Citation
Characterization and Use of Bacteriophages Associated with Citrus Bacterial Pathogens for Disease Control

Material Information

Title:
Characterization and Use of Bacteriophages Associated with Citrus Bacterial Pathogens for Disease Control
Creator:
BALOGH, BOTOND
Copyright Date:
2008

Subjects

Subjects / Keywords:
Bacteria ( jstor )
Bacteriophages ( jstor )
Diseases ( jstor )
Grapefruits ( jstor )
Leaves ( jstor )
Lesions ( jstor )
Pathogens ( jstor )
Tomatoes ( jstor )
Xanthomonas ( jstor )
Xanthomonas axonopodis ( jstor )
City of Miami ( local )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Botond Balogh. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
3/1/2007
Resource Identifier:
658230418 ( OCLC )

Downloads

This item has the following downloads:

balogh_b ( .pdf )

balogh_b_Page_014.txt

balogh_b_Page_013.txt

balogh_b_Page_037.txt

balogh_b_Page_015.txt

balogh_b_Page_045.txt

balogh_b_Page_098.txt

balogh_b_Page_058.txt

balogh_b_Page_066.txt

balogh_b_Page_052.txt

balogh_b_Page_043.txt

balogh_b_Page_046.txt

balogh_b_Page_019.txt

balogh_b_Page_007.txt

balogh_b_Page_090.txt

balogh_b_Page_032.txt

balogh_b_Page_071.txt

balogh_b_Page_057.txt

balogh_b_Page_075.txt

balogh_b_Page_083.txt

balogh_b_Page_049.txt

balogh_b_Page_068.txt

balogh_b_Page_086.txt

balogh_b_Page_080.txt

balogh_b_Page_084.txt

balogh_b_Page_004.txt

balogh_b_Page_072.txt

balogh_b_Page_061.txt

balogh_b_Page_109.txt

balogh_b_Page_059.txt

balogh_b_Page_060.txt

balogh_b_Page_016.txt

balogh_b_Page_039.txt

balogh_b_Page_067.txt

balogh_b_Page_020.txt

balogh_b_Page_021.txt

balogh_b_Page_036.txt

balogh_b_Page_030.txt

balogh_b_Page_005.txt

balogh_b_Page_003.txt

balogh_b_Page_026.txt

balogh_b_Page_064.txt

balogh_b_Page_104.txt

balogh_b_Page_008.txt

balogh_b_Page_035.txt

balogh_b_Page_093.txt

balogh_b_Page_100.txt

balogh_b_Page_094.txt

balogh_b_Page_108.txt

balogh_b_Page_069.txt

balogh_b_Page_047.txt

balogh_b_Page_053.txt

balogh_b_Page_012.txt

balogh_b_Page_097.txt

balogh_b_Page_056.txt

balogh_b_Page_081.txt

balogh_b_Page_017.txt

balogh_b_Page_105.txt

balogh_b_Page_018.txt

balogh_b_Page_044.txt

balogh_b_Page_102.txt

balogh_b_Page_033.txt

balogh_b_Page_050.txt

balogh_b_Page_054.txt

balogh_b_Page_092.txt

balogh_b_Page_089.txt

balogh_b_Page_063.txt

balogh_b_pdf.txt

balogh_b_Page_112.txt

balogh_b_Page_074.txt

balogh_b_Page_091.txt

balogh_b_Page_011.txt

balogh_b_Page_023.txt

balogh_b_Page_110.txt

balogh_b_Page_010.txt

balogh_b_Page_103.txt

balogh_b_Page_088.txt

balogh_b_Page_038.txt

balogh_b_Page_029.txt

balogh_b_Page_028.txt

balogh_b_Page_027.txt

balogh_b_Page_096.txt

balogh_b_Page_055.txt

balogh_b_Page_065.txt

balogh_b_Page_087.txt

balogh_b_Page_073.txt

balogh_b_Page_001.txt

balogh_b_Page_040.txt

balogh_b_Page_095.txt

balogh_b_Page_034.txt

balogh_b_Page_077.txt

balogh_b_Page_078.txt

balogh_b_Page_051.txt

balogh_b_Page_041.txt

balogh_b_Page_024.txt

balogh_b_Page_101.txt

balogh_b_Page_085.txt

balogh_b_Page_099.txt

balogh_b_Page_079.txt

balogh_b_Page_042.txt

balogh_b_Page_009.txt

balogh_b_Page_048.txt

balogh_b_Page_002.txt

balogh_b_Page_062.txt

balogh_b_Page_070.txt

balogh_b_Page_106.txt

balogh_b_Page_025.txt

balogh_b_Page_076.txt

balogh_b_Page_082.txt

balogh_b_Page_031.txt

balogh_b_Page_006.txt

balogh_b_Page_111.txt

balogh_b_Page_107.txt

balogh_b_Page_022.txt


Full Text





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

































Copyright 2006

by

Botond Balogh



































Unokatestv~rem, Dob6 Gdbor eml~kkre









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.









TABLE OF CONTENTS

page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

L IST O F T A B L E S ...................................................................................................... . 8

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

A B S T R A C T ............ ................... ............................................................ 1 1

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

The Citrus Industry .............. .. .................. ........................... .... ..... ......... 13
Sym ptom s, Etiology and Epidem iology ....................................................... .... ........... 14
D ise a se Im p act......................................................................... 16
Disease M management ......................................................... ........... ....... ...... 17
G enetic V ariation of the Pathogen........................................................................... ...... 17
Citrus Canker in Florida .................................... ... .. .......... ....... ..... 18
Citrus Canker Eradication Program ...................................... .......... ............................... 20
Bacteriophages Associated with Citrus Canker............................................. ...............21
Citrus Bacterial Spot............ ...... ....... ...................... ............ 22
P project G oal and O objectives ......................................................................... ....................23

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

E arly H isto ry ......................................................... .................. ................ 3 1
Other Uses of Phages in Plant Pathology .....................................................................32
R eturn of Phage-B ased D disease C ontrol..................................................................... ...... 32
Considerations About Phage Therapy .............................. ... ........... ............... .... 33
Factors Influencing Efficacy of Phages as Biological Control Agents .............................34
C current R research .............................................................................37
Outlook for the Future ................................. ......... ..... .................. 38

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

Intro du action ................... .......................................................... ................ 4 0
M materials and M ethods .................................... ..... .. .. .... ....... ......... 41
B acterial Strains and B acteriophages ........................................ ......................... 41
Standard Bacteriophage Techniques ........................................ .......................... 42
Phage Isolation from Diseased Plant Tissue ....................................... ............... 44
Phage Typing of 81 Xanthomonas Strains .............. ....... ....... ...............45
Electron M icroscopy .................................. .. .. .. ...... .. ............45
M molecular Techniques ...................................... .. .......... ....... .... 45









Results ................... ...... ........ ......... ........ ............... ........ 47
B acteriophage Isolations ....................... ...................... ........................................... 47
Classification of Isolated Phages Based on Chloroform Sensitivity, Plaque
M orphology, Host Range and Virion M orphology.....................................................47
Comparison of Florida Group A Phages and CP2 Based on Genome Size and RFLP
P ro fi le ................................. ................................. ................ 4 8
Phage Typing of 81 Xanthomonas Strains .............. ....... ....... ...............48
D iscu ssion ......... ........... .. .. .......... ..... ........................................... ...... 49

4 CONTROL OF CITRUS CANKER AND CITRUS BACTERIAL SPOT WITH
B A C TE R IO PH A G E S ......... .. ....... .............. ..........................................................63

Introduction ................. ......................................... ............................63
M materials and M ethods ............................ ........... ........ ....... .. ....... 65
B acterial Strains and B acteriophages ........................................... ....................... 65
Plant Material and Cultural Conditions in the Greenhouse...........................................65
Standard Bacteriophage Techniques ........................................ .......................... 66
Disease A ssessm ent and D ata 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
P ark F lo rid a .................................................... ... .. ........................................ 6 9
Citrus Bacterial Spot Disease Control Trials at the Plant Science Unit of the
U university of Florida in Citra, Florida. ................................... ........... ................... 69
R 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-R esistant M utant.................................................... .................71
Effect of Phage and Copper-Mancozeb Applications on Citrus Canker Disease
D evelopm ent in a Citrus N ursery ......................... ................ ............... .... 72
Effect of Bacteriophage Treatment on Citrus Bacterial Spot Disease Development
in a Com m ercial Citrus N ursery ........................ .............. ....... ............... .... 72
Effect of Bacteriophage and Copper-Mancozeb Treatments on Citrus Bacterial Spot
Disease Development in an Experimental Nursery ............................................. 73
D isc u ssio n ................... ........................................................... ................ 7 3

5 INTERACTION OF BACTERIOPAHGES AND THE HOST BACTERIA ON THE
PH YLLOPLANE ..... ............... ............................. .. .. ...... .............. .. 79

Introduction ................. ..................................... ............................79
M material and M methods ........ ...... .............................. ......................... 80
Standard B acteriophage Techniques .............. .... ................. .................. .... 80
Changes in Xanthomonas axonopodis pv. citrumelo Phage Populations on the Field....81
Interaction ofXanthomonasperforans and Its Bacteriophage on the Tomato
F oliage in the G reenhou se ........................................ .............................................82









Interaction ofXanthomonas axonopodis pv. citri and Its Bacteriophages on
G rapefruit Foliage in the G reenhouse.................................. ..................................... 83
R e su lts ...................................... ........................................ ..... ............... 8 5
Persistence of Bacteriophages on Citrus Leaf Surface Under Field Conditions ............85
Ability of Three Phages ofXanthomonas axonopodispv. citri to Multiply on
Grapefruit Foliage in the Presence of Their Bacterial Host, and Their Effect on
Citrus Canker D disease D evelopm ent ...................................................... ................. .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
D isc u ssio n ......... .... .... ..... .......... ......... ..................................................8 8
Overall Sum m ary and Conclusions ............................................... ............................. 89

APPENDIX

A CALCULA TION S ......... ....................................................... ............................ 98

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

L IST O F R EFER EN CE S .................................................................................................... 101

B IO G R A PH IC A L SK E T C H .............................................................................. ... ............ 112









LIST OF TABLES


Table page

3-1. B bacterial strains used in this study .............................................. .............................. 51

3-2. B acteriophages 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 (A U D P C ) ............................................................................. ........... ........ 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 (A U D P C ) ............................................................................. ........... ........ 78

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

5-2. Effect of a curative application of bacteriophage OXv3-3-13h on tomato bacterial spot
disease development, incited by Xanthomonasperforans strain AopgH: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









LIST OF FIGURES


Figure page

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

1-2. Severe citrus canker infection on K ey lim e.................................... ....................... .......... 25

1-3. Citrus canker lesions on lem on 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 A rgentina........................................................... ............... 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) OXaacAl, C)
OXaacFl, D) OXaacF2, E) OXaacF3, F) OXaacF5, G) OXaacF8...................................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. U ndigested bacteriophage D N A ........................................ .............................................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









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 Xanthomonasperforans ...........................95

5-4. Populations of bacteriophage (DXv3-3-13h on tomato foliage in the presence or absence
of its host Xanthomonas perforans strain AopgH:ME-B .............. .....................................96

A-1. Visualization of the disease 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









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









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.









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 400 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. medical 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 Sdo 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 xP. trifoliata)

(111).









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









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

temperature for disease development is 20-300C (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









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,









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 Sdo 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









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 Sdo 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









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, avrGfl, 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 medical) that was probably









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









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 ofpilA 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









in Taiwan, tested for their ability to lyse Taiwanese strains ofXac, 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 ofXac 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









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.


































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




































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
































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


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
































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


Figure 1-6. Covered citrus nursery in Argentina.




























































figure 1-/. Wmincreaks outline an orange grove in Argentina.























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


I


J
r


*- *1.*


.a.....


'" ~
1



















*. *

.:.'
*2
i '.: 'LI, .- '
a ..h` ,.
~t \ A:


4m



... ''. .


'.* .
**'. *. ri.*-
..* *, "* '- ** t
iad *:a


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


>









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

campestrispv. 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









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 ofXanthomonas 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 ofX 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









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.









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 20C. Vidaver (128) suggested that P.
syringae and P. phaseolicola, causal agents of halo blight and brown spot of bean, may be
more prevalent below 22C 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









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









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









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 Xanthomonasperforans, 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 ofErwinia 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 oflysing E.









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 u'1-hlinie'l ia, 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 ofbacteriophages for controlling plant diseases is an emerging field with great

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









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.









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









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 280C. 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

NaC1, 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 -800C in NB supplemented with

30% glycerol. Bacteriophages (Table 3-2) were stored at 40C and protected from light.









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 280C, the phages were eluted by

pouring 5 mL sterilized tap water into the 100 mmxl5 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 40C 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 [tL of concentrated

bacterial suspension in empty Petri dishes and then 12 mL warm (480C) 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 280C

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









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 280C 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









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 Tecnologia 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 ofbacteriophages by spotting 20 tiL 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 280C, the phage was purified by three

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









Phage Typing of 81 Xanthomonas Strains

Twelve phages were used in the phage typing study: a-MME, (5536, (DXacm4-11,

(DXv3-21, (XaacAl, (XaacFl, (DXaacF8, (Dccl9-1, (Dccl3-2, 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 unweightedd 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, 37C) to digest any contaminating bacterial nucleic acids. Subsequently, samples were

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

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









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

IL 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 iL and then the solution was subjected to chloroform-isoamyl alcohol

extraction once by adding 250 [L 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 [L

sodium acetate (3 M, pH 5.2) and 800 [L of cold (-40C) 95% ethanol to the tube, vortexing and

incubating at -800C for 30 min. The sample was then centrifuged (10,000 g, 20 min) at 40C 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 iL

sterile DI water and stored at 40C.

Digestion ofDNA / i/h restriction enzymes. One microliter of the phage DNA was mixed

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

in a 1.5 mL microcentrifuge tube. The mixture was incubated at 370C 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.









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, DXaacF8, was found to be filamentous, similar to members oflnoviridae (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,









OXaacAl, 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, OXaacF1 (group 1), OXaacF2 (group 7), OXaacF3 (group 2) and

OXaacF5 (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 ofX. 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









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 OXaacF 1 and OXaacF8 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 OXaacAl were identical and were almost identical to that of OXaacF 1: the only

difference was OXaacF 1 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









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.











Table 3-1. Bacterial strains used in this study


X axonopodis pv. aurantifolii
78-4-3-2-4B
XC 90

X axonopodis pv. citrumelo
Xacm 36
Xacm 45
Xacm 47
S4


Origin


Genotype a


Strain
X axonopodis pv. citri
1311
1319
1322
1604
1617
1635
2525
1528-7-3
1622-4
1660-1
306
94-358-1
BV38
BV42
RF2f
Xac5
Xac6
Xac7
Xac8
Xac15
Xac25
Xac30
Xac31
Xac41
Xac51
Xac65
Xac66
Xac71
Xc05-2592
XC62
XC63
XQ05-1-2


Pakistan
India
Brazil
Argentina
Argentina
Argentina
Argentina
Argentina
Argentina
Paraguay
Brazil
Argentina
Argentina
Argentina
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Argentina
Japan
Japan
Florida



Argentina
Mexico


Florida
Florida
Florida
Florida


Provided by

Canteros, B.I.b
Canteros, B.I.
Canteros, B.I.
Canteros, B.I.
Canteros, B.I.
Canteros, B.I.
Canteros, B.I.
Canteros, B.I.
Canteros, B.I.
Canteros, B.I.
Hartung, J.S.c
Canteros, B.I.
Canteros, B.I.
Canteros, B.I.
This study
Sun, X.d
Sun, X.
Sun, X.
Sun, X.
Sun, X.
Sun, X.
Sun, X.
Sun, X.
Sun, X.
Sun, X.
Sun, X.
Sun, X.
Sun, X.
Canteros, B.I.
Hartung, J.S.
Hartung, J.S.
Sun, X.



Canteros, B.I.
Hartung, J.S.


This study
This study
This study
This study


Miami
Miami
Miami
Miami
Manatee
Wellington
Miami
Miami
Miami
Manatee
Miami
Miami
Wellington
Wellington



Miami









Table 3-1. Continued
Strain
S4mg


Origin
Florida


Genotypea


X vesicatoria
MME
X perforans
AopgH:ME-B


Provided by
This study


Jones, J.B.e

Jones, J.B.


a Genotype ofX. axonopodis pv. citri strains isolated in Florida.
b Institute Nacional de Tecnologia Agropecuaria, Bella Vista, Corrientes, Argentina.
' 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.
SUniversity of Florida, Plant Pathology Department, Gainesville, FL.
fRF2 is a phage-resistant mutant of Xac65.
g Chloramphenicol and Rifamycin-resistant mutant of S4.










