

metabolized solely by enzymes localized on the Ti plasmid (Chet et al., 1993). The

close correlation between these two traits is probably the major reason for the success of

this control system. Several lines of evidence have proven a direct role for bacteriocin

production in the disease control process. Transfer of the plasmid encoding agrocin 84 to

another Agrobacterium non-producer strain enables the transformed strain to produce the

bacteriocin and control crown gall simultaneously, while curing of the plasmid led to the

subsequent loss of both traits. Also, only strains sensitive to the bacteriocin are

controlled by strain K84 (Lam and Gaffney, 1993).

Despite its commercial success, the use of agrocin 84 has also encountered some

problems. The first of which is its restricted specificity, which is limited to A.

tumefaciens strains harboring agropine-specific enzymes. Under natural conditions,

transfer of the agrocin plasmid into the pathogen presents a potential problem for

biocontrol. Studies on the use of strain K84 (Thompson, 1993) during field experiments

in Greece, showed that some galls were formed by agrocin-resistant A. tumefaciens

strains. This problem was solved by a deletion of a 5.9 kb region of the agrocin plasmid

which contained the genes necessary for plasmid mobilization. The modified strain has

subsequently undergone substantial field testing and is marketed as the first genetically

engineered organism to be used as a pesticide (Thompson, 1993).









The use of molecular biology techniques offers the potential for improving biocontrol

strains to enhance production of antimicrobial compounds and improve host colonization

potential, to introduce novel traits into organisms and to develop procedures to detect

and minimize the risk of gene transfer. Constitutive expression of antimicrobial genes

and the use of tissue specific promoters to facilitate the expression is likely to improve

their efficacy as was shown in the case of cecropin. Introduction of genes conferring the

ability of a strain to metabolize new substrates could potentially improve its ability to

colonize host plants. Genes for sucrose metabolism were introduced into Pseudomonas

fluorescens, that improved the organisms ability to colonize sugarcane (Thompson,

1993). Strains of Pseudomonasputida, an excellent root colonizer, were transformed

with genes required for chitinase biosynthesis. Constitutive expression of chitinase in

these strains resulted in increased biocontrol activity compared to controls (Chet et al.,

1993). The risk of gene transfer between strains is one of the biggest concerns relating to

public perception and the use of genetically modified organisms. The most common

mechanisms by which this occurs have been identified, and steps to reduce this risk such

as deletion of genes required for conjugal transfer have been applied. In addition, the

use of selectable markers now makes it easier to track released organisms in natural

habitats.

A recent innovation for improving biocontrol technology is the direct incorporation of

genes capable of conferring disease resistance into plants. This has been fueled by the

rapid development of improved methods for plant transformation and increased

identification and molecular characterization of potential target genes for transformation

(During, 1996). One of the strategies used is the use of antibacterial genes to engineer









plants for disease resistance. Among the first examples of this was the transformation of

tobacco plants with cecropin B (Nordeen et al., 1992). Cecropins are lytic peptides that

were isolated from the giant silk moth Hyalophora cecropia, which insert into the outer

membranes of bacterial and eukaryotic cells and disrupt membrane structure. They are

small peptides containing between 35-37 amino acids which exhibit a broad spectrum of

activity against both Gram-negative and Gram-positive bacteria. Studies on the toxicity

of cecropins have provided data on the exact quantities of the toxin which are required to

kill several types of plant pathogenic bacteria. Toxicity of cecropin to several plant

protoplasts has also been studied (Nordeen et al., 1992). In most cases the levels of

lethal concentrations for plant protoplasts are significantly higher than for selected

corresponding bacterial pathogens. One notable exception is tomato for which the lethal

concentration is approximately 4.5 [IM, compared to 3.4 jaM for Clavibacter

michiganense subsp. michiganense, an important pathogen of tomato. This highlights

the first key step in plant transformation with foreign genes and that is the necessity for

tight control of the expression of incorporated genes.

Several differences exist between expression of genes in plants and bacteria,

therefore before plant transformations with cecropin were made, modifications to enable

correct inplanta expression of the genes were made. Initial constructs carried the

cauliflower mosaic virus (CaMV) 35S promoter, a modified cecropin B gene and a

nopaline synthase (NOS) 3' polyadenylation signal. Transformation of tobacco with

these constructs did not result in resistance to P. solanacearum despite endogenous

detection of the cecropin peptide (Jaynes et al., 1993). Citing probable problems with

the levels of expression, a new construct, shiva-1, was designed in which the cecropin









structural gene was even further modified and now driven by the wound-inducible

proteinase inhibitor II promoter. Higher levels of the cecropin peptide were detected in

transgenic plants expressing this construct, and these plants also exhibited enhanced

resistance to P. solanacearum (Jaynes et al., 1993). Similar results were obtained when

using another modified cecropin construct, MB39, to engineer tobacco plants for

resistance to Pseudomonas syringae pv. tabaci (Huang et al., 1997). The biocontrol

"plague" of inconsistent results also occurs with this type of strategy. A number of

researchers using similar constructs have found very low or indetectable levels of

endogenous cecropin and no significant disease protective effects (During, 1996). One

possible explanation for this variability was forwarded by Mills et al. (1994) who

showed that Cecropin B activity is reduced by the activity of endogenous plant

endopeptidases.

Transformation of plants with the bacteriolytic enzyme, lysozyme, has also

conferred some disease protective effects. In one study, the bacteriophage T4 lysozyme

gene was fused to a plant signal peptide and incorporated into potato plants. The

molecular weight of this protein is approximately 18,700, and therefore represents

transformation with a larger protein compared to cecropin (During et al., 1993). Despite

its larger size and bacterial derivation, lysozyme T4 was actively produced and correctly

processed in transgenic plants. Proof of its extracellular release was obtained when the

protein was localized to the intercellular spaces of transgenic plants. Potato plants

expressing this construct demonstrated a significant reduction in susceptibility to Erwinia

carotovora subsp. atroseptica, having some 65% less maceration of tuber slices relative

to non-transformed controls (During et al., 1993).






26


Development of transgenic plants carrying antimicrobial genes is still in the initial

stages. Progress towards better understanding of how these genes are correctly expressed

in plants still needs to be made. Many commercially useful crops are non-transformable;

efforts also need to be focused in this direction. One of the biggest challenges facing the

commercial release of transgenics in the future is the public's skepticism towards their

use.















CHAPTER 3
PRODUCTION OF MULTIPLE
BACTERIOCIN-LIKE COMPOUNDS

The characteristics of inhibition of tomato race-one (TI) strains of Xanthomonas

campestris pv. vesicatoria by tomato race-three (T3) strains of X. campestris pv.

vesicatoria most closely resemble those of bacteriocins (Tudor, 1995). They were

characterized by a narrow inhibitory spectrum, restricted to X campestris strains, by the

presence of a biologically active protein moiety, inducibility by mitomycin C and non-

self inhibition.

Reports on the production of bacteriocin-like compounds by phytopathogenic bacteria

are scattered. The first report was by Okabe in 1954 who reported that strains of

Pseudomonas solanacearum were inhibitory only to other Pseudomonas strains (Cuppels

et al., 1978). Since then the production of these types of compounds has been reported

for the following plant pathogenic genera: Agrobacterium (Kerr and Htay, 1974),

Clavibacter (Echandi, 1976), Erwinia (Echandi and Moyer, 1979), Pseudomonas (Frey

et al., 1996) and Xanthomonas (Xu and Gonzalez, 1991; Fett et al., 1987). Besides

agrocin production by A. radiobacter, the most characterized of these are from

pseudomonad species. Strains of P. syringae pv. syringae produce at least two high

molecular weight particulate bacteriocins which resemble bacteriophage tails, while P.

glycinea and P. phaseolicola produce low molecular weight non-particulate bacteriocins

(Vidaver, 1972). More recently a comprehensive study on bacteriocin production by P.

solanacearum revealed that 121 of the 149 strains tested produced bacteriocin-like









compounds (Frey et al., 1996). These were placed into nine groups based on differential

inhibition of 22 sensitive indicator strains. In this study the presence of bacteriocin

groups was correlated with genomic variability as assessed by two different genomic

fingerprinting methods. The use ofbacteriocins to assess genome variability, known as

bacteriocin-typing, has been one of the most prevalent uses of bacteriocins by

phytobacteriologists.

Few xanthomonads have been reported to be bacteriocin producers. Strains of X.

oryzae pv. oryzae produce at least four bacteriocin-like compounds (Xu and Gonzalez,

1991). One report indicates that at least two of them are low molecular weight

compounds with variable specificities (Huang et al., 1996). Two independent reports

have suggested the production of multiple bacteriocin-like compounds by single strains

of X. campestris pv. glycines (Fett et al., 1987; Ahn and Cho, 1996).

Due to the limited research which has been conducted on the production of

bacteriocin-like compounds by Xanthomonas spp., very little is known of their potential

use in biocontrol strategies. This study seeks to further characterize the production of

these types of compounds by tomato race-three (T3) strains ofX. campestris pv.

vesicatoria. Previous studies have suggested the presence of multiple antagonistic

compounds by these strains (Tudor, 1995). Further proof of this will involve a screen for

additional compounds, identification of differences between these compounds and

genetic analysis. Up to this point, only in vitro assays have been conducted to observe

antagonism between strains (Tudor, 1995); in this study, experiments involving in

plant antagonism were performed.









Materials and Methods

Bacterial Strains, Plasmids, and Culture Conditions

The bacterial strains and plasmids used in this study are listed in Appendix A. All

strains of X. campestris pv. vesicatoria, were grown on nutrient agar (BBL) at 28 C.

Strains of Escherichia. coli were grown on Lauria-Bertani (LB) medium (Miller, 1972)

at 37 C. All strains were stored in capped tubes containing 2.0 ml sterile tap water at

room temperature or in nutrient broth containing 30% glycerol at -70 C. Antibiotics

were used to maintain selection for resistance markers at the following concentrations:

tetracycline (Tet) 10 [g/ml; rifamycin (Rif) 100 [g/ml; spectinomycin (Spec),

kanamycin (Kan) 50 pjg/ml and nalidixic acid (Nal) 20 pjg/ml

In Vitro Antagonism Assays

In vitro inhibitory activity was detected using two different techniques. Strains to be

tested for antagonism were grown overnight at 28 C in 4 ml nutrient broth tubes. The

cells were pelleted and resuspended into sterile tap water. Cell suspensions were then

standardized to an OD of 0.3 at 600 nm (concentration equal to 5 X 108 CFU/ml).

Twenty-five microliters were then spotted onto nutrient agar plates, five strains per plate,

and allowed to grow for 18 hr at 28 C. Both a positive and a negative control were

plated onto each test plate. In the first technique a 5 X 107 CFU/ml cell suspension of the

indicator was sprayed over the surface of the plate using a Sigma aerosol spray unit

(Sigma Chemical, St. Louis, MO). The alternative technique involved incubation of the

plates for 24 hr, killing the test strains by inverting glass plates over 2-3 ml of

chloroform until all of the chloroform was evaporated, aerating the plates for 1 hr, and

overlaying the agar surface with 3.5 ml of 0.7% water agar (50 C) which contained 200









pl of a 5 X 107 CFU/ml cell suspension of the indicator strain. Tomato race 1 X

campestris pv. vesicatoria strain 91-106, was used as the indicator strain in these studies.

A clear zone of inhibition around test colonies after 24 48 hr was considered indicative

of antagonism and scored as bacteriocin-like (BCN +) activity. The diameter of inhibition

zones was measured as the distance from the edge of the test culture to the end of the

zone. Three distinct phenotypes were scored as follows: non- inhibition, weak

inhibition (zone diameters less than or equal to 4mm) and strong inhibition (zone

diameters greater than 4mm). Each test was replicated four times.

Cell-free extracts were screened for BCN + activity by growing test cultures for 18 hr

in nutrient broth followed by centrifugation to pellet cells. The supernatant was then

sterilized using a low protein binding microcon filters (Amicon, Beverly, MA) with a

pore size of 0.2 pm and analyzed for antagonistic activity by the well diffusion method

(Tagg and McGiven, 1971). Ten mm diameter wells were cut into nutrient agar plates

containing 20 ml of the test medium. Wells were filled with 200 Pl of test filtrates and

left for 12 hr to allow diffusion of the liquid into the medium. Plates were then overlaid

with 3.5 ml 0.7% molten agar containing 200 pl of a 5 x 10s CFU/ml cell suspension of

the indicator strain. Strain 91-106 was used as the negative control for the test strain and

also as the indicator strain for these studies. Plates were examined after 24 hr at 28 C

for growth around wells. Each test was replicated four times.

In Planta Antagonism Assays

Six-week-old seedlings of the tomato cultigen Florida 7060 were infiltrated with

bacterial suspensions of both the producer and the indicator strains using a hypodermic

syringe as described previously (Jones et al., 1998a). Prior to inoculation strains were









grown in nutrient broth for 18 hr, harvested by centrifugation, and resuspended in sterile

tap-water. Suspensions of producer and non-producer strains were infiltrated into fully

expanded leaflets at the same time or 12 hr prior to infiltration of the same area with a

sensitive indicator strain. Concentration of the test strains used were 5xl07 CFU/ml and

indicator strains were inoculated at a concentration of 5x106 CFU/ml. Each treatment

consisted of three replications. Plants were incubated between 240 C and 280 C. To

examine the levels of indicator strain populations in leaflets of each treatment, 1-cm2 leaf

disks were removed from inoculated areas, macerated in lml sterile tap-water, and

dilution-plated onto nutrient agar amended with the appropriate antibiotic. Estimates

were made at 12-hour intervals for 48 hr. Each experiment was conducted twice.

DNA Manipulations

Plasmids from E. coli strains were extracted using the alkaline lysis method described

by Sambrook et al. (Sambrook et al., 1989). Restriction endonuclease digestions were

performed using the manufacturers specifications for each enzyme. All enzymes were

obtained from Promega (Madison, WI). Restricted fragments were separated by

electrophoresis in agarose gels (Seakem GTG [FMC Bio Products, Rockland, ME]) in

TAE buffer at 8V/cm. Gels were stained for 20 min. with ethidium bromide (0.5|pg/ml)

and photographed over a UV transilluminator (Fotodyne INC., New Berlin, WI).

Screening of Cosmid Library

A cosmid library of purified DNA from X campestris pv. vesicatoria strain 91-118

(G. V. Minsavage, University of Florida, Gainesville, Florida) was used for screenings.

The library was constructed in the cosmid vector pLAFR3 (Tet r) and maintained in E.

coli strain DH5ci. Seven hundred and fifty clones each containing 15-30-kb of insert









DNA were screened for antagonistic activity after cosmid clones were conjugated into X

campestris pv. vesicatoria strain ME90 (Rif r, Kan r). Conjugations were performed

using a triparental mating scheme in which cosmid clones (Tet r) were used as donor

strains, strain ME90 (Rifr, Kan r) as the recipient with pRK2073 (Spec r) supplied as the

helper plasmid. All 3 strains were mixed on nutrient yeast-glucose agar (Daniels et al.,

1984) and incubated at 28 C for 24 hr. Matings were then suspended in 2 ml sterile tap

water and plated on nutrient agar amended with rifamycin and tetracycline. Purified

transconjugants were then transferred onto nutrient agar plates, 5 per plate, and

screened for antagonism against strain 91-106. Transconjugants around which clear

zones of inhibition were visible were selected for further study. Expression of BCN+

activity was also tested in four other recipients and against several other indicator strains.

