1

CHARACTERIZATION AND USE OF BACTERIOPHAGES ASSOCIATED WITH CITRUS BACTERIAL PATHOGENS FOR DISEASE CONTROL

By

BOTOND BALOGH

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

2006

2

Copyright 2006

by

Botond Balogh

3

Unokatestvérem, Dobó Gábor emlékére

4

ACKNOWLEDGMENTS

I would like to thank my committee members, Drs. Jeffrey B. Jones, Robert E. Stall, Timur

M. Momol, Donna H. Duckworth and Paul A. Gulig, for their support, constructive criticism and

guidance through the entire research project and preparation of this manuscript. I especially

appreciate Dr. Jones’ goodwill and loyalty that he showed in the time of need.

I would like to thank all those who helped me in this project: Aaron Hert, Jason Hong,

Frank Figueiredo, Mizuri Marutani, Fanny Iriarte, Ellen Dickstein, Jerry Minsavage, Nelly

Canteros, Alberto Gochez, Debra Jones, Xiaoan Sun, Amber Totten, Tanya Stevens, Mark

Gooch, Jake, Hans-W Ackermann, Donna Williams, Jim Dilley, Henry Yonce, Lee E. Jackson

and the employees of the OmnyLitics Inc., Justyna Kowara, Scott Taylor and the stuff of the

Citra Plant Science Unit, Terry Davoli, Ulla Benny, Kris Beckham, Gary Marlow, Patricia

Rayside, Mark Ross, Eldon Philman, Vanessa Ivanovski, Chandrika Ramadugu and the ones I

forgot to mention.

I would like to thank all those who, while not contributing directly to the project, helped

me during the time I was working on it: Abby Guerra, Gail Harris, Jim Barrel and Aleksa

Obradovic.

5

TABLE OF CONTENTS page

ACKNOWLEDGMENTS ...............................................................................................................4

LIST OF TABLES...........................................................................................................................8

LIST OF FIGURES .........................................................................................................................9

ABSTRACT...................................................................................................................................11

1 CITRUS CANKER.................................................................................................................13

The Citrus Industry .................................................................................................................13 Symptoms, Etiology and Epidemiology .................................................................................14 Disease Impact........................................................................................................................16 Disease Management ..............................................................................................................17 Genetic Variation of the Pathogen..........................................................................................17 Citrus Canker in Florida .........................................................................................................18 Citrus Canker Eradication Program........................................................................................20 Bacteriophages Associated with Citrus Canker......................................................................21 Citrus Bacterial Spot...............................................................................................................22 Project Goal and Objectives ...................................................................................................23

2 THE USE OF BACTERIOPHAGES FOR CONTROLLING PLANT DISEASES..............31

Early History...........................................................................................................................31 Other Uses of Phages in Plant Pathology ...............................................................................32 Return of Phage-Based Disease Control.................................................................................32 Considerations About Phage Therapy ....................................................................................33 Factors Influencing Efficacy of Phages as Biological Control Agents ..................................34 Current Research ....................................................................................................................37 Outlook for the Future ............................................................................................................38

3 CHARACTERIZATION OF BACTERIOPHAGES ASSOCIATED WITH CITRUS CANKER IN FLORIDA AND ARGENTINA ......................................................................40

Introduction.............................................................................................................................40 Materials and Methods ...........................................................................................................41

Bacterial Strains and Bacteriophages ..............................................................................41 Standard Bacteriophage Techniques ...............................................................................42 Phage Isolation from Diseased Plant Tissue ...................................................................44 Phage Typing of 81 Xanthomonas Strains ......................................................................45 Electron Microscopy .......................................................................................................45 Molecular Techniques .....................................................................................................45

6

Results.....................................................................................................................................47 Bacteriophage Isolations .................................................................................................47 Classification of Isolated Phages Based on Chloroform Sensitivity, Plaque

Morphology, Host Range and Virion Morphology......................................................47 Comparison of Florida Group A Phages and CP2 Based on Genome Size and RFLP

Profile...........................................................................................................................48 Phage Typing of 81 Xanthomonas Strains ......................................................................48

Discussion...............................................................................................................................49

4 CONTROL OF CITRUS CANKER AND CITRUS BACTERIAL SPOT WITH BACTERIOPHAGES.............................................................................................................63

Introduction.............................................................................................................................63 Materials and Methods ...........................................................................................................65

Bacterial Strains and Bacteriophages ..............................................................................65 Plant Material and Cultural Conditions in the Greenhouse.............................................65 Standard Bacteriophage Techniques ...............................................................................66 Disease Assessment and Data Analysis ..........................................................................66 Greenhouse Citrus Canker Control Trials .......................................................................67 Citrus Canker Nursery Trials in Argentina .....................................................................68 Citrus Bacterial Spot Disease Control Trials at Dilley and Son Nursery in Avon

Park, Florida.................................................................................................................69 Citrus Bacterial Spot Disease Control Trials at the Plant Science Unit of the

University of Florida in Citra, Florida. ........................................................................69 Results.....................................................................................................................................71

Effect of Bacteriophage Application and Use of Protective Skim Milk Formulation on Citrus Canker Disease Development in the Greenhouse ........................................71

Effect of Phage Application on Disease Severity Incited by a Phage-Sensitive Xac Strain and Its Phage-Resistant Mutant.........................................................................71

Effect of Phage and Copper-Mancozeb Applications on Citrus Canker Disease Development in a Citrus Nursery ................................................................................72

Effect of Bacteriophage Treatment on Citrus Bacterial Spot Disease Development in a Commercial Citrus Nursery ..................................................................................72

Effect of Bacteriophage and Copper-Mancozeb Treatments on Citrus Bacterial Spot Disease Development in an Experimental Nursery .....................................................73

Discussion...............................................................................................................................73

5 INTERACTION OF BACTERIOPAHGES AND THE HOST BACTERIA ON THE PHYLLOPLANE....................................................................................................................79

Introduction.............................................................................................................................79 Material and Methods .............................................................................................................80

Standard Bacteriophage Techniques ...............................................................................80 Changes in Xanthomonas axonopodis pv. citrumelo Phage Populations on the Field....81 Interaction of Xanthomonas perforans and Its Bacteriophage on the Tomato

Foliage in the Greenhouse ...........................................................................................82

7

Interaction of Xanthomonas axonopodis pv. citri and Its Bacteriophages on Grapefruit Foliage in the Greenhouse..........................................................................83

Results.....................................................................................................................................85 Persistence of Bacteriophages on Citrus Leaf Surface Under Field Conditions .............85 Ability of Three Phages of Xanthomonas axonopodis pv. citri to Multiply on

Grapefruit Foliage in the Presence of Their Bacterial Host, and Their Effect on Citrus Canker Disease Development ...........................................................................86

Ability of a Xanthomonas perforans Phage to Multiply on Tomato Foliage in the Presence of Its Bacterial Host, and Its Effect on Tomato Bacterial Spot Disease Development ................................................................................................................87

Discussion...............................................................................................................................88 Overall Summary and Conclusions ........................................................................................89

APPENDIX

A CALCULATIONS..................................................................................................................98

Conversion of Horfall-Barratt Values to Mean Percentages ..................................................98 Calculation of Area Under the Disease Progress Curve.........................................................98

LIST OF REFERENCES.............................................................................................................101

BIOGRAPHICAL SKETCH .......................................................................................................112

8

LIST OF TABLES

Table page 3-1. Bacterial strains used in this study ........................................................................................51

3-2. Bacteriophages used in the study...........................................................................................53

3-3. Chloroform sensitivity, plaque morphology and host range of bacteriophages originating in Florida, Argentina and Japan .........................................................................54

3-4. Summary of morphological characteristics of phage virions. ...............................................57

3-5. Grouping of Florida bacteriophages based on BamH I RLFP profile. ..................................62

4-1. Effect of bacteriophage treatment and use of skim milk formulation on citrus canker disease development incited by Xanthomonas axonopodis pv. citri strain Xac65 on grapefruit plants in the greenhouse .......................................................................................77

4-2. Effect of bacteriophage treatment on citrus canker disease development incited by phage sensitive Xanthomonas axonopodis pv. citri strain Xac65 and its phage-resistant mutant, RF2 on grapefruit plants in the greenhouse .............................................................77

4-3. Effect of phage and copper-mancozeb application on citrus canker disease development, incited by Xanthomonas axonopodis pv. citri strain Xc05-2592 in nursery trial in Bella Vista, Corrientes, Argentina ...............................................................77

4-4. Effect of bacteriophage treatment on citrus bacterial spot disease development incited by naturally occurring Xanthomonas axonopodis pv. citrumelo strains at Dilley and Son Nursery in Avon Park, Florida in 2004, as measured by area under the disease progress curve (AUDPC)......................................................................................................78

4-5. Effect of bacteriophage and copper-mancozeb treatment on citrus bacterial spot disease development incited by Xanthomonas axonopodis pv. citrumelo strain S4m at the UF Plant Science Unit in Citra, Florida in 2006, as measured by area under the disease progress curve (AUDPC)......................................................................................................78

5-1. Effect of a phage treatments on citrus canker disease development, incited by Xanthomonas axonopodis pv. citri strain Xac65, as measured by disease severity..............94

5-2. Effect of a curative application of bacteriophage ΦXv3-3-13h on tomato bacterial spot disease development, incited by Xanthomonas perforans strain ΔopgH:ME-B, as measured by area under the disease progress curve (AUDPC). ...........................................97

A-1. Horsfall-Barratt scale, the corresponding disease intervals and midpoint values. ...............99

9

LIST OF FIGURES

Figure page 1-1. Young citrus canker lesions on grapefruit. ............................................................................24

1-2. Severe citrus canker infection on Key lime...........................................................................25

1-3. Citrus canker lesions on lemon fruit......................................................................................26

1-4. Asian leafminer (Phyllocnistis citrella) tunnels on Swingle citrumelo foliage.....................26

1-5. Citrus canker infection in Asian leafminer tunnels. ..............................................................27

1-6. Covered citrus nursery in Argentina......................................................................................27

1-7. Windbreaks outline an orange grove in Argentina................................................................28

1-8. Citrus bacterial spot lesions on grapefruit leaves. .................................................................29

1-9. Citrus bacterial spot lesions in Asian leafminer tunnels........................................................30

3-1. Transmission electron micrographs of representative phages. A) CP2, B) ΦXaacA1, C) ΦXaacF1, D) ΦXaacF2, E) ΦXaacF3, F) ΦXaacF5, G) ΦXaacF8. .....................................56

3-2. Dendogram and phage sensitivity matrix showing relationship amongst Xanthomonas strains causing citrus canker and citrus bacterial spot based on similarity of sensitivity profile against a battery of 12 phages.. .................................................................................59

3-3. Undigested bacteriophage DNA............................................................................................60

3-4. Bacteriophage DNA digested with BamH I ..........................................................................61

3-5. Representatives of the 7 RFLP groups based on BamH I. digestion profile .........................62

4-1. Experimental citrus nursery, the location of the 2006 citrus canker trials, at the INTA research station in Bella Vista, Corrientes, Argentina..........................................................75

4-2. Disease control plots in Dilley and Son Nursery, Avon Park, Florida..................................75

4-3. Field plots at the Plant Science Research and Education Unit of the University of Florida in Citra, Florida ........................................................................................................76

5-1. Bacteriophage populations and sunlight irradiation in Citra, FL on July 18 and 19, 2006...91

5-2. Changes in bacteriophage populations on grapefruit foliage in the presence or absence the host, Xanthomonas axonopodis pv. citri strain Xac65....................................................92

10

5-3. Changes in bacteriophage populations on grapefruit foliage in the presence or absence the host, Xanthomonas axonopodis pv. citri strain Xac65....................................................93

5-4. Symptoms of tomato bacterial spot, incited by Xanthomonas perforans..............................95

5-4. Populations of bacteriophage ΦXv3-3-13h on tomato foliage in the presence or absence of its host Xanthomonas perforans strain ΔopgH:ME-B......................................................96

A-1. Visualization of the disese progress curve and of the area under the disease progress curve. A) Disease progress curve is prepared by plotting the diseases percentage values (y-axis) against the time of disease assessment (x-axis). B) Area under the disease progress curve (AUDPC) is one value describing the overall disease progress throughout the season. ........................................................................................................100

11

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

CHARACTERIZATION AND USE OF BACTERIOPHAGES ASSOCIATED WITH CITRUS BACTERIAL PATHOGENS FOR DISEASE CONTROL

By

Botond Balogh

December 2006

Chair: Jeffrey B. Jones Major Department: Plant Pathology

Citrus canker, incited by Xanthomonas axonopodis pv. citri and X. axonopodis pv.

aurantifolii, is one of the most damaging citrus diseases in the world. Citrus canker was

reintroduced to Florida in the 1990s and threatens the state’s $9 billion citrus industry. This work

focused on a biological control approach to use bacteriophages for reducing bacterial pathogen

populations and disease severity on citrus. Bacteriophages isolated from citrus canker lesions in

Florida and Argentina were evaluated based on plaque morphology, chloroform sensitivity, host

range, genome size, DNA restriction profile and virion morphology. The phage isolates showed a

lack of diversity, as 61 of 67 bacteriophages were nearly identical and the remaining six were

identical to each other. Mixtures of bacteriophages were evaluated for controlling citrus canker

in greenhouse trials in Florida and in nursery trials in Argentina. Bacteriophages reduced citrus

canker disease severity both in greenhouse and field trials. The level of control was inferior to

chemical control with copper bactericides. The combination of bacteriophage and copper

treatments did not result in increased control. Citrus canker field trials in Florida have been

prohibited until recently, as the disease was under eradication. For this reason we evaluated the

efficacy of phage treatment on a similar bacterial citrus disease, citrus bacterial spot, incited by

X. axonopodis pv. citrumelo. Bacteriophages reduced citrus bacterial spot severity. The level of

12

control was equal or inferior to chemical control with copper bactericides. The combination of

bacteriophage and copper treatments did not result in increased control. In experiments

monitoring the fate of bacteriophages on the citrus foliage following bacteriophage application,

phage populations stayed steady on the foliage during nighttime but were drastically reduced

within hours after sunrise. The rate of reduction varied among the phages. The ability of

bacteriophages to multiply on the plant foliage in the presence of their bacterial host was

investigated. Phages varied in their ability to multiply, and the ones that successfully increased in

populations on the bacterial host on the leaf surface also reduced disease severity, whereas the

ones that were unable to multiply in the target environment did not reduce disease severity. In

summary, bacteriophages show significant promise as part of an integrated management strategy

for controlling citrus canker.

13

CHAPTER 1 CITRUS CANKER

The Citrus Industry

Citrus species originate from the southeast Asia-India region. They were introduced to the

Americas by Portuguese and Spanish explorers in the 16th century (21). Today citrus is produced

in 140 countries (122), mainly between the North and South 40º latitudes (121), and citrus fruits

are first among fruit crops in the international trade based on value (122). Citrus production has

been on the rise throughout the second half of the 20th century, and the total citrus production of

the world is around 105 million tons per year (122). Orange (Citrus sinensis) accounts for almost

two thirds of the total citrus production (65%), followed by tangerine (C. reticulata) (21%),

lemon (C. limon) (6%) and grapefruit (C. paradisi) (5.5%) (124). Other significant commercially

grown species are lime (C. aurantifolia), pummelo (C. grandis) and citron (C. medica). The

largest citrus producers are Brazil (20%), United States (14%), China (12%), Mexico (6%) and

the countries of the Mediterranean Basin (15%) (122). Most citrus fruits are produced for fresh

market consumption and only around 30% is processed (122). Most of the processing is orange

juice production that is carried out almost exclusively in São Paulo state of Brazil and in Florida

(122).

Citrus is a perennial evergreen with around 50 years of commercial production (121).

While originally it was grown on its own root system, in modern commercial production the

producing plants are grafted onto rootstocks (121). Rootstocks influence adaptation to soil types,

tree size, production characteristics and can provide cold hardiness and resistance to diseases,

nematodes and insects (53, 54, 105, 111). Some important rootstock species are sour orange (C.

aurantium L.), trifoliate orange (Poncirus trifoliata) and citrumelo (C. paradisi ×P. trifoliata)

(111).

14

The total citrus production of the United States was 11.6 million tons ($2.68 billion value)

in the 2005-06 growing season (125). Florida is the biggest citrus producer in the country

providing 68% of the total production, followed by California (28%), Texas (4%) and Arizona

(4%).

Florida’s $9 billion citrus industry (49) is in a rather precarious situation presently. It was

heavily hit in 2004 and 2005 by hurricanes; as a result the total citrus production fell 40%

compared to the pre-hurricane 2003-04 season and the producing area shrunk to the lowest since

1994 with 576,400 bearing acres (125). Another effect of the two consecutive extreme hurricane

seasons was the large scale dissemination of the citrus canker bacterium throughout the state,

which eliminated hopes of eradicating the disease (34). The permanent presence of citrus canker

is expected not only to reduce production volume, increase prices and cause market losses (139),

but also to completely eliminate grapefruit production from the state (49). Moreover, citrus

greening, another devastating citrus disease, was detected in Florida in 2005 (60) and is known

to be present in 12 counties in south Florida (36). Citrus greening is caused by a phloem limited

bacterium, Liberibacter asiaticus, and unlike citrus canker it kills the infected trees (60). Due to

its epidemiology and the presence of its vector, the Asian citrus psyllid (Diaphorina citri) in

Florida, its eradication is not feasible (35).

Symptoms, Etiology and Epidemiology

Citrus canker is one of the most devastating diseases affecting citrus production worldwide

(49). Its center of origin is the southeast Asia-India region (21), similar to its host plants. By the

20th century it was present in most citrus growing areas around the globe: Asia, South and

Central Africa, North and South America, Australia, New Zealand and the Pacific islands (71).

Citrus canker causes erumpent lesions on fruit, foliage and young stems (Figures 1-1, 1-2,

1-3) (21). The earliest symptoms on leaves are minuscule, slightly raised blister-like lesions that

15

appear around 7 days after inoculation under optimum conditions (Figure 1-1). Optimum

temperature for disease development is 20-30°C (71). The lesions expand over time and their

color turns light tan and then tan-to-brown (49). Subsequently, water-soaked margins appear

around the lesions frequently surrounded by a chlorotic halo (Figure 1-2) (49). The center of the

lesion becomes raised and corky and the touch of the lesions feels like sand paper (106).

Eventually the lesions form a crater-like appearance (106) and in some hosts they fall out leaving

a shothole appearance (106). Severe disease can cause defoliation (47), dieback and fruit drop

(49, 71).

A single, plasmid-encoded bacterial protein, PthA, is responsible for inciting the symptoms

(30). It is delivered into the cytoplasm of plant cell by the type 3 secretion system of the

pathogen, from where it is transported into the nucleus (136). This protein is believed to be

important in inciting cell division, cell enlargement and cell death (20, 30).

The bacterium multiplies to high populations in the lesions and in the presence of free

moisture the cells will ooze out to the plant surface. Rain water collected from diseased leaves

contains high bacterial concentrations ranging from 104 to 105 cfu/mL (108). The inoculum is

then dispersed by wind driven rain to new growth on the same plant or to other plants (19,49).

Wind blown inoculum was disseminated up to 32 meters in Argentina (107); however, in Florida

rainstorms move the pathogen much further with estimates of up to 7 miles (50). Long distance

spread of inoculum occurs mostly as a result of human involvement either by moving diseased

plant material or on contaminated equipment (50). Extreme weather conditions, such as

hurricanes and tornadoes have also been shown to facilitate long distance disease spread (51,57).

Bacteria enter the tissues through stomatal openings (48,49) or wounds. Wind speeds

above 18 mph aid in penetration of bacterial cells through the stomatal pores (48). Wounding

16

often occurs by thorns, blowing sand or insect damage (49). Asian leafminer (Phyllocnistis

citrella), an insect that was introduced to Florida in 1993, creates wounds that expose the

mesophyll tissues allowing entry of bacterial inoculum (Figure 1-4, 1-5) (51). Leafminer itself is

not a vector of the disease (17), but its actions can lead to significant field infection even on

highly resistant cultivars and species (49).

Bacteria multiply in the expanding lesions (49) and survive only in the margins of old

lesions. The pathogen persists in lesions in the leaves or fruits until the tissues decompose;

however, long term survival, up to a few years, occurs in lesions in woody tissues (49). Outside

of plant tissues the bacterium is quickly eliminated by desiccation or sunlight irradiation (49),

and it cannot survive longer than a few days in the soil, probably due to competition with

saprophytes (43).

The two main determinants of host susceptibility are the stage of leaf expansion and the

resistance of mesophyll tissue (48,58). All above ground citrus tissues of susceptible genotypes

are sensitive when they are young, especially at the second half of the expansion phase of growth

(108).

Disease Impact

Citrus canker impacts the citrus industry on several levels. It reduces both fruit yield and

quality. While yield reduction has an effect in all spheres of citrus production, the quality

problems affect mostly the fresh fruits as the blemishes caused by the disease (Figure 1-3) render

the fruit aesthetically undesirable and thus unmarketable (50). Additional economic damages

result from market losses due to the disease exclusion policies in canker-free citrus growing

regions. Those regions and countries where citrus canker is not present ban importation of fruits

from canker inflicted areas because of the fear of introduction of the pathogen (50). Furthermore,

if citrus canker becomes endemic, commercial production of the most susceptible citrus species,

17

such as grapefruit, will cease as it becomes impossible to grow them profitably (50). More

resistant species such as tangerines will likely take their place (71,103).