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.
OXaacF1 to 37 Xac65 citrus, aerial Florida This study
OXaacAl to 30 BV42 citrus, aerial Argentina This study
ccQ5 Xac65 Jackson, L.E.b
ccD7 Xac65 Jackson, L.E.
cc 13-2 Xac65 Jackson, L.E.
cc 19-1 Xac65 Jackson, L.E.
ccD24 Xac65 Jackson, L.E.
ccD25 Xac65 Jackson, L.E.
ccD28 Xac65 Jackson, L.E.
ccD35 Xac65 Jackson, L.E.
ccD36 Xac65 Jackson, L.E.
(XV3-21 Xac65 Jackson, L.E.
a-MME MME lake water Florida Minsavage, G.V.0
(5536 Xacm 47 Jackson, L.E.
DXv3-3-13h AopgH:ME-B Jackson, L.E.
DX44 Xacm47 Jackson, L.E.
PXacm4-4 Xacm36 citrus, aerial Florida This study
PXacm4-16 Xacm36 citrus, aerial Florida This study
PXacm4-11 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.










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 Xacl5
OXaacF 1 Ra clear, 2-3 -c ++ ++
OXaacF2 R clear, 2-3 ++ ++
OXaacF3 R clear, 2-3 ++ ++
(XaacF4 R clear, 2-3 ++ ++
(XaacF5 R clear, 2-3 ++ ++
(XaacF6 R clear, 2-3 ++ ++
(XaacF7 S turbid, 1 ++ -++ ++
(XaacF8 S turbid, 1 ++ -++ ++
ODXaacF6 R clear, 2-3 ++ ++ -
(XaacF7 S turbid, 1 ++ -- ++ ++
(XaacF 1 R clear, 2-3 ++ ++
(XaacF11 S turbid, 1 ++ -- ++ ++
(XaacF12 R clear, 2-3 ++ ++
(XaacF 13 S turbid, ++ ++ ++
(XaacF14 R clear, 2-3 ++ ++
(XaacF 15 R clear, 2-3 ++ ++
(XaacF 1 R clear, 2-3 ++ ++
(XaacF 17 R clear, 2-3 ++ ++
(XaacF 1 R clear, 2-3 ++ ++
(XaacF21 R clear, 2-3 ++ ++
(XaacF22 R clear, 2-3 ++ ++
(XaacF21 R clear, 2-3 ++ ++
OXaacF22 R clear, 2-3 ++ ++
(XaacF23 R clear, 2-3 + ++ +
(XaacF24 R clear, 2-3 ++ ++
(XaacF25 R clear, 2-3 + ++
(XaacF26 R clear, 2-3 ++ ++
(XaacF27 R clear, 2-3 ++ ++
(XaacF28 R clear, 2-3 ++ ++
(XaacF29 R clear, 2-3 ++ ++
(XaacF32 R clear, 2-3 ++ ++
(XaacF33 R clear, 2-3 ++ ++
OXaacF32 R clear, 2-3 ++ ++
(XaacF3 R clear, 2-3 ++ ++
(XaacF34 R clear, 2-3 ++ ++
(XaacF33 R clear, 2-3 ++ ++
(XaacF34 R clear, 2-34 ++ ++
(XaacF37 R clear, 2-34 ++ ++
(XaacA3 R clear, 3-4 ++ ++ ++
(XaacA4 R clear, 3-4 ++ ++ ++
OXaacA3 R clear, 3-4 ++ ++ ++
(XaacA4 R clear, 3-4 ++ ++ ++
(XaacA3 R clear, 3-4 ++ ++ ++
(XaacA4 R clear, 3-4 ++ ++ ++
(XaacA7 R clear, 3-4 ++ ++ ++
(DXaacA8 R clear, 3-4 ++ ++ ++ -
(DXaacA9 R clear, 3-4 ++ ++ ++











Table 3-3.



OXaacA10
(XaacA 11
(XaacA12
(XaacA13
(XaacA14
(XaacA15
(XaacA16
(XaacA17
(XaacA 18
(XaacA19
(XaacA20
(XaacA21
(XaacA22
(XaacA23
(XaacA24
(XaacA25
(XaacA26
(XaacA27
(XaacA28
(XaacA29
(XaacA30
CP1
CP2
CP3

32 phages
30 phages
6 phages
1 phage
1 phage


Continued
Chloroform
sensitivity
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R


clear
clear, 3-4
turbid, 1
clear, 1
clear, 4-5


Plaque typeb,
diameter (mm)
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 3-4
clear, 4-5
clear, 3
clear, 1


BV38
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++


Summary

++
++


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.


Xacl5


Host ranged
XC90 Xac41 Xac65
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++
++ ++

++ ++
-I-
















l0onm A

p
***:-
-'?
'(4k


100nr


100 nm


-
n. .
";1F


IF


S


10Onm


100nmF


S100 nm G
Figure 3-1. Transmission electron micrographs of representative phages. A) CP2, B) OXaacAl,
C) OXaacFl, D) DXaacF2, E) DXaacF3, F) DXaacF5, G) DXaacF8.









Table 3-4. Summary of morphological characteristics of phage virions.
Head
Phage Head shape (Lengtha, nm) Tail shape Classificationb
CP2 symmetrical 50.8 short Podoviridae
OXaacAl symmetrical 53.3 short Podoviridae
OXaacF symmetrical 58.3 short Podoviridae
DXaacF2 symmetrical 51.2 short Podoviridae
DXaacF3 symmetrical 56.6 short Podoviridae
DXaacF5 symmetrical 62.5 short Podoviridae
DXaacF8 Inoviridae
a Length measurement represents the average value obtained by measuring three virions per
phage.
Classification based on virion morphology as described by Ackermann Fig. 4.1 (1).













Tvning Phaaes


Percent Similarity
8 8


S1 2345


6 7 8 9 101112 Strain
,N2003-00012-7
XN2004-00225-2
XS2004-00159
XC2005-00344-1
fll XN2003-00012-3
fll XN2003-00013-2
X2001-00042
fIlff XN2003-00011-1
fIlff XN2003-00011-2
*fll XN2003-00011-8
fl l XC1 59
LMG 7484
S XS1999-00061
JH96E
SXC64
SJH96-2
S XCC88B
JH96C
m XC90
I *MN S4
I XME XC2005-00110
XC2000-00042
X2000-12884
X2001-00006
X2003-01008
XC165
X2001-00032
XC261
XN2003-01036
X2000-12862
flfm X2000-00071
lfm X2003-03516
SNU6B
flf X1999-12813
XC392
flf XCC03-1635
fl l X12000-00075
fl l X12000-00080
fl l X12000-00120
fl l X12006-00204
flf XS1999-00038
mmmmm XC214
flo XC100
flfm X2003-02912
lfm XC2002-00010
I--- -


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:(Xacm4-11, 2:DXV3-21, 3:ccD19-1, 4:D5536, 5:DXaacF8, 6:DCP2, 7:DXaacAl,

8:DXaacF1, 9:ca-MME, 10:CP3, 11:ccD13-2, 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.


Class
Xac A Etrog
Xacm
Xacm
Xac A
Xac A Etrog
Xac A Etrog
Xacm
Xac A Etrog
Xac A Etrog
Xac A Etrog
Xac A
Xa diffenbachlai
Xacm
Xaa B
Xaa B
Xaa B
Xaa B
Xaa B
Xaa D
Xacm
Xacm
Xacm
Xac A Wellingto
Xac A Wellingto
Xac A Wellingto
Xac A*
Xac A Wellingto
Xac A*
Xac A Wellingto
Xacm
Xac A Miami
Xac A Miami
Xac A
Xac A Miami
Xac A
Xac A
Xac A Miami
Xac A Miami
Xac A Miami
Xac A
Xac A Miami
Xac A
Xac A
Xac A Miami
Xac A Miami


Origin
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Pakistan

Florida
Argentina
Japan
Argentina
Argentina
Argentina
Mexico
Florida
Florida
Florida
n Florida
n Florida
n Florida
India
n Florida
Saudi Arabia
n Florida
Florida
Florida
Florida
Korea
Florida
Indonesia
Argentina
Florida
Florida
Florida
Florida
Florida
Talwan
Yemen
Florida
Florida


G1











G2










G3














G4











Typing Phages


Percent Similarity


1 2345 6 7 89101112 Strain Class Origir

m m m m XC2002-00010 Xac A Miami Florid
MEME M B o X12000-00194 XacA Miami Florid;
B MENNEN X2003-00008 XacA Miami Florid;
*fll XC63 Xac A Japan
SXC273 Xac A* Saudi
XN2003-01035 Xac A Wellington Florid;
SE XN2003-00012-1 Xac A Etrog Florid
EE E XN2003-00012-6 Xac A Etrog Florid
E E E 6B XaaB Argen
E E 7B Xaa B Argen
EE 306 XacA Brazil
EE X2000-00067 Xac A Miami Florid.
EE X11999-00112 Xac A Miami Florid.
E X12001-00098 Xac A Miami Florid.
10537C Xaa C Brazil
17C Xaa C Brazil
1C Xaa C Brazil
2C Xaa C Brazil
3C Xaa C Brazil
4C Xaa C Brazil
JH70C Xaa C Brazil
MME X. vesicatoria
10535C Xaa C Brazil
IAPAR9691/90 Xaa C Brazil
91-106 X. euvesicatoria
91-118 X. perforans
XC205 Xac A Taiwa
MEE EE mE 8B XaaB Argen
S E E XC427 XacA Thaila
ME ATCC 49118 XacA Florid.
SmE XC362 XacA Austra
SXC62 XacA Japan
XS1995-00051 Xac A Manatee Florid.
XS1997-00006 Xac A Manatee Florid.
XS1997-00018 Xac A Manatee Florid.
XS1997-00082 Xac A Manatee Florid.
XS2003-00004 Xac A Manatee Florid.


a G4


Arabia G5
a
a IG6
a
tina 1G7
tina

a G8
a
a




G9


n
tina
nd

lia IG10

a
a
a Gb
a
a


Figure 3-2. Continued







































4--


gure 3-3. Undigested bacteriophage DNA. Panel A: Lanes (from left to right) 1 :, 2:PXaacFl,
3:DXaacF2, 4:DXaacF3, 5:DXaacF4, 6:DXaacF5, 7:DXaacF6, 8:DXaacF10,
9:DXaacF12, 10:DXaacF14, 11:DXaacF15, 12:DXaacF16, 13: XaacF17, 14:X.
Panel B: 1:X, 2:CP2, 3:DXaacF18, 4:DXaacF19, 5:DXaacF20, 6:DXaacF21,
7:DXaacF22, 8:DXaacF23, 9:DXaacF24, 10:DXaacF26, 11l:DXaacF27,
12:DXaacF28, 13:X, 14: DXaacF29. Panel C: 1:X, 2:DXaacF30, 3:DXaacF31,
4: XaacF32, 5: XaacF33, 6: XaacF34, 7: XaacF35, 8: XaacF36, 9: XaacF37,
10:X, 11 :PXaacF2. Arrows indicate size marker 21,226 bp.











A--


Figure 3-4. Bacteriophage DNA digested with BamH I. Panel A: Lanes (from left to right) 1 :,
2:DXaacFl, 3:DXaacF2, 4:DXaacF3, 5:DXaacF4, 6:DXaacF5, 7:DXaacF6,
8: XaacF10, 9:DXaacF12, 10:DXaacF14, 11:DXaacF15, 12:DXaacF16,
13: XaacF17, 14:X. Panel B: 1:X, 2:DXaacF18, 3:DXaacF19, 4:DXaacF20,
5: DXaacF21, 6: DXaacF22, 7: DXaacF23, 8: DXaacF24, 9: DXaacF26, 10:CP2,
11:DXaacF27, 12:DXaacF28, 13: DXaacF29, 14: DXaacF30, 15: DXaacF31, 16:X.
Panel C: 1:DXaacF32, 2:DXaacF33, 3:DXaacF34, 4:DXaacF35, 5:X, 6:DXaacF36,
7:PXaacF37, 8: XaacF2. Arrows indicate size marker 21,226 bp.









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
OXaacF 1 OXaacF3 OXaacF5 OXaacF27 OXaacF31 OXaacF34 OXaacF2
OXaacF6 OXaacF4 OXaacF32 OXaacF36
OXaacF 12 OXaacF 10 OXaacF33 OXaacF37
OXaacF 15 OXaacF 14
OXaacF 16 OXaacF 17
OXaacF 18 OXaacF30
OXaacF 19
OXaacF20
OXaacF21
OXaacF22
OXaacF23
OXaacF24
OXaacF26
OXaacF28







A g*


gure 3-5. Representatives of the 7 RFLP groups based on BamH I. digestion profile. Lanes
(from left to right) 1:X, 2:DXaacF24 (group 1), 3:OXaacF30 (group 2), 4:OXaacF5
(group 3), 5:OXaacF27 (group 4), 6:OXaacF31 (group 5), 7:OXaacF34 (group 6),
8:OXaacF2 (group 7), 9:X. Arrow indicates size marker 21,226 bp.









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.









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 Tecnologia 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.









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 280C. For preparation of

bacterial suspensions 24 h cultures were suspended in sterile tap water, their concentration was

adjusted to 5x 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 NaC1, 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 -800C in NB supplemented with 30% glycerol.

Bacteriophages were stored at 40C 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 25-30C.









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 280C 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









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 i ith phage mixture applied iith or i ilthIut 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, ccD7 and cce13-2 (test 1 -

5 x 109 PFU/mL, test 2 1 x 109 PFU/mL, test 3 1 x 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 x 108 cfu/mL, tests 2 and 3 5 x 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.









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 OXaacF2, was

found also to be resistant to CP2, CP31, ccD7 and ccD13-2. In three tests grapefruit plants were

inoculated either with Xac65 or RF2, at 5x 106 cfu/mL 12 h after treatment with the mixture of

CP2, CP31, ccD7 and ccD13-2, each at Ixl08 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

Tecnologia 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

(OXaacAl, ~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.









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 (OXacm2004-4,

OXacm2004-16 (both isolated from the nursery), and 0X44 (OmniLytics Inc., Salt Lake City,

UT), 2.4 x 108 PFU/mL each) blended in a suspension of 0.75% (wt/V) non-fat dry milk powder

using a non-C02, 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









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, (ccD13-2, a-MME and 05536, x 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









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, ccD7 and cce13-2) 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









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

OXaacAl; 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).









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









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

amount 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.


































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


C~ '.'.i,-*0acl.


". 'i.




'a i -

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


























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









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 x 108 cfu/mL in test 1 and 5x106 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 sranefmnit nlants in the greenhouse


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 py. citri strain Xc05-2592 in
nursery trial in Bella Vista Corrientes Argentina


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.


________ -- -- --- -,-- c'o -rj ------ ------ -_. -______________
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


Treatment Disease intensityz
Treatment p
UTCY'x Phage Cu-Mzv Phage+Cu-Mz
Test 1 2.6a 2.0b 1.6c 1.8bc <0.0001
Test 2 3.la 2.1b 1.6c 2.0b <0.0001










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)


SSites 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)


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.


AUDPC
Phage Control pz
Site 1y 173 331 0.0023
Site 2 169 261 0.0357
Site 3 128 154 0.1585


AUDPC
Treatment AUDPC p
_UTCYz Phage Cu-Mzw Phage+Cu-Mz
Test 1 0.53a 0.20b 0.08b 0.13b 0.0115
Test 2 24.3a 14.9b 4.0c 2.0c <0.0001









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









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 280C for 18 h. Then the culture was sterilized, enumerated and stored

at 4C 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 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 ccD13-2, a-MME and 05536, all adjusted to xl 10

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 [iL of chloroform. The tubes were

stored at 40C until all samples were taken and processed and then transported to the Gainesville









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 a-MME, Xacm45

selectively detected phage (5536, and S4m detected all three phages. Titers of phages a-MME

and (5536 were calculated from the plaque number, dilution information and leaf weight and

were expressed as PFU/g leaf tissue. Titer of phage cc(13-2 was determined indirectly, as

described. The efficacy of plating of phage a-MME on strain MME versus strain S4m was

determined concurrently to phage titer determinations by plating known concentration of phage

a-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 (5536

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 ccq(13-2 in each sample was

determined by the following equation: cc(13-2 concentration = total phage concentration (as

determined on S4m) a-MME concentration (calculated) (5536 concentrated (calculated).

Finally, the data were converted by the y=logio(x+l) 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

ofX. perforans strain, AopgH: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 (DXv3-3-13h (108 PFU/mL). The plants were









placed in clear polyethylene bags for 1 day after phage application in a growth chamber (280C).

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 DXv3-3-13h was also used for controlling tomato bacterial

spot severity incited by AopgH:ME-B. Six young tomato plants were inoculated with strain

AopgH:ME-B; and later three of them were sprayed with DXv3-3-13h 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









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 (PXV3-21, OXaacF1 or cce19-1, 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 x 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 OXV3-21, OXaacF and

cce19-1 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 x 106









PFU/mL) and immediately followed by application of the bacterial suspension at a much higher

concentration (1 x 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 OXaacFl; Xacl5 for cce19-1 and Xac30 for OXv3-21.