Southern Hybridization Analysis

DNA to be analyzed by Southern analysis was electrophoresed on 0.75% agarose

according to standard protocols (Sambrook et al., 1989). The gel was denatured in 0.4 N

NaOH and 0.6 M NaCI for 30 min, and neutralized for 30 min in 0.5 M Tris, 1.5 M

NaCl. The denatured DNA was then transferred by the procedure of Southern (1975) to a

nylon membrane (Schleicher & Schuell, Keen, MA). Hybridized DNA was detected

with the Genius nonradioactive DNA labeling and detection kit (Boehringer Mannheim,

Indianapolis, IN) as described by the manufacturer. Probes were labeled by the random

primer method (Feinberg and Vogelstein, 1983) for incorporation of digoxigenin-11-

deoxyuridine-triphosphate (DIG-dUTP). Probes were denatured by boiling at 100 C

before use. Hybridization was performed at 68 C for 24 hr and the membrane washed









with 0. X SSC, 0.1% W/V SDS. BCN clones were used in cross-hybridization

experiments with each other.

Transposon Mutagenesis

Insertion mutagenesis of BCN + clones was performed using the Tn3HoHo derivative,

pHoKmGus (D. Dahlbeck and B. Staskawicz, University of California, Berkley, CA).

The protocol used was as follows: E. coli strain HB 101 (pHoKmGus) was transformed

with plasmid DNA isolated from a BCN + clone. Transformants (Tetr Kanr) were then

mated with E. coli C2110 (Nalr ) using pRK2073 as the helper plasmid. Matings were

incubated on LB agar plates amended with tetracycline, kanamycin and nalidixic acid.

Insertion derivatives were mobilized into X campestris pv. vesicatoria strain ME90 (Rif,

Kan r) as previously described and screened for loss of BCN + activity against strain 91-

106. Confirmation of insertions was made by restriction endonuclease analysis.

Marker Exchange Mutagenesis

Triparental matings were performed using BCN insertion mutants (Tet r Kan r) as

donors, 91-118R (Rifr) as the recipient and pRK2073 as the helper plasmid. Rifamycin-

resistant 91-118r was isolated by plating 109 cells onto NA amended with rifamycin,

followed by selection for resistant colonies with wild-type BCN+ and pathogenicity

characteristics. Transconjugants were selected on nutrient agar amended with

tetracycline, kanamycin and rifamycin. They were then serially transferred each day for

10 days onto nutrient agar amended with kanamycin and rifamycin. At the end of the

transfer schedule dilutions were plated onto phosphate media (10g/L NaH2PO4)

containing kanamycin and rifamycin. Single colonies were screened on media containing

all three antibiotics. Colonies which were unable to grow on this medium (Tet s Kan r









Rif ) were selected to represent those in which a successful recombination event had

occurred with concomitant loss of the introduced plasmid (pLAFR3). These were

screened for loss of BCN + activity. Confirmation of the procedure was obtained by

probing marker exchange mutants a probe derived from the transposable element used to

generate the mutants.

Chemical Mutagenesis and Complementation Analysis

Chemical mutagenesis was performed using N-methyl-N'-nitro-N-nitrosoguanidine

(NTG) as described by Carlton and Brown (Carlton and Brown, 1981). Cells from 18 hr

cultures of wild-type T3 strain 91-118 were harvested by centrifugation and resuspended

in an equal volume of Tris-maleic acid buffer (6 g/1 Tris, 5.8 g/1 Maleic acid, 1.0 g/1

(NH4)2S04, 0.25 g/1 FeSO4.7H20, 0.1 g/1 MgSO4.7H20 and 5 mg/1 Ca(N03)2, pH 6.0).

Addition of NTG to a final concentration of 100 pjg/ml was followed by incubation at

370C for 30 or 60 min. Cells were again harvested by centrifugation, resuspended in 1

ml sterile tap water, plated onto nutrient agar and incubated at 280C. Mutants were

subsequently screened for in vitro antagonism against a sensitive indicator strain and

complementation with BCN + clones. For complementation analysis, BCN + clones were

mobilized into 91-118R using a triparental mating scheme.



Results

Screen of Cosmid Library

Ten additional clones conferring BCN + activity were isolated from the X campestris

pv. vesicatoria T3 library when screened against indicator strain 91-106. These were

different from pXV120, pXV442, and pL3XV344, which were isolated in a previous









study (Tudor, 1995). Together, these could be grouped into three subsets of activity

based on cross-hybridization experiments (Fig. 1): BCN-A, consisting of pXV501,

pXV519, pXV711, and pXV754; BCN-B, consisting of pXV442 and pXV699; and

BCN-C consisting solely of pXV120. Clones pXV717, pXV878, and pXV933 were

omitted from further studies due to inconsistencies in plate assays.

Differences Between BCN+ Groups

Differences between in vitro activity of clones were observed between, but not

within groups (Fig. 2A). BCN-A clones consistently yielded the largest zones of

inhibition, even larger than those of the wild-type (wt) T3 strain, while BCN-C+ activity

consistently resulted in the smallest zones of inhibition. BCN-A+ activity was the only

activity to be detected in unconcentrated cell-free extracts (Fig. 2B). BCN+ activity could

not be detected in wild-type extracts.

When tested against selected X campestris pv. vesicatoria indicator strains, BCN

groups showed differential patterns of inhibition (Table 1). All of the tomato race 1(T1)

strains tested were variably susceptible to at least one of the bacteriocin-like activities,

while T2 strains were sensitive only to BCN-A+ activity. Some TI strains, for example,

strain 91-106 which was routinely used as the standard indicator strain for most

experiments, were susceptible to all 3 groups. BCN groups were not uniformly

expressed in all genetic backgrounds, with only BCN-B activity detected in all

backgrounds tested (Table 2). A wild-type T3 strain was not susceptible to any of the

bacteriocin-like activities.
















1 2 3 4 5 6 7 8 9 10 11 12 13
A.



me- ab-




B.

.






C.


m


Figure 1. Southern hybridization analysis of BCN clones. Probes used were
plasmid DNA from A. pXV12.1 (12.1-kb EcoRI BCN-A +subfragment of
pXV519), B. pXV8.9 (8.9-kb KpnI BCN-B +subfragment of pXV442), and
C. pXV5.1 (5.1-kb EcoRI/HinDIII BCN-C +subfragment of pXV120). Lane
designations 1-13 plasmid DNA from: 1. pXV120, 2. pXV442, 3. pXV501,
4. pXV503, 5. PXV519, 6. pXV586, 7. pXV699, 8. pXV711, 9.
pXV717, 10. pXV754, 11. pXV878, 12. pXV933, 13. pXV12.1.


12-kb -I






8-kb -I









5-kb -*


























A


Figure 2. Assays of bacteriocin-like activity against Xanthomonas campestris
pv. vesicatoria strain 91-106. Included in the assays were: A. assay of viable
cell cultures; B. assay of cell-free supernatants. Plate designations were as
follows: 1. 91-118 (BCN ), 2. pXV519 (in ME90) (BCN-A ), 3. pXV442 (in
ME90) (BCN-B ), 4. pXV120 (in ME90) (BCN-C ), and 5. pLAFR3 (in
ME90)









Table 1. Differential inhibition of selected Xanthomonas campestris pv. vesicatoria
indicator strains by the three identified groups of bacteriocin-like activity, BCN-A,
BCN-B and BCN-C.
aIndicator strain Tomato Race BCN-A BCN-B BCN-C


10 T2 b+
XV56 T2 + -
19 T2 + -
92-51 T + +
93-25 Tl ++ -
M103B TI ++ + +
92-48 T2 +
91-106 TI ++ + +
91-108 TI ++ + +
91-111 TI + -
93-25 Tl + +
92-40 TI + -
93-29 TI ++ + +
XV14 TI ++ -
91-118 T3

a Source of strains: Dr. R. E. Stall, Plant Pathology Department, University of Florida,
Gainesville, FL 32611.
b Diameter of inhibition zones measured as distance from edge of test culture to the end
of the zone (mm). Measurements were used to assign three phenotypes which were
scored as: non- inhibition (-), weak inhibition (zone diameters less than or equal to
4mm) (+) and strong inhibition (zone diameters greater than 4mm) (++). Each test was
replicated four times.










Table 2. Bacteriocin expression in different Xanthomonas campestris pv. vesicatoria
recipient strains (BCN-) against X. campestris pv. vesicatoria indicator strain 91-106.

Recipient strain BCN-A BCN-B BCN-C

M103B a + + +
ME90 ++ + +
75-3 +
93-1 + + +
TED-3 + +

a Diameter of inhibition zones measured as distance from edge of test culture to the end
of the zone (mm). Measurements were used to assign three phenotypes which were
scored as: non- inhibition (-), weak inhibition (zone diameters less than or equal to
4mm) (+) and strong inhibition (zone diameters greater than 4mm) (++). Each test was
replicated four times.









Mutagenesis

Transposon insertion mutants deficient in all three of the BCN+ phenotypes were

selected following a screen against indicator strain 91-106. Using these, marker

exchange mutants for each of the wild-type loci were obtained and designated ME-A,

ME-B, and ME-C. These mutants still retained detectable levels of BCN+ activity

(Fig.3) despite confirmation of successful marker-exchange by transposon-probing (data

not shown).

Six hundred NTG-generated mutants were screened for BCN activity. Two were

found to be BCN in standard plate assays and were given the designations NTG-1 and

NTG2-7 (Table 3). One of these, NTG2-7, was complemented by all of the BCN-A

clones, but by none of the members of the other groups. BCN+ activity could not be

restored to NTG-1 with any of the cloned BCN determinants (Table 3).

In Planta Studies

Inplanta T3 strain 91-118 significantly reduced the population of a sensitive strain by

approximately 4 log units relative to the water control, after a period of 48-hr (Fig.4).

This effect was, however, only observed when the T3 strain was applied 12 hr prior to

inoculation with the indicator strain and not when the two were co-inoculated. The

presence of a BCN TI strain did not significantly inhibit growth of the indicator strain

in any of the treatments although, relative to the water control growth was slower.

The NTG-generated BCN mutant (NTG2-7) of strain 91-118 was not capable of in

plant growth inhibition of a sensitive strain (Fig. 5A) as was its wild-type counterpart.

Wild-type T3 strain 91-118 again caused an approximately 4 log unit decrease in the

indicator strain population level relative to the negative control, while NTG2-7 caused










































Figure 3. Assays of residual bacteriocin-like activity in marker exchange
mutants of wild-type BCN + Xanthomonas campestris pv.vesicatoria T3 strain
91-118, against X campestris pv. vesicatoria strain 91-106. Included in the
assay were: wt-T3 (strain 91-118), ME-A (BCN-A-), ME-B (BCN-B-), and
ME-C (BCN-C-). All marker exchange mutants retained BCN-type activity.









Table 3. Screen of NTG mutants for BCN activity and complementation with BCN-A ,
BCN-B+ or BCN-C+ clones against Xanthomonas campestris pv. vesicatoria indicator
strain 91-106.


Characteristic NTG-1 NTG2-7

BCN+ a-
BCN-A ++
BCN-B -
BCN-C -


a Diameter of inhibition zones measured as distance from edge of test culture to the end
of the zone (mm). Measurements were used to assign three phenotypes which were
scored as: non- inhibition (-), weak inhibition (zone diameters less than or equal to
4mm) (+) and strong inhibition (zone diameters greater than 4mm) (++). Each test was
replicated four times.







43

















8






U-
0 4
0)
0

2




0
0 10 20 30 40 50 60
Time(hours)



Figure 4. Time course of growth ofXanthomonas campestris pv. vesicatoria
indicator strain 91-106 in tomato cultigen 7060 when inoculated at the same
time(- -) or 12 hours after treatments (-). Bacterial populations in leaves
were sampled every 12 hours for 48 hours following infiltration with: sterile
tap water (*), 75-3 (BCN -) (m), and 91-118 (BCN ) (A). Results represent
averages of three independent experiments.





























0 10 20 30
TIME(HOURS)


10 20 30
TIME(HOURS)


40 50 60


40 50 60


Figure 5. A. Time course of growth ofXanthomonas campestris pv.
vesicatoria indicator strain 91-106 in tomato cultigen 7060. Bacterial
populations in leaves were sampled every 12 hours for 48 hours following
infiltration with: TI strain 75-3 (BCN -) (*), NTG2-7 (BCN -) (U), NTG2-
7(BCN-A) (BCN-A ) (A), and T3 Wild-type BCN strain 91-118 (X). B.
Time course of growth ofX. campestris pv.vesicatoria test strains listed
above in tomato cultigen 7060.


2.5 +









no significant decrease relative to the negative control. However, when complemented

with pXV519 (BCN-A ) a reduction of approximately 1.5 log units relative to the

negative control was observed (Fig. 5A). Growth of the test strains inoculated prior to

the indicator strain was also monitored for a 48-hour time period (Fig. 5B). There were

no significant differences in growth amongst these strains over the same time period.



Discussion

The results of this study indicate that T3 strains of X. campestris pv. vesicatoria

produce more than one bacteriocin-like compound. At least three clear BCN activities

could be differentiated by hybridization and in vitro antagonism analyses. Three

additional, potential BCN activities were isolated from the T3 genomic library;

however, because of inconsistencies in plate assays they were not studied any further.

Similar inconsistencies were observed in the bacteriocin-like activity of Pseudomonas

solanacearum (Frey et al., 1996). Retention of bacteriocin-like activity in marker-

exchange mutants for each of the groups confirms the production of multiple compounds

by T3 strains as inactivation of single wild type loci was insufficient to abolish in vitro

antagonistic abilities. Furthermore, it indicates that expression of the three determinants

occurs independently of each other.

The production of multiple bacteriocin-like compounds by a single strain is rare

(Vidaver, 1983); however, it has been suggested to occur in strains of X campestris pv.

glycines. Fett et al, (1987) suggested that strains of X. campestris pv. glycines produced

multiple bacteriocins. This was based on findings that the bacteriocin activities (of a

single strain) detected towards the different indicator strains were variably sensitive to









protease and heat treatments. However, these activities were never resolved and

demonstrated to occur independently. In another study five cosmid clones conferring

bacteriocin-like activity were isolated from aX. campestris pv. glycines cosmid library

(Ahn and Cho, 1996). Cross-hybridization experiments with these clones revealed the

presence of multiple bacteriocin-like activities within the single strain. However,

differences between these groups were not further examined.

Three distinct bacteriocin-like activities were resolved from T3 strain 91-118, which

could be differentiated by inhibitory spectra. A similar situation was observed in strain

TA33a of Leuconostoc mesenteroides, which produced three bacteriocins with different

inhibitory spectra (Papathanasopoulos et al., 1997). Bacteriocin specificity is in part

determined by the presence of specific receptors on the surface of sensitive cells to

facilitate binding and entry of bacteriocin molecules into the cell (Lakey, et al., 1994).