Disease Management

Countries that are free of citrus canker apply quarantine measures to stop the introduction

of the pathogen. As a disease exclusion measure, they prohibit importation of fruits and plant

material from canker inflicted areas (50). Introduction of the disease is generally met by

eradication programs in which all infected and exposed citrus is destroyed (50). The disease was

subject to eradication with varying degrees of success; it was successfully eradicated from

Mozambique, New Zealand, Australia, South Africa, the Fiji Islands and twice from the US,

while eradication efforts failed in Argentina, Uruguay, Paraguay, and more recently in Florida

(98,103,106). The disease is currently under active eradication in Australia and São Paulo state

in Brazil (39,103). In endemic situations the emphasis has shifted to implementing integrated

management programs (76). The disease is under management in all Asian countries, Argentina,

Uruguay, Paraguay, several states of Brazil and Florida (49,76,109). These programs rely on

• planting resistant citrus cultivars (76), • production of disease-free nursery stock by locating nurseries outside of citrus canker areas

and/or indoors (Figure 1-6) (76), and • restricting disease spread by establishing windbreaks (Figure 1-7) and fences around

groves, using preventative copper bactericides (76,109) and by controlling Asian leafminer.

Additionally, in order to ensure that fresh fruit destined for internal and export markets is

disease free, producing groves are regularly inspected for the presence of citrus canker and

sanitation protocols are established in the packing houses (75).

Genetic Variation of the Pathogen

Citrus canker is not a disease caused by a single pathogen but rather a of group similar

diseases caused by closely related Xanthomonas species (49). The Asiatic type (Canker A) is

18

caused by X. axonopodis pv. citri (Xac) strains that originated from Asia. This is the most

geographically widespread pathogen, which also has the widest host range and by far the biggest

impact. Almost all commercially grown citrus varieties are susceptible to this bacterium to a

certain level. Grapefruit, Key lime and lemon are most susceptible. Sweet oranges range from

highly susceptible (Hamlins and Navels) to moderately susceptible (Valencias). Tangerines and

mandarins are moderately resistant while only calamondin (C. mitus) and kumquats (Fortunella

spp.) are highly resistant. The B type of citrus canker (cancrosis B or false canker) was present in

Argentina, Uruguay and Paraguay (103) from the 1920s to 1980s, and eventually was eliminated

by strains of the A pathotype. The causal agent, X. axonopodis pv. aurantifolii (Xaa) , has

limited host range compared to canker A: it is only pathogenic on lemon, Key lime, sour orange,

and pummelo. It was not pathogenic on grapefruit. The C type of citrus canker (cancrosis C) is

also caused by X. axonopodis pv. aurantifolii. It has only been found in São Paulo state in Brazil

and its only two known hosts are Key lime and sour orange. A fourth group of strains (A*

pathotype) was isolated in southwest Asia in the 1990s (126). These strains constitute a subgroup

of the A pathotype, X. axonopodis pv. citri, but their host range is limited to Key lime (5). The

different pathotypes can be distinguished by host range, cultural and physiological characteristics

(49), bacteriophage sensitivity (21), serology (6), plasmid fingerprints (94), DNA-DNA

homology (32), and by various RFLP and PCR analyses (26).

Citrus Canker in Florida

Citrus canker was first introduced into Florida in 1912 on rootstocks imported from Japan

(29). It spread to all Gulf states between Texas and South Carolina. A strict eradication program

was implemented and the disease was eliminated from Florida by 1933 and from the entire

United States by 1947 (29). The second eradication program began in 1984, when Xanthomonas

was isolated from lesions from nursery stock (102,103). Research later showed that this pathogen

19

differed from xanthomonads causing citrus canker. The bacterium was determined to be

endemic, exclusive to Florida, and considerably less aggressive than Xac (103,106). The disease

was named citrus bacterial spot (CBS), the pathogen was named X. axonopodis pv. citrumelo,

and the eradication efforts for CBS were cancelled in 1990 (103). By that time 20 million citrus

plants were destroyed and $94 million was spent (103). In 1986 genuine citrus canker was

discovered in the Tampa Bay area. An eradication effort was begun and the pathogen was

declared eradicated in 1994 (103). Another introduction was discovered in urban Miami in 1995.

The pathogen, called Miami genotype, is determined to be genetically related to Xac strains from

several geographical areas from Southeast Asia and South America (50). The Miami genotype

later spread throughout the state despite the eradication program (26,103) and was responsible

for the majority of post 1997 outbreaks (50). In 1997 there was a new outbreak in the Tampa Bay

area. The pathogen, called Manatee genotype, was identical to Xac strains from China and

Malaysia, based on rep-PCR analysis (26). Also, it was indistinguishable from Xac strains that

caused the outbreak in the 1980s in the Tampa Bay region (26,103), implying that the earlier

eradication effort was not successful. Interestingly, the Manatee genotype disappeared again in

1999 and then reappeared in 2005 (113). In 2000, a new genotype of Xac was identified and was

designated the Wellington genotype. It was discovered in the Palm Beach area. Its host range

was limited to Key Lime (114). Wellington genotype is closely related to the A* strains and thus

likely originates from South-west Asia (26). It was later determined that the inability of this

genotype to infect grapefruit was due to the presence of an avirulence gene, avrGf1, in its

genome (98). The latest introduction of an exotic strain was in 2003 in Orange County. These

strains were found in a residential area on an Etrog citron tree (Citrus medica) that was probably

20

brought to Florida from Pakistan illegally (Etrog genotype) (X. Sun, D. Jones, R.E. Stall,

personal communication).

Citrus Canker Eradication Program

The Citrus Canker Eradication Program (CCEP) was established in 1995 in response to the

Miami outbreak by the Florida Department of Agriculture and Consumer Services (FDACS),

Division of Plant Industry (DPI) and the USDA, Animal and Plant Health Inspection Service

(APHIS) (49). The original quarantine area was 14 square miles in the urban Miami area (103).

After the discovery of citrus canker in the Tampa Bay area in 1997, the quarantine area was

extended to that region as well (49). Citrus trees infected by the disease or located within the

exposure area were uprooted and burned in the commercial groves, or cut down and chipped in

urban areas (49). The exposure area was originally 125 feet radius of a diseased tree, based on

data collected from Argentina (108). Later research showed that the inoculum spreads further

than 125 feet under the Florida conditions (50,51), and the exposure radius was extended to 1900

feet in January 2000 based on epidemiological data collected in the Miami area (50,51). The

eradication program was hindered by strong public resistance; the ensuing legal battles often

delayed the surveying and tree removal (103). The legal challenges were overcome in spring

2004, but then the hurricanes of 2004 and 2005 spread the pathogen from areas awaiting

eradication widely throughout the citrus growing area. On January 10, 2006, APHIS

discontinued the CCEP (24,123), because its continuation in the new situation was judged

infeasible (i) due to financial constraints associated with the tree removal and reimbursement

programs, and (ii) because the citrus industry could not survive the loss of such a large

production area (24,123). During the program 16.5 million trees were destroyed, including 11.3

million trees from commercial groves, equaling 15% of the total bearing acreage (34). The total

costs of CCEP exceeded $600 million (138). In March 2006 APHIS released a Citrus Health

21

Response Plan, which outlined procedures for managing citrus production in the permanent

presence of the disease (34).

The impact of living with the endemic citrus canker situation in the Florida is estimated at

$254.2 million annually (139) and may result in elimination of grapefruit production (49). The

acceptance of citrus canker as an endemic disease will also result in losses in interstate and

international commerce of the state’s fresh citrus fruit, which currently represents 20% of the

state’s $9 billion dollar citrus industry (88).

Bacteriophages Associated with Citrus Canker

There are several reports on bacteriophages found in association with citrus canker. Phages

CP1 and CP2 have been isolated in Japan (45). Goto found that both CP1 and CP2 had wide host

ranges and more than 97% of Xac strains present in Japan were sensitive to one of these two

phages (44). In fact, the Japanese Xac strains comprised two groups: the strains in the first one

originated mostly from Unshu orange (Citrus unshu) and were sensitive to CP2 only, whereas

the members of the second group had a variety of hosts and were sensitive only to CP1. These

phages have been used for detection of the pathogen (45). Bacteriophage CP3, which was also

isolated in Japan, had a tadpole shape with a spherical head and long tail (46). Goto et al. (46)

found that strains of the B pathotype (X. axonopodis pv. aurantifolii) could be distinguished from

type A strains (Xac) based on sensitivity to phage CP3, with all B strains being sensitive to CP3

and all A strains being resistant. Canker C strains were differentiated from canker A strains by

their resistance to both CP1 and CP2 (106).

Filamentous phage Cf was isolated in Taiwan (27). It has a very narrow host range (44),

contains single stranded DNA that is approximately 1 kb long, produces small and clear plaques

and is a temperate phage (27). The product of pilA gene, a type 4 prepilin of the host bacterium is

required for infection of Cf (27). Phages CP115 and CP122, isolated from citrus canker lesions

22

in Taiwan, tested for their ability to lyse Taiwanese strains of Xac, lysed 97.8% of them when

used in combination (135). These phages, however, did not lyse Xanthomonas strains that did not

cause citrus canker, or any other bacteria tested. The authors concluded that these phages could

be used for specific detection of Xac strains in Taiwan. Temperate phage PXC7 that was isolated

from Japanese Xac strain XCJ18 produces small turbid plaques with irregular borders and is

sensitive to chloroform (134). When Xac strain XCJ19 was lysogenized with the phage, its

colony morphology changed from smooth to dwarf and became resistant to phage CP2 (134).

Bacteriophages were also found in citrus canker lesions in Argentina in 1979 (R.E. Stall,

personal communication).

Citrus Bacterial Spot

Citrus bacterial spot (CBS) was discovered in 1984 as a new Xanthomonas disease of

citrus nursery stock that causes canker-like symptoms (102). Citrus bacterial spot differs from

citrus canker in that it causes flat or sunken lesions (Figure 1-8) instead of the corky, raised ones

and is generally less aggressive than citrus canker (106). The pathogen, Xanthomonas

axonopodis pv. citrumelo (Xacm) only exists in Florida (103) and is genetically different than

Xanthomonas strains causing citrus canker (26). Xacm strains are genetically diverse (26) and

comprise three groups based on aggressiveness as measured by rate of lesion expansion and the

ability to multiply in citrus leaves (56). Only the most aggressive strains are able to maintain

high populations in the lesions. Graham et al. (56) questioned if the less aggressive strains

should be considered a citrus pathogen at all. The most aggressive strains are disseminated by

wind driven rain, whereas strains of intermediate and low aggressiveness are spread mainly by

mechanical means (56). Wounding caused by thorns and citrus leafminer (Figure 1-9) facilitates

pathogen entry and increases CBS disease incidence and severity. The disease causes a reduction

in photosynthetically active leaf area and in severe cases can induce leaf drop. Xacm infection

23

also causes bud failure in Swingle citrumelo plants (55). The CBS bacteria are most aggressive

on trifoliate orange, Swingle citrumelo and grapefruit (103,106).

Project Goal and Objectives

The goal of this project was to develop a bacteriophage-based disease control strategy that

could be used as part of an integrated management program against citrus canker in Florida. The

objectives were (i) establishment of a bacteriophage collection against Xac strains present in

Florida, (ii) collection and characterization of bacteriophage associated with citrus canker in

Florida and Argentina; (iii) determination of bacteriophage sensitivity of Xac strains present in

Florida, and (iv) evaluation of bacteriophages for suppression of citrus canker in greenhouse

experiments in Florida and in field experiments in Argentina, and for suppression of citrus

bacterial spot, as a model disease system, in field experiments in Florida.

24

Figure 1-1. Young citrus canker lesions on grapefruit.

25

Figure 1-2. Severe citrus canker infection on Key lime.

26

Figure 1-3. Citrus canker lesions on lemon fruit.

Figure 1-4. Asian leafminer (Phyllocnistis citrella) tunnels on Swingle citrumelo foliage.

27

Figure 1-5. Citrus canker infection in Asian leafminer tunnels.

Figure 1-6. Covered citrus nursery in Argentina.

28

Figure 1-7. Windbreaks outline an orange grove in Argentina.

29

Figure 1-8. Citrus bacterial spot lesions on grapefruit leaves.

30

Figure 1-9. Citrus bacterial spot lesions in Asian leafminer tunnels.

31

CHAPTER 2 THE USE OF BACTERIOPHAGES FOR CONTROLLING PLANT DISEASES

Early History

Bacteriophages were discovered in the beginning of the 20th century independently by

Twort in 1915 and by d’Herelle in 1917 (112). There were differences in interpretation about the

nature and origin of this “lytic principle”. Twort proposed that a bacterial enzyme caused the

lysis, while d’Herelle speculated that a virus was responsible for the phenomenon. A direct

consequence of d’Herelle’s concept was the idea of using phages for controlling bacterial

diseases. Soon after the first medical (112) and veterinary (28) applications, phages were

evaluated for control of plant diseases.

In 1924 Mallman and Hemstreet (79) isolated the “cabbage-rot organism”, Xanthomonas

campestris pv. campestris, from rotting cabbage and demonstrated that the filtrate of the liquid

collected from the decomposed cabbage inhibited in vitro growth of the pathogen. The following

year Kotila and Coons (73) isolated bacteriophages from soil samples that were active against the

causal agent of blackleg disease of potato, Erwinia carotovora subsp. atroseptica. They

demonstrated in growth chamber experiments that co-inoculation of E. carotovora subsp.

atroseptica with phage successfully inhibited the pathogen and prevented rotting of tubers (73).

These workers also isolated phages against Erwinia carotovora subsp. carotovora and

Agrobacterium tumefaciens from various sources such as soil, rotting carrots and river water

(25). Thomas (120) treated corn seeds that were infected with Pantoea stewartii, the causal agent

of Stewart’s wilt of corn, with bacteriophage isolated from diseased plant material. The seed

treatment reduced disease incidence from 18% to 1.4%. Despite the promising early work, phage

therapy did not prove to be a reliable and effective means of controlling phytobacteria. Several

workers questioned if positive results were possible. In 1963, Okabe stated, “in general, the

32

phage seems to be ineffective for [controlling] the disease development” (91). Three decades

later, Goto concluded, “practical use of phages for control of bacterial plant disease in the field

has not been successful.” (44). Chemical control with antibiotics and copper compounds became

the standard for controlling bacterial plant diseases (31,81).

Other Uses of Phages in Plant Pathology

Bacteriophages still remained in use in plant pathology and have been used as tools for

detection, identification, classification, and enumeration of pathogenic bacteria and were also

used for disease forecasting. Phage typing, as a method of differentiating different races or

pathovars of the same bacterial species became a standard method in plant epidemiological

studies (70). Phages CP1 and CP2 of Xanthomonas axonopodis pv. citri, the causal agent of

citrus canker, were used for species-specific identification and classification of strains of the

pathogen in Japan (44). These two phages in combination with phage CP3 were used for

differentiating worldwide strains causing citrus canker (46). Wu et al. (135) used phages CP115

and CP122 for identification of X. axonopodis pv. citri strains in Taiwan. Phages were used for

detection of the host bacterium from crude samples and seed lots (68) and directly from lesions

on the plant foliage (91) by monitoring increases in homologous phage concentration. Okabe and

Goto (91) demonstrated that phages could be used for quantifying bacterial cells based on the

number of newly produced phages and average burst size. They developed a method for

indirectly forecasting bacterial leaf blight by monitoring phage titers in rice fields (91).

Reliability of this latter method was questioned later by Civerolo (22).

Return of Phage-Based Disease Control

Several factors have contributed to the re-evaluation of phage therapy for plant disease

control. The use of antibiotics has been largely discontinued in agriculture due to the emergence

of antibiotic-resistant bacteria in the field (80,84,119) and because of concerns of possible

33

transfer of antibiotic resistance from plant pathogens to human pathogens. The feasibility of

reliance on copper compounds is questioned, because of the emergence of copper-tolerant strains

among phytobacteria (81,129); phytotoxicity caused by ionic copper (85,109) and soil

contamination from extended heavy use (72). Additionally, concerns about food safety and

environmental protection and the goal of achieving sustainable agriculture necessitated

development of safer, more specific and environment-friendly pesticides (96). These factors,

together with the expanding knowledge base about phage application in medicine (13,31,74), led

to renewed interest in bacteriophage-based disease control in modern agriculture.

Considerations About Phage Therapy

Greer (59) and Kutter (74) identified the several advantages of using phages for disease

control.

1. Phages are self-replicating and self-limiting; they replicate only as long as the host bacterium is present in the environment, but are quickly degraded in its absence (74).

2. Bacteriophages are natural components of the biosphere; they can readily be isolated from everywhere bacteria are present, including soil, water, plants, animals (2,52,133) and the human body (93).

3. Phages could be targeted against bacterial receptors that are essential for pathogenesis, so resistant mutants would be attenuated in virulence (74).

4. Bacteriophages are non-toxic to the eukaryotic cell (59). Thus, they can be used in situations where chemical control is not allowed due to legal regulations, such as for treatment of peach fruit before harvest (137) or for control of human pathogens in fresh-cut produce (77,78).

5. Phages are specific or highly discriminatory, eliminating only target bacteria without damaging other, possibly beneficial members of the indigenous flora. Thus their use can also be coupled with the application of antagonistic bacteria for increased pressure on the pathogen (117); or they can be used to promote a desired strain against other members of the indigenous flora (14).

6. Phage preparations are fairly easy and inexpensive to produce and can be stored at 4 ºC in complete darkness for months without significant reduction in titer (59). Application can be carried out with standard farm equipment and since phages are not inhibited by the majority of agrochemicals (9, 137), they can be tank-mixed with them without significant loss in titer. Copper-containing bactericides have been shown to inactivate phages (9,66); however, inhibition was eliminated if phages were applied at least three days after copper (66).

A number of disadvantages and concerns have been raised in relation to phage therapy

(59,74,127); moreover, additional problems specific to agricultural applications have surfaced.

34

1. Limited host range can be a disadvantage, as often there is diversity in phage types of the target bacterium (59). Several approaches have been tried for addressing this problem: using broad host range phages (99,115), using host range mutant phages (37,38,89), applying phages in mixtures (38) or even to breed them (62).

2. The requirement of threshold numbers of bacteria (104-106 cfu/mL) may limit the impact of phages (131).

3. Emergence of phage resistant mutants can render phage treatment ineffective. However, using mixtures of phages that utilize distinct cell receptors can suppress the emergence of resistance (118). Also, phage resistance often comes at some metabolic cost to the bacteria. Loss of virulence was observed with phage-resistant mutants of Ralstonia solanacearum (61), Xanthomonas campestris pv. pruni (95), and Pantoea stewartii (120).

4. Environmental effects, such as temperature, pH and physiology of bacterium can hinder control. Civerolo (22) observed that Xanthomonas phaseoli phages attacked Xanthomonas and Pseudomonas species only at temperatures above 20ºC. Vidaver (128) suggested that P. syringae and P. phaseolicola, causal agents of halo blight and brown spot of bean, may be more prevalent below 22ºC because of phage resistance. Leverentz (78) noted that phage treatment caused a significant population reduction of the Listeria monocytogenes on melons but not on apples, because phages were unstable on apple slices, possibly due to low pH (4.37 in apple vs. 5.77 in melon) (78).

5. Unavailability of target organism can hinder control. Plant pathogenic bacteria often occur in non-homogenous masses surrounded by extracellular polysaccharides that protect them from phage attachment (44,91), or reside in protected spaces on the surface, or inside the plant and unavailable for the control agents (22).

6. There is a concern that phages have the potential of transducing undesirable characteristics, such are virulence factors, between bacteria (127).

7. Lysogenic conversion, alteration of phenotypic characteristics of lysogenized bacteria by their prophages, have been found to have undesirable consequences, such as resistance to bacteriophages, toxin production or even increased virulence. When Xanthomonas axonopodis pv. citri strain XCJ19 was lysogenized with temperate phage PXC7, it became resistant to phage CP2 (134). Phage-associated toxin production has not been documented amongst phytobacteria, but such cases are known amongst human pathogens (130) and in bacteria of plant associated nematodes (92). Goto (44) reported that Xanthomonas campestris pv. oryzae strains lysogenized by phages Xf or Xf-2 became more virulent on rice.

8. Consumer perception of adding viruses to food products also could become an issue (59). 9. Vidaver (127) raised the concern that “transducing phages can introduce active prokaryotic

genes into plant and animal cells”. 10. Despite the generally narrow host ranges of phages, negative side effects due to inhibition of

beneficial bacteria are possible. Examples for negative phage impact in agriculture include studies in which the phage-incited reduction in symbiotic nitrogen fixing bacteria reduced growth and nitrogen content of cowpea (4) and in which biocontrol ability of Pseudomonas flourescens was abolished by a lytic bacteriophage (69).

Factors Influencing Efficacy of Phages as Biological Control Agents

Goodridge stated that the efficacy of phage therapy depends on the ability of a phage to

find its host before it is destroyed (42). According to Johnson (67) the success of a particular

35

biocontrol treatment is influenced by agent and target densities. A component of Johnson’s

model is the possibility that the target resides in spatial refuges where the biocontrol agent

cannot penetrate. Gill and Abedon (40) proposed several additional factors specifically in

relation to phage therapy: the location in which the target pathogen resides; the presence of

adequate water as a medium for virus diffusion; rates of virion decay; timing of phage

application; phage in situ multiplication ability; and relative fitness of phage-resistant bacterial

mutants.

Gill and Abedon (40) looked at the factors that could contribute to success or failure of

phage therapy in the rhizosphere and in the phyllosphere. They suggested that phage therapy

might meet with more success in the rhizosphere because phages are readily isolated from there

and can survive longer in the soil than on the leaf surface. However, they identified several

factors that can hinder success of disease control in the rhizosphere. The rate of diffusion through

the heterogeneous soil matrix is low and changes as a function of available free water. Phages

can become trapped in biofilms (110) and reversibly adsorb to particles of the soil, such as clay

(132). Low soil pH can also inactivate them (116). Physical refuges can protect bacteria from

coming into contact with phages, and due to the low rates of phage diffusion and high rates of

phage inactivation only a low number of viable phages is available to lyse target bacteria (40).