Results

Persistence of Bacteriophages on Citrus Leaf Surface Under Field Conditions

The fate of three phages, ccD13-2, a-MME and 05536, 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 x 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).

05536 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,

03356 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 cce13-2 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 D3356, 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 cce13-2 at site 1 and site 2, respectively.

DMME populations were lower from the beginning compared to the other two phages (Figure









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, QDMME 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

OXV3-21 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 OXV3-21 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 OXaacF 1 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, OXaacF 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,

cce19-1 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 cc019-1

populations were higher in the presence of the host than in its absence (35% vs. 9.2%); however









at nine hours no cc019-1 was detected in the presence of the host (Figure 5-2C). In the second

experiment the presence of Xac65 cce19-1 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, OXV3-21 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: cce19-1-treated plants had significantly

less disease than DXV3-21-treated plants and significantly more disease than OXaacFl-treated

ones (Table 5-1). It was suspected that the nutrient broth present in phage suspensions

contributed to the disease increase on the OXV3-21 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 OXaacF1 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, PXv3-3-13h, to multiply on tomato foliage in the presence of

its host, AopgH:ME-B, was evaluated in greenhouse studies. DXv3-3-13h 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

AopgH:ME-B DXv3-3-13h populations increased 11 and 87 fold in the first 24 h, in experiments









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 PXv3-3-13h was evaluated for its ability to reduce bacterial spot of tomato disease

development incited by AopgH:VME-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.









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









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.














7 05536

6 ---. ---. MME
v -' ...-- ccq13-2
S. -- V\ 5 .....,
., 4
S3
a-
2 -


0
0 -
8:15 PM 7:45 AM 9:00 AM 10:00 AM 11:00 AM


B
7 05536

6 ---O--- MME
5 N- cc, 13-2
5 .
4 4
3 .. -.
a-
2 2
0 1

0
0 -----------------
8:15 PM 7:45 AM 9:00 AM 10:00 AM 11:00 AM



C 700
600

E 500
400
300
200

100
0






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.
























Hours after inoculation


C 160
C 160 ---o--- cc019-1
140 ------- cc. 19-1 with host
S120
S100
C.
o 80 -
60
40
20 -
0 + .....

0 3 6 9
Hours after inoculation

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 (XV3-21. B) Phage OXaacF 1. C) Phage ccD 19-1. Percent population =
percent of population average at 0 hour. Error bars indicate the standard error.
























Hours after inoculation


B 900
800
c 700
o
S600
500
400
8 300
200
100
0


C 200
180-
160 -
140 -
*5 120 -
C.
o 100-
80-
S60-
0 40-
20 -
0


- qXaacF1
- OXaacF1 with host








h _:::_--'-I
~------ -- -


-iI


Hours after inoculation


Hours after inoculation

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 (XV3-21. B) Phage OXaacF 1. C) Phage ccD 19-1. Percent population =
percent of population average at 0 hour. Error bars indicate the standard error.


---o--- ccQ19-1
---o--- ccq19-1 with host







.' ,.-
*^~.~~~....,-;..v.,,,;;:;:,









Table 5-1. Effect of treatment with three phages on citrus canker disease development, incited
by Xanthomonas axonopodispv. citri strain Xac65, as measured by disease severity.
Average disease severity
Treatment Controly ccC 19-1 OXaacF 1 XV3-21 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.

































Figure 5-4. Symptoms of tomato bacterial spot, incited by Xanthomonasperforans.











































1 2 3 4 5 6 7 8
Days after application


Figure 5-4. Populations of bacteriophage PXv3-3-13h on tomato foliage in the presence or
absence of its host Xanthomonasperforans strain AopgH:ME-B. A) Experiment 1. B)
Experiment 2. Error bars indicate the standard error.










Table 5-2. Effect of a curative application of bacteriophage OXv3-3-13h on tomato bacterial
spot disease development, incited by Xanthomonasperforans strain AopgH: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.









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-i). 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-i) 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-i).

The area under this curve consists of several trapezoids. The area of each trapezoid is determined

as (y2+y1)/2*(x2-x1), where (xl,yi) 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.











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
SMidpoint values are as published in the Elanco Conversion Tables for Barratt-Horsfall Rating
Numbers (ELANCO Products Co., Indianapolis, IN).















30

25
u,
S20

S15

10
0(


0 7 14
Days after first rating


30

25

S20

" 15
c-

2 10
a.


Days after first rating


Figure A-i. Visualization of the disease 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.


-Area Under the Disease Progress Curve




Full Text

PAGE 1

1 CHARACTERIZATION AND USE OF BACTER IOPHAGES ASSOCIATED WITH CITRUS BACTERIAL PATHOGENS FOR DISEASE CONTROL By BOTOND BALOGH A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2006

PAGE 2

2 Copyright 2006 by Botond Balogh

PAGE 3

3 Unokatestvrem, Dob Gbor emlkre

PAGE 4

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. Gu lig, for their support, co nstructive 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 Maruta ni, Fanny Iriarte, Ellen Dickst ein, Jerry Minsavage, Nelly Canteros, Alberto Gochez, Debra Jones, Xiao an 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, Ga ry Marlow, Patricia Rayside, Mark Ross, Eldon Philman, Vanessa Ivanovski, Chandrika Ramadugu and the ones I forgot to mention. I would like to thank a ll those who, while not c ontributing directly to the project, helped me during the time I was working on it: Abby Guerra, Gail Harris, Jim Barrel and Aleksa Obradovic.

PAGE 5

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 Phag es as Biological Control Agents..................................34 Current Research............................................................................................................... .....37 Outlook for the Future......................................................................................................... ...38 3 CHARACTERIZATION OF BACTERIOP HAGES 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

PAGE 6

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 a nd 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 Cank er Control Trials.......................................................................67 Citrus Canker Nursery Trials in Argentina.....................................................................68 Citrus Bacterial Spot Disease Control Tr ials 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 A pplication and Use of Protective Skim Milk Formulation on Citrus Canker Disease Deve lopment 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 A pplications on Citrus Canker Disease Development in a Citrus Nursery................................................................................72 Effect of Bacteriophage Treatment on Citr us Bacterial Spot Disease Development in a Commercial Citrus Nursery..................................................................................72 Effect of Bacteriophage and Copper-Manco zeb Treatments on Citrus Bacterial Spot Disease Development in an Experimental Nursery.....................................................73 Discussion..................................................................................................................... ..........73 5 INTERACTION OF BACTERIOPAHGES A ND 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

PAGE 7

7 Interaction of Xanthomonas axonopodis pv. citri and Its Bacteriophages on Grapefruit Foliage in the Greenhouse..........................................................................83 Results........................................................................................................................ .............85 Persistence of Bacteriophages on Citr us Leaf Surface Under Field Conditions.............85 Ability of Three Phages of Xanthomonas axonopodis pv. citri to Multiply on Grapefruit Foliage in the Presence of Th eir 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 E ffect on Tomato Bacterial Spot Disease Development................................................................................................................87 Discussion..................................................................................................................... ..........88 Overall Summary and Conclusions........................................................................................89 APPENDIX A CALCULATIONS..................................................................................................................98 Conversion of Horfall-Barratt Va lues to Mean Percentages..................................................98 Calculation of Area Under th e Disease Progress Curve.........................................................98 LIST OF REFERENCES.............................................................................................................101 BIOGRAPHICAL SKETCH.......................................................................................................112

PAGE 8

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 mo rphology and host range of bacteriophages originating in Florida, Argentina and Japan.........................................................................54 3-4. Summary of morphological char acteristics of phage virions................................................57 3-5. Grouping of Florida bacteriophages based on Bam H I RLFP profile...................................62 4-1. Effect of bacteriophage treatment and us e 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 pl ants 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 Vist a, Corrientes, Argentina...............................................................77 4-4. Effect of bacteriophage treatment on citr us bacterial spot dise ase 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 citr us canker disease development, incited by Xanthomonas axonopodis pv. citri strain Xac65, as measured by disease severity..............94 5-2. Effect of a curative a pplication of bacteriophage Xv3-3-13h on tomato bacterial spot disease development, incited by Xanthomonas perforans strain opgH:ME-B, as measured by area under the dis ease progress curve (AUDPC)............................................97 A-1. Horsfall-Barratt scale, the corresponding disease intervals and midpoint values................99

PAGE 9

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 le sions 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 or ange grove in Argentina................................................................28 1-8. Citrus bacterial spot le sions on grapefruit leaves..................................................................29 1-9. Citrus bacterial spot lesi ons in Asian leafminer tunnels........................................................30 3-1. Transmission electron micrographs of representative phages. A) CP2, B) XaacA1, C) XaacF1, D) XaacF2, E) XaacF3, F) XaacF5, G) XaacF8......................................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 Bam H I..........................................................................61 3-5. Representatives of the 7 RFLP groups based on Bam H I. digestion profile.........................62 4-1. Experimental citrus nursery, the location of the 2006 citrus canker trials, at the INTA research station in Bella Vi sta, 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 Resear ch and Education Unit of the University of Florida in Citra, Florida...................................................................................................... ..76 5-1. Bacteriophage populations a nd sunlight irradiation in Citr a, FL on July 18 and 19, 2006...91 5-2. Changes in bacteriophage populations on grap efruit foliage in the presence or absence the host, Xanthomonas axonopodis pv. citri strain Xac65....................................................92

PAGE 10

10 5-3. Changes in bacteriophage populations on grap efruit 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 Xv3-3-13h on tomato foliage in the presence or absence of its host Xanthomonas perforans strain opgH:ME-B......................................................96 A-1. Visualization of the disese progress cu rve 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

PAGE 11

11 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CHARACTERIZATION AND USE OF BACTER IOPHAGES 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 di seases in the world. Citrus canker was reintroduced to Florida in the 1990s and threatens the stat es $9 billion citrus industry. This work focused on a biological control approach to us e bacteriophages for reducing bacterial pathogen populations and disease severity on citrus. Bacteriophages isolated fr om citrus canker lesions in Florida and Argentina were evaluated based on plaque morphology, chloroform sensitivity, host range, genome size, DNA restriction profile a nd 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 a nd in nursery trials in Argentin a. Bacteriophages reduced citrus canker disease severity both in greenhouse and fiel d trials. The level of c ontrol was inferior to chemical control with copper bactericides. The combination of bacteriophage and copper treatments did not result in incr eased control. Citrus canker fiel d trials in Florida have been prohibited until recently, as the disease was under er adication. For this reason we evaluated the efficacy of phage treatment on a similar bacterial ci trus disease, citrus ba cterial spot, incited by X. axonopodis pv. citrumelo. Bacteriophages reduced citrus bact erial spot severity. The level of

PAGE 12

12 control was equal or inferior to chemical control with copper b actericides. The combination of bacteriophage and copper treatments did not re sult in increased control. In experiments monitoring the fate of bacteriophages on the citr us foliage following b acteriophage application, phage populations stayed steady on the foliage du ring nighttime but were drastically reduced within hours after sunrise. The rate of reduc tion varied among the phages. The ability of bacteriophages to multiply on the plant foliage in the presence of th eir 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.

PAGE 13

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 a nd South 40 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 produc tion (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 proce ssed (122). Most of th e processing is orange juice production that is carried out almost exclusiv ely in So Paulo state of Brazil and in Florida (122). Citrus is a perennial evergreen with ar ound 50 years of commerci al production (121). While originally it was grown on its own root system, in modern commercial production the producing plants are grafted onto r ootstocks (121). Rootstocks infl uence adaptation to soil types, tree size, production characteristic s and can provide cold hardine ss and resistance to diseases, nematodes and insects (53, 54, 105, 111). Some impor tant rootstock specie s are sour orange ( C. aurantium L.), trifoliate orange ( Poncirus trifoliata ) and citrumelo ( C. paradisi P. trifoliata ) (111).

PAGE 14

14 The total citrus production of the United Stat es was 11.6 million tons ($2.68 billion value) in the 2005-06 growing season (125 ). Florida is the biggest ci trus producer in the country providing 68% of the total production, followed by California (28%), Texas (4%) and Arizona (4%). Floridas $9 billion citrus industr y (49) is in a rath er precarious situa tion presently. It was heavily hit in 2004 and 2005 by hurricanes; as a re sult 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 eff ect of the two consecutive extreme hurricane seasons was the large scale dissemination of th e citrus canker bacterium throughout the state, which eliminated hopes of eradica ting the disease (34). The permanent presence of citrus canker is expected not only to reduce pr oduction volume, increase prices and cause market losses (139), but also to completely eliminate grapefruit pr oduction from the state (49). Moreover, citrus greening, another devastating citrus disease, was detected in Fl orida 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 citr us production worldwide (49). Its center of origin is the southeast Asia-India region (21), si milar to its host plants. By the 20th century it was present in most citrus gr owing 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 fru it, 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

PAGE 15

15 appear around 7 days after inoculation unde r optimum conditions (Figure 1-1). Optimum temperature for disease development is 20-30C (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 surrou nded by a chlorotic halo (Figure 1-2) (49). The center of the lesion becomes raised and cor ky and the touch of the lesion s feels like sand paper (106). Eventually the lesions form a crater-like appearan ce (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, Pt hA, is responsible for inciting the symptoms (30). It is delivered into the cytoplasm of pl ant cell by the type 3 secretion system of the pathogen, from where it is transporte d 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 surfa ce. Rain water collected from diseased leaves contains high bacterial c oncentrations 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 mete rs 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 equipmen t (50). Extreme weather conditions, such as hurricanes and tornadoes have also been shown to facilitate long di stance disease spread (51,57). Bacteria enter the tissues through stomat al openings (48,49) or wounds. Wind speeds above 18 mph aid in penetrati on of bacterial cells through th e stomatal pores (48). Wounding

PAGE 16

16 often occurs by thorns, blowing sand or insect damage (49). Asian leafminer ( Phyllocnistis citrella ), an insect that was introduced to Flor ida in 1993, creates wounds that expose the mesophyll tissues allowing entry of bacterial inocul um (Figure 1-4, 1-5) (51). Leafminer itself is not a vector of the disease (17) but its actions can lead to si gnificant field in fection 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, o ccurs in lesions in woody tissues (49). Outside of plant tissues the bacterium is quickly eliminat ed by desiccation or sun light 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 susceptibil ity 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 th e second half of the expansion phase of growth (108). Disease Impact Citrus canker impacts the citrus industry on se veral 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 bl emishes 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 ba n importation of fruits from canker inflicted areas because of the fear of introduction of the pathogen (50). Furthermore, if citrus canker becomes endemic, commercial produc tion of the most susceptible citrus species,

PAGE 17

17 such as grapefruit, will cease as it becomes impossible to grow them profitably (50). More resistant species such as tangerine s will likely take their place (71,103). Disease Management Countries that are free of ci trus canker apply quarantine m easures 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). In troduction of the diseas e 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 Afri ca, the Fiji Islands a nd twice from the US, while eradication efforts failed in Argentina, Uruguay, Paraguay, and more recently in Florida (98,103,106). The disease is currentl y under active eradication in Australia and So Paulo state in Brazil (39,103). In endemic situations the em phasis has shifted to implementing integrated management programs (76). The disease is under ma nagement in all Asian countries, Argentina, Uruguay, Paraguay, several states of Brazil a nd Florida (49,76,109). These programs rely on planting resistant citrus cultivars (76), production of disease-free nursery stock by locating nurseries out side of citrus canker areas and/or indoors (Fig ure 1-6) (76), and restricting disease spread by establishing windbreaks (Fi gure 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 insp ected for the presence of citrus canker and sanitation protocols are establishe d in the packing houses (75). Genetic Variation of the Pathogen Citrus canker is not a diseas e caused by a single pathogen but rather a of group similar diseases caused by closely related Xanthomonas species (49). The Asiati c type (Canker A) is

PAGE 18

18 caused by X. axonopodis pv. citri ( Xac ) strains that originated fr om Asia. This is the most geographically widespread pathoge n, which also has the widest host range and by far the biggest impact. Almost all commercially grown citrus va rieties are susceptible to this bacterium to a certain level. Grapefruit, Key lime and lemon ar e most susceptible. Sweet oranges range from highly susceptible (Hamlins and Navels) to modera tely 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 ca nker (cancrosis B or fals e 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. Th e C type of citrus canker (cancrosis C) is also caused by X axonopodis pv aurantifolii It has only been found in So 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 sout hwest Asia in the 1990s (126). Th ese 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 dist inguished by host range, cultural a nd physiological characteristics (49), bacteriophage sensitivity (21), serology (6), plasmi d fingerprints (94), DNA-DNA homology (32), and by various RF LP and PCR analyses (26). Citrus Canker in Florida Citrus canker was first introduced into Flor ida in 1912 on rootstocks imported from Japan (29). It spread to all Gulf states between Texa s and South Carolina. A st rict eradication program was implemented and the diseas e was eliminated from Florida by 1933 and from the entire United States by 1947 (29). The second erad ication program began in 1984, when Xanthomonas was isolated from lesions from nursery stock (10 2,103). Research later showed that this pathogen