Differential distribution of these receptors among strains may be the reason for the

specificities observed in the inhibition spectra for these three compounds.

A strain capable of producing several compounds against competitors likely has a

selective advantage in natural ecosystems. In fact, studies on the changing tomato-race

distribution of X campestris pv. vesicatoria populations in Florida (Jones et al., 1998a)

may be a reflection of this. Up until 1991 T3 strains were not found in Florida; however,

in field-surveys over the following 4 years, T3 strains have become increasingly

predominant, and as of 1995 represented some 55% of the total strains isolated from

tomato fields (Jones et al., 1998a). Since all of the Tl strains tested were susceptible to

at least one of the bacteriocin-like activities, and TI strains are the only other race found

on tomato in Florida, it is very tempting to speculate that the antagonism between these









strains has played a role in this race shift. In addition no other obvious factors commonly

associated with sudden population race shifts could be identified in this situation (Jones

et al., 1998a).

Results of inplanta data show that T3 strains are inhibitory to Tl indicator strains in

leaves of a susceptible tomato genotype. This provides evidence of a direct selective

advantage for T3 strains in the infection court in which they both exist. Studies have

documented the coexistence of multiple races within a single leaf-lesion (O'Garro and

Tudor, 1994); an indication that site competition is likely to occur between these two

races in a susceptible host. The reduction in indicator strain population level is

comparable to that achieved in the hypersensitive response of tomato line Hawaii 7998 to

Tl strains (Whalen et al., 1993). Similar growth levels of test strains indicate inhibition

of the indicator strains is not the result of a plant effect. Results of NTG2-7 ( BCN )

mutant analysis inplanta indicate that production of these bacteriocin-like compounds is

important in mediating the observed antagonism. The fact that NTG-mutagenesis has the

potential to generate mutations at multiple sites within the genome (Carlton and Brown,

1981) may account for the BCN phenotype in NTG-1, however, complementation of

BCN+ activity in NTG2-7 by BCN-A+ clones indicates that it is a simpler mutation

event. Antagonism, however, only occurred when T3 strains were pre-emptive. In plate

assays similar results were obtained ( Tudor, 1995). Results such as this have prompted

the speculation that the major use of these types of compounds in biocontrol strategies

would be prophylactic rather than curative (Vidaver, 1983). In a recent study, a non-

pathogenic T3 strain of X. campestris pv. vesicatoria was shown to reduce the severity of

bacterial spot disease in the field when applied prophylactically (Liu, 1998).






48


This study and others (Jones et al., 1998a; Liu, 1998) have provided some initial

insights as to the potential use of bacteriocin-like producer strains in the biocontrol of

bacterial spot disease on tomato. The lack of effective control measures to manage this

disease justifies a further investigation of this strategy, perhaps involving a more

thorough screen for these types of compounds. In this case, identification of a

compound with activity against T3 strains would be particularly useful.















CHAPTER 4
CHARACTERIZATION OF GENETIC DETERMINANTS AND
EVALUATION OF THEIR ROLE IN ANTIBIOSIS

Bacteriocins and bacteriocin-like compounds encompass a vast array of structurally

different, ribosomally synthesized polypeptides (Riley, 1999). Typically, several

genes, which are often encoded in operons, are involved in their production. These have

been extensively characterized for bacteriocins produced by Escherichia coli and several

Gram positive bacterial species. In contrast, little is known about the genetic

determinants involved in bacteriocin production by phytopathogenic bacteria.

Purification of bacteriocins produced by Pseudomonas syringae and P. solanacearum

from culture supematants, indicated that these were high molecular weight compounds

(Vidaver, 1976); however, there has been no subsequent isolation of the genes involved

in their biosynthesis. There are other examples where the approximate molecular weight

of these compounds have been roughly determined, but there is a similar lack of

information on the genes involved (Matsuo et al., 1981; Chuang et al., 1999). The

genes involved in agrocin 84 synthesis and immunity were located on a 48-kb conjugal

plasmid designated pAgK84. Partial characterization of the locus revealed five

complementation groups, which spanned a 21-kb region of the plasmid (Wang et al.,

1994). Similarities of these genetic determinants to those of other bacteriocins are

expected to be mininal since this atypical bacteriocin-like compound lacks a biologically

active protein component. More recent reports on agrocin 84 now refer to it as an

antibiotic (Wang et al., 1994). Genes involved in the regulation of bacteriocin









production by Erwinia herbicola (Vanneste and Yu, 1996) and E. carotovora subsp.

carotovora (Chuang et al., 1999) have been isolated. However, the corresponding

bacteriocin structural and immunity genes were not isolated in either of these studies.

Recently, a gene involved in the synthesis of a bacteriocin produced by Xanthomonas

campestris pv. glycines strain 8ra has been cloned and sequenced (Heu and Cho, 1997).

Studies on the use of bacteriocin-producing strains for the biocontrol of plant diseases

have provided little proof as to the actual role of antibiosis in the process. The direct

detection of antibiotic activity in plant tissue provided the earliest evidence for this.

Syringacin W-1, a bacteriocin produced by Pseudomonas syringae was purified from

infected red kidney bean stems. This along with the decrease in the population of a

bacteriocin-sensitive strain in plant tissues provided evidence that syringacin W-1

production was involved in competition among P. syringae strains in bean plants (Smidt

and Vidaver, 1982). The identification of genes involved in antibiotic production would

provide a new basis for evaluating the contribution of antibiosis in biocontrol strategies.

Disruption of antibiotic genes followed by an assessment of their residual competitive

potential is now one of the most frequently used strategies to examine the role of

antibiosis. Tn3-generated mutants of Erwinia herbicola which did not produce antibiotic

activity against E. amylovora were less effective at suppressing disease development than

their antibiotic-producing counterparts (Kearns and Mahanty, 1998).

Novel biotechnological applications such as the development of "super" biocontrol

strains and transgenic plants expressing antimicrobial compounds, requires knowledge of

the mechanism by which specific gene products are involved in the biosynthesis of these

compounds. The genes involved in the biosynthesis of cecropin and a number of other









antibacterial compounds have been well characterized and successfully expressed in

transgenics (During, 1996).

Knowledge of the genetic determinants involved in antibiotic production is

advantageous as it provides information on their potential use in biotechnological

applications and facilitates easier evaluation of their role in biocontrol strategies. The

goals of this study were therefore, to characterize the genetic determinants involved in

the production of bacteriocin-like compounds secreted by Xanthomonas campestris pv.

vesicatoria T3 strains. At least three different bacteriocin-like activities are associated

with T3 strains. Individual genes and corresponding mutations were used to evaluate the

relative contribution of each to competition between strains in plant. Initial evaluation

of the expression of one of these compounds was examined inplanta and in vitro using

Tn3-gus promoter fusions.



Materials and Methods

Bacterial Strains, Plasmids, and Culture Conditions

The bacterial strains and plasmids used in this study are listed in Appendix A. All

strains of X campestris pv. vesicatoria, and X campestris pv. glycines were grown on

nutrient agar (BBL) at 28 C. Strains of E. coli were grown on Lauria-Bertani (LB)

(Miller, 1972) medium at 37 C. Antibiotics were used to maintain selection for

resistance markers at the following concentrations: tetracycline (Tet) 10 [g/ml;

rifamycin (Rif) and ampicillin (Amp) 100 [g/ml; Spectinomycin (Spec), kanamycin

(Kan) 50 pjg/ml and nalidixic acid (Nal) 20 [g/ml.









In vitro Antagonism and Immunity Assays

Inhibitory assays for X campestris pv. vesicatoria strains were conducted as

previously described (Tudor, 1995). Bacteriocin production by strains of X campestris

pv. glycines was detected as described before (Fett et al., 1987). Producer strains were

grown overnight at 280C in test tubes containing 4 ml nutrient broth. Cells were

harvested by centrifugation and resuspended in sterile tap water to final concentration of

5 x 10s CFU/ml. Aliquots of 20 pl were inoculated, (five per plate), onto nutrient agar

plates and incubated at 200C for 48 hr. Producer cells were killed by exposure to

chloroform vapors for 1 hour followed by overlaying the agar surface with 10 ml sterile

0.7% water agar, containing 200 pl of a 5 x 10s CFU/ml suspension of actively growing

indicator cells. Plates were incubated at 280C for 24 hr. Bacteriocin-like activity was

scored as clear zones of inhibition surrounding producer colonies while the lack of

inhibition zones was indicative of immunity. Each producer-indicator strain combination

was replicated four times.

In Planta Antagonism Assays

Assays for inplanta antagonism were conducted as previously described (refer to

Chapter 3). Representative clones from each of the bacteriocin-like activities and marker

exchange mutants with singly inactivated wild type loci were screened for their ability to

inhibit a sensitive indicator strain when inoculated 12 hr prior to the indicator strain.

Wild type T3 strain 91-118 was routinely used as a positive control and a BNC- TI strain

as the negative control. Antibiotic resistance markers were used to facilitate selective

monitoring of indicator strain populations. Each treatment consisted of three replications

and each experiment was repeated twice. Population data were analyzed by the general









linear model procedure and means were compared by Fisher's LSD test at a P = 0.05.

All statistical analysis was performed using the Statistical Analysis System (SAS

production release 6.12, SAS Institute, Inc., Cary, NC).

DNA Extractions

Plasmids from E. coli strains were extracted using the alkaline lysis method described

by Sambrook et al. (Sambrook et al., 1989). Plasmid extractions from xanthomonad

strains were conducted using the method of Kado and Liu (1981).

Total genomic DNA was extracted from xanthomonad strains using the protocol

described by Ausubel et al. (1987), with minor modifications. Bacterial cells were

harvested from 1.5 ml broth cultures, washed in 1 ml distilled water and resuspended in

567 jl of TE buffer (10 mM Tris-HCl, pH 8.0; 1 mM EDTA, pH 8.0). Cells were then

lysed for one hour in a solution containing proteinase K (100 pjg/ml) (Boehringer

Mannheim, Indianapolis, IN) and 0.5% sodium dodecyl sulfate (SDS) (Sigma Chemical,

St. Louis, MO). NaCl (0.7M) and hexadecyltrimethylammonium bromide (1.0%) were

added and the mixture was incubated at 65 C for 20 min. DNA was extracted by

sequential chloroform-isoamyl alcohol (24:1) and phenol-chloroform-isoamyl alcohol

(25:24:1) extractions and precipitated with 70% ethanol. Pellets were resuspended in 100

jpl TE and stored at 4 C.

DNA Manipulations

Standard techniques for molecular cloning were as described by Sambrook et al.

(1989). Restriction endonuclease digestions were performed using manufacturers

specifications. All enzymes were obtained from Promega (Madison, WI). The

QIAquick gel extraction kit (Quiagen, Valencia, CA) was used to gel purify restricted









fragments. T4 DNA ligase was used according to manufacturers specifications

(Promega, Madison, WI). Constructs were transformed into competent DH5c cells.

Competent cells were prepared using rubidium chloride as follows. One ml of an

overnight culture of DH5ca cells grown at 370C in LB medium, was inoculated into 100

ml of the same medium. Cultures were grown at 370C with vigorous aeration, until an

OD550 = 0.4 (approximately 5 x 107 CFU/ml). Cells were harvested by centrifugation at

2000g for 10 min at 40C. Cells were then gently resuspended in 10 ml ice-cold buffer 1

(10 mM MOPs pH 7, 10mM RbCl2) and centrifuged again under the same conditions.

The pellet was resuspended in 5 ml buffer 2 (100 mM MOPs pH 6.5, 50 mM CaCl2, 10

mM RbCl2) and placed on ice for 30 min. The centrifugation steps were repeated and the

cellular pellet was resuspended in 5 ml of ice-cold buffer 3 (100 mM MOPs, pH 6.5, 50

mM CaC12, 20% glycerol).

Bacterial Conjugations

All recombinant clones generated were screened after conjugation into X campestris

pv. vesicatoria, usually strains ME90 (Rif, Kanr) or 93-1R (Rif). Conjugations were

performed using the previously described triparental mating technique (refer to chapter

3). A spontaneous rif-resistant mutant (XP 144R) of X campestris pv. glycines strain

XP144 was generated for use as a recipient. Rifamycin-resistant strains were isolated

after plating approximately 109 cells onto NA amended with rifamycin, followed by

selection for resistant colonies.

Southern Hybridization Analysis

Southern hybridization analysis was conducted as previously described (refer to

Chapter 3). A 2-kb Sall, BCN-A-specific probe (see Fig. 4 for corresponding Sall sites)









was labeled by random primer incorporation of digoxigenin-11-deoxyuridine-

triphosphate (DIG-dUTP) (Feinberg and Vogelstein, 1983) and used to probe plasmid

and genomic DNA of selected X campestris pv. vesicatoria and X campestris pv.

glycines strains. Genomic DNA was digested with Sall, while plasmid DNA was

undigested. DNA to be analyzed by Southern analysis was electrophoresed on 0.75%

agarose according to standard protocols (Sambrook et al., 1989).

Physical Characterization of BCN+ Cosmid Clones

Restriction endonuclease mapping and Southern hybridization analysis were used to

determine regions of overlap between cosmid clones. Restriction endonuclease mapping

was also used to plot the approximate insertion points of Tn3 into cosmid clones. Based

on restriction mapping of clones and approximate localization of Tn3 inserts, several

deletion constructs of BCN+ clones were made utilizing available restriction sites.

Deletion constructs were transformed into the plasmid vector pLARF 119 and screened

for in vitro inhibition and immunity activity.

DNA Sequencing

The smallest deletion derivatives retaining BCN+ activity and immunity were

sequenced. A 10.5-kb region derived from pXV519 (BCN-A ) and a 5.1-kb EcoRI

fragment derived from pXV120 (BCN-C ) were sequenced. For sequence analysis,

fragments were cloned into the phagemid vector, pBluescript II KS using appropriate

enzymes and sequenced using the standard T3 and T7 and custom designed

oligonucleotides. The primer-walking strategies used for sequencing are shown in

figures 1 and 2 and the sequences of custom designed oligonucleotides shown in

appendix B. DNA sequencing was performed by the DNA Sequencing Core Laboratory






















T3___
136L_0
E B
I I


138 1
1410I
K 151 152 .
I


139
- 137
T7


144 1_. 132
140 123
-- T7 T7


1 kb



Figure 1. Primer walking strategy used to sequence pXV12.1 (BCN-A+). Subclones derived from pXV12.1
were cloned into pBluescript II KS before sequencing. Subclones generated for sequencing were: pBXV4.5
(4.5-kb BamHI/EcoRl fragment); pBXV6.2 (6.2-kb BamHI fragment); and pBXV3.0 (3.0-kb BamHI/Kpnl
fragment). E (EcoRI), B (BamHl), K (Kpnl).


T3





















T3
157
H 1 166 E

169 167
161
44 156


1 kb


Figure 2. Primer walking strategy used to sequence pXV5.1 (BCN-C ), a 5.1-kb EcoRI/HinDIII subclone
derived from pXV120 (BCN-C ) cloned into pBluescript II KS. H (HinDII), E(EcoRI).









of the Interdisciplinary Center for Biotechnology Research, University of Florida,

Gainesville FL. The exact location of Tn3-gus promoter fusions were determined by

sequencing using an oligonucleotide complementary to a sequence in the N-terminus of

the GUS gene (5' GATTTCACGGGTTGGGGTTTC 3'). DNA sequence analysis was

performed with the Sequaid II program (Kansas State University, Manhattan, KS).