An additional problem is the need for high population of both phage and bacterium in order to

start the “chain reaction” of bacterial lysis (40).

The phyllosphere is a harsh environment, because of high UV and visible light irradiation

and desiccation (40). It has been noted that phages were harder to isolate from aerial plant tissue

than from soil even for pathogens of aerial tissues (37,41,91). Phages applied to aerial tissues

degrade extremely rapidly during the day (8,10,23,66,83). Additionally, the lack of moisture on

36

the leaf surfaces does not allow phage dispersion except for temporary leaf wetness periods after

application, during rain events or when dew is present on the leaves at night and early morning.

There were several approaches to increase efficacy of control in the phyllosphere environment,

including applying treatments in the evening or early morning (10,38), using protective

formulations that increase phage longevity on the foliage (8,10,66,89) and using carrier bacteria

for phage propagation in the target environment (115). Using phages isolated from aerial tissues

might be advantageous, as they might be better adapted for surviving and multiplying on the

plant surfaces. A phyllosphere phage investigated by Iriarte et al. (66) turned out to be resistant

to desiccation.

Timing of bacteriophage applications relative to the arrival of the pathogen influenced

efficacy of disease control in several instances. Civerolo and Keil (23) achieved a marked

reduction of peach bacterial spot only if phage treatment was applied one hour or one day before

inoculation with the pathogen. There was a slight disease reduction when phage was applied one

hour after inoculation and no effect if applied one day later. Civerolo (22) suggested that bacteria

were inaccessible to phage in the intercellular spaces, or there were not enough phages reaching

the pathogen. Schnabel et al. (101) achieved a significant reduction of fire blight on apple

blossoms when the phage mixture was applied at the same time as the pathogen, Erwinia

amylovora. In contrast, disease reduction was not significant when phages were applied a day

before inoculation. Berghamin Filho (18) investigated the effect of timing on the efficacy of

phage treatment in greenhouse trials with two pathosystems: black rot of cabbage, caused by

Xanthomonas campestris pv. campestris and bacterial spot of pepper, caused by Xanthomonas

campestris pv. vesicatoria. Phage treatment was applied once varying from 7 days before to 4

days after pathogen inoculation. On cabbage significant disease reduction was achieved if the

37

phage treatment was applied 3 days before to 1 day after inoculation, whereas on pepper it was

achieved when applied from 3 days before to the day of inoculation. The greatest disease

reduction occurred with application of phages the same day as inoculation in both pathosystems.

The effect of phage concentration on disease control efficacy has also been investigated.

Balogh (8) treated tomato plants with phage mixtures of 104, 106 and 108 PFU/mL before

inoculating with Xanthomonas perforans, causal agent of tomato bacterial spot. The two higher

concentrations significantly reduced disease severity, whereas the lowest concentration did not.

Current Research

There has been considerable amount of work on the use of phages for control of bacterial

spot of peach, caused by Xanthomonas campestris pv. pruni. Civerolo and Keil (23) reduced

bacterial spot severity on peach leaves under greenhouse conditions with a single application of a

single-phage suspension. Zaccardelli et al. (137) isolated eight phages active against the

pathogen, screened them for host range and lytic ability, and selected a lytic phage with the

broadest host range for disease control trials. Biweekly spray-applications of the phage

suspension in producing orchards significantly reduced bacterial spot incidence on fruits (99).

Tanaka et al. (117) treated tobacco bacterial wilt, caused by Ralstonia solanacearum, by

co-application of an antagonistic avirulent R. solanacearum strain and a bacteriophage that was

active against both the pathogen and the antagonist. The avirulent strain alone reduced the ratio

of wilted plants from 95.8% to 39.5%, whereas the co-application of the avirulent strain and the

phage resulted in 17.6% wilted plants.

Control of Erwinia amylovora, the fire blight pathogen of apple, pear and raspberry, with

bacteriophages is currently under investigation in Canada and the USA. Schnabel et al. (101)

used a mixture of three phages for controlling fire blight on apple blossoms and achieved

significant (37%) disease reduction. Gill et al. (41) isolated 47 phages capable of lysing E.

38

amylovora and categorized them based on plaque morphology and host range. Later the phages

were evaluated for disease control ability in pear blossom bioassays, and the ones with broad

host ranges and best disease control ability were selected for subsequent orchard trials (115).

Pantoea agglomerans, a bacterial antagonist that was also sensitive to the phages, was used to

deliver and propagate them on the leaf surface. Disease control comparable to streptomycin was

achieved (115).

There has been extensive research on suppressing tomato bacterial spot with phage.

Flaherty et al. (38) effectively controlled the disease in greenhouse and field experiments with a

mixture of four host-range mutant phages active against the two predominant races of the

pathogen, X. campestris pv. vesicatoria. Balogh et al. (10) enhanced the efficacy of phage

treatment with protective formulations that increased phage persistence on tomato foliage.

Obradovic et al. (89,90) used formulated phages in combination with other biological control

agents and systemic acquired resistance inducers, as a part of integrated an disease management

approach. Phage-based integrated management of tomato bacterial spot is now officially

recommended to tomato growers in Florida (85), and bacteriophage mixtures against the

pathogen are commercially available (Agriphage from OmniLytics Inc. Salt Lake City, UT, EPA

Registration # 67986-1).

Other important work includes the reduction of incidence of bacterial blight of geranium

with foliar applications of a mixture of host-range mutant phages (37), disinfection of

Streptomyces scabies-infected seed potatoes using a wide host range phage (82), and a reduction

in the loss of cultivated mushrooms caused by bacterial blotch with phage applications (86,87).

Outlook for the Future

Use of bacteriophages for controlling plant diseases is an emerging field with great

potential. The concern about environment-friendly sustainable agriculture and the rise of organic

39

production necessitates improvements in biological disease control methods, including the use of

bacteriophages against bacterial plant pathogens. On the other hand, the lack of knowledge about

the biology of phage-bacterium-plant interaction and influencing factors hinders progress in the

field. Much research in these areas is needed before phages can become effective and reliable

agents of plant disease management.

40

CHAPTER 3 CHARACTERIZATION OF BACTERIOPHAGES ASSOCIATED WITH CITRUS CANKER

IN FLORIDA AND ARGENTINA

Introduction

In order to develop a phage based disease control strategy it is necessary to (i) establish a

collection of phages that could be used for control; (ii) evaluate diversity of the target organism,

and, (iii) determine if there are phages associated with the pathogen in nature and if so, assess

their impact.

Xanthomonas strains causing citrus canker have been classified based on phage sensitivity

in the past. Goto (44) categorized Xanthomonas axonopodis pv. citri (Xac) strains present in

Japan, causing A type of citrus canker, into two groups based on their sensitivity to phages CP1

and CP2. More than 97% of Xac strains were sensitive to one of these two phages, but none to

both of them. The strains that were sensitive to CP2 originated mostly from Unshu orange

(Citrus unshu), whereas the ones sensitive only to CP1 had a variety of hosts. Goto et al. (46)

found that strains of the B pathotype of citrus canker (X. axonopodis pv. aurantifolii (Xaa))

could be distinguished from type A strains (Xac) based on sensitivity to phage CP3, with all B

strains sensitive to CP3 and all A strains resistant. Strains of the C pathotype (Xaa) were also

differentiated from A strains by their resistance to both CP1 and CP2 (106).

There is known diversity amongst Xac strains in Florida. Four distinct genotypes have

caused outbreaks since 1993 (49): the Miami, the Manatee, the Wellington and the Etrog

genotypes. Cubero et al. (26) used rep-PCR protocol for distinguishing these genotypes and to

determine their geographic origin. They determined that (i) the Miami genotype was related to

Xac strains from several geographical areas in southeast Asia and South America; (ii) the

Manatee genotype was identical to Xac strains from China and Malaysia and also to Xac that was

present in Florida in the 1980s and was supposedly eradicated; and (iii) the Wellington genotype

41

was related to A* strains (Xac) from southwest Asia. There has not been any genetic analysis

published in relation to the Etrog strains, but it is suspected that they were brought into Florida

on plant material from Pakistan.

The objective of this project was to assemble a phage collection active against citrus

canker, partly from academic and commercial sources, and partly by isolating them from plant

tissue showing citrus canker symptoms in Florida, where the disease was under eradication, and

in Argentina, where the disease is endemic. Once a collection was established and the phages

were grouped based on host range, they were used to type the prevalent Xac strains in Florida.

The typing results provided information about the diversity of the pathogen and determined

which phages could be used for disease control. Changes in phage sensitivity of Xac strains

isolated in different years may provide information on what impact naturally occurring phages

have on the pathogen.

Materials and Methods

Bacterial Strains and Bacteriophages

Bacterial strains were grown on nutrient agar (NA) medium (0.8% (wt/V) nutrient broth

(NB) (BBL, Becton Dickinson and Co., Cockeysville, MD) and 1.5% (wt/V) Bacto Agar (Difco,

Becton Dickinson and Co., Sparks, MD)) at 28°C. For bacteriophage detection and propagation

either semisolid nutrient agar yeast extract medium (NYA), (0.8% Nutrient Broth, 0.6% Bacto

Agar and 0.2% Yeast Extract (Difco, Becton Dickinson and Co., Sparks, MD)) or liquid nutrient

broth medium was used. Sterilized tap water or SM buffer (0.05 M Tris-HCl (pH 7.5), 0.1 M

NaCl, 10 mM MgSO4 and 1% (w/V) gelatin) was used for preparing phage suspensions.

Bacterial strains used in this study (Table 3-1) were stored at -80ºC in NB supplemented with

30% glycerol. Bacteriophages (Table 3-2) were stored at 4ºC and protected from light.

42

Standard Bacteriophage Techniques

Purification and storage. Phages were purified by three subsequent single plaque

isolations. Single plaque isolations were carried out by transferring phages from isolated plaques

to a fresh lawn of the host bacterium using sterile toothpicks and then quadrant streaking them

with sterile plastic transfer loops. Following purification the phages were propagated by mass

streaking on fresh lawns of the host. After a 24-h incubation at 28ºC, the phages were eluted by

pouring 5 mL sterilized tap water into the 100 mm×15 mm Petri dishes (Fischer Scientific Co.

LLC, Suwannee, GA) and gently shaking the plates (~20 rpm) for 30 min. The eluate was

centrifuged (10,000 g, 10 min), treated with chloroform or filter-sterilized, depending on the

phage, then quantified as described below, and stored in 2-mL plastic vials at 4°C in complete

darkness. The concentrations of these suspensions were approximately 109 plaque forming units

(PFU) per mL.

Determination of titer. Phage concentrations were determined by dilution-plating-plaque-

count assay on NYA plates without bottom agar as previously described (97). One hundred

microliter aliquots of dilutions of phage suspensions were mixed with 100 µL of concentrated

bacterial suspension in empty Petri dishes and then 12 mL warm (48°C) NYA medium was

poured in each dish. The dishes were gently swirled to evenly distribute the bacteria and the

phages within the medium. After the medium solidified, the plates were transferred to 28°C

incubators and the plaques were counted on the appropriate dilutions after 24 or 48 hours. The

phage concentration was calculated from the plaque number and specific dilution and was

expressed as PFU/mL.

Phage propagation. Phages were recovered from storage, purified by single plaque

isolations and then mass streaked on the freshly prepared lawn of the propagating host. The next

day phages were eluted from the plate, sterilized and enumerated, as described above. The eluate

43

was used for infecting 500 mL actively growing culture of the propagating strain (108 cfu/mL)

grown in NB liquid medium in 1 liter flasks, at 0.1 multiplicity of infection (MOI), (i.e., the

phage concentration at the beginning of the incubation was 107 PFU/mL). After addition of the

phage and 5-min incubation on the bench top, the culture was shaken at 150 rpm at 28ºC for 18

h. The resulting culture was sterilized; phages were enumerated and stored at 4 ºC in the dark

until use. This method yielded phage titers of approximately 1010 PFU/mL.

Phage concentration by high speed centrifugation. High titer phage lysates (~1010

PFU/mL) were concentrated and purified according to methods described by Hans-W.

Ackermann (personal communication). One hundred milliliters of the sterilized lysate was

centrifuged at 10,000 g for 10 min to sediment the bacterial debris. Forty milliliters of the

supernatant was transferred to a new centrifuge tube and centrifuged at 25,000 g for 60 min to

sediment the phage particles. The supernatant was discarded and replaced with 0.1 M ammonium

acetate solution (pH 7.0). Following an additional centrifugation (60 min, 25,000 g) the

supernatant was discarded and the pellet was resuspended in 1.5 mL SM buffer. The final phage

concentration was approximately 1012 PFU/mL.

Evaluation of bacterial sensitivity. Sensitivity of a bacterial strain to phages was

determined based on the ability of the phage to produce plaques on the bacterial lawn, and the

level of sensitivity was evaluated based on efficiency of plating (EOP) on the test strain in

comparison with the propagating host strain of the phage as follows. A phage suspension of

known concentration was plated simultaneously on the test and the host strains and EOP was

calculated as the number of plaques on the test strain divided by the number of plaques on the

host strain. For example, if the phage produced 55 plaques on the host and 36 plaques on the test

strain, then EOP = 36/55 = 0.65. The higher the EOP, the more efficient the phage is in initiating

44

disease of the bacterium. Consequently, the more sensitive the bacterial strain is to the phage. If

the EOP was higher than or equal to 0.1, the test strain was considered sensitive; if the EOP was

less than 0.1, but more than or equal to 0.01 the strain was considered moderately sensitive. If

EOP was less than 0.01, the strain was considered resistant.

Phage Isolation from Diseased Plant Tissue

Bacteriophages were isolated from leaves and fruits with characteristic citrus canker

lesions in Florida and in Argentina. In Florida, phage was isolated from diseased tissue received

from the Florida Department of Agriculture & Consumer Services, Division of Plant Industry,

Gainesville, FL, between May and August 2003, as a part of the Citrus Canker Eradication

Program (53). In Argentina, diseased tissue was collected directly from infected trees located at

the Instituto Nacional de Tecnología Agropecuaria (INTA) research center in Bella Vista,

Corrientes, and from commercial citrus groves in Corrientes province. The tissue samples were

placed in plastic freezer bags or 125 mL flasks and after the addition of 50 mL deionized (DI) or

sterilized tap water were shaken for 20 min. Two milliliters were collected and centrifuged at

10,000 g for 10 min to remove debris. The supernatants were either treated with chloroform or

filter-sterilized and then were checked for the presence of bacteriophages by spotting 20 µL onto

freshly prepared lawns of the indicator bacteria. In Florida three Xac strains were used for

detection: Xac65, a Miami type; Xac60, a Manatee type, and Xac66, a Wellington-type. In

Argentina a battery of 11 Xac strains of diverse origins plus one Xaa strain were used (1622-4,

1528-7-3, 1635, 1660-1, 94-358-1, 1319, 1617, 1604, 1322, 2525, 78-4-3-2-4B and 1311) (Table

3-1). If lysis was observed after 24 h incubation at 28°C, the phage was purified by three

successive single plaque isolations and then propagated and stored, as described above.

45

Phage Typing of 81 Xanthomonas Strains

Twelve phages were used in the phage typing study: α-MME, Φ5536, ΦXacm4-11,

ΦXv3-21, ΦXaacA1, ΦXaacF1, ΦXaacF8, Φcc19-1, Φcc13-2, CP1, CP2 and CP3 (Table 3-2).

The bacterium-phage interactions were scored as sensitive = 2, moderately sensitive = 1 and

resistant = 0. Similarity matrix was calculated from the phage typing scores using the Pearson

correlation, and a dendrogram of relatedness was prepared in which the clustering was achieved

by UPGMA (unweighted pair group method using arithmetic averages) with Bionumerics

software package version 3.0 (Applied Math, Kotrijk, Belgium).

Electron Microscopy

Transmission electron microscopy (TEM) was carried out at the Electron Microscopy

Laboratory of the Microbiology and Cell Science Department of the University of Florida,

Gainesville, FL. The phages were visualized using negative staining protocol with 1% aqueous

uranyl acetate, as follows. A drop of the phage suspension was applied to a 300 mesh formvar-

coated copper grid. After 2 min the liquid was blotted away and the grid was rinsed with DI

water. A 1% uranyl acetate solution was applied to the grid and blotted away after 1 min. The

phages were observed and photographed on a Zeiss EM-10CA transmission electron microscope

operating at 100 kV.

Molecular Techniques

Phage DNA extraction. Fifteen hundred microliters of concentrated and purified

bacteriophage suspensions (~1012 PFU/mL, SM buffer) were treated with nuclease (12.3

unit/sample DNase I (Qiagen Inc., Valencia, CA) and 6.3 units/sample RNase A (Qiagen), 30

min, 37ºC) to digest any contaminating bacterial nucleic acids. Subsequently, samples were

divided into 500 µL subsamples, placed in 1.5 mL microcentrifuge tubes, and 375 µL of a

phenol-chloroform-isoamyl alcohol mixture (25:24:1) was added. The samples were vortexed

46

and centrifuged for 5 min at 10,000 g. The top, aqueous layer was transferred into a new

microfuge tube and the organic layer was discarded. The sample volume was brought up to 500

µL again with the addition of sterile DI water and was subjected to phenol-chloroform-isoamyl

alcohol extraction two more times. Sterilized DI water was added to the sample to bring the

volume up to 500 µL and then the solution was subjected to chloroform-isoamyl alcohol

extraction once by adding 250 µL of a chloroform-isoamyl alcohol mixture (24:1), vortexing and

then centrifuging for 5 min at 10,000 g. The top, aqueous phase was saved and transferred to a

new tube, while the remainder was discarded. The DNA was then precipitated by adding 40 µL

sodium acetate (3 M, pH 5.2) and 800 µL of cold (-4°C) 95% ethanol to the tube, vortexing and

incubating at -80°C for 30 min. The sample was then centrifuged (10,000 g, 20 min) at 4°C and

the supernatant was discarded and replaced with 1 mL of 70% ethanol. After another

centrifugation (10,000 g, 5 min) the supernatant was gently removed by pipetting and the pellet

containing the precipitated DNA was allowed to air dry. The pellet was resuspended in 20 µL

sterile DI water and stored at 4°C.

Digestion of DNA with restriction enzymes. One microliter of the phage DNA was mixed

with 7 µL sterile DI water, 1 µL enzyme buffer and 1 µL restriction enzyme (EcoRI or BamHI)

in a 1.5 mL microcentrifuge tube. The mixture was incubated at 37°C for 90 minutes.

Immediately following the incubation, the entire mixture was subjected to agarose gel

electrophoresis (100) using 1% agarose gel (SeqKem GTG agarose, Cambrex Bio Science

Rockland Inc., Rockland, ME) to separate restriction fragments. EcoRI plus HinDIII digested

Lambda phage DNA (Promega Co., Madison, WI) was used as size marker. The gel was

photographed after 15 min ethidium bromide staining.

47

Results

Bacteriophage Isolations

Bacteriophages were isolated from leaf samples exhibiting citrus canker symptoms in

spring and summer 2003, in Florida. Phages were detected in 37 of the 138 samples (27%).

Bacteriophages were isolated from leaf and fruit tissues with characteristic citrus canker lesions

in December 2003 in Corrientes province, Argentina. Phages were detected in 30 of 56 samples

tested (54%). An attempt was made to isolate phages in Florida in 2006; none were detected in

62 samples.

Classification of Isolated Phages Based on Chloroform Sensitivity, Plaque Morphology, Host Range and Virion Morphology

Chloroform sensitivity, plaque morphology and host range were determined for each

newly isolated phage and for CP1, CP2 and CP3 phages originating in Japan. The phages

isolated in Florida comprised two groups. Thirty one of 37 phages (group A) were resistant to

chloroform, produced clear plaques of 2-3 mm in diameter on strain Xac65 and had identical

host ranges (Table 3-3). Virion morphology of four members of this group was determined: they

were all tailed phages with symmetrical heads and short tails (Figure 3-2, Table 3-4), which is

characteristic of phages in the Podoviridae family. The remaining six phages (group B) were

chloroform sensitive, produced 1 mm turbid plaques on Xac65 and had an identical host range,

which was different than the host range of group A phages. Virion morphology of one phage in

this group, ΦXaacF8, was found to be filamentous, similar to members of Inoviridae (Figure 3-1,

Table 3-4). The 30 phages collected in Argentina were all chloroform resistant, produced 3-4

mm clear plaques on Xac65 and had an identical host range, which only differed from the host

range of group A phages in the reaction on Argentina Xac strain BV42. The Argentina phages

lysed BV42 whereas the Florida group A phages did not (Table 3-3). One Argentina phage,

48

ΦXaacA1, belonged to Podoviridae based on virion morphology (Figure 3-1, Table 3-4). CP1,

CP2 and CP3 were all resistant to chloroform. CP1 did not lyse any strains tested, while CP3

only lysed its propagating strain, XC90. CP2 had the same host range as the Florida phage group

A, produced clear, 3 mm plaques on Xac65 (Table 3-3) and belonged to Podoviridae based on

virion morphology (Figure 3-2, Table 3-4).

Comparison of Florida Group A Phages and CP2 Based on Genome Size and RFLP Profile

Genome sizes of all 31 members of Florida phage group A and CP2 were identical and

were approximately 22 kb (Figure 3-3). However, the group A phages constituted seven groups

based on RFLP profile after BamHI digestion (Figures 3-4, 3-5, Table 3-5). The RFLP profile of

CP2 was different from that of all seven groups (Figure 2-4, panel B). Restriction profiles based

on EcoRI digestion resulted in six groups, five of which matched five BamHI groups; whereas

the sixth one contained two BamHI groups (data not shown). Virion morphology of members of

different BamHI groups, ΦXaacF1 (group 1), ΦXaacF2 (group 7), ΦXaacF3 (group 2) and

ΦXaacF5 (group 3) did not differ considerably (Figure 3-1, Table 3-4).