PAGE 19

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 cancell ed in 1990 (103). By that time 20 million citrus plants were destroyed and $94 million was spen t (103). In 1986 genuine citrus canker was discovered in the Tampa Bay area. An eradi cation effort was begun and the pathogen was declared eradicated in 1994 (103). Another in troduction was discovered in urban Miami in 1995. The pathogen, called Miami genotype, is dete rmined 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 stat e despite the eradication progr am (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 Manate e genotype, was identical to Xac strains from China and Malaysia, based on rep-PCR analys is (26). Also, it was indistingui shable from Xac strains that caused the outbreak in the 1980s in the Tamp a Bay region (26,103), implying that the earlier eradication effort was not successful. Interestin gly, the Manatee genotype disappeared again in 1999 and then reappeared in 2005 (1 13). In 2000, a new genotype of Xac was identified and was designated the Wellington genotype. It was discovere d 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 determin ed that the inability of this genotype to infect grapef ruit was due to the presence of an avirulence gene, avrGf1 in its genome (98). The latest introduction of an exo tic strain was in 2003 in Orange County. These strains were found in a residentia l area on an Etrog citron tree ( Citrus medica ) that was probably

PAGE 20

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 establishe d 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 Heal th 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 (4 9). Citrus trees infected by the disease or located within the exposure area were uprooted and burned in the co mmercial groves, or cut down and chipped in urban areas (49). The exposure ar ea 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 th e 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 sp read the pathogen fr om areas awaiting eradication widely throughout the citrus growing area. On January 10, 2006, APHIS discontinued the CCEP (24,123), because its cont inuation in the new situation was judged infeasible (i) due to financial constraints asso ciated with the tree removal and reimbursement programs, and (ii) because the citrus industr y could not survive the loss of such a large production area (24,123). During th e program 16.5 million trees we re 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 Ma rch 2006 APHIS released a Citrus Health

PAGE 21

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 can ker situation in the Florida is estimated at $254.2 million annually (139) and may result in e limination of grapefruit production (49). The acceptance of citrus canker as an endemic diseas e will also result in losses in interstate and international commerce of the states fresh citrus fruit, which currently represents 20% of the states $9 billion dollar citrus industry (88). Bacteriophages Associated with Citrus Canker There are several reports on bacteriophages found in associat ion 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 pa thogen (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 stra ins 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 na rrow 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

PAGE 22

22 in Taiwan, tested for their ability to lyse Taiwanese strains of Xac, lysed 97.8% of them when used in combination (135). Thes e phages, however, did not lyse Xanthomonas strains that did not cause citrus canker, or any othe r bacteria tested. The authors c oncluded 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 pl aques with irregula r 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 canke r lesions in Argentina in 1979 (R.E. Stall, personal communication). Citrus Bacterial Spot Citrus bacterial spot (CBS) wa s 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 th an citrus canker (106). The pathogen, Xanthomonas axonopodis pv. citrumelo ( Xacm ) only exists in Florida (103) a nd 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 th e 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 pa thogen at all. The most aggressi ve strains are disseminated by wind driven rain, whereas strain s of intermediate and low aggressiveness are spread mainly by mechanical means (56). Wounding caused by thorns a nd citrus leafminer (Figure 1-9) facilitates pathogen entry and increases CBS disease incide nce and severity. The di sease causes a reduction in photosynthetically active leaf area and in severe cases can induce leaf drop. Xacm infection

PAGE 23

23 also causes bud failure in Swingle citrumelo plants (55). The CBS bacteria are most aggressive on trifoliate orange, Swingle ci trumelo and grapefruit (103,106). Project Goal and Objectives The goal of this project was to develop a bact eriophage-based disease control strategy that could be used as part of an integrated manageme nt program against citrus canker in Florida. The objectives were (i) establishment of a bacteriophage collection against Xac strains present in Florida, (ii) collection and char acterization of bacteriophage asso ciated with citrus canker in Florida and Argentina; (iii) determin ation 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 experiment s in Argentina, and for suppression of citrus bacterial spot, as a model disease syst em, in field experiments in Florida.

PAGE 24

24 Figure 1-1. Young citrus ca nker lesions on grapefruit.

PAGE 25

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

PAGE 26

26 Figure 1-3. Citrus canker lesions on lemon fruit. Figure 1-4. Asian leafminer ( Phyllocnistis citrella ) tunnels on Swingle citrumelo foliage.

PAGE 27

27 Figure 1-5. Citrus canker infect ion in Asian leafminer tunnels. Figure 1-6. Covered citrus nursery in Argentina.

PAGE 28

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

PAGE 29

29 Figure 1-8. Citrus bacterial s pot lesions on grapefruit leaves.

PAGE 30

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

PAGE 31

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 dHerelle in 1917 (112). There were differences in interpretation about the nature and origin of this lyt ic principle. Twort proposed that a bacterial enzyme caused the lysis, while dHerelle speculated that a viru s was responsible for the phenomenon. A direct consequence of dHerelles concept was the idea of using phages for controlling bacterial diseases. Soon after the first medical (112) a nd veterinary (28) appl ications, phages were evaluated for control of plant diseases. In 1924 Mallman and Hemstreet (79) is olated the cabbage-rot organism, Xanthomonas campestris pv. campestris from rotting cabbage and demonstrat ed 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) isolat ed bacteriophages from soil samp les that were active against the causal agent of blackleg disease of potato, Erwinia carotovora subsp. atroseptica They demonstrated in growth chamber e xperiments that co-inoculation of E. carotovora subsp. atroseptica with phage successfully inhibited the pat hogen and prevented rotting of tubers (73). These workers also isolated phages against Erwinia carotovora subsp. carotovora and Agrobacterium tumefaciens from various sources such as so il, rotting carrots and river water (25). Thomas (120) treated corn seeds that were infected with Pantoea stewartii the causal agent of Stewarts 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 re liable and effective means of c ontrolling phytobacteria. Several workers questioned if positive re sults were possible. In 1963, Ok abe stated, in general, the

PAGE 32

32 phage seems to be ineffective for [controlling] the disease developmen t (91). Three decades later, Goto concluded, practical use of phages for control of bact erial plant disease in the field has not been successful. (44). Chemical contro l with antibiotics and copper compounds became the standard for controlling b acterial plant diseases (31,81). Other Uses of Phages in Plant Pathology Bacteriophages still remained in use in plant pathology and ha ve been used as tools for detection, identification, classifi cation, and enumeration of pathoge nic bacteria and were also used for disease forecasting. Phage typing, as a method of differentiati ng different races or pathovars of the same bacterial species becam e 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 iden tification 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 fr om 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 quantif ying bacterial cells based on the number of newly produced phages and averag e burst size. They developed a method for indirectly forecasting bacterial leaf blight by monitoring phage titers in rice fields (91). Reliability of this la tter method was questioned later by Civerolo (22). Return of Phage-Based Disease Control Several factors have contributed to the re-e valuation 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

PAGE 33

33 transfer of antibiotic resistan ce 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 ca used 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 achie ving sustainable agriculture necessitated development of safer, more specific and enviro nment-friendly pesticides (96). These factors, together with the expanding knowledge base abou t 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 seve ral 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 bi osphere; they can readil y 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 agains t bacterial receptors that are essential for pathogenesis, so resistant mutants would be at tenuated 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 lega l regulations, such as for treatment of peach fruit before harvest (137) or for control of human pathogens in fr esh-cut produce (77,78). 5. Phages are specific or highly discriminator y, eliminating only targ et bacteria without damaging other, possibly beneficial members of the indigenous fl ora. Thus their use can also be coupled with the application of antagonis tic bacteria for increased pressure on the pathogen (117); or they can be us ed to promote a desired strain against other members of the indigenous flora (14). 6. Phage preparations are fairly easy and inexpens ive to produce and can be stored at 4 C in complete darkness for months without significan t 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-mixe d with them without si gnificant loss in titer. Copper-containing bactericides have been s hown to inactivate phages (9,66); however, inhibition was eliminated if phages were appl ied at least three da ys 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.

PAGE 34

34 1. Limited host range can be a disadvantage, as of ten 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 re nder phage treatment ineffective. However, using mixtures of phages that utilize distinct cell receptors can suppress the emergence of resistance (118). Also, phage resistance often co mes at some metabolic cost to the bacteria. Loss of virulence was observed w ith phage-resistant mutants of Ralstonia solanacearum (61), Xanthomonas campestris pv. pruni (95), and Pantoea stewartii (120). 4. Environmental effects, such as temperatur e, pH and physiology of bacterium can hinder control. Civerolo (22) observed that Xanthomonas phaseoli phages attacked Xanthomonas and Pseudomonas species only at temperatures above 20C. Vidaver (128) suggested that P. syringae and P. phaseolicola causal agents of halo blight and brown spot of bean, may be more prevalent below 22C because of phage resistance. Leverentz (78) noted that phage treatment caused a significan t population reduction of the Listeria monocytogenes on melons but not on apples, because phages were unstabl e 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 co ntrol. Plant pathogenic ba cteria often occur in non-homogenous masses surrounded by extracellular po lysaccharides that protect them from phage attachment (44,91), or reside in protecte d 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 ch aracteristics of lysogenized bacteria by their prophages, have been found to have undesi rable consequences, such as resistance to bacteriophages, toxin production or even increased virulence. When Xanthomonas axonopodis pv. citri strain XCJ19 was lysogenized w ith temperate phage PXC7, it became resistant to phage CP2 (134). Phage-associat ed toxin production has not been documented amongst phytobacteria, but such cases are known amongst human pathogens (130) and in bacteria of plant associated nemat odes (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 tra nsducing 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 possibl e. Examples for negative phage impact in agriculture include studies in which the phage-inc ited reduction in symbiotic nitr ogen fixing bacteria reduced growth and nitrogen content of cowpea (4) and in which bi ocontrol ability of Pseudomonas flourescens was abolished by a lyti c bacteriophage (69). Factors Influencing Efficacy of Phages as Biological Control Agents Goodridge stated that the effi cacy of phage therapy depends on the ability of a phage to find its host before it is destroyed (42). Accordin g to Johnson (67) the success of a particular

PAGE 35

35 biocontrol treatment is influenced by agent a nd target densities. A component of Johnsons model is the possibility that the target reside s in spatial refuges where the biocontrol agent cannot penetrate. Gill and Abedon (40) proposed several additional f actors specifically in relation to phage therapy: the lo cation in which the target pat hogen resides; the presence of adequate water as a medium for virus diffusi on; rates of virion decay; timing of phage application; phage in situ multiplication ability; and relative fi tness 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 phy llosphere. 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 le af surface. However, they identified several factors that can hinder success of disease control in the rhizosphe re. 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 numbe r 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, becau se 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 ti ssues (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

PAGE 36

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 eveni ng 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 targ et environment (115). Using phage s isolated from aerial tissues might be advantageous, as they might be bett er 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 applica tions 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 tr eatment was applied one hour or one day before inoculation with the pathogen. Th ere was a slight disease reducti on 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 ther e were not enough phages reaching the pathogen. Schnabel et al. (101) achieved a significant redu ction 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 si gnificant when phages were applied a day before inoculation. Berghamin Filho (18) investig ated 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 sign ificant disease reductio n was achieved if the

PAGE 37

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 applicati on of phages the same day as i noculation 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 redu ced disease severity, whereas th e 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 gr eenhouse conditions with a single application of a single-phage suspension. Zaccardelli et al. (137) isolated eight pha ges active against the pathogen, screened them for host range and lytic ability, and selected a lytic phage with the broadest host range for diseas e control trials. Biweekly spra y-applications of the phage suspension in producing orchards si gnificantly reduced bacterial sp ot 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 th e co-application of the avirulent strain and the phage resulted in 17.6% wilted plants. Control of Erwinia amylovora the fire blight pathogen of a pple, pear and raspberry, with bacteriophages is currently under investig ation in Canada and the USA. Schnabel et al. (101) used a mixture of three phages for controlli ng fire blight on apple blossoms and achieved significant (37%) disease reduction. Gill et al. (41) isolated 47 phages capable of lysing E.

PAGE 38

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 contro l ability were selected for s ubsequent 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. Di sease control comparable to streptomycin was achieved (115). There has been extensive research on suppr essing 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 activ e against the two predominant races of the pathogen, X. campestris pv. vesicatoria Balogh et al. (10) enhanced the efficacy of phage treatment with protective formul ations that increased phage persistence on tomato foliage. Obradovic et al. (89,90) used formulated phages in comb ination 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 th e reduction of incidence of b acterial 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 bact erial 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 enviro nment-friendly sustainable agricu lture and the rise of organic

PAGE 39

39 production necessitates improvements in biological disease control methods, including the use of bacteriophages against bacterial plant pathogens. On the other ha nd, 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.

PAGE 40

40 CHAPTER 3 CHARACTERIZATION OF BACTERIOPHAGES ASSOCIATED WITH CITRUS CANKER IN FLORIDA AND ARGENTINA Introduction In order to develop a phage based disease contro l 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 na ture and if so, assess their impact. Xanthomonas strains causing citrus canke r 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 tw o 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 pa thotype 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 di stinguishing 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

PAGE 41

41 was related to A* strains ( Xac ) from southwest Asia. There has not been any genetic analysis published in relation to the Etrog st rains, but it is suspected that they were brought into Florida on plant material from Pakistan. The objective of this project was to assemb le 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 Florid a, 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 th e diversity of the pathogen and determined which phages could be used for disease c ontrol. Changes in ph age sensitivity of Xac strains isolated in different years may provide inform ation 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., Cockeysvill e, MD) and 1.5% (wt/V) Bacto Agar (Difco, Becton Dickinson and Co., Sparks MD)) at 28C. For bacteriopha ge detection and propagation either semisolid nutrient agar yeast extract medium (NYA), (0.8% Nutrient Broth, 0.6% Bacto Agar and 0.2% Yeast Extract (Dif co, Becton Dickinson and Co., Sparks, MD)) or liquid nutrient broth medium was used. Sterilized tap wate r 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 -80C in NB supplemented with 30% glycerol. Bacteriophages (Table 3-2) were stored at 4C and pr otected from light.

PAGE 42

42 Standard Bacteriophage Techniques Purification and storage. Phages were purified by thr ee subsequent single plaque isolations. Single plaque isolati ons were carried out by transferring phages from isolated plaques to a fresh lawn of the host bacterium using ster ile toothpicks and then quadrant streaking them with sterile plastic transfer loops. Following purification th e phages were propagated by mass streaking on fresh lawns of the host. After a 24-h incubation at 28C, the phages were eluted by pouring 5 mL sterilized tap water into the 100 mm mm Petri dishes (Fis cher Scientific Co. LLC, Suwannee, GA) and gently shaking the pl ates (~20 rpm) for 30 min. The eluate was centrifuged (10,000 g, 10 min), treated with chloro form or filter-sterilized, depending on the phage, then quantified as describe d below, and stored in 2-mL pl astic vials at 4C in complete darkness. The concentrations of thes e suspensions were approximately 109 plaque forming units (PFU) per mL. Determination of titer. Phage concentrations were dete rmined by dilution-plating-plaquecount assay on NYA plates without bottom agar as previously described (97). One hundred microliter aliquots of dilutions of phage suspensi ons were mixed with 100 L of concentrated bacterial suspension in empty Petri dishes and then 12 mL warm (48C) NYA medium was poured in each dish. The dishes were gently swir led to evenly distribut e the bacteria and the phages within the medium. After the medium solidified, the plates were transferred to 28C 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 stor age, purified by single plaque isolations and then mass streaked on the freshly prepared lawn of the pr opagating host. The next day phages were eluted from the pl ate, sterilized and enumerated, as described above. The eluate

PAGE 43

43 was used for infecting 500 mL actively grow ing culture of the propagating strain (108 cfu/mL) grown in NB liquid medium in 1 liter flasks, at 0.1 multiplicity of inf ection (MOI), (i.e., the phage concentration at the be ginning 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 28C for 18 h. The resulting culture was sterilized; phages were enumerated and stored at 4 C 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 ster ilized lysate was centrifuged at 10,000 g for 10 min to sediment th e bacterial debris. Forty milliliters of the supernatant was transferred to a new centrifuge tube and centr ifuged 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 addi tional centrifugation (60 min, 25,000 g) the supernatant was discarded and the pellet was resu spended 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 pla ques 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 effici ent the phage is in initiating

PAGE 44

44 disease of the bacterium. Consequently, the more se nsitive the bacterial stra in is to the phage. If the EOP was higher than or equal to 0.1, the test strain was considered se nsitive; if the EOP was less than 0.1, but more than or equal to 0.01 the st rain 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 a nd fruits with charact eristic citrus canker lesions in Florida and in Argentin a. 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, diseas ed tissue was collected directly from infected trees located at the Instituto Nacional de Tecnologa Agropecuar ia (INTA) research cen ter in Bella Vista, Corrientes, and from commercial citrus groves in Corrientes provi nce. The tissue samples were placed in plastic freezer bags or 125 mL flasks and after the addi tion of 50 mL deionized (DI) or sterilized tap water were shaken for 20 min. Tw o 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 th e presence of bacteriophages by spotting 20 L onto freshly prepared lawns of the indi cator bacteria. In Florida three Xac strains were used for detection: Xac65, a Miami type; Xac60, a Mana tee type, and Xac66, a Wellington-type. In Argentina a battery of 11 Xac strains of divers e 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 incuba tion at 28C, the phage was purified by three successive single plaque isolations and then propagated and stored, as described above.