Database searches were performed using TBLAST (Altschul et al., 1997).

Construction of Promoter Fusions and Assays for 13-glucuronidase Activity

Promoter fusions to the 13-glucronidase (gusA) gene were obtained after random

insertion mutagenesis of pXV12.1 in E. coli with a promoterless Tn3-gus construct.

Mutants were transferred to X. campestris pv. vesicatoria strain 93-1R by a triparental

mating scheme before testing for loss of BCN-A and detection of 13-glucronidase

activity.

In vitro 3-glucuronidase activity of Xanthomonas strains was measured after 24 hr of

growth strains in nutrient broth (NB), King's medium B (KMB), KMB amended with

two levels of FeSO4 (0.01 mM or o.1 mM) and XVM2. Activity was also measured after

growth in susceptible tomato genotype Florida 7060 at 24 hr intervals for 96 hr. For

assays of strains grown in broth, cells were harvested and resuspended in gus-assay

buffer (50 mM NaPO4 [pH7.0], 10 mM Dithiothreitol, 1 mM EDTA, 0.1% Sarcosyl,

0.1% Triton X100 and 2 mM 4-methylumbelliferyl 3-D-glucuronide[MUG]). For assays

of bacteria grown in plants, tomato leaflets were inoculated with 5 x 108 CFU/ml in

sterile water as previously described. At 24 hour intervals, leaf discs (0.5 cm2) were

macerated in 0.5 ml sterile distilled water and a 50 pl aliquot was removed for the assay.

The number of bacteria per assay was calculated by plating appropriate dilutions of leaf









extracts or media aliquots onto NA amended with rifamycin. Each treatment was

replicated twice and each experiment was conducted at least twice. Fluorometric assays,

using 4-methylumbelliferyl 3-D-glucuronide as a substrate were conducted to test j3-

glucuronidase activity (Jefferson et al., 1987) using a CytoflorlI fluorescence multiwell

plate reader (Perseptive Biosystems, Foster City, CA). A construct containing the lacZ

promoter driving the expression of the 13-glucuronidase gene and another with a stop

codon inserted within the gene, were transferred to X. campestris pv. vesicatoria strain

93-1 and used as positive and negative controls respectively. One unit of (3-

glucuronidase was defined as nanomoles of MUG released per minute and activity is

reported as units per CFU.

Assay data were analyzed by the mixed procedure of Statistical Analysis System

(SAS production release 6.12, SAS Institute, Inc., Cary, NC) and means were

compared by Fisher's LSD test at aP = 0.05.

RNA Extraction and RT-PCR

The acid guanidinium-phenol-chloroform method was used to isolate total RNA from

Xanthomonas (Chomczynski and Sacchi, 1987) with minor modifications. One ml of

cells from overnight cultures were harvested by centrifugation at 40C (8 000g) for 2 min

and resuspended in 400 jpl in solution D (4 M guanidinium thiocyanate, 25 mM sodium

citrate [pH 7], 0.5% sarcosyl, 0.1 M 2-mercaptoethanol). Sequentially, 50 pl of 2 M

sodium acetete, pH 4, 500 jpl phenol and 150 jpl chloroform-isoamyl alcohol mixture

(24:1) were added to the homogenate followed by incubation on ice for 15 min. Samples

were centrifuged at 10 000g for 20 min at 40C. The aqueous phase was transferred to a

fresh tube, mixed with and equal volume of isoproponol and precipitated at 800C for









20 min. RNA was pelleted by centrifugation at 10 000g for 20 min at 40C and

resuspended in DNase reaction mix containing 30 U RQ1 Rnase-Free Dnase and 1U

RNasein ribonuclease inhibitor (Promega, Madison, WI) in a 100 Pl volume. Reactions

were incubated for 30 min at 370C. Residual DNase was removed by phenol-chloroform

extraction and the resultant pellet resuspended in 20 ipl sterile diethyl pyrocarbonate

(DEPC)-treated water.

Reverse transcription (RT)-PCR was conducted using primers specific to the putative

coding region of the gene involved in BCN-A+ activity. Primers custom primersl40 and

144 (appendix B) were used for first strand synthesis, followed by the addition of

primers 158 and 165 for product amplification. The annealing positions and sequences of

these primers can be seen in Fig. 4 and appendix B respectively. The Enhanced-Avian

RT-PCR kit (Sigma Chemical, St. Louis MO) was used for all reactions.. The reverse

transcriptase reaction was carried out in 50 mM Tris-HCl, pH 8.3, 40 mM KCl, 8 mM

MgCl2, 1 M DTT, 500 jtM each dNTP, 1 jpm reverse primers, 0.25 ag/Pl template

RNA and 25 U Enhanced Avian RT. In the negative control reverse transcriptase was

omitted. The reaction mixtures were incubated at 420C for 50 min. Half (10 Pl)of the

cDNA was removed and used for PCR amplification with the addition of the following

components to the final concentrations of: 50 mM Tris-HCl, pH 8.3, 15 mM ammonium

sulfate, pH 9.3; 2.5 mM MgCl2; 0.1% Tween 20; 200 jtM of each dNTP; 500 nM

forward primers; 0.05 U/jpl AccuTaq LA DNA polymerase. PCR was carried out in 35

cycles of 940C for 1 min, 500C for 1 min and 720C forl min, finishing with an

extension at 720C for 10 min. RT-PCR products were digested with restriction enzymes









known to have sites within the region, for confirmation of the product. Samples were

electrophoresed on 1.25% (wt/vol) agarose gels as described previously.

Ammonium Sulfate Fractionation and Ultrafiltration

BCN-A activity was partially purified from cell-free extracts of stationary phase

cultures by ammonium sulfate precipitation. Solid ammonium sulfate was slowly added

to the culture supernatant to 30% saturation at 40C with constant stirring for 2 hr.

Precipitated proteins were pelleted by centrifugation at 10 000g for 30 min at. 4'C and

resuspended in one tenth volume 10 mM phosphate buffer pH 6.8. The fraction was

assayed for in vitro inhibition activity and passed through a series of membrane filters

(Amicon) with molecular weight cut-off limits of 10-, 30-, 50- and 100- kDa,

respectively. Resultant filtrates and retentates were also assayed for in vitro inhibition

activity.



Results

Physical Characterization of BCN-A+ Clones

The four cosmid clones identified in this group, pXV501, pXV519, pXV711 and

pXV754 contained 23-kb to 28-kb inserts ofX. campestris pv. vesicatoria race T3 DNA.

Southern hybridization and restriction endonuclease mapping using EcoRl, HinDIII and

BamHI, were used to delimit an approximately 15-kb region of overlap between the

clones. Each of the four clones contained all of the information necessary for expression

of BCN-A+ activity in a heterologous X campestris pv. vesicatoria TI strain (ME90). In

addition, mobilization of these clones into this BCN-A-sensitive strain rendered it









immune to BCN-A activity, but not to BCN-B BCN-C or wild-type BCN This

phenotype, which conferred resistance to BCN-A was designated IMM-A+.

Tn3-insertion mutagenesis was conducted on clone pXV519 and the approximate

locations of insertion points mapped to EcoR1 fragments using EcoR1, HinDIII and

BamHI (Fig.3). Five EcoRI fragments with approximate sizes of 12.1-, 5.0-, 3.8-, 3.7-,

3.4 and 1.1-kb respectively were mapped to pXV519. Insertions, which led to a loss of

BCN-A+ activity all mapped to the largest, 12.1-kb, EcoRI fragment and were

differentially distributed between two BamHI/Kpnl and BamHI/EcoRI subfragments.

Other pXV519::Tn3 insertions, which retained BCN-A+ activity, were used to delimit

an approximate 9.5-kb region required for activity. The 12.1-kb EcoRI fragment was

cloned into pLARF 119 generating pXV12.1, further subcloning from this generated

pXV8.0 (8.0-kb EcoRI/Kpnl fragment) which retained both the BCN-A and IMM-A

phenotypes. Although further subcloning of pXV8.0 yielded no smaller BCN-A+

subclones using PstI, Sall, EcoRV, Xbal, Sinal, or SstI, pXV4.5 (4.5-kb

BamHI/EcoRI fragment) was obtained which was sufficient to confer the IMM-A+

phenotype.

Sequence Analysis of pXV12.1 (BCN-A )

A 10.5-kb segment of pXV12.1 was sequenced and the complete sequence is shown

in Fig.4. The sequence has a 54% GC base composition. The DNA sequence when

translated into all possible open reading frames (ORFs), revealed 19 possible ORFs

greater then 250 base pairs (bp). Deletion analysis and mapping of Tn3 insertion sites,

eliminated all but seven of these from involvement in BCN-A+ activity, and all but 5

from IMM-A+ activity (Fig.3). ORF 1 was contained in the minimum region required for


















2 17* 22 33 26 25 28
7 5 14 20 30 16 3
58* 55 1 34 23 15 24
10 11


E B K B E E E E ElH

ORF 1 bcnA t 2 3 4 5 6
ORFs 1 kb

I I (pXV12.1) BCN-A+ IMM-A

I (pXV8.0) BCN-A+ IMM-A

I (pXV4.5) BCN-A- IMM-A

(pXV6.2) BCN-A- IMM-A-

(pXV3.5) BCN-A- IMM-A-

(pXV4.8) BCN-A- IMM-A-


Figure 3. Partial physical map of pXV519 (approximately 29-kb) (BCN-A+), subclones, and approximate
locations of Tn3-gus inserts. Inserts enclosed by a circle all resulted in a loss of BCN-A+ function. The
precise location of some of these was determined by sequencing and, where known is indicated by an asterisk.
The orientation of the lacZ gene in one of these is shown by an arrow. The location and phenotypes of
selected sunclones is also shown here. The restriction endonucleases are abbreviated using the following
code: E (EcoRI), B (BamHIl), K (Kpnl) and H (HinDIII).








64



1 GGATCCGGCCACGCCGGACGCCCCGCTGGTGCCGGTGGAGTTGCAAGACGCACGCCTGGG 60
61 CGTGGCCGGCACCTTGAAGGCCTGGGCCGCGATCGGCCGCGCCTCCCTGGAGCGCGATGG 120
121 TCAGCAGGCCGAGCTGGTCTTCGACAGCCGCGGCAACGACCAACGTGCACAGCTCAAGCA 180
181 GGTGCAGGCAAAAACGCCGGGCGGCGCGCTGGATCTCACCGGTGAAGTGGCCTGGGCGCC 240
241 GGAGCTGCAATGGGATGTCAGCGCGCAACTGGCCAAGTTCGACCCGGGCTATTTCGCGCC 300
301 GGGTTGGAACGGCAACCTGTCCGGCAAGATCGCGTCCAAGGGTCGCCAATTGCCGGCACC 360
361 TGCCGGCGGTGTCTCGCCGGGGTTCGAGGCCACTGCAGAGATACCCAGCCTGACCGGGCA 420
421 GCTGCGTCAACGCGCGCTGTCGGGCAACGGCAAGTTCGCGTTGCGCGGCGAACAGGGCGA 480
481 AGGCGAATTGCAACTGGCGCTGGGCAATAGCCGCATCAACGCCAAGGGCAAGGTGGGCGA 540
541 CCAGCTCGATATCGCCGCGCAGCTGCAGCCGTTGCAGCTGGACGACGTGCTGCCTGGCGC 600
601 CACCGGCATCGTGCGCGGGCAACTGCAGGTCAGCGGCAAGCGCGACGCACCCGACATCAC 660
661 CGCCGACATTGCCGGCAACGGCTGCGTTGGGACACTATAGCGCCAGACATCAGCCTGCTG 720
721 GCCGGCTGCCGTGGCGTGGCAGCGATGGGCAACTGGCCCTGCAAGGCACCGCCATCGAAG 780
781 CCGGCGTGGTGCTGGATAGCGTGCGCGTGCAGGCGCGCGGCGCGGTGGAGGCCTTGCGAC 840
841 TGGATGCGGACATCGCCAACAGCATGGCCAGCGTCGCCCTGCAGGGCGATGTGCGGCGCA 900
901 ACGGCGAGCGCTGGCAGGGGCAGGTCGCCACGCTGCGCATCGCACCGGCCAAGGGCGATG 960
961 CCTGGGCGCTGCGCAAACCGGCGCAGTTCAGCACCGATGGCGCCGCGTTTACCTTCTCCG 1020
1021 -ACACCTGCCTGGGCGCGGCCACCGGCGGCGCGCTGTGCGCCAGCGCCAACTGGCCGCGCG 1080
1081 -AGGGCATGGTGGTGCACGGCGACGCGTTGCCGCTGTCGCTGGTGCAGCCGTGGCTGCCCA 1140
1141 -AGCAGGAAGGTCGGCAGATCTATCTGCGCGGCGAACTGTCGCTGGACGGCAGCTTCAAGC 1200
1201 -CGCGCGGCAACGCCTGGGAAGGCAGCCTGCGCATCGCCTCGCCCGAAGGCGGCATTCGCT 1260
1261 -TGGGCGAAACCCGCTACGCGGCAGTGGCCGGCAATCCCAATCGCGGCGAGTTGCTGCGCT 1320
1321 -ACGACCAATTCAGCGTGCAGGCCGATTTCACCCAGCAACAGATCCAGGGCAAGCTCGGTA 1380
1381 -TCGGTTTCCAGGGCGCCGGCTTTGTCGATGCCAAGTTCAATACCGGCTGGGATGCGTACG 1440
1441 -CGCCGCTCAACGGCGAGCTGTATCTGAACATGTCGCGGCTGTACTGGCTGGAGCTGGTGA 1500
1501 -TTGCCGACGTGGTGCGGCCGAAGGGGCTGGTCGAGGGCCACGTGAGCCTGCGCGGCACCC 1560
1561 -GCGACAAACCGCTGCTCGGCGGCGATGCCACCTTGAGCGATTTCACTGCCGAATACCCGT 1620
1621 -CGATGGGTTTGACGCTGAGCGAGGGCAAGGGCCGCTTCGATGCGCTGCCGGACGGCTCGG 1680
1681 -CCAAGATCACCGCCTCGGCCAAGTCCGGCCCCGGCACGCTCACCGTCGATGGTGGGTTGT 1740
1741 -CGTGGTTCGGAACCAGCACCCCGTTGCTGCTCAATATCCGCGGCGACAACGTGCTGGCCT 1800
1801 -ACAACACCAGCGAGTTGCGCATCATCGCCAACCCGGACATGCAGTTCGGCATTACCGACA 1860
1861 -ACACCATGCAGTTGCGCGGCAAGGTCACCGTGCCCGAAGCGGACATCGACCTGGAACGCC 1920
1921 -TGGACCGTGGCACCTCGGTGTCCGAAGACGTGGTGGTGCTGGACCCGGTGGACCCGGAAC 1980




Figure 4. Nucleotide sequence of a 10.5-kb portion of pXV12.1 containing all of the
genes required for BCN-A+ expression. The deduced amino acid sequence is given for
bcnA The putative ribosome binding site (rbs) is underlined and the promoter-like -10
and -35 regions are overlined. The positions of two Sail sites and annealing points of
four primers (140, 144, 158, and 165) used for RT-PCR are indicated.