Phage Typing of 81 Xanthomonas Strains

Sixty nine xanthomonads of worldwide origin causing citrus canker were phage typed on a

battery of 12 phages: 49 Xac strains of pathotype A, three Xac strains of type A*, eight strains of

type B (Xaa) and nine strains of type C (Xaa). Also included were eight strains of X. axonopodis

pv. citrumelo (Xacm), causal agent of citrus bacterial spot, one strain each of X. perforans, X.

vesicatoria and X. euvesicatoria, causal agents of bacterial spot of tomato, and one strain of X.

axonopodis pv. dieffenbachiae, causal agent of anthurium bacterial blight.

In general, citrus canker strains of the same genotype did tend to cluster together (Figure

3-2). On the other hand, Xacm strains showed considerable diversity. Most of Etrog strains were

in one group (G1), clustering together with a Pakistani A strain. Two Etrog strains comprised an

49

independent group (G6), however. Most of the B strains clustered together (G2) and were

sensitive to phage CP3, however strains 6B, 7B (G7) and 8B differed from the rest and were not

sensitive to CP3. Wellington strains clustered together with the A* strains in groups G3 or G5.

The majority of Miami strains and the majority of A strains of various geographical origins

clustered together and were sensitive to at least 8 of 12 phages (G4). However, three Miami

strains and one Brazilian A strain constituted another group (G8) and were resistant to all but two

phages, including ΦXaacF1 and ΦXaacF8 that are naturally present in Florida. All C strains

clustered in group G9 and were resistant to CP1 and CP2. None of the Manatee strains were

sensitive to any phages tested (G11). On the other hand, strain ATCC 49118 (G10), which was

isolated from the 1980s Tampa Bay outbreak was sensitive to 4 phages. Two A strains, isolated

in Florida in 2005 (XC2005-00344-1) and in 2006 (XI2006-00204), have not been classified into

any of the Florida genotypes by genetic methods. One of them had an identical profile to most

Miami types (G4), whereas the other one was resistant to 11 of 12 phages tested. Host ranges of

phages CP2 and ΦXaacA1 were identical and were almost identical to that of ΦXaacF1: the only

difference was ΦXaacF1 was not able to lyse strain XC427 from Thailand.

Discussion

In this study bacteriophages were frequently associated with citrus canker lesions in

Argentina, where the disease is endemic. Also phages were widespread in Florida in 2003, where

citrus canker was under eradication and was located in relatively concise location, mainly in the

Miami-Palm Beach area. It seems that there are near identical population of phages around the

world, as all phages isolated from Argentina and the majority of the ones isolated in Florida were

almost identical to phage CP2 originating in Japan. It is possible that CP2, or a progenitor of CP2

spread around the world with the Xac strains. Interestingly, attempts to isolate phages in Florida

in 2006 were unsuccessful. Since 2003 citrus canker was spread by hurricanes widely within

50

Florida, and it may be that the bacteria managed to escape its pathogens in the process. Also, this

study showed the presence of strains resistant to the phages isolated in Florida and it is possible

that their prevalence resulted in decline of phage.

Results of this study also showed that phages can be used for distinguishing Xac genotypes

present in Florida. Our results generally supported the findings of Cubero et al. (26); however

based on phage typing results, the representative strain associated with the 1980s outbreak and

strains from the 1997 Manatee outbreak were different. This finding contradicts the theory that

the 1997 citrus canker outbreak in Manatee County was caused by strains of the 1980s outbreak

that survived the eradication program. Strain XC2005-00344-1, which is an A strain isolated in

Polk County in 2005 and is unclassified genetically, was resistant to 11 of 12 phages, which

indicates the possibility of a population shift to phage resistance. That in turn could mean that the

naturally occurring phages pose a selection pressure on the pathogen populations, that is, they

have a significant impact on the pathogen’s life. Thus, artificial introduction of exotic

bacteriophages, phage therapy, may well be effective for reducing pathogen populations and

suppressing disease.

The slight differences in host ranges of the Argentina phages, Florida group A phages and

CP2 could be caused by minor differences in phage DNA sequences: the presence or absence of

DNA sequence patterns in the phage genome that are recognized by the host bacterial

restriction-modification system.

51

Table 3-1. Bacterial strains used in this study Strain Origin Genotypea Provided by X. axonopodis pv. citri 1311 Pakistan Canteros, B.I.b

1319 India Canteros, B.I. 1322 Brazil Canteros, B.I. 1604 Argentina Canteros, B.I. 1617 Argentina Canteros, B.I. 1635 Argentina Canteros, B.I. 2525 Argentina Canteros, B.I. 1528-7-3 Argentina Canteros, B.I. 1622-4 Argentina Canteros, B.I. 1660-1 Paraguay Canteros, B.I. 306 Brazil Hartung, J.S.c

94-358-1 Argentina Canteros, B.I. BV38 Argentina Canteros, B.I. BV42 Argentina Canteros, B.I. RF2f Florida Miami This study Xac5 Florida Miami Sun, X.d

Xac6 Florida Miami Sun, X. Xac7 Florida Miami Sun, X. Xac8 Florida Manatee Sun, X. Xac15 Florida Wellington Sun, X. Xac25 Florida Miami Sun, X. Xac30 Florida Miami Sun, X. Xac31 Florida Miami Sun, X. Xac41 Florida Manatee Sun, X. Xac51 Florida Miami Sun, X. Xac65 Florida Miami Sun, X. Xac66 Florida Wellington Sun, X. Xac71 Florida Wellington Sun, X. Xc05-2592 Argentina Canteros, B.I. XC62 Japan Hartung, J.S. XC63 Japan Hartung, J.S. XQ05-1-2 Florida Miami Sun, X. X. axonopodis pv. aurantifolii 78-4-3-2-4B Argentina Canteros, B.I. XC 90 Mexico Hartung, J.S. X. axonopodis pv. citrumelo Xacm 36 Florida This study Xacm 45 Florida This study Xacm 47 Florida This study S4 Florida This study

52

Table 3-1. Continued Strain Origin Genotypea Provided by S4mg Florida This study X. vesicatoria MME Jones, J.B.e

X. perforans ΔopgH:ME-B Jones, J.B.

a Genotype of X. axonopodis pv. citri strains isolated in Florida. b Instituto Nacional de Tecnología Agropecuaria, Bella Vista, Corrientes, Argentina. c United States Department of Agriculture, Agricultural Research Service, Fruit Laboratory, Beltsville, MD. d Florida Department of Agriculture and Consumer Services, Division of Plant Industry, Gainesville, FL. e University of Florida, Plant Pathology Department, Gainesville, FL. f RF2 is a phage-resistant mutant of Xac65. g Chloramphenicol and Rifamycin-resistant mutant of S4.

53

Table 3-2. Bacteriophages used in the study Phage Host Source Origin Provided by CP1 XC62 Japan Hartung, J.S.a

CP2 XC63 Japan Hartung, J.S. CP3 XC90 soil Japan Hartung, J.S. CP31 Xac65 Hartung, J.S. ΦXaacF1 to 37 Xac65 citrus, aerial Florida This study ΦXaacA1 to 30 BV42 citrus, aerial Argentina This study ccΦ5 Xac65 Jackson, L.E.b

ccΦ7 Xac65 Jackson, L.E. ccΦ13-2 Xac65 Jackson, L.E. ccΦ19-1 Xac65 Jackson, L.E. ccΦ24 Xac65 Jackson, L.E. ccΦ25 Xac65 Jackson, L.E. ccΦ28 Xac65 Jackson, L.E. ccΦ35 Xac65 Jackson, L.E. ccΦ36 Xac65 Jackson, L.E. ΦXV3-21 Xac65 Jackson, L.E. α-MME MME lake water Florida Minsavage, G.V.c

Φ5536 Xacm 47 Jackson, L.E. ΦXv3-3-13h ΔopgH:ME-B Jackson, L.E. ΦX44 Xacm47 Jackson, L.E. ΦXacm4-4 Xacm36 citrus, aerial Florida This study ΦXacm4-16 Xacm36 citrus, aerial Florida This study ΦXacm4-11 Xacm36 citrus, aerial Florida This study

a United States Department of Agriculture, Agricultural Research Service, Fruit Laboratory, Beltsville, MD. b OmniLytics, Inc., Salt Lake City, UT. c University of Florida, Plant Pathology Department, Gainesville, FL.

54

Table 3-3. Chloroform sensitivity, plaque morphology and host range of bacteriophages originating in Florida, Argentina and Japan

Chloroform Plaque typeb, Host ranged

sensitivity diameter (mm) BV38 XC90 Xac41 Xac65 Xac15 ΦXaacF1 Ra clear, 2-3 -c - ++ ++ - ΦXaacF2 R clear, 2-3 - - ++ ++ - ΦXaacF3 R clear, 2-3 - - ++ ++ - ΦXaacF4 R clear, 2-3 - - ++ ++ - ΦXaacF5 R clear, 2-3 - - ++ ++ - ΦXaacF6 R clear, 2-3 - - ++ ++ - ΦXaacF7 S turbid, 1 ++ - - ++ ++ ΦXaacF8 S turbid, 1 ++ - - ++ ++ ΦXaacF9 S turbid, 1 ++ - - ++ ++ ΦXaacF10 R clear, 2-3 - - ++ ++ - ΦXaacF11 S turbid, 1 ++ - - ++ ++ ΦXaacF12 R clear, 2-3 - - ++ ++ - ΦXaacF13 S turbid, 1 ++ - - ++ ++ ΦXaacF14 R clear, 2-3 - - ++ ++ - ΦXaacF15 R clear, 2-3 - - ++ ++ - ΦXaacF16 R clear, 2-3 - - ++ ++ - ΦXaacF17 R clear, 2-3 - - ++ ++ - ΦXaacF18 R clear, 2-3 - - ++ ++ - ΦXaacF19 R clear, 2-3 - - ++ ++ - ΦXaacF20 R clear, 2-3 - - ++ ++ - ΦXaacF21 R clear, 2-3 - - ++ ++ - ΦXaacF22 R clear, 2-3 - - ++ ++ - ΦXaacF23 R clear, 2-3 - - + ++ - ΦXaacF24 R clear, 2-3 - - ++ ++ - ΦXaacF25 S turbid, 1 ++ - - ++ ++ ΦXaacF26 R clear, 2-3 - - ++ ++ - ΦXaacF27 R clear, 2-3 - - ++ ++ - ΦXaacF28 R clear, 2-3 - - ++ ++ - ΦXaacF29 R clear, 2-3 - - ++ ++ - ΦXaacF30 R clear, 2-3 - - ++ ++ - ΦXaacF31 R clear, 2-3 - - ++ ++ - ΦXaacF32 R clear, 2-3 - - ++ ++ - ΦXaacF33 R clear, 2-3 - - ++ ++ - ΦXaacF34 R clear, 2-3 - - ++ ++ - ΦXaacF35 R clear, 2-3 - - ++ ++ - ΦXaacF36 R clear, 2-3 - - ++ ++ - ΦXaacF37 R clear, 2-3 - - ++ ++ - ΦXaacA1 R clear, 3-4 ++ - ++ ++ - ΦXaacA2 R clear, 3-4 ++ - ++ ++ - ΦXaacA3 R clear, 3-4 ++ - ++ ++ - ΦXaacA4 R clear, 3-4 ++ - ++ ++ - ΦXaacA5 R clear, 3-4 ++ - ++ ++ - ΦXaacA6 R clear, 3-4 ++ - ++ ++ - ΦXaacA7 R clear, 3-4 ++ - ++ ++ - ΦXaacA8 R clear, 3-4 ++ - ++ ++ - ΦXaacA9 R clear, 3-4 ++ - ++ ++ -

55

Table 3-3. Continued Chloroform Plaque typeb, Host ranged sensitivity diameter (mm) BV38 XC90 Xac41 Xac65 Xac15 ΦXaacA10 R clear, 3-4 ++ - ++ ++ - ΦXaacA11 R clear, 3-4 ++ - ++ ++ - ΦXaacA12 R clear, 3-4 ++ - ++ ++ - ΦXaacA13 R clear, 3-4 ++ - ++ ++ - ΦXaacA14 R clear, 3-4 ++ - ++ ++ - ΦXaacA15 R clear, 3-4 ++ - ++ ++ - ΦXaacA16 R clear, 3-4 ++ - ++ ++ - ΦXaacA17 R clear, 3-4 ++ - ++ ++ - ΦXaacA18 R clear, 3-4 ++ - ++ ++ - ΦXaacA19 R clear, 3-4 ++ - ++ ++ - ΦXaacA20 R clear, 3-4 ++ - ++ ++ - ΦXaacA21 R clear, 3-4 ++ - ++ ++ - ΦXaacA22 R clear, 3-4 ++ - ++ ++ - ΦXaacA23 R clear, 3-4 ++ - ++ ++ - ΦXaacA24 R clear, 3-4 ++ - ++ ++ - ΦXaacA25 R clear, 3-4 ++ - ++ ++ - ΦXaacA26 R clear, 3-4 ++ - ++ ++ - ΦXaacA27 R clear, 3-4 ++ - ++ ++ - ΦXaacA28 R clear, 3-4 ++ - ++ ++ - ΦXaacA29 R clear, 3-4 ++ - ++ ++ - ΦXaacA30 R clear, 3-4 ++ - ++ ++ - CP1 R clear, 4-5 - - - - - CP2 R clear, 3 - - ++ ++ - CP3 R clear, 1 - ++ - - -

Summary 32 phages R clear - - ++ ++ - 30 phages R clear, 3-4 ++ - ++ ++ - 6 phages S turbid, 1 ++ - - ++ ++ 1 phage R clear, 1 - ++ - - - 1 phage R clear, 4-5 - - - - - a S: sensitive, R: resistant. b Plaque morphology was evaluated on the propagating host (Table 3-2) after 36 h. c ++: sensitive, -: resistant. d Strains Xacm36, Xacm45, Xacm47 and 306 were resistant to all phages. Strain BV42 had identical profile to Xac65.

56

A B

C D

E F

G

Figure 3-1. Transmission electron micrographs of representative phages. A) CP2, B) ΦXaacA1, C) ΦXaacF1, D) ΦXaacF2, E) ΦXaacF3, F) ΦXaacF5, G) ΦXaacF8.

57

Table 3-4. Summary of morphological characteristics of phage virions.

Phage Head shape Head

(Lengtha, nm) Tail shape Classificationb

CP2 symmetrical 50.8 short Podoviridae ΦXaacA1 symmetrical 53.3 short Podoviridae ΦXaacF1 symmetrical 58.3 short Podoviridae ΦXaacF2 symmetrical 51.2 short Podoviridae ΦXaacF3 symmetrical 56.6 short Podoviridae ΦXaacF5 symmetrical 62.5 short Podoviridae ΦXaacF8 Inoviridae

a Length measurement represents the average value obtained by measuring three virions per phage. b Classification based on virion morphology as described by Ackermann Fig. 4.1 (1).

58

100

806040200 Phi

Xac

m4-

11

Phi

Xv3

-21

ccP

hi19

-1

Phi

5536

Phi

Xaa

cF8

CP2

Phi

Xaa

cA1

Phi

Xaa

cF1

alph

a-M

ME

CP3

ccP

hi13

-2

CP1

XN2003-00012-7XN2004-00225-2

XS2004-00159

XC2005-00344-1

XN2003-00012-3XN2003-00013-2

X2001-00042XN2003-00011-1

XN2003-00011-2

XN2003-00011-8

XC159LMG 7484

XS1999-00061

JH96E

XC64

JH96-2XCC88B

JH96CXC90

S4

XC2005-00110XC2000-00042

X2000-12884

X2001-00006

X2003-01008

XC165X2001-00032

XC261

XN2003-01036

X2000-12862X2000-00071

X2003-03516SNU6B

X1999-12813

XC392

XCC03-1635XI2000-00075

XI2000-00080

XI2000-00120

XI2006-00204

XS1999-00038XC214

XC100X2003-02912

XC2002-00010XI2000 00194

Xac A EtrogXacm

Xacm

Xac A Miami

Xac A EtrogXac A Etrog

XacmXac A Etrog

Xac A Etrog

Xac A Etrog

Xac AXa diffenbachiae

Xacm

Xaa B

Xaa B

Xaa BXaa B

Xaa BXaa D

Xacm

XacmXacm

Xac A Wellington

Xac A Wellington

Xac A Wellington

Xac A*Xac A Wellington

Xac A*

Xac A Wellington

XacmXac A Miami

Xac A MiamiXac A

Xac A Miami

Xac A

Xac AXac A Miami

Xac A Miami

Xac A Miami

Xac A Miami

Xac A MiamiXac A

Xac AXac A Miami

Xac A MiamiX A Mi i

FloridaFlorida

Florida

Florida

FloridaFlorida

FloridaFlorida

Florida

Florida

Pakistan

Florida

Argentina

Japan

ArgentinaArgentina

ArgentinaMexico

Florida

FloridaFlorida

Florida

Florida

Florida

IndiaFlorida

Saudi Arabia

Florida

FloridaFlorida

FloridaKorea

Florida

Indonesia

ArgentinaFlorida

Florida

Florida

Florida

FloridaTaiwan

YemenFlorida

Florida

Fl id Figure 3-2. Dendogram and phage sensitivity matrix showing relationship amongst

Xanthomonas strains causing citrus canker and citrus bacterial spot based on similarity of sensitivity profile against a battery of 12 phages. Typing phages: 1:ΦXacm4-11, 2:ΦXV3-21, 3:ccΦ19-1, 4:Φ5536, 5:ΦXaacF8, 6:ΦCP2, 7:ΦXaacA1, 8:ΦXaacF1, 9:α-MME, 10:CP3, 11:ccΦ13-2, 12:CP1. Percent similarity values were calculated using the Pearson correlation, and clustering was achieved by UPGMA using Bionumerics software package version 3.0. Black rectangle = sensitive; grey rectangle = moderately sensitive; no rectangle = resistant. Xac=Xanthomonas axonopodis pv. citri; Xaa: Xanthomonas axonopodis pv. aurantifolii; Xacm= Xanthomonas axonopodis pv. citrumelo.

Strain Class Origin

G1

G2

G4

G3

Typing Phages_____ 1 2 3 4 5 6 7 8 9 10 11 12

kjkj

Percent Similarity

59

100

806040200 PhiX

acm

4-11

PhiX

v3-2

1

ccPh

i19-

1

Phi5

536

PhiX

aacF

8

CP

2

PhiX

aacA

1

PhiX

aacF

1

alph

a-M

ME

CP

3

ccPh

i13-

2

CP

1

X2003-02912

XC2002-00010

XI2000-00194

X2003-00008

XC63

XC273

XN2003-01035

XN2003-00012-1XN2003-00012-6

6B

7B

306

X2000-00067XI1999-00112

XI2001-00098

10537C

17C

1C

2C

3C

4C

JH70C

MME

10535C

IAPAR9691/90

91-106

91-118

XC205

8B

XC427

ATCC 49118

XC362

XC62

XS1995-00051

XS1997-00006

XS1997-00018

XS1997-00082

XS2003-00004

Xac A Miami

Xac A Miami

Xac A Miami

Xac A Miami

Xac A

Xac A*

Xac A Wellington

Xac A EtrogXac A Etrog

Xaa B

Xaa B

Xac A

Xac A MiamiXac A Miami

Xac A Miami

Xaa C

Xaa C

Xaa C

Xaa C

Xaa C

Xaa C

Xaa C

X. vesicatoria

Xaa C

Xaa C

X. euvesicatoria

X. perforans

Xac A

Xaa B

Xac A

Xac A

Xac A

Xac A

Xac A Manatee

Xac A Manatee

Xac A Manatee

Xac A Manatee

Xac A Manatee

Florida

Florida

Florida

Florida

Japan

Saudi Arabia

Florida

FloridaFlorida

Argentina

Argentina

Brazil

FloridaFlorida

Florida

Brazil

Brazil

Brazil

Brazil

Brazil

Brazil

Brazil

Brazil

Brazil

Taiwan

Argentina

Thailand

Florida

Australia

Japan

Florida

Florida

Florida

Florida

Florida Figure 3-2. Continued

Strain Class Origin

G4

G5

G6

G7

G8

G9

G11

G10

Typing Phages_____ 1 2 3 4 5 6 7 8 9 10 11 12Percent Similarity

60

Figure 3-3. Undigested bacteriophage DNA. Panel A: Lanes (from left to right) 1:λ, 2:ΦXaacF1,

3:ΦXaacF2, 4:ΦXaacF3, 5:ΦXaacF4, 6:ΦXaacF5, 7:ΦXaacF6, 8:ΦXaacF10, 9:ΦXaacF12, 10:ΦXaacF14, 11:ΦXaacF15, 12:ΦXaacF16, 13:ΦXaacF17, 14:λ. Panel B: 1:λ, 2:CP2, 3:ΦXaacF18, 4:ΦXaacF19, 5:ΦXaacF20, 6:ΦXaacF21, 7:ΦXaacF22, 8:ΦXaacF23, 9:ΦXaacF24, 10:ΦXaacF26, 11:ΦXaacF27, 12:ΦXaacF28, 13:λ, 14: ΦXaacF29. Panel C: 1:λ, 2:ΦXaacF30, 3:ΦXaacF31, 4:ΦXaacF32, 5:ΦXaacF33, 6:ΦXaacF34, 7:ΦXaacF35, 8:ΦXaacF36, 9:ΦXaacF37, 10:λ, 11:ΦXaacF2. Arrows indicate size marker 21,226 bp.