PAGE 45

45 Phage Typing of 81 Xanthomonas Strains Twelve phages were used in the phage typing study: -MME, 5536, Xacm4-11, Xv3-21, XaacA1, XaacF1, XaacF8, cc19-1, cc13-2, CP1, CP2 and CP3 (Table 3-2). The bacterium-phage interactions were scored as sensitive = 2, modera tely sensitive = 1 and resistant = 0. Similarity matrix was calculated from the phage typing sc ores using the Pearson correlation, and a dendrogram of relatedness was pr epared in which the clustering was achieved by UPGMA (unweighted pair group method usi ng 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 a nd Cell Science Department of the University of Florida, Gainesville, FL. The phages were visualized usin g negative staining protoc ol with 1% aqueous uranyl acetate, as follows. A drop of the phage suspension was applied to a 300 mesh formvarcoated copper grid. After 2 min the liquid was bl otted 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 Zei ss 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, 37C) to digest any contam inating bacterial nucleic acids. Subsequently, samples were divided into 500 L subsamples, placed in 1.5 mL microcentrifuge tubes, and 375 L of a phenol-chloroform-isoamyl alcohol mixture (25:24:1) was adde d. The samples were vortexed

PAGE 46

46 and centrifuged for 5 min at 10,000 g. The top, a queous layer was transferred into a new microfuge tube and the organic layer was disc arded. The sample volume was brought up to 500 L again with the addition of sterile DI water and was subjected to phe nol-chloroform-isoamyl alcohol extraction two more times Sterilized DI water was adde d to the sample to bring the volume up to 500 L and then the solution was subjected to chloroform-isoamyl alcohol extraction once by adding 250 L of a chloroform-i soamyl alcohol mixture (24:1), vortexing and then centrifuging for 5 min at 10,000 g. The top, a queous phase was saved and transferred to a new tube, while the remainder was discarded. The DNA was then precipitated by adding 40 L sodium acetate (3 M, pH 5.2) and 800 L of co ld (-4C) 95% ethanol to the tube, vortexing and incubating at -80C for 30 min. The sample wa s then centrifuged (10,000 g, 20 min) at 4C and the supernatant was discarded and replaced wi th 1 mL of 70% etha nol. After another centrifugation (10,000 g, 5 min) the supernatant was gently remove d by pipetting and the pellet containing the precipitated DNA was allowed to air dry. The pellet was resuspended in 20 L sterile DI water and stored at 4C. Digestion of DNA with restriction enzymes. One microliter of the phage DNA was mixed with 7 L sterile DI water, 1 L enzy me buffer and 1 L restriction enzyme ( Eco RI or Bam HI) in a 1.5 mL microcentrifuge tube. The mixt ure was incubated at 37C for 90 minutes. Immediately following the incubation, the enti re 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. Eco RI plus Hin DIII digested Lambda phage DNA (Promega Co., Madison, W I) was used as size marker. The gel was photographed after 15 min ethi dium bromide staining.

PAGE 47

47 Results Bacteriophage Isolations Bacteriophages were isolated from leaf sa mples 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 pha ges 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 tw o 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 mo rphology of four members of this group was determined: they were all tailed phages with symmetrical heads an d short tails (Figure 32, 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 gr oup A phages. Virion morphology of one phage in this group, XaacF8, was found to be filament ous, 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 identica l 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,

PAGE 48

48 XaacA1, 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 CP 2 Based on Genome Size and RFLP Profile Genome sizes of all 31 members of Florid a phage group A and CP2 were identical and were approximately 22 kb (Figure 3-3). Howeve r, the group A phages constituted seven groups based on RFLP profile after Bam HI 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). Re striction profiles based on Eco RI digestion resulted in six gr oups, five of which matched five Bam HI groups; whereas the sixth one contained two Bam HI groups (data not shown). Vi rion morphology of members of different Bam HI groups, XaacF1 (group 1), XaacF2 (group 7), XaacF3 (group 2) and XaacF5 (group 3) did not differ cons iderably (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 we re eight strains of X. axonopodis pv. citrumelo ( Xacm ) causal agent of citrus bacter ial spot, one st rain 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 strain s of the same genotype did tend to cluster together (Figure 3-2). On the other hand, Xacm strains showed considerable dive rsity. Most of Etrog strains were in one group (G1), clustering together with a Pa kistani A strain. Two Etr og strains comprised an

PAGE 49

49 independent group (G6), however. Most of the B strains clustered together (G2) and were sensitive to phage CP3, however strains 6B, 7B (G 7) and 8B differed from the rest and were not sensitive to CP3. Wellington strains clustered togeth er with the A* strains in groups G3 or G5. The majority of Miami strains and the majority of A strains of vari ous geographical origins clustered together and were se nsitive to at least 8 of 12 phages (G4). However, three Miami strains and one Brazilian A strain constituted anothe r group (G8) and were resistant to all but two phages, including XaacF1 and XaacF8 that are naturally presen t in Florida. All C strains clustered in group G9 and were resistant to CP 1 and CP2. None of the Manatee strains were sensitive to any phages tested (G11). On th e other hand, strain ATCC 49118 (G10), which was isolated from the 1980s Tampa Bay outbreak was se nsitive 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. On e of them had an iden tical profile to most Miami types (G4), whereas the other one was resi stant to 11 of 12 phages tested. Host ranges of phages CP2 and XaacA1 were identical and were almost identical to that of XaacF1: the only difference was XaacF1 was not able to lyse strain XC427 from Thailand. Discussion In this study bacteriophages were frequently associated with citr us 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 locate d in relatively concise location, mainly in the Miami-Palm Beach area. It seems that there are near identical populat ion of phages around the world, as all phages isolated from Argentina and th e 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. Interestingl y, attempts to isolat e phages in Florida in 2006 were unsuccessful. Since 2003 citrus canker was spread by hurricanes widely within

PAGE 50

50 Florida, and it may be that the bacteria managed to escape its pathogens in the process. Also, this study showed the presence of strain s resistant to the phages isolated in Florida and it is possible that their prevalence resulte d in decline of phage. Results of this study also showed that phages can be used for distinguishing Xac genotypes present in Florida. Our results genera lly supported the findings of Cubero et al. (26); however based on phage typing results, the representative strain associat ed with the 1980s outbreak and strains from the 1997 Manatee outb reak were different. This findi ng 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 genetica lly, was resistant to 11 of 12 phages, which indicates the possibility of a populat ion shift to phage resistance. That in turn could mean that the naturally occurring phages pose a selection pressure on the pathoge n populations, that is, they have a significant impact on th e pathogens life. Thus, artifi cial introduction of exotic bacteriophages, phage therapy, may well be e ffective 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 ph age DNA sequences: the presence or absence of DNA sequence patterns in the phage genome th at are recognized by the host bacterial restriction-modification system.

PAGE 51

51 Table 3-1. Bacterial stra ins 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

PAGE 52

52 Table 3-1. Continued Strain Origin Genotypea Provided by S4mg Florida This study X. vesicatoria MME Jones, J.B.e X. perforans opgH:ME-B Jones, J.B. a Genotype of X. axonopodis pv citri strains isolated in Florida. b Instituto Nacional de Tecnologa Agropecuar ia, Bella Vista, Corrientes, Argentina. c United States Department of Agriculture, Agri cultural Research Service, Fruit Laboratory, Beltsville, MD. d Florida Department of Agriculture and C onsumer Services, Division of Plant Industry, Gainesville, FL. e University of Florida, Plant Pat hology Department, Gainesville, FL. f RF2 is a phage-resistant mutant of Xac65. g Chloramphenicol and Rifamycin-resistant mutant of S4.

PAGE 53

53 Table 3-2. Bacteriopha ges 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. XaacF1 to 37 Xac65 citrus, aerial Florida This study XaacA1 to 30 BV42 citrus, aerial Argentina This study cc 5 Xac65 Jackson, L.E.b cc 7 Xac65 Jackson, L.E. cc 13-2 Xac65 Jackson, L.E. cc 19-1 Xac65 Jackson, L.E. cc 24 Xac65 Jackson, L.E. cc 25 Xac65 Jackson, L.E. cc 28 Xac65 Jackson, L.E. cc 35 Xac65 Jackson, L.E. cc 36 Xac65 Jackson, L.E. XV3-21 Xac65 Jackson, L.E. -MME MME lake water Florida Minsavage, G.V.c 5536 Xacm 47 Jackson, L.E. Xv3-3-13h opgH:ME-B Jackson, L.E. X44 Xacm47 Jackson, L.E. Xacm4-4 Xacm36 citrus, aerial Florida This study Xacm4-16 Xacm36 citrus, aerial Florida This study Xacm4-11 Xacm36 citrus, aerial Florida This study a United States Department of Agriculture, Agri cultural Research Service, Fruit Laboratory, Beltsville, MD. b OmniLytics, Inc., Salt Lake City, UT. c University of Florida, Plant Pat hology Department, Gainesville, FL.

PAGE 54

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 XaacF1 Ra clear, 2-3 -c ++ ++ XaacF2 R clear, 2-3 ++ ++ XaacF3 R clear, 2-3 ++ ++ XaacF4 R clear, 2-3 ++ ++ XaacF5 R clear, 2-3 ++ ++ XaacF6 R clear, 2-3 ++ ++ XaacF7 S turbid, 1 ++ ++ ++ XaacF8 S turbid, 1 ++ ++ ++ XaacF9 S turbid, 1 ++ ++ ++ XaacF10 R clear, 2-3 ++ ++ XaacF11 S turbid, 1 ++ ++ ++ XaacF12 R clear, 2-3 ++ ++ XaacF13 S turbid, 1 ++ ++ ++ XaacF14 R clear, 2-3 ++ ++ XaacF15 R clear, 2-3 ++ ++ XaacF16 R clear, 2-3 ++ ++ XaacF17 R clear, 2-3 ++ ++ XaacF18 R clear, 2-3 ++ ++ XaacF19 R clear, 2-3 ++ ++ XaacF20 R clear, 2-3 ++ ++ XaacF21 R clear, 2-3 ++ ++ XaacF22 R clear, 2-3 ++ ++ XaacF23 R clear, 2-3 + ++ XaacF24 R clear, 2-3 ++ ++ XaacF25 S turbid, 1 ++ ++ ++ XaacF26 R clear, 2-3 ++ ++ XaacF27 R clear, 2-3 ++ ++ XaacF28 R clear, 2-3 ++ ++ XaacF29 R clear, 2-3 ++ ++ XaacF30 R clear, 2-3 ++ ++ XaacF31 R clear, 2-3 ++ ++ XaacF32 R clear, 2-3 ++ ++ XaacF33 R clear, 2-3 ++ ++ XaacF34 R clear, 2-3 ++ ++ XaacF35 R clear, 2-3 ++ ++ XaacF36 R clear, 2-3 ++ ++ XaacF37 R clear, 2-3 ++ ++ XaacA1 R clear, 3-4 ++ ++ ++ XaacA2 R clear, 3-4 ++ ++ ++ XaacA3 R clear, 3-4 ++ ++ ++ XaacA4 R clear, 3-4 ++ ++ ++ XaacA5 R clear, 3-4 ++ ++ ++ XaacA6 R clear, 3-4 ++ ++ ++ XaacA7 R clear, 3-4 ++ ++ ++ XaacA8 R clear, 3-4 ++ ++ ++ XaacA9 R clear, 3-4 ++ ++ ++

PAGE 55

55 Table 3-3. Continued Chloroform Plaque typeb, Host ranged sensitivity diameter (mm) BV38 XC90 Xac41 Xac65 Xac15 XaacA10 R clear, 3-4 ++ ++ ++ XaacA11 R clear, 3-4 ++ ++ ++ XaacA12 R clear, 3-4 ++ ++ ++ XaacA13 R clear, 3-4 ++ ++ ++ XaacA14 R clear, 3-4 ++ ++ ++ XaacA15 R clear, 3-4 ++ ++ ++ XaacA16 R clear, 3-4 ++ ++ ++ XaacA17 R clear, 3-4 ++ ++ ++ XaacA18 R clear, 3-4 ++ ++ ++ XaacA19 R clear, 3-4 ++ ++ ++ XaacA20 R clear, 3-4 ++ ++ ++ XaacA21 R clear, 3-4 ++ ++ ++ XaacA22 R clear, 3-4 ++ ++ ++ XaacA23 R clear, 3-4 ++ ++ ++ XaacA24 R clear, 3-4 ++ ++ ++ XaacA25 R clear, 3-4 ++ ++ ++ XaacA26 R clear, 3-4 ++ ++ ++ XaacA27 R clear, 3-4 ++ ++ ++ XaacA28 R clear, 3-4 ++ ++ ++ XaacA29 R clear, 3-4 ++ ++ ++ XaacA30 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.

PAGE 56

56 A B C D E F G Figure 3-1. Transmission elec tron micrographs of represen tative phages. A) CP2, B) XaacA1, C) XaacF1, D) XaacF2, E) XaacF3, F) XaacF5, G) XaacF8.

PAGE 57

57 Table 3-4. Summary of morphological characterist ics of phage virions. Phage Head shape Head (Lengtha, nm)Tail shape Classificationb CP2 symmetrical 50.8 short Podoviridae XaacA1 symmetrical 53.3 short Podoviridae XaacF1 symmetrical 58.3 short Podoviridae XaacF2 symmetrical 51.2 short Podoviridae XaacF3 symmetrical 56.6 short Podoviridae XaacF5 symmetrical 62.5 short Podoviridae XaacF8 Inoviridae a Length measurement represents the average value obtained by measuring three virions per phage. b Classification based on vi rion morphology as described by Ackermann Fig. 4.1 (1).

PAGE 58

58 100 80 60 40 20 0 PhiXacm4-11 PhiXv3-21 ccPhi19-1 Phi5536 PhiXaacF8 CP2 PhiXaacA1 PhiXaacF1 alpha-MME CP3 ccPhi13-2 CP1 XN2003-00012-7 XN2004-00225-2 XS2004-00159 XC2005-00344-1 XN2003-00012-3 XN2003-00013-2 X2001-00042 XN2003-00011-1 XN2003-00011-2 XN2003-00011-8 XC159 LMG 7484 XS1999-00061 JH96E XC64 JH96-2 XCC88B JH96C XC90 S4 XC2005-00110 XC2000-00042 X2000-12884 X2001-00006 X2003-01008 XC165 X2001-00032 XC261 XN2003-01036 X2000-12862 X2000-00071 X2003-03516 SNU6B X1999-12813 XC392 XCC03-1635 XI2000-00075 XI2000-00080 XI2000-00120 XI2006-00204 XS1999-00038 XC214 XC100 X2003-02912 XC2002-00010 XI200000194 Xac A Etrog Xacm Xacm Xac A Miami Xac A Etrog Xac A Etrog Xacm Xac A Etrog Xac A Etrog Xac A Etrog Xac A Xa diffenbachiae Xacm Xaa B Xaa B Xaa B Xaa B Xaa B Xaa D Xacm Xacm Xacm Xac A Wellington Xac A Wellington Xac A Wellington Xac A* Xac A Wellington Xac A* Xac A Wellington Xacm Xac A Miami Xac A Miami Xac A Xac A Miami Xac A Xac A Xac A Miami Xac A Miami Xac A Miami Xac A Miami Xac A Miami Xac A Xac A Xac A Miami Xac A Miami XAMii Florida Florida Florida Florida Florida Florida Florida Florida Florida Florida Pakistan Florida A rgentina Japan A rgentina A rgentina A rgentina Mexico Florida Florida Florida Florida Florida Florida India Florida Saudi Arabia Florida Florida Florida Florida Korea Florida Indonesia A rgentina Florida Florida Florida Florida Florida Taiwan Yemen Florida Florida Flid Figure 3-2. Dendogram and phage sensitiv ity matrix showing relationship amongst Xanthomonas strains causing citrus canker and citrus bacterial spot based on similarity of sensitivity profile agains t a battery of 12 phages. Typing phages: 1: Xacm4-11, 2: XV3-21, 3:cc 19-1, 4: 5536, 5: XaacF8, 6: CP2, 7: XaacA1, 8: XaacF1, 9: -MME, 10:CP3, 11:cc 13-2, 12:CP1. Percent similarity values were calculated using the Pearson correlati on, 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