65





1981 -AAACGCCCGCATCGCCATTGGACATGGACCTGGCCATCGTGCTCGGCGACAAGGTCAACA 2040
2041 -TGAGCGGCTTCGGCCTCAAGGGTGGACTGAGCGGGCAGATGCAGATCCGCGCGCGCCCTG 2100
2101 -GCCGCGAAATGACCGCCAATGGCGGGCTGGATGTGCGTGGGCGCTACAAGGCCTATGGCC 2160

2161 -AGGACCTGACCATCACCCGTGGCCAGCTGACCTGGAACAACAATATCGTCTCCGACCCGC 2220
2221 -GCGTGAGCCTGCGCGCCGAACGCAAGATCGGCGATGTCACCGCCGGCATCGACGTCAGCG 2280
2281 -GCCGCGCCGAATCGCCGCGTGCCGATGTGTGGTCCGAGCCGGCGATGTCGCAATCGGAGG 2340

2341 -CGCTTTCGTACCTGGTCCTCGGCCGCGGGCTGTCCACTGCCAGCAGCGACGAAACTCAGC 2400
2401 -AGGTCAGCGCCGCCTCGGCGGCACTGTCGGCCGGCAGCAGCCTGATCGCCTCGCAGATCG 2460
2461 -GCGCCAAGCTCGGCCTGGACGAAGCCGGCGTCAGTCAATCAAGCACGCTGGGCTCGGTGG 2520

2521 -TGGGTTTCGGCAAATATCTCTCGCCCAAACTCTACGTCGGCTACGGCGTGTCGATGGTCG 2580

2581 -GCGGCGGCTCGGTACTGACGCTCAAGTACCTGCTCAGCCGCGGCTTCGATATCGAAGCCG 2640
2641 -AATCCAGCACCATCGAGACCAAGGGGTCGGTGAACTGGCGGCGGGAGAAGTAGGCTATCG 2700
2701 -GCAAACTACAGCGGATTGACAAAGCTGACTCAGCTGTTCATGCTGGCTGACAAGTGTGGT 2760

2761 -ATCCGTCACGCAGGCTTCTTCAGGGGACCGAAGAGAGTTGCCTGCCGCCAATCAATCAGG 2820
2821 -GGAAGGAAAAGTGTCTTACCAAGGACGCTCGGGTTTGCGGTACCTGTTGCTAACAGCGAC 2880
2881 -GTTGGTCAGCGGCGGCGCCAGCGCTGTTGAAGTAACCGCCAACGGCAAAATCGCTTTCGT 2940

2941 -GGAAAACGGTTGGTATGGAGAGGGGCTGTCCCTCGCACACTCCGCGCCCGTAGCAGGCTG 3000
3001 -CACCGGTGGCGCAAACAATTACGTCATCGATAAGAACCATCCGTCTTACCAGGAGTTGCT 3060
3061 -TGCGATCGCATTGACCGCATACACCCGCGAACTTGATGTCAAGCTGATCGTGGAGCCGGG 3120

3121 -CGTCTGCGGATTTGGTGGGCGCACGAAGATACTGGCGATGCGACTCAGTAAATGAAGTCA 3180
3181 -CTTAAAAATCTTTCGCTCTATAAGCCAGTATCCAAATAAATATTCTAGCAGATGGCTTGA 3240
3241 -TAGCCGAGCCTATCTTAGGCGCCTGGAGATAAATCGAAAATGAGAAATATATCAAAAACT 3300

3301 -CTAACGCTGTTTGGGTTCGCATTCCTCTGTGGCAAGGCTGAAGCACTCCAGTACATTTCT 3360
3361 -GGGCGAGTGACATCTATCGAGGCGACATACATGCCTGCGCAAATCCCTTTTGTTCTAACG 3420
3421 -GCTGGCAATGCAGCATGCCCTGCTGGAAAGCCAGTTTATTGGGCAAATCCCAACTTGGAT 3480

3481 -AACAACAAAGCAATTTACGCAGCTTTAGTAACTTCTTTAACGACCGGCAAAAAATAACAT 3540
3541 -TCATCATTGATGATAATGATGTGTCATGCAATGGGCGCTTCTTATATCTTGAAGACTAAC 3600
3601 -AGGCGCATCCAAATCTAGCATGCGACTCAAAGATATCGCTCCATAATAAAAAACCTAGAA 3660

3661 -AAATTTAGATAATTGCATTTTTATCCTGCAGTCTTCAATACAGCCTAAGTATCCGGGGGA 3720
3721 -CCAAGAAAATATGTCGGGCGTAAATTATAAAAGCGTCGTTGCGCTAATATTTTTGATTTC 3780
3781 -AGCGCTGCCCTTGCAGGCACAGATGGTTTCTCCCGAATCAGAGTTCGACAAACGCGTACG 3840

3841 -TAGCTCTCAATCGGTGCAGCCACTTGGCGATAAGCCCTTTGGCGAGCTGGTTGATCTTTA 3900
3901 -TACGGGGGCAGTCTCATTTAGGCAGGTGGACATAGTTCTAGAAGGACAGGGTCTTCCCAT 3960
3961 -TGAGCTTTCTAGAACGTTAGAAATCAATGATCGCAGTTTCGCGCAAACTGATTCGCGTGC 4020

4021 -TTTTACGGACTGGCGTCTGGAAGTCCCGCGAATTGAAACGATAATTGCTGATAGATTCAA 4080


Figure 4, continued














4081 -TTTGAGATGCGCGAACATTTCAGGCGCACCGACTATAGTGTTGAATAGCAGTGCCTCTTC 4140
4141 -TAGTCCGATTTTCTATCCCCCTGAGCAGTGGTGGGGCGGCTATCAGCTGGTCATTCCTGG 4200

4201 -AAGTGGCCGCCAAGATTTATTAAAGCGCGATGCATCGAACGGACTTGCGCCGCAACCCAT 4260


158
4261 -GAGTGTGGACGGTGGCGTCGTCGATTTTCCGGTCCTGACAAAAAATCAATGGATGATCGG 4320

4321 -TTGCCTCCAAAAAACTCTAACGGTCTGGCGGGAAGCGACGGATTCTTCGTGGTTTCCCCG 4380
4381 -GAAGGTACGAAATACTGGATGGACTGGGCAGTCACCAGGCGGTACGACAACGTCGGCTTC 4440
4441 -GCTCAACCCAACTTCGGTGGCGTTGCGCGAAATATGGCGATGCGGCTTGTTTCCAGGATA 4500
4501 -GAAGATAAATTCGGCAACGCGCTAACTTACCAATACGACGCTGCCGGAAATCTGACCAGA 4560
4561 -ATTTCCTCTAGTGACGGCCGAGTACTTTCCATAAATTACGAGTCGTGGTCCCAGCGACTG 4620
4621 -GCATCTGGTGCAACCGAATCTGGGGCACGCATCACGTCGGCGGTGCTGCAGACCACAGAC 4680
4681 -ACCCCACCTAGGACGTGGTCATACCGATACGGCGTGCTTGCAGATGGGCGTACGCATCTT 4740
4741 -GTGGCGCTTACACAGCCTGACGGTTCCTCATGGGCATTTGATGTCGGAGGCTTTAACACC 4800

-35
4801 GCGCCTGGATACTCGGCCATCGTGTATCCAACCGATAATGGATGTTCCTTTGTTACGAAT 4860

-10 165
4861 ACAAGCGCGGTTACCGCTACAGGAACCATGCGCCACCCGTCGGGGTTGACAGGCATATTT 4920
rbS M R H P S G L T G I F
Sall
4921 ACAGTAGGGTCGACAATCAGAGGCCGTTCGTACGTTCCTTACGTTTGTGATAATTCATTT 4980
T V G S T I R G R S Y V P Y V C D N S F
4981 AATCGCAGAACGCTCGTCCATCCAAATGTCTACCCTTCTCGCACCCTTGCAAGGAAAGAA 5040
N R R T L V H P N V Y P S R T L A R K E
5041 TTTTATGGTCCGGGGGTCGCATCTCGGGTTTGGTCATACGCTTACTCACCGCCAAACAAT 5100
F Y G P G V A S R V W S Y A Y S P P N N
5101 AGCTGGAGTAAAGACTGTAGTGCTGGCTGCGCCTCCACAGTTTGGACCGATGTTGTGGCT 5160

S W S K D C S A G C A S T V W T D V V A


144
5161 CCTAACGGTAGCGCAACCCGCTATACGTATAGCAATAGATTTGATGTGAGCGAGAGCCAA 5220

P N G S A T R Y T Y S N R F D V S E S Q

5221 CTGCTTAGGTCTGAGCAGTTCTCAGGAACTGTGGGAACTACCTTGGTGAGAACTCTGCAA 5280

L L R S E Q F S G T V G T T L V R T L Q
5281 AACTCGTACGCTCCTGCTACCAATGGTCCTTGGCCTGCTCGTTACGGTTTAAATTTTGCG 5340
N S Y A P A T N G P W P A R Y G L N F A


Figure 4, continued
















5341 AGTAATCTAAATTCGGATCAGACGACACGTGTCTCCCCACTCAGCTCACAAGTGATAGGT 5400
S N L N S D Q T T R V S P L S S Q V I G
5401 CAGGACAGTACGGCATATACATATTCGGTACGCAGCTACGACGCTTTTGCAAGACCGACA 5460
Q D S T A Y T Y S V R S Y D A F A R P T
5461 GCGGTTGCGAGGTATACCCCGTGGCATGGACGTACCGATGTTACTTCGTATTATGACAAT 5520
A V A R Y T P W H G R T D V T S Y Y D N
5521 ACGCAAAAGTGGGTGCTGGGCCAAGAAGCAAGCTCTACAAACAGCGATACCGGTTTAATG 5580
T Q K W V L G Q E A S S T N S D T G L M
5581 GAGCGGCAGATTACTTATAATAGTAATGCCCAGCCAATCTCCATCACGGCTTTCGGAAAG 5640
E R Q I T Y N S N A Q P I S I T A F G K
5641 CTAAGACAAACCATTGGTTATAACAGTGATGGGACAGTTGCTTATGTTGCCGACGCGAAT 5700
L R Q T I G Y N S D G T V A Y V A D A N
5701 AATAACTCGACTACGCTTGGCTCCTGGAAGAGAGGCGTGCCGCAATCTGTTCGTTTTGCG 5760

N N S T T L G S W K R G V P Q S V R F A
140
5761 GACGGGACCTCAATTTCTGCAAATGTCAACGACAATGGTGGATTAGTTCCGTAACGGAT 5820

D G T S I S A N V N D N G W I S S V T D

5821 CAGAATGGTTATCAAACGAACTATAGTTATGATCAGATGGGCCGTCTCTCAACAATTGCG 5880
Q N G Y Q T N Y S Y D Q M G R L S T I A
5881 TATCCTGCGAATGACAGCACGGCTTGGAATATAAAGTCGCAATCTTTTGAGAGAGTCGGC 5940
Y P A N D S T A W N I K S Q S F E R V G
5941 AATGCTGAATACGATATTGCGGCAGATCATTGGCGCCAAACAGTGATCAACGGAAACGCT 6000
N A E Y D I A A D H W R Q T V I N G N A
6001 CGTAAGATTACGTACTTTGATGCAATGTTGCAGCCGCTTCTGACGCGCGAATATGACGTA 6060
R K I T Y F D A M L Q P L L T R E Y D V
6061 GCAAATGAATCCGGGACACAGAGATTCCAGCGCCTTTCCTACGACACCATGGGTCAACTA 6120
A N E S G T Q R F Q R L S Y D T M G Q L
6121 ACATTCTCTTCCTATCCGGGAAATAACAGCTCGCTGAGTACAGGTTATCGAAACAGTTAT 6180
T F S S Y P G N N S S L S T G Y R N S Y
6181 GACGTTCTAGGCCGATCGATTTCTTCTTCTCAAGATTCAGAGCTCGGCTCATTAACCACG 6240
D V L G R S I S S S Q D S E L G S L T T
6241 ACCACGAACTATCTGTCAGGAAATCAAACTAGCGTTACTGACCCGCGTGGGCAGGTCACC 6300
T T N Y L S G N Q T S V T D P R G Q V T


Figure 4, continued








68





6301 GTGACTGGCTATCAGGTATTCGACCAACCCGCGTATGACACGCCTGTCTGGATCCAACAT 6360
V T G Y Q V F D Q P A Y D T P V W I Q H
6361 CCAGAAGGCACCTACACTGATATAGCTAGAAATATTTTCGGCAAGCCAATATCGATAACC 6420
P E G T Y T D I A R N I F G K P I S I T
6421 CGCCGCAACGCCAATGCTTCCCAGGCGTTGACTCGTACTTATGCTTACAACAGCAATCAG 6480
R R N A N A S Q A L T R T Y A Y N S N Q
6481 GAGCTTTGTCGTTCGGTGGAGCCTGAAACGGGTGCTACTTTGATGGGATACGACGCTGCC 6540
E L C R S V E P E T G A T L M G Y D A A
6541 GGAAACATGAAGTGGTCAGCTGCAGGCTTGGCCTCGGATATCGGGTGTGATCAGACTGGC 6600
G N M K W S A A G L A S D I G C D Q T G
6601 GACAGTGCGACAATCGCAACGCGTCGCGTGGATCGAGCTTATGATTCGCGCAACCGTTTA 6660
D S A T I A T R R V D R A Y D S R N R L
6661 AATGCATTGTTCTTCCCTGATGGCAACGGAAACCAGCGCTGGACGTACTGGCCGGATGGT 6720
N A L F F P D G N G N Q R W T Y W P D G
6721 TTAGTCAAGCAAATAACTACGATCAACAGCGGTGTAGCCAGTTACAACAGCTATGCCTAT 6780
L V K Q I T T I N S G V A S Y N S Y A Y
6781 AACAAGCGTCGTCTTCTTGTTGGCGAGAGTCAAGGGCAAGCAGATGGAGAAACCTGGGCG 6840
N K R R L L V G E S Q G Q A D G E T W A
6841 ATCAACTCCATCTACAACGCGAACGGTGACTTGGCGATGCATCGTTATCCAACTGGGATG 6900

I N S I Y N A N G D L A M H R Y P T G M
Sall
6901 ACGGTCGACTACGCGCCAAACGCGCTGGGCCAAGCCACGCAAGCCGGCGGCTACGCGACG 6960

T V D Y A P N A L G Q A T Q A G G Y A T
6961 GGCGCCAGTTATTATCCGAACGGAGCGCTCAGGCAGTTTACATATGGAAATGGCATTGTG 7020
G A S Y Y P N G A L R Q F T Y G N G I V
7021 CACAGCATGGTGCAAAACGCACGCCAGCTGCCTGACACCAGTGAAGATGCCTTCGGTGGC 7080
H S M V Q N A R Q L P D T S E D A F G G
7081 ACTGCAGTATTAAGCGATGGTTATGACTATGACGCAAACGGTAATGTTGCGGCCATAACA 7140
T A V L S D G Y D Y D A N G N V A A I T
7141 GACGGGGCAAGCGGTCGCAATCAGCGCGGCAACCGAACGATGACGTATGACGGGCTAGAT 7200
D G A S G R N Q R G N R T M T Y D G L D
7201 CGTTTGAGATCTACGGTATCTCCAATGTTTGGCACCGCGGAGTACCGCTACGACTCCCTG 7260
R L R S T V S P M F G T A E Y R Y D S L
7261 GATAATCTGACCTACGTAAGAGCACCAGGACGGGAGCATTACTACTGTTACGATCCGTAT 7320
D N L T Y V R A P G R E H Y Y C Y D P Y