A

C

B

1 2 3 4 5 6 7 8 9 10 11 12 13 14

1 2 3 4 5 6 7 8 9 10 11

1 2 3 4 5 6 7 8 9 10 11 12 13 14

61

Figure 3-4. Bacteriophage DNA digested with BamH I. Panel A: Lanes (from left to right) 1:λ,

2:ΦXaacF1, 3:ΦXaacF2, 4:ΦXaacF3, 5:ΦXaacF4, 6:ΦXaacF5, 7:ΦXaacF6, 8:ΦXaacF10, 9:ΦXaacF12, 10:ΦXaacF14, 11:ΦXaacF15, 12:ΦXaacF16, 13:ΦXaacF17, 14:λ. Panel B: 1:λ, 2:ΦXaacF18, 3:ΦXaacF19, 4:ΦXaacF20, 5:ΦXaacF21, 6:ΦXaacF22, 7:ΦXaacF23, 8:ΦXaacF24, 9:ΦXaacF26, 10:CP2, 11:ΦXaacF27, 12:ΦXaacF28, 13: ΦXaacF29, 14: ΦXaacF30, 15: ΦXaacF31, 16:λ. Panel C: 1:ΦXaacF32, 2:ΦXaacF33, 3:ΦXaacF34, 4:ΦXaacF35, 5:λ, 6:ΦXaacF36, 7:ΦXaacF37, 8:ΦXaacF2. Arrows indicate size marker 21,226 bp.

A

C

B

1 2 3 4 5 6 7 8 9 10 11 12 13 14

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

1 2 3 4 5 6 7 8

62

Table 3-5. Grouping of Florida bacteriophages based on BamH I RLFP profile. Group 1 Group 2 Group 3 Group 4 Group 5 Group 6 Group 7 ΦXaacF1 ΦXaacF3 ΦXaacF5 ΦXaacF27 ΦXaacF31 ΦXaacF34 ΦXaacF2ΦXaacF6 ΦXaacF4 ΦXaacF32 ΦXaacF36 ΦXaacF12 ΦXaacF10 ΦXaacF33 ΦXaacF37 ΦXaacF15 ΦXaacF14 ΦXaacF16 ΦXaacF17 ΦXaacF18 ΦXaacF30 ΦXaacF19 ΦXaacF20 ΦXaacF21 ΦXaacF22 ΦXaacF23 ΦXaacF24 ΦXaacF26 ΦXaacF28

Figure 3-5. Representatives of the 7 RFLP groups based on BamH I. digestion profile. Lanes

(from left to right) 1:λ, 2:ΦXaacF24 (group 1), 3:ΦXaacF30 (group 2), 4:ΦXaacF5 (group 3), 5:ΦXaacF27 (group 4), 6:ΦXaacF31 (group 5), 7:ΦXaacF34 (group 6), 8:ΦXaacF2 (group 7), 9:λ. Arrow indicates size marker 21,226 bp.

1 2 3 4 5 6 7 8 9

63

CHAPTER 4 CONTROL OF CITRUS CANKER AND CITRUS BACTERIAL SPOT WITH

BACTERIOPHAGES

Introduction

The traditional control strategies for managing bacterial plant diseases have utilized

chemical control with reliance on antibiotics and copper bactericides (3,81,119). However, the

use of antibiotics has been largely discontinued in agriculture due to the emergence of antibiotic-

resistant bacteria in the field (119) and due to concerns that phytobacteria can harbor and transfer

antibiotic resistance to human pathogens. The extended heavy use of copper compounds led to

soil contamination (72) and the emergence of copper-tolerant strains amongst phytobacteria (81)

led to reduced efficacy of copper treatments. High doses of ionic copper also can cause

phytotoxicity (85,109). These factors have prompted a search for identifying novel, effective

means for managing bacterial diseases of plants. Additionally, concerns about food safety and

environmental protection, and the growing organic production necessitate development of safer,

more specific and environment-friendly pesticides (96).

Bacteriophages have been used effectively for controlling several Xanthomonas diseases.

Severity of bacterial spot of peach, caused by X. campestris pv. pruni, has been reduced both on

foliage (23) and on fruits (99) with phage treatment. Phage applications decreased geranium

bacterial blight severity, caused by X. campestris pv. pelargonii (37). Tomato bacterial spot

disease severity was reduced and fruit yield was increased by phage applications (10,38,89,90).

Phage-based integrated management of tomato bacterial spot is recommended to tomato growers

in Florida (85), and bacteriophage mixtures against the pathogen are commercially available

(Agriphage from OmniLytics Inc., Salt Lake City, UT, EPA Registration # 67986-1). However,

phage therapy has not been evaluated for the control of citrus diseases.

64

Our objective was to evaluate phage therapy for treatment of citrus canker. Additionally,

we wished to investigate the effect of protective formulations on efficacy of bacteriophage

treatment and also to use phages in combination with chemical bactericides as a part of an

integrated approach. Skim milk-based protective formulations have been used to increase phage

persistence on plant foliage and contributed to enhanced control of tomato bacterial spot (10).

Ionic copper is toxic to all cells, since it interacts with sulfhydryl (-SH) groups of amino acids

and causes denaturation of proteins ((3), p.345). Ionic copper reduces bacteriophage populations

(11, 66); however, it only persists on plant foliage a short time and does not affect bacteriophage

survival if applied three or more days in advance of phage application (66).

Citrus canker until recently has been under eradication in Florida, and field trials in groves

could not be carried out in the state. Therefore, we conducted disease control trials (i) in the

greenhouse of the citrus canker quarantine facility of the Florida Department of Agriculture &

Consumer Services, Division of Plant Industry (DPI) and (ii) in Argentina, where disease canker

is endemic in cooperation with researchers of the Instituto Nacional de Tecnología Agropecuaria

(INTA). Additionally, we carried out field trials in Florida for suppressing citrus bacterial spot

(CBS), another Xanthomonas disease of citrus (102). CBS, incited by X. axonopodis pv.

citrumelo (Xacm), is similar to citrus canker, as both diseases cause lesions on citrus leaves and

stems, infect young tissue, spread by wind driven rain and their severity is greatly increased by

Asian leafminer (Phyllocnistis citrella) tunneling activity. CBS differs from citrus canker in that

it does not affect full grown plants, causes flat or sunken lesions (Figure 1-8) instead of the

corky, raised ones of canker (Figure 1-1) and is generally less aggressive than citrus canker

(106). Additionally, CBS is not regulated allowing the implementation of field trials.

65

Materials and Methods

Bacterial Strains and Bacteriophages

Bacterial strains (Table 3-2) were grown on nutrient agar (NA) medium (0.8% (wt/V)

Nutrient Broth (NB) (BBL, Becton Dickinson and Co., Cockeysville, MD) and 1.5% (wt/V)

Bacto Agar (Difco, Becton Dickinson and Co., Sparks, MD)) at 28°C. For preparation of

bacterial suspensions 24 h cultures were suspended in sterile tap water, their concentration was

adjusted to 5×108 cfu/mL (A600=0.3), and then were diluted appropriately. Bacterial inoculations

were carried out by misting the bacterial suspension on the plant foliage with a hand-held

sprayer. For detection and propagation of bacteriophages (Table 3-1) either semisolid, soft

nutrient agar yeast extract medium (NYA), (0.8% Nutrient Broth, 0.6% Bacto Agar and 0.2%

Yeast Extract (Difco, Becton Dickinson and Co., Sparks, MD)), or liquid nutrient broth medium

was used. Sterilized tap water or SM buffer (0.05 M Tris-HCl (pH 7.5), 0.1 M NaCl, 10 mM

MgSO4 and 1% (w/V) gelatin) was used for preparing phage suspensions. Bacterial strains used

in this study (Table 3-2) were stored at -80ºC in NB supplemented with 30% glycerol.

Bacteriophages were stored at 4ºC and protected from light.

Plant Material and Cultural Conditions in the Greenhouse

Duncan grapefruit plants were grown from seed in 15 cm plastic pots in Terra-Lite®

agricultural mix (Scott Sierra Horticultural Products Co., Marysville, OH) and fertilized as

needed using Osmocote® Outdoor & Indoor Smart Release® Plant Food (Scott Sierra

Horticultural Products Co.) controlled release fertilizer (NPK 19-6-12). The plants were kept in

the glasshouse of the citrus canker quarantine facility of the Division of Plant Industry in

Gainesville, Florida at temperature ranging from 25-30ºC.

66

Standard Bacteriophage Techniques

Determination of titer. Phage concentrations were determined by the dilution-plating-

plaque count assay on NYA plates without bottom agar as previously described (97).

Propagation. Phages were recovered from storage, purified by single plaque isolations and

then mass streaked on the freshly prepared lawn of the propagating host. The following day the

phages were eluted from the plate, sterilized and enumerated. The eluate was used for infecting

500 mL of actively growing culture of the propagating strain (108 cfu/mL) grown in NB liquid

medium in 1 liter flasks, at 0.1 multiplicity of infection (MOI). After addition of the phage and 5

min incubation on the bench top, the culture was shaken at 150 rpm at 28ºC for 18 h. Then the

culture was sterilized, enumerated and stored at 4 ºC in the dark until use. This method yielded

phage titers of approximately 1010 PFU/mL.

Phage concentration by high speed centrifugation. High titer phage lysates (~1010

PFU/mL) were concentrated and purified according to methods described by Hans-W.

Ackermann (personal communication). One hundred milliliters of the sterilized lysate was

centrifuged at 10,000 g for 10 min to sediment the bacterial debris. Forty milliliters of the

supernatant was transferred to a new centrifuge tube and centrifuged at 25,000 g for 60 min to

sediment the phage particles. The supernatant was discarded and replaced with 0.1M ammonium

acetate solution (pH 7.0). Following an additional centrifugation (60 min, 25,000 g) the

supernatant was discarded and the pellet was resuspended in 1.5 mL of SM buffer. The final

phage concentration was approximately 1012 PFU/mL.

Disease Assessment and Data Analysis

The ratio of diseased leaf surface area was estimated using the Horsfall-Barratt (HB) scale

(12): 1=0%, 2=0-3%, 3=3-6%, 4=6-12%, 5=12-25%, 6=25-50%, 7=50-75%, 8=75-87%, 9=87-

94%, 10=94-97%, 11=97-100%, 12=100%. The HB values were converted to estimated mean

67

percentages by using the Elanco Conversion Tables for Barratt-Horsfall Rating Numbers

(ELANCO Products Co., Indianapolis, IN), as described in Appendix A. If disease was assessed

more than once, the area under the disease progress curve (AUDPC) was computed by

trapezoidal integration of disease percentage values taken at different timepoints using the

formula proposed by Shaner and Finnley (104), as described in Appendix A.

Analysis of variance (ANOVA) and subsequent separation of sample means by the Waller-

Duncan K-ratio t Test (for balanced data) or by least square means method (for unbalanced data)

was carried out using SAS System for Windows program release 8.02 (Cary, NC).

Greenhouse Citrus Canker Control Trials

Control with phage mixture applied with or without skim milk formulation. Duncan

grapefruit plants were heavily pruned and fertilized to induce a new flush of growth that is highly

susceptible to citrus canker infection. Approximately three weeks later the emerging-uniform-

new-foliage was treated with a equal mixture of phages CP2, CP31, ccΦ7 and ccΦ13-2 (test 1 -

5×109 PFU/mL, test 2 - 1×109 PFU/mL, test 3 - 1×108 PFU/mL of each phage) in the evening

and the plants were covered with white plastic bags. Phages were applied with or without skim

milk formulation (0.75% (wt/V) non-fat dry milk powder). The goal of placing in plastic bags

was to maintain high relative humidity, simulate dew conditions and also to contribute to the

opening of the stomata that increases the penetration of bacteria into the leaf tissues. On the next

morning the bags were removed from the plants, and the plants were inoculated with Xac65 (test

1 - 1×108 cfu/mL, tests 2 and 3 - 5×106 cfu/mL). After inoculation the plants were placed inside

the bags for an additional 24 h of high moisture. After removal from the bags and after the

foliage was allowed to dry the plants were arranged in a completely randomized pattern on a

greenhouse bench. The disease was assessed 3-4 weeks after inoculation. Four plants were used

per treatment.

68

Control of disease caused by phage-sensitive strain and its phage-resistant mutant. A

mutant of Xac65, designated RF2, which was selected for resistance to phage ΦXaacF2, was

found also to be resistant to CP2, CP31, ccΦ7 and ccΦ13-2. In three tests grapefruit plants were

inoculated either with Xac65 or RF2, at 5×106 cfu/mL 12 h after treatment with the mixture of

CP2, CP31, ccΦ7 and ccΦ13-2, each at 1×108 PFU/mL. These tests were carried out similarly to

the previous ones.

Citrus Canker Nursery Trials in Argentina

The trials were conducted at the experimental nursery of the Instituto Nacional de

Tecnología Agropecuaria (INTA) research center in Bella Vista, Corrientes (Figure 4-1) from

January to April 2006 in cooperation with scientists of INTA. The experiment was set up in a

split-plot design: there were two main treatments, phage and no phage, and within them there

were two sub-treatments, copper-mancozeb and no copper-mancozeb. The sub-treatments were

randomly replicated four times within the main treatments. The replications were 5-m long plots.

The main plots were replicated twice. The plants were inoculated with a copper-resistant Xac

strain, Xc05-2592 in January 2006 and then re-inoculated in February. The phage treatment

(ΦXaacA1, ~106 PFU/mL) was mixed with a protective formulation (0.75% (wt/V) non-fat dry

milk powder) and spray-applied twice weekly in. Copper-mancozeb (3 g/L Caurifix WG (BASF)

(a.i. 84% copper oxychloride) plus 2 g/L Dithane M80 (a.i. 80% mancozeb)) was also spray-

applied twice weekly. Disease assessment was carried out once, after 12 weeks of phage and

copper-mancozeb applications. All diseased leaves were collected and the average number of

lesions per leaf was determined. This experiment was carried out at two sites at the nursery

simultaneously.

69

Citrus Bacterial Spot Disease Control Trials at Dilley and Son Nursery in Avon Park, Florida

The experiment was carried out at Roland L. Dilley & Son, Inc., a citrus nursery in Avon

Park, FL (Figure 4-2), which had a long history of CBS, caused by X. axonopodis pv. citrumelo.

The cultural methods and the high plant density in the nursery resulted in active plant growth,

long leaf-wetness periods and easy pathogen spread, all contributing to CBS development. Three

test sites were established in the nursery: two were set in a section containing Valencia orange

scions on Volkamer lemon (Citrus volkameriana) rootstock (Figure 4-2), and the third one was

set in a section of Ray Ruby grapefruit scions on Cleopatra mandarin (C. reticulata Blanco)

rootstocks. In each test site a row was divided into six 6-m-long plots, spaced 3 m apart, with

three receiving phage-treatment (phage), and three not (control). The phage-treated plots were

sprayed twice-weekly at dawn with a mixture of three bacteriophages (ΦXacm2004-4,

ΦXacm2004-16 (both isolated from the nursery), and ΦX44 (OmniLytics Inc., Salt Lake City,

UT), 2.4 × 108 PFU/mL each) blended in a suspension of 0.75% (wt/V) non-fat dry milk powder

using a non-CO2, backpack sprayer. The treatment was applied from August 9 until November

11, 2004. The high winds associated with the hurricanes caused a great deal of physical damage

to the plants opening up entry for pathogen ingress. Disease severity, caused by natural

inoculum, was evaluated five times during the experiment (biweekly between September 16 and

November 11).

Citrus Bacterial Spot Disease Control Trials at the Plant Science Unit of the University of Florida in Citra, Florida.

Two trials were carried out during summer and fall of 2006 at the Plant Science Unit of the

University of Florida in Citra, Florida (Figure 4-3). Swingle citrumelo plants, donated by Dilley

and Son, Inc. (Avon Park, FL) were transplanted in April 2005 and were fertilized monthly with

Osmocote® Outdoor & Indoor Smart Release® Plant Food (Scott Sierra Horticultural Products

70

Co., Marysville, OH) and irrigated as needed using an overhead irrigation system. The

experiment was set up in a split-plot design: there were two main treatments, phage and no

phage, and within them there were two sub-treatments, copper-mancozeb and no treatment. The

sub-treatments were randomly replicated five times within the main treatments. Each replicate

was one plot consisting of nine plants spaced 25 cm apart. The plots were separated by 2 m. The

plants were pruned and heavily fertilized to achieve uniform emerging young foliage before

inoculation with S4m, an antibiotic resistant mutant of a Xacm strain, S4 (Table 3-2). Inoculation

was carried out by spraying plants with the suspension of the pathogen (107 cfu/mL) amended

with 0.025% Silwet L-77 (Loveland Industries, Co., Greely, CO), a wetting agent, for increased

inoculum penetration. In the first trial every plant was inoculated, and pruned back after

developing disease symptoms, and then the emerging foliage rated for disease. This was done to

simulate the situation similar to that of the Dilley’s nursery, where the pathogen is present in the

nursery but is not directly inoculated on the plants. Since the disease pressure was extremely low

using this method, in the second trial the middle plant of every plot was inoculated and then the

disease was assessed on the inoculated foliage.

Phage and copper-mancozeb applications started before inoculations and continued until

the last rating was taken. The phage treatment was applied twice weekly in the evenings. It

consisted of a mixture of three phages, (ccΦ13-2, α-MME and Φ5536, 1×108 PFU/mL each)

applied with skim milk formulation (0.75% (wt/V) non-fat dry milk powder) using a non-CO2,

backpack sprayer. The copper-mancozeb treatment (2.5 g/L Manzate 75DF (Griffin Co.,

Valdosta, GA) (a.i. 75% mancozeb) and 3.6 g/L Kocide 2000 ( E. I. du Pont de Nemours & Co.)

(a.i. 53.8% copper hydroxide)) was applied once a week, 2 days apart from the phage

71

applications. Disease severity was assessed weekly using the Horsfall-Barratt scale, as described

above.

Results

Effect of Bacteriophage Application and Use of Protective Skim Milk Formulation on Citrus Canker Disease Development in the Greenhouse

In order to evaluate whether the presence of high populations of bacteriophages on the

grapefruit foliage suppresses citrus canker disease development, a mixture of four bacteriophages

(CP2, CP31, ccΦ7 and ccΦ13-2) was applied on grapefruit plants 12 h before inoculation with X.

axonopodis pv. citri (Xac) strain, Xac65. The phage mixture was applied with or without a skim

milk formulation. Three tests were carried out and in two of them there were significant

differences between treatments (Table 4-1). In both of those tests the application of phage

mixture without skim milk significantly reduced disease severity (Table 4-1). No phages could

be recovered two days after application from plants if they were applied without skim milk

formulation, whereas if applied with the formulation phage populations ranged from 104 to 107

PFU/mL. However, interestingly, formulated phage treatment did not decrease disease severity

(Table 4-1).

Effect of Phage Application on Disease Severity Incited by a Phage-Sensitive Xac Strain and Its Phage-Resistant Mutant

In order to reduce possible cross-treatment interference caused by phage spread between

plants, the effect of phage treatment on the development of citrus canker incited by a phage

sensitive strain, Xac65, and its phage resistant mutant, RF2 were compared in three experiments.

The disease caused by the two strains was not significantly different (Table 4-2), indicating that

the mutation to phage resistance did not affect pathogenicity of RF2. Phage application

significantly reduced the disease severity incited by Xac65 in all three experiments (Table 4-2).

Interestingly, the phage treatment also decreased disease severity incited by the RF2 in of one of

72

three tests (Table 4-2). It is likely that this resulted from an error during the execution of the

experiment: the plants may have been inoculated with Xac65 instead of RF2, or the phages may

have been contaminated with a phage that was able to lyse RF2.

Effect of Phage and Copper-Mancozeb Applications on Citrus Canker Disease Development in a Citrus Nursery

The effect of bacteriophage treatment, copper-mancozeb treatment, and their combination

was evaluated for reducing citrus canker disease development under field conditions in an

experimental citrus nursery in Argentina (Figure 4-1). Nursery plants were inoculated with a

copper tolerant Xac strain and were treated (i) twice weekly with a suspension of phage

ΦXaacA1; or (ii) twice weekly with a copper-mancozeb solution, or (iii) both with phage and

copper-mancozeb. Phage application alone significantly reduced citrus canker disease severity,

but it was not as effective as the copper-mancozeb treatment (Table 4-3). The combination of

phage and copper-mancozeb treatments did not result in increased control (Table 4-3); to the

contrary it reduced the efficacy of copper treatment in one of two trials.

Effect of Bacteriophage Treatment on Citrus Bacterial Spot Disease Development in a Commercial Citrus Nursery

Disease control trials were set up in the summer and fall of 2004 in a commercial citrus

nursery in Florida to evaluate the effect of preventative bacteriophage applications on CBS

disease development incited by naturally occurring Xacm populations. A mixture of three phages

was applied twice weekly in skim milk formulation at three sites. Two sites were established in

sections of Valencia orange, which is moderately sensitive to CBS, whereas the third site was in

a grapefruit section, which is highly susceptible to the disease. Phage treatment contributed to a

significant disease reduction in both Valencia sites (Table 4-4). Some disease suppression also

occurred at the grapefruit site, but it was not significant (p = 0.1585) (Table 4-4).

73

Effect of Bacteriophage and Copper-Mancozeb Treatments on Citrus Bacterial Spot Disease Development in an Experimental Nursery

The effects of bacteriophage treatment, copper-mancozeb treatment and their combination

were evaluated for reducing CBS disease development in an experimental citrus nursery. Phage

treatment contributed to significant disease reduction (Table 4-5); however, it provided less

control than the copper-mancozeb in one of two trials. The combination of phage and copper

treatments did not result in increased disease control, but did not result in reduced control, either.