PAGE 59

59 100 80 60 40 20 0PhiXacm4-11 PhiXv3-21 ccPhi19-1 Phi5536 PhiXaacF8 CP2 PhiXaacA1 PhiXaacF1 alpha-MME CP3 ccPhi13-2 CP1 X2003 02912 XC2002-00010 XI2000-00194 X2003-00008 XC63 XC273 XN2003-01035 XN2003-00012-1 XN2003-00012-6 6B 7B 306 X2000-00067 XI1999-00112 XI2001-00098 10537C 17C 1C 2C 3C 4C JH70C MME 10535C IAPAR9691/90 91-106 91-118 XC205 8B XC427 A TCC 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 Etrog Xac A Etrog Xaa B Xaa B Xac A Xac A Miami Xac 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 Florida Florida A rgentina A rgentina Brazil Florida Florida Florida Brazil Brazil Brazil Brazil Brazil Brazil Brazil Brazil Brazil Taiwan A rgentina Thailand Florida A ustralia 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

PAGE 60

60 Figure 3-3. Undigested b acteriophage DNA. Panel A: Lanes (from left to right) 1: 2: XaacF1, 3: XaacF2, 4: XaacF3, 5: XaacF4, 6: XaacF5, 7: XaacF6, 8: XaacF10, 9: XaacF12, 10: XaacF14, 11: XaacF15, 12: XaacF16, 13: XaacF17, 14: Panel B: 1: 2:CP2, 3: XaacF18, 4: XaacF19, 5: XaacF20, 6: XaacF21, 7: XaacF22, 8: XaacF23, 9: XaacF24, 10: XaacF26, 11: XaacF27, 12: XaacF28, 13: 14: XaacF29. Panel C: 1: 2: XaacF30, 3: XaacF31, 4: XaacF32, 5: XaacF33, 6: XaacF34, 7: XaacF35, 8: XaacF36, 9: XaacF37, 10: 11: XaacF2. 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

PAGE 61

61 Figure 3-4. Bacteriopha ge DNA digested with Bam H I. Panel A: Lanes (from left to right) 1: 2: XaacF1, 3: XaacF2, 4: XaacF3, 5: XaacF4, 6: XaacF5, 7: XaacF6, 8: XaacF10, 9: XaacF12, 10: XaacF14, 11: XaacF15, 12: XaacF16, 13: XaacF17, 14: Panel B: 1: 2: XaacF18, 3: XaacF19, 4: XaacF20, 5: XaacF21, 6: XaacF22, 7: XaacF23, 8: XaacF24, 9: XaacF26, 10:CP2, 11: XaacF27, 12: XaacF28, 13: XaacF29, 14: XaacF30, 15: XaacF31, 16: Panel C: 1: XaacF32, 2: XaacF33, 3: XaacF34, 4: XaacF35, 5: 6: XaacF36, 7: XaacF37, 8: XaacF2. 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

PAGE 62

62 Table 3-5. Grouping of Flor ida bacteriophages based on Bam H I RLFP profile. Group 1 Group 2 Group 3 Group 4 Group 5 Group 6 Group 7 XaacF1 XaacF3 XaacF5 XaacF27 XaacF31 XaacF34 XaacF2 XaacF6 XaacF4 XaacF32 XaacF36 XaacF12 XaacF10 XaacF33 XaacF37 XaacF15 XaacF14 XaacF16 XaacF17 XaacF18 XaacF30 XaacF19 XaacF20 XaacF21 XaacF22 XaacF23 XaacF24 XaacF26 XaacF28 Figure 3-5. Representatives of the 7 RFLP groups based on Bam H I. digestion profile. Lanes (from left to right) 1: 2: XaacF24 (group 1), 3: XaacF30 (group 2), 4: XaacF5 (group 3), 5: XaacF27 (group 4), 6: XaacF31 (group 5), 7: XaacF34 (group 6), 8: XaacF2 (group 7), 9: Arrow indicates size marker 21,226 bp. 1 2 3 4 5 6 7 8 9

PAGE 63

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 antibioticresistant bacteria in the field ( 119) and due to concerns that phyt obacteria can harbor and transfer antibiotic resistance to human pathogens. The ex tended 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 prompt ed a search for identifying novel, effective means for managing bacterial dis eases of plants. Additionally, c oncerns about food safety and environmental protection, and the growing organi c production necessitate development of safer, more specific and environmentfriendly pesticides (96). Bacteriophages have been used e ffectively 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 tr eatment. Phage applications decreased geranium bacterial blight severity, caused by X. campestris pv. pelargonii (37). Tomato bacterial spot disease severity was reduced a nd fruit yield was increased by phage applications (10,38,89,90). Phage-based integrated management of tomato b acterial spot is recommended to tomato growers in Florida (85), and bacteriophage mixtures ag ainst the pathogen are commercially available (Agriphage from OmniLytics Inc., Salt Lake City, UT, EPA Registra tion # 67986-1). However, phage therapy has not been evaluated fo r the control of citrus diseases.

PAGE 64

64 Our objective was to evaluate phage therapy fo r treatment of citrus canker. Additionally, we wished to investigate the effect of prot ective formulations on efficacy of bacteriophage treatment and also to use phages in combination with chemical bactericid es as a part of an integrated approach. Skim milk-bas ed protective formulations have been used to increase phage persistence on plant foliage and c ontributed to enhanced control of tomato bacterial spot (10). Ionic copper is toxic to all ce lls, since it interacts with sulfhydryl (-SH) groups of amino acids and causes denaturation of protei ns ((3), p.345). Ionic copper redu ces 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 eradic ation in Florida, and field trials in groves could not be carried out in the state. Therefore, we conducted di sease control tria ls (i) in the greenhouse of the citrus canker qua rantine facility of the Florida Department of Agriculture & Consumer Services, Division of Pl ant Industry (DPI) and (ii) in Argentina, where disease canker is endemic in cooperation with researchers of th e Instituto Nacional de Tecnologa 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 ra in 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 (Figur e 1-8) instead of the corky, raised ones of canker (Figure 1-1) and is generally less aggressi ve than citrus canker (106). Additionally, CBS is not regulated allowing the impl ementation of field trials.

PAGE 65

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 Dickins on and Co., Cockeysville, MD) and 1.5% (wt/V) Bacto Agar (Difco, Becton Dick inson and Co., Sparks, MD)) at 28C. For preparation of bacterial suspensions 24 h cultures were suspended in sterile tap water, their concentration was adjusted to 5108 cfu/mL (A600=0.3), and then were diluted appr opriately. Bacteria l inoculations were carried out by misting the bacterial susp ension on the plant foliage with a hand-held sprayer. For detection and propa gation 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 buffe r (0.05 M Tris-HCl (pH 7.5), 0.1 M NaCl, 10 mM MgSO4 and 1% (w/V) gelatin) was used for prepar ing phage suspensions. Bacterial strains used in this study (Table 3-2) we re stored at -80C in NB s upplemented with 30% glycerol. Bacteriophages were stored at 4C 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 Ho rticultural Products Co., Mary sville, OH) and fertilized as needed using Osmocote Outdoor & Indoor Smart Release Plant Food (Scott Sierra Horticultural Products Co.) contro lled release fertilizer (NPK 196-12). The plants were kept in the glasshouse of the citrus canker quarantine facility of the Divisi on of Plant Industry in Gainesville, Florida at temper ature ranging from 25-30C.

PAGE 66

66 Standard Bacteriophage Techniques Determination of titer. Phage concentrations were de termined by the dilution-platingplaque count assay on NYA plates without bo ttom 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 plat e, sterilized and enumerated. Th e eluate was used for infecting 500 mL of actively growing cultur e of the propagating strain (108 cfu/mL) grown in NB liquid medium in 1 liter flasks, at 0.1 multiplicity of infection (MOI). Af ter addition of the phage and 5 min incubation on the bench top, the culture was shaken at 150 rpm at 28C for 18 h. Then the culture was sterilized, en umerated and stored at 4 C 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 ster ilized lysate was centrifuged at 10,000 g for 10 min to sediment th e bacterial debris. Forty milliliters of the supernatant was transferred to a new centrifuge tube and centr ifuged 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 addi tional centrifugation (60 min, 25,000 g) the supernatant was discarded and the pellet was re suspended 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 es timated 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=8794%, 10=94-97%, 11=97-100%, 12=100%. The HB va lues were converted to estimated mean

PAGE 67

67 percentages by using the Elanco Conversion Tables for Barratt-Horsfall Rating Numbers (ELANCO Products Co., Indianapolis IN), as described in Appendi x A. If disease was assessed more than once, the area under the diseas e progress curve (AUDPC) was computed by trapezoidal integration of disease percentage values taken at differe nt 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 WallerDuncan K-ratio t Test (for bala nced data) or by least square m eans 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. Approxi mately three weeks late r the emerging-uniformnew-foliage was treated with a e qual mixture of phages CP2, CP31, cc 7 and cc 13-2 (test 1 59 PFU/mL, test 2 19 PFU/mL, test 3 18 PFU/mL of each phage) in the evening and the plants were covered with white plastic ba gs. Phages were applied with or without skim milk formulation (0.75% (wt/V) no n-fat dry milk powder). The goa l 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 l eaf tissues. On the next morning the bags were removed fr om the plants, and the plants were inoculat ed with Xac65 (test 1 18 cfu/mL, tests 2 and 3 56 cfu/mL). After inoculation th e 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 arra nged in a completely randomized pattern on a greenhouse bench. The disease was a ssessed 3-4 weeks after inocula tion. Four plants were used per treatment.

PAGE 68

68 Control of disease caused by phage-sensiti ve strain and its phage-resistant mutant. A mutant of Xac65, designated RF2, which was selected for resistance to phage XaacF2, was found also to be resistant to CP2, CP31, cc 7 and cc 13-2. In three tests grapefruit plants were inoculated either with Xac65 or RF2, at 56 cfu/mL 12 h after treatme nt with the mixture of CP2, CP31, cc 7 and cc 13-2, each at 18 PFU/mL. These tests were carried out similarly to the previous ones. Citrus Canker Nursery Trials in Argentina The trials were conducted at the experiment al nursery of the Instituto Nacional de Tecnologa Agropecuaria (INTA) re search center in Bella Vista, Corrientes (Figure 4-1) from January to April 2006 in cooperation with scientis ts of INTA. The experiment was set up in a split-plot design: there were two main treatment s, 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 tr eatments. The replications were 5-m long plots. The main plots were replicated twice. The pl ants were inoculated with a copper-resistant Xac strain, Xc05-2592 in January 2006 and then re-ino culated in February. The phage treatment ( XaacA1, ~106 PFU/mL) was mixed with a protective formulation (0.75% (wt/V) non-fat dry milk powder) and spray-applied tw ice weekly in. Copper-mancozeb (3 g/L Caurifix WG (BASF) (a.i. 84% copper oxychloride) pl us 2 g/L Dithane M80 (a.i. 80% mancozeb)) was also sprayapplied twice weekly. Disease assessment was carried out once, after 12 weeks of phage and copper-mancozeb applications. All diseased leaves were collected and the averag e number of lesions per leaf was determined. This experiment was carried out at tw o sites at the nursery simultaneously.

PAGE 69

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 ci trus 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 sprea d, all contributing to CBS development. Three test sites were established in th e nursery: two were set in a se ction 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 grape fruit 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 ( Xacm2004-4, Xacm2004-16 (both isolated from the nursery), and X44 (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 hurri canes caused a great deal of physical damage to the plants opening up entry for pathogen i ngress. 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 (Figur e 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 Pl ant Food (Scott Sierra Horticultural Products

PAGE 70

70 Co., Marysville, OH) and irrigated as needed using an overhead ir rigation system. The experiment was set up in a split-plot design: there were two main treatments, phage and no phage, and within them there we re two sub-treatments, copper-mancozeb and no treatment. The sub-treatments were randomly replicated five ti mes within the main trea tments. 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 unifo rm emerging young foliage before inoculation with S4m, an antib iotic 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 (Lovela nd Industries, Co., Greely, CO), a wetting agent, for increased inoculum penetration. In the fi rst trial every plant was inoculated, and pruned back after developing disease symptoms, and then the emergi ng foliage rated for disease. This was done to simulate the situation similar to that of the Dilleys nur sery, where the pathogen is present in the nursery but is not directly inoculated on the plan ts. Since the disease pressure was extremely low using this method, in the second trial the middle pl ant of every plot was inoculated and then the disease was assessed on the inoculated foliage. Phage and copper-mancozeb applications starte d 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, (cc 13-2, -MME and 5536, 18 PFU/mL each) applied with skim milk formulation (0.75% (wt/V) non-fat dry milk powder) using a non-CO2, backpack sprayer. The coppe r-mancozeb treatment (2.5 g/L Manzate 75DF (Griffin Co., Valdosta, GA) (a.i. 75% mancozeb) and 3.6 g/L Ko cide 2000 ( E. I. du Pont de Nemours & Co.) (a.i. 53.8% copper hydroxide)) was applied onc e a week, 2 days apart from the phage

PAGE 71

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 citr us canker disease development, a mixture of four bacteriophages (CP2, CP31, cc 7 and cc 13-2) was applied on grapefruit plan ts 12 h before inoculation with X. axonopodis pv. citri ( Xac ) strain, Xac65. The phage mixture wa s applied with or without a skim milk formulation. Three tests we re carried out and in two of them there were significant differences between treatments (Table 4-1). In both of those tests th e application of phage mixture without skim milk signif icantly reduced disease severity (Table 4-1). No phages could be recovered two days after application from pl ants if they were ap plied 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 de velopment of citrus canker incited by a phage sensitive strain, Xac65, and its ph age resistant mutant, RF2 were co mpared in three experiments. The disease caused by the two strain s was not significantly different (Table 4-2), indicating that the mutation to phage resistance did not aff ect 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 in cited by the RF2 in of one of

PAGE 72

72 three tests (Table 4-2). It is likely that this resulted from an error du ring 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 Ap plications on Citrus Canker Disease Development in a Citrus Nursery The effect of bacteriophage treatment, coppermancozeb treatment, and their combination was evaluated for reducing citrus canker dise ase development under field conditions in an experimental citrus nursery in Argentina (Figure 4-1). Nursery pl ants were inoculated with a copper tolerant Xac strain and were treated (i) twice weekly with a suspension of phage XaacA1; or (ii) twice weekly with a copper-mancozeb solu tion, or (iii) both with phage and copper-mancozeb. Phage application alone significan tly reduced citrus ca nker 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 c opper treatment in one of two trials. Effect of Bacteriophage Treatment on Citr us Bacterial Spot Disease Development in a Commercial Citrus Nursery Disease control trials were set up in the su mmer and fall of 2004 in a commercial citrus nursery in Florida to evaluate the effect of preventative b acteriophage applications on CBS disease development incite d by naturally occurring Xacm populations. A mixture of three phages was applied twice weekly in skim milk formulati on at three sites. Two s ites were established in sections of Valencia orange, which is moderately sensitive to CBS, whereas the third site was in a grapefruit section, which is high ly susceptible to the disease. Phage treatment contributed to a significant disease reduction in bo th Valencia sites (Table 4-4) Some disease suppression also occurred at the grapefruit site, but it wa s not significant (p = 0.1585) (Table 4-4).

PAGE 73

73 Effect of Bacteriophage and Copper-Mancoz eb Treatments on Citrus Bacterial Spot Disease Development in an Experimental Nursery The effects of bacteriophage treatment, coppe r-mancozeb treatment and their combination were evaluated for reducing CBS disease developm ent in an experimental citrus nursery. Phage treatment contributed to significant disease re duction (Table 4-5); how ever, 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 cont rol, but did not result in reduced control, either. Discussion Phage treatments proved to be effective mean s of disease reduction both under greenhouse and field conditions, both for citr us 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 spec ificities. In these stud ies 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 exampl e 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 mutati on. It indicates that th ose four phages had the same receptor specificity. Mutation to resistan ce by RF2 resulted in abolishment of disease control by the phage treatment w ithout loss in aggressiveness of the pathogen. This phenomenon also underscores the need for constant stringent monitoring of pathogen populations in a phagebased disease management program. Skim milk formulation, despite its effect of increasing phage longevity on plant foliage, totally abolished disease control achieved by the pha ges. 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

PAGE 74

74 also served as a carbon source fo r the bacterium. There is clearl y 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 fre quent for a practical application in a commercial grove setting. The relation of application frequency and control efficacy is an economically important factor a nd should be investigated before considering commercialization. The combination of phage treatment a nd 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 a pplication 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 re ducing 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 ar e to be used as a part of an integrated pest management program, they likely will perform better together with ot her biological control methods or with chemical methods of different modes of action such as systemic acquired resistance inducers.

PAGE 75

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 D illey and Son Nursery, Avon Park, Florida.

PAGE 76

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

PAGE 77

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 18 cfu/mL in test 1 and 56 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 phageresistant mutant, RF2 on grapef ruit 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-man cozeb 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 acco rding to analysis of variance. x UTC = untreated control y Means within the same row followed by the same letter are not signifi cantly different based on least square means differences at p=0.05 level. z Disease intensity = average number of lesions on diseased leaves.

PAGE 78

78 Table 4-4. Effect of bacteriopha ge 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 Valenc ia orange, site 3 of grapefruit. z Probability of equality of sa mple means according to t test. Table 4-5. Effect of bacteri ophage and copper-mancozeb treatme nt 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 signifi cantly different based on least square means differences at p=0.05 level.

PAGE 79

79 CHAPTER 5 INTERACTION OF BACTERIOPAHGES A ND THE HOST BACTERIA ON THE PHYLLOPLANE Introduction Bacteriophages, when used as biological control agents for foliar plant dis eases interact with the target organism on the leaf surface, the phylloplane. The phylloplane is a constantly changing environment; there are changes in temper ature, 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, especi ally 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 leve ls of relative humidity have different effects on phage longevi ty. Water is necessary for di sseminating phage virions and it provides the medium for phage-bacterium interac tions. It also contribut es to virion stability, since many phages are sensitive to desiccation. Microbial activity and en zymes degrade phages. Modern intensive agriculture relies heavily on ap plication of pesticides for control of plant diseases and insect pests. Most chemicals test ed did not affect bacter iophage persistence (9, 137); however, copper compounds are clearly detrimental (9,66). Materials that increase the longevity of vi ruses in the field have been identified (7,8,15,16,63,64,65). The use of such protective mate rials also increased disease control achieved with bacteriophages (10). Timing phage tr eatments to apply in the evenings and early mornings when sunlight UV irradiation is minima l, also increased efficacy (10). One important factor that remains unexplored is the phages them selves. Bacteriophages used in previous studies and also for commercial phage produ cts available for agricultural us e were selected strictly based on in vitro tests (10, 89, 90, 99) without a ny evaluation in the target environment. It is not known

PAGE 80

80 whether the different phages included in the mixture persist similarly on the plant surface, whether the ratio of active ingred ients in the phage treatments changes in the target environment after application. It is not known if different pha ges in the phage mixture contribute equally to disease control. It is not known if there are such characteristics th at make a phage more suitable to disease control on a leaf surf ace than others. The goal was to try to shed some light in this darkness. The two objectives were (i) to mon itor the fate of bacteriophages, individual constituents of a phage mixture used for diseas e control in a real fiel d situation after being applied to the leaf surface, and (i i) to test if the ability to mu ltiply 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 de termined by the dilution-platingplaque-count assay on nutrient ag ar yeast extract plates without bottom agar, as previously described (97). Phage propagation. Phages were recovered from stor age, purified by single plaque isolations and then mass streaked on the freshl y prepared lawn of th e propagating host. After overnight incubation the phages were eluted from the plate, sterilized and enumerated, as described above. The eluate was used for infec ting 500 mL of actively gr owing 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 inc ubation on the bench top, the culture was shaken at 150 rpm at 28C for 18 h. Then the culture was sterilized, enumerated and stored at 4C 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.

PAGE 81

81 Ackermann (personal communication). One hundred milliliters of the ster ilized lysate was centrifuged at 10,000 g for 10 min to sediment th e bacterial debris. Forty milliliters of the supernatant was transferred to a new centrifuge tube and centr ifuged 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 additi onal centrifugation step (60 min, 25,000 g) the supernatant was discarded and the pellet was resu spended 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 Sc ience Unit of the University of Florida in Citra, Florida. A mixture of th ree bacteriophages was sprayed onto Swingle citrumelo plants at 8 PM on July 18, 2006. The phages used were cc 13-2, -MME and 5536, all adjusted to 18 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 tr ifoliate 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 laborat ory. The bags were then weighe d and 50 mL of sterilized tap water was poured into each bag. The bags were shaken on a wrist acti on 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 co ntaining 20 L of chloroform. The tubes were stored at 4C until all samples were taken and processed and then transported to the Gainesville

PAGE 82

82 laboratory for determination of phage titer. Ph ages were enumerated by dilution plating and plaque counts on three b acterial strains: MME ( X. vesicatoria ), Xacm45 ( Xanthomonas axonopodis pv. citri ) and S4m ( Xacm ). MME selectively detected phage -MME, Xacm45 selectively detected phage 5536, and S4m detected all thr ee phages. Titers of phages -MME and 5536 were calculated from the plaque number, dilution information and leaf weight and were expressed as PFU/g leaf tissue. Titer of phage cc 13-2 was determined indirectly, as described. The efficacy of plating of phage -MME on strain MME ve rsus strain S4m was determined concurrently to phage titer determ inations by plating known concentration of phage -MME on both strains (5 replication each) a nd dividing the average number of plaques produced on MME by the averag e number of plaques produced on S4m. EOP of phage 5536 on Xacm45 versus S4m was determined similarl y. The EOP values were multiplied with the phage concentrations determined on the appropr iate selective host to calculate the phage concentration on S4m for each sample. The concentration of cc 13-2 in each sample was determined by the following equation: cc 13-2 concentration = total phage concentration (as determined on S4m) -MME concentration (calculated) 5536 concentrated (calculated). Finally, the data were converted by the y=log10(x+1) function and were graphed. The sunlight irradiation data were downloaded from th e 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, opgH:ME-B (105 cfu/mL), amended with 0.025% Silwet L-77 (Loveland Industries, Co., Greely, CO). Five da ys later three of the plants we re sprayed with sterilized tap water and three with a su spension of bacteriophage Xv3-3-13h (108 PFU/mL). The plants were

PAGE 83

83 placed in clear polyethylene bags for 1 day after phage application in a growth chamber (28C). Before removing the bags, the plants were tr ansferred to a greenhouse where they remained throughout the experiment in a completely ra ndomized arrangement. Phage populations were monitored daily for a week by removing three leaflets and recovering the phages from the leaflets. Disease control study. Phage Xv3-3-13h was also used for c ontrolling tomato bacterial spot severity incited by opgH:ME-B. Six young tomato plants were inoculated with strain opgH:ME-B; and later three of them were sprayed with Xv3-3-13h and three were sprayed with sterilized tap water. Met hods of bacterial inoculation and phage application were the same as for the population studies. The ratio of dis eased leaf tissue was estimated on the bottom three fully expanded leaves of each plant twice, 19 and 22 days after inoculation using the HorsfallBarratt (HB) scale (12): 1=0%, 2=0-3%, 3=3-6%, 4=6-12%, 5=1225%, 6=25-50%, 7=50-75%, 8=75-87%, 9=87-94%, 10=94-97%, 11= 97-100%, 12=100%. The HB va lues 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 com puted by trapezoidal integration of disease percentage values taken in different timepoi nts using the formula proposed by Shaner and Finnley (104), as described in Appendix A. The sample means of the phage-treated and nontreated plants were compared by t-test usi ng 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 Servi ces, Division of Plant Industry, citrus canker

PAGE 84

84 quarantine facility. Duncan grapefruit plants were heavily pruned and fertilized to induce the simultaneous boom of a new flush that is suscepti ble to citrus canker inf ection. Three weeks later the emerging uniform new foliage was treate d with on of three di fferent single-phage suspensions ( XV3-21, XaacF1 or cc 19-1, 5x109 PFU/mL) or with steri lized 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 16 cfu/mL). After inoculation the plants were placed inside the bags for an addi tional 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. Th e disease was assessed 3-4 weeks after inoculation. Four plants were us ed per treatment. In the first e xperiment the phage suspensions were prepared by diluting high titer lysates, but b ecause of concerns that the presence of nutrient broth in the phage lysate may c ontribute to increased disease se verity, in the second experiment the phage lysates were concentrated in order to remove nutrient br oth. The ratio of diseased leaf surface area was estimated using the Horsfall-Barr att (HB) scale (12). The HB values were converted to estimated mean pe rcentages by using the Elanco Conversion Tables for BarrattHorsfall Rating Numbers (ELANCO Products Co., I ndianapolis, IN), as described in Appendix A. Analysis of variance (ANOVA) and subsequent separation of sample means by the WallerDuncan K-ratio t Test was car ried out using SAS System for Windows program release 8.02 (Cary, NC). Population studies. In order to determine if bacteriophages XV3-21, XaacF1 and cc 19-1 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 (56

PAGE 85

85 PFU/mL) and immediately followed by application of the bacterial suspension at a much higher concentration (18 cfu/mL) or with sterilized tap wate r for control. Phage populations were monitored by removing three leaflets 0, 3, 6 a nd 9 h after application, recovering the phages from the leaflets and determining concentratio ns. 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 se nsitive to one of thr ee phages. Strain Xac41 was used for specific detection of XaacF1; Xac15 for cc 19-1 and Xac30 for Xv3-21. Results Persistence of Bacteriophages on Citr us Leaf Surface Under Field Conditions The fate of three phages, cc 13-2, -MME and 5536, was monitored on citrus foliage. The mixture of these three phages was used for bi ological control of citr us bacterial spot, and was applied at 8 PM. The three phages were applied at equal concentrations, at 18 PFU/mL, in skim milk formulation. The sunlight irradia tion at the sites was 0 until 6:30 AM then gradually increased to approximately 400 W/m2 at 9:45 AM (Figure 5-1). Th en the irradiation suddenly jumped to almost 600 W/m2. The three phages persisted differe ntly on the foliage (Figure 5-1). 5536 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 sudden ly higher sunlight irradiation, 3356 populations sharply dropped between 10 AM and 11AM, decreasing 2159 and 801 folds, in sites 1 and 2, respectively, and were prac tically eliminated by 11 AM. The populations of phage cc 13-2 decreased 32 and 12 fold between 8 PM and 10 AM, in sites 1 and 2, respectively (Figure 5-1). The rate of popul ation decline increased between 10 AM and 11 AM, but it was much lower than that of 3356, 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 cc 13-2 at site 1 and site 2, respectively. MME populations were lower from the beginni ng compared to the other two phages (Figure

PAGE 86

86 5-1). The nighttime reduction was low, but the popul ations decreased earlie r in the morning than other two phages. In site 1 there was no popul ation reduction between 8 PM and 7:45 AM, but between 7:45 and 9 AM there was a 2490 fold d ecrease and from 9 AM or later the populations were hardly detectable, 19 PFU /leaf (Figure 5-1). In site 2, MME 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 ab le 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 pr esence of high populations of the host bacterium, and to (ii) reduce citrus canker disease severity under greenhouse conditions. The populations of phage XV3-21 decreased in the absence of Xac65 to 1.6 and 1.9% of the orig inal populations after nine hours on the grapefruit leav es in experiments 1 and 2, resp ectively (Figures 5-2A and 53A). In the presence of Xac65 XV3-21 populations persisted better than in the absence of the host, although they still decrea sed over time and after 9 hours were 4.2 and 26.2% of the starting populations. Populations of phage XaacF1 also decreased in the absence of the Xac65 and were 4.2 and 5.8% at the end of the experiment in expe riment 1 and 2, respectively (Figures 5-2B and 5-3B). However, in the presence of the host, XaacF1 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, cc 19-1 decreased in populations in the absen ce 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 cc 19-1 populations were higher in the pr esence of the host than in its absence (35% vs. 9.2%); however

PAGE 87

87 at nine hours no cc 19-1 was detected in the presence of the host (Figure 5-2C). In the second experiment the presence of Xac65 cc 19-1 populations were at 45% after 9 h (Figure 5-2C). These three phages were used to treat grap efruit plants before inoculating with the pathogen, Xac65, in two experiments. In the firs t experiment none of the phages reduced citrus canker disease severity (Tab le 5-1). On the contrary, XV3-21 application resulted in significant disease increase compared to the water control (Table 5-1). However there were significant differences between plants treat ed with different phages: cc 19-1-treated plants had significantly less disease than XV3-21-treated plants and sign ificantly more disease than XaacF1-treated ones (Table 5-1). It was suspect ed that the nutrie nt broth present in phage suspensions contributed to the di sease increase on the XV3-21 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 XaacF1 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 To mato Bacterial Spot Disease Development To investigate if the correlation between a phages ability to multiply on the plant surface and effectively reduce disease is a general and pres ent in other pathosystems, the interactions of a X. perforans strain, the causal agen t of tomato bacterial spot (Fi gure 5-4), and its bacteriophage were studied. The ability of phage, Xv3-3-13h, to multiply on tomato foliage in the presence of its host, opgH:ME-B, was evaluated in greenhouse studies. Xv3-3-13h populations continuously decreased in the ab sence of the bacterial host and were under the detection limit 5 or 6 days after application, in experiments 1 a nd 2, respectively (Figure 5-5). In the presence of opgH:ME-B Xv3-3-13h populations increased 11 and 87 fold in the first 24 h, in experiments

PAGE 88

88 1 and 2, respectively, and then declined at slower rate than without the hos t (Figure 5-5). After 7 days, the populations were only 8.5 and 17 times lower than at the time of application. Phage Xv3-3-13h was evaluated for its ability to reduce bacterial spot of tomato disease development incited by opgH:ME-B. Disease severity wa s significantly reduced on phagetreated 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 th ree phages to two and by late morning to one. Experiments with formulations showed that incr easing phage persistence c ontributes to increased disease control ability. Thus it seems likely that using phages th at 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 w ould 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 mult iply efficiently enough to even maintain their populations could not reduce the dis ease. We only showed this co rrelation with a limited number of phages and only in two pathosystems, and it woul d be interesting to con tinue such evaluations on a broader scale. If it is true that those phages are more effectiv e, which persist longer and can multiply efficiently in the phyllosphere than th e 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 disc ouraging, however, that in this study we found no variability amongst the phyllosphere phages in host ranges. Howeve r, our results also showed that the phages sens itivity profiles of Xanthomonas species overlap, thus phage diversity may be increased by isolating phages from different Xanthomonas pathosystems.

PAGE 89

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 di seases in the world. Citrus canker was reintroduced to Florida in the 1990s and threatens the stat es $9 billion citrus industry. This work focused on a biological control approach to us e 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 near ly identical to each other and phage CP2 of Japan. Mixtures of bacteriophages were evalua ted for controlling citrus canker in greenhouse trials in Florida and in nursery trials in Argent ina. Bacteriophages reduced citrus canker disease severity both in greenhouse and fiel d trials. The level of control was inferior to chemical control with copper bactericides. The combination of b acteriophage 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. Fo r 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 bact erial 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 cont rol. In experiments m onitoring the fate of bacteriophages on the citrus foliage following bacteriophage application, phage populations stayed steady on the foliage during nighttime but were drastically redu ced 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 bacter ial host was investigated. Phages varied in their ability to multiply, and the ones that successfully increased in populations on the bacterial host

PAGE 90

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 seve ral areas before commer cial phage application will become a feasible option. First of all, the di versity of the citrus canker phage library has to be increased, either by further is olations or by changing host range s 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, in cluding 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. Identifica tion of formulations that provid e protection without the adverse effects could potentially result in increased di sease control efficacy. The optimal frequency of application and phage concentratio n also remains to be determine d. 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 di sease control efficacy of individual bacteriophages by sele ction for increased persistence on the leaf surfaces and higher in vivo multiplication ability. Also, rapid screening methods coul d be developed to predict disease control abili ty of phages.

PAGE 91

91 0 1 2 3 4 5 6 7 8:15 PM7:45 AM9:00 AM10:00 AM11:00 AMLog10 PFU/g leaf tissue 5536 MME cc 13-2 A 0 1 2 3 4 5 6 7 8:15 PM7:45 AM9:00 AM10:00 AM11:00 AMLog10 PFU/g leaf tissue 5536 MME cc 13-2 B 0 100 200 300 400 500 600 7008 :0 0 P M 9:00 PM 1 0 :00 PM 11:00 PM 1 2: 0 0 A M 1:00 AM 2:00 AM 3 :0 0 A M 4:00 AM 5 :0 0 A M 6 :0 0 A M 7 :0 0 A M 8:0 0 A M 9 :0 0 A M 10:00 AM 1 1: 0 0 A MSunlight (W/m2) C Figure 5-1. Bacteriophage populatio ns and sunlight irradiation in Citra, FL on July 18 and 19, 2006. A) Bacteriophage populations at site 1. B) Bacteriopha ge populations at site 2. C) Sunlight irradiation. Phage was applied at 8:00PM. Error bars indicate the standard error.

PAGE 92

92 0 20 40 60 80 100 120 140 160 0369 Hours after inoculationPercent population XV3-21 XV3-21 with host A 0 100 200 300 400 500 600 0369 Hours after inoculationPercent population XaacF1 XaacF1 with host B 0 20 40 60 80 100 120 140 160 0369 Hours after inoculationPercent population cc 19-1 cc 19-1 with host C Figure 5-2. Changes in bacter iophage populations on grapefruit foliage in the presence or absence of the host, Xanthomonas axonopodis pv. citri strain Xac65. Experiment 1. A) Phage XV3-21. B) Phage XaacF1. C) Phage cc 19-1. Percent population = percent of population average at 0 hour. Error bars indicate the standard error.

PAGE 93

93 0 20 40 60 80 100 120 140 0369 Hours after inoculationPercent population XV3-21 XV3-21 with host A 0 100 200 300 400 500 600 700 800 900 0369 Hours after inoculationPercent population XaacF1 XaacF1 with host B 0 20 40 60 80 100 120 140 160 180 200 0369 Hours after inoculationPercent population cc 19-1 cc 19-1 with host C Figure 5-3. Changes in bacter iophage populations on grapefruit foliage in the presence or absence the host, Xanthomonas axonopodis pv. citri strain Xac65. Experiment 2. A) Phage XV3-21. B) Phage XaacF1. C) Phage cc 19-1. Percent population = percent of population average at 0 hour. Error bars indicate the standard error.

PAGE 94

94 Table 5-1. Effect of treatment with three phages on citrus canke r disease development, incited by Xanthomonas axonopodis pv. citri strain Xac65, as meas ured by disease severity. Average disease severity Treatment Controly cc 19-1 XaacF1 XV3-21 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 signifi cantly 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.

PAGE 95

95 Figure 5-4. Symptoms of toma to bacterial spot, incited by Xanthomonas perforans

PAGE 96

96 0 1 2 3 4 5 6 7 8 9 10 12345678 Days after applicationLog10 PFU/leaflet Host No Host A 0 1 2 3 4 5 6 7 8 9 10 12345678 Days after applicationLog10 PFU/leaflet Host No Host B Figure 5-4. Populations of bacteriophage Xv3-3-13h on tomato foliage in the presence or absence of its host Xanthomonas perforans strain opgH:ME-B. A) Experiment 1. B) Experiment 2. Error bars indi cate the standard error.

PAGE 97

97 Table 5-2. Effect of a curative application of bacteriophage Xv3-3-13h on tomato bacterial spot disease development, incited by Xanthomonas perforans strain opgH:ME-B, as measured by area under the dis ease progress curve (AUDPC). Treated Control pz AUDPC 12.5 27.3 0.0026 z Probability of equality of sa mple means according to t test.

PAGE 98

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-H orsfall 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 dete rmined by taking the arithmetic mean of the percentages of seve ral ratings taken at one time, where each individual percentage was taken as the midpoint of the appropriate percen tage interval (Table A1). For example, if the following four HB ratings were assigned to a gr apefruit 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 differe nt timepoints using the formula proposed by Shaner and Finnley (104). Dis ease 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 trap ezoids. 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 exam ple if four disease ratings were taken one week apart and the disease percentages are 5, 10, 20 and 25, then there wi ll be three trapezoids. The area of the first trapezoid is (10+5)/2*7=52.5, the area of the s econd is 105, the area of th e third one is 157.5, so the total area of the dise ase progress curve is 315.

PAGE 99

99 Table A-1. Horsfall-Barratt scale, the corres ponding disease interval s 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).

PAGE 100

100 A 0 5 10 15 20 25 30 071421 Days after first ratingPercent disease Disease Progress Curve B 0 5 10 15 20 25 30 071421 Days after first ratingPercent disease Area Under the Disease Progress Curve Figure A-1. Visualization of the disese progress curve and of th e area under the disease progress curve. A) Disease progress curve is prep ared by plotting the diseases percentage values (y-axis) against the time of diseas e assessment (x-axis). B) Area under the disease progress curve (AUDPC) is one valu e describing the overall disease progress throughout the season.

PAGE 101

101 LIST OF REFERENCES 1. Ackermann H. 2005. Bacteriophage classifi cation. Pages 67-89 in: Ba cteriophages: biology and applications. in: Kutter E, and Sulakvelid ze A, eds. CRC Press, Boca Raton, Florida. 2. Adams MH. 1959. Bacteriophages. Intersci ence Publishers, New York, New York. 3. Agrios GN. 2004. Plant pathology. 5th ed. Elseiv er Academic Press, San Diego, California. 4. Ahamd MH, and Morgan V. 1994. Characterization of a cowpea ( Vigna undulata ) rhizobiophage and its effects on cowpea nodulat ion and growth. Biol Fertil Soils 18:297-301. 5. Al-saadi A. 2005. Phenotypic character ization 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, Po llard 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]. Gaines ville, 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 agen ts 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 bacterio phages 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 c ontrol. Phytopathology 94 6. 12. Barratt RW, and Horsfall JG. 1945. An impr oved grading system for measuring plant disease. Phytopathology 35:655. 13. Barrow PA. 2001. The use of bacteriophages fo r treatment and prev ention of bacterial disease in animals and animal models of hu man infection. J Chem Technol Biotechnol 76:677682. 14. Basit HA, Angle JS, Salem S, and Gewaily EM. 1992. Phage coating of soybean seeds reduces nodulation by indigenous soil br adyrhizobia. Can J Microbiol 38:1264-1269.

PAGE 102

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 formulati ons. 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, ed itor. Interaction be tween bacteria and bacteriophages on plant surfaces and in plant tiss ues. Proceedings of the third international conference of plant pathogenic bacteria; April 14-21 1971; Wage ningen. 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 canke r eradication not feasible Online: Florida Department of Agriculture and Consumer Servic es 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 transmissibl e lytic principle (bacteriophage) in relation to plant pathogens. Phytopathology 15:357-370. 26. Cubero J, and Graham JH. 2002. Genetic relationship among wo rldwide strains of Xanthomonas causing canker in citrus species and desi gn of new primers for their identification by PCR. Appl Environ Microbiol 68:1257-1264. 27. Dai H, Chiang KS, and Kuo TT. 1980. Characteriz ation of a new filamentous phage Cf from Xanthomonas citri J Gen Virol 46:277-289. 28. d'Herelle F. 1921. Le bactriophage: Sone rle dans limmunit. Masson et Cie, Paris.

PAGE 103

103 29. Dopson RN. 1964. The eradication of c itrus 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 pl ant cells elicits division, enlargement, and cell death. Mol Plant-Microb e Interact 12:556-560. 31. Duckworth DH, and Gulig PA. 2002. Bacteriopha ges: potential trea tment for bacterial infections. BioDrugs 16:57-62. 32. Egel DS, Graham JH, and Stal l 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 Indus try. 2006 August 29. Available from: http://www.doacs.state.fl.us/p i/canker/pdf/cankerflorida.pdf. 35. FDACS DPI.. Florida Huangl ongbing Science Panel Report Onlin e: Florida Department of Agriculture and Consumer Services, Divisi on of Plant Industry. 2006 January 31. Available from: http://www.doacs.state.fl.us/pi/ch rp/greening/hlbsci encereport1-31-06.pdf. 36. FDACS DPI. Huanglongbing statewide survey map Online: Florida Department of Agriculture and Consumer Services, Divisi on of Plant Industry. 2006 October 25. Available from: http://www.doacs.state.fl.us/p i/chrp/greening/maps/cgsit_map.pdf. 37. Flaherty JE, Harbaugh BK, Jones JB, So modi GC, and Jackson LE. 2001. H-mutant bacteriophages as a potential biocontrol of bact erial blight of gerani um. HortScience 36:98-100. 38. Flaherty JE, Jones JB, Harbaugh BK, So modi GC, and Jackson LE. 2000. Control of bacterial spot on tomato in the greenhouse and fi eld 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 internati onal 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/on line/feature/phages/. 41. Gill JJ, Svircev AM, Smith R, and Castle AJ. 2003. Bacteriophages of Erwinia amylovora Appl Environ Mi crobiol 69:2133-2138. 42. Goodridge LD. 2004. Bacteriophage biocontrol of plant pathogens: fact or fiction? Trends in biotechnology 22:384-385.

PAGE 104

104 43. Goto M. 1992. Citrus canker. Pages 250-269 in: Plant Diseases of Inte rnational 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. St udies 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 Phytopath ological Society of Japan 46:329-338. 47. Goto M, and Yaguchi Y. 1979. Relationship be tween defoliation and disease severity in citrus canker ( Xanthomonas citri ). Ann Phytopathol Soc Japan 45:689-694. 48. Gottwald TR, and Graham JH. 1992. A devi ce for precise and nondisruptive stomatal inoculation of leaf tissue with bact erial pathogens. Phytopathology 82:930-935. 49. Gottwald TR, Graham JH, and Schubert TS. C itrus 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.plantmanagementnetwor k.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 regulato ry eradication policy fo r an invasive species. Phytopathology 91:30-34. 51. Gottwald TR, Sun X, Riley T, Graham JH, Ferrandino F, and Taylor EL. 2002. Georeferenced 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. 20 03. 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 St ate 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.

PAGE 105

105 57. Graham JH, Gottwald TR, Cubero J, and Achor DS. 2004. Pathogen profile: Xanthomonas axonopodis pv. citri : factors affecting successful eradicatio n of citrus canker. Mol Plant Pathol 5:1-15. 58. Graham JH, Gottwald TR, Riley TD, and Ac hor 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 Huanglongbi ng 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 bact eriophage. Int J Food Microbiol 34:197-207. 63. Ignoffo CM, and Garcia C. 1996. Simulated s unlight-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/h eterocyclic 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. 199 7. Sunlight stability and rain-fastness of formulations of Baculovirus he liothis. 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 surf aces. 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 surv ival 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. Akadmiai Kiad, Budapest.

PAGE 106

106 71. Koizumi M. 1985. Citrus canker: the world si tuation. Pages 2-7 in: Citrus canker: An international perspective. Timme r 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, Wash ington: The Evergreen State College. 1997 Novermber 15. Available from: http://www.evergreen.edu/phage/p hagetherapy/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 ma nagement of the citrus bacterial canker disease caused by Xanthomonas campestris pv. citri in the State of Paran, Brazil. Crop Protection 9:3-7. 77. Leverentz B, Conway WS, Alavidze Z, Jani siewicz WJ, Fuchs Y, Camp MJ, Chighladze., and Sulakvelidze A. 2001. Examination of b acteriophage 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, Janisiew icz 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 bacterio cin. Appl Environ Mi crobiol 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 streptomyc in-resistant strains of Erwinia amylovora in Israel and occurrence of blossom blight in the autu mn. Phytoparasitica 26:223-230. 81. Marco GM, and Stall RE. 1983. Control of bacter ial spot of pepper in itiated 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.

PAGE 107

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 practic al considerations. Pages 595-602 in: Science and cultivation of edible fungi, Vol II: Proceedings of the 14th international cong ress. Elliot TJ, ed. 87. Munsch P, Olivier JM, and Houdeau G. 1991. E xperimental control of bacterial blotch by bacteriophages. Pages 389-396 in: Science and cu ltivation of edible fungi. in: Maher MJ, ed. Balkema, Rotterdam, The Netherlands. 88. Muraro RP, Roka F, and Spr een TH. 2000.An overview of Argen tinas 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 Ag riculture 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 fo liar applications of bacteriopha ges 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. Associat ion of bacteriophage particles with toxin production by Clavibacter toxicus the causal agent of annual ry egrass toxicity. Phytopathology 83:676-681. 93. Osawa S, Furuse K, and Watanabe I. 1981. Di stribution of ribonucleic acid coliphages in animals. Appl Environ Microbiol 45:164-168. 94. Pruvost O, Hartung JS, Civerolo EL, D ubois C, and Perrier X. 1992. Plasmid DNA fingerprints distingu ish pathotypes of Xanthomonas campestris pv. citri the causal agent of citrus bacterial canker dis ease. Phytopathology 82:485-490. 95. Randhawa PS, and Civerolo EL. 1986. Interaction of Xanthomonas campestris pv. pruni with pruniphage and epiphytic bacteria on de tached peach leaves. Phytopathology 76:549-553. 96. Reichelderfer K, and Barry JW. 1995. Introduc tion. Pages 28-35 in: Bior ational pest control agents: formulation and delivery. in: Hall FR, a ndBarry JW, eds.American Chemical Society, Washington, DC.

PAGE 108

108 97. Rizvi S, and Mora PT. 1963. Bacteriophage pla que-count assay and conf luent lysis on plates without bottom agar layer. Nature 200:1324-1325. 98. Rybak MA. 2005. Genetic determinants of host ra nge specificity of th e Wellington strain of Xanthomonas axonopodis pv citri [dissertation]. Gainesville, Flor ida: 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, J ones AL, and Jackson LE. 1999. Bacteriophage of Erwinia amylovora and their potential for biocontrol. 102. Schoulties CL, Civerolo EL, Miller JW, St all RE, Krass CJ, Poe SR, and Ducharme EP. 1987. Citrus canker in Flor ida. Plant Dis 71:388-395. 103. Schubert TS, Rizvi SA, Sun X, Gottwald TR, Graham JH, and Dixon WN. 2001. Meeting the challenge of eradicati ng 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 K nox 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 relati ng to the recent outbreak of citrus canker in Florida. Annu Rev Phytopathol:399-420. 107. Stall RE, Marco GM, and Canteros de Ec henique 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 Canter os de Echenique BI. 1980. Population dynamics of Xanthomonas citri causing cancrosis of citrus in Ar gentina. 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. Persisten ce of two model enteri c viruses (B40-8 and MS-2 bacteriophages) in wate r distribution pipe biofilms. Water Science and Technology 43:133-138.

PAGE 109

109 111. Stover E, Castle B, and Hebb J. Citrus root stock usage in the Indian river region Online: Electronic Data Information Source (EDIS) HS -852, Horticultural Science Department, IFAS, University of Florida. 2002 April. Ava ilable from: http://e dis.ifas.ufl.edu/HS129. 112. Summers WC. 2005. Bacteriophag e research: Early History. Pa ges 5-27 in: Bacteriophages: biology and applications. in: Kutter E, and Sulakv elidze A, eds. CRC Press, Boca Raton, Florida. 113. Sun X, Jones D, and Duan YP. 2005.Detecti on 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, Gottw ald TR, Graham JH, Di xon WN, Schubert TS, Chaloux PH, Stromberg VK, Lacy GH, and Sutton BD. 2004. Detection and characterization of a new strain of citrus canker ba cteria 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 bacterioph ages. Proceedings of the 1st international symposium on biological contro l of bacterial plant diseases; October 23-26, 2005; Seeheim/Darmstadt, Germany. Berlin, Germany: Die Deutsche Bibliothek CIPEinheitsaufnahme 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. Contro l 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 64:270274. 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 Stewarts 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 in fection, 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 Co mmodities Area [cited 2006 October 31]. Available from: http://www.unctad.org/infocomm/ anglais/orange/market.htm#prod.

PAGE 110

110 123. USDA APHIS. Citrus canker Online: United Stat es Department of Agriculture Animal and Plant Health Inspection Service. 2006 October 5. Available from: http://www.aphis.usda.gov/newsroom/hot_issu es/citrus_canker/citrus_canker.shtml. 124. USDA FAS. Situation and outlook 2005/06 On line: United States Department of Agriculture Foreign Agricu ltural Service. 2006 February 20. Available from: http://www.fas.usda.gov/htp/Hort_Circular/ 2006/02-06/02-20-06%20C itrus%20Feature.pdf. 125. USDA NASS. Citrus fruits 2006 Summar y Online: United States Department of Agriculture National Agricultu ral Statistics Service. 2006 Se ptember 21. Available from: http://usda.mannlib.cornell.edu/usda/c urrent/CitrFrui/Cit rFrui-09-21-2006.pdf. 126. Verniere C, Hartung JS, Pruvost OP, Civerolo EL, Alvarez AM, Maestri P, and Luisetti J. 1998. Characterization of phenotypi cally 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 phyt opathogenic 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. 2 002. Bacteriophage control of b acterial virulence. Infection and immunity 70:3985-3993. 131. Wiggins BA, and Alexander M. 1985. Minimu m bacterial density for bacteriophage replication: implications for signi ficance of bacteriophages in na tural ecosystems. Appl Environ Microbiol 49:19-23. 132. Williams ST, Mortimer AM, and Manchest er 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. Is olation and Partial Ch aracterization 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, an d Wang LY. 1993. Use of phages fo r 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.

PAGE 111

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 Informati on 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 bene fits. Electronic Data Information Source IFAS, University of Florida; Online 2005 March. Report No.: FE 532.

PAGE 112

112 BIOGRAPHICAL SKETCH Botond Balogh was born to Balogh Antal and G yuris Borbla, in Kalocsa, Hungary, on March 25, 1976. He graduated from the Sgvri Endre High School in Szeged in 1994 with mathematics major. He attended the University of Agronomy, College of Agricultural Sciences, in Gdll from 1994 to 1998, majoring in horticulture and plant breeding genetics. He participated in a research pr ogram for pepper breeding via anther and microspore culture under the guidance of Judith Mityk in the Ag ricultural Biotechnol ogical Center, Gdll from 1996 to 1998. In 1998 he took part in a year-long ag ricultural 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 graduate d with the degree of Bachelor of Science with the plant scien ce 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 ge nome of several strains of Xanthomonas campestris pv. vesicatoria a plant pathogenic bacterium. From A ugust 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 doctora l studies at the Plant Pathology department and conducted a research project on ev aluating bacteriophages for contro lling citrus canker and citrus bacterial spot.