Figure 4, continued














7321 TGGCATCTTACCAATATCAAGATCAATGAGTGCTCAGGAAGCACCGTCGTCGGTCTCGCT 7380

W H L T N I K I N E C S G S T V V G L A
7381 TACGATTTGCAGGGCAATCTCACCAACAAGAATGGGCAGGCGTACGTCTTCGACTTCGGG 7440
Y D L Q G N L T N K N G Q A Y V F D F G

7441 AATCGTCTGCGTGCAGCAACAAATAAGGAAAGCTACCGGTACGACGGTAACGGCAGGCGA 7500

N R L R A A T N K E S Y R Y D G N G R R
7501 ACCCAAGCGATACAGTCGGGCGGCAGCGTTGGGTCGATGTATGACCAGGCTGGAGTTTTG 7560

T Q A I Q S G G S V G S M Y D Q A G V L

7561 CGCTTTCAGAAAAATCAACGCCTTTCCAGAATGACGGAGTATGTTTTGCTGGGCGGCAGC 7620
R F Q K N Q R L S R M T E Y V L L G G S
7621 AACGTTGCGGAAGTTGAGTGGACGTTTGGACAAGTACCTGCAATGAAGGACGCTTTGACC 7680

N V A E V E W T F G Q V P A M K D A L T
7681 TGGGCCGCCACTTCTGGAAGCGTGCGCTATGTGGTCGAGGAAAGTATCGACGGACTGACG 7740
W A A T S G S V R Y V V E E S I D G L T

7741 TGGACGTCCGTTTACGAAGGTGATCAAACCACGTGGACCTCATTGTCCCGCCCCTCTGGA 7800
W T S V Y E G D Q T T W T S L S R P S G
7801 GCCTATTCTTATCGTGTACTTGCCTGTACACAGGACGGGACATGCACAGCGCTGCCGGGG 7860

A Y S Y R V L A C T Q D G T C T A L P G
7861 GTTTCGCATGTCAAGCGTTCTGCCGCAGATATCATTCCGCTGCTCTATCAGCTTTTGCTG 7920
V S H V K R S A A D I I P L L Y Q L L L

7921 AACTGACTAGCGCAAGTTTTTCATTTCGAAAGAGCTAACGCCCAATGATAGTGCTTCACA 7980
N *

7981 CGCGGTATCCACAATTTAGCCAAAAAAAGCTATTTATGACGCTGTGCATTTTTTTGCCAC 8040

8041 TTCTATGTGCTAAAGCGTCCGCAGCACCTTACGTTGTCATGGGAAATATCGTAACCCGTC 8100
8101 TCAATACAGGCTGGGGTGCTGAGGGTTTGTATGTCGAAACAACCGGAGCTTCCCAGAGCG 8160
8161 CAACAGGATGTGGTACGGGCAATATATTTTTCATAGAGCAAGGAGGCTCGATGAATAAAG 8220

8221 AGATGACATCCTTATTAATGATGGCTATGCAGAATGACTTTCCGGTTGACCTCTATGTAG 8280
8281 ATGGATGCGTAAACAGTGCGATGAGACTGAAAGCGGTCTATATTAATAAGAGGCGATGAT 8340
8341 ATGAGGTTTTACCGAATTAGCCTGTTGGCGCTAATCATTTTTGCGTCGCCCCGGGCATCA 8400

8401 GCGAACGAATCGGTGCGAGATGTGGAGGTTGTGCAACTCGGTACGTATCAGGCCTCAACA 8460
8461 ATACACTTCGTCTGGTTTGGATCTCTTTCTCAAGAATGTAAGAGCTCCCCAACAATGTAT 8520

8521 TTCGATGACGCGAAACCAGGTGGGAAAGCGTTAATGGCGGTACTGACCGCTGCACTGATA 8580
8581 AATAAAAGAAAAAGTGGATTATCAGGCTGCCGGATGCGACATTGTTGAGGTCTATCTTAAA 8640
8641 TAATCTAATGATAGCTAATGAGCTAGTTTTGTTGCGACAAAATTGATTTGTACGTAGTGG 8700
8701 ATTTGGCAGAAGTTAAGTCGCGGCTTACATTTGCACTGGGGCGGGGAAATGAATAAATGT 8760


Figure 4, continued








70





8761 TCGGATGCTTATGGTATTTATTTGCGTACTTTATTTGTATTCTTTATGTATACTTTGTTT 8820

8821 TGCACAAGTGCTTCGGCTCAAGTGATCCGTTATATACACACGGACGGACTTGGTTCTGTC 8880
8881 TCAGTAGTGACAGATGCGATTCGTAATATTCTTGAGCGGCGTGAGTACGAGCCTTATGGA 8940
8941 ACGACTTTAGGAGAGACCAAAGGTGGGAGTGGCTATATAGGGCATGTGATGGACTCTGTG 9000

9001 ACCGGGCTGACATACATGCAACAAAGATATTACGATGCGGACGTTGGAAGATTTCTATCG 9060

9061 GTAGATCCTATCAAAGCCTATGATAGTCCGGAAAAATCTTTTAACCGATATTGGTACGCT 9120
9121 GATAATAATCCTTACAGATTTTATGATTTGGACGGTAGAGAGTCAAAAGTAGTTCAGGAT 9180

9181 ATGAAGCCAGTAACAGCTGTTGCTCCTCCGCCAACCCCTACATTTCAGCCGGCTATATTA 9240

9241 ACTTTGGGTGCGGTCCTGGTGGAAAGGACTGCGCCGGCTGCCGTAAGCTGGGTTTCGCCA 9300
9301 ATTATGGGCTCAATAGTATTGGCGCCAGTATTATATTTTGCGTCGCCCGATGCTTGCGGT 9360
9361 GGAGCACGCTGTGGTGAGCTGAGCTGGCAAGAGTCTCGTCGGCATCCGCCGGGCTATTGG 9420

9421 CCTGCGGATCGAGGTTCAGAGGAATGGGGACGCCGTAGGGGAGTTGGAGCAATCGAGGGG 9480
9481 CGGCGAATATTTCATGGTATAAAGCAGAAGGATAAAGGCCAGGGTGGCGGCAAGGGACTA 9540
9541 GACAATTATGGGGTAAATCCCGATACCGGTGACATAGTTAATCCCGAGGGTGAGGTCATC 9600

9601 GGCAACTTGGAAAACGGGTAAAATATGAACGGAAATGAATTTTCTATTGAAGTAGCTCTT 9660
9661 TACCTCAAAGGTGATGAGCTTGAACCAAGCGTGGTGACTGAGGTGTTGGGTGTTGAACCA 9720
9721 ACTAAGTCACATTCAAAAGGAAAAATTTGGCTAACTTCTAGCGGTAAAAAGGTTGTTGAA 9780

9781 AAAATAGGTCTTTGGAAATTCTCATTGCGAGGAGCGGCTGAAGAATATTCCGAGATTATA 9840
9841 GCGAGCGTATGCGAAGTCGCCGTCAGGGGTATTGTTCCGTGCTCCACGATCCCTGGAGTA 9900
9901 ACTGAGGCGTTTCTAGATATTTTTTTGGTAAAAAATTCCGATGAATTTGGTGGCGGAACA 9960

9961 TTTGAGCTTTTTTCAGACGCATCCATTATCCGCTCGCTAGCGGAGACAGGTCTTCCGATT 10020
10021 -CACTCCACATTGGCGATCGTTCCGAGGCCGCCGTCCGAGTAGGAGTCGGGTTTTGAGTAC 10080
10081 -TGCGTTGACTGTCCGCTTCCAGAGAAGCCCTTTAGCTATAGATGAATGTTGCGGACGCGC 10140

10141 -TTGCCAAACAAGCCTAACAAGGGCTAGACGATCTTGTCGAAGCCGAACTAGTGAAATGAG 10200
10201 -GGCGGCTTGCTTTTGTCGCCCGTACGTACCCTTGATCAATGTATTATAGAGGTCTCCAGG 10260
10261 -CAAACTAGTCCCGAGGGGTCAACCGATGCGCAAGAGCAAATTCATCTAGAGCACGATCGT 10320

10321 -TACGACATTGAAGCCGGAAGACAGGTCAAGGATGTTTGTCGACAACTGGGTATTTCCGAC 10380
10381 -GCGACGTACTACGTCTGGAAACGTAAGCTCCCCCGATCCGCAGACACCGCACGGAATGGG 10440
10441 -GTCGGAATGGGGTCAGGTTGACTTTTTAGGTGAACTAGACACCGTGTCGCGGTTTCTGGG 10500

10501 -AATTC 10505


Figure 4, continued









both functions, however, no transposon inserts were mapped to this ORF, therefore it

remains unknown whether this ORF is involved. The largest ORF identified (3, 618 bp)

was located directly downstream of ORF 1. Activities of both the deletion derivatives

and mapped transposon insertion mutants, indicate that the putative product of this large

ORF is directly involved in BCN-A+ activity. The ORF was therefore given the

designation bcnA. The putative product of this ORF contains 1012 amino acids with a

theoretical molecular weight of approximately 111-kDa and approximate pI of 6.81 as

calculated using ExPASy proteomics tools (Swiss Institute of Bioinformatics, Geneva,

Switzerland). The ATG start codon of bcnA is proceeded by a consensus putative

ribosome-binding site (RBS) sequence (5' AGGA 3') and potential -10 and -35

promoter-like sequences (Fig. 4). No obvious transcription termination signal was found

downstream of the stop codon for this ORF. Five non-overlapping ORFs (ORFs 2-6) all

located in pXV4.5 (BCN-A-, IMM-A ) were located directly downstream of bcnA. The

potential start codons of these ORFs are at positions 7,965, 8,341, 8,749, 9,625 and

10,196 respectively (Fig. 4). If translated, these ORFs would potentially encode

products containing 124, 100, 290, 145 and 88 amino acids respectively. ORFs 3, 4

and 5 are encoded in the same frame as bcnA. Putative consensus-like rbs sequences

were identified for ORFs 2, 3, 5 and 6. No transposon insertions were mapped to this

region, therefore, the exact ORF involved in immunity to BCN-A+ remains unknown.

Evidence for Transcription and Translation of bcnA

A transcriptional frame-shift at position 4309 in ORF 1 would potentially generate a

larger bcnA ORF. RT-PCR of total RNA from X campestris pv. vesicatoria strain ME90

transformed with pXV12.1 using primer pairs 144/158 and 140/165 was used to









determine whether this frame shift occurred. The orientation and annealing positions of

these primers are shown in Fig. 4. If the frame shift does occur, primer pair 144/158

would have amplified a fragment of approximately 1-kb, however, no product was

generated with this pair while primer pair 140/165 generated a product of approximately

the same size indicating that the frameshift did not occur (Fig. 5). These results also

confirm that ORF bcnA is transcribed.

BCN-A+ activity was detected in unconcentrated cell-free extracts of strains

expressing bcnA, but not in mutants with known insertions into this ORF. Activity was

precipitated with 30% ammonium sulfate. Ultrafiltration of solubilized precipitated

extracts and crude-cell free extracts indicate that the product has an apparent molecular

weight between 50- and 100-kDa (Table 1).

Homology Search Results and bcnA Characteristics

Homology searches using all of the identified reading frames revealed no significant

homology to any previously sequenced bacteriocin determinants. ORF 1 had no

significant homology to any sequences in the databases searched. Both the nucleic acid

and protein sequences of bcnA showed significant homology to wapA (wall associated

protein A of Bacillus subtilis), Rhs elements (rearrangement hot spot elements of E. coli)

and a hypothetical protein from Coxiella burnetii. The C-terminal half of the wapA

protein has two regions containing a total of 31 imperfect copies of an amino acid repeat

sequence (xxxxGxxxx(Y,F)xYDxxGxxx) with a general periodicity of 21 (Foster, 1993).

An almost identical repeat sequence (xxGxxxRYxYDxxGRL{I or T}xxxx) with a similar

periodicity was identified in Rhs core elements with 28 repetitions arranged in four

blocks of 16, 3, 5 and 3 motifs (Hill et al., 1994). The regions of homology between


































1-kb










Figure 5. RT-PCR using primer pairs 144/158 and 140/165. Lane
designations. X (X HinDIII/EcoRI); 1. RT-PCR using primer pair 140/165;
2. Control minus addition of reverse transcriptase; L. 1-kb-ladder; 3 RT-
PCR using primer pairs 144/158; 4. Control minus addition of reverse
transcriptase.









Table 1. Recovery of BCN-A activity by ultrafiltration and 30% ammonium sulfate
fractionation from cell-free extracts ofXanthomonas campestris pv. vesicatoria strain
ME90 harboring pXV519.

Molecular Weight Cut-off a BCN-A+ activity
(kDa) Filtrate Retentate
10 +-

30 +
50 +
100 + +


a BCN-A+ activity as detected by in vitro inhibition assays.









bcnA and these two proteins lie within the C-terminal 500 residues of bcnA and the

repeat regions of these two proteins. Examination of the bcnA sequence reveals the

presence of a similar motif repeated seven times with an approximate periodicity of 24.

Two similar motifs were found outside of this repeat region at positions 431 and 549.

The sequences and positions of these motifs are shown in Fig. 6. Similar motifs have

also been found in other ligand-binding proteins and are involved in carbohydrate binding

(Vlazny and Hill, 1995).

Hydrophobicity determination and search for signal sequences revealed that the

putative bcnA product should be hydrophillic, with no potential transmembrane regions

as calculated by the Kyte and Doolittle method (Kyte and Doolittle, 1982) and should be

devoid of any known signal sequences (Nielsen et al., 1997).

ORFs 2, 3 and 5 had no significant homology to any sequences in the databases

searched. ORF 4 had significant homology with a hypothetical protein from

Streptomyces coelicolor, wapA, an insecticidal toxin complex (TccC) protein from

Photorhabdus luminescens, and RhsA. The specific regions of homology were with the

N-terminal residues 130 of the putative ORF 4 protein and the same regions containing

the repeated motif, however, a similar motif was not identified in this region. ORF 6

was significantly homologous to a transposase (ORFA) identified in Salmonella

enteritidis and Yersinia enterocolitica.

Expression of bcnA

Translational bcnA::gusA fusions were generated. The position and orientation of

fusion 78, which was used for these studies, can be seen in Fig. 3. The orientation of

lacZ gene is the same as that of bcnA. This construct was mobilized into X c.




