Discussion

Phage treatments proved to be effective means of disease reduction both under greenhouse

and field conditions, both for citrus canker and citrus bacterial spot. Their efficacy would likely

be further increased if the phage mixtures were carefully designed and consisted of higher

number of phages with distinct cell receptor specificities. In these studies the number of phages

included in the mixture ranged from one to four and their receptor specificity was not evaluated

at all.

The greenhouse trials provided a good example of what could happen if phages with

similar receptor specificity are used. Strain RF2 became resistant to all four phages used in the

greenhouse trials with presumably a single mutation. It indicates that those four phages had the

same receptor specificity. Mutation to resistance by RF2 resulted in abolishment of disease

control by the phage treatment without loss in aggressiveness of the pathogen. This phenomenon

also underscores the need for constant stringent monitoring of pathogen populations in a phage-

based disease management program.

Skim milk formulation, despite its effect of increasing phage longevity on plant foliage,

totally abolished disease control achieved by the phages. It is possible that skim milk, which acts

as a wetting agent and breaks down surface tension, helped in the pathogen ingress. It could have

74

also served as a carbon source for the bacterium. There is clearly a need for further research on

developing protective formulations that give protection without the adverse effects.

The twice weekly spray schedule used in this study might be too frequent for a practical

application in a commercial grove setting. The relation of application frequency and control

efficacy is an economically important factor and should be investigated before considering

commercialization.

The combination of phage treatment and copper-mancozeb did not prove to be

advantageous. It was expected that the presence of ionic copper may reduce phage efficacy as it

generally inactivates proteins. However, phage application also reduced copper efficacy in these

trials. It is possible that phage virions and the proteins in the skim milk formulation tied up a

significant part of free copper ions thus reducing the copper concentration available for

eliminating the bacterial pathogen. The phage sprays also could have washed off a considerable

amont of copper from the plant foliage. If phages are to be used as a part of an integrated pest

management program, they likely will perform better together with other biological control

methods or with chemical methods of different modes of action such as systemic acquired

resistance inducers.

75

Figure 4-1. Experimental citrus nursery, the location of the 2006 citrus canker trials, at the

INTA research station in Bella Vista, Corrientes, Argentina.

Figure 4-2. Disease control plots in Dilley and Son Nursery, Avon Park, Florida.

76

Figure 4-3. Field plots at the Plant Science Research and Education Unit of the University of

Florida in Citra, Florida.

77

Table 4-1. Effect of bacteriophage treatment and use of skim milk formulation on citrus canker disease development incited by Xanthomonas axonopodis pv. citri strain Xac65 on grapefruit plants in the greenhouse

Percent disease severity Treatment Phagey Formulated phage Water pz

Test 1x 53b 75a 72a 0.0116 Test 2 6b 12ab 16a 0.1967 Test 3 13b 23a 24a 0.0211

x Inoculum concentration was 1×108 cfu/mL in test 1 and 5×106 cfu/mL in tests 2 and 3. y Means within the same row followed by the same letter are not significantly different according to the Waller-Duncan K-ratio t Test at p=0.05 level. z p = Probability that there are no differences in treatment means according to analysis of variance. Table 4-2. Effect of bacteriophage treatment on citrus canker disease development incited by

phage sensitive Xanthomonas axonopodis pv. citri strain Xac65 and its phage-resistant mutant, RF2 on grapefruit plants in the greenhouse

Percent disease severity Pathogen Xac65 RF2 Treatment Phagey Water Phage Water pz

Test 1 16b 55a 47a 51a 0.0454 Test 2 17b 39a 8b 48a 0.0052 Test 3 5b 42a 46a 36a 0.0262

y Means within the same row followed by the same letter are not significantly different according to the Waller-Duncan K-ratio t Test at p=0.05 level. z p = Probability that there are no differences in treatment means according to analysis of variance. Table 4-3. Effect of phage and copper-mancozeb application on citrus canker disease

development, incited by Xanthomonas axonopodis pv. citri strain Xc05-2592 in nursery trial in Bella Vista, Corrientes, Argentina

Disease intensityz Treatment UTCy,x Phage Cu-Mzv Phage+Cu-Mz

pw

Test 1 2.6a 2.0b 1.6c 1.8bc <0.0001 Test 2 3.1a 2.1b 1.6c 2.0b <0.0001

v Cu-Mz = copper-mancozeb. w p = Probability that there are no differences in treatment means according to analysis of variance. x UTC = untreated control y Means within the same row followed by the same letter are not significantly different based on least square means differences at p=0.05 level. z Disease intensity = average number of lesions on diseased leaves.

78

Table 4-4. Effect of bacteriophage treatment on citrus bacterial spot disease development incited

by naturally occurring Xanthomonas axonopodis pv. citrumelo strains at Dilley and Son Nursery in Avon Park, Florida in 2004, as measured by area under the disease progress curve (AUDPC)

AUDPC Phage Control pz

Site 1y 173 331 0.0023 Site 2 169 261 0.0357 Site 3 128 154 0.1585

y Sites 1 and 2 consisted of Valencia orange, site 3 of grapefruit. z Probability of equality of sample means according to t test.

Table 4-5. Effect of bacteriophage and copper-mancozeb treatment on citrus bacterial spot disease development incited by Xanthomonas axonopodis pv. citrumelo strain S4m at the UF Plant Science Unit in Citra, Florida in 2006, as measured by area under the disease progress curve (AUDPC)

AUDPC Treatment UTCy,z Phage Cu-Mzw Phage+Cu-Mz

px

Test 1 0.53a 0.20b 0.08b 0.13b 0.0115 Test 2 24.3a 14.9b 4.0c 2.0c <0.0001

w Cu-Mz = copper-mancozeb. x p = Probability that there are no differences in treatment means according to analysis of variance. y UTC = untreated control. z Means within the same row followed by the same letter are not significantly different based on least square means differences at p=0.05 level.

79

CHAPTER 5 INTERACTION OF BACTERIOPAHGES AND THE HOST BACTERIA ON THE

PHYLLOPLANE

Introduction

Bacteriophages, when used as biological control agents for foliar plant diseases interact

with the target organism on the leaf surface, the phylloplane. The phylloplane is a constantly

changing environment; there are changes in temperature, sunlight irradiation, leaf wetness,

relative humidity, osmotic pressure, pH, microbial flora, and in the case of agricultural plants,

also chemical compounds (40). These factors are harmful to bacteriophages to varying extents.

Sunlight irradiation, especially in the UV A and B range, is highly detrimental to

microorganisms in general (4,65) and is mainly responsible for eliminating bacteriophages

within hours of application (66). Different temperatures and levels of relative humidity have

different effects on phage longevity. Water is necessary for disseminating phage virions and it

provides the medium for phage-bacterium interactions. It also contributes to virion stability,

since many phages are sensitive to desiccation. Microbial activity and enzymes degrade phages.

Modern intensive agriculture relies heavily on application of pesticides for control of plant

diseases and insect pests. Most chemicals tested did not affect bacteriophage persistence (9,

137); however, copper compounds are clearly detrimental (9,66).

Materials that increase the longevity of viruses in the field have been identified

(7,8,15,16,63,64,65). The use of such protective materials also increased disease control

achieved with bacteriophages (10). Timing phage treatments to apply in the evenings and early

mornings when sunlight UV irradiation is minimal, also increased efficacy (10). One important

factor that remains unexplored is the phages themselves. Bacteriophages used in previous studies

and also for commercial phage products available for agricultural use were selected strictly based

on in vitro tests (10, 89, 90, 99) without any evaluation in the target environment. It is not known

80

whether the different phages included in the mixture persist similarly on the plant surface,

whether the ratio of active ingredients in the phage treatments changes in the target environment

after application. It is not known if different phages in the phage mixture contribute equally to

disease control. It is not known if there are such characteristics that make a phage more suitable

to disease control on a leaf surface than others. The goal was to try to shed some light in this

darkness. The two objectives were (i) to monitor the fate of bacteriophages, individual

constituents of a phage mixture used for disease control in a real field situation after being

applied to the leaf surface, and (ii) to test if the ability to multiply on the leaf surface influences

phage disease control efficacy and makes a detectable difference on disease control.

Material and Methods

Standard Bacteriophage Techniques

Determination of titer. Phage concentrations were determined by the dilution-plating-

plaque-count assay on nutrient agar yeast extract plates without bottom agar, as previously

described (97).

Phage propagation. Phages were recovered from storage, purified by single plaque

isolations and then mass streaked on the freshly prepared lawn of the propagating host. After

overnight incubation the phages were eluted from the plate, sterilized and enumerated, as

described above. The eluate was used for infecting 500 mL of actively growing culture of the

propagating strain (108 cfu/mL) grown in NB liquid medium in 1 liter flasks at 0.1 multiplicity of

infection (MOI). After addition of the phage and 5-min incubation on the bench top, the culture

was shaken at 150 rpm at 28ºC for 18 h. Then the culture was sterilized, enumerated and stored

at 4ºC in the dark until use. This method yielded phage titers of approximately 1010 PFU/mL.

Phage concentration by high speed centrifugation. High titer phage lysates (~1010

PFU/mL) were concentrated and purified according to methods described by Hans-W.

81

Ackermann (personal communication). One hundred milliliters of the sterilized lysate was

centrifuged at 10,000 g for 10 min to sediment the bacterial debris. Forty milliliters of the

supernatant was transferred to a new centrifuge tube and centrifuged at 25,000 g for 60 min to

sediment the phage particles. The supernatant was discarded and replaced with 0.1 M ammonium

acetate solution (pH 7.0). Following an additional centrifugation step (60 min, 25,000 g) the

supernatant was discarded and the pellet was resuspended in 1.5 mL SM buffer. The final phage

concentration was approximately 1012 PFU/mL.

Changes in Xanthomonas axonopodis pv. citrumelo Phage Populations on the Field

The experiment was conducted at the Plant Science Unit of the University of Florida in

Citra, Florida. A mixture of three bacteriophages was sprayed onto Swingle citrumelo plants at 8

PM on July 18, 2006. The phages used were ccΦ13-2, α-MME and Φ5536, all adjusted to 1×108

PFU/mL as determined on their common propagating bacterium, Xanthomonas axonopodis pv.

citrumelo (Xacm) strain, S4m. The mixture was applied in skim milk formulation (0.75% (wt/V)

non-fat dry milk powder) using a non-CO2, Solo backpack sprayer. The phage populations were

monitored in the evening and the next morning as follows. Leaf samples were taken at 8:15 PM,

7:45 AM, 9:00 AM, 10:00 AM and 11:00 AM. For each sample three trifoliate leaves,

originating from a singe plant, were removed and placed into a plastic freezer bag (Hefty OneZip

quart size, Pactiv Co., Lake Forest, IL). At each timepoint four samples were collected from

plants within the same plot. The samples were placed on ice in a portable plastic cooler and

immediately carried to the laboratory. The bags were then weighed and 50 mL of sterilized tap

water was poured into each bag. The bags were shaken on a wrist action shaker (Burrel Co.,

Oakland, CA) for 20 min and then 1 mL of leaf-wash was removed from each bag and

transferred into 1.5 mL microcentrifuge tubes containing 20 µL of chloroform. The tubes were

stored at 4ºC until all samples were taken and processed and then transported to the Gainesville

82

laboratory for determination of phage titer. Phages were enumerated by dilution plating and

plaque counts on three bacterial strains: MME (X. vesicatoria), Xacm45 (Xanthomonas

axonopodis pv. citri) and S4m (Xacm). MME selectively detected phage α-MME, Xacm45

selectively detected phage Φ5536, and S4m detected all three phages. Titers of phages α-MME

and Φ5536 were calculated from the plaque number, dilution information and leaf weight and

were expressed as PFU/g leaf tissue. Titer of phage ccΦ13-2 was determined indirectly, as

described. The efficacy of plating of phage α-MME on strain MME versus strain S4m was

determined concurrently to phage titer determinations by plating known concentration of phage

α-MME on both strains (5 replication each) and dividing the average number of plaques

produced on MME by the average number of plaques produced on S4m. EOP of phage Φ5536

on Xacm45 versus S4m was determined similarly. The EOP values were multiplied with the

phage concentrations determined on the appropriate selective host to calculate the phage

concentration on S4m for each sample. The concentration of ccΦ13-2 in each sample was

determined by the following equation: ccΦ13-2 concentration = total phage concentration (as

determined on S4m) – α-MME concentration (calculated) - Φ5536 concentrated (calculated).

Finally, the data were converted by the y=log10(x+1) function and were graphed. The sunlight

irradiation data were downloaded from the Florida Automated Weather Network website

(http://fawn.ifas.ufl.edu).

Interaction of Xanthomonas perforans and Its Bacteriophage on the Tomato Foliage in the Greenhouse

Population study. Six young tomato plants, in 3-4 leaf stage, were dipped in a suspension

of X. perforans strain, ΔopgH:ME-B (105 cfu/mL), amended with 0.025% Silwet L-77 (Loveland

Industries, Co., Greely, CO). Five days later three of the plants were sprayed with sterilized tap

water and three with a suspension of bacteriophage ΦXv3-3-13h (108 PFU/mL). The plants were

83

placed in clear polyethylene bags for 1 day after phage application in a growth chamber (28ºC).

Before removing the bags, the plants were transferred to a greenhouse where they remained

throughout the experiment in a completely randomized arrangement. Phage populations were

monitored daily for a week by removing three leaflets and recovering the phages from the

leaflets.

Disease control study. Phage ΦXv3-3-13h was also used for controlling tomato bacterial

spot severity incited by ΔopgH:ME-B. Six young tomato plants were inoculated with strain

ΔopgH:ME-B; and later three of them were sprayed with ΦXv3-3-13h and three were sprayed

with sterilized tap water. Methods of bacterial inoculation and phage application were the same

as for the population studies. The ratio of diseased leaf tissue was estimated on the bottom three

fully expanded leaves of each plant twice, 19 and 22 days after inoculation using the Horsfall-

Barratt (HB) scale (12): 1=0%, 2=0-3%, 3=3-6%, 4=6-12%, 5=12-25%, 6=25-50%, 7=50-75%,

8=75-87%, 9=87-94%, 10=94-97%, 11=97-100%, 12=100%. The HB values were converted to

estimated mean percentages by using the Elanco Conversion Tables for Barratt-Horsfall Rating

Numbers (ELANCO Products Co., Indianapolis, IN), as described in Appendix A. The area

under the disease progress curve (AUDPC) was computed by trapezoidal integration of disease

percentage values taken in different timepoints using the formula proposed by Shaner and

Finnley (104), as described in Appendix A. The sample means of the phage-treated and non-

treated plants were compared by t-test using SAS System for Windows program release 8.02

(Cary, NC).

Interaction of Xanthomonas axonopodis pv. citri and Its Bacteriophages on Grapefruit Foliage in the Greenhouse

Disease control studies. Two experiments were carried out at the greenhouse of the

Department of Agriculture & Consumer Services, Division of Plant Industry, citrus canker

84

quarantine facility. Duncan grapefruit plants were heavily pruned and fertilized to induce the

simultaneous boom of a new flush that is susceptible to citrus canker infection. Three weeks later

the emerging uniform new foliage was treated with on of three different single-phage

suspensions (ΦXV3-21, ΦXaacF1 or ccΦ19-1, 5x109 PFU/mL) or with sterilized tap water. Fifty

milliliters of the phage suspension was spayed on each plant in the evening, and then the plants

were placed in white plastic bags. On the next morning the bags were removed from the plants,

and the plants were inoculated with Xac strain Xac65 (test 1×106 cfu/mL). After inoculation the

plants were placed inside the bags for an additional 24 h of high moisture. After removal from

the bags and after the foliage was allowed to dry the plants were arranged in a completely

randomized pattern on a greenhouse bench. The disease was assessed 3-4 weeks after

inoculation. Four plants were used per treatment. In the first experiment the phage suspensions

were prepared by diluting high titer lysates, but because of concerns that the presence of nutrient

broth in the phage lysate may contribute to increased disease severity, in the second experiment

the phage lysates were concentrated in order to remove nutrient broth. The ratio of diseased leaf

surface area was estimated using the Horsfall-Barratt (HB) scale (12). The HB values were

converted to estimated mean percentages by using the Elanco Conversion Tables for Barratt-

Horsfall Rating Numbers (ELANCO Products Co., Indianapolis, IN), as described in Appendix

A. Analysis of variance (ANOVA) and subsequent separation of sample means by the Waller-

Duncan K-ratio t Test was carried out using SAS System for Windows program release 8.02

(Cary, NC).

Population studies. In order to determine if bacteriophages ΦXV3-21, ΦXaacF1 and

ccΦ19-1 are able to multiply on the grapefruit foliage in the presence of their host, Xac65,

grapefruit plants were sprayed with a mixture of the three phages at low concentration (5×106

85

PFU/mL) and immediately followed by application of the bacterial suspension at a much higher

concentration (1×108 cfu/mL) or with sterilized tap water for control. Phage populations were

monitored by removing three leaflets 0, 3, 6 and 9 h after application, recovering the phages

from the leaflets and determining concentrations. In order to determine populations of the

individual phages, the leaf washes were plated on three Xac strains that specifically detected

each of the tree phages. These strains were only sensitive to one of three phages. Strain Xac41

was used for specific detection of ΦXaacF1; Xac15 for ccΦ19-1 and Xac30 for ΦXv3-21.

Results

Persistence of Bacteriophages on Citrus Leaf Surface Under Field Conditions

The fate of three phages, ccΦ13-2, α-MME and Φ5536, was monitored on citrus foliage.

The mixture of these three phages was used for biological control of citrus bacterial spot, and

was applied at 8 PM. The three phages were applied at equal concentrations, at 1×108 PFU/mL,

in skim milk formulation. The sunlight irradiation at the sites was 0 until 6:30 AM then gradually

increased to approximately 400 W/m2 at 9:45 AM (Figure 5-1). Then the irradiation suddenly

jumped to almost 600 W/m2. The three phages persisted differently on the foliage (Figure 5-1).

Φ5536 populations reduced slightly from 8 PM until 10 AM, 132 and 135 fold in sites 1 and 2,

respectively (Figure 5-1). However, concurrently with the suddenly higher sunlight irradiation,

Φ3356 populations sharply dropped between 10 AM and 11AM, decreasing 2159 and 801 folds,

in sites 1 and 2, respectively, and were practically eliminated by 11 AM. The populations of

phage ccΦ13-2 decreased 32 and 12 fold between 8 PM and 10 AM, in sites 1 and 2, respectively

(Figure 5-1). The rate of population decline increased between 10 AM and 11 AM, but it was

much lower than that of Φ3356, with 66 and 30 fold in sites 1 and 2, respectively (Figure 5-1).

At 11 AM still there were, 632 and 142 PFU per leaf ccΦ13-2 at site 1 and site 2, respectively.

ΦMME populations were lower from the beginning compared to the other two phages (Figure

86

5-1). The nighttime reduction was low, but the populations decreased earlier in the morning than

other two phages. In site 1 there was no population reduction between 8 PM and 7:45 AM, but

between 7:45 and 9 AM there was a 2490 fold decrease and from 9 AM or later the populations

were hardly detectable, 19 PFU/leaf (Figure 5-1). In site 2, ΦMME populations reduced slightly,

6 folds, between 8PM and 7:45AM, and more than 500 fold between 7:45 and 10 AM to 121

PFU/leaf (Figure 5-1).

Ability of Three Phages of Xanthomonas axonopodis pv. citri to Multiply on Grapefruit Foliage in the Presence of Their Bacterial Host, and Their Effect on Citrus Canker Disease Development

In order to elucidate if a phage need to be able to multiply on the leaf surface for effective

disease control, three phages active against Xac were evaluated for their ability to (i) multiply on

grapefruit leaf surface in the presence of high populations of the host bacterium, and to (ii)

reduce citrus canker disease severity under greenhouse conditions. The populations of phage

ΦXV3-21 decreased in the absence of Xac65 to 1.6 and 1.9% of the original populations after

nine hours on the grapefruit leaves in experiments 1 and 2, respectively (Figures 5-2A and 5-

3A). In the presence of Xac65 ΦXV3-21 populations persisted better than in the absence of the

host, although they still decreased over time and after 9 hours were 4.2 and 26.2% of the starting

populations. Populations of phage ΦXaacF1 also decreased in the absence of the Xac65 and were

4.2 and 5.8% at the end of the experiment in experiment 1 and 2, respectively (Figures 5-2B and

5-3B). However, in the presence of the host, ΦXaacF1 increased in population, to 324 and 557%,

after nine hours in experiments 1 and 2, respectively (Figures 5-2B and 5-3B). The third phage,

ccΦ19-1 decreased in populations in the absence of Xac65 to 0.4 and 1.5% by the end of the

experiment (Figures 5-2B and 5-3B). In experiment 1, 6 hours after application ccΦ19-1

populations were higher in the presence of the host than in its absence (35% vs. 9.2%); however

87

at nine hours no ccΦ19-1 was detected in the presence of the host (Figure 5-2C). In the second

experiment the presence of Xac65 ccΦ19-1 populations were at 45% after 9 h (Figure 5-2C).

These three phages were used to treat grapefruit plants before inoculating with the

pathogen, Xac65, in two experiments. In the first experiment none of the phages reduced citrus

canker disease severity (Table 5-1). On the contrary, ΦXV3-21 application resulted in significant

disease increase compared to the water control (Table 5-1). However there were significant

differences between plants treated with different phages: ccΦ19-1-treated plants had significantly

less disease than ΦXV3-21-treated plants and significantly more disease than ΦXaacF1-treated

ones (Table 5-1). It was suspected that the nutrient broth present in phage suspensions

contributed to the disease increase on the ΦXV3-21 treated plants and it could have also caused

the loss of control provided by the phages. Thus, for the second experiment nutrient broth was

removed from phage suspensions before applications. In this experiment ΦXaacF1 treatment

significantly reduced disease severity, while the other two phages did not have any significant

effect.

Ability of a Xanthomonas perforans Phage to Multiply on Tomato Foliage in the Presence of Its Bacterial Host, and Its Effect on Tomato Bacterial Spot Disease Development

To investigate if the correlation between a phage’s ability to multiply on the plant surface

and effectively reduce disease is a general and present in other pathosystems, the interactions of

a X. perforans strain, the causal agent of tomato bacterial spot (Figure 5-4), and its bacteriophage

were studied. The ability of phage, ΦXv3-3-13h, to multiply on tomato foliage in the presence of

its host, ΔopgH:ME-B, was evaluated in greenhouse studies. ΦXv3-3-13h populations

continuously decreased in the absence of the bacterial host and were under the detection limit 5

or 6 days after application, in experiments 1 and 2, respectively (Figure 5-5). In the presence of

ΔopgH:ME-B ΦXv3-3-13h populations increased 11 and 87 fold in the first 24 h, in experiments

88

1 and 2, respectively, and then declined at slower rate than without the host (Figure 5-5). After 7

days, the populations were only 8.5 and 17 times lower than at the time of application.

Phage ΦXv3-3-13h was evaluated for its ability to reduce bacterial spot of tomato disease

development incited by ΔopgH:ME-B. Disease severity was significantly reduced on phage-

treated plants compared to water-treated plants (Table 5-2).

Discussion

Phages constituting a phage mixture were shown to persist differently in the phyllosphere.

The phages in the mixture had different levels of persistence and in the morning soon after

sunrise the phage diversity was reduced from three phages to two and by late morning to one.

Experiments with formulations showed that increasing phage persistence contributes to increased

disease control ability. Thus it seems likely that using phages that can persist longer on the

foliage either by to resisting desiccation and sunlight irradiation better, or being more effective in

multiplying in the presence of the pathogen would be better suited for disease control.

The phage that was effective in multiplying on the leaf surface did control the disease.

Conversely, the phages that were not able to multiply efficiently enough to even maintain their

populations could not reduce the disease. We only showed this correlation with a limited number

of phages and only in two pathosystems, and it would be interesting to continue such evaluations

on a broader scale. If it is true that those phages are more effective, which persist longer and can

multiply efficiently in the phyllosphere than the phages naturally inhabiting the phyllosphere

should be the most effective ones. Thus the phyllosphere may be the best place for finding

phages for disease control purposes. It is discouraging, however, that in this study we found no

variability amongst the phyllosphere phages in host ranges. However, our results also showed

that the phages sensitivity profiles of Xanthomonas species overlap, thus phage diversity may be

increased by isolating phages from different Xanthomonas pathosystems.

89

Overall Summary and Conclusions

Citrus canker, incited by Xanthomonas axonopodis pv. citri and X. axonopodis pv.

aurantifolii, is one of the most damaging citrus diseases in the world. Citrus canker was

reintroduced to Florida in the 1990s and threatens the state’s $9 billion citrus industry. This work

focused on a biological control approach to use bacteriophages for reducing bacterial pathogen

populations and disease severity on citrus. Bacteriophages isolated from citrus canker lesions in

Florida and Argentina were evaluated based on plaque morphology, chloroform sensitivity, host

range, genome size, DNA restriction profile and virion morphology. There was low diversity

among the isolated phages, as 61 of 67 were nearly identical to each other and phage CP2 of

Japan. Mixtures of bacteriophages were evaluated for controlling citrus canker in greenhouse

trials in Florida and in nursery trials in Argentina. Bacteriophages reduced citrus canker disease

severity both in greenhouse and field trials. The level of control was inferior to chemical control

with copper bactericides. The combination of bacteriophage and copper treatments did not result

in increased control. Citrus canker field trials in Florida have been prohibited until recently, as

the disease was under eradication. For this reason we evaluated the efficacy of phage treatment

on a similar bacterial citrus disease, citrus bacterial spot, incited by X. axonopodis pv. citrumelo.

Bacteriophages reduced citrus bacterial spot severity. The level of control was equal or inferior

to chemical control with copper bactericides. The combination of bacteriophage and copper

treatments did not result in increased control. In experiments monitoring the fate of

bacteriophages on the citrus foliage following bacteriophage application, phage populations

stayed steady on the foliage during nighttime but were drastically reduced within hours after

sunrise. The rate of reduction varied among the phages. The ability of bacteriophages to multiply

on the plant foliage in the presence of their bacterial host was investigated. Phages varied in their

ability to multiply, and the ones that successfully increased in populations on the bacterial host

90

on the leaf surface also reduced disease severity, whereas the ones that were unable to multiply

in the target environment did not reduce disease severity.

However there is a need for research in several areas before commercial phage application

will become a feasible option. First of all, the diversity of the citrus canker phage library has to

be increased, either by further isolations or by changing host ranges and receptor specificities of

available phages. The protective formulation also need improvement: the 0.75% skim milk

powder formulation increases phage longevity on the foliage, but it also provides carbon source

to the microorganisms on the leaf surface, including the pathogen. Additionally, the protein

content of skim milk may tie up a considerable amount of ionic copper thus reducing the efficacy

of copper bactericides. Identification of formulations that provide protection without the adverse

effects could potentially result in increased disease control efficacy. The optimal frequency of

application and phage concentration also remains to be determined. Both of these factors greatly

influence the cost of phage treatment, what in turn determines feasibility. In these field tests we

applied phages twice weekly at ~108 PFU/ml, but it is not known if less frequent applications

and lower phage concentrations would have achieved similar control. Combining phage

treatment with other biological control methods and chemicals with different modes of action

than copper also needs to be evaluated. It may be possible to increase disease control efficacy of

individual bacteriophages by selection for increased persistence on the leaf surfaces and higher

in vivo multiplication ability. Also, rapid screening methods could be developed to predict

disease control ability of phages.

91

0

1

2

3

4

5

6

7

8:15 PM 7:45 AM 9:00 AM 10:00 AM 11:00 AM

Log 1

0 PFU

/g le

af ti

ssue

Φ5536ΦMMEccΦ13-2

A

0

1

2

3

4

5

6

7

8:15 PM 7:45 AM 9:00 AM 10:00 AM 11:00 AM

Log 1

0 PFU

/g le

af ti

ssue

Φ5536ΦMMEccΦ13-2

B

0

100

200

300

400

500

600

700

8:00 P

M

9:00 P

M

10:00

PM

11:00

PM

12:00

AM

1:00 A

M

2:00 A

M

3:00 A

M

4:00 A

M

5:00 A

M

6:00 A

M

7:00 A

M

8:00 A

M

9:00 A

M

10:00

AM

11:00

AM

Sun

light

(W/m

2 )

C

Figure 5-1. Bacteriophage populations and sunlight irradiation in Citra, FL on July 18 and 19,

2006. A) Bacteriophage populations at site 1. B) Bacteriophage populations at site 2. C) Sunlight irradiation. Phage was applied at 8:00PM. Error bars indicate the standard error.

92

0

20

40

60

80

100

120

140

160

0 3 6 9

Hours after inoculation

Per

cent

pop

ulat

ion

ΦXV3-21ΦXV3-21 with host

A

0

100

200

300

400

500

600

0 3 6 9

Hours after inoculation

Per

cent

pop

ulat

ion

ΦXaacF1ΦXaacF1 with host

B

0

20

40

60

80

100

120

140

160

0 3 6 9

Hours after inoculation

Per

cent

pop

ulat

ion

ccΦ19-1ccΦ19-1 with host

C

Figure 5-2. Changes in bacteriophage populations on grapefruit foliage in the presence or

absence of the host, Xanthomonas axonopodis pv. citri strain Xac65. Experiment 1. A) Phage ΦXV3-21. B) Phage ΦXaacF1. C) Phage ccΦ19-1. Percent population = percent of population average at 0 hour. Error bars indicate the standard error.

93

0

20

40

60

80

100

120

140

0 3 6 9

Hours after inoculation

Per

cent

pop

ulat

ion

ΦXV3-21ΦXV3-21 with host

A

0

100

200

300

400

500

600

700

800

900

0 3 6 9

Hours after inoculation

Perc

ent p

opul

atio

n

ΦXaacF1ΦXaacF1 with host

B

020406080

100120140160180200

0 3 6 9

Hours after inoculation

Per

cent

pop

ulat

ion

ccΦ19-1ccΦ19-1 with host

C

Figure 5-3. Changes in bacteriophage populations on grapefruit foliage in the presence or

absence the host, Xanthomonas axonopodis pv. citri strain Xac65. Experiment 2. A) Phage ΦXV3-21. B) Phage ΦXaacF1. C) Phage ccΦ19-1. Percent population = percent of population average at 0 hour. Error bars indicate the standard error.

94

Table 5-1. Effect of treatment with three phages on citrus canker disease development, incited by Xanthomonas axonopodis pv. citri strain Xac65, as measured by disease severity.

Average disease severity Treatment Controly ccΦ19-1 ΦXaacF1 ΦXV3-21 pz

Test 1 22bc 33b 13c 51a 0.0018 Test 2 28ab 24bc 18c 32a 0.0049

y Means within the same row followed by the same letter are not significantly different based on least square means differences at p=0.05 level. z p = Probability that there are no differences in treatment means according to analysis of variance.

95

Figure 5-4. Symptoms of tomato bacterial spot, incited by Xanthomonas perforans.

96

0123456789

10

1 2 3 4 5 6 7 8

Days after application

Log 1

0 PFU

/leaf

let

HostNo Host

A

0123456789

10

1 2 3 4 5 6 7 8

Days after application

Log 1

0 P

FU/le

afle

t

HostNo Host

B

Figure 5-4. Populations of bacteriophage ΦXv3-3-13h on tomato foliage in the presence or

absence of its host Xanthomonas perforans strain ΔopgH:ME-B. A) Experiment 1. B) Experiment 2. Error bars indicate the standard error.

97

Table 5-2. Effect of a curative application of bacteriophage ΦXv3-3-13h on tomato bacterial

spot disease development, incited by Xanthomonas perforans strain ΔopgH:ME-B, as measured by area under the disease progress curve (AUDPC).

Treated Control pz

AUDPC 12.5 27.3 0.0026 z Probability of equality of sample means according to t test.

98

APPENDIX A CALCULATIONS

Conversion of Horfall-Barratt Values to Mean Percentages

The Horsfall-Barratt values (12) were converted to estimated mean percentages by using

the Elanco Conversion Tables for Barratt-Horsfall Rating Numbers (ELANCO Products Co.,

Indianapolis, IN). The estimated mean percentage of certain experimental unit, one plant in the

greenhouse experiments or one plot in the field trials, was determined by taking the arithmetic

mean of the percentages of several ratings taken at one time, where each individual percentage

was taken as the midpoint of the appropriate percentage interval (Table A-1). For example, if the

following four HB ratings were assigned to a grapefruit plant, 10, 7, 5 and 10, then the estimated

mean percentage was calculated as (95.31 + 62.5 + 18.75 + 95.31) / 4 = 78.5.

Calculation of Area Under the Disease Progress Curve

The area under the disease progress curve (AUDPC) (Figure A-1) was computed by the

trapezoidal integration of disease percentage values taken in different timepoints using the

formula proposed by Shaner and Finnley (104). Disease progress curve is graphed by plotting the

diseases percentage values (y-axis) against the time of disease assessment (x-axis) (Figure A-1).

The area under this curve consists of several trapezoids. The area of each trapezoid is determined

as (y2+y1)/2*(x2-x1), where (x1,y1) are the coordinate of the earlier and (x2,y2) are the coordinates

of the later observation. For example if four disease ratings were taken one week apart and the

disease percentages are 5, 10, 20 and 25, then there will be three trapezoids. The area of the first

trapezoid is (10+5)/2*7=52.5, the area of the second is 105, the area of the third one is 157.5, so

the total area of the disease progress curve is 315.

99

Table A-1. Horsfall-Barratt scale, the corresponding disease intervals and midpoint values.

HB value Disease Percentage Midpointz

1 0 0 2 0-3 2.34 3 3-6 4.68 4 6-12 9.37 5 12-25 18.75 6 25-50 37.5 7 50-75 62.5 8 75-88 81.25 9 87-94 90.63 10 94-97 95.31 11 97-100 97.66 12 100 100

z Midpoint values are as published in the Elanco Conversion Tables for Barratt-Horsfall Rating Numbers (ELANCO Products Co., Indianapolis, IN).

100

A

0

5

10

15

20

25

30

0 7 14 21

Days after first rating

Per

cent

dis

ease

Disease Progress Curve

B

0

5

10

15

20

25

30

0 7 14 21

Days after first rating

Per

cent

dis

ease

Area Under the Disease Progress Curve

Figure A-1. Visualization of the disese progress curve and of the area under the disease progress

curve. A) Disease progress curve is prepared by plotting the diseases percentage values (y-axis) against the time of disease assessment (x-axis). B) Area under the disease progress curve (AUDPC) is one value describing the overall disease progress throughout the season.

101

LIST OF REFERENCES

1. Ackermann H. 2005. Bacteriophage classification. Pages 67-89 in: Bacteriophages: biology and applications. in: Kutter E, and Sulakvelidze A, eds. CRC Press, Boca Raton, Florida.

2. Adams MH. 1959. Bacteriophages. Interscience Publishers, New York, New York.

3. Agrios GN. 2004. Plant pathology. 5th ed. Elseiver Academic Press, San Diego, California.

4. Ahamd MH, and Morgan V. 1994. Characterization of a cowpea (Vigna undulata) rhizobiophage and its effects on cowpea nodulation and growth. Biol Fertil Soils 18:297-301.

5. Al-saadi A. 2005. Phenotypic characterization and sequence analysis of pthA homologs from five pathogenic variant groups of Xanthomonas citri [dissertation]. Gainesville, Florida: University of Florida.

6. Alvarez AM, Benedict AA, Mizumoto CY, Pollard LW, and Civerolo EL. 1991. Analysis of Xanthomonas campestris pv. citri and X. c. citrumelo with monoclonal antibodies. Phytopathology 81:857-865.

7. Arthurs SP, Lacey LA, and Behle RW. 2006. Evaluation of spray-dried lignin-based formulations and adjuvants as solar protectants for the granulovirus of the codling moth, Cydia pomonella (L). Journal of invertebrate pathology 93:88-95. Epub: 2006 Jun 13.

8. Balogh B. 2002. Strategies of improving the efficacy of bacteriophages for controlling bacterial spot of tomato [thesis]. Gainesville, Florida: University of Florida.

9. Balogh B, Jones JB, Momol MT, and Olson SM. 2005. Momol MT, Ji P, Jones JB, editors. Persistence of bacteriophages as biocontrol agents in the tomato canopy. Proceedings of the 1st IS on tomato diseases; June 21-24, 2004; Orlando, Florida. ISHS:299-302.

10. Balogh B, Jones JB, Momol MT, Olson SM, Obradovic A, King P, and Jackson LE. 2003. Improved efficacy of newly formulated bacteriophages for management of bacterial spot on tomato. Plant Dis 87:949-954.

11. Balogh B, Jones JB, Stall RE, Canteros BI, and Gochez AM. 2004. Association of bacteriophages with citrus canker and their use for control. Phytopathology 94 6.

12. Barratt RW, and Horsfall JG. 1945. An improved grading system for measuring plant disease. Phytopathology 35:655.

13. Barrow PA. 2001. The use of bacteriophages for treatment and prevention of bacterial disease in animals and animal models of human infection. J Chem Technol Biotechnol 76:677-682.

14. Basit HA, Angle JS, Salem S, and Gewaily EM. 1992. Phage coating of soybean seeds reduces nodulation by indigenous soil bradyrhizobia. Can J Microbiol 38:1264-1269.

102

15. Behle RW, McGuire MR, Gillespie RL, and Shasha BS. 1997. Effects of alkaline gluten on the insecticidal activity of Bacillus thuringiensis. J Econ Entomol 90:354-360.

16. Behle RW, McGuire MR, and Shasha BS. 1996. Extending the residual toxicity of Bacillus thuringiensis with casein-based formulations. J Econ Entomol 89 6:1399-1405.

17. Belasque J,Jr, Parra-Pedrazzoli AL, Neto JR, Yamamoto PT, Chagas MCM, Parra JRP, Vinyard BT, and Hartung JS. 2005. Adult citrus leafminers (Phyllocnistis citrella) are not efficient vectors for Xanthomonas axonopodis pv. citri. Plant Dis 89:590-594.

18. Bergamin Filho A, and Kimati H. 1981. Estudos sobre um bacteriofago isolado de Xanthomonas campestris. II. Seu emprego no controle de X. campestris e X. vesicatoria. Sum Phytopath 7:35-43.

19. Bock CH, Parker PE, and Gottwald TR. 2005. Effect of simulated wind-driven rain on duration and distance of dispersal of Xanthomonas axonopodis pv. citri from canker-infected citrus trees. Plant Dis 89:71-80.

20. Brunings AM, and Gabriel DW. 2003. Xanthomonas citri: breaking the surface. Molecular plant pathology 4:141-157.

21. Civerolo EL. 1984. Bacterial canker disease of citrus. J Rio Grande Valley Hortic Assoc 37:127-146.

22. Civerolo EL. 1972. Maas Geesteranus HP, editor. Interaction between bacteria and bacteriophages on plant surfaces and in plant tissues. Proceedings of the third international conference of plant pathogenic bacteria; April 14-21 1971; Wageningen. Wageningen: Centre for Argricultural Publishing and Documentation 25-37.

23. Civerolo EL, and Keil HL. 1969. Inhibition of bacterial spot of peach foliage by Xanthomonas pruni bacteriophage. Phytopathology 12:1966-1967.

24. Compton L. USDA determines citrus canker eradication not feasible Online: Florida Department of Agriculture and Consumer Services - Division of Plant Industry. 2006 January 11 [cited 2006 October 31]. Available from: http://www.doacs.state.fl.us/press/2006/01112006_2.html.

25. Coons GH, and Kotila JE. 1925. The transmissible lytic principle (bacteriophage) in relation to plant pathogens. Phytopathology 15:357-370.

26. Cubero J, and Graham JH. 2002. Genetic relationship among worldwide strains of Xanthomonas causing canker in citrus species and design of new primers for their identification by PCR. Appl Environ Microbiol 68:1257-1264.

27. Dai H, Chiang KS, and Kuo TT. 1980. Characterization of a new filamentous phage Cf from Xanthomonas citri. J Gen Virol 46:277-289.

28. d'Herelle F. 1921. Le bactériophage: Sone rôle dans limmunité. Masson et Cie, Paris.

103

29. Dopson RN. 1964. The eradication of citrus canker. Pl Dis Rept 48:30-31.

30. Duan YP, Castaneda A, Zhao G, Erdos G, and Gabriel DW. 1999. Expression of a single, host-specific, bacterial pathogenicity gene in plant cells elicits division, enlargement, and cell death. Mol Plant-Microbe Interact 12:556-560.

31. Duckworth DH, and Gulig PA. 2002. Bacteriophages: potential treatment for bacterial infections. BioDrugs 16:57-62.

32. Egel DS, Graham JH, and Stall RE. 1991. Genomic relatedness of Xanthomonas campestris strains causing diseases of citrus. Appl Environ Microbiol 57:2724-2730.

33. Erskine JM. 1973. Characteristics of Erwinia amylovora bacteriophage and its possible role in the epidemiology of fire blight. Can J Microbiol:7.

34. FDACS DPI. Comprehensive Report on Citrus Canker Eradication Program in Florida Through 14 January 2006 Revised Online: Florida Department of Agriculture and Consumer Services Division of Plant Industry. 2006 August 29. Available from: http://www.doacs.state.fl.us/pi/canker/pdf/cankerflorida.pdf.

35. FDACS DPI.. Florida Huanglongbing Science Panel Report Online: Florida Department of Agriculture and Consumer Services, Division of Plant Industry. 2006 January 31. Available from: http://www.doacs.state.fl.us/pi/chrp/greening/hlbsciencereport1-31-06.pdf.

36. FDACS DPI. Huanglongbing statewide survey map Online: Florida Department of Agriculture and Consumer Services, Division of Plant Industry. 2006 October 25. Available from: http://www.doacs.state.fl.us/pi/chrp/greening/maps/cgsit_map.pdf.

37. Flaherty JE, Harbaugh BK, Jones JB, Somodi GC, and Jackson LE. 2001. H-mutant bacteriophages as a potential biocontrol of bacterial blight of geranium. HortScience 36:98-100.

38. Flaherty JE, Jones JB, Harbaugh BK, Somodi GC, and Jackson LE. 2000. Control of bacterial spot on tomato in the greenhouse and field with H-mutant bacteriophages. HortScience 35:882-884.

39. Gambley CF. 2005.The distribution of citrus canker in Emerald, Australia and bacterial survival in citrus trash. Second international citrus canker and huanglongbing research workshop; November 7-11, 2005; Orlando,Florida. 70.

40. Gill JJ, and Abedon ST. 2003. Bacteriophage Ecology and Plants. APSnet November. http://www.apsnet.org/online/feature/phages/.

41. Gill JJ, Svircev AM, Smith R, and Castle AJ. 2003. Bacteriophages of Erwinia amylovora. Appl Environ Microbiol 69:2133-2138.

42. Goodridge LD. 2004. Bacteriophage biocontrol of plant pathogens: fact or fiction? Trends in biotechnology 22:384-385.

104

43. Goto M. 1992. Citrus canker. Pages 250-269 in: Plant Diseases of International Importance. in: Kumar J, Chaube HS, Singh US, and Mukhopadhyay AN, eds., Prentice-Hall, Englewood Cliff, NJ.