431 SLSTGYRNSYDVLGRSISSSQ
549 PETGATLMGYDAAGNMKW

741TAVLSDGYDYDANGNVAAIT
767 NQRGNRTMTYDGLDRLRSTV
788 PMFGTAEYRYDSLDNLTYVR
808 APGREHYYCYDPYWHLTNIK
832 SGSTVVGLAYDLQGNLTNKN
865 AATNKESYRYDGNGRRTQAI
885 QSGGSVGSMYDQAGVLRFQK

xxxxGxxxxYxYDxxGxxxx WAPA consensus motif
F
XxxxGxxxRYxYDxxGRLIx RHS consensus motif
T
IDGxxYYFDxNxG Ligand-Binding Motif




Figure 6. Alignment of bcnA repeating motif region. Number represents position of Y in
YD conserved sequence. X- Conserved in WAPA, RHSA, and ligand-binding
consensus sequences; X- Only found in RHS consensus sequences; X- Only found in
ligand-binding consensus sequences.









vesicatoria strain 93-IR to generate 93-1(78) and expression was tested after growth in a

susceptible tomato genotype. Gus expression was only detected after 48 hr of growth in

plant leaves at a level of 1.2 x 10-6 U/CFU and continued to increase to 3.9 x 10-6 U/CFU

after 96 hr of growth inplanta (Fig. 7). Gus expression was also detected after 24 hr of

growth in NB, KMB and XVM2 media (Table 2). Expression in XVM2 was

significantly higher than in the other two media types at a level of 5.4 x 10-6 U/CFU

which was approximately ten-fold higher than after growth in NB and 100-fold higher

than KMB. Amendments of FeSO4 to KMB had no significant effect on gus expression

levels in this media (Table 3).

Distribution and Genomic Localization of bcnA

A 2-kb internal Sall fragment of bcnA was used to probe chromosomal and plasmid

DNA of fourteen X campestris pv. vesicatoria strains, six X campestris pv. glycines

strains and six other X. campestris pathovars. Several large plasmids were isolated from

the strains tested (data not shown), however, there was no homology of these with the

bcnA-specific probe. Among X campestris pv. vesicatoria strains, only T3 strain

genomic DNA hybridized to the probe (Fig.8). The Probe also hybridized with strain

1717b ofX. campestris pv. glycines. The immunity subclone (pXV4.5) conferred

immunity to both X campestris pv. glycines and X campestris pv. vesicatoria strains

sensitive to the bacteriocin which is produced by this strain (Fig. 9).

Physical Characterization of BCN-B+ Clones

Two clones, pXV442 and pXV699 containing approximately 18-kb of overlapping

DNA were sufficient to confer BCN-B+ activity when mobilized into TI strain ME90,

however, no immunity function was found associated with either of these clones. Four




















6



(O D---
-0 C
0










0 20 40 60 80 100 120

T nm (hours)

Figure 7. Time course of P-glucuronidase activity of Tn3-gusA::bcnA
translational fusions during growth of Xanthomonas campestris pv.
vesicatoria strain 93-1(78) in tomato cultigen 7060.
(9O












vesicatoria strain 93-1(78) in tomato cultigen 7060.









Table 2. The effect of media type on p-Glucuronidase activity of Tn3-gusA: :bcnA
promoter fusions in three different media types.

Media type a p-Glucuronidase activity (10-6U/CFU)

Nutrient Broth b 0.54 a

Kings Medium B 0.006 b

XVM2 5.03 c


aU defined as nanomoles of 4-methylumbelliferone released per minute.
b1-Glucuronidase activity determined after 24 hours of growth in selected media type.
b Means followed by different letters significantly different at P = 0.05.









Table 3. p-Glucuronidase activity of Tn3-gusA::bcnA promoter fusions in nutrient broth
and Kings Medium B with amendments of FeSO4.

Media type a p-Glucuronidase activity (10-6U/CFU)

Nutrient Broth b 0.60 a

Kings Medium B 0.010 b

Kings Medium B (0.01 mM FeSO4) 0.018 b

Kings Medium B (0.1 mM FeSO4) 0.016 b


a U defined as nanomoles of 4-methylumbelliferone released per minute.
b1-Glucuronidase activity determined after 24 hours of growth in selected media type.
Means followed by different letters significantly different at P = 0.05.























1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18


2.0-kb a




2.0-kb- a M -


Figure 8. Southern hybridization analysis of plasmid (A) and genomic (B)
DNA from Xanthomonas campestris pv. vesicatoria and other X campestris
pathovars using a 2-kb bcnA specific probe. Lane desigantions: 1. pXV12.1
(BCN-A ); 2. pXV442 (BCN-B +); 3. pXV120 (BCN-C +); 4-9. X c. pv.
glycines strains XP144, B83, J3-27-1A, 1717a, 1717b; 10-12. X c. pv.
vesicatoria T3 strains NTG2-7, T3-L2, T3-L3; 13. X c. pv. alfalfae; 14. X c.
pv. diffienbachiae, 15. X c. pv. phaseoli, 16. X gardaneri, 17. X c. pv.
begoniae, 18. X. c. pv. holcicola.











































Figure 9. Assays of bacteriocin activity of Xanthomonas campestris pv. glycines
strains 1717b, J3-27-1A, B83 and XP144, againstX, c. pv. glycines indicator strain
XP144 with or without pXV4.5 (IMM-A) The IMM-A phenotype is expressed in
this strain against the bacteriocin produced by strain 1717b of X c. pv. glycines.









BCN-B- mutants were obtained after Tn3 insertion mutagenesis of pXV442. All of these

inserts were contained in 8.9-kb KpnI fragment which was subcloned into pUFR051

generating pXV8.9. This derivative retained BCN-B activity.



Physical Characterization of BCN-C+ Clone

Only pXV120 exhibited BCN-C+ activity when mobilized into TI strain ME90,

however, no immunity function was found associated with this clone. Two BCN-C-

mutants were obtained after Tn3 insertion mutagenesis of pXV120. All of these inserts

were contained in 5.1-kb HinDIII/EcoRl fragment which was subcloned into pUFR051

generating pXV5.1. This derivative retained BCN-C activity. It was also cloned into

pBluescript and sequenced.

Sequence Analysis of pXV5.1 (BCN-C )

The entire 5.1-kb subclone was sequenced and the complete sequence is shown in

Fig. 10. The sequence has a 67% GC base composition. The DNA sequence when

translated into all possible open reading frames (ORFs), revealed 9 possible ORFs

greater then 250 bp (ORFs 1-9). A Tn3-BCN-C- insertion was localized to position 4643

in the sequence by sequencing from the inserted transposable element (Fig. 10). No open

reading frames were identified in this region, however, it was positioned 98 bp

downstream of ORF 9 at positions 3915-4545. ORF 9 showed significant homology to a

extracellular metalloproteases secreted by Aeromonas hydrophila, Armillaria mellea,

Pleutorus ostreatus, Grifola frondosa, Aspergillus fumigatus and Penicillum citrinum.








84



1 AAGCTTGGCTGCAGGTCGACGGATCTACCACGTCATCGTCCAGGCCACCCCGCGCTTGCG 60
61 CGAGCGCCTGTGCAAGGCAGGCACCTGCGATGCCGCCGGCGTGCCGGTGGAATTGCCCAG 120
121 GCCGGCCACCTTCGAGGCTGCGGTCAAGCGCTGATGCGCGGGCGGCGCTGCGGCGATGGC 180
181 GGGCGAGCTGGCGCGTGATGCGGTTGCCGCATCGCCGCTCGACCTGCCCGCCTGGCGTCG 240
241 CCACCGCCATGTGCAATCGCGCCTGCGAACAGCCGCAGTCCCCAGCCGTTGGTGTGGTGC 300
301 TCGGATGGGCCGTAGCGGAACGATGCGGTGACCGCGCACGGCAGCACGTTGCAGCCCAGA 360
361 CTGCAGCTGCGCACGGTTGTTCGCGCTACGGGGCGCCGGGAACGACGGAATCGTCGATGG 420
421 GTCCCAGGCAGAGAACACCCCGGGCAGGCGGCCGACACGCAATCACGCAGCAGGGATGGC 480
481 ATCTGCTCGCCGAGCGCGCAGGAGCCGCTGCGAGATTGCGCCTATGCTGACCGGCACCCC 540
541 GCTCCTGTTGCCAGCCACCCCGCCCGGCCCGGTGCTGCTGGCCTATAGCGGCGGGATGGA 600
601 TTCCAGCGTGCTGCTGCATCTGCTGGCGGCGACGCCACGCTATCGGCACGCAGGATTTCG 660
661 CGCGCTGCATGTGCATCATGGGCTGCATGCCGATGCCGATGCCTGGGCCGCGCATTGCCA 720
721 ACGCACCTGCGACACGCTGCAGGTGCCGTTGCAGATCGTGCGGGTGCAGGTGGCACGCGA 780
781 CAGCGGTCTTGGGTTGGAAGCGGCAGCACGCCACGCGCGCCATGCGGCGTTCGCGCAGGC 840
841 GCTCACTGCTGGCGAATGGCTGGCGCTGGCGCACCATCGCGACGACCAGGCGGAAACCTT 900
901 CCTGCTGCGTGCGCTGCGCGCTTCCGGCCCCGAAGGCCTGGCGGCGATGCGGCCGCAGCG 960
961 CCCATTCGCCAGCGGCACGCTGTGGCGCCCGATGCTGGCGCACGCCCGCGCCGATCTGCT 1020
1021 -CGCCTACGCGCATGCGCAAGGGTTGCGCTGGATCGAAGATCCCAGCAATGCCGACCCACG 1080
1081 -CCATGACCGCAACTTCCTGCGCAGCCAGGTCGTGCCGTTGCTGCAGCAACGCTGGCCGCA 1140
1141 -AGCACCCGACGCGTTGGCGCGCAGTGCGCAACTGAGCGCCGATGCCAGCGCATTACTGCT 1200
1201 -GCAGCAGGACATGGCGTTGTTGCCGAGCGTGCGGACCGCAAGCGGCGCGCTGGATCTGCA 1260
1261 -GGCATTGCGCGCGCAACCGGTGGAACGAAGGGCACGGCTGTTGCGCGCCTGGGTGAGTGC 1320
1321 -GGCGCACGCGCCGCCGCTGCCTGCGCAAGGCGTGGCCGCGCTGGAACGGGAAATCGACAA 1380
1381 -TCACGCTGCAGATCGGCAGGCCTGCTTTGCCTGGCAACAGGTAGAAGTACGCCGCTGGCG 1440
1441 -CCAGTACCTGTTTCTGCATCGCCACGCTGCGGCCTGGCCGTCGGACTGGCACGCGCAGTG 1500
1501 -GGACGGCGCAGCGCCCCTGCACCTGCCCGACGGTGCGCAGTTGCGGTTGCTCGGCGCACC 1560
1561 -CGGGCTGCGCTTTGCGCAACCGCTGCTGGTGCGCGCACGCCGGGGCGGCGAGCGCATCGT 1620
1621 -GCTGCCGCAGCGCACGCACTCGCACCAACTCAAACACCTGCTGCAGGCGCTGGATCTACC 1680
1681 -GCCGTGGGAGCGCGGGCGCCTGCCGATCCTGTGGGACGGCACGCAGGTGCTCGCCGCCGG 1740
1741 -CGACTGCATCATTTCCGCCACCTTGGACGCATGGTTGCAGACCAATGCCGCCGCGCTGCA 1800
1801 -ATGGCGCGCGGCTGCCGACGCGAATTGACCCGCCTGCGCGCGCGCCGCACACTTTAGCGA 1860
1861 -TGGCCAAGAAGTCTTTGAACGAAACGTCCCCGGTTGCCCGCTTCGAACAGTCGCTGGAAG 1920
1921 -AACTCGAGCAACTGGTGCAGAAGATGGAAGTCGGCGATCTGAGCCTGGAGCAATCGCTCA 1980
1981 -CGGCCTACGAGCGCGGCATCGGGTTGTACCGTGATTGCCAGCAGGCGCTGGAACAGGCCG 2040




Figure 10. Nucleotide sequence of pXV5.1 containing all of the genes required for BCN-
C+ expression. Insertion point of Tn3 which resulted in a loss of BCN-C+ expression is
indicated by an arrow and the start and stop codons of ORF 9 are underlined.








85



2041 -AACTGCGCGTGCGCTTGTTGACCGACCCCGCCCGCCCCGAGCTTGCCGAAGCCTTCGAAC 2100
2101 -CGCCGTCGCTGGACGGCTGAGGTGAGTTCGGCGCTGTTCGACCACTGGATCGCCCGCACC 2160
2161 -GAACGCAGCCTGGAGGCCGGCCTGCCGCGCGCCACGCATGCACCGCAGCGCCTGCATGCA 2220
2221 -GCGATGCGCCATGCCGTGCTGGGCGGGGGCAAGCGCATGCGTCCGCTGCTGGTCTACGCA 2280

2281 -AGCGGCGCTTTGTTCGGCGCTGCTGAGGATCAGCTCGACACGCCCGCGGTGGCGGTAGAG 2340
2341 -CTGATTCACGCCTATTCGCTGGTGCACGACGACTTGCCGGCGATGGACGACGACGCGCTG 2400
2401 -CGGCGCGGCCAGCCTACCGTGCATATCGCCTTCGATGAGGCCACCGCGATTCTGGCCGGC 2460

2461 -GACGCCTTGCAGACGCGCGCGTTCGAATTACTCGCCACTGCCTCTGCAAGCGCCGAACTG 2520

2521 -CGCGTCAGCTGGATGCAGAGCCTGGCCACTGCTGCGGGCGCGGCCGGCATGTGCGGCGGC 2580
2581 -CAGGCGCTGGATATCGACGCGACCGGGCAACTGCAGTCGCTGCAGGATCTGCAACGTATG 2640

2641 -CATGCGCTCAAGACCGGCGCATTGATCCGTGCGGCGGTCCGGATGGGCGCACTCACCGGA 2700

2701 -TGCGCCGCGGCGGCCGACCAGCAGCGTCTGGACGAATTTGCCGACGCCTTGGGCCTGGCC 2760
2761 -TTCCAGGTACGCGACGACATCCTGGATGTGGAATCGAGCTCGGCGCAGCTGGGCAAGACC 2820
2821 -GCCGGCAAGGACGCGGCGCAAGCCAAATCCACCTACCCCGCCCTGCTGGGGATGGATGGC 2880

2881 -GCCAAGGCGAAGCTGGCCGAGCTGGCCGCGCGCATGCACGACGTCCTGCAGCCCTATGGG 2940
2941 -CAACCAGGCGAGACGTTGGCCACCTTGGGCCGGTTTGCGGTGAACCGCGCGCACTAGGCG 3000
3001 -CAGTTGCCAACGCCGCTACACCGCAGCGAGCGGGAACGGCCGATGCCCGAATCGATCTGA 3060

3061 -AGTTTCCCGCGTGCCGCAATGGCGCGCTGGTGGCTCGCGCAGCTACATCTGGACGATTTG 3120
3121 -CGACATAAAGCGGACGCGCCGGACGCGATTTGCGCACTCCGTGCGCCAGCGCACGGTCAA 3180
3181 -CGTCTGCACCGGAGGCGTGCAAGCAGCGAACTCGGTCACACATTGCAACTTTTCCTGGCT 3240