44. Goto M. 1992. Fundamentals of bacterial plant pathology. Academic Press, Inc., San Diego, California.

45. Goto M, Serizawa S, and Morita M. 1971. Studies on citrus canker disease. II. Leaf infiltration technique for detection of Xanthomonas citri (Hasse) Dowson, with special reference to comparison with phage method.

46. Goto M, Takahashi T, and Messina MA. 1980. A comparative study of the strains of Xanthomonas campestris pv. citri isolated from citrus canker in Japan and cancrosis B in Argentina. Annals of the Phytopathological Society of Japan 46:329-338.

47. Goto M, and Yaguchi Y. 1979. Relationship between defoliation and disease severity in citrus canker (Xanthomonas citri). Ann Phytopathol Soc Japan 45:689-694.

48. Gottwald TR, and Graham JH. 1992. A device for precise and nondisruptive stomatal inoculation of leaf tissue with bacterial pathogens. Phytopathology 82:930-935.

49. Gottwald TR, Graham JH, and Schubert TS. Citrus canker: The pathogen and its impact. Plant Health Progress. 2002 2002 August 12 doi:10.1094/PHP-2002-0812-01-RV:2006 October 31. http://www.plantmanagementnetwork.org/pub/php/review/citruscanker.

50. Gottwald TR, Hughes G, Graham JH, Sun X, and Riley T. 2001. The citrus canker epidemic in Florida: The scientific basis of regulatory eradication policy for an invasive species. Phytopathology 91:30-34.

51. Gottwald TR, Sun X, Riley T, Graham JH, Ferrandino F, and Taylor EL. 2002. Geo-referenced spatiotemporal analysis of the urban citrus canker epidemic in Florida. Phytopathology 92:361-377.

52. Goyal SM, Gerba CP, and Bitton G, editors. 1987. Phage ecology. Wiley, New York.

53. Graham JH. 1995. Root regeneration and tolerance of citrus rootstocks to root rot caused by Phytophthora nicotianae. Phytopathology 85:111-117.

54. Graham JH, Bright DB, and McCoy CW. 2003. Phytophthora - Diaprepes weevil complex: Phytophthora spp. relationship with citrus rootstocks. Plant Dis 87 1:85-90.

55. Graham JH, Drouillar DL, Bannwart DL, and Dilley JR. 2000. Bud failure on Swingle citrumelo in a Florida citrus nursery caused by Xanthomonas axonopodis pv. citrumelo. Proceedings of the 13th annual meeting of the Florida State Horticultural Society 112:60-68.

56. Graham JH, and Gottwald TR. 1990. Variation in aggressiveness of Xanthomonas campestris pv. citrumelo associated with citrus bacterial spot in Florida citrus nurseries. Phytopathology 80:190-196.

105

57. Graham JH, Gottwald TR, Cubero J, and Achor DS. 2004. Pathogen profile: Xanthomonas axonopodis pv. citri : factors affecting successful eradication of citrus canker. Mol Plant Pathol 5:1-15.

58. Graham JH, Gottwald TR, Riley TD, and Achor D. 1992. Penetration through leaf stomata and strains of Xanthomonas campestris in citrus cultivars varying in susceptibility to bacterial diseases. Phytopathology 82:1319-1325.

59. Greer GG. 2005. Bacteriophage control of foodborne bacteria. J Food Prot 68:1102-1111.

60. Halbert S. 2005.The discovery of Huanglongbing in Florida. 2nd international citrus canker/Huanglongbing research workshop; November 7-11, 2005; Orlando,Florida. 50.

61. Hendrick CA, and Sequeira L. 1984. Lipopolysaccharide-defective mutants of the wilt pathogen Pseudomonas solanacearum. Appl Environ Microbiol 48:94-101.

62. Hibma AM, Jassim SAA, and Griffiths MW. 1997. Infection and removal of L-forms of Listeria monocytogenes with bred bacteriophage. Int J Food Microbiol 34:197-207.

63. Ignoffo CM, and Garcia C. 1996. Simulated sunlight-UV sensitivity of experimental dust formulations of the nuclear polyhedrosis virus of Helicoverpa/Heliothis. J Invertebr Pathol 67:192-194.

64. Ignoffo CM, and Garcia C. 1995. Aromatic/heterocyclic amino acids and the stimulated sunlight-ultraviolet inactivation of the Heliothis/Helicoverpa baculovirus. Environ Entomol 24:480-482.

65. Ignoffo CM, Garcia C, and Saathoff SG. 1997. Sunlight stability and rain-fastness of formulations of Baculovirus heliothis. Environ Entomol 26:1470-1474.

66. Iriarte FB, Balogh B, Momol MT, and Jones JB. In press. Factors affecting survival of bacteriophage on tomato leaf surfaces. Appl Environ Microbiol.

67. Johnson KB. 1994. Dose-response relationships and inundative biological control. Phytopathology 84:780-784.

68. Katznelson H, and Sutton MD. 1951. A rapid phage plaque count method for the detection of bacteria as applied to the demonstration of internally borne bacterial infection of seed. J Bacteriol 61:689-701.

69. Keel C, Ucurum Z, Michaux P, Adrian M, and Haas D. 2002. Deleterious impact of a virulent bacteriophage on survival and biocontrol activity of Pseudomonas fluorescens Strain CHA0 in Natural Soil. Mol Plant-Microbe Interact 15:567-576.

70. Klement Z, Rudolf K, and Sands DC, editors. 1990. Methods in phytobacteriology. Akadémiai Kiadó, Budapest.

106

71. Koizumi M. 1985. Citrus canker: the world situation. Pages 2-7 in: Citrus canker: An international perspective. Timmer LW, ed. Citrus Research & Education Center, University of Florida, Lake Alfred.

72. Koller W. 1998. Chemical approaches to managing plant pathogens. in: Handbook of integrated pest management. Ruberson JR, ed. Dekker.

73. Kotila JE, and Coons GH. 1925. Investigations on the blackleg disease of potato. Michigan Agr Exper Sta Tech Bull 67:3-29.

74. Kutter E. Phage therapy: bacteriohages as antibiotics Olympia, Washington: The Evergreen State College. 1997 Novermber 15. Available from: http://www.evergreen.edu/phage/phagetherapy/phagetherapy.htm.

75. Leite JRP. 2005. Integrated management of citrus canker in Southern Brazil. Second international citrus canker and huanglongbing research workshop; November 7-11, 2005; Orlando,Florida.

76. Leite JRP, and Mohan SK. 1990. Integrated management of the citrus bacterial canker disease caused by Xanthomonas campestris pv. citri in the State of Paraná, Brazil. Crop Protection 9:3-7.

77. Leverentz B, Conway WS, Alavidze Z, Janisiewicz WJ, Fuchs Y, Camp MJ, Chighladze., and Sulakvelidze A. 2001. Examination of bacteriophage as a biocontrol method for Salmonella on fresh-cut fruit: a model study. J Food Prot 64:1116-1121.

78. Leverentz B, Conway WS, Camp MJ, Janisiewicz WJ, Abuladze T, Yang M, Saftner R, and Sulakvelidze A. 2003. Biocontrol of Listeria monocytogenes on fresh-cut produce by treatment with lytic bacteriophages and a bacteriocin. Appl Environ Microbiol 69:4519-4526.

79. Mallmann WL, and Hemstreet CJ. 1924. Isolation of an inhibitory substance from plants. Arg Res 28:599-602.

80. Manulis S, Zutra D, Kleitman F, Dror O, David I, Zilberstaine M, and Shabi E. 1998. Distribution of streptomycin-resistant strains of Erwinia amylovora in Israel and occurrence of blossom blight in the autumn. Phytoparasitica 26:223-230.

81. Marco GM, and Stall RE. 1983. Control of bacterial spot of pepper initiated by strains of Xanthomonas campestris pv. vesicatoria that differ in sensitive to copper. Plant Dis 67:779-781.

82. McKenna F, El-Tarabily KA, Hardy GESTJ, and Dell B. 2001. Novel in vivo use of a polyvalent Streptomyces phage to disinfest Streptomyces scabies-infected seed potatoes. Plant Pathol 50:666-675.

83. McNeil DL, Romero S, Kandula J, Stark C, Stewart A, and Larsen S. 2001. Bacteriophages: A potential biocontrol agent against walnut blight (Xanthomonas campestris pv. juglandis). N Z Plant Prot 54:220-224.

107

84. Minsavage GV, Canteros BI, and Stall RE. 1990. Plasmid-mediated resistance to streptomycin in Xanthomonas campestris pv. vesicatoria. Phytopathology 80:719-723.

85. Momol MT, Jones JB, Olson SM, Obradovic A, Balogh B, King P. Integrated management of bacterial spot on tomato in Florida. EDIS, Institute of Food and Agricultural Sciences, University of Florida, Online 2002 September. Report No.: PP110.

86. Munsch P, and Olivier JM. 1995. Biocontrol of bacterial blotch of the cultivated mushroom with lytic phages: Some practical considerations. Pages 595-602 in: Science and cultivation of edible fungi, Vol II: Proceedings of the 14th international congress. Elliot TJ, ed.

87. Munsch P, Olivier JM, and Houdeau G. 1991. Experimental control of bacterial blotch by bacteriophages. Pages 389-396 in: Science and cultivation of edible fungi. in: Maher MJ, ed. Balkema, Rotterdam, The Netherlands.

88. Muraro RP, Roka F, and Spreen TH. 2000.An overview of Argentina’s citrus canker control program with applicable costs for a similar program in Florida. Proceedings of the international citrus canker research workshop; June 20-22, 2000; Ft. Pierce FL. Online: Division of Plant Industry, Florida Department of Agriculture and Consumer Services.

89. Obradovic A, Jones JB, Momol MT, Balogh B, and Olson SM. 2004. Management of tomato bacterial spot in the field by foliar applications of bacteriophages and SAR inducers. Plant Dis 88:736-740.

90. Obradovic A, Jones JB, Momol MT, Olson SM, Jackson LE, Balogh B, Guven K, and Iriarte FB. 2005. Integration of biological control agents and systemic acquired resistance inducers against bacterial spot on tomato. Plant Dis 89:712-716.

91. Okabe N, and Goto M. 1963. Bacteriophages of plant pathogens. Annu Rev Phytopathol 1:397-418.

92. Ophel KM, Bird AF, and Kerr A. 1993. Association of bacteriophage particles with toxin production by Clavibacter toxicus, the causal agent of annual ryegrass toxicity. Phytopathology 83:676-681.

93. Osawa S, Furuse K, and Watanabe I. 1981. Distribution of ribonucleic acid coliphages in animals. Appl Environ Microbiol 45:164-168.

94. Pruvost O, Hartung JS, Civerolo EL, Dubois C, and Perrier X. 1992. Plasmid DNA fingerprints distinguish pathotypes of Xanthomonas campestris pv. citri , the causal agent of citrus bacterial canker disease. Phytopathology 82:485-490.

95. Randhawa PS, and Civerolo EL. 1986. Interaction of Xanthomonas campestris pv. pruni with pruniphage and epiphytic bacteria on detached peach leaves. Phytopathology 76:549-553.

96. Reichelderfer K, and Barry JW. 1995. Introduction. Pages 28-35 in: Biorational pest control agents: formulation and delivery. in: Hall FR, andBarry JW, eds.American Chemical Society, Washington, DC.

108

97. Rizvi S, and Mora PT. 1963. Bacteriophage plaque-count assay and confluent lysis on plates without bottom agar layer. Nature 200:1324-1325.

98. Rybak MA. 2005. Genetic determinants of host range specificity of the Wellington strain of Xanthomonas axonopodis pv. citri [dissertation]. Gainesville, Florida: University of Florida.

99. Saccardi A, Gambin E, Zaccardelli M, Barone G, and Mazzucchi U. 1993. Xanthomonas campestris pv. pruni control trials with phage treatments on peaches in the orchard. Phytopathol Mediterr 32:206-210.

100. Sambrook J, Fritsch EF, and Maniatis T. 1989. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York.

101. Schnabel EL, Fernando WGD, Meyer MP, Jones AL, and Jackson LE. 1999. Bacteriophage of Erwinia amylovora and their potential for biocontrol.

102. Schoulties CL, Civerolo EL, Miller JW, Stall RE, Krass CJ, Poe SR, and Ducharme EP. 1987. Citrus canker in Florida. Plant Dis 71:388-395.

103. Schubert TS, Rizvi SA, Sun X, Gottwald TR, Graham JH, and Dixon WN. 2001. Meeting the challenge of eradicating citrus canker in Florida - Again. Plant Dis 85:340-356.

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

105. Shapiro JP, and Gottwald TR. 1995. Resistance of eight cultivars of citrus rootstock to a larval root weevil, Diaprepes abbreviatus L. (Coleoptera: Curculionidae). J Econ Entomol 88:148-154.

106. Stall RE, and Civerolo EL. 1991. Research relating to the recent outbreak of citrus canker in Florida. Annu Rev Phytopathol:399-420.

107. Stall RE, Marco GM, and Canteros de Echenique BI. 1982. Importance of mesophyll in mature-leaf resistance to cancrosis of citrus. Phytopathology 72:1097-1100.

108. Stall RE, Miller JW, Marco GM, and Canteros de Echenique BI. 1980. Population dynamics of Xanthomonas citri causing cancrosis of citrus in Argentina. Proc Fla Hort Soc 93:10-14.

109. Stein B, Ramallo J, Foguet L, and Morandini M. 2005.Chemical control of citrus canker in lemons (Citrus limon, (L.) Burm. f) in Tucuman, Argentina (Abstr.). Second international citrus canker and huanglongbing research workshop; November 7-11, 2005; 25.

110. Storey MV, and Ashbolt NJ. 2001. Persistence of two model enteric viruses (B40-8 and MS-2 bacteriophages) in water distribution pipe biofilms. Water Science and Technology 43:133-138.

109

111. Stover E, Castle B, and Hebb J. Citrus rootstock usage in the Indian river region Online: Electronic Data Information Source (EDIS) HS-852, Horticultural Science Department, IFAS, University of Florida. 2002 April. Available from: http://edis.ifas.ufl.edu/HS129.

112. Summers WC. 2005. Bacteriophage research: Early History. Pages 5-27 in: Bacteriophages: biology and applications. in: Kutter E, and Sulakvelidze A, eds. CRC Press, Boca Raton, Florida.

113. Sun X, Jones D, and Duan YP. 2005.Detection of Manatee genotypes of citrus canker bacteria in Hillsborough Co, Florida. Second international citrus canker and huanglongbing research workshop; November 7-11, 2005; Orlando,Florida. 32.

114. Sun X, Stall RE, Jones JB, Cubero J, Gottwald TR, Graham JH, Dixon WN, Schubert TS, Chaloux PH, Stromberg VK, Lacy GH, and Sutton BD. 2004. Detection and characterization of a new strain of citrus canker bacteria from Key/Mexican lime and alemow in south Florida. Plant Dis 88:1179-1188.

115. Svircev AM, Lehman SM, Kim W-, Barszcz E, Schneider KE, and Castle AJ. 2006. Zeller W, Ullrich C, editors. Control of the fire blight pathogen with bacteriophages. Proceedings of the 1st international symposium on biological control of bacterial plant diseases; October 23-26, 2005; Seeheim/Darmstadt, Germany. Berlin, Germany: Die Deutsche Bibliothek - CIP-Einheitsaufnahme 259-261.

116. Sykes IK, Lanning S, and Williams ST. 1981. The effect of pH on soil actinophage. J Gen Microbiol 122:271-280.

117. Tanaka H, Negishi H, and Maeda H. 1990. Control of tobacco bacterial wilt by an avirulent strain of Pseudomonas solanacearunm M4S and its bacteriophage. Ann Phytopath Soc Japan 56:243-246.

118. Tanji Y, Shimada T, Yoichi M, Miyanaga K, Hori K, and Unno H. 2004. Toward rational control of Escherichia coli O157:H7 by a phage cocktail. Appl Microbiol Biotechnol 64:270-274.

119. Thayer PL, and Stall RE. 1961. A survey of Xanthomonas vesicatoria resistance to streptomycin. Proc Fla Hort Soc 75:163-165.

120. Thomas RC. 1935. A bacteriophage in relation to Stewart’s disease of corn. Phytopathology 25:371-372.

121. Timmer LW, Zitko SE, Gottwald TR, and Graham JH. 2000. Phytophthora brown rot of citrus: Temperature and moisture effects on infection, sporangium production, and dispersal. Plant Dis 84:157-163.

122. UNCTAD INFO COMM Citrus fruit Online: United Nations Conference on Trade and Development, Market Information in the Commodities Area [cited 2006 October 31]. Available from: http://www.unctad.org/infocomm/anglais/orange/market.htm#prod.

110

123. USDA APHIS. Citrus canker Online: United States Department of Agriculture Animal and Plant Health Inspection Service. 2006 October 5. Available from: http://www.aphis.usda.gov/newsroom/hot_issues/citrus_canker/citrus_canker.shtml.

124. USDA FAS. Situation and outlook 2005/06 Online: United States Department of Agriculture Foreign Agricultural Service. 2006 February 20. Available from: http://www.fas.usda.gov/htp/Hort_Circular/2006/02-06/02-20-06%20Citrus%20Feature.pdf.

125. USDA NASS. Citrus fruits - 2006 Summary Online: United States Department of Agriculture National Agricultural Statistics Service. 2006 September 21. Available from: http://usda.mannlib.cornell.edu/usda/current/CitrFrui/CitrFrui-09-21-2006.pdf.

126. Verniere C, Hartung JS, Pruvost OP, Civerolo EL, Alvarez AM, Maestri P, and Luisetti J. 1998. Characterization of phenotypically distinct strains of Xanthomonas axonopodis pv. citri from southwest Asia. Eur J Plant Pathol 104:477-487.

127. Vidaver AK. 1976. Prospects for control of phytopathogenic bacteria by bacteriophages and bacteriocins. Annu Rev Phytopathol:14.

128. Vidaver AK, and Schuster ML. 1960. Characterization of Xanthomonas phaseoli bacteriophages. J Virol 4:300-308.

129. Voloudakis AE, Reignier TM, and Cooksey DA. 2005. Regulation of resistance to copper in Xanthomonas axonopodis pv. vesicatoria. Appl Environ Microbiol 71:782-789.

130. Wagner PL, and Waldor MK. 2002. Bacteriophage control of bacterial virulence. Infection and immunity 70:3985-3993.

131. Wiggins BA, and Alexander M. 1985. Minimum bacterial density for bacteriophage replication: implications for significance of bacteriophages in natural ecosystems. Appl Environ Microbiol 49:19-23.

132. Williams ST, Mortimer AM, and Manchester L. 1987. Ecology of soil bacteriophages. Pages 157-179 in: Phage ecology. in: Goyal SM, Gerba CP, and Bitton G, eds. John Wiley and Sons, New York.

133. Woods TL, Israel HW, and Sherf AF. 1981. Isolation and Partial Characterization of a bacteriophage of Erwinia stewartii from the corn flea beetle, Chaetocnema pulicaria. Prot.Ecol 3:229-236.

134. Wu WC. 1972. Phage-induced alterations of cell disposition, phage adsorption and sensitivity, and virulence in Xanthomonas citri. Ann Phytopath Soc Japan 38:333-341.

135. Wu WC, Lee ST, Kuo HF, and Wang LY. 1993. Use of phages for indentifying the citrus canker bacterium Xanthomonas campestris pv. citri in Taiwan. Plant Pathol 42:389-395.

136. Yang Y, and Gabriel DW. 1995. Xanthomonas avirulence/pathogenicity gene family encodes functional plant nuclear targeting signals. MPMI 8:627-631.

111

137. Zaccardelli M, Saccardi A, Gambin E, and Mazzucchi U. 1992. Xanthomonas campestris pv. pruni bacteriophages on peach trees and their potential use for biological control. Phytopathol Mediterr 31:133-140.

138. Zansler ML, Spreen TH, Muraro RP. Florida's citrus canker eradication program (CCEP): Benefit-cost analysis. Electronic Data Information Source, IFAS, University of Florida, Online 2005 March. Report No.: FE 531.

139. Zansler ML, Spreen TH, Muraro RP. Florida's citrus canker eradication program (CCEP): Summary of annual costs and benefits. Electronic Data Information Source , IFAS, University of Florida; , Online 2005 March. Report No.: FE 532.

112

BIOGRAPHICAL SKETCH

Botond Balogh was born to Balogh Antal and Gyuris Borbála, in Kalocsa, Hungary, on

March 25, 1976. He graduated from the Ságvári Endre High School in Szeged in 1994 with

mathematics major. He attended the University of Agronomy, College of Agricultural Sciences,

in Gödöllő from 1994 to 1998, majoring in horticulture and plant breeding – genetics. He

participated in a research program for pepper breeding via anther and microspore culture under

the guidance of Judith Mitykó in the Agricultural Biotechnological Center, Gödöllő, from 1996

to 1998. In 1998 he took part in a year-long agricultural exchange program organized by the

Communicatin for Agriculture Exchange Program. In the summer and fall of 1999 he attended

Brevard Community College in Cocoa, FL. He was admitted to the University of Florida,

College of Agricultural and Life Sciences, in January 2000, and graduated with the degree of

Bachelor of Science with the plant science major and plant pathology minor in August 2000.

During his undergraduate studies he participated in a research project characterizing a mobile

genetic element found in the genome of several strains of Xanthomonas campestris pv.

vesicatoria, a plant pathogenic bacterium. From August 2000 he enrolled to the graduate

program of the University of Florida, College of Agricultural and Life Sciences, Department of

Plant Pathology. He graduated with a degree of Master of Scence in May 2002. The title of his

thesis was: “Strategies For Improving The Efficacy Of Bacteriophages For Controlling Bacterial

Spot Of Tomato”. He continued with his doctoral studies at the Plant Pathology department and

conducted a research project on evaluating bacteriophages for controlling citrus canker and citrus

bacterial spot.