3241 -GCATTTGTGTCGTGCGTTGACACATTCGGAGTTGCAGCACATGACCGATCGCACATTTTA 3300
3301 -AAGCTGAATTGTCGACATCCGTGCTATTTCGGCGGTGTTCATAAGTTCGCTACGAACAAA 3360
3361 -AATATTCGCAATTGAATAACTTGTCGCGTCCGCTCGTCCTGCCTAGGTTTGGCTTGCCCG 3420

3421 -CATCCGCGGACATCCCCCGATCACACGGAGAGTCACTGGTGAAGAACGTTTTTCTCGCAT 3480
3481 -CGTTTGCAGCAGGCACGCTGGCCGTTGTTGGCGTCCTCGGATCGGCCCAGGCGCAGTCCG 3540
3541 -TGCGCGGTCCGCTTCCGTTGACCATCGAACTGTCGCCAGGGCCGACCAGGCTGGGCGTCA 3600

3601 -TCAGGGCAAGATCGCCGTCACCGTCACCAATAACGGCAGCCAGACCGCACGCGTGCCGAC 3660
3661 -GTACCAGCTGCCGCTGAAGTCGCTGGATAACGGCATTCTGGAAGTGAGCCGTGACGGCAA 3720
3721 -GCCGGTGGACTACACCGGGCGCCTGGTCAAGCGCGGCCTGCCCAAGGCCGCCGATTTCAC 3780

3781 -CATCCTGCAGCCAGGCCAGAGCGTGAAGGGCGAGGTCGATCTGGCCGGCGCCTACGATCT 3840

3841 -GTCCACCAGCGGCAACTACACCATCCAGGTGCGCTCGGCGCTGCAGTACGCCTCCTTTTC 3900

3901 -CGACGGCAGCCTGATGAAAGCCGCCAGCGGCGAGCCGGCCGTGGCCACCAGCACCCCGCT 3960


Figure 10, continued















3961 -CACCGTGTGGCTGGACGGCGTCAACCGTGGCGTACAGCGCCAGCTCGCAGTCGGCCCGAC 4020

4021 -CGCCGTGGTCAACGGCATCAATTACCTCAACTGCAGCACCACCCGCACCAGCCAGATCGC 4080

4081 -CAGTGCCGTCACTGCCGCGCGCAACTACTCGCAGAACGCGCGCAACTACCTCAATGCCGG 4140

4141 -CAGCACCGGCGCGCGCTACACCACCTGGTTTGGTGCCTATAACGCCTCGCGCTACAGCCG 4200

4201 -GGTCAGCTCGAACTTCGTCAACATCGACAATGCGCTGGACCAGAACAACGGCCAGATTAC 4260

4261 -CATCAATTGCGGTTGCACCGACAGTGCCTTCGCGTACGTCTATGCCAACGCGCCTTACGA 4320

4321 -AATCTATGTCTGCAATGCGTTCTGGAGTGCATCGACCACCGGCACCGACTCCAAGGCCGG 4380

4381 -CACGCTGGTACACGAGATGAGTCACTTCACGGTCGTCGCCGGCACCCAGGACCGCGTCTA 4440

4441 -TGGCCAGTCCGGCGCGCGCAGCCTGGCGATCAGCAACCCGGCGCAGGCCATCACCAACGC 4500

4501 -CGATAGCCATGAGTACTTCGCCGAAAACACCCCGGCGCAGAACTGATCGCTCCATCGCCT 4560

4561 -GATACGACAAAGGCCCGGCATCGCTGCCGGGCCTTTGTTGTTTCCACCGCGCCGATTACT 4620

4621 -TGATCAGGCGCAGCGCGAACGGGTAACGGTAATGCTCGCCGTTGTTGGCCTTGACTGCGG 4680

4681 -CCAGGATGCACAGCACCAGGTTGGCAATCCACACCAGCGTCGGCAAGAAGAACAGGATGC 4740

4741 -CGAACGAGACAATGGTCAGAATGATCGCTGCCACGGTGGCAATGAGCACGGTGATCTGGA 4800

4801 -AATTCAGCGCTTCCTTAGCTTGATCGGTCGCGAACGGCTTGGACGGGCTGGCGTCCTTGC 4860

4861 -TGATCAACCAGATTACCAACGCACCGACGAACGAGGTCACGATGCCGAGCAGATGCGCGG 4920

4921 -CCAGCGCGAGCGTACGCTCCTCCGACGGACTGGTGCCGACCGGTGGCGGTGGCGGTGCGG 4980
4981 -CGTGCGATTCGAATTC 4996


Figure 10, continued









Relative Contribution of Bacteriocin-like Activities to In Planta Competition

The smallest active subclones corresponding to all three of the bacteriocin-like

activities were mobilized into X campestris pv. vesicatoria strain ME90 and tested for

their ability to inhibit a sensitive Tl strain in susceptible tomato cultigen 7060 when

applied 12 hr before. Both the wt T3 strain and recombinant BCN-A+ strain limited the

growth of indicator strain 91-106 by approximately 10-fold and 100-fold respectively,

over a 72 hour period of growth after infiltration into tomato leaves (Fig. 11). In contrast,

the populations of indicator strains increased in the presence of a wt TI (BCN-), and

recombinant BCN-B and BCN-C strains.

T3 strains with single mutations at each of the bacteriocin-like wt loci were tested for

their ability to inhibit a sensitive TI strain when applied 12 hr before. Growth of the

indicator strain was restricted in the presence of the wt T3 strain, and marker exchange

mutants inactivated at the BCN-B and BCN-C wt loci (Fig. 12). However, a wt T3

strain mutated at the BCN-A locus no longer inhibited the growth of the sensitive strain.






















8



6 -- TT1
_j
-e- BCN-A
D --BCN-B
U-

d- -A- T3
_j
2 -



0
0 20 40 60 80
TIME (HRS)

Figure 11. Time course of growth ofXanthomonas campestris pv. vesicatoria
indicator strain 91-106 in tomato cultigen 7060. Bacterial populations in
leaves were sampled every 12 hours for 48 hours following infiltration with:
wild-type TI strain 75-3 (BCN -) T3 strain 91-118 (BCN ), pXV519 (BCN-
A), pXV442 (BCN-B ), and pXV120 (BCN-C ). All plasmids were
transferred to X campestris pv. vesicatoria strain ME90 prior to testing.







89

















7

6 -UT1
-0 ME-A
5 -* ME-B
SME-C
0 -A- T3
24
U-


-j
2-

1


0
0 10 20 30 40 50 60 70 80
TIME(HOURS)


Figure 12. Time course of growth ofXanthomonas campestris pv.vesicatoria
indicator strain 91-106 in tomato cultigen 7060. Bacterial populations in
leaves were sampled every 12 hours for 48 hours following infiltration with
strains 75-3 (T1)(BCN -), 91-118 (T3) (BCN +), ME-A (BCN-B C ), ME-B
(BCN-A+ C), and ME-C (BCN-A+ B).









Discussion

Three distinct genomic regions of X. campestris pv. vesicatoria T3 strain 91-118 have

been identified for their ability to encode bacteriocin-like activities. Sequencing of the

two major regions involved, confirmed that they have no DNA or amino acid similarity.

Genes associated with immunity to these bacteriocin-like compounds were only found for

one of the regions.

Analysis of the overlap region of clones with BCN-A activity, deletion analysis, and

mapping of transposon insertion points, has shown that the gene(s) responsible for the

BCN-A+ phenotype are located within an 8.0-kb segment of DNA. The immunity

function was localized to a 4.5-kb fragment, which was expressed independently of

BCN-A+ activity. Sequence analysis of this region revealed the presence of seven open

reading frames with potential involvement in the BCN-A+ phenotype. ORF 1 is located

within this region, however, its involvement in BCN-A+ expression remains unknown.

Sequence analysis to map the exact insertion points of Tn3-gus transposable-elements,

revealed that multiple insertions into bcnA ORF resulted in complete loss of that

phenotype. Thus, the product of this bcnA is directly involved in the synthesis or export

of this compound. This ORF potentially encodes a protein of up to 111-kDa. Evidence

that a high-molecular weight component is involved in BCN-A-specific antibiosis was

obtained when ultra filtration analysis of crude and precipitated activity predicted an

apparent molecular weight of between 50- and 100-kDa. However, this is a rough

estimate of molecular weight and may be overestimated by aggregation of molecules.

Helveticin J, a bacteriocin produced by Lactobacillus helveticus, was shown to exist in

aggregates in excess of 300-kDa in culture supernatants, when the actual molecular









weight of the compound was determined to be approximately 37-kDa (Jeorger and

Klaenhammer, 1986). Previous studies have indicated that the inhibitory compound

secreted by X campestris pv. vesicatoria T3 strains is heat-labile and trypsin-resistant

(Tudor, 1995), two properties which are suggestive of a high molecular weight

component. Many of the bacteriocins that have been identified in Gram-negative bacteria

are high molecular weight compounds. Colicins of E. coli and pyocins of Pseudomonas

aeruginosa range from 29- to 81-kDa (Braun et al., 1995).

A good consensus-like ribosome binding sequence was found directly upstream of the

bcnA start codon. Direct evidence for the transcription of this ORF was obtained after

generation of a product following reverse transcription. In addition, the presence of an

active promoter 5' to the start of this ORF is deduced from the 13-glucuronidase activity

detected in Tn3-gus::bcnA translational fusions.

Homology searches using the deduced amino acid sequence of this ORF revealed

significant homology to only two known proteins, WAPA and RHS. The function of

WAPA is unknown, however, it is constitutively produced in abundance and is non-

essential for normal cellular functions (Foster, 1993). A smaller ORF lies directly

downstream of wapA, however, the function of this ORF is also unknown. A mutant

insertionally-activated in the the wapA gene of Bacillus subtilius, had no distinguishable

phenotype apart from lack of the wapA gene product. The wapA gene encodes a 258-kDa

precursor, which is processed into three overlapping wall-bound and released forms of

220-, 109- and 58-kDa. Wall binding was associated with the N-terminus of the

molecule, deletion constructs minus this region were unable to bind cell walls. The two

larger wapA derivatives are wall-bound, while, the smallest is actively secreted into the









media. This 58-kDa derivative consists of the C-terminus of the molecule and contains

the repeated peptide motif. Homology between wapA and bcnA is specific to this region.

Rhs elements are composite genetic elements which are repeated in the genome, and

are widely distributed among natural E. coli strains. They were first identified because of

the high degree of rearrangement that occurred in these regions. They all share a GC-rich

core region of approximately 3.7-kb followed by a variable AT-rich core-extenstion.

This uncharacteristically high GC rich region is believed to have recently been introduced

into the E. coli genome from another organism with high GC content. The largest ORF

of these elements spans the core and extension regions and is approximately 4.1-kb in

length. This ORF is directly followed by three smaller ORFs (Hill et al., 1994). The

core-protein like the putative bcnA gene product, is high molecular weight, hydrophilic,

devoid of a signal sequence, and contains 28 copies of an almost identical motif to that

found in the C-terminal region of BCNA and WAPA. A number of secreted ligand-

binding proteins have been identified as having similar motifs, which are believed to be

involved in carbohydrate binding. These include a number of toxins, where the motifs

are involved in target recognition (Foster, 1993). The C-terminal repeating units of

ToxA, a toxin secreted by Clostridium difficile, are involved in interactions with the

oligosaccharide components of receptor molecules on target cells (von Eichel-Streiber et

al., 1992). Several outer membrane proteins have been implicated in bacteriocin binding

to target cells (Lakey et al., 1994); however, only recently has a role for core

lipopolysaccharide in bacteriocin binding been demonstrated. Binding of bacteriocin 28,

a bacteriocin produced by Serratia marcescens, to sensitive cells was blocked in rfaQ

mutants which are impaired in core LPS biosynthesis (Enfedaque et al., 1996). This may









provide a clue as to the role of the putative carbohydrate-binding motifs identified in the

bcnA gene product. The key elements of these motifs (i.e., a conserved core of aromatic

residues followed generally by an asparagine) are present in BCNA. This motif has been

identified in a discrete C-terminal region of the BCNA, which may be suggestive of a

domain organization for this protein. Several high molecular weight toxin molecules,

including bacteriocins, are organized into discrete domains, which each have a different

function. The domains usually have distinct binding and catalytic functions.

Although initial reports on Rhs elements of E. coli attributed no known function to

these elements, a parallel between these elements and the genetic determinants for

bacteriocin production by E. coli was noted (Hill et al., 1994). Colicins are large

polypeptides, notably devoid of signal sequences, whose release and immunity are

mediated by genes directly downstream of the structural gene. Evidence that they may

indeed encode a bacteriocin-like function was obtained when deletion derivatives of the

RhsA element lacking the ORFs downstream of the core ORF, were found to impart a

toxic effect on E. coli strains used for routine culturing (Vlazny et al., 1995). A short

(72-base pair) ORF, located within the C-terminus of the core ORF, was found to be

sufficient to confer toxicity. Similarly, the toxic effects of most Gram-negative

bacteriocins are localized in the C-termini of these molecules. Toxicity was only

observed after cells had reached the stationary phase of growth. Interestingly, the

translation product of dsORF-al, which lies directly downstream of the core ORF,

suppresses toxicity, a structure that mirrors the mechanism of immunity to colicins and

immunity to BCN-A+ activity. The exact ORF involved in immunity to BCN-A+ was not

identified; however, ORF 4 potentially encodes a protein with homology to WapA,









RhsA, and an insecticidal toxin complex, suggesting that the activity of this and the

bcnA ORF may be linked. Genes which confer immunity to bacteriocins are almost

always only protective to the bacteriocin with which they are associated.

Carnobacterium piscicola strain LV17 produces two bacteriocins, carnobacteriocins

BM1 and B2. The gene for immunity to B2, was located downstream of the B2

structural gene, and conferred immunity only to this bacteriocin (Herbin et al., 1997).

Rhs elements, like wapA, are non-essential to the cells that produce them (Feulner et

al., 1990); however, they have remained highly conserved over a considerable period of

evolution. Since they are non-essential for regular cellular functions and are not

universally distributed among E. coli strains, it has been proposed that they may play a

role in the natural ecology of the cell. Perhaps the finding that they are toxic provides

evidence for this. BCN-A determinants are also not universally distributed among

Xanthomonas campestris pathovars and BCN-A- mutants are viable, indicating a non-

essential role for this compound.

Another ORF, designated H-rpt, which lies down stream of the core and dsORF-al

ORFs, is believed to be an insertion sequence with homologies to genes involved in the

generation of variation in Xanthobacter autotrophicus, Vibrio cholerae, and Salmonella

entericia. ORF 6, one of the ORFs downstream of bcnA, has high similiarity to

transposase ORFs from Salmonella enteritidis and Yersinia enterocolitica. The genetic

determinants for Pyocin AP41, a bacteriocin produced by Pseudomonas aeruginosa,

were isolated from the chromosome and found to be part of a transposon-like structure.

The similarity in structure and function of the genes involved in RhsA and BCN-A+

expression, along with the observation that rhs loci are believed to be a recent addition to




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs