ABSTRACT ADHIKARI, PRAGYA. Mapping QTL Derived From

ABSTRACT

ADHIKARI, PRAGYA. Mapping QTL Derived from Solanum pimpinellifolium LA3707 for Bacterial Spot Disease Resistance and Fruit Morphology in Tomato (Under the direction of Dr. Dilip R. Panthee and Dr. Chris G. Gunter).

Tomato (Solanum lycopersicum L.), is the second most consumed vegetable in the US. The large-scale production of both fresh-market and processing tomatoes is challenged by several diseases including bacterial spot. Bacterial spot disease in tomato is caused by at least four species and four races of Xanthomonas species (Xs) with the distinct geographical distribution. Change in species structure and evolution of new races and strains is a continuous process. In this research, we sought to study the phenotypic and genotypic diversity of Xs strains in NC and characterize the predominant Xs races and species. We then explored the genetic resistance in tomato against the predominant race of Xs in NC. Additionally, we also discerned the genetic basis of diverse shapes, sizes, and color in the cultivated tomato as these traits often determine the market value and culinary purposes of the tomato.

To characterize the bacterial spot pathogens, we collected 284 Xs strains from major tomato production regions of western NC. Copper and streptomycin sensitivity tests revealed that over 95% of the Xs strains were copper resistant while 25% and 45% were streptomycin resistant in 2016 and 2015 respectively. DNA fingerprint profiling with BOX repetitive element polymerase chain reaction (BOX-PCR) assay detected four haplotypes among Xs strains. The study on a subset of representative Xs strains (n = 45) identified Xs strains in NC as a single species, X. perforans (Xp) based on highly conserved hrpB7 gene sequences, while virulence of

Xs strains on tomato differential cultivars confirmed ~91% and 9% of Xs strains were races T4 and T3, respectively. Additionally, phylogenetic and comparative sequence analysis of six genomic regions (fusA, gapA, gltA, gyrB, lacF, and lepA) suggested that 74% and 13% of Xp strains of NC were similar with tomato races T4 and T3 from Florida, respectively. This suggested that the best bacterial spot management practices in tomato in NC should be implemented with major focus on introducing host resistance against race T4 and by considering the challenges currently posed by the intense use of copper-based bactericides.

To identify genomic regions associated with the genetic resistance to race T4, quantitative trait loci (QTL) analysis was performed in an intraspecific recombinant inbred lines (RIL) population NC 10204 at F5 and F6 generations developed from a cross between two elite breeding lines- NC 30P x NC-22L (2008). Experiments were conducted in four environments including two locations over two years. Five major QTL on chromosome 1, 4, 6, 11, and 12, and one minor QTL on chromosome 2 were identified. The QTL on chromosomes 1, 2, 4, and 6 were also validated in an independent inter-specific population developed from the crossing of NC 1CELBR x PI

270443. The QTL on chromosome 6 explained the most substantial phenotypic variance (up to

26%) followed by the QTL on chromosome 1 (up to 23%) and the QTL on chromosome 4 (up to

15%). Since the donor of the resistance associated with these QTL is a released superior breeding line NC 30P, the donor parent and the QTL information will be useful to breed tomato against the bacterial spot disease.

The NC 10204 population was also used to identify genomic regions controlling fruit size, shape, and color in tomato using precision phenotyping software Tomato Analyzer (TA). Four fruit size attributes on chromosome 2 explaining up to 20% of the phenotypic variance; three fruit shape attributes on chromosome 10 and 12 explaining up to 25% phenotypic variance; and three- color attributes on chromosome 4, 9, and 6 explaining up to 21% of phenotypic variance were detected. Our study identified novel QTL controlling fruit shape attributes on chromosome 10 and

12, and also confirmed the previously detected genetic loci controlling fruit size and color in inter- specific tomato population. This information will be useful to improve fruit shape in cultivated tomato.

by Pragya Adhikari

A dissertation submitted to the Graduate Faculty of North Carolina State University in partial fulfillment of the requirements for the degree of Doctor of Philosophy

Horticultural Science

Raleigh, North Carolina 2018

APPROVED BY:

______Dr. Dilip R. Panthee Dr. Christopher C. Gunter Committee Co-Chair Committee Co-Chair

______Dr. Frank J. Louws Dr. Hamid Ashrafi Minor Advidor (Plant Pathology)

______Consuelo Arellano

DEDICATION

I dedicate this work:

In memory of my father Ghanshyam Adhikari, who believed in my ability to be successful in the

academic arena when I was just four years old. “You left your six years old daughter and

couldn’t see her long journey to this stage, but I am sure you would have been very proud of

your little girl”.

To my mother, Laxmi Sharma Adhikari; grandfather, Dwarika P. Adhikari; brother, Shishir

Adhikari; and my aunt, Shova Adhikari, without their guidance, love, and support, I would not

have reached where I stand today.

To my husband, Niraj Rayamajhi, who always had confidence in me and offered me

encouragement and support in all my endeavors. “Finally, the time has come to be with you”.

BIOGRAPHY

Pragya Adhikari was born in Nawalparasi, Nepal. She pursued her primary and secondary education in Nepal. She completed her undergraduate degree in Agricultural Science with an elective in plant breeding and genetics, from the Institute of Agriculture and Animal Science,

Tribhuvan University, Nepal in 2011. She served as a research assistant in Nepal Agriculture

Research Council after earning her bachelor’s degree for 10 months, where she was involved in a rice and wheat breeding program. Then, she came to the United States to pursue her graduate degrees. She earned her Master’s degree of Science in Plant Science from Missouri State

University, Springfield, Missouri in December 2014. In 2015, she joined North Carolina State

University as a graduate research and teaching assistant in the Department of Horticultural

Science to pursue her Doctor of Philosophy in Horticultural Science with a major in plant breeding and a minor in plant pathology.

iii

ACKNOWLEDGMENTS

I wish to express my sincere appreciation to my advisor Dr. Dilip R. Panthee for accepting me as his graduate student, and for his guidance, support (both professionally and personally), advice, and encouragement throughout my graduate career, and prompt edits on the manuscripts. I am fortunate to have Dr. Frank J. Louws in my committee as a minor advisor, which lead to a minor degree in plant pathology. Thank you, Dr. Louws, for providing me with wealth of experiences and knowledges in plant pathology by giving me opportunities to learn through a major project and expose myself to a different laboratory outside of the University in

Florida. I owe much of my doctoral learning in these experiences that you were not required to do but did it anyway for my learning. I cannot fully express my gratitude for your time and guidance in the laboratory work, your encouraging and motivating words during stressful points, and your constant support throughout my study period. I am also grateful to my co-chair Dr.

Christopher C. Gunter, and my committee members Dr. Hamid Ashrafi and Dr. Consuelo

Arellano for their suggestions and guidance during the research projects. Thank you, Dr.

Consuelo, for assisting me with the data analysis, and thank you for your patience and every SAS code that you emailed me, especially since I was located off-campus at the research station. I would also like to thank my extension mentor for mentoring me in the extension activities.

I gratefully acknowledge the funding sources- National Science Foundation grant IOS-

1546625 to Dr. Dilip R. Panthee; and Department of Horticultural Science, who supported my graduate study and research projects. I would like to thank all those who helped me in various ways to complete my Ph.D. projects:

• Dr. Jeff Jones (University of Florida) for providing the lab space, tomato differentials,

and control bacteria to characterize the race and species structure of bacterial strains of

NC causing bacterial spot disease in tomato; and Dr. Sujan Timilsina (University of

Florida) for assisting me to conduct experiments at the University of Florida and helping

me with the phylogenetic analysis of the bacterial strains.

• Dr. David Ritchie for providing the control bacteria and protocol for copper and

streptomycin sensitivity test of bacterial strains.

• Craig Mauney for helping me to coordinate with tomato growers to identify fields for the

collection of symptomatic leaf samples.

• Thomas Ingram and Kerstin Bieler for helping me collect the symptomatic leaf samples

and isolating the bacteria. I appreciate the help from Thomas in teaching me to use

various software.

• Dr. Ralph A. Dean for providing me lab space in Raleigh; and Dr. Yeon Yee Oh for her

guidance and suggestions in the lab work.

• Pradeep Neupane for his suggestions and help to troubleshoot the PCR problems.

• Dr. Tika B. Adhikari for his valuable suggestions in finishing up my bacterial projects

and constructive comments in the manuscripts.

• Ann Piotrowski for her support in the field research and Amelia Heintz-Botz for helping

me with the fruit scanning.

• Mario De Jesus Velasco Alvarado and Navin Shrestha for helping me to record the

disease data.

• Dr. Zianbo Zhang for his help in the data analysis.

• Field crews of the Mountain Horticultural Crops Research and Extension Center

(MHCREC) and Piedmont Research Station (PRS) for preparing the field, tomato

planting, and taking good care of the field throughout the seasons.

• Dr. Peter Balint-kurti for providing me greenhouse space in Method Road Greenhouse to

conduct my preliminary disease screening experiment.

Additionally, I wish to express my gratitude to Dr. Julia Kornegay for her encouraging words and support in every step and providing opportunities towards my professional development. I would also like to thank all the helping hands in Raleigh. I am very grateful to the faculties, staffs, and students located at MHCREC, who provided a family environment and made my stay at MHCREC memorable. I would like to thank Dr. Thomas Ranney and Amira

Ranney for their affections and friendly company during my stay at MHCREC. Thank you,

Nathan Lynch, for your suggestions and help both related and unrelated to my thesis work and thank you for being friendly. I am also thankful to Nepali communities in Raleigh and Ashville area, who did not let me feel alone and made my life easy and joyful during my study period.

Special thanks go to Pratima Karki and Apil Tamang for their constant love and support during my stay in Raleigh.

My husband’s, advice, encouragement, love and support both in the personal and professional life were unlimited. The love and encouragement of my family were the driving force behind my success. Thank you, all awesome people.

TABLE OF CONTENTS

LIST OF TABLES ...... ix LIST OF FIGURES ...... xi Chapter 1: Literature Review ...... 1

1.1. Advances and Challenges in Bacterial Spot Resistance Breeding in Tomato ...... 1 Causal Organism ...... 2 Disease Symptoms ...... 3 Pathogen Taxonomy and Species Complexity ...... 4 Current Understanding of Race Structure and Distribution ...... 7 Disease Epidemiology ...... 9 Disease Management Strategies ...... 11 Current Understanding of Plant-Pathogen Interactions and Host Resistance ...... 17 Breeding for Qualitative and Quantitative Disease Resistance to Pathogenic Races ...... 19 Breeding Tomato for Resistance to Xanthomonas gardneri ...... 30 Transgenic Resistance to Bacterial Spot ...... 31 1.2. Breeding Tomatoes for Fruit Morphology and Color Traits ...... 33 1.3. Research Objectives ...... 38 References ...... 40

Chapter 2: Phenotypic and Genetic Diversity of Xanthomonas populations causing bacterial spot of tomato in North Carolina ...... 63

Abstract ...... 63 Introduction ...... 65 Materials and Methods ...... 67 Sampling Scheme and Sample Collection ...... 67 Copper and Streptomycin Sensitivity Assay ...... 69 Genomic DNA Extraction and Quantification ...... 70 BOX Polymerase Chain Reaction (BOX-PCR) Assay ...... 70 Multiplex Quantitative Real-Time PCR (qPCR) Assay ...... 72 Virulence Assay and Race Identification ...... 72 Multi-locus Sequence Analysis (MLSA) ...... 73 Results ...... 77 Copper and Streptomycin Sensitivity Assays ...... 77 BOX-PCR Assay ...... 81 Multiplex Quantitative Real-Time PCR (qPCR) Assay ...... 83 Virulence Assay and Race Identification ...... 84 Multi-locus Sequence Analysis (MLSA) ...... 87 Discussion ...... 92 References ...... 98

Chapter 3: Detection and Validation of QTL Controlling Bacterial Spot Disease Resistance Race T4 in an Intra-Specific Tomato Population ...... 106

Abstract ...... 106 Introduction ...... 107

vii

Materials and Methods ...... 109 Population Development and Experimental Design ...... 109 Inoculum Preparation, Inoculation, and Disease Evaluation ...... 111 Statistical Analysis ...... 112 DNA Extraction and Genotyping ...... 113 Linkage Map Construction ...... 113 QTL Analysis ...... 114 Results ...... 114 Phenotypic Variation ...... 114 Genetic Linkage Map ...... 121 QTL Analysis ...... 122 QTL Validation ...... 125 Discussion ...... 127 References ...... 132

Chapter 4: Mapping Quantitative Trait Loci (QTL) Controlling Fruit Morphology And Color Parameters in an Intra-Specific Tomato Population ...... 137

Abstract ...... 137 Materials and Methods ...... 141 Population Development ...... 141 Phenotypic Analysis ...... 143 Data Analysis ...... 144 DNA Extraction and Genotyping ...... 144 Linkage Map Construction and QTL Analysis ...... 145 Results ...... 146 QTL in F2:3 Generation ...... 152 QTL in F5:6 Generation ...... 152 QTL in Both Generations ...... 156 Discussion ...... 162 References ...... 166

Chapter 5: Future Directions for the Current Research on a Bacterial Spot Disease Resistance in Tomato ...... 173

What are the Next Steps? ...... 174 References ...... 178

viii

LIST OF TABLES

Table 1.1 Summary of breeding efforts against each race of Xanthomonas causing bacterial spot disease in tomato ...... 20

Table 2.1 Information on Xanthomonas spp (Xs) isolated from North Carolina ...... 69

Table 2.2 Summary of representative strains of Xanthomonas spp (Xs) from North Carolina used in the race identification and multiplex quantitative real-time polymerase chain reaction (qPCR) assay...... 75

Table 2.3 Xanthomonas spp (Xs) strains collected from North Carolina in 2015 and used for copper and streptomycin sensitivity assays...... 78

Table 2.4 Xanthomonas spp (Xs) strains collected from North Carolina in 2016 and used for copper and streptomycin sensitivity assays ...... 80

Table 2.5 Summary of haplotypes based on polymerase chain reactions (PCR) using repetitive sequence-based PCR with the BOX fingerprint profiles detected among Xanthomonas spp (Xs) strains from North Carolina in 2015 and 2016 ...... 82

Table 2.6 Virulence assay and hypersensitive response (HR) of representative Xanthomonas perforans (Xp) strains from major tomato-producing regions of North Carolina...... 86

Table 2.7 Sequence statistics of six housekeeping genes (fusA, gapA, gltA, gyrB, lacF, lepA) in Xanthomonas perforans strains from North Carolina ...... 88

Table 2.8 Summary of haplotypes and single nucleotide polymorphism (SNP) variation among Xanthomonas perforans strains from North Carolina based on concatenated sequences of six housekeeping genes (fusA, gapA, gltA, gyrB, lacF, and lepA) ...... 89

Table 3.1 Descriptive statistics (mean, minimum, maximum, and standard deviation) for bacterial spot disease data of NC 10204, heritability (H2), and analysis of variance (ANOVA) for two years (2016 and 2017), and two locations (MHCREC and PRS) ...... 116

Table 3.2 The correlation coefficients of the disease response between two generations and two locations in the intra-specific population NC 10204 ...... 119

Table 3.3 Descriptive statistics (mean, minimum, maximum, and standard deviation) for bacterial spot disease data of NC 13666, heritability (H2), and type 3 analysis of variance of fixed effect for two locations (MHCREC and PRS) in the year 2017 ...... 120

Table 3.4 Summary of the linkage maps of NC 10204 and NC13666 showing 12 chromosomes along with the number of markers per chromosome and the

length of each chromosome ...... 121

Table 3.5 QTL associated with bacterial spot disease caused by Xanthomonas perforans race T4 resistance identified in an intra-specific mapping population NC10204 of tomato derived from NC 30P x NC 22L-1(2008) in different environment conditions ...... 124

Table 3.6 QTL associated with bacterial spot disease caused by Xanthomonas perforans race T4 resistance identified in an inter-specific validation population NC13666 of tomato derived from NC 1CELBR x PI 270443 in different environment conditions ...... 126

Table 3.7 Comparison of physical map distances of the detected QTL associated with bacterial spot resistance in mapping population (NC 10204) and validation population (NC 13666) of tomato ...... 127

Table 4.1 Descriptive statistics for fruit morphological and color attributes measured by Tomato Analyzer (TA) in two generations (F2:3 and F5:6), significance test of genotypic effects on the attributes measured, and the correlation coefficients between generations in the intra-specific population NC 10204 of tomato. Table also presents the heritability estimates (H2) for all fruit quality traits ...... 147

Table 4.2 Pearson’s correlation coefficient among fruit morphology and color attributes in the intra-specific tomato population NC 10204 in F2:3 generation ...... 150

Table 4.3 Pearson’s correlation coefficient among fruit morphology and color traits in the intra-specific tomato population NC 10204 in F5:6 generation ...... 151

Table 4.4 Summary of QTL information obtained in NC10204 population in F2:3 generation controlling fruit morphology and color ...... 153

Table 4.5 Summary of QTL information obtained in NC10204 population in F5:6 generation controlling fruit morphology and color ...... 155

Table 4.6 Summary of consistent QTL information obtained in NC10204 population of tomato controlling fruit morphology and color in two years ...... 157

LIST OF FIGURES

Figure 1.1 Typical culture of Xanthomonas spp on YDC (Yeast extract-dextrose-CaCO3) medium ...... 2

Figure 1.2 Symptoms of bacterial spot disease on tomato: A) lower surface of leaves B) upper surface of a leaf C) stems D) calyx E) fruit surface, and F) whole plant ...... 3

Figure 1.3 Distribution of Xanthomonas spp causing bacterial spot disease in tomato and pepper ...... 7

Figure 1.4 Disease cycle of bacterial spot of tomato ...... 10

Figure 2.1 Map of North Carolina showing counties (star symbols) from where Xanthomonas spp strains were collected and analyzed in this study ...... 68

Figure 2.2 Overall proportions (A) and county wise proportions of Xanthomonas spp (Xs) strains collected from diverse tomato cultivars and farms across eight counties in North Carolina during 2015 (B) and 2016 (C) seasons, that were resistant to copper and streptomycin ...... 79

Figure 2.3 Proportions of diverse types of Xanthomonas spp (Xs) strains observed during two-year survey in major tomato-producing regions of North Carolina ...... 81

Figure 2.4 Hierarchical clustering dendrogram based repetitive sequence-based polymerase chain reaction (rep-PCR) with the BOX element primer set for Xanthomonas spp (Xs) strains analyzed in this study ...... 83

Figure 2.5 Image showing amplifications of 45 representative Xanthomonas spp (Xs) strains along with four type strains belonging to each species analyzed by quantitative real time polymerase chain (qPCR) assay ...... 84

Figure 2.6 Hypersensitive response (HR) induced on pepper (A) and tomato (B) by X. perforans race T4 and T3 respectively ...... 85

Figure 2.7 Maximum likelihood phylogenetic analysis of X. perforans strains isolated from NC using fusA gene sequence (A), gapA gene sequence (B), gltA gene sequence (C), gyrB gene sequence (D) lacF gene sequence (E), and lepA gene sequence (F) ...... 90

Figure 2.8 Maximum likelihood phylogenetic analysis of X. perforans strains isolated from North Carolina using concatenated sequence (A) and its visualization with counties, BOX-PCR fingerprint profiles, copper and streptomycin sensitivity, and race profile using T-BAS v2 (B) ...... 91

Figure 3.1 Distribution of Area Under Disease Progress Curve (AUDPC) for bacterial spot disease severity among recombinant inbred lines (RIL) population of NC10204 across two locations (Mountain Horticultural Crops Research & Extension

Center at Mills River- MHCREC, and Piedmont Research Station at Salisbury- PRS) and years- MHCREC-2016 (A), PRS-2016 (B), MHCREC-2017 (C), and PRS-2017 (D) ...... 118

Figure 3.2 Distribution of AUDPC for bacterial spot disease severity among RIL population of NC13666 across two locations- MHCREC (a), and PRS (b) in 2017 ...... 119

Figure 3.3 The logarithm of odds (LOD) graph for the major QTLs detected in NC 10204 that are validated in NC 13666 population ...... 123

Figure 4.1 The fruit images of NC 22L-1 (2008) (A) and NC 30P (B) ...... 142

Figure 4.2 The scanned tomato fruit images that are imported into Tomato Analyzer (TA) Software ...... 143

Figure 4.3 The logarithm of odds (LOD) graph for the tomato fruit size QTLs detected in NC 10204 ...... 159

Figure 4.4 The logarithm of odds (LOD) graph for the tomato fruit shape QTLs detected in NC 10204 ...... 160

Figure 4.5 The logarithm of odds (LOD) graph for the tomato fruit color QTLs detected in NC 10204 ...... 161

xii

CHAPTER 1: LITERATURE REVIEW

The cultivated tomato (Solanum lycopersicum L) is the second most consumed vegetable after potato in the United States, where the per capita consumption of fresh market and processed tomato in 2017 were 20.3 lb (9.2 kg) and 73.3 lb (33.3 kg), respectively (Agricultural Marketing

Resource Center, 2018). Tomato is consumed in various forms (fresh or processed) and is rich in vitamins A and C, as well as lycopene, an important constituent associated with decreased odds of cancer and heart disease (Merk et al., 2012). Moreover, tomato is also a model plant for basic research in plant science including fleshy fruit development because of its genetic tractability and economic value. Cultivated tomatoes have diverse shapes, sizes, and color, which often determine their culinary purposes and market values. Because of wide-use and nutritional values, there is a high demand for both fresh-market and processing tomato varieties. However, the large-scale production of both fresh-market and processing tomatoes is challenged by several abiotic and biotic stresses. Among biotic stresses, tomato diseases including bacterial spot are a major threat to tomato production.

1.1. Advances and Challenges in Bacterial Spot Resistance Breeding in Tomato

Bacterial spot is an economically important disease of fresh market tomato in North

Carolina (NC) and several other states. Bacterial spot pathogen primarily infects tomato and pepper, but it also causes disease in other crops of the Solanaceae family (Potnis et al., 2015).

The disease can cause 17 to 66% yield losses in tomato depending on the stage of infection, and reduction in fruit quality (Pohronezny and Volin, 1983). In Florida, the yield loss due to bacterial spot disease was estimated to be $3,000 per acre ($7,413 per ha) (Vallad et al., 2013). Likewise,

~ $4 M to $12.5 M yield losses have been reported in mid-west of the USA (Ma et al., 2011). In

recent decades, extensive research on pathogen taxonomy, genomics, plant-pathogen interaction, disease resistance, and disease management strategies has been conducted throughout the world.

Causal Organism

Bacterial spot of tomato is caused by at least four species of Xanthomonas. These are X. euvesicatoria, X. vesicatoria, X. perforans, and X. gardneri (Jones et al., 2004). The word

‘Xanthomonas’ is originated from the Greek word, where ‘Xanthos’ means ‘yellow,’ and

‘monas’ means ‘entity.’ As the name indicates, it is a yellow-pigmented bacterium due to the pigment called xanthomonadin (Büttners and Bonas, 2010), the colonies of which appear large, smooth, mucoid- fluidal and yellow on the nutrient agar (Figure 1.1). Xanthomonas is a large genus of plant-pathogenic bacteria which contains 19 species and more than 140 pathovars, collectively infecting around 124 monocots and 268 dicots (Büttners and Bonas, 2010).

Xanthomonas are gram-negative, rod-shaped cells with a single polar flagellum and aerobic and occur either individually or in pairs with the optimum growth temperature of 25 oC to 30oC

Figure 1.1: Typical culture of Xanthomonas spp on YDC (Yeast extract-dextrose-CaCO3) medium.

(Ryan et al., 2011). Several pathogenesis-related gene clusters are present in the Xanthomonas genome. These include i) xps and xcs gene cluster encoding a type II secretion system (T2SS), ii) regulation of pathogenicity factors (rpf) cluster, iii) hypersensitivity response and pathogenicity

(hrp) gene clusters encoding a type III secretion system (T3SS), and iv) gum genes encoding synthesis of extracellular polysaccharides called ‘xanthan’ (Ryan et al., 2011).

Disease Symptoms

Bacterial spot affects all the above-ground parts of the plant including leaves, calyx, stems, and fruits (Figure 1.2). The symptoms begin with small and numerous yellow-green angular spot on the young leaves mainly on leaf edges and tips, followed by circular, dark, water-soaked greasy lesions (Vallad et al., 2004). Sometimes, lesions are surrounded by a yellow

Figure 1.2: Symptoms of bacterial spot disease on tomato: A) lower surface of leaves B) upper surface of a leaf C) stems D) calyx E) fruit surface, and F) whole plant.

halo. These lesions expand their size and coalesce to give necrotic and a blighted appearance, resulting in severe defoliation exposing the fruits to sun-scald, cracking and making them unmarketable (Scott and Jones, 1986; Sherf and MacNab, 1986). X. perforans causes a shot-hole appearance on the leaves (Stall et al., 2009). Infected fruits are characterized by small, dark brown, irregular spots forming bumps or a wart-like appearance. All species of Xanthomonas cause similar symptoms in tomato plants, which makes it difficult to distinguish among species based on visual symptoms alone.

Pathogen Taxonomy and Species Complexity

Tomato bacterial spot was first discovered in 1914 in South Africa and defined the causal pathogen as Bacterium vesicatorium (Doidge, 1921). Later, Gardner and Kendrick (1921) identified similar leaf spot on tomato in Indiana, USA and named it as B. exitosum, but their designation was rejected as their strain was considered similar to that identified by Doidge

(1921). However, these two strains had different amylolytic activity, i.e., the strain of Doidge

(1921) being non-amylolytic and strain of Gardner and Kendrick (1921) being strongly amylolytic. Later, these strains were reclassified as X. vesicatoria and again as X. campestris pv. vesicatoria. In the early 1990s, two different groups, A and B, at the species level were noted in

X. campestris pv. vesicatoria concerning pectolytic and amylolytic activity, and DNA homology and genomic fingerprints (Vauterin et al., 1995; Stall et al., 1994; Louws et al., 1995). This resulted in reclassification, which reclassified group B strains under X. campestris pv. vesicatoria and gave the new name to group A as X. axonopodis pv. vesicatoria (Vauterin et al., 1995).

Bouzar et al. (1996) noticed the strong amylolytic activity for all the group B strains studied.

Additionally, two more groups of xanthomonads were isolated from tomato: one in Yugoslavia initially named as Pseudomonas gardneri (Šutic, 1957) and another in Florida in the early 1990s

(Jones et al., 1995), which were later characterized as group D and group C, respectively (Jones et al., 2000). At first, no differences were observed between P. gardneri and X. vesicatoria both in lab and in the plant itself, hence concluded to be synonymous to each other (Dye, 1966).

Later, P. gardneri was identified to be a Xathomonas and found genetically distinct from other xanthomonads based on DNA-DNA hybridization results (De Ley, 1978). Therefore, P. gardneri was renamed to X. gardneri and kept in a separate group. The bacterial classification was then revised in which X. axonopodis pv. vesicatoria consisted of group A and C, X. vesicatoria consisted of group B, and group D was placed under X. gardneri (Jones et al., 2000). Based on the DNA-DNA homology, group C was again moved from X. axonopodis pv. vesicatoria to a separate species, called X. perforans and group A was renamed to X. euvesicatoria (Jones et al.,

2004).

Current taxonomy of Xanthomonas causing the bacterial spot on tomato includes four species. These are X. euvesicatoria (group A), X. vesicatoria (group B), X. perforans (group C), and X. gardneri (group D). However, the Xanthomonas taxonomy is still debatable. For instances, Potnis et al. (2015) observed a close phylogenetic relationship between X. perforans and X. euvesicatoria based on mutilocus sequence analysis (MLSA). Constantin et al. (2016) proposed to merge X. perforans and X. euvesicatoria as a single species X. euvesicatoria.

Recently, a unique group of bacterial strains was reported in Nigeria, which showed the same pathogenic reactions on tomato and pepper and had an identical hrpB2 sequence as that of X. perforans but belong to X. euvesicatoria based on the six housekeeping gene sequences (Jibrin et al., 2015). In Tanzania and Australia, strains of X. arboricola were found to be associated with a bacterial spot of tomato (Mbega et al. 2012; Roach et al. 2018). In addition, a different type of

Xanthomonas strain with a fyuA genotype, not belonging to any of the previously defined species, has been detected to cause bacterial leaf spot of tomato in Tanzania (Mbega et al.,

2012). Xanthomonas campestris pv. raphani has also been reported to cause leaf spots in tomato in Russia, which were amylolytic, pectolytic, and showed similar biochemical properties as that of X. vesicatoria (Punina et al., 2009). This suggests the taxonomy of Xanthomonas, or a better delineation of the species causing a bacterial spot of tomato might need more revisions in the future.

Based on the current taxonomy, the scientific classification of Xanthomonas causing a bacterial spot of tomato (Potnis et al., 2015) is:

Kingdom: Bacteria

Phylum: Proteobacteria

Class: Gammaproteobacteria

Order: Xanthomonadales

Family: Xanthomonadaceae

Genus: Xanthomonas

Species: euvesicatoria; vesicatoria; perforans; gardneri

The high genetic variation within Xanthomonas species might be due to the plasmid with genes encoding for type 3 secretion system (T3SS) and type IV secretion system (T4SS), and insertion sequence (IS) elements located on plasmids, which are transposable DNA elements jumping from one species to other (Ryan et al., 2011). This is also supported by the report of frequent exchange of the plasmids and horizontal gene transfer among the Xanthomonas strains

(Timilsina et al., 2015). Insertion sequence (IS) elements are associated with genomic breaks, gene islands, and gene clusters that are specific to particular species, thereby causing genome diversification in the Xanthomonas (Ryan et al., 2011). The recombination among the species has also been reported as a causal factor for the diversity observed among the bacterial spot pathogens (Timilsina et al., 2015).

Current Understanding of Race Structure and Distribution

The tomato bacterial spot pathogens are distributed worldwide with the distinct distribution of species and pathogenic races (Figure 1.3) (Potnis et al., 2015). At present, four pathogenic races (T1, T2, T3, and T4) associated with differential responses on tomato cultivars have been reported. The T1 race has been reported in X. euvesicatoria, T2 race in X. vesicatoria and in X. gardneri, while T3 and T4 races are reported in X. perforans (Jones et al., 2004).

Mutation in effector genes avrXv3 and avrXv4 of races T3 and T4 belonging to X. perforans have also been identified, which was speculated to be race T5 (Minsavage et al.,

2003). However, isolation of race T5 under field condition has not been reported yet.

Figure 1.3: Distribution of Xanthomonas spp causing bacterial spot disease in tomato and pepper in the world, as adapted from Potnis et al., 2015.

X. euvesicatoria was the only species, present as race 1 in Florida until 1991 before the X. perforans race T3 strain was reported. While X. vesicatoria was never a problem in Florida and the Southeastern USA, it is prominent in the Midwest production region (Timilsina et al., 2015;

Louws et al., 1995). X. perforans race T4 strain evolved in 1998 (Astua-Monge et al., 2000;

Minsavage et al., 2003), and thereafter has been detected at higher numbers in field surveys in

Florida. In a recent survey, only X. perforans race T4 were detected in the samples collected from different parts of Florida, indicating a major shift in the race structure within X. perforans strains from previous surveys, where T4 and T3 races were in the ratio of 3:1 (Horvath et al.,

2012). X. perforans race T4 strains has been recently reported in Louisiana (Lewis et al., 2016).

One of the first USA reports of X. gardneri appeared in the Midwest USA and Ontario Canada on a contaminated seed lot in 1991 and has since become widespread (Cuppels et al., 2006).

Both X. gardneri and X. perforans are predominant in Ontario, Canada (Abbasi et al., 2015). X. gardneri has been reported from tomato fields in Pennsylvania (Kim et al., 2010), and was responsible for the epidemics in the Midwest (Ma et al., 2011).

The introduction of X. gardneri and X. perforans have also been reported in some parts of

Asia. For instances, X. gardneri has been reported for the first time in Malaysia (Rashid et al.,

2015); X. perforans in Korea (Myung et al., 2009) and Iran (Osdaghi et al., 2017). Likewise, X. perforans has been reported for the first time in Australia (Roach et al. 2018). In a survey of

Ethiopia, all species except X. euvesicatoria were detected from tomato samples (Kebede et al.,

2014). X. perforans T3 was observed in greater number than X. perforans T4 strains in this

Ethiopia study. X. gardneri is predominant in central west Brazil (Quezado-Duval et al., 2004).

The low polymorphism observed in the population of X. gardneri in central west Brazil suggested that the introduction of this pathogen is recent, but this needs to be confirmed. This shift in pathogen distribution observed in several tomato production regions throughout the world might be due to the import of contaminated seeds (Quezado-Duval et al., 2004), which highlights the importance of seed testing. Good knowledge of the pathogen distribution is

required to manage the disease and to optimize the breeding programs to improve host resistance against the disease.

Disease Epidemiology

Bacterial pathogens enter the plant tissue through natural openings such as stomata and hydathodes, wounds, and through the openings caused by the hair loss, in the case of fruits

(Scott, 1997). The pathogens grow on the plant leaf surfaces epiphytically before they enter into the host tissues through such openings. Xanthomonads either invade host tissues systemically through vascular systems if they enter through hydathodes or colonize mesophyll parenchyma if entering through stomata, after which, Hrp is highly expressed (Ryan et al., 2011; Potnis et al.,

2015). The fluctuation in humidity facilitates the bacterial entry through hydathodes. During high humid condition, bacteria colonize the water droplets exuded by the hydathodes, and then get infiltrated into the plant along with the droplets when the humidity decreases (Ryan et al., 2011).

Therefore, the disease is favored by warm, humid condition and high rainfall with the optimum temperature of 24oC to 30oC (Gardner and Kendrick, 1921; Araujo et al., 2010). On the contrary,

X. gardneri strains prefer cooler temperature (Jones et al., 1998). X. gardneri has been reported to cause the highest disease severity at a temperature of 20 °C (Araujo et al., 2010).

The bacterial spot pathogens as seed-borne, and it is believed that the pathogens introduce in the field through the contaminated seeds (Burlakoti et al., 2018). The pathogens can persist on the surface of tomato seeds, which gives rise to infected tomato seedlings the next season (Figure 1.4). About 1 in 100 commercial seeds obtained from diseased fields have a potential to give rise to diseased seedlings (Gardner and Kendrick, 1921). It appeared that the pathogen was disseminated over a long distance and introduced into new fields and geographical regions through such infected-seeds and transplants (Gardner and Kendrick, 1921). The

secondary spread of the pathogen within field and greenhouse occurs through rain-splash, overhead irrigation, sprinkler irrigation, and wind-blown rain (Ryan et al., 2011). Sometimes, they are also spread through the contaminated farm equipment and the worker activities.

Additionally, bacterial spot pathogens can survive in infected plant debris, volunteer plants, and weeds (Jones et al., 1986; Ryan et al., 2011). They can survive in crop residue from a few months to a few years depending on the climatic condition, but only a few days to a week in the soil after the crop residue decomposes (Jones et al., 1986). Xanthomonads infecting tomato plants were also reported from sprouting weeds of tomato fields in Brazil (Araújo et al., 2015).

Removal of infected crop debris, volunteer plants, and weeds help prevent further spread of the bacterial spot disease incidence in tomato fields.

Figure 1.4: Disease cycle of bacterial spot of tomato, adapted from, http://www.apsnet.org/edcenter/intropp/lessons/prokaryotes/pages/bacterialspot.aspx

Disease Management Strategies

Practical management of bacterial spot is challenging in commercial tomato production due to limited efficacy of current disease management strategies and lack of commercial resistant tomato cultivar. The high diversity of the pathogens and their rapid multiplication under favorable environmental condition, pathogen gene mutation against host resistant genes and bactericides, horizontal gene transfer of bactericides resistant genes among pathogens, and evolution of new species and races have further complicated the disease management (Obradovic et al., 2008). Therefore, current management programs are based on integrating multiple tactics including the best cultural practices, chemicals (such as copper), antibiotics, bacteriophages, plant growth promoting bacteria, and plant activators.

As alluded to previously, Xanthomonas strains are commonly introduced to tomato production fields through contaminated seeds and transplants (Burlakoti et al., 2018) or overwintered host debris and volunteer weeds (McCarter et al., 1983; Jones et al., 1986; Araújo et al., 2015). Therefore, tomato production should begin with the certified pathogen-free seed and transplants; followed by elimination of volunteer plants, host debris, weed, and crop residues during and end of the season; and crop rotation with non-hosts in the next season to prevent inoculum carry over of the bacterial spot pathogen (Goode and Sasser, 1980). High levels of soil potassium and the osmotic potential of plant tissue might be useful to prevent the disease (Stall et al., 2009). Additionally, weekly sprays of neem oil and fish emulsion (Abbasi et al., 2003); foliar sprays with compost water extracts (Al-Dahmani et al., 2003); sprays of low molecular weight chitosan (Clmw) obtained from crustacean shells, at the rate of 3 mg⁄ml three days prior to bacterial inoculation (Coqueiro et al., 2011) were reported to reduce disease severity caused by bacterial spot in tomato. The chitosan treatment was found to be associated with the increased

accumulation of flavonoids, suggesting the role of flavonoids in the control of bacterial spot disease in tomato (Coqueiro et al., 2011).

Chemical control mostly relies on the antibiotic streptomycin and copper-based bactericides. Streptomycin has been used for the management of bacterial diseases in plants since the late 1950s (Minsavage et al., 1990). The application of copper-based bactericides started even earlier in the 1920s (Stall et al., 1986), and continue to be routinely used as a standard treatment to manage foliar bacterial diseases of tomato including a bacterial spot in several tomato production regions. However, both streptomycin and copper resistant Xanthomonas strains were reported soon after their widespread use and are now present in many tomato production regions (Abbasi et al., 2015; Griffin et al., 2017). The use of copper bactericide in combination with mancozeb or maneb (ethylene-bis-dithiocarbamates/EBDC) showed better effect than copper alone but was not effective to control the disease during favorable environmental condition (Marco and Stall, 1983). Also, EBDC has a high risk of human health

(Stall et al., 2009).

Streptomycin and copper resistance in Xanthomonas strains were first reported in 1961

(Thayer and Stall, 1961) and 1983, respectively, both from Florida (Marco and Stall, 1983).

However, the examination of older cultures suggested that the resistance had been present in populations of the pathogen in Florida since 1968 (Voloudakis et al., 1993). Both copper and streptomycin resistant Xanthomonas populations were then reported from the Caribbean and

Central America in 1999 (Bouzar et al., 1999); Florida (Vallad et al., 2013); Brazil (Araújo et al.,

2012); and Ethiopia (Kebede et al., 2014). The widespread existence of copper-resistant

Xanthomonas strains was reported in Ontario, Canada (Abbasi et al., 2015); Tanzania. (Shenge et al., 2014); and Tennessee (Mixon, 2012). Limited efficacy of copper-based bactericides for the

management of bacterial spot disease of tomato has also been observed in Ohio (Miller and

Mera, 2011).

Most of the genes associated with copper resistance from plant-pathogenic bacteria are plasmid encoded. Copper resistance genes in Xanthomonas campestris pv. vesicatoria were detected on 188- to 200-kb self-transmissible plasmids pXvCu and pXV10a in strains from

Florida (Stall et al., 1986) and Oklahoma (Bender et al., 1990), respectively. DNA hybridization experiments indicated that the resistance genes on pXV10a and pXvCu plasmids share nucleotide sequence homology and may have a common origin (Bender et al., 1990). Another

100-kb non-self-transmissible plasmid containing loci for copper resistance was also discovered in strains from California. Chromosome encoded copper resistance gene clusters have also been reported in a 7,652-bp XbaI/EcoRI chromosomal fragment in X. campestris pv. vesicatoria strain

XvP26 from Taiwan (Basim et al., 2005). This study identified five open reading frames (ORFs) yielding highly similar amino acid sequence as those obtained from previously reported copper resistance genes (Basim et al., 2005). Recently, the sequencing of six tomato and pepper copper resistant strains belonging to different species of Xanthomonas showed that all six strains harbored the previously described copper resistance gene called copLAB on a conjugative plasmid (Richard et al., 2017). The plasmid was highly conserved among five strains (dating from 1955 to 2010) suggesting extensive events of horizontal gene transfer at the niche level

(Richard et al., 2017). The transfer of plasmids between copper-resistant and copper-sensitive strains through conjugation has also been demonstrated under laboratory conditions (Stall et al.,

1986; Bender and Cooksey, 1986).

Resistance to streptomycin occurs either due to the mutation in the binding affinity of ribosomal proteins for this antibiotic or due to the action of periplasmic enzymes encoded by plasmid-borne genes on the antibiotic. Both chromosomal and plasmid-mediated streptomycin

resistances in plant-pathogenic bacteria have been reported (Minsavage et al., 1990).

Streptomycin resistance might occur spontaneously because of the direct interaction of antibiotic molecules with the small ribosomal subunit. Besides, horizontal gene transfer through conjugation might also spread resistance to streptomycin in bacterial populations (Behlau et al.,

2012). Once the plasmids harboring either copper or streptomycin resistance genes or both are spread among the bacterial population, they persist within the population and decreasing the use of copper or streptomycin doesn’t help to reduce the persistence of resistant bacterial strains

(reviewed by Örmälä and Jalasvuori, 2013).

Besides copper and streptomycin, the photocatalytic nanoscale titanium dioxide (TiO2) doped with zinc (TiO2/Zn) at ≈500 to 800 ppm has been reported to significantly reduce bacterial spot disease incidence caused by X. perforans (Paret et al., 2013). The antibacterial activity of a silver-based nanocomposite, Ag-dsDNA-GO was demonstrated against X. perforans in the greenhouse, suggesting its potential use as an alternative to copper and streptomycin for the management of the bacterial spot in tomato during the transplant phase (Strayer et al., 2016).

Another antibiotic named kasugamycin demonstrated similar efficacy as that of standard copper– mancozeb treatment for the control of bacterial spot (Vallad et al., 2010), but bacteria are equally likely to develop resistance against the kasugamycin, as kasugamycin has a similar mode of action as that of streptomycin (Huang et al., 2012). Although alternative products to streptomycin and copper-based bactericides are reported, the practical utility of such compounds in the field needs further investigations.

Plant activators such as Acibenzolar-S-methyl (ASM) (Actigard 50WG, Syngenta Crop

Protection, Greensboro, NC, USA), which induces systemic acquired resistance (SAR) in plants, has been used as an alternative to copper-based bactericides for field management of bacterial spot. ASM is an important management tool in tomato fields harboring copper and streptomycin

resistant bacterial strains (Louws et al., 2001). Different studies have reported different concentrations of ASM for the management of bacterial spot disease. Initial studies tested the efficacy of a higher concentration of ASM against the bacterial spot disease. For instances,

Louws et al. (2001) demonstrated reduced foliar disease severity in 14 of the 15 bacterial spot experiments through the application of ASM at a rate of 35 g a.i./ ha every 7 -10 days, which was as effective as that of standard copper bactericides for the control of bacterial spot. Graves and

Alexander (2002) obtained equal or better control of the bacterial spot disease in tomato through the spray of ASM at a rate of 10.5 g a.i./ha compared to standard copper-based bactericides in

Virginia. Later, Huang et al. (2012) tested the efficacy of lower concentration of ASM for the management of the bacterial spot disease in tomato, where they obtained significant reduction in bacterial spot disease in the field with weekly sprays of ASM at lower rates ranging from 1.58 g a.i/ha to 4.2 g a.i./ha without compromising the yield, compared to standard copper-EDBC treatment. However, in this study, the effect of low concentration ASM was not tested in the seedlings (Huang et al., 2012). Briceno-Montero and Miller (2005) also obtained better control of the bacterial spot disease through ASM compared to standard copper-mancozeb treatment. All these studies were based on the foliar application of ASM. Recently, the efficacy of soil application of ASM through drip irrigation for the management of bacterial spot disease in tomato has also been demonstrated (Huang and Vallad, 2018). This study obtained better control of bacterial spot disease through the soil application of ASM at the rate of 10.0 mg/l compared to

18.8 mg/l of foliar spray and 0.84 mg/l of soil application. The authors argue that soil application of ASM is likely to offer protection for longer period than foliar application as the main metabolite of ASM, i.e., CGA210007, which is a plant defense activator, remains longer in soil

(up to 10 days) than in tomato plants (up to 3 days) after the application of ASM (Huang and

Vallad, 2018).

Although several successes of ASM have been reported in the studies, there are some concerns regarding the practical application of ASM in the fields. First, the induced resistance by

ASM is a host response, which is affected by the environment, host genotype, plant age, plant health, the level of induced plant defense before application of ASM, and application methods of

ASM (Walters and Fountaine, 2009; Huang et al., 2012). This results in inconsistent and incomplete disease control. Second, induced resistance through ASM application is costly to the hosts, potentially resulting in reduced yield and phytotoxicity (Vallad and Goodman, 2004;

Walter et al., 2005). The third concern is the cost-benefit of ASM due to low efficacy and potential yield loss (Huang et al., 2012).

Several researches have also been conducted to identify biological control agents for the management of bacterial spot disease of tomato. The spray of antagonistic bacteria like

Pseudomonas syringae Cit7 and Pseudomonas putida B56 (Byrne et al., 2005); bacteriophages

(Balogh et al., 2003; Obradovic et al., 2004; Jones et al., 2007), and plant growth promoting rhizobacteria (PGPR) such as Pseudomonas fluorescens 89B-61 and Bacillus pumilus SE34 (Ji et al., 2006) are some biological control agents reported for the management of bacterial spot of tomato. A better effect was observed from the combined application Pseudomonas syringae Cit7 and PGPR Pseudomonas fluorescens 89B-61 compared to the individual application of each agent (Ji et al., 2006). A study identified a hrp mutant (75-3S hrpG) of Xanthomonas when used as a foliar spray reduced mean disease severity by ~76 % in multiple trials conducted in multiple states (Moss et al., 2007). The 75-3S hrpG mutant performed better than P. syringae Cit7 and was comparable to the plant activator ASM, suggesting the effectiveness of this mutant for the management of bacterial spot of tomato (Moss et al., 2007). However, the biocontrol agents have yielded inconsistent results, and the efficacy of such biological control agents depends on the environmental conditions that impact disease development (Huang et al., 2012). The biocontrol

agents effective in lab conditions may not translate to eficacy in field conditions. In addition, bacteriophages have a narrow bacterial host range (strains/race/species specific), are sensitive to heat and UV light, and require optimization for ideal phage population size, which limits their practical application (Balogh et al, 2010).

Although, several chemicals and products such as compost water extract, neem oil, chitosan, nanoparticles, and bacteriophages are reported as alternative options for the management of bacterial spot disease, only copper-based bactericides, streptomycin, ASM, bacteriophages, and PGPR are highly studied. Studies have reported that different products with different modes of actions if used in an integrated way, offers better control than if used alone.

For instances, the disease control was more effective when ASM was used in combination with bacteriophage than obtained through a bacteriophage and ASM alone (Obradovic et al., 2004;

Obradovic et al., 2005). Similarly, Briceno-Montero and Miller (2005) obtained better control by combining biological control agents with other treatments compared to biological control agents alone. The disease severity was significantly reduced with the combined application of famoxadone, cymoxanil, ASM, copper, and mancozeb in replicated trials compared to untreated controls and copper-mancozeb standard (Roberts et al., 2008). The foliar spray with a tank mix of copper hydroxide and Bacillus subtilis QST 713 provided a better reduction in disease severity compared to Bacillus subtilis QST 713 alone (Abbasi and Weselowski, 2015).

Current Understanding of Plant-Pathogen Interactions and Host Resistance

Both plant and pathogen possess a suite of genes that enable them to communicate with each other. Such rapidly evolving genes are those that encode effectors (small secreted proteins) from the pathogen and resistance (R) proteins from the host. Host R proteins mediate a response to effectors and activate plant defense responses such as effector-triggered immunity (ETI),

which is characterized by localized cell death termed as a hypersensitive response (HR)

(Stukenbrock and McDonald, 2009). At least, 45 effectors have been reported in bacterial spot xanthomonads, and among them, AvrBs2, XopD, XopF1, XopK, XopL, XopN, XopQ, XopR,

XopX, XopZ1, and XopAD represent the core effectors required during pathogenesis (Potnis et al., 2015). A large number of R genes in plants encode the nucleotide-binding site-leucine-rich repeat (NLR) proteins. In tomato, NLRome containing 355 NLRs is made available, and the role of each NLR in the disease resistance is being studied (Piquerez et al., 2014). Unfortunately, only a few R genes against bacterial spot disease have been identified in tomato, and even so, they could not be fully utilized in the breeding program.

Both the qualitative resistance characterized by HR and quantitative resistance have been observed in several tomato species against bacterial spot disease. For instance, HR resistance has been identified to race T1 in S. lycopersicum accessions HI 7998; to race T3 in S. lycopersicum accessions HI 7981 and S. pimpenellifolium accessions PI128216 and PI 126932; to race T4 in S. pennellii accession LA716; and to X. gardneri in S. pimpenellifolium accessions LA2533,

LA1936, and PI 128216 (Scott and Jones, 1986; Scott et al., 1995; Astua-Monge et al., 2000;

Liabeuf et al., 2015) (Table 1.1). Similarly, quantitative resistance against bacterial spot has been identified in S. lycopersicum var. cerasiformae accession PI 114490 (all races); S. pimpenellifolium accessions PI 126428, PI 340905-S, and PI 155372 to race T3; and S. pimpenellifolium accessions PI 128216 and PI 126932 to race T4 (Wang et al., 1990; Scott et al.,

1995; Scott et al., 1997). Despite available sources of resistance to bacterial spot, no commercial resistant tomato cultivars are available so far. This indicates the desperate need of tomato genotype with combined resistance to all races to offer a practical solution to manage the bacterial spot disease in commercial tomato production fields.

Breeding tomato for resistance to bacterial spot has been challenging due to the evolution of new races of the pathogen to overcome the identified resistance genes, multigenic control of the resistance, and non-additive gene effects (Hutton et al., 2010b). New races of bacterial spot have evolved without any deployment of commercial resistant cultivars, suggesting the independent evolution of these races from the selection pressure from host resistance (Table 1.1).

Breeding for Qualitative and Quantitative Disease Resistance to Pathogenic Races

The race T1 includes strains eliciting HR in a tomato differential HI 7998; all T1 strains identified so far are X. euvesicatoria (Scott et al., 1986). The resistance against T1 race in HI

7998 is controlled by multiple non-dominant and independent loci (Wang et al., 1994). Three loci Rx1, Rx2, and Rx3, were revealed to control HR in HI 7998 with additive effects using an interspecific backcross population derived from a cross of S. lycopersicum HI 7998 (recurrent parent) and S. penelli LA716 (Yu et al., 1995). All three resistance loci in HI 7998 interact with a common bacterial effector avrRxv (Whalen et al., 1993). The avrRxv effector of T1 strain was found to induce HR only on HI 7998, indicating race-specific resistance (Whalen et al., 1993).

On the other hand, the field resistance of HI 7988 is reported to be controlled by additional factors in combination with the hypersensitivity response (Scott and Jones, 1989). Rx1 and Rx2 were located on the opposite arm of chromosome 1 and Rx3 on chromosome 5 in HI

7998, but further works were limited due to the inheritance of polymorphic markers from wild parent and association of the markers with susceptibility in this study (Yu et al., 1995). Yang et al. (2005b) carried out a follow-up study to identify polymorphic markers in linkage with the resistance in HI 7998 and derived from the cultivated tomato species background. This study used F2 population and advanced backcross population derived from the cross of HI 7998 with S. lycopersicum Ohio 88119 (elite cultivar) to characterize the resistant loci detected in HI 7998.

Table 1.1: Summary of breeding efforts against each race of Xanthomonas causing bacterial spot disease in tomato.

Race Resistance Resistant Bacterial Species Chromosome Markers References Type Genotype Genes Effector Scott and Jones, 1986; Whalen et al., 1993; Wang Rx3 -L1, SP5, T1 et al., 1994; Yu HI 7998 TOM196 (SSR), (before S. lycopersicum Rx3 5 avrRxv et al., 1995; (HR) TOM144 (SSR), 1989) Yang et al., COSOH57(SNP) 2005; Wang et al., 2011; Sim et al., 2015 Scott and Jones, 1986; Whalen et Rx2 1 al., 1993; Wang et al., 1994; Yu et al., 1995 Scott and Jones, 1986; Whalen et Rx1 1 al., 1993; Wang et al., 1994; Yu et al., 1995 S. lycopersicum PI114490 quantitative 2,3,10,11 Scott et al., 2015 var. cerasiformae

Wang et al., 1990; Whalen, et deletion al., 1993; Scott T2 S. lycopersicum of 680 bp PI114490 quantitative 2,3,10,11 et al., 1997; (1989) var. cerasiformae region of Scott et al., avrRxv 2003; Scott et al., 2015

Table 1.1 (continued). Jones et al., 1995; Scott et T3 HI 7981 cLEC-24-C3 (SNP), al., 1995; Scott S. lycopersicum Xv3 11 avrXv3 (1991) (HR) SL10029 (SNP) et al., 1996; Wang et al., 2011 Scott et al., PI 128216 S. 1995; Robbins et Rx4 11 avrXv3 pcc12 (HR) pimpenellifolium al., 2009; Pei, et al., 2012 Hutton, 2008; PI 126932 S. Rx4 11 avrXv3 Wang et al., (HR) pimpenellifolium 2011 Austua-Monge LA716 (HR) S. pennelii Xv4 avrXv4 et al., 2000 LA 1589 S. Rx 11 avrXv3 Sun et al., 2011a (HR) pimpenellifolium LA1589 S. lycopersicum PI114490 quantitative 2,3,10 - - Scott et al., 2015 var. cerasiformae PI 126428 quantitative - - - Scott et al., 1995 PI 340905-S quantitative - - - Scott et al., 1995 PI 155372 quantitative - - - Scott et al., 1995 Rx3 and Wang et al., Fla7600 breeding line - - Xv3 2011 Hutton et al., Fla 8233 breeding line quantitative 2010a Hutton et al., Fla 8517 breeding line quantitative 2010a Hutton et al., Fla 8326 breeding line quantitative 2010a

Table 1.1 (continued). Austua-Monge et al., 2000; Stall T4 LA 716 S. penneli RXopJ4 6 XopJ4 J350 & J352 et al., 2009; (1998) (HR) Sharlach et al., 2013 Hutton et al., S. lycopersicum C2_Atlg10050 for PI114490 quantitative 2,3,10,11 - 2010b; Scott et var. cerasiformae QTL on chr 11 al., 2015 S. PI 128216 quantitative - - - Scott et al., 2006 pimpenellifolium S. PI 126932 quantitative - - - Scott et al., 2006 pimpenellifolium Hutton et al., Fla 8233 breeding line quantitative 11 - - 2010a; Hutton et al., 2010b Hutton et al., Fla 8517 breeding line quantitative 3, 11 - - 2010a; Hutton et al., 2010b Hutton et al., Fla 8326 breeding line quantitative 11 - - 2010a; Hutton et al., 2010b X. LA2533 S. Liabeuf et al., - - - - gardneri (HR) pimpenellifolium 2015 LA1936 S. Liabeuf et al., - - - - (HR) pimpenellifolium 2015 PI 128216 S. Liabeuf et al., - - - - (HR) pimpenellifolium 2015

However, in this study, markers linked with the Rx1 loci did not co-segregate with field resistance nor an HR assay in the greenhouse, indicating no role of the Rx1 loci in the disease resistance. The Rx2 loci could not be further studied since the polymorphic markers linked to

Rx2 loci could not be detected (Yang et al., 2005b). The Rx3 loci on chromosome 5 were significantly associated with the bacterial spot disease resistance explaining 25% of the phenotypic variation for the disease resistance in the F2 generation and 41% in an advanced backcross population in replicated field trials. In addition, a locus from HI 7998 linked to susceptibility was also detected on chromosome 4 that explained 11 % of the total phenotypic variation (Yang et al., 2005b). Other factors such as plant maturity (6%) and growth habit (6%) also contributed to some degree of observed variation in the bacterial disease resistance. Plants with late maturity and indeterminate growth habit provided a higher level of bacterial spot disease resistance (Yang et al., 2005b). The marker Rx3 -L1 linked to Rx3 loci would be useful in the tomato breeding program to select for race T1 resistance in elite elite crosses and to pyramid multiple resistance loci in a tomato cultivar. However, no association was detected between the marker Rx3 -L1 and Rx3 loci in a complex breeding population; instead, the marker

SP5 was found to be linked with Rx3 loci in this population (Sim et al., 2015). Additionally,

QTL have been detected on chromosome 1, 4, 6, and 7 for race T1 resistance in this complex breeding population (Sim et al., 2015). The race T1 was displaced by the race T3 before the development and deployment of any race T1 resistant cultivar.

The T2 race of bacterial spot not causing HR in HI 7998 was first reported in Brazil

(Wang et al., 1990), and later detected globally (Hutton et al., 2010b). T2 race differed from race

T1 by the deletion of 680 bp region of avrRxv (Whalen et al., 1993), and was amylolytic and pectolytic (Stall et al., 1994). Originally all T2 strains belonged to X. vesicatoria (Stall et al.,

2009), but T2 races belonging to X. gardneri have also been reported. The quantitative resistance

to race T2, belonging to X. vesicatoria, has been identified in S. lycopersicum var. cerasiformae

PI 114490 (Scott et al., 1997). The T2 resistance in PI 114490 is controlled by two genes on chromosome 11 with additive gene effects (Scott et al., 2003). In addition to race T2, PI 114490 also confers resistance to races T1, T3, and T4. The T2 resistance of PI 114490 was highly correlated to race T1 resistance in PI 114490 but was poorly correlated with T3 resistance (Scott et al., 2003). This suggests T2 and T3 resistance is controlled by different genes in PI 114490

(Scott et al., 2003). Seven QTL conferred T2 resistance in PI 114490 but with minor effects

(<15%) (Yang et al., 2005a). This might be either due to a limited number of polymorphic markers used in this study or due to the polygenic nature of T2 resistance with minor effects

(Yang et al., 2005a). A common region on chromosome 11 of PI 114490 conferred partial resistance to T2, T3, and T4 races that explained 14.6 %, 63.8 % and 44.3% of the variation respectively, but this resistance locus alone was not sufficient to recover the full resistance of PI

114490 (Yang et al., 2005a). The partial resistance to race T2 has also been observed in HI

7981, the cultivar ‘Loica’ and Loica-derived breeding lines (LB97, LB99, LB60, and LB76), and the cultivar ‘Ohio 8245’ in Uruguay (Berrueta et al., 2016). However, race T2 has not become a problem in the southeast region of the US including North Carolina.

The T3 race did not induce HR in HI 7998 and was first reported in 1991 (Jones et al.,

1995). T3 is antagonistic to T1 race (Jones et al., 1988) and displaced T1 race in Florida even before the development of T1 resistant cultivar (Jones et al., 1998). When T1 and T3 are applied together on both susceptible and resistant lines, the number of T1 strains reduced compared to the application of T1 only. However, the population of T3 when applied alone or in combination with T1 did not change, suggesting competitive nature of T3 (Jones et al., 1988). The greater fitness of T3 strains might be due to bacteriocin-like substances produced by the T3 strain that suppresses the growth of the T1 strains (Tudor-Nelson et al., 2003; Hert et al., 2005).

Both hypersensitivity response and field resistance to T3 has been observed in several tomato accessions such as S. lycopersicum HI 7981, and S. pimpinellifolium PI 126932, PI

128216, LA1589 while other accessions including PI 114490, PI 126428, PI 340905-S, and PI

155372 were resistant compared to a susceptible variety (Scott et al., 1995; Sun et al., 2011a). A common locus conferred HR in PI 126932 and PI 128216, whereas a different locus was responsible for HR in HI 7981 (Hutton, 2008). HI 7981 had the highest level of resistance to T3 race, while selections of PI 126932 and PI 128216 were partially resistant to this race (Scott et al., 1995). The HR in HI 7981 is controlled by an incompletely dominant gene Xv3 on chromosome 11 (Scott et al., 1996), but the field resistance in HI 7981 is controlled quantitatively by Xv3 and other resistant loci (Scott et al., 2001). The resistance in PI 128216 is controlled by a single dominant locus Rx-4 on chromosome 11 with additive gene action

(Robbins et al., 2009). Both Xv3 and Rx4 loci were mapped in the same region on the chromosome 11 (Wang et al., 2011). The resistance in LA1589 is also controlled by a single dominant gene RxLA1589 on chromosome 11 (Sun et al., 2011a). Allelism test revealed that Xv3 loci, Rx-4 loci, and T3 resistance in PI 126932 are either closely linked (within 0.1cM) or allelic or the same gene (Wang et al., 2011). Further study identified no sequence variation in the region of Rx4 in HI 7981, PI128216, and LA1589, suggesting R genes in these lines are the same

(Baimei et al., 2015). Not surprisingly, the R genes in all these resistant lines recognize the common bacterial effector avrXv3 of T3 strain. The breeding line Fla.7600 possessing both Rx-3 and Xv-3 mediated resistance has been developed, which could be a useful resource in gene pyramiding for multiple races (Wang et al., 2011). To pyramid both Rx-3 and Xv-3 loci, Wang et al. (2011) identified SSR markers TOM196 and TOM144 and the SNP marker COSOH57 for the selection of HI 7998 derived Rx3 locus, and SNP marker cLEC-24-C3 and SL10029 for the selection of Xv3 locus.

The Rx4 gene identified in PI 128216 has been further fine mapped to a 45.1 kb region on chromosome 11 spanning by the markers pcc17 and pcc14 using F2 population obtained from the resistant accession PI 128216 and susceptible variety OH 88119 (Pei et al., 2012). The resistant

Rx4 loci differed from the susceptible loci in OH 88119 by 6-bp insertion/deletion (InDel) and eight SNPs. High-resolution map and genome sequence information allowed to map the Rx4 gene 0.07 cM away from the marker pcc12, while further refinement using only three markers

(pcc14, pcc12, and cLEC-24-c3) mapped the Rx4 gene exactly at the pcc12 marker site. This gene was found to be of NBS-LRR class of resistance gene (Pei et al., 2012). This study also tested resistance and susceptibility of 12 tomato lines using the InDel marker pcc12, where all resistant lines had the same size of PCR products as that of PI 128216 and all susceptible lines have the same size of PCR product as that of OH 88119 (Pei et al., 2012). This suggests and confirms the usefulness of the marker in tomato breeding program and, also in the cloning of Rx4 gene. However, Rx4 and Xv3 resistance couldn’t be effectively utilized in the southeast US due to the evolution of pathogens that lacked the avrXv3 effector. Again, the pathogen evolved before the development and deployment of the T3 resistant cultivar, which might be due to the selection pressure imposed by the cultivation of grape tomatoes that may have originated from S. pimpinellifolium accessions (Wang et al., 2011)

Hypersensitivity response with T3 strain was also observed in S. pennellii LA716, but the resistance of LA716 was governed by a different gene Xv4 that corresponds with the avirulence gene avrXv4 of the pathogen (Astua-Monge et al., 2000). LA 716 also confers HR resistance to race T4. The non-hypersensitive resistance observed in PI 114490 against race T3 was controlled by multiple QTL on chromosome 1, 3, 8, and 11, with QTL on chromosome 11 contributing

56.5% of the phenotypic variance (Sun et al., 2011b). The QTL controlling T3 resistance was also detected on chromosome 2, 4, and 6 in a complex breeding population in tomato (Sim et al.,

2015). In addition, three breeding lines Fla. 8326, Fla. 8233, and Fla., 8517 with race T3 have also been developed (Hutton et al., 2010a). Fla. 8233 and Fla. 8326 are large-fruited fresh- market tomato, and Fla. 8517 is a plum tomato. These three lines have HI 7998 (in all three),

PI128216 (Fla. 8233 and Fla. 8517), and PI 114490 (Fla. 8517), and PI 126932 (Fla.8326) in their pedigree (Hutton et al., 2010a).

The hypersensitive resistance of HI 7998, PI 128216, and PI 126932 against race T3

(avrXv3 effector) was overcome by race T4, as the resistance in these three lines acted on the common bacterial effector avrXv3 (Astua-Monge et al., 2000; Minsavage et al., 2003). X. perforans strains, containing another effector xopJ4 (previously called avrXv4) but lacking avrXv3 that induces an HR in S. penelli LA716 and any other tomato genotype with RxopJ4

(Xv4) resistant locus were identified; such strains are designated as race T4 (Astua-Monge et al.,

2000; Stall et al., 2009; Sharlach et al., 2013). The T4 race evolved due to the mutation in avrXv3 gene. According to a preliminary study in Florida, the xopJ4 effector was conserved among X. perforans isolates, which makes this effector a potential target to achieve the durable resistance (Sharlach et al., 2013). This can also be linked to the fact that the pathogen effectors in

T1 and T3 race evolved within five years in Florida, but the T4 race is persisting since 1998 (20 years).

The resistance locus RXopJ4 was mapped to a 20-cM segment on chromosome 3 (Astua-

Monge et al., 2000). Later, re-characterization led the RXopJ4 resistance locus from S. pennelli

LA716 mapped to a 190- kb segment on the long arm of chromosome 6 between the markers

J350 and J352 (Sharlach et al., 2013). These markers will be useful to further fine map and reach to the candidate gene controlling T4 resistance in LA 716. However, RXopJ4 loci in LA 716 is linked with several negative traits. As a result, the introgression lines containing this resistance showed autogenous leaf necrosis, low fruit yield, and small fruit size (Sharlach et al., 2013). The

linkage drags associated with this resistance are limiting the utilization LA 716 in the tomato breeding program against T4 race. Moreover, slow and weak HR along with inconsistent phenotypes were observed in the RXopJ4 heterozygotes (Sharlach et al., 2013).

The three T3 resistant advanced breeding lines (Fla. 8233, Fla. 8326, and Fla. 8517) were also reported with non-hypersensitive T4 resistance. Fla 8233 and Fla 8517 have moderate to high level of T4 resistance, and Fla 8326 has a moderate level of T4 resistance (Hutton et al.,

2010a). Originally, these three lines were bred for T1 resistance, followed by the integration of

T3 and T4 resistance (Hutton et al., 2010b). The T4 resistance in these lines was found to be mostly dominant with additive and epistatic effect controlled by multiple loci with moderate effects, which is limiting the development of resistant cultivar from these resistant advanced breeding lines. In addition, these lines provide only partial resistance, which may not be sufficient during environmental conditions that favor plant disease progress (Hutton et al.,

2010a).

Hutton et al. (2010b) conducted QTL analysis and identified T4 resistance loci and their linked molecular markers on chromosomes 1, 3, 10, and 11 in PI 114490 derived inbred backcross (IBC) populations. Among these, the QTL on chromosome 11 located near the centromere behaved as a major QTL that explained 29.4% of phenotypic variance. The QTL on chromosome 1, 3, and 10 represented minor loci and explained 2.8 to 6.3% of the phenotypic variation to T4 resistance. However, none of the IBC lines recovered the full resistance of PI

114490 (Hutton et al., 2010b). The QTL on chromosomes 11 and 3 detected in PI 114490 derived IBC population were also confirmed in the populations derived from the three breeding lines Fla. 8517, Fla. 8233, and Fla. 8326 through selective genotyping approach. The donor source for QTL on chromosome 11 was identified as PI 114490 and HI 7998 in all three breeding lines (Hutton et al., 2010b). Although HI 7998 is susceptible to T2, T3, and T4 races,

the broad-spectrum resistance against multiple races observed in the HI 7998-derived QTL might be due to the epistatic gene present in the HI 7998 derived lines but absent in the HI 7998

(Hutton et al., 2010b). However, the epistatic gene could not be identified in this study. Also, the donor source for QTL on chromosome 3 could not be determined (Hutton et al., 2010b).

To better characterize the genetics of BS resistance in PI 114490 and to capture the lost portion of PI 114490 resistance in the previous study, Scott et al. (2015) crossed PI 114490 with two susceptible lines Fla. 7324 and Fla.7613. The F2 plants were first screened with a marker associated with the self-pruning gene (sp) to select only determinate lines followed by screening with a chromosome 11 CAPS marker developed from C2_Atlg10050 to select lines that have homozygous, heterozygous, and susceptible alleles for the QTL detected on chromosome 11 in

PI 114490 against bacterial spot disease. The lines showing the highest and the lowest level of

T4 resistance were selected in F3 and F4 generations. The ninety-five F4 selections (resistant or susceptible) were genotyped using the SolCAP array, and the ninety-two F5 selections were tested for T4 race resistance in Florida, and for T1, T2, T3 race, and X. gardneri resistance in

Ohio. Single marker-trait analysis detected strong QTL effects on chromosome 2, 3, 10, and 11, which were significantly associated with the resistance to all four races and X. gardneri except

QTL on chromosome 11. Unlike in the studies conducted by Yang et al. (2005a) and Sun et al.

(2011b), the PI 11490 derived QTL on chromosome 11 did not show a significant effect on race

T3 but had the highest effect on race T4. The QTL on chromosome 3 showed the highest effect against all races (Scott, et al., 2015). However, none of the lines were highly resistant to all the races. The most resistant lines are being re-evaluated in Florida and Ohio to identify the QTL conferring better resistance and to check if the PI 114490 derived resistant lines have recovered the full resistance of PI 114490 (Scott et al., 2015).

Recently, Bhattarai et al. (2017) identified the moderate level of resistance against race

T4 in S. pimpinellifolium derived breeding lines including 74L-1W (2008), NC2CELBR, 081-12-

1X-gsms, NC22L-1 (2008), and 52LB-1, in the field and greenhouse studies. Further work is needed to characterize the resistance observed in these lines and to evaluate if the observed resistance in these lines is sufficient to manage the disease before integrating the resistance in the cultivar.

Breeding Tomato for Resistance to Xanthomonas gardneri

The outbreaks of X. gardneri is recent, and there are not many breeding efforts against this species. X. gardneri has been reported from three states of US- Ohio, Michigan, and

Pennsylvania (Kim et al., 2010; Ma et al., 2011), Costa Rica (Bouzar et al., 1999), Canada and

Brazil (Quezado-Duval et al., 2004), Russia (Kornev et al., 2007), the Indian Ocean regions

(Hamza et al., 2010), and Bulgaria (Aleksandrova et al., 2014).

Although resistance genes against X. gardneri in tomato have not been characterized so far, some tomato lines showing resistance against X. gardneri, have been identified. Three S. pimpinellifolium accessions PI 128216, LA2533, and LA1936 showed both HR in the greenhouse and field resistance against X. gardneri (Liabeuf et al., 2015). These accessions also showed resistance against X. perforans race T3; HR to X. perforans was observed before the HR to X. gardneri. Among the three resistant accessions, PI 128216 and LA 2533 were used to develop four segregating population. The disease resistance in these populations was possibly controlled by one to four loci with moderate heritability (Liabeuf et al., 2015). Another study identified two varieties- IZK Alya (cherry type) and Nikolina F1 (determinate large fruit type) showing resistance to X. gardneri, but no HR was observed in this study (Aleksandrova et al.,

2014). Conversely, two resistance genes Bs7 and Bs3 have been reported against X. gardneri

avirulence genes avrBs7 and AvrHah1, respectively in pepper, another member of the

Solanaceae family (Schornack et al., 2008; Potnis et al., 2012).

Transgenic Resistance to Bacterial Spot

Transgenic resistance offers a promising tool to obtain bacterial spot resistance in tomato.

This approach seeks the solutions to eliminate linkage drags observed in the lines developed from backcrossing. Transgenic resistance offers a way to incorporate bacterial spot resistance genes from another crop like pepper into a tomato, which can’t be achieved through conventional breeding methods. The Bs2 gene targeting the bacterial core effector avrBs2 from pepper has been deployed in tomato through the transgenic approach (Tai et al., 1999). The Bs2 gene confers resistance to all field strains of Xanthomonas infecting tomatoes. In a multi-year replicated field trial in Florida in commercial tomato production regions, Bs2 transgenic tomatoes showed the highest disease resistance, against current field races of Xanthomonas, among all the genotypes tested with significantly higher yields compared to controls in the absence of bactericidal compounds (Horvath et al., 2012). The Bs2 gene when incorporated into a highly susceptible California tomato cultivar ‘VF36’ significantly reduced the disease severity and increased the yield compared to non-transformed lines without any adverse effects in the transgenic tomatoes (Horvath et al., 2012). Likewise, the commercial parent lines and hybrids of tomato from University of Florida breeding programs with Bs2 gene yielded double compared to the near-isogenic non-transgenic lines under Florida field conditions (Horvath et al., 2013).

Considering potential concerns of the transgenic products, Horvath et al. (2012) also highlighted the safety profiles of Bs2 transgenic tomatoes as- i) Bs2 protein belongs to the one of the largest family of plant proteins (NB-LRR family); ii) Bs2 protein naturally occurs in pepper, which has been widely consumed with no known adverse effects on human; iii) there is a low risk of

transgene escape as tomato is >99% self-pollinating; iv) sexually compatible relatives of tomato has not been reported in North America. Besides transgenic tomatoes, other transgenic disease resistant crops such as papaya and squash have been commercialized and widely adopted with no safety issues (Fuchs and Gonsalves, 2007).

Notably, a mutation in avrBs2 effector has been observed in rare Xanthomonas strains that could overcome the resistance conferred by the Bs2 gene (Dangl et al., 2013). This suggests that multiple targets for such core effectors are necessary to achieve durable resistance. Two recessive genes in pepper bs5 and bs6 are reported to confer resistance to all races of bacterial spot in peppers. These genes are predicted to last longer than the dominant resistance genes as these genes don’t involve specific gene-for-gene interaction with the bacterial effectors (Vallejos et al., 2010). Cloning of both bs5 and bs6 genes will enhance the tomato breeding program for bacterial spot disease resistance through the transgenic approach.

Additionally, the transgenic approach could also allow integration of Bs3 and Bs7 genes from pepper into a tomato to achieve resistance against X. gardneri. Moreover, transgenic expression of EF-Tu receptor (EFR), a pattern recognition receptor (PRR) from the cruciferous plant Arabidopsis thaliana, also conferred broad-spectrum bacterial resistance, including X. perforans in tomato. However, the resistance was less efficient against X. perforans compared to other bacterial pathogens and compared to that conferred by Bs2 transgenic tomato plants

(Lacombe et al., 2012). Therefore, deployment of Bs2 gene in combination with other novel resistance genes would provide durable resistance to bacterial spot disease in tomato.

Although the effectiveness of Bs2 transgenic tomatoes to manage bacterial spot is successfully demonstrated in multiple field trials, the Bs2 transgenic tomatoes could not be commercialized because of public concerns towards genetically engineered food products and due to the lack of funding to complete expensive regulatory studies required by the USDA. The

public awareness about GM food products, its process, and the safety is critical to change the negative attitudes and perceptions of consumer towards GM food.

1.2. Breeding Tomatoes for Fruit Morphology and Color Traits

The origin of domestication for tomato is in the Andean region of South America and was spread to Europe as part of the Columbian exchange. During domestication and breeding, the tomato was selected for various fruit shapes, sizes, and color. As a result, today cultivated tomatoes have diverse shapes, sizes, and color. The size of cultivated tomatoes ranges from small cherry size to medium, large and extra-large fruited tomatoes. The shape of tomato can be classified into 8 categories-flat, round, rectangular, ellipsoid, heart, long, obovoid, and oxheart

(Rodriguez et al., 2011). The color of cultivated tomatoes ranges from green, pink, yellow, orange, red, to dark red and purple (Paran and van der Knaap, 2007).

The tomato fruit morphology and color determine their market values, appearance, and consumer acceptability. Growers demand highly profitable tomatoes, which is correlated with the tomato size. Consumers demand red, visually flawless tomatoes that are firm and sweet tasting with different shapes and sizes depending on their culinary purposes (Piombino et al.,

2013). For instances, small cherry tomatoes are preferred for salad, while large round tomatoes are preferred for slicing in sandwiches and burgers. The processing industry prefers rectangular and blocky tomatoes as these shapes prevent the fruit from rolling from conveyor belts during mechanical harvesting and have high solids content (Visa et al., 2014). In addition, the consumers tend to judge tomatoes first on visual appearance, and taste second, although both traits are important (Barrett et al., 2010). The importance of tomato fruit shape, size, and color to consumers makes the genetic improvement of those traits a priority for tomato breeders. A better

understanding of genetic basis of fruit morphology and appearance will aid in their genetic improvements through breeding efforts.

However, such traits are quantitative and controlled by multiple genes that make it difficult to fully elucidate the genetic basis controlling fruit size, shape, and color in tomato.

Fruit color is the function of different pigments including carotenoids, chlorophyll, flavonoids, and anthocyanins, among which carotenoid have a primary role (Wang et al., 2015). The mutations in the carotenoid pathway have resulted in different colors in tomato (Ballester et al.,

2010). Several genes regulating carotenoid content, hence fruit color have been identified in tomato. For instance, phytoene synthase gene 1(Psy1) on chromosome 3, encoding the r (yellow flesh) mutant phenotype confers yellow flesh color in ripe tomato (Fray and Grierson, 1993). The single dominant lycopene β-cyclase (LcyB) gene on chromosome 6 increases β-carotene in the tomato fruit that gives the orange color of the fruits, while old-gold (og), a recessive mutation of the LcyB, destroys β-carotene and increases lycopene giving red color in tomato fruit (Ronen et al., 2000). Likewise, lycopene ε-cyclase (εLCY) gene on chromosome 12 encoding DEL

(DELTA, reddish orange) mutant phenotype and carotenoid isomerase (CRITSO) gene on chromosome 10 encoding t (tangerine) mutant phenotype confers orange fruit and flesh color

(Ronen et al., 1999; Isaacson et al., 2002). Other genes increasing fruit carotenoids and pigmentation include UV-damaged DNA binding protein 1 (DDB1) and zeaxanthin epoxidase

(ZE) on chromosome 2 associated with high pigment mutations hp-1, hp-1w, and hp-3; and deetiolated 1(DET1) on chromosome 1 associated with high pigment mutations hp-2, hp-2i, and hp-2dg (reviewed by Levin and Schaffer, 2013). A recessive locus y on the short arm of chromosome 1 was responsible for a colorless fruit epidermis that results in pink fruit in tomato when combined with red flesh (Ballester et al., 2010). Further study revealed that ripening dependent downregulation of the naringenin chalcone in the fruit peel is responsible for the

development of pink fruit in tomato, and the gene encoding MYB transcription factor SlMYB12 is a potential candidate gene for this trait (Ballester et al., 2010). Similarly, the green fruit of tomato is due to amino acid substitution in homologs of STAY-GREEN protein of rice (Barry et al., 2008).

Genetic studies on tomato fruit shapes and sizes started during 1900’s (Grandillo et al.,

1996), and the QTL analysis on fruit morphology was conducted in 1988 (Weller et al., 1988).

After that, several QTL controlling tomato fruit shapes and sizes have been identified, however, only few genes have been cloned. Fruit shape genes include i) SUN on short arm of chromosome

7 (Xiao et al., 2008), ii) OVATE on chromosome 2 (Liu et al., 2002), iii) FASCIATED (FAS) on chromosome 11 (Cong et al., 2008), and iv) LOCULE NUMBER (LC) on chromosome 2 (Muños et al., 2011). Fruit size genes include i) Cell Number Regulator (CNR)/FW2.2 on chromosome 2

(Frary et al., 2000) and ii) SlKLUH/FW3.2 on chromosome 3 (Chakrabarti et al., 2013). FAS and

LC also affect fruit sizes in addition to fruit shape by changing the locule number (Tanksley,

2004).

Both SUN and OVATE confer fruit elongation, but SUN’s effect is more prominent than that of OVATE. The SUN locus in the domesticated species evolved from the gene duplication event, which increased the gene expression in the domesticated species resulting in more elongated shape compared to wild species (Xiao et al., 2008). SUN confers uniform elongation to maintain bilateral symmetry as observed in most commercially grown tomatoes, heirloom, and oxheart tomatoes, whereas OVATE is responsible for the asymmetric elongation causing neck constriction or pear shape as observed in ellipsoid and obovoid varieties of grape tomato

(Gonzalo and van der Knapp, 2008). The effect of SUN is determined after the pollination during cell multiplication phase of fruit development, while the effect of OVATE is observed before pollination during early floral development (van der Knapp and Tanksley, 2001). SUN encodes a

member of the IQ domain, and OVATE encodes for ovate family protein (OFP) (Liu et al., 2002;

Xiao et al., 2008).

Likewise, both FAS and LC control locule number and flat fruit shape, but FAS mutation has a higher effect compared to LC. A fas mutant has a potential to produce more than 15 locules and unfused carpels that not only change the fruit shape but also results into increase in the number of floral organs such as sepals, petals, and stamens (Tanksley, 2004). The fas mutation is present in most multilocular fresh market tomatoes, whereas large and extra-large fruited tomatoes carry mutations in both fas and lc loci. An epistatic interaction between fas and lc loci has been reported as a causal factor for the extremely large size of tomatoes mediated through high locule number as reviewed by van der Knaap et al. (2014). FAS encodes member of the

YABBY family, while LC encodes for the members of WOX family (Cong et al., 2008; Munos et al., 2011).

FW2.2 is the first weight or fruit size QTL cloned from fruits and vegetables (Frary et al.,

2000). The size of tomato is determined by the number of cells within ovary before fertilization, the number of fertilizations, the number of cell divisions occurred post-fertilization inside the fruit, and the cell enlargement (Lippman and Tanksley, 2001). The FW2.2 gene controls fruit size by encoding a protein of the cell number regulatory (CNR) family that negatively regulates the cell number (Guo and Simmons, 2011). FW3.2, the second cloned fruit weight locus, regulates the fruit size by increasing cell layers and delaying fruit ripening (Chakrabarti et al.,

2013). The FW3.2 locus encodes a cytochrome P450 of the CYP78A class and the likely ortholog of Arabidopsis KLUH. Therefore, the corresponding gene for FW3.2 was renamed as

SlKLUH (Chakrabarti et al., 2013; Zhang et al., 2012;). A regulatory SNP M9 in the promoter region of ORF6 showed highly significant association with the increased fruit size regulated by

FW3.2 (Chakrabarti et al., 2013).

Recently, another fruit weight gene Cell Size Regulator (CSR) has been cloned and characterized (Mu et al., 2017). The locus of this gene was previously identified as a minor QTL that explained 8% of the phenotypic variance and was fine mapped near to the FAS gene on chromosome 11 (Huang and van der Knaap, 2011; Mu et al., 2017). Additionally, some major loci controlling tomato fruit shape and size have been fine mapped. The locus fw11.2 controlling fruit size/weight has been fine mapped to a region of ~750 kb containing 66 candidate genes above FAS and fw11.3 loci (Illa-berenguer et al., 2015). A locus controlling fruit shape fs8.1, has been fine mapped to 3 Mb regions on chromosome 8 near centromere containing 122 candidate genes (Clevenger, 2012). fs8.1 determines the pattern of tomato carpel shape in early flower development stage and confers blocky and slightly elongated shape in processing tomatoes

(Grandillo et al., 1996; Ku et al. 2000). The two fruit shape loci sov1 and sov2, which function by suppressing the OVATE mutation have been mapped to chromosome 10 (within 1.2Mb interval) and 11(within 2.4 Mb interval) respectively in an intra-specific population (Rodriguez et al., 2013). While sov1 was responsible for both obovoid and elongated shape, sov2 mainly conferred elongated fruit shape. (Rodriguez et al., 2013). Fine-mapping of sov1 on chromosome

10 identified two candidate genes (Monteforte et al., 2014). Other less characterized QTL controlling tomato fruit morphology include fw1.1, fw2.1, fw3.1, fw3.3, fw4.1, and fw11.2 for size trait and where fs2.1, lcn2.4, lcn5.1, and lcn6.1 for shape trait (Grandillo et al., 1996; Illa- berenguer et al., 2015).

The presence of other loci controlling fruit shapes and sizes in addition to the cloned genes provides the evidence that there might be other undetected loci controlling tomato fruit morphology. Likewise, several QTL controlling fruit color in tomato have been detected on other genomic locations apart from the genes involved in carotenoid pathway, suggesting there are other more genes/loci controlling fruit color in tomato in addition to the known genes involved in

carotenoid biosynthesis pathway (Saliba-Colmbani et al., 2001; Liu et al., 2003; Liu et al., 2017).

Therefore, detection of additional genetic factors underlying tomato fruit morphology and color will help to explain fully the observed variation among fruit shapes, sizes, and color in cultivated tomato species. The genes and loci associated with tomato fruit morphology and color are mostly identified in the inter-specific population derived from the cross of wild and domesticated tomato species due to unavailability of sufficient polymorphic markers. The confounding effect of different genetic backgrounds and deleterious linkage drag have made the introgression of potentially beneficial alleles mapped in inter-specific populations difficult (Lecomte et al. 2004).

The QTL mapping study using intra-specific population will minimize linkage drags and allow detection of the minor allele effects, as most large effect alleles are homozygous in the population derived from closely related parents (Rodriguez et al., 2013).

1.3 Research Objectives

In this research, we sought to characterize the Xs strains in NC, and then utilize the pathogen information to identify genetic loci resistance to the predominant races/species of Xs prevalent in NC. We also sought to study the genetic basis of diverse shapes, sizes, and color in tomato as these traits often determine the market value and culinary purposes of the tomato. A better understanding of the loci influencing bacterial spot disease resistance, fruit shape, size, and color in tomato will help improve these traits as desired by the growers and consumers in released hybrids.

The specific objectives of this research were:

1. Chapter 2: i) To evaluate and determine the proportion of field strains of Xs that are

resistant to copper and streptomycin; ii) assess the genetic diversity of Xs populations

using BOX-PCR assay and species-specific hrpB7 gene probes; iii) to identify pathogenic

races of Xs strains from NC, and iv) to determine the phylogenetic relationships and

genetic distinctness of strains of Xs in NC and compare them with the other reported

strains of Xs from different regions.

2. Chapter 3: i) To identify QTL associated with the bacterial spot disease resistance

against race T4 within the intra-specific recombinant inbred line (RIL) population of

tomato-derived from two elite breeding lines; ii) to verify detected QTL in an

independent segregating population of tomato.

3. Chapter 4: To detect segregating QTL that influence tomato fruit i) shape, ii) size, and

iii) color in the intra-specific recombinant inbred line (RIL) population of tomato-derived

from two elite breeding lines.

REFERENCES

Abbasi, P. A., and Weselowski, B. (2015). Efficacy of Bacillus subtilis QST 713 formulations,

copper hydroxide, and their tank mixes on bacterial spot of tomato. Crop Protection

74:70-76.

Abbasi, P. A., Khabbaz, S. E., Weselowski, B., and Zhang, L. (2015). Occurrence of copper-

resistant strains and a shift in Xanthomonas spp. causing tomato bacterial spot in Ontario.

Canadian Journal of Microbiology 61:753-761.

Abbasi, P., Cuppels, D., and Lazarovits, G. (2003). Effect of foliar applications of neem oil and

fish emulsion on bacterial spot and yield of tomatoes and peppers. Canadian Journal of

Plant Pathology 25:41-48.

Agricultural Marketing Resource Center. (2018). Tomatoes. Retrieved from

https://www.agmrc.org/commodities-products/vegetables/tomatoes

Al-Dahmani, J. H., Abbasi, P. A., Miller, S. A., and Hoitink, H. A. (2003). Suppression of

bacterial spot of tomato with foliar sprays of compost extracts under greenhouse and field

conditions. Plant Disease 87:913-919.

Aleksandrova, K., Ganeva, D., and Bogatzevska, N. (2014). Xanthomonas gardneri–

characterization and resistance of Bulgarian tomato varieties. Türk Tarım ve Doğa

Bilimleri 7:1540-1545.

Araujo, E. R., Pereira, R. C., Ferreira, M. A. S. V., Café-Filho, A. C., Moita, A. W., and

Quezado-Duval, A.M. (2010). Effect of temperature on pathogenicity components of

tomato bacterial spot and competition between Xanthomonas perforans and X. gardneri.

In III International Symposium on Tomato Diseases 914:39-42.

Araújo, E. R., Pereira, R. C., Ferreira, M. A. S. V., Quezado-Duval, A.M., and Moita, A. W.

(2012). Sensitivity of xanthomonads causing tomato bacterial spot to copper and

streptomycin and in vivo infra-specific competitive ability in Xanthomonas perforans

resistant and sensitive to copper. Journal of Plant Pathology 94:79-87.

Araújo, E., Costa, J., Pontes, N., and Quezado-Duval, A., (2015). Xanthomonas perforans and X.

gardneri associated with bacterial leaf spot on weeds in Brazilian tomato fields.

European Journal of Plant Pathology 14:543-548.

Astua-Monge, G., Minsavage, G. V., Stall, R. E., Vallejos, C. E., Davis, M. J., and Jones, J. B.

(2000). Xv4-Avrxv4: A new gene-for-gene interaction identified between Xanthomonas

campestris pv. vesicatoria race T3 and the wild tomato relative Lycopersicon pennellii.

Molecular Plant-Microbe Interactions 13:1346-1355.

Baimei, Z., Haipeng, C., Junjie, D., and Wencai, Y. (2015). Allelic tests and sequence analysis of

three genes for resistance to Xanthomonas perforans race T3 in tomato. Horticultural

Plant Journal 1:41-47.

Ballester, A. R., Molthoff, J., de Vos, R., te Lintel Hekkert, B., Orzaez, D., Fernández-Moreno,

J. P., Tripodi, P., Grandillo, S., Martin, C., Heldens, J., and Ykema, M. (2010).

Biochemical and molecular analysis of pink tomatoes: deregulated expression of the gene

encoding transcription factor SlMYB12 leads to pink tomato fruit color. Plant Physiology

152:71-84.

Balogh, B., Jones, J. B., Iriarte, F. B., and Momol, M. T. (2010). Phage therapy for plant disease

control. Current Pharmaceutical Biotechnology 11:48-57.

Balogh, B., Jones, J. B., Momol, M. T., Olson, S. M., Obradovic, A., King, P., and Jackson, L. E.

(2003). Improved efficacy of newly formulated bacteriophages for management of

bacterial spot on tomato. Plant Disease 87:949-954.

Barrett, D. M., Beaulieu, J. C., and Shewfelt, R. (2010). Color, flavor, texture, and nutritional

quality of fresh-cut fruits and vegetables: desirable levels, instrumental and sensory

measurement, and the effects of processing. Critical Reviews in Food science and

Nutrition 50:369-389.

Barry, C. S., McQuinn, R. P., Chung, M. Y., Besuden, A., and Giovannoni, J. J. (2008). Amino

acid substitutions in homologs of the STAY-GREEN protein are responsible for the

green-flesh and chlorophyll retainer mutations of tomato and pepper. Plant Physiology

147:179–187. doi:10.1104/pp.108.118430

Basim, H., Minsavage, G. V., Stall, R. E., Wang, J. F., Shanker, S., and Jones, J. B. (2005).

Characterization of a unique chromosomal copper resistance gene cluster from

Xanthomonas campestris pv. vesicatoria. Applied and Environmental Microbiology

71:8284-8291.

Behlau, F., Jones, J. B., Myers, M. E., and Graham, J. H. (2012). Monitoring for resistant

populations of Xanthomonas citri subsp. citri and epiphytic bacteria on citrus trees

treated with copper or streptomycin using a new semi-selective medium. European

Journal of Plant Pathology 132:259-270.

Bender, C. L., Malvick, D. K., Conway, K. E., George, S., and Pratt, P. (1990). Characterization

of pXV10A, a copper resistance plasmid in Xanthomonas campestris pv. vesicatoria.

Applied and Environmental Microbiology 56:170-175.

Bender, C., and Cooksey, A. A. (1986). Indigenous plasmids in Pseudomonas syringae pv.

tomato: conjugative transfer and role in copper resistance. Journal of Bacteriology

165:534-541.

Berrueta, M. C., Gimenez, G., Galván, G. A., and Borges, A. (2016). New sources of partial

resistance to bacterial spot race T2 in processing tomatoes. Horticultura Brasileira

34:326-332.

Bhattarai, K., Louws, F. J., Williamson, J. D., and Panthee, D. R., (2017). Resistance to

Xanthomonas perforans race T4 causing bacterial spot in tomato breeding lines. Plant

Pathology 66:1103-1109.

Bouzar, H., Jones, J. B., Somodi, G. C., Stall, R. E., Daouzli, N., Lambe, R. C., Gastelum, R. F.,

and Correa, R. T. (1996). Diversity of Xanthomonas campestris pv. vesicatoria race

variation in tomato and pepper fields of Mexico. Canadian Journal of Plant Pathology

18:75-77.

Bouzar, H., Jones, J. B., Stall, R. E., Louws, F. J., Schneider, M., Rademaker, J. L. W., De

Bruijn, F. J., and Jackson, L. E. (1999). Multiphasic analysis of xanthomonads causing

bacterial spot disease on tomato and pepper in the Caribbean and Central America:

evidence for common lineages within and between countries. Phytopathology 89:328-

355.

Briceno-Montero, G., and Miller, S.A. (2005). Evaluation of biological control options for

bacterial spot management during tomato transplant production. Acta Horticulturae

695:357-365.

Burlakoti, R. R., Hsu, C. F., Chen, J. R., and Wang, J. F. (2018). Population dynamics of

Xanthomonads associated with bacterial spot of tomato and pepper during 27 years

across Taiwan. Plant Disease. 102: 1348-1356.

Büttner, D., and Bonas, U. (2010). Regulation and secretion of Xanthomonas virulence factors.

FEMS Microbiology Reviews 34:107-133.

Byrne, J. M., Dianese, A. C., Ji. P., Campbell, H. L., Cuppels, D. A., Louws, F. J., Miller, S. A.,

Jones, J. B., and Wilson, M. (2005). Biological control of bacterial spot of tomato under

field conditions at several locations in North America. Biological Control 32: 408-418.

Chakrabarti, M., Zhang, N. A., Sauvage, C., Muños, S., Blanca, J., Cañizares, J., Diez, M. J.,

Schneider, R., Mazourek, M., McClead, J., and Causse, M. (2013). A cytochrome P450

regulates a domestication trait in cultivated tomato. Proceedings of the National Academy

of Sciences 110:17125-17130.

Clevenger J. (2012). Metabolic and genomic analysis of elongated fruit shape in tomato

(Solanum lycopersicum). M.S. thesis dissertation, Ohio State University, OH, USA.

Cong, B., Barrero, L. S, and Tanksley, S. D. (2008). Regulatory change in YABBY-like

transcription factor led to evolution of extreme fruit size during tomato domestication.

Nature Genetics 40:800–804.

Constantin, E. C., Cleenwerck, I., Maes, M., Baeyen, S., Van Malderghem, C., De Vos, P. and

Cottyn, B. (2016). Genetic characterization of strains named as Xanthomonas axonopodis

pv. dieffenbachiae leads to a taxonomic revision of the X. axonopodis species complex.

Plant Pathology 65:792-806.

Coqueiro, D. S. O., Maraschin, M., and Piero, R. M. D. (2011). Chitosan reduces bacterial spot

severity and acts in phenylpropanoid metabolism in tomato plants. Journal of

Phytopathology 159:488-494.

Cuppels, D. A., Louws, F. J., and Ainsworth, T. (2006). Development and evaluation of PCR

based diagnostic assays for the bacterial speck and bacterial spot pathogens of tomato.

Plant Disease 90:451-458.

Dangl, J. L., Horvath, D. M., and Staskawicz, B.J. (2013). Pivoting the plant immune system

from dissection to deployment. Science 341:746-751.

De Ley, J. (1978). Modern molecular methods in bacterial taxonomy: evaluation, application,

prospects. Proceedings of the 4th International Conference on Plant pathogenic Bacteria,

Angers 347-357.

Doidge, E. (1921). A tomato canker. Annals of Applied Biology 7:407-430.

Dye, D. (1966). Cultural and biochemical reactions of additional Xanthomonas spp. New

Zealand Journal of Science 9:913.

Frary, A., Nesbitt, T. C., Frary, A., Grandillo, S., Van Der Knaap, E., Cong, B., Liu, J., Meller,

J., Elber, R., Alpert, K. B., and Tanksley, S. D. (2000). fw2. 2: a quantitative trait locus

key to the evolution of tomato fruit size. Science 289:85-88.

Fray, R. G, and Grierson, D. (1993). Identification and genetic-analysis of normal and mutant

phytoene synthase genes of tomato by sequencing, complementation and co-suppression.

Plant Molecular Biology 22:589–602.

Fuchs, M., and Gonsalves, D. (2007). Safety of virus-resistant transgenic plants two decades

after their introduction: lessons from realistic field risk assessment studies. Annual

Review of Phytopathology 45:173-202.

Gardner, M., and Kendrick, J. (1921). Bacterial spot of tomato. Journal of Agricultural Research

21:123-156.

Gonzalo, M. J., and Van Der Knaap, E. (2008). A comparative analysis into the genetic bases of

morphology in tomato varieties exhibiting elongated fruit shape. Theoretical and Applied

Genetics 116:647-656.

Goode, M. J., and Sasser, M. (1980). Prevention-the key to controlling bacterial spot and speck

of tomato. Plant Disease 64: 831-834.

Grandillo, S., Ku, H. M., and Tanksley, S. D. (1996). Characterization of fs8. 1, a major QTL

influencing fruit shape in tomato. Molecular Breeding 2:251-260.

Graves, A. S., and Alexander, S.A. (2002). Managing bacterial speck and spot of tomato with

acibenzolar-S-methyl in Virginia. Plant Health Progress doi:10.1094/PHP-2002-0220-01-

RS.

Griffin, K., Gambley, C., Brown, P., and Li, Y. (2017). Copper-tolerance in Pseudomonas

syringae pv. tomato and Xanthomonas spp. and the control of diseases associated with

these pathogens in tomato and pepper. A systematic literature review. Crop Protection 96:

144-150.

Guo, M., and Simmons, C.R. (2011). Cell number counts–the fw2. 2 and CNR genes and

implications for controlling plant fruit and organ size. Plant Science 181:1-7.

Hamza, A. A., Robène-Soustrade, I., Jouen, E., Gagnevin, L., Lefeuvre, P., Chiroleu, F., and

Pruvost, O. (2010). Genetic and pathological diversity among Xanthomonas strains

responsible for bacterial spot on tomato and pepper in the southwest Indian Ocean region.

Plant Disease 94:993-999.

Hert, A. P., Roberts, P. D., Momol, M. T., Minsavage, G. V., Tudor-Nelson, S. M., and Jones, J.

B. (2005). Relative importance of bacteriocin-like genes in antagonism of Xanthomonas

perforans tomato race 3 to Xanthomonas euvesicatoria tomato race 1 strains. Applied and

Environmental Microbiology 71:3581-3588.

Horvath, D. M., Pauly, M. H., Hutton, S. F., Vallad, G. E., Scott, J. W., Jones, J. B., Stall, R. E.,

Dahlbeck, D., Staskawicz, B. J., Tricoli, D., and Deynze, A.V. (2013). The pepper Bs2

gene confers effective field resistance to bacterial leaf spot and yield enhancement in

Florida tomatoes. In IV International Symposium on Tomato Diseases 1069:47-51.

Horvath, D. M., Stall, R. E., Jones, J. B., Pauly, M. H., Vallad, G. E., Dahlbeck, D., Staskawicz,

B. J., and Scott, J.W. (2012). Transgenic resistance confers effective field level control of

bacterial spot disease in tomato. PLoS One 7:e42036.

Huang, C. H., and Vallad, G. E. (2018). Soil applications of acibenzolar-S-methyl induce

defense gene expression in tomato plants against bacterial spot. European Journal of

Plant Pathology 150:971-981.

Huang, C. H., Vallad, G. E., Zhang, S., Wen, A., Balogh, B., Figueiredo, J. F. L., Behlau, F.,

Jones, J. B., Momol, M. T., and Olson, S. M. (2012). Effect of application frequency and

reduced rates of acibenzolar-S-methyl on the field efficacy of induced resistance against

bacterial spot on tomato. Plant Disease 96:221-227.

Huang, Z., and van der Knaap, E. (2011). Tomato fruit weight 11.3 maps close to fasciated on

the bottom of chromosome 11. Theoretical and Applied Genetics 123:465-474.

Hutton, S. (2008). Inheritance and mapping of resistance to bacterial spot race T4 (Xanthomonas

perforans) in tomato, and its relationship to race T3 hypersensitivity, and inheritance of

race T3 hypersensitivity from PI 126932. Doctoral dissertation, University of Florida.

Hutton, S. F., Scott, J. W., and Jones, J. B. (2010a). Inheritance of resistance to bacterial spot

race T4 from three tomato breeding lines with differing resistance backgrounds. Journal

of the American Society for Horticultural Science 135:150-158.

Hutton, S. F., Scott, J. W., Yang, W., Sim, S. C., Francis, D. M., and Jones, J. B. (2010b).

Identification of QTL associated with resistance to bacterial spot race T4 in tomato.

Theoretical and Applied Genetics 121:1275-1287.

Illa-Berenguer, E., Van Houten, J., Huang, Z., and van der Knaap, E. (2015). Rapid and reliable

identification of tomato fruit weight and locule number loci by QTL-seq. Theoretical and

Applied Genetics 128:1329-1342.

Isaacson, T., Ronen, G., Zamir, D., and Hirschberg, J. (2002). Cloning of tangerine from tomato

reveals a carotenoid isomerase essential for the production of β-Carotene and

xanthophylls in plants. Plant Cell 14:333–342.

Ji, P., Campbell, H. L., Kloepper, J. W., Jones, J. B., Suslow, T.V., and Wilson, M. (2006).

Integrated biological control of bacterial speck and spot of tomato under field conditions

using foliar biological control agents and plant growth-promoting rhizobacteria.

Biological Control 36:358-367.

Jibrin, M. O., Timilsina, S., Potnis, N., Minsavage, G. V., Shenge, K. C., Akpa, A. D., Alegbejo,

M. D., Beed, F., Vallad, G. E., and Jones, J. B. (2015). First report of atypical

Xanthomonas euvesicatoria strains causing bacterial spot of tomato in Nigeria. Plant

Disease 99:415-415.

Jones, J. B., Bouzar, H., Somodi, G. C., Stall, R. E., Pernezny, K., El-Morsy, G., and Scott, J. W.

(1988). Evidence for the preemptive nature of tomato race 3 of Xanthomonas campestris

pv. vesicatoria in Florida. Phytopathology 88:33-38.

Jones, J. B., Bouzar, H., Stall, R. E., Almira, E. C., Roberts, P. D., Bowen, B. W., Sudberry, J.,

Strickler, P. M., and Chun, J. (2000). Systematic analysis of xanthomonads

(Xanthomonas spp) associated with pepper and tomato lesions. International Journal of

Systematic and Evolutionary Microbiology 50:1211-1219.

Jones, J. B., Jackson, L. E., Balogh, B., Obradovic, A., Iriarte, F. B., and Momol, M. T. (2007).

Bacteriophages for plant disease control. Annual Review of Phytopathology 45:245-262.

Jones, J. B., Lacy, G. H., Bouzar, H., Stall, R. E., and Schaad, N. W. (2004). Reclassification of

the xanthomonads associated with bacterial spot disease of tomato and pepper.

Systematic and Applied Microbiology 27:755-762.

Jones, J. B., Stall, R. E., and Bouzar, H. (1998). Diversity among xanthomonads pathogenic on

pepper and tomato. Annual Review of Phytopathology 36:41-58.

Jones, J. B., Stall, R. E., Scott, J. W., Somodi, G. C., Bouzar, H., and Hodge, N. C. (1995). A

third tomato race of Xanthomonas campestris pv. vesicatoria. Plant Disease 79:395-398.

Jones, J., Pohroneznt, K. L., Stall, R., and Jones, J. P. (1986). Survival of Xanthomonas

campestris pv. vesicatoria in Florida on tomato crop residue, weeds, seeds, and volunteer

tomato plants. Phytopathology 76:430-434.

Kebede, M., Timilsina, S., Ayalew, A., Admassu, B., Potnis, N., Minsavage, G. V., Goss, E. M.,

Hong, J. C., Strayer, A., Paret, M., and Jones, J. B. (2014). Molecular characterization of

Xanthomonas strains responsible for bacterial spot of tomato in Ethiopia. European

Journal of Plant Pathology140:677-688.

Kim, S. H., Olson, T. N., Peffer, N. D., Nikolaeva, E. V., Park, S., and Kang, S. (2010). First

report of bacterial spot of tomato caused by Xanthomonas gardneri in Pennsylvania.

Plant Disease 95:638.

Kornev, K. P., Matveeva, E. V., Pekhtereva, E. S., Polityko, V. A., Ignatov, A. N., Punina, N. V.,

and Schaad, N.W. (2007). Xanthomonas species causing bacterial spot of tomato in the

Russian Federation. In II International Symposium on Tomato Diseases 808:243-246.

Ku, H. M., Grandillo, S., and Tanksley, S. D. (2000). fs8. 1, a major QTL, sets the pattern of

tomato carpel shape well before anthesis. Theoretical and Applied Genetics 101:873-878.

Lacombe, A., Wu, V. C., White, J., Tadepalli, S., and Andre, E. E. (2012). The antimicrobial

properties of the lowbush blueberry (Vaccinium angustifolium) fractional components

against foodborne pathogens and the conservation of probiotic Lactobacillus rhamnosus.

Food Microbiology 30:124-131.

Lecomte, L., Duffé, P., Buret, M., Servin, B., and Causse, M. (2004). Marker-assisted

introgression of five QTLs controlling fruit quality traits into three tomato lines revealed

interactions between QTLs and genetic backgrounds. Theoretical and Applied Genetics

109:658-68.

Levin, I., and Schaffer, A. A. (2013). Molecular mapping of complex traits in tomato. In

Genetics, Genomics, and Breeding of Tomato (eds Liedl, B. E. et al.) Ch. 5.

Lewis Ivey, M., Strayer, A., Sidhu, J., and Minsavage, G. (2016). Bacterial leaf spot of tomato

(Solanum lycopersicum) in Louisiana is caused by Xanthomonas perforans, tomato race

4. Plant Disease 100:1233-1233.

Liabeuf, D., Francis, D., and Sim, S. (2015). Screening cultivated and wild tomato germplasm

for resistance to Xanthomonas gardneri. Acta Horticulturae 1069:65-70

Lippman, Z., and Tanksley, S.D. (2001). Dissecting the genetic pathway to extreme fruit size in

tomato using a cross between the small-fruited wild species Lycopersicon

pimpinellifolium and L. esculentum var. Giant Heirloom. Genetics 158:413-422.

Liu, J., Van Eck, J., Cong, B., and Tanksley, S. D. (2002). A new class of regulatory genes

underlying the cause of pear-shaped tomato fruit. Proceedings of the National Academy

of Sciences 99:13302-13306.

Liu, X., Geng, X., Zhang, H., Shen, H., and Yang, W. (2017). Association and genetic

identification of loci for four fruit traits in tomato using InDel markers. Frontiers in Plant

Science 8:1269.

Liu, Y. S., Gur, A., Ronen, G., Causse, M., Damidaux, R., Buret, M., Hirschberg, J., and Zamir,

D. (2003). There is more to tomato fruit colour than candidate carotenoid genes. Plant

Biotechnology Journal 1:195-207.

Louws, F. J., Fulbright, D. W., Stephens, C. T., and de Bruijn, F. J. (1995). Differentiation of

genomic structure by rep-PCR fingerprinting to rapidly classify Xanthomonas campestris

pv. vesicatoria. Phytopathology 85:528-536.

Louws, F. J., Wilson, M., Campbell, H. L., Cuppels, D. A., Jones, J. B., Shoemaker, P. B., Sahin,

F., and Miller, S.A. (2001). Field control of bacterial spot and bacterial speck of tomato

using a plant activator. Plant Disease 85:481-488.

Ma, X., Ivey, M. L., and Miller, S. (2011). First report of Xanthomonas gardneri causing

bacterial spot of tomato in Ohio and Michigan. Plant Disease 95:1584-1584.

Marco, G. M., and Stall, R. E. (1983). Control of bacterial spot of pepper initiated by strains of

Xanthomonas campestris pv. vesicatoria that differ in sensitivity to copper. Plant Disease

67:779-781.

Mbega, E. R., Mabagala, R. B., Adriko, J., Lund, O. S., Wulff, E. G., and Mortensen, C. N.

(2012). Five species of xanthomonads associated with bacterial leaf spot symptoms in

tomato from Tanzania. Plant Disease 96:760-761.

McCarter, S. M., Jones, J. B., Gitaitis, R. D., and Smitley, D. R. (1983). Survival of

Pseudomonas syringae pv. tomato in association with tomato seed, soil, host tissue, and

epiphytic weed hosts in Georgia. Phytopathology 73:1393-98.

Merk, H. L., Ashrafi, H., and Foolad, M. R. (2012). Selective genotyping to identify late blight

resistance genes in an accession of the tomato wild species Solanum pimpinellifolium.

Euphytica. 187:63–75.

Miller, S. A., and Mera, J. (2011). Evaluation of fungicides and bactericides for the control of

foliar and fruit diseases of processing tomatoes. Plant Disease Management Reports

6:V061.

Minsavage, G. V., Canteros, B. I., and Stall, R. E. (1990). Plasmid-mediated resistance to

streptomycin in Xanthomonas capestris pv. vesicatoria. Phytopathology 80:719-723.

Minsavage, G., Balogh, B., Stall, R., and Jones, J. B. (2003). New tomato races of Xanthomonas

campestris pv. vesicatoria associated with mutagenesis of tomato race 3 strains.

Phytopathology 93: S62. (Abstr.).

Mixon, J. T. (2012). Prevalence of copper resistance among foliar bacterial pathogens of tomato

in Tennessee. Master's Thesis, University of Tennessee.

Monforte, A. J., Diaz, A., Caño-Delgado, A., and Van Der Knaap, E. (2014). The genetic basis

of fruit morphology in horticultural crops: lessons from tomato and melon. Journal of

Experimental Botany 65:4625-4637.

Moss, W. P., Byrne, J. M., Campbell, H. L., Ji, P., Bonas, U., Jones, J. B., and Wilson, M.

(2007). Biological control of bacterial spot of tomato using hrp mutants of Xanthomonas

campestris pv. vesicatoria. Biological Control 41:199-206.

Mu, Q., Huang, Z., Chakrabarti, M., Illa-Berenguer, E., Liu, X., Wang, Y., Ramos, A., and van

der Knaap, E. (2017). Fruit weight is controlled by cell size regulator encoding a novel

protein that is expressed in maturing tomato fruits. PLoS genetics 13:p.e1006930.

Muños, S., Ranc, N., Botton, E., Bérard, A., Rolland, S., Duffé, P., Carretero, Y., Le Paslier,

M.C., Delalande, C., Bouzayen, M., and Brunel, D. (2011). Increase in tomato locule

number is controlled by two SNPs located near WUSCHEL. Plant Physiology 156:2244-

2254.

Myung, I. S., Moon, S. Y., Jeong, I. H., Lee, Y. K., Lee, Y. H., and Ra, D. S. (2009). Bacterial

spot of tomato caused by Xanthomonas perforans, a new disease in Korea. Plant Disease

93:1349-1349.

Obradovic, A., Jones, J. B., Momol, M. T., Balogh, B., and Olson, S. M. (2004). Management of

tomato bacterial spot in the field by foliar applications of bacteriophages and SAR

inducers. Plant Disease 88:736-740.

Obradovic, A., Jones, J. B., Momol, M. T., Olson, S. M., Jackson, L. E., Balogh, B., Guven, K.,

and Iriarte, F.B. (2005). Integration of biological control agents and systemic acquired

resistance inducers against bacterial spot on tomato. Plant Disease 89:712-716.

Obradovic, A., Jones, J., Balogh, B., and Momol, M. (2008). Integrated management of tomato

bacterial spot. In integrated management of diseases caused by fungi, phytoplasma and

bacteria, Springer, Dordrecht 211-223.

Örmälä, A. M., and Jalasvuori, M. (2013). Phage therapy: should bacterial resistance to phages

be a concern, even in the long run? Bacteriophage 3:e24219.

Osdaghi, E., Taghavi, S. M., Hamzehzarghani, H., Fazliarab, A., and Lamichhane, J. R. (2017).

Monitoring the occurrence of tomato bacterial spot and range of the causal agent

Xanthomonas perforans in Iran. Plant Pathology 66:990-1002.

Paran, I., and van der Knaap, E. (2007). Genetic and molecular regulation of fruit and plant

domestication traits in tomato and pepper. Journal of Experimental Botany 58:3841-3852.

Paret, M. L., Vallad, G. E., Averett, D. R., Jones, J. B., and Olson, S. M. (2013). Photocatalysis:

effect of light-activated nanoscale formulations of TiO2 on Xanthomonas perforans and

control of bacterial spot of tomato. Phytopathology 103:228-236.

Pei, C., Wang, H., Zhang, J., Wang, Y., Francis, D. M., and Yang, W. (2012). Fine mapping and

analysis of a candidate gene in tomato accession PI128216 conferring hypersensitive

resistance to bacterial spot race T3. Theoretical and Applied Genetics 124:533-542.

Piombino, P., Sinesio, F., Moneta, E., Cammareri, M., Genovese, A., Lisanti, M. T., Mogno, M.

R., Peparaio, M., Termolino, P., Moio, L., and Grandillo, S. (2013). Investigating

physicochemical, volatile and sensory parameters playing a positive or a negative role on

tomato liking. Food Research International 50:409-19.

Piquerez, S. J., Harvey, S. E., Beynon, J. L., and Ntoukakis, V. (2014). Improving crop disease

resistance: lessons from research on Arabidopsis and tomato. Frontiers in Plant Science

5:671

Pohronezny, K., and Volin, R. B. (1983). The effect of bacterial spot on yield and quality of

fresh market tomatoes [Xanthomonas campestris]. HortScience 18:69-70.

Potnis, N., Minsavage, G., Smith, J. K., Hurlbert, J. C., Norman, D., Rodrigues, R., Stall, R. E.,

and Jones, J. B. (2012). Avirulence proteins AvrBs7 from Xanthomonas gardneri and

AvrBs1. 1 from Xanthomonas euvesicatoria contribute to a novel gene-for-gene

interaction in pepper. Molecular Plant-Microbe Interactions 25:307-320.

Potnis, N., Timilsina, S., Strayer, A., Shantharaj, D., Barak, J. D., Paret, M. L., Vallad, G. E., and

Jones, J. B. (2015). Bacterial spot of tomato and pepper: diverse Xanthomonas species

with a wide variety of virulence factors posing a worldwide challenge. Molecular Plant

Pathology 16:907-920.

Punina, N.V., Ignatov, A. N., Pekhtereva, E. S., Kornev, K. P., Matveeva, E. V., Polityko, V. A.,

Budenkov, N. I., and Schaad, N. W. (2009). Occurrence of Xanthomonas campestris pv.

raphani on tomato plants in the Russian Federation. In H. Saygili, F. Sahin, and Y. Aysan

(Eds.), II International symposium on tomato diseases 808:287-290.

Quezado-Duval, A. M., Leite Jr, R. P., Truffi, D., and Camargo, L. E. (2004). Outbreaks of

bacterial spot caused by Xanthomonas gardneri on processing tomato in central-west

brazil. Plant Disease 88:157-161.

Rashid, T. S., Kamaruzaman, S., Golkhandan, E., Nasehi, A., and Awla, H. K. (2015). First

report of Xanthomonas gardneri causing bacterial spot of tomato in Malaysia. Plant

Disease 100:854.

Richard, D., Boyer, C., Vernière, C., Canteros, B. I., Lefeuvre, P., and Pruvost, O. (2017).

Complete genome sequences of six copper-resistant Xanthomonas strains causing

bacterial spot of solaneous plants, belonging to X. gaardneri, X. euvesicatoria, and X.

vesicatoria, using long-read technology. Genome Announcements 5:e01693-16.

Roach, R., Mann, R., Gambley, C. G., Shivas, R. G., and Rodoni, B. (2018). Identification of

Xanthomonas species associated with bacterial leaf spot of tomato, capsicum and chilli

crops in eastern Australia. European Journal of Plant Pathology 150:595-608.

Robbins, M. D., Darrigues, A., Sim, S. C., Masud, M. A. T., and Francis, D. M. (2009).

Characterization of hypersensitive resistance to bacterial spot race T3 (Xanthomonas

perforans) from tomato accession PI 128216. Phytopathology 99:1037-1044.

Roberts, P. D., Momol, M. T., Ritchie, L., Olson, S. M., Jones, J. B., and Balogh, B. (2008).

Evaluation of spray programs containing famoxadone plus cymoxanil, acibenzolar-S-

methyl, and Bacillus subtilis compared to copper sprays for management of bacterial spot

on tomato. Crop Protection 27:1519-1526.

Rodríguez, G. R., Kim, H. J., and van der Knaap, E. (2013). Mapping of two suppressors of

OVATE (sov) loci in tomato. Heredity 111:256.

Rodríguez, G. R., Muños, S., Anderson, C., Sim, S. C., Michel, A., Causse, M., Gardener, B. M.,

Francis, D., and van der Knaap, E. (2011). Distribution of SUN, OVATE, LC and FAS in

the tomato germplasm and the relationship to fruit shape diversity. Plant Physiology 110.

Ronen, G., Carmel-Goren, L., Zamir, D., and Hirschberg, J. (2000). An alternative pathway to

beta-carotene formation in plant chromoplasts discovered by map-based cloning of beta

and old-gold color mutations in tomato. Proceedings of the National Academy of

Sciences of the United States of America 97:11102-7

Ronen, G., Cohen, M., Zamir, D., and Hirschberg, J. (1999). Regulation of carotenoid

biosynthesis during tomato fruit development: expression of the gene for lycopene

epsilon-cyclase is downregulated during ripening and is elevated in the mutant Delta. The

Plant Journal 17:341–351

Ryan, R. P., Vorhölter, F. J., Potnis, N., Jones, J. B., Van Sluys, M. A., Bogdanove, A. J., and

Dow, J. M. (2011). Pathogenomics of Xanthomonas: understanding bacterium–plant

interactions. Nature Reviews Microbiology 9:344-355.

Saliba-Colombani, V., Causse, M., Langlois, D., Philouze, J., and Buret, M. (2001). Genetic

analysis of organoleptic quality in fresh market tomato. 1. Mapping QTLs for physical

and chemical traits. Theoretical and Applied Genetics 102:259-272.

Schornack, S., Minsavage, G. V., Stall, R. E., Jones, J. B., and Lahaye, T. (2008).

Characterization of AvrHah1, a novel AvrBs3‐like effector from Xanthomonas gardneri

with virulence and avirulence activity. New Phytologist 179:546-556.

Scott, J. W., Hutton, S. F., Jones, J. B., Francis, D. M., and Miller, S.A. (2006). Resistance to

bacterial spot race T4 and breeding for durable, broad-spectrum resistance to other races.

Rpt Tomato Genet Coop 56:33-36.

Scott, J. W, Miller, S. A., Stall, R. E., Jones, J. B., and Somodi, G. C., Barbosa, V., Francis, D.

L., and Sahin, F. (1997). Resistance to race T2 of the bacterial spot pathogen in tomato.

HortScience 32:724-727.

Scott, J. W., and Jones, J. (1986). Sources of resistance to bacterial spot in tomato. HortScience

21:304-306.

Scott, J. W., and Jones, J. (1989). Inheritance of resistance to foliar bacterial spot of tomato

incited by Xanthomonas campestris pv. vesicatoria. Journal of the American Society for

Horticultural Science 114:111-114

Scott, J. W., Francis, D. M., Miller, S. A., Somodi, G. C., and Jones, J. B. (2003). Tomato

bacterial spot resistance derived from PI 114490; inheritance of resistance to race T2 and

relationship across three pathogen races. Journal of the American Society for

Horticultural Science 128:698-703.

Scott, J. W., Hutton, S. F., Shekasteband, R., Sim, S. C., and Francis, D. M. (2015).

Identification of tomato bacterial spot race T1, T2, T3, T4, and Xanthomonas gardneri

resistance QTLs derived from PI 114490 populations selected for race T4. Acta

Horticulturae 1069:53-58.

Scott, J. W., Jones, J., and Somodi, G. (2001). Inheritance of resistance in tomato to race T3 of

the bacterial spot pathogen. Journal of the American Society for Horticultural Science

126:436-441.

Scott, J. W., Stall, R. E., Jones, J. B., and Somodi, G. C. (1996). A single gene controls the

hypersensitive response of Hawaii7981 to race 3 (T3) of the bacterial spot pathogen.

Tomato Genetics Cooperative 46:23.

Scott, J., Jones, J., Somodi, G., and Stall, R. (1995). Screening tomato accessions for resistance

to Xanthomonas campestris pv. vesicatoria, race T3. Hortscience 30:579-581.

Sharlach, M., Dahlbeck, D., Liu, L., Chiu, J., Jiménez-Gómez, J. M., Kimura, S., Koenig, D.,

Maloof, J. N., Sinha, N., Minsavage, G. V., and Jones, J. B. (2013). Fine genetic mapping

of RXopJ4, a bacterial spot disease resistance locus from Solanum pennellii LA716.

Theoretical and Applied Genetics 126:601-609.

Shenge, K., Mabagala, R. B., Mortensen, C. N., and Wydra, K. (2014). Resistance of

Xanthomonas campestris pv. vesicatoria isolates from Tanzania to copper and

implications for bacterial spot management. African Journal of Microbiology Research

8:2881-2885.

Sherf, A. F., and MacNab, A. A. (1986). Vegetable diseases and their control. John Wiley and

Sons. New York, United States 728:1986.

Sim, S. C., Robbins, M. D., Wijeratne, S., Wang, H., Yang, W., and Francis, D. M. (2015).

Association analysis for bacterial spot resistance in a directionally selected complex

breeding population of tomato. Phytopathology 105:1437-1445.

Stall, R. E., Beaulieu, C., Egel, D., Hodge, N. C., Leite, R. P., Minsavage, G. V., Bouzar, H.,

Jones, J. B., Alvarez, A. M., and Benedict, A. A. (1994). Two genetically diverse groups

of strains are included in Xanthomonas campestris pv. vesicatoria. International Journal

of Systematic and Evolutionary Microbiology 44:47-53.

Stall, R. E., Jones, J. B., and Minsavage, G. V. (2009). Durability of resistance in tomato and

pepper to xanthomonads causing bacterial spot. Annual Review of Phytopathology

47:265-284.

Stall, R. E., Loschke, D. C., and Jones, J. B. (1986). Linkage of copper resistance and avirulence

loci on a self-transmissible plasmid in Xanthomonas campestris pv. vesicatoria.

Phytopathology 76:240-243.

Strayer, A., Ocsoy, I., Tan, W., Jones, J., and Paret, M. (2016). Low concentrations of a silver-

based nanocomposite to manage bacterial spot of tomato in the greenhouse. Plant Disease

100:1460-1465.

Stukenbrock, E. H., and McDonald, B. A. (2009). Population genetics of fungal and oomycete

effectors involved in gene-for-gene interactions. Molecular Plant-Microbe Interactions

22:371-380.

Sun, H., Liu, X., Li, W., and Yang, W. (2011a). Preliminary mapping of a gene in tomato

accession LA1589 conferring resistance to race T3 of bacterial spot [J]. Journal of

Agricultural University of Hebei 6:012.

Sun, H., Zhang, J. Y., Wang, Y. Y., Scott, J. W., Francis, D. M., and Yang, W. (2011b). QTL

analysis of resistance to bacterial spot race T3 in tomato. Acta Horticulturae Sinica

38:2297-2308

Šutic, D. (1957). Tomato bacteriosis. Posebna Izd. Inst. Zasht. Bilja, Beograd [Spec. Edit. Inst.

Plant Prot., Beograd] 6:65.

Tai, T. H., Dahlbeck, D., Clark, E. T., Gajiwala, P., Pasion, R., Whalen, M. C., Stall, R. E., and

Staskawicz, B. J. (1999). Expression of the Bs2 pepper gene confers resistance to

bacterial spot disease in tomato. Proceedings of the National Academy of Sciences

96:14153-14158.

Tanksley, S.D. (2004). The genetic, developmental, and molecular bases of fruit size and shape

variation in tomato. The Plant Cell 16:S181-S189.

Thayer, P. L., and Stall, R. E. (1961). A survey of Xanthomonas vesicatoria resistance to

streptomycin. Proceedings of the Florida State Horticultural Society 75:163-165.

Timilsina, S., Jibrin, M. O., Potnis, N., Minsavage, G. V., Kebede, M., Schwartz, A., Bart, R.,

Staskawicz, B., Boyer, C., Vallad, G. E., and Pruvost, O. (2015). Multilocus sequence

analysis of xanthomonads causing bacterial spot of tomato and pepper plants reveals

strains generated by recombination among species and recent global spread of

Xanthomonas gardneri. Applied and Environmental Microbiology 81:1520-1529.

Tudor-Nelson, S., Minsavage, G., Stall, R., and Jones, J. (2003). Bacteriocin-like substances

from tomato race 3 strains of Xanthomonas campestris pv. Vesicatoria. Phytopathology

93:1415-1421.

Vallad, G. E., and Goodman, R. M. (2004). Systemic acquired resistance and induced systemic

resistance in conventional agriculture. Crop Science 44:1920-1934.

Vallad, G. E., Pernezny, K. L., Balogh, B., Wen, A., Figueiredo, J. F. L., Jones, J. B., Momol, T.,

Muchovej, R. M., Havranek, N., Abdallah, N., and Olson, S. (2010). Comparison of

kasugamycin to traditional bactericides for the management of bacterial spot on tomato.

HortScience 45:1834-1840.

Vallad, G. E., Timilsina, S., Adkison, H., Potnis, N., Minsavage, G., Jones, J., and Goss, E.

(2013). A recent survey of xanthomonads causing bacterial spot of tomato in Florida

provides insights into management strategies. TomaTo Proceedings 25.

Vallejos, C. E., Jones, V., Stall, R. E., Jones, J. B., Minsavage, G. V., Schultz, D. C., Rodrigues,

R., Olsen, L. E., and Mazourek, M. (2010). Characterization of two recessive genes

controlling resistance to all races of bacterial spot in peppers. Theoretical and Applied

Genetics 121:37-46. van Der Knaap, E., and Tanksley, S. D. (2001). Identification and characterization of a novel

locus controlling early fruit development in tomato. Theoretical and Applied Genetics

103:353–358. doi: 10.1007/s001220100623 van der Knaap, E., Chakrabarti, M., Chu, Y., Clevenger, J. P., Illa-Berenguer, E., Huang, Z.,

Keyhaninejad, N., Mu, Q., Sun, L., Wang, Y., and Wu, S. (2014). What lies beyond the

eye: the molecular mechanisms regulating tomato fruit weight and shape. Frontiers in

Plant Science 5:227. doi:10.3389/fpls.2014.00227

Vauterin, L., Hoste, B., Kersters, K., and Swings, J. (1995). Reclassification of Xanthomonas.

International Journal of Systematic Bacteriology 45:472–489.

Visa, S., Cao, C., Gardener, B.M., and van der Knaap, E. (2014). Modeling of tomato fruits into

nine shape categories using elliptic fourier shape modeling and Bayesian classification of

contour morphometric data. Euphytica 200:429-439.

Voloudakis, A. E., Bender, C. L., and Cooksey, D. A. (1993). Similarity between copper

resistance genes from Xanthomonas campestris and Pseudomonas syringae. Applied and

Environmental Microbiology 59:1627-1634.

Walters, D. R., and Fountaine, J. M. (2009). Practical application of induced resistance to plant

diseases: an appraisal of effectiveness under field conditions. Journal of Agricultural

Science 147:523-535.

Walters, D., Walsh, D., Newton, A., and Lyon, G. (2005). Induced resistance for plant disease

control: maximizing the efficacy of resistance elicitors. Phytopathology 95:1368-1373.

Wang, H., Hutton, S. F., Robbins, M. D., Sim, S. C., Scott, J. W., Yang, W., Jones, J. B., and

Francis, D. M. (2011). Molecular mapping of hypersensitive resistance from tomato

‘Hawaii 7981’ to Xanthomonas perforans race T3. Phytopathology 101:1217-1223.

Wang, J. F., Jones, J. B., Scott, J. W., and Stall, R. E. (1990). A new race of the tomato group of

strains of Xanthomonas campestris pv. vesicatoria. Phytopathology 80:1070.

Wang, J., Stall, R., and Vallejos, C. (1994). Genetic analysis of a complex hypersensitive

reaction to bacterial spot in tomato. Phytopathology 84:126-132.

Wang, L., Li, J., Zhao, J., and He, C. (2015). Evolutionary developmental genetics of fruit

morphological variation within the Solanaceae. Frontiers in Plant Science 6:248.

Weller, J. I., Soller, M., and Brody, T. (1988). Linkage analysis of quantitative traits in an

interspecific cross of tomato (Lycopersicon esculentum x Lycopersicon pimpinellifolium)

by means of genetic markers. Genetics 118:329-339.

Whalen, M. C., Wang, J. F., Carland, F. M., Heiskell, M. E., Dahlbeck, D., Minsavage, G.V.,

Jones, J. B., Scott, J. W., Stall, R. E., and Staskawicz, B. J. (1993). Avirulence gene

avrRxv from Xanthomonas campestris pv. vesicatoria specifies resistance on tomato line

HI 7998. Molecular Plant-Microbe Interactions 6:616-627.

Xiao, H., Jiang, N., Schaffner, E., Stockinger, E. J., and van der Knaap, E. (2008). A

retrotransposon-mediated gene duplication underlies morphological variation of tomato

fruit. Science 319:1527-1530.

Yang, W., Miller, S. A., Scott, J. W., Jones, J. B., and Francis, D.M. (2005a). Mining tomato

genome sequence databases for molecular markers: application to bacterial resistance and

marker assisted selection. Acta Horticulturae 695:241-250.

Yang, W., Sacks, E. J., Lewis Ivey, M. L., Miller, S. A., and Francis, D. M. (2005b). Resistance

in Lycopersicon esculentum intraspecific crosses to race T1 strains of Xanthomonas

campestris pv. vesicatoria causing bacterial spot of tomato. Phytopathology 95:519-527.

Yu, Z., Wang, J., Stall, R., and Vallejos, C. (1995). Genomic localization of tomato genes that

control a hypersensitive reaction to Xanthomonas campestris pv. vesicatoria (Doidge)

dye. Genetics 141:675-682.

Zhang, N., Brewer, M. T., and van der Knaap, E. (2012). Fine mapping of fw3. 2 controlling

fruit weight in tomato. Theoretical and Applied Genetics 125:273-284.

CHAPTER 2: PHENOTYPIC AND GENETIC DIVERSITY OF XANTHOMONAS

POPULATIONS CAUSING BACTERIAL SPOT OF TOMATO IN NORTH CAROLINA

ABSTRACT

Bacterial spot caused by Xanthomonas species (Xs) is the serious disease of tomato in

North Carolina (NC). In total, 293 Xs strains from NC, representing spatial and tempooral diversity, were analyzed for phenotypic and genetic diversity. Copper and streptomycin sensitivity assays revealed that over 95% of the Xanthomonas strains were copper resistant while

25% and 45% were streptomycin resistant in 2016 and 2015 respectively. A BOX repetitive element (BOX) polymerase chain reaction (BOX-PCR) assay detected four haplotypes (H1, H2,

H3, and H4) among Xs strains. The study on a subset of representative Xs strains (n = 45) using the multiplex quantitative real-time polymerase chain reaction (qPCR) targeting highly conserved hrpB7 gene identified Xs strains in NC as a single species X. perfornas (Xp). The hypersensitive reaction (HR) of representative strains (n = 45) of Xp on tomato differential cultivars confirmed ~91% and 9% of Xp strains were races T4 and T3, respectively.

Additionally, phylogenetic and comparative sequence analysis of six genomic regions (fusA, gapA, gltA, gyrB, lacF, and lepA) grouped 59% of Xp strains of NC with the Xp race T4 reference strain GEV839 from Florida. Similarly, 25% of Xp strains from North Carolina were similar with Xp race T4 reference strain 17-12; 13% with Xp race T3 reference strain 91-118 and

ICMP166690-767 and the remaining 3% as a separate group, suggesting that the majority of Xp strains in North Carolina have a monophyletic relationship with the recent strains of Xp in

Florida. Overall, our results suggested that the bacterial spot management practices in tomato in

NC should be implemented with significant focus on introducing host resistance against race T4

and considering the challenges currently posed by the widespread existence of copper resistant bacterial spot pathogens.

INTRODUCTION

Bacterial spot (Xanthomonas spp) is an economically important disease of fresh market tomato (Solananum lycopersicum L.) in North Carolina (NC) and several other states and affect all-above ground parts of the plant. Fruit lesions can either make the fruit unmarketable or can cause severe defoliation and poor-quality fruits (Burlakoti et al., 2018). At least four species within the genus Xanthomonas- X. vesicatoria, X. euvesicatoria, X. perforans, and X. gardneri have been reported as causal agents of bacterial spot (Jones et al. 2004). In addition, four races

(T1, T2, T3, and T4) associated with differential hosts have been reported (Jones et al. 2004). X. euvesicatoria, X. vesicatoria, and X. perforans are favored by warm temperatures (25-30oC), while X. gardneri strains prefer cooler temperature of 20 °C (Araujo et al., 2010). The pathogen is introduced in the tomato fields through contaminated seed or transplants, on overwintered host debris or through epiphytic multiplication on the phyllosphere of the crop plant (Burlakoti et al.

2018). About 1 in 100 commercial seeds obtained from diseased fields have potential to give rise to diseased seedlings the next cropping cycle (Gardner and Kendrick 1921). Therefore, the pathogens are more likely to be disseminated over a long distance and introduced into new fields and geographical regions through such infected-seeds and transplants (Gardner and Kendrick

1921). Xanthomonas spp. Can survive epiphytically on volunteer plants and weeds (Jones et al.

1986). Bacterial spot xanthomonads causing symptoms on tomato plants were also reported from sprouting weeds within tomato fields in Brazil (Araújo et al. 2015). The secondary spread of the pathogen within field and greenhouse occur through rain-splash, overhead irrigation, sprinkler irrigation, and wind-blown rain (Ryan et al. 2011).

Practical management of bacterial spot is challenging in commercial production fields due to limited efficacy of current disease management strategies and lack of commercial resistant cultivars available against bacterial spot disease. Bacterial disease management is based on

integrated use of cultural practices and application of chemicals such as copper, antibiotics

(streptomycin) and plant activators (Louws et al. 2001; Louws 2018; Potnis et al. 2015).

However, copper and streptomycin resistant Xanthomonas isolates have already been widely reported in USA (Florida, Tennessee, Ohio), Canada, Brazil, Ethiopia, and Tanzania (Araújo et al. 2012; Miller and Mera 2011; Mixon 2012; Vallad et al. 2013; Shenge et al. 2014; Abbasi et al. 2015). Copper-based bactericides and streptomycin have also been routinely used as a standard treatment to manage foliar bacterial diseases of tomato in NC. As a result, tomato growers from NC have reported failure of these compounds for the management of bacterial disease in their fields. However, failure is associated with the bacterial resistance to copper or streptomycin has not been investigated in NC.

The evolution of new Xanthomonas species and races over time has hindered the development and implementation of the host resistance for the management of the bacterial spot disease. For example, X. euvesicatoria was the only species, present as race 1, in Florida until

1991 before the X. perforans race T3 strain was reported (Timilsina et al. 2015). X. perforans race T4 strain emerged in 1998 (Astua-Monge et al. 2000; Minsavage et al. 2003), and thereafter it has been detected in higher number in Florida. In a recent survey, only X. perforans race 4 were detected from different parts of Florida, indicating a major shift in the race composition within X. perforans strains from previous surveys, where races T4 and T3 were in the ratio of 3:1

(Horvath et al. 2012). Also, bacterial spot pathogens have distinct distribution of species and races across different geographical regions. X. euvesicatoria and X. vesicatoria are distributed throughout the world, whereas X. perforans and X. gardneri are mostly reported from North

America, South America, Africa and Europe (Potnis et al. 2015). One of the first USA reports of

Xg appeared in the Midwest USA on a contaminated seed lot in 1991 and had recently caused the epidemics in the Midwest (Cuppels et al. 2006; Ma et al. 2011). Although X. vesicatoria is

prominent in Midwest production, it was never a problem in Florida and the Southeastern USA

(Louws et al., 1995). Therefore, it is critical to understand the local pathogen species and race structure, and their diversity to develop disease management and breeding strategies in the targeted regions. However, the race profile, species structure, and genetic diversity of

Xanthomonas population in NC have not been investigated so far in a systematic manner.

We investigated phenotypic characteristics and genetic diversity of the bacterial spot pathogen in NC for the first time in a systematic manner. Based on growers’ experience in NC, and population genetics analysis in Southeastern USA (Potnis et al. 2015; Timilsina et al. 2015;

Vallad et al. 2013), we hypothesized that Xanthomonas populations in NC would exhibit resistance to copper and streptomycin. We also expect that X. perforans to be the major bacterial spot causing pathogen in tomato based on our preliminary studies and the prevalence of this species in other states of Southeastern states. The objectives of this study were: i) to evaluate and determine the proportion of tomato field strains of Xanthomonas with resistance to copper and streptomycin; ii) assess the genetic diversity of Xanthomonas population causing bacterial spot disease in tomato in NC using BOX-PCR assay and species-specific hrpB7 gene probes; iii) to identify pathogenic races of Xanthomonas strains from NC, and iv) to determine the phylogenetic relationships and genetic distinctness of Xanthomonas strains in NC compared to other reported strains of Xanthomonas from different regions.

MATERIALS AND METHODS

Sampling Scheme and Sample Collection

Field surveys were conducted across major tomato growing regions of NC during the

2015 and 2016 seasons (Figure 2.1). Ten farms were surveyed in 2015 and thirteen farms were

surveyed in 2016 for bacterial spot. A modified random sampling scheme (5 field sites with four leaves sampled per site) was employed to collect 20 symptomatic leaves systematically from each field for the isolation of Xs strains. Fields represented a diversity of tomato cultivars (Table

2.1). Leaf samples were placed in a plastic sealable bag and labelled with the date of collection, location, and tomato cultivar. The leaf samples were held at 4°C until the bacteria were isolated.

Fresh lesions were chosen to avoid the mixture of other pathogens or saprophytic bacteria and cut (~ 4 mm2) with a sterile scalpel. The infected leaves were surface sterilized with 10% chlorox and rinsed with sterile water three times. Xs strains were isolated on Difco yeast dextrose agar (YDC) medium (Thermo Scientific, Wilmington, DE) and incubated at temperature of 27°C for 24-48 hours. Subsequently, plates were examined for the presence of yellow mucous

Xanthomonas colonies. Only putative Xanthomonas colonies based on colony color and morphology were sub-cultured by streaking to obtain pure cultures. Other non-Xs colonies or fungi were discarded. In all, 293 strains of Xs (n = 163 in 2015, and n = 127 in 2016) were isolated from 23 fields across eight counties of NC, respectively (Table 2.1). All pure cultures originating from single colonies were stored in 30% glycerol at -80°C for further work.

Figure 2.1: Map of North Carolina showing counties (star symbols) from where Xanthomonas species (Xs) strains were collected and analyzed in this study.

Table 2.1: Information on Xanthomonas spp (Xs) isolated from North Carolina.

Number Number of Counties of fields isolates Year Tomato cultivar sampled cultured Plum Regal, BHN784, Mountain Fresh, 2015 and Henderson 5 61 Mountain Majesty, Mountain Merit, 2016 Taste-lee 2015 and Madison 4 43 Resolute, Red Defender, grafted plants 2016 2015 and Mountain Majesty, Roma, Mountain Haywood 6 54 2016 Magic, Mountain Fresh Buncombe 1 20 2015 Biltmore Jackson 1 23 2015 Plum Regal Swain 1 19 2015 Plum Regal Red Mountain, Picus, mixed breeding Rowan 3 42 2016 lines Macon 2 31 2016 Plum Regal, Tasty Lee

Copper and Streptomycin Sensitivity Assays

A total of 290 strains were assessed for copper and streptomycin sensitivity. Sensitivity assay was performed on sucrose peptone agar (SPA) medium amended with different concentrations of copper sulfate (100 ppm, 200 ppm, and 300 ppm) or streptomycin sulfate (20 ppm, 50 ppm, and 100 ppm) using the protocol modified from Marco and Stall (1983). The culture medium was prepared using 15.0 g sucrose, 5.0 g peptone, 0.50 g dibasic potassium phosphate, 0.25 g magnesium sulfate, and 15.0 g Difco agar in 1000 ml of distilled water. The pH was adjusted to 7.2-7.4 before autoclaving. The copper (Cu) ions were obtained from copper sulfate (CuSO4, Sigma-Aldrich Corporation, St Louis, MO; 59% CU a.i)). Streptomycin was obtained from streptomycin sulfate dry (Sigma-Aldrich, St. Louis, MO). Fresh stock solution

(10,000 ppm) of copper sulfate and streptomycin sulfate were prepared, and filter sterilized using

0.22 μm pore size filter under laminar flow. Then, an appropriate amount of stock solution of copper and streptomycin was added in SPA medium after the medium was autoclaved and cooled down to get the desired concentration of copper and streptomycin. Bacteria were cultured

in Luria-Bertani (LB) medium (Thermo Scientific, Wilmington, DE) for 24- 48 hrs and suspended in sterile water to adjust the concentration to 108 colony forming units per milliliter

(CFU/ml) (Marco and Stall 1983). Five microliters of the suspension were pipetted on the SPA medium amended with different concentrations of copper and streptomycin and incubated for 48 h at 27°C. Dr. David Ritchie, Department of Entomology and Plant Pathology, North Carolina

State University, Raleigh, kindly provided strains of Xs known to be sensitive and resistant to streptomycin and included as controls. Controls used for sensitivity assays for Xs strains were:

Xcv 611 (CuR SpS), Xcv 43 (CuR SpR), and Xcv 135 (CuS SpS). The growth of the bacteria on the amended medium was visually observed and recorded. Strains capable of growing on SPA medium containing 200 ppm of CuSO4 or 100 ppm of streptomycin were considered as copper and streptomycin resistant, respectively.

Genomic DNA Extraction and Quantification

Bacterial strains were grown in LB media for 24 to 48 h at 27°C. The total genomic DNA was extracted with guanidium thiocyanate as described by Pitcher et al. (1989) and slightly modified by Luc Vauterin (Pitcher et al. 1989; Rademaker and de Brujin 1997). DNA was quantified using a NanoDrop 2000 Spectrophotometer (Thermo Scientific, Wilmington, DE).

Working DNA concentration of 10 ng/µl was prepared for genetic diversity analysis.

BOX Polymerase Chain Reaction (BOX-PCR) Assay

We have chosen BOX-PCR assay for the genetic diversity analysis because this assay was robust and yielded more amplifications compared to enterobacterial repetitive intergenic consensus (ERIC) PCR (ERIC-PCR) and repetitive extragenic palindromic (REP) PCR (REP-

PCR) assays (Louws et al. 1995). Importantly, BOX-PCR assay has been successfully utilized

for the characterization of bacterial pathogens infecting tomatoes (Araújo et al. 2017; Louws et al. 1995). A subset of 278 strains of Xathomonas were characterized by BOX-PCR in this study.

BOX-PCR (BOXA1R: 5’-CTACGGCAAGGCGACGCTGACG-3’) amplification was performed in a 25 µl reaction mixture as described (Louws et al. 1995; Louws and Cuppels

2001) previously with some modifications. One initial cycle of denaturation at 95°C for 5 min was carried out, followed by 40 cycles of denaturation at 95°C for 30 s; annealing at 48°C for 1 min; extension at 68°C for 10 min; and a final cycle of extension at 68°C for 10 min. The PCR- products (4-μl) were analyzed by running gel electrophoresis on 1.5% agarose (Genesee

Scientific Corporation, Research Triangle Park, NC) gel containing ethidium bromide (10 mg/ml, Sigma-Aldrich) in 1× Tris-acetate-EDTA (TAE, Thermo Scientific, Wilmington, DE) at

35V for 10 to 12 h. The gel image was visualized using a gel documentation system (BioRad Gel

Doc, Hercules, CA).

Fingerprint profiles were scored visually: ‘0’ for the absence of the band and ‘1’ for the presence of the band to generate a binary matrix. The unique haplotypes were determined based on the fingerprint profiles and used to generate the cluster dendrogram. A hierarchical agglomerative cluster analysis was conducted to generate hierarchical clusters from 71uclidean distance matrices using a hierarchy module within the SciPy cluster package in python (Jones et al. 2014). The linkage method with group average (unweighted-pair-group) technique was used to define the hierarchical clustering by comparing data points, and the cophenetic correlation coefficient was calculated to determine how well the cluster dendrogram preserved the original distances between data points (Saraçli et al. 2013).

Multiplex Quantitative Real-Time PCR (qPCR) Assay

A multiplex real-time TaqMan PCR assay was conducted on a subset of 45 representative

Xanthomonas strains based on phenotypic, genomic fingerprint, and temporal/spatial diversity.

The assay differentiates the four species of Xanthomonas (Table 2.2). Genomic DNA was extracted by boiling the bacterial suspensions as mentioned by Strayer et al. (2016). Four species-specific probes and two primer sets targeting the highly conserved hrpB7 gene regions were used. The multiplex master mix was prepared, and the bacterial DNA was amplified using the qPCR protocol. Reference strains used for representing four different species associated with the bacterial spot of tomato were included: Xv 144, Xe 157, Xp 1484, and Xg GA2 (Strayer et al.

2016).

Virulence Assay and Race Identification

A subset of representative strains (n=45) of Xanthomonas was evaluated for a hypersensitive response (HR) on tomato differentials (Table 2.2). HR test was conducted on five- weeks-old tomato and pepper plants to identify pathogenic races of Xs strains in NC. The pepper differential tomato genotypes used for HR assay of Xanthomonas strains were: Bonny Best

(susceptible to all races), Hawaii 7998 (effective against avrRxv; race T1), and FL 216 (effective against avrXv3; race T3); and pepper differentials used were: Early Calwonder (ECW) 10R

(effective against avrBs1), ECW 20R (effective against avrBs2), ECW 30R (effective against avrBs3), and ECW (susceptible to all races). The resistant genotype LA716 carrying RxopJ4 gene against XopJ4 effector of race T4 could not be used as a tomato differential, because of the extreme necrosis observed in this genotype. Bacterial suspensions were prepared by suspending

24 h cultures grown on nutrient agar in sterile tap water, and the concentration was adjusted to

8 the absorbance at 600 nm (A600 = 0.3 OD, approximately 5 × 10 CFU/ml). The reference strain

X. perforans 91-118 (isolated in 1991) and X. euvesicatoria E3 (isolated in 1961) were used as controls for race T3 and race T1, respectively, and were obtained from Dr. Jeff Jones,

Department of Plant Pathology, University of Florida, Gainesville, FL. E3 also contains avrBs1 and avrBs2 effectors. Bacterial suspensions were then infiltrated into the abaxial leaf surfaces with the aid of a hypodermic syringe (1 ml volume capacity) on five-week-old tomato and pepper seedlings. The disease reactions on different genotypes were evaluated for HR after 24 h and the susceptible reaction up to 2 days after infiltration. HR was characterized by water- soaking areas followed by confluent tissue necrosis occurring up to 24 h after infiltration.

Multi-locus Sequence Analysis (MLSA)

Forty-two representative Xanthomonas strains out of 45 were subjected to MLSA using the six housekeeping genes fusA (elongation factor G), gapA (glyceraldehyde-3-phosphate dehydrogenase A), gltA (citrate synthase), gyrB (gyrase B), lacF (ABC transporter sugar permease), and lepA (GTP binding protein) as described previously (Timilsina et al. 2015).

These genes were amplified using six primer pairs designed for the multilocus sequence typing of Xanthomonas species (Almeida et al. 2010). The amplification was performed in a 50 µl reaction mixture containing PCR reagents of 2.5 U of Phusion high fidelity polymerase (Biorad),

1× Taq buffer, 0.2 mM dNTP, 0.4 µm of each forward and reverse primers, and water in a Peltier element-based thermocycler using two-step cycling. One initial cycle of denaturation at 98°C for

2 min was carried out, followed by 35 cycles of denaturation at 98°C for 20 s; and extension at

72°C for 1 min; and a final cycle of extension at 72°C for 10 min. The PCR-product (4 μl) was analyzed by running gel electrophoresis on 1.0 % agarose gel. The gel image was visualized using a gel documentation system (BioRad). The PCR products with positive bands on gel images were submitted for Sanger platinum sequencing in Genomic Sciences Laboratory (GSL),

North Carolina State University, Raleigh, NC. These housekeeping genes were phylogenetically compared with reference Xanthomonas strains publicly available in sequence databases of Plant-

Associated and Environmental Microbes Database (PAMDB; www.pamdb.org) or National

Center of Biotechnology Information (NCBI; www.ncbi.nlm.nih.gov).

Sequences of Xanthomonas strains from NC and sequences of Xanthomonas reference strains were aligned using MUSCLE within Molecular Evolutionary Genetics Analysis

(MEGA) software version 7.0 using default parameters (Kumar et al. 2016). The nucleotide sequences for each gene were trimmed at both ends such that the final sizes of genes were: 591 positions for fusA, 444 positions for gapA, 501 positions for gltA, 411 positions for gyrB, 408 positions for lacF, and 390 positions for lepA. Individual gene sequences were concatenated in the alphabetical order of the six genes to give the total length of 2745 positions, i.e., 1-591 for fusA, 592-1035 for gapA, 1036-1536 for gltA, 1537-1947 for gyrB, 1948-2355 for lacF, and

2356-2745 for lepA. All sequences were deposited to GenBank (Accession Nos.

MH887544 to MH887747). Phylogenetic analysis based on concatenated gene sequences and individual gene sequences were conducted using maximum likelihood method within MEGA software version 7.0 (Kumar et al. 2016). The general time reversible model with gamma- distributed invariant sites was used to construct a maximum likelihood tree with 1000 bootstrap replicates. Haplotypes for each gene were determined based on the alignment files and the single nucleotide polymorphisms (SNPs) variation among haplotypes were detected. The phylogenetic relationship of the strains along with other metadata (geographic location, BOX-PCR profile, race profile, copper and streptomycin sensitivity) were examined using T-BAS v2 (Carbone et al.

2016).

Table 2.2: Summary of representative strains of Xanthomonas spp (Xs) from North Carolina used in the race identification and multiplex quantitative real-time polymerase chain reaction (qPCR) assay.

Strains County Year of Tomato Copper Streptomycin Box- collection cultivarw Resistantx Resistanty PCR assayz NC-10 Henderson 2015 Plum Regal + + H2 NC-11 Henderson 2015 Plum Regal + - H2 NC-12 Henderson 2015 Plum Regal + + H2 NC-14 Henderson 2015 Plum Regal + - H2 NC-22 Henderson 2015 Plum Regal + + H1 NC-28 Madison 2015 Red Defender + - H1 NC-37 Madison 2015 Red Defender + + H1 NC-42 Madison 2015 Red Defender + + H1 NC-47 Haywood 2015 Mountain + - H1 Majesty NC-67 Buncombe 2015 Biltmore + - H1 NC-79 Buncombe 2015 Biltmore + - H1 NC-80 Buncombe 2015 Biltmore + - H1 NC-87 Henderson 2015 BHN 786 + - H1 NC-101 Henderson 2015 BHN 786 + - H2 NC-110 Madison 2015 Grafted plants + - H1 NC-112 Madison 2015 Grafted plants + + H3 NC-124 Madison 2015 Grafted plants + + H1 NC-126 Jackson 2015 Plum Regal + - H1 NC-145 Jackson 2015 Plum Regal + - H1 NC-157 Haywood 2015 Roma + + H1 NC-180 Swain 2015 Plum Regal + - H1 NC-189 Swain 2015 Plum Regal + - H1 NC-193 Haywood 2015 Roma + - H1 NC-204 Haywood 2015 Roma + + H1 NC-232 Henderson 2016 Mountain + + H1 Majesty NC-242 Henderson 2016 Mountain + - H1 Merit NC-248 Henderson 2016 Plum Regal + - H1

Table 2.2 (continued).

NC-252 Macon 2016 Plum Regal + + H1 NC-264 Macon 2016 Plum Regal + - H1 NC-280 Macon 2016 Tasty Lee - - H2 NC-282 Macon 2016 Tasty Lee - - H4 NC-283 Macon 2016 Tasty Lee + - H2 NC-289 Madison 2016 Resolute + + H1 NC-293 Madison 2016 Resolute + - H1 NC-304 Rowan 2016 Red Mountain + - H2 NC-309 Rowan 2016 Red Mountain + + H2 NC-310 Rowan 2016 Red Mountain + - H2 NC-312 Rowan 2016 Red Mountain n/a n/a n/a NC-318 Rowan 2016 Picus + - H1 NC-330 Rowan 2016 Picus + - H2 NC-350 Haywood 2016 Mountain + - H2 Magic NC-352 Haywood 2016 Mountain + - H2 Magic NC-354 Haywood 2016 Mountain + - H2 Magic NC-366 Haywood 2016 Mountain + - H1 Magic NC-373 Henderson 2016 Tasty Lee + - H1 w Tomato cultivar from which Xs strains were isolated.

x Copper resistance of each strain; “+” indicates copper resistant; “-” indicates copper sensitive, and n/a indicates not applicable. y Streptomycin resistance of each strain; “+” indicates streptomycin resistant; “-” indicates streptomycin sensitive, and n/a indicates not applicable.

z Haplotype of each strain based on their BOX-PCR fingerprinting profile. n/a indicates not applicable.

RESULTS

Copper and Streptomycin Sensitivity Assays

The results documented over 95% (16/163 in 2015 and 123/127 in 2016) of Xanthomonas strains were resistant to copper and 45% (73/163) and 25% (32/127) were resistant to streptomycin in 2015 and 2016 respectively (Figure 2.2A). The distribution of coper resistant

Xanthomonas strains was uniform across year and counties i.e. nearly all farms had 100% copper resistant Xanthomonas strains in both years (Table 2.3; Table 2.4). However, the distribution of streptomycin resistant Xanthomonas strains was not as uniform throughout the counties and years as copper. A lower percentage of streptomycin resistant Xanthomonas strains were observed in 2016 compared to 2015 (Figures 2.2B, 2.2C). No streptomycin resistant

Xanthomonas strains were observed in Haywood County in 2016, but 83% Xs strains were resistant to streptomycin in the same county in year 2015. Most Xanthomonas strains collected were either resistant to both copper and streptomycin or resistant (Figure 2.3). Xanthomonas strains susceptible to both copper and streptomycin were not observed in 2015 but five strains susceptible to both copper and streptomycin were detected in 2016 (Figure 2.3). Only one strain

(isolated in 2015) was copper susceptible, but streptomycin resistant. All the Xs strains were also tested on 300 ppm of copper sulfate, but the results did not change much. Only eight additional

Xs strains were susceptible in 2015 and three in 2016 (data not shown) when tested on 300 ppm of copper sulfate as compared to 200 ppm of copper sulfate.

Table 2.3: Xanthomonas spp (Xs) strains collected from North Carolina in 2015 and used for copper and streptomycin sensitivity assays.

Number of Number of Total Tomato strains strains Field County strains cultivarsa resistant to resistant to analyzed copperb streptomycinc I Henderson Plum Regal 19 18 11

II BHN784 15 14 0

III Madison Red Defender 20 20 18

IV Grafted plants 17 17 16

V Haywood Mountain 2 2 0 Majesty VI Roma 16 16 15

VII Roma 12 12 10

VIII Buncombe Biltmore 20 20 2

IX Jackson Plum Regal 23 23 0

X Swain Plum Regal 19 19 1

Total 163 161 73

a Tomato cultivar planted in each field from which Xs strains were isolated. b Total number of Xs strains that were resistant to 200 ppm copper sulfate. c Total number of Xs strains that were resistant to 100 ppm of streptomycin sulfate.

Copper and Streptomycin Sensitivity Test A 99 97 100 80 strains 60 45 40 25 (%) 20 Resistant 0 % Resistant to Cu % Resistant to Sp Xanthomonas Year 2015 Year 2016

Year 2015 % Resistant to Cu % Resistant to Sp

strains 100 100 100 100 100 94 92 100 83 B 80 60 (%) 40 32 Xanthomonas 20 10 0 5 0 Henderson Madison Haywood Buncombe Jackson Swain Resistant Counties

Year 2016 C % Resistant to Cu % Resistant to Sp 100 100 100 100 100 87 80 60 48 Xanthomonas 40 33 33

strains (%) 20 7 0 0

Resistant Henderson Haywood Madison Rowan Macon

Counties

Figure 2.2: Overall proportions (A) and county wise proportions of Xanthomonas spp (Xs) strains collected from diverse tomato cultivars and farms across eight counties in North Carolina during 2015 (B) and 2016 (C) seasons, that were resistant to copper and streptomycin.

Table 2.4: Xanthomonas spp (Xs) strains collected from North Carolina in 2016 and used for copper and streptomycin sensitivity assays.

Total Number of Number of Tomato Field County strains strains resistant strains resistant cultivarsa analyzed to copperb to streptomycinc I Henderson Mountain Fresh 7 7 0

I Mountain Majesty 6 6 2

I Mountain Merit 4 4 0 I Plum regal 7 7 0 II Mountain Fresh 0 0 0 III Tasty-Lee 3 3 0

IV Madison Resolute 6 6 2 V Red defender 0 0 0

VI Haywood Mountain Magic 6 6 0 VII Mountain Magic 18 18 0 VIII Mountain Fresh 0 0 0

IX Rowan Red Mountain 12 12 9 X Picus 17 17 0 Mixed Breeding XI 10 10 4 lines

XII Macon Plum Regal 19 18 7 XIII Tasti- Lee 12 9 8 Total 127 123 32 a Tomato cultivar planted in each field from which Xs strains were isolated. b Total number of Xs strains that were resistant to 200 ppm copper sulfate. c Total number of Xs strains that were resistant to 100 ppm of streptomycin sulfate.

100 91 89 Year 2015 72 75 Year 2016

strains (%) strains 50 32 25 4 1 0 1 0

Xanthomonas Cu-R, Sp-S Cu-R,Sp-R Cu-S, Sp-R Cu-S,Sp-S

Phenotypes

Figure 2.3: Proportions of diverse types of Xanthomonas spp (Xs) strains observed during two- year survey in major tomato-producing regions of North Carolina, where, Cu-R, Sp-S = Resistant to copper, but sensitive to streptomycin; Cu-R, Sp-R= Resistant to both copper and streptomycin; Cu-S, Sp-S= Sensitive to both copper and streptomycin; Cu-S, Sp-R= Resistant to streptomycin, but sensitive to copper.

BOX-PCR Assay

A total of 278 Xs strains isolated during 2015 and 2016 were subjected for BOX-PCR genomic fingerprinting, which generated 13 amplicons ranging from ~300 bp to 2 kb. Based on the BOX-PCR scores, four haplotypes (H1, H2, H3, and H4) were determined- 258 strains had

H1 haplotype (92.8%), 18 strains had H2 haplotype (6.4%), one strain was determined for H3

(0.4%) and H4 haplotypes (0.4%) (Table 2.5). Strains with H1 haplotype were detected in all eight counties in both years. H2 haplotypes were detected in the strains collected from

Henderson County (NC-1, NC-7, NC-10, NC-11, NC-12, NC-14, NC-101) in 2015, and Macon

(NC-279, NC-280, NC-283), Rowan (NC-304, NC-309, NC-310, NC-329, NC-330) and

Haywood (NC-350, NC-352, and NC-354) counties in 2016. The strain NC-112 isolated from

Madison county in 2015 was defined by H3 haplotype and the strain NC-282 isolated from

Macon county in 2016 represented H4 haplotype (Table 2.5). The hierarchical agglomerative cluster analysis of Xs strains grouped four haplotypes in four clusters/groups- cluster1 contained

H2 haplotypes; cluster 2 contained H1 haplotypes; cluster 3 contained H4 haplotype, and cluster

4 contained H3 haplotype (Figure 2.4). The cophenetic correlation coefficient of the clustering was 0.99, which indicates the cluster dendrogram preserved the original distances between data points. The distance between cluster 1 and cluster 2 was ~0, whereas cluster 3 and 4 showed the greater distance between each other and between cluster 1 and cluster 2 (Figure 2.4), indicating all Xs strains tested in this study except two strains in cluster 3 and 4 (NC-282 and NC-112) were similar.

Table 2.5: Summary of haplotypes based on polymerase chain reactions (PCR) using repetitive sequence-based PCR with the BOX fingerprint profiles detected among Xanthomonas spp (Xs) strains from North Carolina in 2015 and 2016.

Haplotyp Amplicon detected in each haplotypew Yearx Countyy #strainsz

e 1 2 3 4 5 6 7 8 9 10 11 12 13 2015, All 8 counties 25 H1 1 1 1 1 1 1 1 1 1 1 1 1 1 2016 8 Henderson, 2015, Macon, H2 1 1 1 0 1 1 1 1 1 1 1 1 1 18 2016 Haywood, Rowan H3 0 0 0 0 0 0 0 0 1 1 1 1 1 2015 Madison 1 H4 0 0 0 0 0 0 0 0 1 0 1 0 0 2016 Macon 1 w A total of 13 amplicons were generated from BOX-PCR. “1” represents the presence of particular amplicon and “0” represents the absence of particular amplicon in the given haplotype. x Year of isolation of the strains belonging to each haplotype. y Counties from where the strains belonging to each haplotype detected. z Total number of strains belonging to each haplotype.

Figure 2.4: Hierarchical clustering dendrogram based repetitive sequence-based polymerase chain reaction (rep-PCR) with the BOX element primer set for Xanthomonas spp (Xs) strains analyzed in this study. These clusters were based on Euclidean distance matrices. The linkage method with group average (unweighted-pair-group) technique was used to define the hierarchical clustering. Only strains with representative haplotypes are shown for clarity purposes. 93% of total Xs strains were grouped into cluster 2, 6.5% of Xs strains were grouped into cluster 1. Clusters 3 and 4 contained only one strain. The cophenetic correlation coefficient of the clustering was 0.99, which indicated the cluster dendrogram preserved the original distances between datat points.

Multiplex Quantitative Real-Time PCR (qPCR) Assay

Reference strains for Xe, Xv, Xp, and Xg were amplified and represented by different colors in the multiplex real-time TaqMan PCR assay (Figure 2.5). All the strains amplified were represented by the same color as that of Xp, indicating all Xanthmonas strains belonged to X.

perforans (Figure 2.5). Therefore, we will denote Xs strains of NC as Xp in the following sections.

Figure 2.5: Image showing amplifications of 45 representative Xanthomonas spp (Xs) strains along with four type strains belonging to each species analyzed by quantitative real time polymerase chain (qPCR) assay. Green, blue, red and purple lines represent X. perforans, X. euvesicatoria, X. gardneri, and X. vesicatoria, respectively.

Virulence Assay and Race Identification

All representative Xp strains except two (NC-14 and NC-350) used in this study induced

HR (Figure 2.6) on all four pepper differentials (Table 2.6), suggesting non-host resistance of pepper and indicates these strains are only pathogenic on tomato. Isolates NC-14 and NC-350

induced HR only on ECW 20R (contains Bs2 gene), but not on ECW 10R and ECW 30R, suggesting NC-14 and NC-350 might be pathogenic on pepper too. Among 45 isolates, only three (NC-79, NC-80, and NC-112) induced HR on FL 216 (Figure 2.6), suggesting these three strains belong to race T3 (Table 2.6). The remainder of Xp strains did not induce HR on HI 7998 nor FL 216.

Figure 2.6: Hypersensitive response (HR) induced on pepper (A) and tomato (B) by X. perforans race T4 and T3 respectively.

Table 2.6: Virulence assay and hypersensitive response (HR) of representative Xanthomonas perforans (Xp) strains from major tomato-producing regions of North Carolina. HR; + indicates induction of HR on specific differentials, and - indicates the absence of HR on specific tomato differential cultivars.

Hypersensitive Response (HR) on Strains HI FL Bonny ECW ECW ECW ECW Race Virulence 7998 216 Best 10R 20R 30R identified assay NC-10, NC-11, - - - + + + + T4 Tomato NC-12, NC-22, NC-28, NC-37, NC-42, NC-47, NC-67, NC-87, NC-101, NC-110, NC-124, NC-126, NC-145, NC-157, NC-180, NC-189, NC-193, NC-204, NC-232, NC-242, NC-248, NC-252, NC-264, NC-280, NC-282, NC-283, NC-289, NC-293, NC-304, NC-309, NC-310, NC-318, NC-330, NC-352, NC-354, NC-366, NC-373

NC-14, NC-350 - - - - + - - T4 Tomato, Pepper

NC-079, NC-80, - + - + + + + T3 Tomato NC-112, NC-312

Xp 91-118 - + - + + + + T3 Tomato (control) Xe E3 (control) + - - + + - - T1 Tomato, Pepper Water ------

Multi-locus Sequence Analysis (MLSA)

Of the 42 representative Xp strains assessed, sequence data were obtained only for 41 strains for gyrB gene, 39 strains for gltA gene, 38 strains for gapA gene, 37 strains for fusA gene, and 35 strains for both lacF and lepA genes, because of either poor amplification in PCR assay or poor sequence quality. A total of 32 strains of Xp were used for the construction of a phylogenetic tree based on concatenated sequences of six housekeeping genes, whereas phylogenetic tree based on individual genes were constructed using their respective sequenced samples. Consistent with the BOX-PCR fingerprinting profile and hrp7 based real-time PCR assay, MLSA based phylogenetic analysis did not show much diversity among Xp strains. All Xp strains were grouped with reference Xp strains (Figure 2.7;). No variation was observed among sequenced Xp strains in fusA, lacF, and lepA gene sequences, while two, three and four haplotypes were observed in gltA, gyrB and gapA gene sequences, respectively (Figure 2.7;

Table 2.7). Although, the rest of the strains did not show sequence variation in gltA gene sequence, strain NC-282 showed a SNP variation at nucleotide sites 27 and 78, which grouped

NC-282 into a different group along with X. euvesicatoria type strain 85-10 (isolated from

Florida in 1985) and atypical Nigerian strain Xs-N1 (Timilsina et al. 2015) (Figure 2.7). The gyrB gene differentiated Xp strains into two groups- 23 strains grouped together with Xp reference strains GEV839 race T4 (isolated from Florida in 2012) and 4-20 race T4 (isolated from Florida in 2006) and Xe type strain 85-10, while 18 strains grouped with Xp type strains

ICMP 16690 767 race T3 (isolated from Florida in 1991), 17-12 race T4 (isolated from Florida in 2006), and 91-118 race T3 (isolated from Florida in 1991) (Timilsina et al. 2015). The gapA gene differentiated Xp strains into three groups-24 strains grouped with Xp reference strain

GEV839 and Xe reference strain 85-10, 9 strains were grouped with Xp reference strain 17-12, and 5 strains were grouped together with Xp reference strains ICMP 16690 767, 4-20, and 91-

118. The number of segregating sites and their nucleotide positions for these genes were estimated for each locus and gapA had more mutation or loss and yielded more haplotypes compared to other loci (Table 2.7).

Table 2.7: Sequence statistics of six housekeeping genesw (fusA, gapA, gltA, gyrB, lacF, and lepA) in Xanthomonas perforans strains from North Carolina.

Gene sequence Nx Hy Sz Nucleotide position of S fusA 591 1 0 n/a gapA 444 3 4 186,195,283,429 gltA 501 2 2 27, 78 gyrB 411 2 1 39 lacF 408 1 0 n/a lepA 390 1 0 n/a Concatenated 2745 4 7 777, 786, 874, 1020, 1062, 1113, 1575 w fusA (elongation factor G), gapA (glyceraldehyde-3-phosphate dehydrogenase A), gltA (citrate synthase), gyrB (gyrase B), lacF (ABC transporter sugar permease), and lepA (GTP binding protein). x N = number of nucleotides present. y H = number of haplotypes detected. z S = total number of segregating sites

Based on the concatenated sequences, Xp strains were divided into two groups- G1 (13 strains; 41%) and G2 (19 strains; 59%). G1 was differentiated into three subgroups G1-A (8 strains; 25% of total), G1-B (4 strains; 13% of total), and G1-C (1 strain; 3% of total) (Figure

2.8). The total of four haplotypes (Ha1, Ha2, Ha3, and Ha4) with seven segregating sites were obtained based on the concatenated sequences (Table 2.8). The SNPs and nucleotide positions for the segregating sites are summarized in Table 2.8. Ha1 belonged to group2, whereas Ha2,

Ha3, and Ha4 belonged to group 1 (Table 2.8). The Ha1 haplotype in group2 differed from Ha2,

Ha3, and Ha4 in group1 by four, five, and seven SNPs respectively. In group1, the Ha2 haplotype differed from Ha3 and Ha4 by one and three SNPs, respectively. The Ha3 haplotype differed from Ha4 haplotype by two SNPs. Most NC strains were similar to the Xp T4 strain

GEV839 isolated from Florida in 2012 (Timilsina et al. 2016) and were grouped in G2 (Figure

2.8). NC strains in G1-A were similar to the Xp T4 reference strain 17-12 isolated from Florida in 2006 (Figure 2.7) (Timilsina et al. 2015). Only four strains were grouped into G1-B and were similar to the reference strain Xp 91-118 from Florida in 1991 (Timilsina et al. 2015) and

ICMP16690-767. The strain NC-282 was placed under G1-C (Figure 2.8).

Table 2.8: Summary of haplotypes and single nucleotide polymorphism (SNP) variation among Xanthomonas perforans strains from North Carolina based on concatenated sequences of six housekeeping genes (fusA, gapA, gltA, gyrB, lacF, and lepA).

Nucleotide Haplotypes position (bp) Group Strains 777 786 874 1020 1062 1113 1575 MRS-30P-11, NC-28, NC-47, NC-67, NC-87, NC-126, NC-145, NC- 189, NC-193, NC-232, Ha1 T G A G A T T G2 NC-242, NC-248, NC- 252, NC-289, NC-309, NC-318, NC-352, NC- 354, NC-373, Xp- GEV839

NC-10, NC-11, NC-14, NC-22, NC-101, NC- Ha2 G C G G A T C G1-A 304, NC-330, NC-350, Xp 17-12

NC-12, NC-42, NC- 112, NC-204, Xp-91- Ha3 G C G C A T C G1-B 118, Xp ICMP16690- 767

Ha4 G C G C C C C G1-C NC-282

MRS-30P-011 NC-10 Xe-85-10 MRS-30P-011 NC-11 A B Xp-GEV839 C NC-10 NC-12 NC-373 NC-11 NC-14 NC-354 NC-12 NC-22 NC-352 NC-14 NC-28 NC-318 NC-22 NC-42 NC-309 NC-28 NC-47 NC-289 NC-42 NC-67 NC-252 NC-47 NC-87 93 NC-248 NC-67 NC-101 NC-242 NC-87 NC-112 NC-232 NC-101 NC-126 NC-193 NC-112 NC-145 NC-189 NC-126 NC-189 NC-145 NC-145 NC-193 NC-126 NC-189 NC-204 NC-87 NC-193 61 NC-232 MRS-30P-011 88 NC-204 NC-242 NC-28 NC-232 NC-248 NC-47 NC-242 98 NC-252 NC-67 NC-248 NC-282 Xp ICMP16690 767 NC-289 NC-289 Xp 4-20 NC-304 NC-304 Xp-91-118 NC-309 NC-309 61 NC-282 NC-318 NC-318 NC-204 NC-330 NC-330 NC-112 NC-350 99 NC-42 NC-350 99 NC-352 NC-352 NC-12 NC-354 99 NC-10 NC-354 NC-373 NC-11 NC-373 Xp-91-118 NC-14 Xp 17-12 72 Xp 17-12 NC-22 Xp-GEV839 Xp-GEV839 NC-101 Xp ICMP16690 767 Xp ICMP16690 767 NC-304 Xs N1 Xp 4-20 NC-330 Xp 4-20 fusA NC-252 NC-350 Xp-91-118 NC-282 Xp 17-12 Xe-85-10 Xe-85-10 Xs N1 Xv-ATCC 11551 (196) Xs N1 Xv-ATCC 11551 (196) Xg ICMP16689 766 Xv-ATCC 11551 (196) Xg ICMP16689 766 Xg ICMP16689 766 0.0050 0.0100 0.0100

MRS-30P-011 NC-252 MRS-30P-011 NC-10 NC-289 NC-10 NC-11 D NC-248 E NC-11 F NC-12 NC-242 NC-12 NC-14 NC-232 NC-14 NC-22 NC-193 NC-22 NC-28 NC-189 NC-28 NC-42 NC-145 NC-42 NC-47 NC-126 NC-47 NC-67 NC-87 NC-67 NC-87 NC-67 NC-87 NC-101 NC-47 NC-101 NC-112 NC-28 NC-112 NC-126 MRS-30P-011 NC-126 NC-145 99 NC-309 NC-145 NC-189 NC-318 NC-189 NC-193 NC-352 NC-193 99 NC-204 NC-354 99 NC-204 NC-232 NC-373 NC-232 NC-242 Xe-85-10 NC-242 NC-248 Xp-GEV839 NC-248 NC-252 Xp 4-20 NC-252 NC-282 NC-10 NC-282 NC-289 NC-11 NC-289 NC-304 NC-12 NC-304 NC-309 NC-14 NC-309 NC-318 NC-22 NC-318 99 NC-330 NC-42 NC-330 95 NC-350 NC-101 NC-350 NC-352 NC-112 NC-352 NC-354 NC-204 NC-354 NC-373 63 NC-282 NC-373 Xp 17-12 NC-304 Xp-91-118 Xp-GEV839 NC-330 Xp 17-12 Xp ICMP16690 767 NC-350 Xp-GEV839 Xp-91-118 Xp-91-118 Xp ICMP16690 767 Xs N1 Xp 17-12 Xp 4-20 Xe-85-10 Xp ICMP16690 767 Xe-85-10 51 Xp 4-20 Xs N1 Xs N1 67 99 Xv-ATCC 11551 (196) Xg ICMP16689 766 Xg ICMP16689 766 Xg ICMP16689 766 Xv-ATCC 11551 (196) Xv-ATCC 11551 (196) 0.0100 0.0100 0.020

Figure 2.7: Maximum likelihood phylogenetic analysis of X. perforans strains isolated from NC using fusA gene sequence (A), gapA gene sequence (B), gltA gene sequence (C), gyrB gene sequence (D) lacF gene sequence (E), and lepA gene sequence (F). Reference strains of X. euvesicatoria (Xe), X. gardneri (Xg), X. perforans (Xp), X. vesicatoria (Xv), and atypical Nigerian strain (N1) were downloaded from the Plant Associated and Environmental Microbes Database and the National Center for Biotechnology Information Database. Values on the branches indicate bootstrap values for each branch, expressed as percentages. The scale bar indicates the number of substitutions per site.

Meta Data NC-373 A Xp-GEV839 B NC-354 NC-352 NC-318 NC-309 NC-289 NC-252 NC-248 98 NC-242 Group 2 NC-232 NC-193 NC-189 MRS-30P-11 NC-28 NC-47 NC-67 NC-87 NC-126 92 NC-145 Xp-91-118 Xp ICMP16690 767 45 NC-204 NC-112 Group 1-B NC-42 NC-12 NC-282 Group 1-C 96 NC-10 NC-11 NC-14 43 NC-22 NC-101 Group 1-A 100 NC-304 NC-330 NC-350 Xp 17-12 Xp 4-20 Xe-85-10 61 Xs N1 Xv-ATCC 11551 (196) Xg ICMP16689 766

0.0100 Figure 2.8: Maximum likelihood phylogenetic analysis of X. perforns strains isolated from North Carolina using concatenated sequence (A) and its visualization with counties, BOX-PCR fingerprint profiles, copper and streptomycin sensitivity, and race profile using T-BAS v2 (B). Reference strains of X. euvesicatoria, X. gardneri, X. perforans, and X. vesicatoria were downloaded from the Plant Associated and Environmental Microbes Database and the National Center for Biotechnology Information Database. Values on both Figures A and B branches indicate bootstrap values for each branch, expressed as percentages. The scale bar on A indicates the number of substitutions per site. The tips of branches on B are colored to indicate four haplotypes generated from multilocus sequence analysis (MLSA) using of six housekeeping genes: fusA (elongation factor G), gapA (glyceraldehyde-3-phosphate dehydrogenase A), gltA (citrate synthase), gyrB (gyrase B), lacF (ABC transporter sugar permease), and lepA (GTP binding protein) and the outer rings indicate the meta data- BOX-PCR haplotypes, streptomycin sensitivity, copper sensitivity, race, and county surveyed respectively from inward to outward. 91

DISCUSSION

Bacterial spot is currently a widespread disease in tomato-producing regions in NC. In this study, we have combined conventional and multi-locus sequencing approaches to rapidly secure information about the Xp - tomato pathosystem. The results documented that there is a widespread existence of copper resistance and a modest level of streptomycin resistance among

Xp strains, which is consistent with many grower observations that copper sprays were not effective to limit bacterial spot incidence. The higher incidence of bacteria resistant to copper than streptomycin observed in this study indicates there have been more selection pressure for copper resistance, presumably associated with frequent copper use (Cooksey 1987). In planta virulence on tomato and pepper differentials indicated that majority of Xp strains belong to race

T4 followed by race T3. The qPCR assay and phylogenetic analysis further allowed comparisons between closely related strains and suggested that although strains within Xp were mostly geographically isolated, yet a majority of the strains were genetically highly homogenous. In addition, the majority of Xp strains in NC were similar to the recently characterized Xp strains in

FL. A productive avenue of future research is to investigate and compare effectors in the Xp strains collected from NC and FL.

Resistance to copper and streptomycin appeared to be a global challenge due to the extensive use of bactericides and antibiotics, which favors the emergence and spread of bacterial strains resistant to copper bactericides (Cooksey 1987). Streptomycin and copper resistance in Xs populations were first reported in 1961 and 1983, respectively from Florida (Thayer and Stall

1961; Marco and Stall 1983); however, examination of older collections suggested that resistance had been present in populations of the pathogen in Florida since 1968 (Voloudakis et al. 1993). Both copper and streptomycin resistant Xs populations were also reported from

Caribbean and Central America in 1999 (Bouzar et al. 1999), Florida (Vallad et al. 2013), Brazil

(Araújo et al. 2012) and Ethiopia (Kebede et al. 2014). The widespread existence of copper resistant Xs strains were reported from Ontario, Canada (Abbasi et al. 2015); Tanzania (Shenge et al. 2014); and Tennessee (Mixon 2012). Limited efficacy of copper-based bactericides for the management of bacterial spot disease of tomato has also been observed in Ohio (Miller and Mera

2011). Notably, all fields sampled in NC harbored Xs strains either resistant to copper or streptomycin. Although Ritchie and Dittapongpitch (1991) reported 63% copper resistant and

30% streptomycin-resistant strains of Xanthomonas species in NC, their findings were focused on pepper and collections from tomato were not well represented. It is not yet clear whether the association of copper-resistance with bacterial strains results of this study is primarily from the spread of already established multiple antibiotics resistant strains or rather from the independent acquisition of resistance determinants by susceptible strains of the same clone. We believe that different selection pressures and genetic pools providing resistance determinants and instability of some resistance determinants (Diancour et al. 2010), all could contribute in explaining the severe outbreaks of bacterial spot disease in NC in recent years.

Most of the genes associated with copper resistance in plant-pathogenic bacteria are plasmid encoded (Basim et al. 2005; Cooksey 1987; Richard et al. 2017). Copper resistance genes in Xanthomonas campestris pv. vesicatoria strains were detected on 188- to 200-kb self- transmissible plasmids pXvCu and pXV10a in strains from Florida (Bender et al. 1990; Stall et al. 1986). DNA hybridization indicated that the resistance genes on pXV10A and pXvCu plasmids share nucleotide sequence homology and may have a common origin (Bender et al.

1990). Another 100-kb non-self-transmissible plasmid containing loci for copper resistance was also discovered in a strain from California (Cooksey 1987). Chromosome encoded copper resistance gene clusters have also been reported in a 7,652-bp XbaI/EcoRI chromosomal fragment in X. campestris pv. vesicatoria strain XvP26 from Taiwan (Basim et al. 2005).

Recently, the sequencing of six tomato and pepper-copper resistant pathogens belonging to different species of Xanthomonas showed that all six strains carried the previously described copper resistance gene system copLAB on a conjugative plasmid (Richard et al. 2017). The plasmid was highly conserved among five strains (dating from 1955 to 2010), suggesting extensive events of horizontal gene transfer at the niche level (Richard et al. 2017). The transfer of plasmids between copper-resistant and copper-sensitive strains through conjugation has also been demonstrated under laboratory conditions (Stall et al. 1986). The homology with the copper resistance (cop) operon previously cloned from Pseudomonas syringae pv. tomato strain PT23 was detected in other saprophytic bacteria on tomato plants, suggesting these saprophytic bacteria as a potential reservoir of copper resistance genes that could be acquired by copper- sensitive pathogens (Cooksey et al. 1990). The presence of a copper-resistant P. putida strain with a plasmid borne cop homolog in a commercial tomato seed lot also suggests that saprophytic bacteria could contribute to the spread of resistant bacterial populations between fields and different geographical areas (Cooksey et al. 1990). Although only one source of copper (copper sulfate) for Xs strains was examined, evaluating and comparing the efficacy of newly developed copper composites such as core-shell copper (CS-Cu), multivalent copper

(MV-Cu), and fixed quaternary ammonium copper FQ-Cu (Strayer-Scherer et al. 2017) and understanding mechanisms of resistance to copper in Xp strains from NC would be potentially useful to reduce rapidly growing copper-resistant strains and to manage bacterial spot in tomato.

In the case of streptomycin, resistance occurs either due to the mutation in the binding affinity of ribosomal proteins for this antibiotic or due to the action of periplasmic enzymes encoded by plasmid-borne genes on the antibiotic (Springer et al. 2001). Both chromosomal and plasmid-mediated streptomycin resistance in plant-pathogenic bacteria have been reported

(Minsavage et al. 1990). Streptomycin resistance might occur spontaneously because of the

direct interaction of antibiotic molecules with the small ribosomal subunit. Besides, horizontal gene transfer through conjugation might also develop resistance to streptomycin (Behlau et al.

2012). The possible reasons for the widespread existence of copper resistance and moderate streptomycin resistance in Xs populations in NC are unknown. However, since the pathogens are seed-borne, it is likely that resistance plasmid was introduced in a common seed source and then distributed throughout different regions through pathogen-contaminated seed. In addition, tomato growers in NC usually purchase transplants from nurseries instead of raising seeding themselves.

Therefore, it is also possible that copper and streptomycin resistant bacteria are introduced into different areas of NC through common transplant nurseries. Many nurseries use streptomycin twice in the transplant production phase of tomato production which would select for streptomycin reistant strains.

Among four species of Xanthomonas causing bacterial spot disease in tomato, this work identified a single species X. perforans in NC. Based on HR assay, 91% of them belonged to race

T4 and 9% belonged to race T3, which indicates race T3 has not been completely displaced by race T4 in NC as observed in Florida (Vallad et al. 2013). Our results suggest the emergence of tomato race T4 strains of Xp in NC carrying mutations in the avrXv3 gene (Astua-Monge et al.

2000; Horvath et al. 2012). Although X. perforans has been reported to infect tomato only, two strains in our study did not elicit HR in pepper differentials except ECW 20 R (contains Bs2 resistant gene against highly conserved effector avrBs2). This indicates the presence of some other genetic factors in these strains that defines its host range on the pepper.

Both BOX-PCR fingerprinting and MLSA detected very low genetic diversity within Xp in NC and yielded consistent results. Similarly, phylogenetic tree based on concatenated sequences of six housekeeping genes also grouped Xp strains of NC in two groups that differed by few numbers of nucleotide per site. Strains with H1 haplotype based on BOX-PCR

fingerprinting were detected with haplotype one based on concatenated sequences of six housekeeping genes except for NC-204, NC-42, which had Ha3 haplotype, and NC-22, which had haplotype 1 based on MLSA. Similarly, strains with haplotype 2 based on BOX-PCR fingerprinting were detected with Ha2 haplotype based on concatenated sequences of six housekeeping genes except for NC-309, NC-352, NC-354, which had haplotype 1 and NC-12, which had haplotype 3 based on concatenated sequences of six housekeeping genes. Although haplotypes H1 and H2 contained 93% and 6% strains, respectively, these two haplotypes were genetically highly homogenous (d = ~ 0).

The MLSA studies of Xp populations from Florida based on six genes showed the presence of two distinct groups and appeared to be clonal within the lineage (Timilsina et al.,

2015). The largest group in the phylogenetic tree that contained 59% of Xp strains of NC included the Xp T4 strain GEV839 isolated from Florida in 2012, suggesting the majority of Xp strains in NC are in the same phylogenetic group as that of Xp strains in Florida (Timilsina et al.

2016). Similarly, 25% of X. perforans strains in NC were grouped with Xp T4 strain Xp17-12 isolated from Florida in 2006. A few Xp strains in NC (13%) were grouped with Xp T3 strain 91-

118 and ICMP166690-767. The strain NC-12, NC-42, and NC-204, which were detected as race

T4 based on HR assays were grouped together with race T3 type strain in the phylogenetic tree.

In addition to strain-level variation, allelic diversification in type III effectors might resolve their race profile. However, both results indicate the presence of race T3 strains in NC. The strain NC-

282 with Ha4 haplotype based on BOX-PCR grouped separately from the rest of the strains in the phylogenetic tree constructed using concatenated sequences of six housekeeping genes in this study. This strain had gltA gene sequence similar to that of X. euvesicatoria, and other five gene sequences similar to that of X. perforans. Intriguingly, the strain NC-282 was sensitive to both copper and streptomycin. Previously, one atypical Nigerian strain has been reported that had

fusA and gyrB gene identical with X. perforans, but gltA and lacF genes were identical to X. euvesicatoria (Timilsina et al. 2015). However, none of the NC Xp strains grouped with this atypical Nigerian strain in the phylogenetic analysis using concatenated gene sequences. Overall,

MLSA revealed Xp strains in NC were similar with Xp strains in Florida, suggesting that Xp strains from NC might have a common origin to the Florida strains. We hypothesize that exchange of contaminated-seeds and tomato transplants might be responsible for the introduction of pathogen throughout the state. However, this observation requires further investigations to determine whether seed and/or transplants are the main carriers for long distance movement.

In summary, our study demonstrates the phenotypic and genomic diversity of Xp populations in NC. We confirm that Xp race T3 is rarely present while race T4 is widely distributed in major tomato-growing regions in NC. The majority of the Xp strains were resistant to copper-based bactericides. These results suggest that the use of copper or streptomycin alone in NC would be ineffective to manage bacterial spot in tomato and highlight the need for a broadly effective strategy. Tomato growers should implement best-management practices such as the use of plant activators that induce systemic acquired resistance (e. g., Actigard), biopesticides, and use of Xp-free-seed and tomato transplants (Burlakoti et al., 2018; Louws

2018; Miller et al. 2014). Alternatively, effector-based breeding strategies (Timilsina et al. 2016) supported by new genomic tools for improving and deploying multiple R proteins in tomato varieties should be exploited to combat emerging races of Xp in NC.

REFERENCES

Abbasi, P. A., Khabbaz, S. E., Weselowski, B., and Zhang, L. 2015. Occurrence of copper-

resistant strains and a shift in Xanthomonas spp. causing tomato bacterial spot in Ontario.

Can. J. Microbiol. 61:753-761.

Almeida, N. F., Yan, S., Cai, R., Clarke, C. R., Morris, C. E., Schaad, N. W., Schuenzel, E. L.,

Lacy, G. H., Sun, X., Jones, J. B., and Castillo, J. A. 2010. PAMDB: a multilocus sequence

typing and analysis database and website for plant-associated microbes. Phytopathology

100:208-215.

Araújo, E. R., Costa, J. R., Ferreira, M. A., and Quezado‐Duval, A. M. 2017. Widespread

distribution of Xanthomonas perforans and limited presence of X. gardneri in Brazil. Plant

Pathol. 66:159-168.

Araújo, E., Costa, J., Pontes, N., and Quezado-Duval, A. 2015. Xanthomonas perforans and X.

gardneri associated with bacterial leaf spot on weeds in Brazilian tomato fields. Eur. J.

Plant Pathol. 14:543-548.

Araujo, E. R., Pereira, R. C., Ferreira, M. A. S. V., Café-Filho, A. C., Moita, A. W., and Quezado-

Duval, A.M. 2010. Effect of temperature on pathogenicity components of tomato bacterial

spot and competition between Xanthomonas perforans and X. gardneri. In III International

Symposium on Tomato Diseases 914:39-42.

Araújo, E., Pereira, R., Ferreira, M., Quezado-Duval, A., and Café-Filho, A. 2012. Sensitivity of

Xanthomonads causing tomato bacterial spot to copper and streptomycin and in vivo infra-

specific competitive ability in Xanthomonas perforans resistant and sensitive to copper. J.

Plant Pathol. 94:79-87.

Astua-Monge, G., Minsavage, G. V., Stall, R., Vallejos, C. E., Davis, M., and Jones, J. B. 2000.

Xv4-vrxv4: A new gene-for-gene interaction identified between Xanthomonas campestris

pv. vesicatoria race T3 and the wild tomato relative Lycopersicon pennellii. Mol. Plant

Microbe Interact. 13:1346-1355.

Basim, H., Minsavage, G. V., Stall, R. E., Wang, J. F., Shanker, S., and Jones, J. B. 2005.

Characterization of a unique chromosomal copper resistance gene cluster from

Xanthomonas campestris pv. vesicatoria. Appl. Environ. Microbiol. 71:8284-8291.

Behlau, F., Jones, J. B., Myers, M. E., and Graham, J. H. 2012. Monitoring for resistant

populations of Xanthomonas citri subsp. citri and epiphytic bacteria on citrus trees treated

with copper or streptomycin using a new semi-selective medium. Eur. J. Plant Pathol.

132:259-270.

Bender, C. L., Malvick, D. K., Conway, K. E., George, S., and Pratt, P. 1990. Characterization of

pXV10A, a copper resistance plasmid in Xanthomonas campestris pv. vesicatoria. Appl.

Environ. Microbiol. 56:170-175.

Bouzar, H., Jones, J., Stall, R., Louws, F., Schneider, M., Rademaker, J., de Bruijn, F. J., and

Jackson, L. E. 1999. Multiphasic analysis of Xanthomonads causing bacterial spot disease

on tomato and pepper in the Caribbean and Central America: Evidence for common

lineages within and between countries. Phytopathology 89:328-355.

Burlakoti, R. R., Hsu, C. F., Chen, J. R., and Wang, J. F., 2018. Population Dynamics of

Xanthomonads Associated with Bacterial Spot of Tomato and Pepper during 27 Years

across Taiwan. Plant Dis. PDIS-04.

Carbone, I., White, J. B., Miadlikowska, J., Arnold, A. E., Miller, M. A., Kauff, F., U'ren, J. M.,

May, G., and Lutzoni, F. 2016. T-BAS: Tree-Based Alignment Selector toolkit for

phylogenetic-based placement, alignment downloads and metadata visualization: an

example with the Pezizomycotina tree of life. Bioinformatics 33:1160-1168.

Cooksey, D. A. 1987. Characterization of a copper resistance plasmid conserved in copper-

resistant strains of Pseudomonas syringae pv. tomato. Appl. Environ. Microbiol. 53:454-

456.

Cooksey, D. A., Azad, H. R., Cha, J. S., and Lim, C. K. 1990. Copper resistance gene homologs in

pathogenic and saprophytic bacterial species from tomato. Appl. Environ. Microbiol.

56:431-435.

Cuppels, D. A., Louws, F. J., and Ainsworth, T. 2006. Development and evaluation of PCR based

diagnostic assays for the bacterial speck and bacterial spot pathogens of tomato. Plant Dis.

90:451-458.

Diancour, L., Passet, V., Nemec, A., Dijkshoorn, L., and Brisse, S. 2010. The population structure

of Acinetobacter baumannii: Expanding multi-resistant clones from an ancestral

susceptible genetic pool. PLoS ONE 5:e10034.

Gardner, M., and Kendrick, J. 1921. Bacterial spot of tomato. J. Agric. Res. 21:123-156.

Horvath, D. M., Stall, R. E., Jones, J. B., Pauly, M. H., Vallad, G. E., Dahlbeck, D., Staskawicz,

B. J., and Scott, J. W. 2012. Transgenic resistance confers effective ﬁeld level control of

bacterial spot disease in tomato. PLoS ONE 7:e42036.

Jones, E., Oliphant, T., and Peterson, P. 2014. {SciPy}: open source scientific tools for {Python}.

https://www.scipy.org/.

Jones, J., lacy, G. H., Bouzar, H., Stall, R. R., and Schaad, N. W. 2004. Reclassification of the

Xanthomonads associated with bacterial spot disease of tomato and pepper. Syst. Appl.

Microbiol. 27:755-762.

Jones, J., Pohroneznt, K. L., Stall, R., and Jones, J. P. 1986. Survival of Xanthomonas campestris

pv. vesicatoria in Florida on tomato crop residue, weeds, seeds, and volunteer tomato

plants. Phytopathology 76: 430-434.

100

Kebede, M., Timilsina, S., Ayalew, A., Admassu, B., Potnis, N., Minsavage, G., Goss, E. M.,

Hong, J. C., Strayer, A., Paret, M., Jones, J. B., and Vallad, G. E. 2014. Molecular

characterization of Xanthomonas strains responsible for bacterial spot of tomato in

Ethiopia. Eur. J. Plant Pathol. 140:677-688.

Kumar, S., Stecher, G., and Tamura, K. 2016. MEGA7: molecular evolutionary genetics analysis

version 7.0 for bigger datasets. Mol. Biol. Evol. 33:1870-1874.

Louws, F. J. 2018. Evaluation of biopesticides and biorationals on bacterial canker and bacterial

spot disease levels in tomato fresh-market production in North Carolina. ISHS Acta

Horticulturae 1207: V International Symposium on Tomato Diseases: Perspectives and

Future Directions in Tomato Protection. E. Moriones, R. Fernández-Muñoz, and C.R.

Beuzón (eds), Málaga, Spain.

Louws, F. J., Fulbright, D. W., Stephens, C. T., and de Bruijn, F. J. 1995. Differentiation of

genomic structure by rep-PCR fingerprinting to rapidly classify Xanthomonas campestris

pv. vesicatoria. Phytopathology 85:528-536.

Louws, F., and Cuppels, D. A. 2001. Molecular techniques. In N. Schaad, J. Jones, and W. Chun

(Eds.), Laboratory guide for identification of plant pathogenic bacteria (Third ed., pp. 321-

328). American Phytopathological Society, St. Paul, Minnesota, APS Press.

Louws, F. J., Wilson, M., Campbell, H. L., Cuppels, D. A., Jones, J. B., Shoemaker, P. B., Sahin,

F., and Miller, S. A. 2001. Field control of bacterial spot and bacterial speck of tomato

using a plant activator. Plant Dis. 85:481-488.

Ma, X., Ivey, M. L., and Miller, S. 2011. First report of Xanthomonas gardneri causing bacterial

spot of tomato in Ohio and Michigan. Plant Dis. 95:1584-1584.

101

Marco, G. M., and Stall, R. E. 1983. Control of bacterial spot of pepper initiated by strains of

Xanthomonas campestris pv. vesicatoria that differ in sensitivity to copper. Plant Dis.

67:779-781.

Miller, S. A., and Mera, J. 2011. Evaluation of fungicides and bactericides for the control of foliar

and fruit diseases of processing tomatoes. Plant Dis. Manag. Rep.6:V061.

Miller, S., Mera, J. R., Pantigoso, H. A., Rehm, N. R., and Vrisman C. M. 2014. Evaluation of

bactericides and a plant activator for the control of bacterial leaf spot of processing

tomatoes. Plant Dis. Manage. Rep. 9:V055.

Minsavage, G. V., Canteros, B. I., and Stall, R. E. 1990. Plasmid-mediated resistance to

streptomycin in Xanthomonas capestris pv. vesicatoria. Phytopathology 80: 719-723.

Minsavage, G., Balogh, B., Stall, R., and Jones, J. 2003. New tomato races of Xanthomonas

campestris pv. vesicatoria associated with mutagenesis of tomato race 3 strains.

Phytopathology 93:S62.

Mixon, J. T. 2012. Prevalence of copper resistance among foliar bacterial pathogens of tomato in

Tennessee. Master's Thesis, University of Tennessee. Retrieved from

http://trace.tennessee.edu/utk_gradthes/1187.

Pitcher, D. G., Saunders, N. A., and Owen, R. 1989. Rapid extraction of bacterial genomic DNA

with guanidium thiocyanate. Lett. Appl. Microbiol. 8:151-156.

Potnis, N., Timilsina, S., Strayer, A., Shantharaj, D., Barak, J. D., Paret, M. L., Vallad, G. E., and

Jones, J. B. 2015. Bacterial spot of tomato and pepper: diverse Xanthomonas species with

a wide variety of virulence factors posing a worldwide challenge. Mol. Plant Pathol.

16:907-920.

102

Rademaker, J. L., and de Bruijn, F. J. 1997. Characterization and classification of microbes by

rep-PCR genomic fingerprinting and computer assisted pattern analysis. DNA markers:

Protocols, applications and overviews 1:151-171.

Richard, D., Boyer, C., Lefeuvre, P., Canteros, B. I., Beni-Madhu, S., Portier, P., and Pruvost, O.

2017. Complete genome sequences of six copper-resistant Xanthomonas strains causing

bacterial spot of Solanaceous plants, belonging to X. gardneri, X. euvesicatoria, and X.

vesicatoria, using long-read technology. Genome Announc. 5: e01693-16.

Ritchie, D., and Dittapongpitch, V. 1991. Copper-and streptomycin-resistant strains and host

differentiated races of Xanthomonas campestris pv. vesicatoria in North Carolina. Plant

Dis. 75:733-736.

Ryan, R. P., Vorhölter, F. J., Potnis, N., Jones, J. B., Van Sluys, M. A., Bogdanove, A. J., and

Dow, J. M. 2011. Pathogenomics of Xanthomonas: understanding bacterium–plant

interactions. Nature Reviews Microbiology 9:344-355.

Saraçli, S., Doğan, N., and Doğan, İ. 2013. Comparison of hierarchical cluster analysis methods

by cophenetic correlation. Journal of Inequalities and Applications 2013:203.

Shenge, K., Mabagala, R. B., Mortensen, C. N., and Wydra, K. 2014. Resistance of Xanthomonas

campestris pv. vesicatoria isolates from Tanzania to copper and implications for bacterial

spot management. Afr. J. Microbiol. Res. 8:2881-2885.

Springer, B., Kidan, Y. G., Prammananan, T., Ellrott, K., Böttger, E. C., and Sander, P. 2001.

Mechanisms of streptomycin resistance: selection of mutations in the 16S rRNA gene

conferring resistance. Antimicrob. Agents Chemother. 45:2877-2884.

Stall, R. E., Loschke, D. C., and Jones, J. B. 1986. Linkage of copper resistance and avirulence

loci on a self-transmissible plasmid in Xanthomonas campestris pv. vesicatoria. Mol. Plant

Pathol. 76:240-243.

103

Strayer, A. L., Jeyaprakash, A., Minsavage, G. V., Timilsina, S., Vallad, G. E., Jones, J. B., and

Paret, M. L. 2016. A multiplex real-time PCR assay differentiates four Xanthomonas

species associated with bacterial spot of tomato. Plant Dis. 100:1660-1668.

Strayer-Scherer, A., Liao, Y. Y., Young, M., Ritchie, L., Vallad, G. E., Santra, S., Freeman, J. H.,

Clark, D., Jones, J. B., and Paret, M. L. 2017. Advanced copper composites against

copper-tolerant Xanthomonas perforans and tomato bacterial spot. Phytopathology

108:196-205.

Thayer, P. L., and Stall, R. E. 1961. A survey of Xanthomonas vesicatoria resistance to

streptomycin. Proc. Fla. State Hort. Soc. 75:163-165.

Timilsina, S., Jibrin, M. O., Potnis, N., Minsavage, G. V., Kebede, M., Schwartz, A., Bart, R.,

Staskawicz, B., Boyer, C., Vallad, G. E., Pruvost, O., Jones, J. B., and Goss, E. M. 2015.

Multilocus sequence analysis of Xanthomonads causing bacterial spot of tomato and

pepper plants reveals strains generated by recombination among species and recent global

spread of Xanthomonas gardneri. Appl. Environ. Microbiol. 81:1520-1529.

Timilsina, S., Abrahamian, P., Potnis, N., Minsavage, G. V., White, F. F., Staskawicz, B. J., Jones,

J. B., Vallad, G. E., and Goss, E.M. 2016. Analysis of sequenced genomes of

Xanthomonas perforans identifies candidate targets for resistance breeding in tomato.

Phytopathology 106:1097-1104.

Vallad, G.E., S. Timilsina, H. Adkison, N. Potnis, G. Minsavage, J. Jones, and E. Goss. 2013. A

recent survey of Xanthomonads causing bacterial spot of tomato in Florida provides

insights into management strategies. Proceedings from the 2013 Florida Tomato Institute,

Naples, FL.

104

Voloudakis, A. E., Bender, C. L., and Cooksey, D. A. 1993. Similarity between copper resistance

genes from Xanthomonas campestris and Pseudomonas syringae. Appl. Environ.

Microbiol. 59:1627-1634.

105

CHAPTER 3: DETECTION AND VALIDATION OF QTL CONTROLLING

BACTERIAL SPOT DISEASE RESISTANCE RACE T4 IN AN INTRA-SPECIFIC

TOMATO POPULATION

ABSTRACT

Bacterial spot of tomato is a serious disease caused by at least four species and four races of Xanthomonas- X. euvesicatoria (race T1), X. vesicatoria (race T2), X. perforans (race T3 and

T4), and X. gardneri, with X. perforans race T4 being predominant in North Carolina. However, practical management of this disease is challenging because of the lack of effective chemicals and commercial resistant cultivar. Identification of genetic resistance is the first step to develop the disease resistant variety. The objective of this study was to identify quantitative trait loci

(QTL) conferring resistance to race T4 in an F2:6 intra-specific population developed by crossing

NC 30P x NC22L-1(2008). Five major QTL located on chromosome 1, 4, 6, 11, and 12 and a minor QTL on chromosome 2 were identified. The QTL on chromosomes 1, 2, 4, and 6 were also validated in an independent inter-specific population developed from the crossing of NC

1CELBR x PI 270443. The QTL on chromosome 6 explained the highest percentage of phenotypic variance (up to 26%) followed by the QTL on chromosome 1 (up to 23%) and the

QTL on chromosome 4 (up to 15%). The donor of the resistance associated with these QTL is a released superior breeding line. Therefore both the donor parent and the QTL information will be useful in tomato breeding program involved in bacterial spot disease resistance.

106

INTRODUCTION

Bacterial spot is one of the major foliar diseases of tomato in North Carolina (NC) and many other tomato-growing regions worldwide. The disease has potential to cause up to 66% of tomato yield loss depending on the growth stage of infection (Pohronezny and Volin 1983). At least four species within the genus Xanthomonas- X. vesicatoria, X. euvesicatoria, X. perforans, and X. gardneri are reported as causal agents of the bacterial spot (Jones et al. 2004). In addition, four races (T1, T2, T3, and T4) associated with differential hosts have been reported (Jones et al.

2004). Practical management of bacterial spot is challenging in commercial production fields due to the limited efficacy of current disease management strategies. Bacterial disease management is based on the integrated use of cultural practices and the application of chemicals such as copper, antibiotics (streptomycin) and plant activators (Potnis et al. 2015). However, copper and streptomycin resistant Xanthomonas isolates have already been widely reported in Florida,

Tennessee, Ohio, Canada, Brazil, Ethiopia, and Tanzania (Araújo et al. 2012; Miller and Mera

2011; Mixon 2012; Vallad et al. 2013; Shenge et al. 2014; Abbasi et al. 2015). We also documented over 95% of bacterial spot strains resistant to copper and up to 44% resistant to streptomycin in NC (Chapter 2).

Although breeding for bacterial spot disease resistance started since 1982, there are no commercial resistant tomato cultivars available (Scott et al. 2015). Breeding for host resistance against a bacterial spot of tomato has been challenging, often impeded by the evolution of new races of the pathogen overcoming the identified resistant germplasm, multigenic control of the resistance, and a low correlation between seedling assays and field resistance (Hutton et al.

2010b). New races of bacterial spot have evolved without any deployment of commercial resistant cultivar, suggesting the independent evolution of these races from the selection pressure due to host resistance. For example, X. euvesicatoria was the only species, present as race 1, in

107

Florida until 1991 before the X. perforans race T3 was reported, (Timilsina et al., 2015). X. perforans race T4 strain emerged in 1998 (Astua-Monge et al. 2000; Minsavage et al. 2003) and thereafter has been detected in higher number in field surveys in Florida. In a recent survey, only

X. perforans race T4 were detected in the samples collected from different parts of Florida, indicating a major shift in the race within X. perforans strains from previous surveys, where T4 and T3 race were in the ratio of 3:1 (Horvath et al. 2012). We conducted a systematic survey for the first time in NC and identified single species X. perforans to be a causal pathogen for bacterial spot disease in tomato fields, within which ~91% were race T4 strains and 9% were race T3 strains (Chapter 2). This suggested that the best bacterial spot management practices in tomato in NC should be implemented with major focus on introducing host resistance against race T4.

The hypersensitive response to race T4 has been identified in S. pennellii LA716 conferred by the locus RXopJ4/Xv4 (Austua-Monge et al. 2000; Sharlach et al. 2013).

Unfortunately, RxopJ4 locus in LA 716 is associated with low fruit yield, small fruit, and autogenous leaf necrosis, therefore not suitable for the tomato breeding program (Sharlach et al.

2013). The non-hypersensitive response against race T4 has been detected in PI 114490, HI

7998, PI 126932, and PI 128216. These lines were used to develop three advanced breeding lines

8233, Fla 8517, and Fla 8326 with moderate to high level of race T4 resistance (Hutton et al.

2010a). However, the T4 resistance in these lines was found to be mostly dominant with epistatic effect controlled by multiple loci with moderate effects, which is limiting the development of resistant cultivar from these resistant advanced breeding lines (Hutton et al. 2010a).

The complex nature of the disease and the lack of effective bactericides and resistant cultivar necessitate the identification of more resistant tomato genotypes to bacterial spot disease and study their underlying genetic architecture. After screening sixty-three tomato lines for T4

108

race resistance in both field and greenhouse in NC, T4 resistance has been identified in several S. pimpinellifolium L3707 derived breeding lines including 74L-1W (2008), NC2CELBR, 081-12-

1X-gsms, NC22L-1(2008), and 52LB1(98) (Bhattarai et al. 2017). Among these, we selected NC

22L-1(2008) and crossed with released breeding line NC 30P to create a recombinant inbred lines (RIL) mapping population NC 10204. The use of elite breeding materials in QTL mapping study will facilitate the detection of QTL of direct relevance to breeders (Würschum 2012). In addition, use of elite breeding materials will allow detection of the minor allele effects, as most large effect alleles are homozygous in the population derived from closely related parents

(Rodriguez et al. 2013).

The present study performed QTL mapping for bacterial spot disease resistance to race

T4 in an intra-specific bi-parental fresh market tomato NC 10204 population. The main objective of this research was to identify QTL for bacterial spot disease resistance against race T4 within the intra-specific recombinant inbred line (RIL) population of tomato-derived from two elite breeding lines. The intra-specific QTL analysis in our study used the Solanaceae Coordinated

Agricultural Project (SolCAP) 7720 SNP array (Hamilton et al. 2012; Sim et al. 2012). We also validated the detected QTL in an independent segregating population of tomato.

MATERIALS AND METHODS

Population Development and Experimental Design

Two mapping populations- NC 10204 and NC 13666 were developed to identify and validate the QTL associated with the bacterial spot disease race T4 resistance respectively. The

F1 hybrid of NC 10204 was created in 2010 by crossing the plum tomato breeding line NC 30P and the grape tomato breeding line NC 22L-1(2008). NC 30P was initially selected as a susceptible parent that belongs to cultivated species and has already been released as a breeding

109

line for its superior horticultural traits from North Carolina State University tomato breeding program (Gardner and Panthee 2010a). NC 22L-1(2008) is an advanced breeding line, which was derived from S. pimpinellifolium L3707. NC 22L-1(2008) has shown partial resistance to the race T4 in the previous field study (Bhattarai et al. 2017). The F1 hybrid was self-pollinated to create a segregating F2 population. The seed of F2 plants were individually extracted to create

F2:3 families. The single seed descent method was used to create F5:6 families at the Mountain

Horticultural Crops Research and Extension Center (MHCREC) in Mills River, NC. Although the mapping population was started with 284 F2 individuals, it was reduced to 121 lines in F4:5 generation and 117 lines in F5:6 generation due to poor seed germination at various generation.

The F1 hybrid of NC 13666 was created in 2013 by crossing tomato breeding line NC

1CELBR and S. pimpinellifolium accession PI 270443. NC 1CELBR is a susceptible parent belonging to cultivated species, which was released as a breeding line for its superior horticultural traits and resistance to early blight and late blight diseases (Gardner and Panthee

2010b). PI 270443 has shown resistance to bacterial spot disease in preliminary disease screening experiments conducted in phytotron and greenhouse. The seed of F2 plants were individually extracted to create F2:3 families. The single seed descent method was used to create

F6:7 families at the MHCREC in Mills River, NC.

For QTL mapping study, a total of 121 F4:5 and 117 F5:6 NC 10204 population along with parents, F1 hybrid, resistant and susceptible controls were evaluated during 2016 and 2017 field seasons respectively in two locations: MHCREC, Mills River, NC; and Piedmont Research

Station (PRS), Salisbury, NC. Similarly, for QTL validation study, a total of 210 F6:7 RIL population of NC 13666 was evaluated for the bacterial disease severity in two locations

MHCREC and PRS in 2017, along with the parents, F1 hybrid, resistant and susceptible controls.

Therefore, NC 10204 was evaluated in four environments, i.e., each combination of location and

110

year, whereas, NC 13666 was evaluated in two environments. Seeds were germinated in 72 cell trays (56 x 28 cm2) in potting mix and grown for six weeks before transplantation in the field. In each trial or environment tested, both populations were planted in a random complete block design with two replications with one six-plant plot per line per replication. Individual plants were grown 45 cm apart within rows, 150 cm apart between rows, in raised bed covered with plastic mulch with drip irrigation. Plants were hand strung and sprayed according to the recommended schedule for fungicides and insecticides (Ivors 2010). No bacterial disease control measures were taken.

Inoculum Preparation, Inoculation, and Disease Evaluation

The inoculum was produced by growing a virulent strain of X. perforans race T4

(#Isolate9) on Yeast extract-dextros-CaCO3 (YDC) media for 24–36 h at 27oC an incubator. The cultured bacteria were washed and suspended in distilled water. The bacterial suspension was

8 standardized to A600= 0.3 (corresponds to a concentration of 5 X 10 CFU/ml). The inoculum was applied at this concentration to susceptible lines NC84713 and Bonny Best in the greenhouse by misting the foliage until runoff with a backpack sprayer one week before field planting. The high humidity was maintained in the greenhouse by misting the foliage and covering the plants with plastic for 48 hours post inoculation and three days before inoculation.

Inoculated plants were allowed to develop the serious disease for a week and then introduced to the fields. Inoculated NC84713 and Bonny Best were introduced in the field to spread the inoculum in 2016 and 2017 respectively. Each inoculated plant was planted at both ends of an individual plot. Plants were rated for disease severity in the field based on the average rating of the plot, using the Horsfall and Barratt scale (1945) with slight modification, where 0 = 0%, 1 = less than 1%, 2 = 1 –3%, 3 = 3–6%, 4 = 6–12%, 5 = 12–25%, 6 = 25–50%, 7 = 50–75%, 8 = 75–

111

87%, 9 = 87–94%, 10 = 94–97%, 11 = 97–100%, and 12 = 100% diseased tissue. The disease was scored five times on a weekly basis starting from 20 days after planting (DAP), and the

AUDPC value was calculated as (Simko and Piepho 2011):

>?3 +,-./ + +,-./ !"#$% = ( )* 0 0235 ∗ (89:/ − 89:/ )= 2 023 0 0@3

th th where, Scorei is a disease score at the i observation, timei is time in days at the i observation, and n is the total number of observations.

Statistical Analysis

Data analysis was carried out in SAS (version 9.4; SAS Institute, Cary, NC). The summary statistics and normal probability plots were calculated using the UNIVARIATE procedure of SAS. Correlation analysis was performed between different environments using

PROC CORR procedure in SAS. The heritability was estimated for each environment by calculating variance components using ‘ASYCOV’ function in PROC MIXED in SAS. Analysis of variance (ANOVA) was performed using the MIXED procedure of SAS for AUDPC data and individual week disease data to determine the differences among genotypes. The least square means (LS Means) of each genotype within each environment were calculated using the same model as that of ANOVA and used for QTL analysis.

For genotype in each location in each year, the following model was run for each individual week disease response and AUDPC value.

Response = μ+ genotype + replication + genotype*replication + error

112

DNA Extraction and Genotyping

The genomic DNA of NC 10204 (192 lines) and NC 13666 population (232 lines) were extracted at F2 and F6 generation respectively. A modified cetyltrimethyl ammonium bromide method was used for DNA extraction and the DNA samples were stored at -20°C in 10 mM

Tris– HCl pH 8.0, and 1 mM EDTA (Kabelka et al. 2002). The genomic DNA of parents of both populations were also extracted. All the extracted DNA were quantified using a NanoDrop 2000

Spectrophotometer (Thermo Scientific, Wilmington, Delaware, USA). The genomic DNA of 192

F2 lines of NC 10204 and 232 lines of NC 13666 along with the four parental lines were genotyped using the SolCAP Illumina Infinium Assay. SNP genotypes were determined using

GenomeStudio version 1.0 (Illumina Inc, San Diego, CA, USA).

Linkage Map Construction

A total of 110 individuals of NC 10204 (either genotypic or phenotypic data were missed) and 210 individuals of NC 13666 were used for the construction of linkage map. Only polymorphic makers were used to construct a final map. The linkage map of NC 10024 was constructed using Joinmap 4.0 (Van Ooijen 2006), whereas NC 13666 linkage map was constructed using R package ‘R/qtl’ (Broman et al. 2003) to check if we can get the consistent

QTL despite different software. The segregation of each marker for the goodness of fit between observed and the expected Mendelian ratio was tested by χ2-test and markers with the distorted segregation at p <0.0001 were discarded. The SNP markers were grouped into their respective chromosome according to tomato genome information SL3.0 available in the Sol Genomics

Network (SGN) (Fernandez-Pozo et al. 2015). The marker orders within each chromosome were calculated using regression mapping algorithm (Stam 1993). Kosambi mapping function was used for estimation of map distances (cM) (Kosambi 1943).

113

QTL Analysis

Windows QTL Cartographer v 2.5 (Wang et al. 2012) software was used for QTL analysis. The Composite Interval Mapping (CIM) method was used with the default parameters

(model 6). A backward regression was used to perform the CIM analysis to enter or remove background markers from the model. The walking speed was set at one cM for the detection of

QTL. A default threshold of logarithm of odds ratio (LOD) score of 2.5 was used to declare the presence of QTL. The additive effect and the proportion of the phenotypic variation (R2) for each

QTL were also obtained using this software. The dominant effect was not reported as individuals used were from RIL population that was mostly in homozygous condition. QTL explaining more than 10% of the phenotypic variance were considered as major QTL. Considering bi-parental mapping population and size of the mapping of the population used in this study, any QTL within 10 cM distance on the same chromosomes were regarded as a single QTL, and if detected in at least two environments considered as a consistent QTL (Pascual et al. 2015).

RESULTS

Phenotypic Variation

The disease data were recorded from four environments for NC 10204 and two environments for NC 13666. The summary statistics of the disease data in NC 10204 and NC

13666, both AUDPC and individual week data for each environment are presented in Table 3.1 and Table 3.2 respectively. In NC 10204 population, the highest mean AUPDC (201) was observed in PRS at Salisbury, NC in 2016. In both years, disease severity was higher in PRS.

The disease severity in MHCREC at Mills River, NC was higher in 2017 compared to 2016, whereas more disease was observed in 2016 in PRS compared to 2017. The disease severity was observed in increasing trend in every week except in PRS in 2017, where disease severity tended

114

to decrease after the third week (Table 3.1). The genotypic effect was significant for the disease severity data at p<0.05 except for the disease data observed in the first three weeks and AUDPC in MHCREC in 2016, and in the second week in PRS in 2017 (Table 3.1). The heritability estimates of disease data were low to moderate in NC 10204 population that ranged from 0.1 to

0.5 (Table 3.1). The AUDPC values were significantly correlated at p<0.05 between years in both locations (MHCREC and PRS), but no correlation was observed between a location within the same year (Table 3.2). In both locations, disease severity data scored in the first and second week were not significantly correlated between years, but the correlation was observed after the second week except in the fourth week in PRS. Although NC-22L-1(2008) was reported to have partial resistance in the field, it has similar or more disease level compared to NC 30P. (Figure

3.1). Instead, in three environments, NC 30P appeared to be more resistant than NC-22L-

1(2008). However, transgressive segregants with a higher level of BS resistance and lower level of resistance than both parents were also observed resulting in a continuous distribution (Figure

3.1). This indicates multiple genes are involved in the disease resistance.

In NC 13666, the highest AUDPC means (172) was observed in MHCREC. There was a significant difference (p < 0.05) among genotypes of RIL population of NC 10204 for bacterial spot resistance evaluated in each environment (Table 3.3). The heritability estimates of disease data were low to moderate in NC 13666 population that ranged from 0.2 to 0.6 (Table 3.3). The significant correlations (p<0.0001) were observed between the two locations’ disease data for individual week and AUDPC. As expected, PI 270443 was more resistant to NC 1CELBR

(Figure 3.2). The transgressive segregants with higher resistance and lower resistance than both parents were also observed resulting in a continuous distribution (Figure 3.2). This confirms the involvement of multiple genes in the disease resistance.

115

Table 3.1: Descriptive statistics (mean, minimum, maximum, and standard deviation) for bacterial spot disease data of NC 10204, heritability (H2), and analysis of variance (ANOVA) for two years (2016 and 2017), and two locations (MHCREC and PRS). Variables used to measure the bacterial spot disease severity. SCORE 1-5 represent disease severity measured according to Horsefall- Barratt (1945) scale for week 1-5 respectively. Y1L1 indicates the MHCREC location and year 2016; Y1L2 indicates PRS location and year 2016; Y2L1 indicates the MHCREC location and year 2017; Y2L2 indicates PRS location and year 2017.

Descriptive Statistics Effect of the fixed variable (ANOVA) Variables Std. H2 Error F Mean Min. Max. Source DF MS Pr > F Dev. DF Value AUDPC_Y1L1 134.7 20.7 81 178 0.1 Genotype 117 456.51 105 1.2 0.2012 AUDPC_Y1L2 200.8 15.3 169 262 0.5 Genotype 104 349.02 97 3 <.0001 AUDPC_Y2L1 149 10.8 112 189 0.5 Genotype 115 178.76 115 3.3 <.0001 AUDPC_Y2L2 186.5 12.9 147 238 0.4 Genotype 116 224.9 107 2.6 <.0001

SCORE1_Y1L1 3.3 1.4 1 6 0 Genotype 117 2.01 105 1 0.5367 SCORE1_Y1L2 6.8 0.7 4 8 0.2 Genotype 104 0.72 97 2.2 <.0001 SCORE1_Y2L1 3.9 0.5 2 5 0.2 Genotype 115 0.28 115 1.5 0.0188 SCORE1_Y2L2 4.7 0.8 2 6 0.2 Genotype 116 0.8 107 1.5 0.0227

SCORE2_Y1L1 4 1.2 1 6 0 Genotype 117 1.31 105 1 0.5587 SCORE2_Y1L2 6.8 0.6 4 8 0.2 Genotype 104 0.41 97 1.6 0.0102 SCORE2_Y2L1 4.6 0.6 3 6 0.2 Genotype 115 0.48 115 2 0.0002 SCORE2_Y2L2 6.1 0.7 4 8 0.1 Genotype 116 0.52 107 1.2 0.1798

SCORE3_Y1L1 4.8 0.9 2 7 0.1 Genotype 117 0.85 105 1.3 0.0918 SCORE3_Y1L2 7 0.8 6 10 0.3 Genotype 104 0.89 97 1.9 0.0008 SCORE3_Y2L1 5 0.6 3 6 0.2 Genotype 115 0.4 115 1.7 0.002 SCORE3_Y2L2 7.6 0.8 5 10 0.3 Genotype 116 0.77 107 2.1 <.0001

116

Table 3.1 (continued).

SCORE4_Y1L1 5.9 0.8 4 7.5 0.2 Genotype 117 0.72 105 1.4 0.0312 SCORE4_Y1L2 7.5 0.9 6 11 0.5 Genotype 104 1.12 97 2.9 <.0001 SCORE4_Y2L1 6.1 0.9 4 9 0 Genotype 115 0.76 115 1.7 0.0017 SCORE4_Y2L2 7.1 0.6 5 10 0.5 Genotype 116 0.57 107 3.1 <.0001

SCORE5_Y1L1 6.3 0.7 5 8 0.3 Genotype 117 0.54 105 1.8 0.0016 SCORE5_Y1L2 7.6 0.9 6 11 0.5 Genotype 104 1.14 97 2.9 <.0001 SCORE5_Y2L1 7.1 0.5 6 9 0.4 Genotype 115 0.36 115 2.3 <.0001 SCORE5_Y2L2 7 0.5 6 9 0.2 Genotype 116 0.36 107 1.6 0.0109

117

Figure 3.1: Distribution of Area Under Disease Progress Curve (AUDPC) for bacterial spot disease severity among recombinant inbred lines (RIL) population of NC10204 across two locations (Mountain Horticultural Crops Research & Extension Center at Mills River- MHCREC, and Piedmont Research Station at Salisbury- PRS) and years- MHCREC-2016 (A), PRS-2016 (B), MHCREC-2017 (C), and PRS-2017 (D). Arrows indicate the AUDPC values for the parents and F1 hybrid. Y1L1 indicates the MHREC location and year 2016; Y1L2 indicates PRS location and year 2016; Y2L1 indicates the MHREC location and year 2017; Y2L2 indicates PRS location and year 2017.

118

Table 3.2: The correlation coefficients of the disease response between two generations and two locations in the intra-specific population NC 10204.

Variables Correlation (r) between years2016 Correlation (r) between and 2017 locations MHCREC and PRS Location=MHCREC Location=PRS Year=2016 Year=2017

AUDPC 0.3**a 0.21* ns ns

SCORE1 nsb ns ns ns

SCORE2 ns ns ns 0.3**

SCORE3 0.3** 0.3** ns ns

SCORE4 0.2* ns 0.2* ns

SCORE5 0.3** 0.4**** 0.4**** ns

Note: a ‘****’ denotes p - value < 0.0001, ‘**’ denotes p - value < 0.01, ‘*’ denotes p - value < 0.05, and b ns represents non-significant correlations.

Figure 3.2: Distribution of AUDPC for bacterial spot disease severity among RIL population of NC13666 across two locations- MHCREC (a), and PRS (b) in 2017. Arrows indicate the AUDPC values for the parents and F1 hybrid. L1 indicates MHCREC location, and L2 indicates PRS location.

119

Table 3.3: Descriptive statistics (mean, minimum, maximum, and standard deviation) for bacterial spot disease data of NC 13666, heritability (H2), and type 3 analysis of variance of fixed effect for two locations (MHCREC and PRS) in the year 2017. Variables used to measure the bacterial spot disease severity. SCORE 1-5 represent disease severity measured according to Horsefall-Baratt (1945) scale for week 1-5 respectively. L1 indicates MHREC location; L2 indicates PRS location.

Variable Descriptive Statistic Effect of fixed variable (ANOVA) H2 Mean Std. Dev Min. Max. Source DF MS Error F Pr > F DF Value AUDPCL1 172.3 19.93104 80.5 217 0.5 Genotype 208 602.91 208 3.28 <.0001

AUDPCL2 162.0 14.91571 119 213.5 0.6 Genotype 206 336.87 180 3.76 <.0001

SCORE1L1 3.7 0.66626 2 8 0.3 Genotype 208 0.56 208 1.76 <.0001

SCORE1L2 3.5 0.83036 1 6 0.2 Genotype 206 0.81 180 1.48 0.0038

SCORE2L1 3.8 0.73764 1 6 0.3 Genotype 208 0.72 208 1.97 <.0001

SCORE2L2 4.9 0.8093 2 8 0.3 Genotype 206 0.83 180 1.95 <.0001

SCORE3L1 4.4 0.74272 1 6 0.3 Genotype 208 0.71 208 1.9 <.0001

SCORE3L2 6.4 0.71137 4 8 0.4 Genotype 206 0.68 180 2.34 <.0001

SCORE4L1 4.9 0.85038 1 7 0.5 Genotype 208 1.06 208 2.96 <.0001

SCORE4L2 6.8 0.7421 5 9 0.5 Genotype 206 0.80 180 3.01 <.0001

SCORE5L1 6.1 0.76242 4 8 0.4 Genotype 208 0.81 208 2.29 <.0001

SCORE5L2 6.6 0.66755 5 10 0.5 Genotype 206 0.64 180 2.89 <.0001

120

Genetic Linkage Map

Out of 7789 total SNP molecular markers that were used for genotyping, only 886 and

1699 polymorphic SNP markers were included in the final linkage map construction of NC

10204 and NC 13666 respectively (Table 3.4). The final linkage map included 12 linkage groups and spanned a total of 739.50 cM in the genetic distance with an average interval of 0.83 cM among 886 SNP polymorphic markers in NC 10204 linkage map (Table 3.4). The NC 13666 linkage map spanned a total of 1395 cM in the genetic distance with an average interval of 1.2 cM among 1699 SNP polymorphic markers (Table 3.4). In both maps, chromosome 1 had the longest length. The chromosome 8 had the shortest length in the NC 10204 linkage map, while chromosome 10 had the shortest length in the NC 13666 linkage map. The highest number of markers was observed in chromosome 4 (212) in NC 10204 linkage map, while in chromosome

11 (526) in NC 13666 linkage map (Table 3.4).

Table 3.4: Summary of the linkage maps of NC 10204 and NC13666 showing 12 chromosomes along with the number of markers per chromosome and the length of each chromosome.

NC 10204 NC 13666 Chromosome Markers Length (cM) Markers Length (cM) 1 58 98.6 99 373.5 2 101 77.4 68 97.4 3 60 52.1 205 92.2 4 212 85.3 348 101.5 5 30 42.6 68 93.1 6 60 66.8 68 167.8 7 36 39.6 75 68.1 8 15 28.8 44 74.4 9 138 94.1 43 102.7 10 31 72.1 101 53.5 11 112 30.1 526 78 12 33 51.9 54 92.7 Total 886 739.4 1699 1394.9

121

QTL Analysis

Five QTL on chromosome 1 (q1-bs1), 4 (q-bs4), 6 (q1-bs6), 11 (q1-bs11), and 12 (q1- bs12) and a minor QTL on chromosome 2 (q1-bs2) were detected in at least two environments in

NC 10204 population based on composite interval mapping (Table 3.6). The q1-bs6 with LOD value of 3-4 (Figure 3.3) explained the highest percentage of phenotypic variance (up to 26%) among the detected QTL, followed by the q1-bs1 (LOD=3) that explained up to 23% of the total phenotypic variance. The QTL region on chromosome 1 was detected at different positions within 45cM regions in two environments in the genetic map, and the positions of the closest markers associated with the QTL were within ~12Mb in the physical map of tomato. The resistance associated with the q1-bs1 is contributed by different parents in different environments

(Table 3.6; Figure 3.3). Since the QTL were detected on different positions of chromosome 1 and contributed by different parents, the QTL detected on chromosome 1 in different environments might represent two different QTL. The q1-bs4 (LOD= 3-4), q1-bs12 (LOD= 3), and q1-bs11

(LOD=3-4) explained up to 15%, 14%, and 11% of total phenotypic variance respectively in NC

10204 (Table 3.6). The QTL on chromosome 2 were detected in three environments- the QTL detected in PRS-2016 and MHCREC-2016 were within 10cM (40cM-48cM) region in the genetic map, whereas, QTL detected in MHCREC-2017 was at the 65cM region. However, the

QTL on chromosome 2 detected in three environments were located within ~8Mb in the physical map. The alleles from NC 22L-1(2008) increased the disease severity for all detected QTL except q1-bs4 in NC 10204 in which the allele from NC 30P increased the disease severity. This suggests, unlike our expectations, NC 30P is the donor of the resistance associated with the genomic regions on chromosomes 1, 6, 11, and 12. The linkage map position of the detected

QTL and their corresponding position in the physical map are presented in Table 3.5. The two

QTL on chromosome 9 (LOD=3) and chromosome 10 (LOD=3) explaining 14% and 15% of

122

total phenotypic variances were detected in only one environment, hence could not be treated as consistent QTL.

Chromosome 6 Chromosome 6 PRS-2016 MHCREC-2017

Chromosome 6 Chromosome 4 PRS-2017 MHCREC 2016

Chromosome 4 Chromosome 1 PRS-2016 PRS-2016

Chromosome 1 PRS-2017

Figure 3.3: The logarithm of odds (LOD) graph for the major QTL detected on chromosome 6, 4, and 1in NC 10204 that are validated in NC 13666 population. The peak represents the QTL. Horizontal axis represents the position of QTL (cM) on the chromomse and vertical axis represents the LOD values.

123

Table 3.5: QTL associated with bacterial spot disease caused by Xanthomonas perforans race T4 resistance identified in an intra- specific mapping population NC10204 of tomato derived from NC 30P x NC 22L-1(2008) in different environment conditions. The table presents the position of the QTL flanking SNP molecular markers, its physical position, LOD score value and phenotypic variation explained by each QTL (R2-value), and additive effect. The negative additive effect indicates the allele from NC-22L-1 (2008) increased the disease severity and positive additive effect indicates the allele from NC 30P increased the disease severity.

Chr. Marker Genetic Map Physical Map Position LOD Additive R2 Position (cM) (bp) (%) PRS-2017 1 solcap_snp_sl_12352 & 49.58 72658153 to 76783695 3 -0.432 23 solcap_snp_sl_21280 PRS-2016 1 solcap_snp_sl_48181 & 95.21 88401832 to 87193106 3 0.395 17 solcap_snp_sl_54783 PRS-2016 2 solcap_snp_sl_29627 & 40.42 37554824 to 37663569 4 -0.182 4 solcap_snp_sl_29658 MHCREC-2017 2 solcap_snp_sl_42441 & 65.05 43574561 to 44069445 4 -0.093 2 solcap_snp_sl_42324 MHCREC-2016 2 solcap_snp_sl_49480 48.71 39278008 3 -0.168 2 PRS-2016 4 solcap_snp_sl_3747 & 72.91 60540713 to 60552797 3 0.496 15 solcap_snp_sl_47165 MHCREC-2016 4 CL009271-0329 84.96 63672498 4 0.522 14 PRS-2016 6 solcap_snp_sl_12765 & Bcyc_868 11.83 36998358 to 42288756 4 -9.185 26 PRS-2017 6 solcap_snp_sl_12765 & Bcyc_868 11.83 36998358 to 42288756 3 -0.35 22 MHCREC-2017 6 Bcyc_868 & solcap_snp_sl_57352 14.33 42288756 to 42343473 4 -0.274 21 MHCREC-2017 11 solcap_snp_sl_9507 & 31.71 5354992 to 5226970 4 -0.256 11 solcap_snp_sl_62748 PRS-2016 11 solcap_snp_sl_724 & 23.13 10018811 to 10015478 3 -0.185 3 solcap_snp_sl_732 PRS-2017 12 solcap_snp_sl_14415 31.06 61833712 3 0.38 14 MHCREC-2017 12 solcap_snp_sl_24755 & 45.48 7801435 to 64210355 3 0.18 9 solcap_snp_sl_31628

124

QTL Validation

To validate the QTL detected in NC10204 population, the QTL analysis was performed in an independent RIL population NC 13666. The disease data were scored for only one year with two environments. The QTL detected in both locations were treated as consistent QTL.

Among six QTL detected in mapping population NC 10204, four QTL on chromosomes 1 (q2- bs1), 4 (q2-bs4), and 6 (q2-bs6) including a minor QTL on chromosome 2 (q2-bs2) were detected in NC 13666 too with LOD values ranging from 3-15 and phenotypic variance ranging from 6 to 23% (Table 3.6). The q2-bs4 (LOD=15) and q2-bs6 (LOD= 10) explained the highest percentage of phenotypic variance of 23%. The q2-bs2 was closer to the q1-bs2 detected in

MHCREC-2017 (Table 3.6; Table 3.7). An additional QTL on chromosome 3 (q2-bs3) (LOD=

4-9) was also detected in NC 13666 population, which explained 6-16% of phenotypic variance.

The donor of the resistance loci on chromosomes 1, 2, 3, and 6 was the resistant parent PI

270443, whereas the susceptible parent NC 1CELBR contributed resistance locus on chromosome 4 (Table 3.6). Consistent with the result from NC 10204, the genomic regions associated with the resistance on chromosome 1 were detected quite far (>100cM) in two environments, but the markers associated with these regions were within 8Mb in the physical map. The genetic map position of the detected QTL along with their corresponding physical map positions are presented in Table 3.6. The comparisons of physical map positions for the QTL detected in NC 10204, and NC 13666 are presented in Table 3.7. The QTL positions on chromosomes 1, 4, 2, and 6 were either overlapped or slightly shifted in two populations (Table

3.7).

125

Table 3.6: QTL associated with bacterial spot disease caused by Xanthomonas perforans race T4 resistance identified in an inter- specific validation population NC13666 of tomato derived from NC 1CELBR x PI 270443 in different environment conditions. The table presents the position of the QTL flanking SNP molecular markers, its physical position, LOD score value and phenotypic variation explained by each QTL (R2-value), and additive effect. The negative additive effect indicates the allele from NC 1CELBR increased the disease severity and positive additive effect indicates the allele from PI 270443 increased the disease severity.

Environment Chr. Marker Genetic Map Physical Map Position LOD Additive R2 Position(cM) (bp) (%) MHCREC-2017 1 solcap_snp_sl_44461 & 360.57 82404532 to 77935539 3 -3.79 6 solcap_snp_sl_43632 PRS-2017 1 solcap_snp_sl_16393 & 211.19 75563097 to 74282365 5 -4.06 9 solcap_snp_sl_1813 MHCREC-2017 2 solcap_snp_sl_36264 & 12.36 45991713 to 44433274 4 -0.18 6 solcap_snp_sl_29920 PRS-2017 2 solcap_snp_sl_36259 & 11.08 46228345 to 45971768 4 -0.18 7 solcap_snp_sl_36265 MHCREC-2017 3 solcap_snp_sl_35650 & 48.24 53874338 to 52856201 4 -0.19 6 solcap_snp_sl_10372 PRS-2017 3 solcap_snp_sl_18952 & 58.18 46842858 to 46651505 9 -0.25 16 CL016221-0325 MHCREC-2017 4 solcap_snp_sl_3104 & 59.27 55101393 to 55514910 15 7.08 23 solcap_snp_sl_2172 PRS-2017 4 solcap_snp_sl_43581 & 63.22 55839329 to 57367706 7 4.61 11 CL017373-0593 MHCREC-2017 6 solcap_snp_sl_19852 & 12.72 43667860 to 41150770 6 -4.95 12 solcap_snp_sl_24450 PRS-2017 6 solcap_snp_sl_19852 & 7.72 43667860 to 41150770 6 -4.74 12 solcap_snp_sl_24450 MHCREC-2017 6 solcap_snp_sl_19852 & 13.72 43667860 to 41150770 10 -0.31 23 solcap_snp_sl_24450 PRS-2017 6 solcap_snp_sl_57594 & 30.2 41150984 to 38384375 7 -0.26 13 solcap_snp_sl_27215

126

Table 3.7: Comparison of physical map distances of the detected QTL associated with bacterial spot resistance in mapping population (NC 10204) and validation population (NC 13666) of tomato.

Physical Map Position (bp) Chromosomes NC 10204 NC 13 666

1 72,658,153 to 88,401,832 (~15Mb) 74,282,365 to 82,404,532 (~8Mb)

2 37554824 to 44069445 (~6Mb) 44433274 to 46228345 (~1.8Mb)

4 60,540,713 to 63,67,2498 (~3Mb) 55,101,393 to 57,367,706 (~2Mb)

6 36,998,358 to 42,288,756 (~5Mb) 41,150,770 to 43,667,860 (~2.5Mb)

DISCUSSION

Bacterial spot disease has been a significant problem in tomato production regions for more than 60 years. However, neither chemical control nor the breeding for the host resistance has achieved success. The new races of the pathogen have evolved overtime, and the distribution of the pathogen species has shifted many times. Currently, X. perforans race T4 is a major problem in tomato production regions in Southeast US including NC (Horvath et al. 2012;

Chapter 2). Therefore, we sought to identify QTL and eventually develop molecular markers associated with the bacterial spot race T4 resistance in an intra-specific population, so that they can be utilized in the breeding program without the concern of linkage drag. The parent that we selected as a resistant parent based on previous field and greenhouse studies by Bhattarai et al.

(2016) to develop intra-specific mapping population turned out to develop bacterial spot disease.

This suggests that screening for disease resistance might require multiple years and locations data to identify resistance genotypes as bacterial spot disease resistance is quantitative and

127

heavily influenced by the environment. Unexpectedly, NC 30P, which we selected as a susceptible parent showed more resistance to bacterial spot disease compared to NC 22L-1

(2208). Therefore, it is expected that both parents may contribute to resistance alleles.

The phenotypic data for disease severity was collected for two years in two locations.

The disease pressure was higher in PRS compared to MHCREC. This might be due to different environmental conditions in two locations. MHCREC is cooler than the PRS, and PRS had perfect weather condition for the bacterial spot disease, i.e. warm, humid condition and high rainfall with the optimum temperature of 24oC to 30oC (Gardner and Kendrick 1921). We identified and validated four QTL on chromosome 1, 2, 4, and 6 associated with bacterial spot disease race T4 resistance in two separate populations derived from S. pimpinellifolium. The most alleles associated with the disease resistance were from NC 30P in NC 10204 population, and from PI 270443 in NC 13666 population. This suggests NC 30P could be a used as a source of bacterial spot resistance in the breeding program.

Plum Regal, an F1 hybrid developed from NC 30P x NC 25P has a good level of tolerance to bacterial spot (Gardner and Panthee, 2010a). It was assumed that the level of tolerance is contributed by NC 25P, which is also resistant to late blight. However, evaluation of both parent lines in 2016 and 2017 in a replicated trial indicated that there was a non-significant level of difference between these two parent lines (with a disease score of 5.7 and 6.0 in NC 25P and NC 30P, respectively) indicating that both parents were contributing for the resistance to BS in Plum Regal (Gardner and Panthee, 2010a). With this observation, it is not surprising to have

NC 30P contributing resistance allele in QTL detected from the NC10204 population in the present study.

Previously, the resistance gene RxopJ4 effective against race T4 race has been mapped to a 190- kb segment on the long arm of chromosome 6 between markers J350 and J352 in S.

128

pennellii LA 716 (Sharlach et al. 2013). The genomic position of J350 is 35036517 to 35036818 and J352 is 35241322 to 35241736 (Sharlach et al. 2013). In our study, the q1-bs6 and q2-bs6 were located on chromosome 6 in the genomic region of 36998358 bp to 42288756 bp in NC

10204 (mapping) population, and in the region of 41150770 bp to 43667860 bp in NC 13666

(validating) population respectively, which is close to the location of RxopJ4, although the sources of this resistance were different. Therefore, the QTL detected in our study on chromosome 6 could be allelic to RxopJ4 gene, but further investigation is required to verify it.

The q1-bs6 also explained the highest percentage of phenotypic variance in NC 10204. This

QTL might be useful in the tomato breeding program if further studied as there will be minimal linkage drag unlike in LA 716.

The q1-bs1 explained the second highest phenotypic variance (up to 23%) in NC 10204, but this QTL explained only up to 9% phenotypic variance in NC 13666 population. The genomic position of QTL on chromosome 1 in physical map overlapped for both NC10204 and

NC 13666, thus validating this QTL. In a previous study, a minor QTL was detected on chromosome 1 for race T4 at TOM 202 marker flanked by LEVCOH11 and LEVCOH12 markers in PI 114490 derived IBC population, but could not be confirmed through selective genotypic approach (Hutton et al. 2010b). A preliminary study identified a locus in chromosome

1 against race T4 that was derived from a susceptible parent in Scott et al. (2015) study. As we detected and validated the QTL on chromosome 1, this represents a novel major QTL against race T4.

Another major QTL was detected in both NC 10204 (q1-bs4) and NC 13666 (q2-bs4) population explaining up to 15% and 23% respectively on chromosome 4. No other QTL have been reported on chromosome 4 against race T4 of bacterial spot disease so far in tomato genotypes. The q1-bs4 and q2-bs4 were located in the genomic region of 60,540,713 bp to

129

63,67,2498 bp in NC 10204 and 55,101,393 to 57,367,706 bp in NC 13666 respectively in the physical map, indicating ~8.5Mb position in the physical map. The alleles associated with the disease resistance on chromosome 4 were derived from NC CELBR in NC 13666 population and

NC 22L-1(2008) in NC 10204 population. The QTL detected on chromosome 4 in our study represents a novel QTL and provides useful information for the breeding of tomatoes against bacterial spot disease race T4. The minor QTL on chromosome 2 detected in NC10204 was also validated in NC 13666 with an overlapping position in the physical map. However, Scott et al.

(2015) identified a QTL with strong effect on chromosome 2 against race T4 in PI 114490 derived population in a preliminary study.

Although q1-bs11 and q2-bs3 were limited to NC 10204 and NC 13666 populations respectively in our study, Hutton et al (2010b) identified QTL on chromosome 11(explained

29.4% phenotyping variance) and 3 (explained 4.8% phenotypic variance) against race T4 in PI

114490 derived inbred backcross population and confirmed through selective genotyping in three populations derived from three breeding lines Fla. 8517, Fla. 8326, and Fla. 8517. A preliminary study identified seven robust QTL- one in chromosome 3, and two in each of chromosome 2, 10, and 11; three weak QTLs on chromosome 8, 9, 12 from PI 114490, whereas QTLs on chromosomes 1, 7, and 9 were from susceptible parent (Scott et al. 2015). In our study, the QTL on chromosome 12 was detected only in NC 10204.

In summary, we identified major QTL associated with the bacterial spot disease resistance on chromosome 1, 4, 6 in an intra-specific mapping population. The donor of the resistance associated with the genomic regions on chromosome 6 that explained the highest percentage of phenotypic variance was NC 30P, a released breeding line with superior horticultural traits. Therefore, NC 30P would be a valuable resource for tomato breeding program to further work on bacterial spot disease resistance breeding. NC 30P can be utilized

130

successfully as a parent without much concern about the linkage drag as observed in the case of

LA 716. NC 30P also contributed to the resistance associated the genomic region on chromosome 1, which explained the largest percentage of phenotypic variation after the locus on chromosome 6. Although the resistance associated with the region on chromosome 4 is obtained from NC 22l-1(2008), this can be useful in stacking the resistance.

131

REFERENCES

Abbasi PA, Khabbaz SE, Weselowski B, Zhang L (2015) Occurrence of copper-resistant strains

and a shift in Xanthomonas spp. causing tomato bacterial spot in Ontario. Can J

Microbiol 61:753-761

Araújo ER, Pereira RC, Ferreira MASV, Quezado-Duval AM, Moita AW (2012) Sensitivity of

xanthomonads causing tomato bacterial spot to copper and streptomycin and in vivo

infra-specific competitive ability in Xanthomonas perforans resistant and sensitive to

copper. J Plant Pathol 94:79-87

Astua-Monge G, Minsavage GV, Stall RE, Vallejos CE, Davis MJ, Jones JB (2000) Xv4-

Avrxv4: A new gene-for-gene interaction identified between Xanthomonas campestris

pv. vesicatoria race T3 and the wild tomato relative Lycopersicon pennellii. Mol Plant

Microbe Interact 13:1346-1355

Bhattarai K, Louws FJ, Williamson JD, Panthee DR (2017) Resistance to Xanthomonas

perforans race T4 causing bacterial spot in tomato breeding lines. Plant Pathol 66:1103-

1109

Broman KW, Wu H, Sen Ś, Churchill GA (2003) R/qtl: QTL mapping in experimental crosses.

Bioinformatics 19:889-890

Fernandez-Pozo N, Menda N, Edwards JD, Saha S, Tecle IY, Strickler SR, Bombarely A,

Fisher-York T, Pujar A, Foerster H, Yan A (2015) The Sol Genomics Network (SGN)—

from genotype to phenotype to breeding. Nucleic Acids Res 43:D1036–D1041.

pmid:25428362

Gardner M, Kendrick J (1921) Bacterial spot of tomato. J Agric Research 21:123-156

Gardner RG, Panthee DR (2010a) 'Plum Regal' Fresh-market plum tomato hybrid and its parents,

NC 25P and NC 30P. HortScience 45:824-825

132

Gardner RG, Panthee DR (2010b) NC 1 CELBR and NC 2 CELBR: Early blight and late blight-

resistant fresh market tomato breeding lines. HortScience 45:975-976

Hamilton JP, Sim S, Stoffel K, Van Deynze A, Buell CR, Francis DM (2012) Single nucleotide

polymorphism discovery in cultivated tomato via sequencing by synthesis. Plant Genome

5:17-29

Horvath DM, Stall RE, Jones JB, Pauly MH, Vallad GE, Dahlbeck D, Staskawicz BJ, Scott JW

(2012) Transgenic resistance confers effective ﬁeld level control of bacterial spot disease

in tomato. PLoS ONE 7:e42036.

Hutton SF, Scott JW, Jones JB (2010a) Inheritance of resistance to bacterial spot race T4 from

three tomato breeding lines with differing resistance backgrounds. J Am Soc Hortic Sci

135:150-158

Hutton SF, Scott JW, Yang W, Sim SC, Francis DM, Jones JB (2010b) Identification of QTL

associated with resistance to bacterial spot race T4 in tomato. Theor Appl Genet

121:1275-1287

Ivors K. (2010). Commercial production of staked tomatoes in the Southeast. North Carolina

State Univ. Coop. Ext., Raleigh, NC.

Jones JB, Lacy GH, Bouzar H, Stall RE, Schaad NW (2004) Reclassification of the

xanthomonads associated with bacterial spot disease of tomato and pepper. Syst Appl

Microbiol 27:755-762

Kabelka E, Franchino B, Francis D (2002) Two loci from Lycopersicon hirsutum LA407 confer

resistance to strains of Clavibacter michiganensis subsp. michiganensis. Phytopathol

92:504-10

Kosambi DD (1944) The estimation of map distances from recombination values. Annals of

Eugenics 12:172-175

133

Miller SA, Mera J (2011) Evaluation of fungicides and bactericides for the control of foliar and

fruit diseases of processing tomatoes. Plant Disease Management Reports 6:V061

Minsavage G, Balogh B, Stall R, Jones JB (2003) New tomato races of Xanthomonas campestris

pv. vesicatoria associated with mutagenesis of tomato race 3 strains. Phytopathol 93: S62

(Abstr.)

Mixon JT (2012) Prevalence of copper resistance among foliar bacterial pathogens of tomato in

Tennessee. Master's Thesis, University of Tennessee

Pascual L, Desplat N, Huang BE, Desgroux A, Bruguier L, Bouchet JP, Le QH, Chauchard B,

Verschave P, Causse M (2015) Potential of a tomato MAGIC population to decipher the

genetic control of quantitative traits and detect causal variants in the resequencing era.

Plant Biotechnol J 13:565-577

Pohronezny K, Volin RB (1983) The effect of bacterial spot on yield and quality of fresh market

tomatoes [Xanthomonas campestris] HortScience 18:69-70

Potnis N, Timilsina S, Strayer A, Shantharaj D, Barak JD, Paret ML, Vallad GE, Jones JB (2015)

Bacterial spot of tomato and pepper: diverse Xanthomonas species with a wide variety of

virulence factors posing a worldwide challenge. Mol Plant Pathol 16:907-920

Rodríguez GR, Kim HJ, van der Knaap E (2013) Mapping of two suppressors of OVATE (sov)

loci in tomato. Heredity 111:256

Scott JW, Hutton SF, Shekasteband R, Sim SC, Francis DM. (2015). Identification of tomato

bacterial spot race T1, T2, T3, T4, and Xanthomonas gardneri resistance QTLs derived

from PI 114490 populations selected for race T4. Acta Horticulturae 1069:53-58

Sharlach M, Dahlbeck D, Liu L, Chiu J, Jiménez-Gómez JM, Kimura S, Koenig D, Maloof JN,

Sinha N, Minsavage GV, Jones JB (2013). Fine genetic mapping of RXopJ4, a bacterial

134

spot disease resistance locus from Solanum pennellii LA716. Theor Appl Genet 126:601-

609

Shenge K, Mabagala RB, Mortensen CN, Wydra K (2014) Resistance of Xanthomonas

campestris pv. vesicatoria isolates from Tanzania to copper and implications for bacterial

spot management. Afr J Microbiol Res 8:2881-2885

Sim S, Durstewitz G, Plieske J, Wieseke R, Ganal MW, Van Deynze A, Hamilton JP, Buell CR,

Causse M, Wijeratne S (2012) Development of a large SNP genotyping array and

generation of high-density genetic maps in tomato. PLoS One 7:e40563

Simko I, Piepho, HP (2012) The area under the disease progress stairs: calculation, advantage,

and application. Phytopathol 102:381-389

Stam P (1993) Construction of integrated genetic linkage maps by means of a new computer

package: Join Map. Plant J 3:739-744

Timilsina S, Jibrin MO, Potnis N, Minsavage GV, Kebede M, Schwartz A, Bart R, Staskawicz

B, Boyer C, Vallad GE, Pruvost O (2015) Multilocus sequence analysis of xanthomonads

causing bacterial spot of tomato and pepper plants reveals strains generated by

recombination among species and recent global spread of Xanthomonas gardneri. Appl

Environ Microbiol 81:1520-1529

Vallad GE, Timilsina S, Adkison H, Potnis N, Minsavage G, Jones J, Goss E (2013) A recent

survey of xanthomonads causing bacterial spot of tomato in Florida provides insights into

management strategies. TomaTo Proceedings 25

Van Ooijen JW (2006) Joinmap 4.0, Software for the calculation of genetic linkage maps in

experimental populations. Plant Research International, Wageningen, Netherlands

135

Wang S, Basten CJ, Zeng ZB (2012) Windows QTL Cartographer 2.5. Department of Statistics,

North Carolina State University, Raleigh, NC

(http://statgen.ncsu.edu/qtlcart/WQTLCart.htm)

Würschum, T (2012) Mapping QTL for agronomic traits in breeding populations. Theor Appl

Genet 125:201-210.

136

CHAPTER 4: MAPPING QUANTITATIVE TRAIT LOCI (QTL) CONTROLLING

FRUIT MORPHOLOGY AND COLOR PARAMETERS IN AN INTRA-SPECIFIC

TOMATO POPULATION

ABSTRACT

Tomato (Solanum lycopersicum L.), is the second most consumed vegetable in the world.

The market value and culinary purpose are often determined by fruit size, shape, and color, which makes the genetic improvement of those traits a priority for tomato breeders. Insufficient polymorphic markers among improved germplasm have required that genetic studies of fruit morphology in tomato employ inter-specific populations, which often possess high number of deleterious alleles due to the linkage drag with the identified quantitative trait loci (QTL). With the release of the Solanaceae Coordinated Agricultural Project (SolCAP) 7720 SNP array and decrease in genotyping costs, tomato breeders can now conduct mapping studies in intra-specific populations. The main objective of the study was to detect QTL associated with the tomato fruit shape, size, and color. The use of elite breeding materials in the genetic mapping studies will facilitate the detection of genetic loci of direct relevance to breeders. We performed QTL analysis in an intra-specific population of tomato developed from a cross between two elite breeding lines NC 30P × NC-22L-1(2008) consisting of 110 recombinant inbred lines (RIL). The precision software Tomato Analyzer (TA) was used to measure 35 fruit morphology attributes and nine color attributes in two environments. The RIL population was genotyped with the

SolCAP 7720 SNP array. We identified novel QTL controlling fruit shape attributes on chromosome 10 and 12 explaining up to 25% phenotypic variance. This information will be useful in improving tomato fruit morphology traits.

137

INTRODUCTION

Tomato is the second most consumed vegetable after potato in the United States, where the per capita consumption of fresh market and processed tomato in 2017 were 20.3 lb (9.2 kg) and 73.3 lb (33.3 kg), respectively (Agricultural Marketing Resource Center, 2018). Tomato is consumed in various forms (fresh or processed) and is rich in vitamins A and C, as well as lycopene, which studies have suggested can decrease odds of cancer and heart disease (Merk et al., 2012). Tomato is also a model species for fleshy fruit development because of its genetic tractability and economic value. The origin of domestication for tomato is in the Andean region of South America and was spread to Europe as part of the Columbian exchange. During domestication and breeding, the tomato was selected for various fruit shapes, sizes, and color. As a result, today cultivated tomatoes have diverse shapes, sizes, and color. The size of cultivated tomatoes ranges from small cherry size to medium, large and extra-large fruited tomatoes. The shape of tomato can be classified into eight categories-flat, round, rectangular, ellipsoid, heart, long, obovoid, and oxheart (Rodriguez et al., 2011). The color of cultivated tomatoes ranges from green, pink, yellow, orange to red, dark red and purple (Paran and van der Knaap, 2007).

The tomato fruit morphology and appearance determine the culinary purposes (fresh consumption, sliced, diced, processed or cooked) and market value of the fruits. Growers demand high profitable tomatoes, which is correlated with the tomato sizes. Consumers demand deep red and visually flawless tomatoes that are firm and sweet tasting (Piombino et al., 2013).

Also, the consumers tend to judge tomatoes first on visual appearance, and taste second, although both traits are important (Barrett et al., 2010). The processing industry prefers rectangular and blocky tomatoes as these shapes prevent the fruit from rolling from conveyor belts during mechanical harvesting (Visa et al., 2014). The importance of tomato fruit shape, size, and color to consumers makes the genetic improvement of those traits a priority for tomato

138

breeders. A better understanding of genetic basis of fruit morphology and appearance will aid in their genetic improvements through breeding efforts.

Early molecular genetic studies in tomato focused on fruit size, shape, and color because of their economic importance and ease of phenotyping. The tomato fruit color is a function of different pigments including carotenoids, chlorophyll, flavonoids, and anthocyanins (Wang et al.,

2015). Several genes involved in the carotenoid pathway such as phytoene synthase gene (Psy1), lycopene β-cyclase (LcyB) gene, lycopene ε-cyclase (εLCY) gene etc. resulting in different fruit colors have been identified in tomato (reviewed by Levin and Schaffer, 2013). However, more genetic loci other than carotenoid genes are involved in the tomato fruit color suggesting tomato fruit color is quantitatively inherited (Liu et al., 2003). Several genes influencing tomato fruit size and shape have been identified such as SUN, OVATE, cell number regulator (CNR)/FW2.2,

SlKLUH/FW3.2, FASCIATED (FAS), and LOCULE NUMBER (LC) (Chakrabarti et al., 2013;

Cong et al., 2008; Frary et al., 2000; Liu et al., 2002; Muños et al., 2011; Xiao et al., 2008). The fas mutation is present in most multilocular fresh market tomatoes, whereas large and extra-large fruited tomatoes carry mutations in both fas and lc loci. Epistatic interaction between fas and lc loci has been reported as a causal factor for the extremely large size of tomatoes mediated through high locule number as reviewed by van der Knaap et al. (2014). SUN confers uniform elongation to maintain bilateral symmetry as observed in most commercially grown tomatoes, heirloom, and oxheart tomatoes, whereas OVATE is responsible for the asymmetric elongation causing neck constriction or pear shape as observed in ellipsoid and obovoid varieties of grape tomato (Gonzalo and van der Knapp, 2008). A major fruit shape QTL fs8.1 confers a blocky and slightly elongated shape in processing tomatoes and have arisen early in the tomato domestication process (Grandillo et al., 1996; Ku et al. 2000; Clevenger, 2012). The two

139

suppressors of OVATE gene sov1 and sov2 were also reported. The sov1 was responsible for both obovoid and elongated shape, sov2 conferred elongated fruit shape (Rodriguez et al., 2013).

These gene discoveries have explained the large proportions of diversity in the fruit morphology and color (Rodriguez et al., 2011) and have enabled the breeding program to manipulate tomato fruit morphologies and appearance traits to produce the fruits demanded by specific market classes. However, it is expected that there are undetected genes/loci controlling tomato fruit shapes, sizes, and color because of the polygenic nature of such traits. Detection of additional genetic factors underlying tomato fruit shape, size and color is the first step in better understanding economically important fruit morphology and appearance traits. Most of the cloned fruit shape and size genes are homozygous in improved germplasm. Therefore, breeders need to detect loci of small effect within improved germplasm. The use of elite breeding materials will allow detection of the minor allele effects, as most large effect alleles are homozygous in the population derived from closely related parents (Rodriguez et al., 2013). The intra-specific population will, therefore, enable to decipher a complete set of genes/loci influencing fruit morphology and appearance traits (Rodriguez et al., 2013). Inter-specific populations, derived from crossing a wild relative to domesticated tomato, have traditionally been used to study fruit morphology and appearance traits due to the relative ease of detecting marker polymorphisms. Introgressing potentially advantageous alleles identified in inter-specific populations is difficult due to the confounding effect of different genetic backgrounds and deleterious linkage drag (Lecomte et al. 2004). On the other hand, the use of elite breeding materials in QTL mapping study will facilitate the detection of QTL of direct relevance to breeders (Würschum, 2012).

The present study performed QTL mapping for tomato fruit shape, size and color traits segregating in an intra-specific bi-parental fresh market tomato NC 10204 derived from the cross

140

of two elite breeding lines. NC 10204 population was segregating for the ovate or pear-shaped trait. The QTL analysis in our study used the Solanaceae Coordinated Agricultural Project

(SolCAP) 7720 SNP array (Sim et al., 2102) for genotyping and the precision phenotyping software Tomato Analyzer (TA) for phenotyping (Brewer et al., 2006). The ovate shape is characterized by three attributes in TA- fruit shape triangle, ovoid, and obovoid. Fruit shape triangle is the ratio of the proximal end width to distal end width (Brewer et al., 2006). Ovoid and obovoid represent the extent of the top and bottom-heaviness of a fruit respectively (Brewer et al., 2006). The effectiveness of TA and the SolCAP array in genetic studies of tomato have been previously demonstrated (Rodriguez et al., 2010a; Rodriguez et al., 2013). Few QTL mapping experiments in tomato have been conducted using intra-specific populations, and to our knowledge no study has identified QTL segregating within an elite breeding material.

Our objectives in this research were to detect QTL that influence fruit i) shape, ii) size, iii) and color. We report novel QTL controlling pear and ovate shaped tomato fruit characterized by ovoid, obovoid, and fruit shape triangle. A better understanding of the loci influencing fruit shape, size, and color will help breeders improve these traits to the benefit of both growers and consumers.

MATERIALS AND METHODS

Population Development

The hybrid NC 10204 was created in 2010 by crossing the plum tomato breeding line NC

30P and the grape tomato breeding line NC 22L-1(2008). NC 30P has the crimson (ogc) gene governing increased lycopene content, and highly elongated fruits with 2-3 locules (Figure 4.1).

It was released as a breeding line for its superior horticultural traits from North Carolina State

University tomato breeding program (Gardner and Panthee, 2010). NC 22L-1(2008), an

141

advanced breeding line derived from S. pimpinellifolium L3707, also elongated fruits with 2-3 locules (Figure 4.1).The F1 hybrid was self-pollinated to create a segregating F2 population, F2 plants were individually harvested to create F2:3 families and so on all the way to create F5:6 families at the Mountain Horticultural Crops Research and Extension Center (MHCREC) in

Mills River, NC. Although the mapping population was started with 284 F2 lines, it was reduced to 129 lines in F2:3 generation and 110 lines in F5:6 generation due to exposure to the deleterious amount of bleach and poor seed germination. Seeds were germinated in 72 cell trays (56 x 28 cm2) in potting mix and grown for six weeks before hand-transplantation in the field. In each generation, the NC 10204 population was planted in a random complete block design with two replications. Individual plants were grown 45 cm apart within rows, 150 cm apart between rows, in plastic mulch with drip irrigation. Plants were hand strung and sprayed according to the recommended schedule for fungicides and insecticides (Ivors, 2010).

Figure 4.1: The fruit images of NC 22L-1 (2008) (A) and NC 30P (B).

142

Phenotypic Analysis

Fruits were phenotyped in the F2:3 and F5:6 generations in 2014 and 2017, respectively using Tomato Analyzer (TA) version 3.0. Ten-twenty fruits per genotype per replication per year were cut proximal to distal (stem end to blossom end), and scanned using a flatbed scanner

(CanoScan 8800F, Canon U.S.A. Inc, Melville, NY, USA). The fruit images were saved as jpeg files and imported into Tomato Analyzer 3.0 for automated phenotypic measurements (Figure

4.2). Images were manually adjusted as needed and analyzed as described by Rodriguez et al.

(2010b). A full description of traits measured by TA can be found in the Tomato Analyzer

Version 3 User Manual (Brewer et al., 2006). Attributes that were segregating within the

Figure 4.2: The scanned tomato fruit images that are imported into Tomato Analyzer (TA) software.

143

populations by visual observation were selected for analysis and exported by Tomato Analyzer for further data analysis. The total of 35 fruit morphology attributes and nine color attributes based on RGB color space (Red, Green, Blue) and CIELAB color space (L*, a*, b*), including two color descriptors (Hue and Chroma) were measured using Tomato Analyzer.

Data Analysis

Data analysis was performed in SAS software (version 9.4, SAS Institute Inc, Cary, NC).

Analysis of variance was performed in SAS using ‘PROC MIXED’. The equation used to estimate variance and least squared (LS) means are the same, which was:

!"#$ = & + () + *+ + ,(./+) + 1()+) + 2)+.

Where !"#$ is the trait value of the kth replication of the ith genotype in the jth year, & is the

th th population mean, () is the fixed effect of the i genotype, *+ is the random effect of j year,

th th ,(./+) is the random effect of k replication in the j year, 1()+) is the random effect of the genotype by year interaction, and 2)+. is the error term (residual). The LS means for each trait were exported for the QTL analysis. Pearson’s correlation coefficients were calculated for each attribute between two generations, and among various attributes within each generation using

‘PROC CORR’. The heritability was estimated for each environment by calculating variance components using ‘ASYCOV’ function in PROC MIXED in SAS.

DNA Extraction and Genotyping

Genomic DNA was obtained from F2 plants and parental lines in the summer of 2013 using a modified cetyltrimethyl ammonium bromide (CTAB) method and stored at -20°C in 10 mM Tris– HCl pH 8.0, and one mM EDTA (Kabelka et al., 2002). Prior to genotyping, DNA was quantified using a NanoDrop 2000 Spectrophotometer (Thermo Scientific, Wilmington, DE,

144

USA). One hundred ninety F2 plants and both parents (in 2013) were genotyped using the

SolCAP Illumina Infinium Assay. SNP genotypes were determined using GenomeStudio version

1.0 (Illumina Inc, San Diego, CA, USA).

Linkage Map Construction and QTL Analysis

A total of 110 individuals (NC 10204) were used for the QTL analysis. All lines genotyped or phenotyped could not be used for QTL analysis because either genotypic or phenotypic data were missed. The genetic map for NC 10204 was constructed using Joinmap 4.0

(Van Ooijen 2006). Regression mapping algorithm (Stam 1993) was used to calculate marker order within each group. Kosambi mapping function was used for estimation of map distances

(cM) (Kosambi 1944). The linkage map was composed of 886 SNP polymorphic markers across

12 chromosomes that spanned 739.50 cM with an average distance of 0.83 cM between markers.

Windows QTL Cartographer v 2.5 software was used for QTL analysis. Composite

Interval Mapping (CIM) method using the default parameters (model 6) was used. A backward regression was used to perform the CIM analysis to enter or remove background markers from the model. The walking speed was set at one cM for the detection of QTL. A default threshold of likelihood of odd (LOD) score of 2.5 was used to declare the presence of QTL. The additive effect and the proportion of the observed phenotypic variation (R2) for each QTL were also obtained using this software. The dominant effect was not estimated in F5:6 generation as lines were in mostly in homozygous condition, hence not presented for both generations. QTLs explaining more than 10% of the phenotypic variance were considered as major QTLs.

Considering bi-parental mapping population and size of the mapping of population used in this study, any QTL within 10 cM distance on the same chromosomes were regarded as a single

QTL, and if detected in both environments considered as a consistent QTL (Pascual et al., 2015).

145

RESULTS

A total of 35 fruit morphological attributes and nine color traits were measured using TA in two generations. The descriptive statistics along with the broad sense heritability based on variance components were presented in Table 4.1. The size attributes are higher in F5:6 generations (Table 4.1). The heritability estimates were 0.7 for size attributes, 0.3-0.6 for color attributes, and 0 to 0.9 for fruit shape attributes (Table 4.1). All the attributes measured were significantly correlated at p<0.05 between F2:3 and F5:6 generations except nine fruit shape attributes (ellipsoid, proximal angle micro, distal angle macro, distal indentation area, distal end protrusion, vertical asymmetry, horizontal asymmetry ovoid, proximal eccentricity, and distal eccentricity) and a color attribute (average A) (Table 4.1). Within each generation, significant correlations were detected between many traits, but only traits with detected QTL will be discussed (Table 4.2 and Table 4.3). For fruit size traits, significant correlations (p<0.001) were observed in both generations between perimeter, area, width mid-height, and maximum width.

Significant negative correlations were observed between fruit shape triangle and ovoid

(p<0.001), and between obovoid and ovoid (p<0.001). A significant correlation (p<0.001) was observed between RGB based color attribute ‘Average Red’ and CIELAB based color attribute

‘Average L’ in both generations. Average Red and Average L were significantly correlated

(p<0.001) with Average B.

146

Table 4.1: Descriptive statistics for fruit morphological and color attributes measured by Tomato Analyzer (TA) in two generations (F2:3 and F5:6), significance test of genotypic effects on the attributes measured, and the correlation coefficients between generations in the intra-specific population NC 10204 of tomato. Table also presents the heritability estimates (H2) for all fruit quality traits.

F2:3(Year 2014) F5:6 (Year 2017) correlation Variable between Std Geno Std Geno Mean Min. Max. H2 Mean Min. Max. H2 generationsb Dev effecta Dev Effect P 207.8 20.38 155.39 281.89 <.0001 0.7 1929 206.19 1403 2566 <.0001 0.7 0.25* A 2836 549.16 1640 5095 <.0001 0.7 2E+05 49705 1E+05 416683 <.0001 0.7 0.25* WMH 50.62 5.39 40.75 66.45 <.0001 0.7 460.9 50.34 300.3 599.83 <.0001 0.7 0.24* MW 51.46 5.4 42.25 67.26 <.0001 0.7 466.2 50.35 304.1 601.75 <.0001 0.7 0.24* HMW 64.46 7.83 44.82 88.38 <.0001 0.7 640.4 87.1 466.2 914.33 <.0001 0.7 0.24* MH 67.64 7.67 47.58 93.02 <.0001 0.7 654.7 86.52 480.3 921.96 <.0001 0.7 0.3** CH 68.07 7.57 48.12 92.92 <.0001 0.7 660.9 85.21 491.3 932.01 <.0001 0.7 0.3** FSI-EII 1.33 0.15 1.06 1.79 <.0001 0.7 1.42 0.19 1.05 2.29 <.0001 0.8 0.3** FSI-EI 1.29 0.17 0.98 1.83 <.0001 0.7 1.4 0.21 0.99 2.34 <.0001 0.9 0.18 CFSH 1.36 0.16 1.09 1.86 <.0001 0.7 1.45 0.2 1.08 2.38 <.0001 0.8 0.28* PFB 0.74 0.04 0.63 0.83 <.0001 0.5 0.69 0.04 0.6 0.8 <.0001 0.7 0.32** DFB 0.64 0.04 0.47 0.73 <.0001 0.6 0.61 0.05 0.5 0.74 <.0001 0.7 0.1 FST 1.18 0.13 0.87 1.77 <.0001 0.6 1.15 0.14 0.87 1.51 <.0001 0.7 0.26* El 0.04 0.01 0.03 0.06 <.0001 0.5 0.04 0.01 0.03 0.06 <.0001 0.8 nsc Cir 0.1 0.03 0.05 0.2 <.0001 0.7 0.11 0.04 0.04 0.26 <.0001 0.8 0.18 Rect 0.53 0.02 0.46 0.57 <.0001 0.6 0.52 0.02 0.48 0.56 <.0001 0.7 0.36*** SH 0.03 0.01 0.01 0.05 <.0001 0.6 0.01 0.01 0 0.04 <.0001 0.7 0.37*** PAMi 201.5 40.92 80.45 249.56 <.0001 0.7 175.8 22.82 117.6 224.18 0 0.3 0.35** PAMa 123 18.11 61.19 158.92 <.0001 0.6 75.63 17.76 27.72 121.97 <.0001 0.9 0.41***

147

Table 4.1 (continued).

PIA 0.07 0.02 0.01 0.13 <.0001 0.6 0.03 0.02 0 0.09 <.0001 0.7 0.56*** DAMa 147.8 13.6 104.16 169.03 0.15 ns 0.2 149.3 10.81 118.5 176.84 0.01 0.2 0.52*** DAMi 107.3 10.15 82.7 132.65 <.0001 0.6 79.82 9.56 43.76 100.73 <.0001 0.8 0.17 DIA 0 0 0 0.01 0 0.4 0 0 0 0 0.08 ns 0.1 0.36*** DEP 0.01 0.01 0 0.09 0.08 ns 0.1 0 0.02 0 0.23 0.35 ns 0 0.24* Ob 0.05 0.05 0 0.24 <.0001 0.6 0.07 0.06 0 0.23 <.0001 0.8 0.48*** Ov 0.12 0.06 0 0.26 <.0001 0.6 0.09 0.06 0 0.21 <.0001 0.7 0.48*** VA 0.1 0.02 0.07 0.16 0.01 0.3 0.06 0.11 0.03 1.23 0.48 0 -0.08 HA-Ob 0.06 0.07 0 0.44 <.0001 0.7 0.06 0.06 0 0.28 <.0001 0.8 0.47*** HA-Ov 0.16 0.08 0 0.42 <.0001 0.5 0.09 0.18 0 1.97 0.14 ns 0.1 0.2 WWP 0.48 0.05 0.38 0.61 <.0001 0.6 0.51 0.04 0.41 0.61 <.0001 0.8 0.46*** Ecc 0.76 0.01 0.73 0.79 <.0001 0.6 0.78 0.01 0.75 0.79 0 0.6 0.26* PE 0.89 0 0.88 0.9 0.38 ns 0.1 0.89 0 0.89 0.91 0 0.3 0.05 DE 0.89 0 0.88 0.89 0.33 ns 0 0.89 0 0.88 0.89 0.40 ns 0 ns FSI-I 1.29 0.17 0.98 1.83 <.0001 0.7 1.41 0.2 1 2.34 <.0001 0.8 0.37*** EAI 0.42 0.01 0.38 0.45 <.0001 0.6 0.39 0.01 0.36 0.43 <.0001 0.7 0.24* Avg-R 175.7 6.67 160.03 199.95 <.0001 0.6 178.5 6.3 165.2 192.98 <.0001 0.6 0.43*** Avg-G 93.79 9.12 68.45 119.53 <.0001 0.6 95.1 7.58 78.51 112.7 <.0001 0.3 0.29** Avg-B 64.14 5.74 49.24 78.88 <.0001 0.4 65.83 4.85 55.88 78.16 <.0001 0.3 0.38*** Avg- 112.7 5.48 99.31 126.72 <.0001 0.5 114.8 4.88 104 126.83 <.0001 0.5 0.44*** Lum Avg-L 43.22 3.05 35.92 51.68 <.0001 0.6 43.85 2.52 38.68 49.24 <.0001 0.4 0.35** Avg-a 32.37 2.71 22.31 40.25 <.0001 0.6 29.82 2.42 24.18 35.62 <.0001 0.3 0.12 Avg-b 28.94 2.09 24.43 37.21 <.0001 0.5 28.49 1.58 24.37 32.22 <.0001 0.5 0.25* Avg-H 41.96 3.77 32.05 55.39 <.0001 0.6 43.98 3.16 36.58 52.02 <.0001 0.3 0.14

148

Table 4.1 (continued).

Avg-C 43.58 1.89 39.19 49.79 <.0001 0.5 41.45 1.83 36.89 46.98 <.0001 0.3 0.15

Note: a The p-value determining the significane of genotypic effects. b ‘***’ denotes p - value < 0.001, ‘**’ p - value < 0.01, ‘*’ p - value < 0.05 c ns represents non-significant

Variables: A= area; P=perimeter; WMH=width mid-height; MW=maximum width; HMW=height mid-width; MH= maximum height; CH=curved height; FSI-ExtII=fruit shape index external II; FSI-ExtI= fruit shape index external I; CFSH=curved fruit shape index; PFB=proximal fruit Blockiness; DFB= distal fruit Blockiness; FST=fruit shape triangle; El=Ellipsoid; Cir=circular;

Rect=Rectangular; SH=shoulder height; PAM =proximal angle micro; PIA=proximal indentation area; DAMi= distal angle micro;

DAMa=distal angle macro; DIA= distal indentation area; DEP = distal end protrusion; Ob=obovoid; Ov=ovoid; VA=vertical asymmetry; HA-Ob=horizontal asymmetry obovoid; HA-Ov=horizontal asymmetry ovoid; WWP=width widest position;

Ecc=eccentricity; PE=proximal eccentricity; DE=distal eccentricity; FSI-Int=fruit shape index internal; EAI=eccentricity area index;

Avg-R=average red; Avg-G=average green; Avg-B=average blue; Avg-Lum=average luminosity; Avg-L=average L value; Avg- a=average a value; Avg-b=average b value; Avg-H= average hue; Avg-C= average chroma

149

Table 4.2: Pearson’s correlation coefficient among fruit morphology and color attributes in the intra-specific tomato population NC 10204 in F2:3 generation.

P A WMH MW FST Ob Ov AvgR AvgL AvgB P 1 A 0.99***a 1 WMH 0.8*** 0.86*** 1 MW 0.81*** 0.86*** 0.99*** 1 FST -0.19* -0.18* ns ns 1 - Ob 0.22* ns -0.18* ns 0.77*** 1 Ov -0.24*** -0.20* ns ns 0.85*** -0.88*** 1 AvgR 0.42*** 0.41*** 0.29** 0.28** -0.24** 0.26** -0.31*** 1 AvgL 0.49*** 0.49*** 0.39*** 0.39*** -0.24** 0.23* -0.28** 0.92*** 1 AvgB nsb ns ns ns -0.23** ns -0.20* 0.56*** 0.55*** 1

Note: a ‘***’ denotes p - value < 0.001, ‘**’ p - value < 0.01, ‘*’ p - value < 0.05 b ns represents non-significant

Variables: P=perimeter, A=area, WMH =width mid-height, MW= maximum width, FST= fruit shape triangle, Ob=Obovoid,

Ov=Ovoid, VA=vertical asymmetry, AvgR=Average Red, AvgL= Average L, and AvgB= Average B

150

Table 4.3: Pearson’s correlation coefficient among fruit morphology and color traits in the intra-specific tomato population NC 10204 in F5:6 generation.

P A WMH MW FST Ob Ov AvgR AvgL AvgB P 1 A 0.98***a 1 WMH 0.70*** 0.79*** 1 MW 0.73*** 0.81*** 0.99*** 1

FST -0.2* ns ns ns 1 - 0.33*** 0.23* ns ns 1 Ob 0.83*** - -0.26** ns ns ns 0.92*** 1 Ov 0.92*** AvgR 0.57*** 0.55*** 0.35*** 0.38*** ns 0.27** ns 1 AvgL 0.54*** 0.54*** 0.41*** 0.42*** ns ns ns 0.88*** 1 AvgB ns ns ns ns -0.23* 0.3** -0.24* 0.6*** 0.5754 1

Note: a ‘***’ denotes p - value < 0.001, ‘**’ p - value < 0.01, ‘*’ p - value < 0.05 b ns represents non-significant

Variables: P=perimeter, A=area, WMH =width mid-height, MW= maximum width, FST= fruit shape triangle, Ob=Obovoid,

Ov=Ovoid, VA=vertical asymmetry, AvgR=Average Red, AvgL= Average L, and AvgB= Average B

151

QTL in F2:3 Generation

A total of 41 putative QTL were detected for all traits in F2:3 generation, 12 QTL describing fruit size variation, 17 explaining fruit shape variation, and 12 explaining fruit color variation. Among 41 QTL, 11 QTL explained less than 10% of phenotypic variance, 20 QTL explained 10% to 20% of the total phenotypic variance, 8 QTL explained 20% to 30% of the total phenotypic variance, and 2 QTL explained greater than 30% of the phenotypic variance

(Table 4.4). The major QTL (>10% of phenotypic variance) were located on chromosomes 4 and

9 for fruit size traits; on chromosomes 2, 4, 5, and 12 for fruit shape traits; and on chromosomes

6, 9, and 11 for fruit color traits (Table 4.4).

QTL in F5:6 Generation

A total of 28 putative QTL were detected for all traits in F5:6 generation, 7 QTL describing fruit size variation, 13 explaining fruit shape variation, and 8 explaining fruit color variation. Among 28 QTL, 7 QTL explained less than 10% of phenotypic variance, 16 QTL explained 10% to 20% of the total phenotypic variance, 4 QTL explained 20% to 30% of the total phenotypic variance, and 1 QTL explained greater than 30% of the phenotypic variance

(Table 4.5). The major QTL (>10% of phenotypic variance) were located on chromosomes 1 and

2 for fruit size traits; on chromosomes 1, 7, 8, 10, and 12 for fruit shape traits; and on chromosomes 4, 6, 9, and 12 for fruit color traits (Table 4.5).

152

Table 4.4: Summary of QTL information obtained in NC 10204 population in F2:3 generation controlling fruit morphology and color.

Category LOD Additive Dominance R2-value Trait Chr. Position(cM) of Trait Score Value Value (%) A 9 51.8 3.4 -303.6 143.4 14.9 A 2 77.4 8.2 -229.8 -210.4 8.4 CH 9 36.2 5.2 -4.2 -2.9 11.4 CH 4 65.5 7.4 -3.9 -3.9 10.4 HMW 9 36.2 6.6 -5.4 -3.3 17.8 Fruit size HMW 4 65.5 6.7 -3.7 -4.2 8.5 MH 4 65.6 7.2 -3.8 -4 9.7 MW 2 77.4 4.7 -1 -2.6 1.4 P 2 77.4 7.4 -8.7 -7.1 8.1 P 4 57.1 5.4 -14.4 0.7 18.4 P 9 51.9 4.2 -12.7 5.8 18.5 WMH 2 77.4 4.9 -1 -2.7 1.4 Cir 5 35.7 3 0 0 34.4 DAMi 5 38.7 3.4 -9 4.1 30.8 Ecc 9 33 3.2 0 0 10 Ecc 6 19.5 4.6 0 0 3.8 FST 10 22.8 3.4 0.1 -0.1 8.5 HA-Ob 12 0.3 4 -0.1 0 24.9 HA-Ov 12 1.3 4.9 0.1 -0.1 26.9 Fruit Ob 12 0.3 4.1 0 0 24.6 shape O 12 0.3 3.7 0 0 24.8 PAMa 4 68.6 5.1 -15.2 16.3 28.2 PAMi 9 36.2 3.8 16 21.7 5.9 PE 8 26.9 3.6 0 0 5.6 PFB 12 0 3 0 0 17.3 PFB 2 46 3.4 0 0 14.9 PIA 2 54.1 3.3 0 0 12 VA 4 51.6 4.4 0 0 11.7 VA 2 53.3 3.3 0 0 10.2 AvgB 11 2 5.2 -3.1 0.6 40.8 AvgB 6 13.8 3.6 1.3 0.5 14.1 AvgBlue 6 0 4.6 -3.8 2.4 22.7 AvgBlue 9 33.8 5.1 -4 -1.1 20.5 Fruit AvgBlue 9 23.5 4 -3.9 -0.1 17.6 Color AvgBlue 4 58.9 4.2 -2.6 -1.4 7.5 AvgG 9 33 3.6 -5.6 -1.7 13.8 AvgHue 6 14.4 5.2 2.5 0.9 17.9 AvgL 9 33.8 3.9 -1.9 -0.8 14.2 AvgLum 9 33.8 4.1 -3.9 -0.9 21.1 AvgRed 9 33.8 4.8 -4.8 -2 18.2 AvgRed 4 59.5 3.2 -1.9 -2.5 2.3

153

Table 4.4 (continued).

Note:

Traits: A = area CH = curved height HMW = height mid-width MH = maximum height MW = maximum width P = perimeter WMH = width mid-height Cir = circular DAMi = distal angle micro Ecc = eccentricity FST = fruit shape triangle HA-Ov = horizontal asymmetry ovoid HA-Ob = horizontal asymmetry obovoid Ov = ovoid Ob = obovoid PAMa = proximal angle macro PAMi = proximal angle micro PE = proximal eccentricity PFB = proximal fruit blockiness PIA = proximal indentation area VA = vertical asymmetry AvgB = average b value AvgBlue = average blue AvgG = average green AvgR = average red AvgLum = average luminosity AvgH = average hue AvgL = average L valu

154

Table 4.5: Summary of QTL information obtained in NC10204 population of tomato in F5:6 generation controlling fruit morphology and color.

Category LOD Additive Trait Chr. Position(cM) R2-value (%) of Trait score Value A 2 66.7 5.6 0 19.6 CH 1 86.9 3.4 89.4 22.1 HMW 1 86.9 3.5 92.5 22.8 Fruit size MH 1 86.9 3.3 89.7 21.6 MW 2 77.4 3.2 -29.1 13.5 P 2 67.7 4.8 -188.4 16.9 WMH 2 77.4 3.2 -29.1 13.4 DEP 1 86.9 4.3 0 14 DFB 10 15.1 4 0 18.5 FST 10 16.1 5.5 0.1 24.2 HA-Ov 2 56.1 4.3 -0.1 3.1 Ob 12 41.8 3 0 6.8 Ov 12 43.5 3.2 0 7.8 Fruit PAMi 8 19.6 4.8 -15.9 16.4 shape PE 1 60.1 4 0 34.1 PFB 7 36.8 3 0 14.1 PFB 1 61.1 3.4 0 3.7 Rect 8 0.5 3.5 0 4 SH 1 73.6 3.4 0 10.2 WWP 12 40.8 3.4 0 11.2 AvgB 4 66.4 3.5 -1.5 21.9 AvgB 4 64.5 3.5 -1.4 21.5 AvgB 6 14.3 4.5 1 20.7 AvgC 12 0 4.1 -1.1 17.9 Color AvgH 11 15.2 3.5 -1.4 7.2 AvgL 4 63.6 3.6 -1.5 15.9 AvgL 9 1.6 3.7 -1.2 12 AvgR 4 59.5 3 -2.7 9.2 Note:

Traits: A= area; CH=curved height; HMW=height mid-width; MH= maximum height; MW=maximum width; P=perimeter; WMH=width mid-height; DEP = distal end protrusion; DFB= distal fruit Blockiness; FST=fruit shape triangle; FST=fruit shape triangle; HA- Ov=horizontal asymmetry ovoid; Ov=ovoid; Ob=obovoid; PAMi =proximal angle micro; PE=proximal eccentricity; PFB=proximal fruit Blockiness; Rect=Rectangular; SH=shoulder height; WWP=width widest position; AvgB=average b value; AvgC= average chroma; AvgH= average hue; AvgL=average L value; AvgR=average red

155

QTL in Both Generations

Although 41 and 28 QTL were detected in an individual generation, only 10 QTL were detected in both generations. The QTL controlling fruit size attributes area, perimeter, maximum width (MW), and width mid-height (WMH) were detected on chromosome 2 with a LOD score value of 3 to 8 that explained up to 20% of the phenotypic variance (Figure 4.3; Table 4.6). The

QTL controlling fruit shape attributes fruit shape triangle (tri), ovoid (Ov), and obovoid (Ob) were identified on chromosomes 10 and 12 with a LOD score value of 3 to 6 explaining up to

25% of the phenotypic variance (Figure 4.4; Table 4.6). The QTL associated with the external fruit color attributes- average Red (AvgR), average L (AvgL), and average B (AvgB) were detected on chromosome 4, 9, and 6 explaining up to 21% of phenotypic variance (Figure 4.5;

Table 4.6). Although QTL associated with AvgL and AvgB were detected, no QTL for hue parameter detected. The position of seven QTL were consistent in both environments- QTLs controlling WMH, MW, AvgR, and AvgB were located within <5cM between two generations and perimeter, area, and tri were located within <10cM between two generations. However, the genomic locations of QTL controlling Ob, Ov, and Avg L were mapped to different locations on the same chromosome in different generations i.e., QTL were detected in opposite ends of the chromosomes in different years (Table 4.6). The positive additive effects were associated with the parent NC 30P and negative additive effects were associated with the parent NC 22L-1(2008)

(Table 4.6). The shape QTL detected in this study on chromosomes 10 and 12 are novel.

156

Table 4.6: Summary of consistent QTL information obtained in NC10204 population of tomato controlling fruit morphology and color in two years. The table presents the position of the QTL in genetic map and physical map, flanking SNP molecular markers, LOD score value and phenotypic variation explained by each QTL (R2-value), additive effect, and if the QTL has been previously identified or novel. Y1 indicates the year 2014 (F2:3 generation) and Y2 indicates the year 2017 (F5:6 generation). The positive additive effects were associated with the parent NC 30P and negative additive effects were associated with the parent NC 22L-1(2008).

Category Traitsa Chr. Closest Markersb Genetic Physical Map LOD Additive R2% of Trait Map Position (bp) Score Position (cM) Fruit Size Yr2_Perimeter 2 solcap_snp_sl_50066 67.7 44783686 5 -188 17 Yr1_Perimeter 2 solcap_snp_sl_21867 77.4 47948927 7 -8.7 8 Yr2_Area 2 solcap_snp_sl_42324 66.7 44069445 6 0 20 Yr1_Area 2 solcap_snp_sl_21867 77.4 47948927 8 -230 8 Yr2_WidthMidheight 2 solcap_snp_sl_21867 77.4 47948927 3 -29 13 Yr1_WidthMidheightt 2 solcap_snp_sl_21867 77.4 47948927 5 -1 1 Yr2_MaxWidth 2 solcap_snp_sl_21867 77.4 47948927 3 -29 14 Yr1_MaxWidth 2 solcap_snp_sl_21867 77.4 47948927 5 -1 1 Fruit Yr2_FruitShapeTriangle 10 solcap_snp_sl_34373, 16.1 3991802;4260136 6 0.1 24 Shape solcap_snp_sl_9598 Yr1_FruitShapeTriangle 10 solcap_snp_sl_9598, 22.8 4260136;57327585 3 0.1 9 solcap_snp_sl_16517 Yr2_Obovoid 12 solcap_snp_sl_14422, 41.8 62120915;7801435 3 0 7 solcap_snp_sl_24755 Yr1_Obovoid 12 solcap_snp_sl_1573, 0.3 4038812;5029856 4 0 25 solcap_snp_sl_58869 Yr2_Ovoid 12 solcap_snp_sl_24755, 43.5 7801435;64210355 3 0 8 solcap_snp_sl_31628 Yr1_Ovoid 12 solcap_snp_sl_1573, 0.3 4038812;5029856 4 0 25 solcap_snp_sl_58869

157

Table 4.6 (continued).

Fruit Yr2_AvgRed 4 solcap_snp_sl_11456, 59.5 58318160;58362612 3 -2.7 9 Color solcap_snp_sl_22228 Yr1_AvgRed 4 solcap_snp_sl_11456, 59.5 58318160;58362612 3 -1.9 2 solcap_snp_sl_22228 Yr2_AvgL 9 solcap_snp_sl_14679, 1.6 743707;756861 4 -1.2 12 solcap_snp_sl_58100 Yr1_AvgL 9 solcap_snp_sl_41490, 33.8 13637843; 4 -1.9 14 solcap_snp_sl_51505 41661711 Yr2_AvgB 6 Bcyc_868, 14.3 42288756; 5 1 21 solcap_snp_sl_57352 42343473 Yr1_AvgB 6 solcap_snp_sl_12765, 13.8 36998357;42288756 4 1.3 14 Bcyc_868

a The phenotypic traits for which QTLs have been detected. Y1= F2:3 generation and Y2=F5:6 generation b The markers located at the closest distances to the QTL positions

158

Chromosome 2 Yr1_Perimeter Chromosome 2 Yr2_Perimeter

Chromosome 1 Chromosome 2 Yr1_Area Yr2_Area

Chromosome 2 Chromosome 2 Yr1_WidthMidheight Yr2_WidthMidheight

Chromosome 2 Chromosome 2 Yr1_MaxWidth Yr2_MaxWidth

Figure 4.3: The logarithm of odds (LOD) graph for the tomato fruit size QTL detected on chromosome 2 in NC 10204. The peak represents the QTL. Horizontal axis represents the position of QTL (cM) on the chromomse and vertical axis represents the LOD values.

. 159

Chromosome 10 Chromosome 10 Yr1_FruitShapeTriangle Yr2_FruitShapeTriangle

Chromosome 12 Chromosome 12 Yr1_Ovoid Yr2_Obovoid

Chromosome 12 Chromosome 12 Yr1_Obovoid Yr2_Ovoid

Figure 4.4: The logarithm of odds (LOD) graph for the tomato fruit shape QTL detected on chromomse 10 and 12 in NC 10204. The peak represents the QTL. Horizontal axis represents the position of QTL (cM) on the chromomse and vertical axis represents the LOD values.

160

Chromosome 9 Chromosome 9 Yr1_AverageL Yr2_AverageL

Chromosome 4 Chromosome 4 Yr1_AverageRed Yr2_AverageRed

Chromosome 6 Chromosome 6 Yr1_AverageB Yr2_AverageB

Figure 4.5: The logarithm of odds (LOD) graph for the tomato external fruit color QTL detected on chromosomes 9, 4, and 6 in NC 10204. The peak represents the QTL. Horizontal axis represents the position of QTL (cM) on the chromomse and vertical axis represents the LOD values.

161

DISCUSSION

Fruit shapes, size, and color are polygenically inherited with effects ranging from small to large. Genetic studies to elucidate molecular mechanisms of fruit morphology and appearance traits in tomato have historically employed inter-specific populations derived from crossing cultivated and wild tomato species to obtain a sufficient number of polymorphic markers

(Rodriguez et al., 2013). However, introgressing QTL from wild/unimproved genotypes into elite lines often suffers from linkage drag and requires extensive backcrossing (Würschum,

2012). Recently intra-specific mapping populations have been used to study the genetic control of tomato fruit shape using the SolCAP SNP array (Rodriguez et al., 2010a; Rodriguez et al.,

2013).

In this study, we identified ten QTL in both the F2:3 and F5:6 generations controlling fruit shapes, sizes, and color traits using an intra-specific bi-parental mapping population derived from the elite tomato breeding lines NC 30P and NC 22L-1(2008). Three attributes of fruit shape traits- fruit shape triangle, ovoid, and obovoid were mapped in this study. Fruit shape triangle

(tri) was mapped to a 7cM region on chromosome 10 that explained 9-24% of the total phenotypic variance. Previously, tri has been mapped to chromosome 7 in three populations, and on chromosome 3 in one population (Brewer et al., 2007). QTL for tri was also detected on chromosomes 1, 2, 3, and 11 in three populations in another study (Gonzalo and van der Knapp,

2008). Consistent with the study of Gonzalo and van der Knapp (2008), in our study the QTL for fruit shape triangle did not overlap with major fruit shape loci on chromosome 7, indicating the fruit shape triangle in NC 10204 is independently controlled by a different locus on chromosome

10. To the extent of our knowledge, fruit shape triangle QTL has not been reported on chromosome 10, indicating that this is a novel QTL.

162

The ovoid and obovoid responsible for the elongated and pear-shaped tomato were co- localized on chromosome 12, which could be attributed to strong correlations between ovoid and obovoid. Previously, the major fruit shape genes SUN and OVATE responsible for fruit elongation were located on chromosomes 7 and 2 respectively (Xiao et al., 2008; Liu et al.,

2002), while the two suppressors of ovate (sov1 and sov2) controlling obovoid and elongated shape were mapped to chromosomes 10 and 11 (Rodriguez et al., 2013). Huang et al. (2013) detected one of the OVATE-like genes (SlOFP16) on the long arm of chromosome 8.

Chusreeaeon et al. (2014) also detected the Slelf1 mutant candidate gene controlling elongated fruit shape on the long arm of chromosome 8, but on different position than that of SlOFP16.

This suggests the QTL associated with the obovoid and ovoid in our study were detected on the genomic regions not identified by prior studies. In our study, both parents contributed equally to the obovoid and ovoid shape traits. However, the detected QTL were not consistent in two generations and detected in two different positions on chromosome 12. In F2:3 generation, both ovoid and obovoid were mapped to genetic map position of 0.3cM and explained 25% of phenotypic variance. In F5:6 generation, they were mapped to genetic map positions of 42cM to

44cM on chromosome 12 explaining 8% of phenotypic variance despite the significant correlations (r=0.5, p<0.001) of these traits between two generations. The markers associated with these QTL in different generations were located on different positions (~62Mb-78Mb in F5:6 and ~40Mb to 50Mb in F2:3) on the chromosome 12 in the physical map too. This suggests, there might be two different QTL controlling ovoid and obovoid fruit shape attributes in NC 10204 on chromosome 12 or the environment influences the expression of ovoid and obovoid.

Four fruit size attributes area, perimeter, maximum width, and width mid-height were mapped to a region between 67.7cM and 77.4cM in the genetic map on chromosome 2 in both generations that explained up to 20% of phenotypic variance. The significant correlations

163

between these four attributes suggest that a single QTL might be controlling all these four traits.

Three cloned genes that influence fruit size are located on chromosome 2 (ovate, fw2.2 and lcn2.1), as well as at least three other putative QTL (Lin et al., 2014). The closest markers associated with the detected QTL for fruit size in our study were located on the chromosome 2 at physical map positions of 44,069,445 to 47,94,8927 bp), which is close to the position of

Fw2.2/CNR (Solyc02g090730) (52, 252, 556 to 52,253,347 bp) according to the gene information in Sol Genomics Network (SGN) (Fernandez-Pozo et al. 2015). This suggests NC

10204 is segregating for major fruit size locus.

Three color attributes AvgR, AvgL, and AvgB were mapped in chromosomes 4, 9, and 6 respectively. AvgR and AvgL were correlated (p<0.0001), whereas AvgB was correlated

(p<0.0001) with both AvgR and AvgL. Previously, QTL affecting external color (L, a, b) have been identified on chromosome 2, 4 (1-5cM near marker TG287), and 9 (1cM near the marker

CT032) in an intra-specific population (Saliba-Colombani et al., 2001). Causse et al. (2002) also detected tomato fruit color QTL on chromosome 4 (66 cM). Celik et al. (2017) detected QTL for external tomato color on chromosomes 1 (67.6 to 67.8 Mb) and 2 (62.3Mb to 62.5 Mb). The

QTL for AvgR in our study was detected on chromosome 4 at 59.5cM in the genetic map and at

~58Mb in the physical map that explained up to 9% of phenotypic variance. Another study identified a QTL for ‘L’ on chromosome 9 at ~23Mb on the physical map that explained 4% phenotypic variation in an association mapping study using InDel markers (Liu et al., 2017). In our study, the average L was mapped to opposite ends of chromosome 9 in two generations at

1.6cM (physical map positions of ~0.7Mb) in F5:6 generation and 33.8cM (physical map position of ~42-43Mb) in F2:3 generation, which explained 12% and 14% of phenotypic variances respectively. The detected positions for the QTL on chromosome 9 in our study were different

164

from that identified by Liu et al. (2017). Based on the above information QTL associated with color detected in the present study from chromosome 4 and 9 may be novel.

The QTL for Average B on chromosome 6 explained the highest percentage of phenotypic variation (14% in F2:3 and 21% in F5:6) of all significant color trait QTL. This QTL was mapped to 14cM in genetic map and 42Mb in the physical map on chromosome 6, which was close to the chromoplast specific LcyB gene (Solyc06g074240) controlling old-gold- crimson (ogc) phenotype located on 45897500 to 45898000 bp on chromosome 6 in tomato according to the gene information in SGN (Fernandez-Pozo et al. 2015). NC 10204 is segregating for the crimson gene (ogc) on chromosome 6, so the QTL segregating for AvgB in this study could be allelic to the ogc. Other studies also identified QTL for color on chromosome

6 (Fulton et al. 1997; Ronen et al., 2000).

Our study identified novel QTL controlling fruit shape attributes on chromosome 10 and

12. At the same time, our study also identified the previously detected genetic loci controlling fruit size and color in inter-specific tomato population. This information will be useful to resolve fruit shape variation in cultivated tomato species fully.

165

REFERENCES

Agricultural Marketing Resource Center. 2018. Tomatoes. Retrieved from

https://www.agmrc.org/commodities-products/vegetables/tomatoes

Barrett, D. M., Beaulieu, J. C., and Shewfelt, R. 2010. Color, flavor, texture, and nutritional

quality of fresh-cut fruits and vegetables: desirable levels, instrumental and sensory

measurement, and the effects of processing. Critical Reviews in Food science and

Nutrition 50:369-389.

Brewer, M. T., Lang, L., Fujimura, K., Dujmovic, N., Gray, S., and van der Knaap E. 2006.

Development of a controlled vocabulary and software application to analyze fruit shape

variation in tomato and other plant species. Plant Physiology 141:15-25.

Brewer, M. T., Moyseenko, J. B., Monforte, A. J., and van der Knaap, E. 2007. Morphological

variation in tomato: a comprehensive study of quantitative trait loci controlling fruit

shape and development. Journal of Experimental Botany 58:1339-1349.

Causse, M., Saliba‐Colombani, V., Lecomte, L., Duffe, P., Rousselle, P., and Buret, M. 2002.

QTL analysis of fruit quality in fresh market tomato: a few chromosome regions control

the variation of sensory and instrumental traits. Journal of Experimental Botany 53:2089-

2098.

Celik, I., Gurbuz, N., Uncu, A. T., Frary, A., and Doganlar, S. 2017. Genome-wide SNP

discovery and QTL mapping for fruit quality traits in inbred backcross lines (IBLs) of

Solanum pimpinellifolium using genotyping by sequencing. BMC Genomics 18:1.

Chakrabarti, M., Zhang, N. A., Sauvage, C., Muños, S., Blanca, J., Cañizares, J., Diez, M. J.,

Schneider, R., Mazourek, M., McClead, J., and Causse, M. 2013. A cytochrome P450

regulates a domestication trait in cultivated tomato. Proceedings of the National Academy

of Sciences 110:17125-17130.

166

Chusreeaeom, K., Ariizumi, T., Asamizu, E., Okabe, Y., Shirasawa, K., and Ezura, H. 2014. A

novel tomato mutant, Solanum lycopersicum elongated fruit1 (Slelf1), exhibits an

elongated fruit shape caused by increased cell layers in the proximal region of the ovary.

Molecular genetics and genomics 289: 399-409.

Clevenger J. 2012. Metabolic and genomic analysis of elongated fruit shape in tomato (Solanum

lycopersicum). M.S. thesis dissertation, Ohio State University, OH, USA.

Cong, B., Barrero, L. S, and Tanksley, S. D. 2008. Regulatory change in YABBY-like

transcription factor led to evolution of extreme fruit size during tomato domestication.

Nature Genetics 40:800–804.

Fernandez-Pozo, N., Menda, N., Edwards, J.D., Saha, S., Tecle, I. Y., Strickler, S. R.,

Bombarely, A., Fisher-York, T., Pujar, A., Foerster, H., and Yan, A. 2015. The Sol

Genomics Network (SGN)—from genotype to phenotype to breeding. Nucleic Acids

Research 43:D1036–D1041. pmid:25428362.

Frary, A., Nesbitt, T. C., Frary, A., Grandillo, S., Van Der Knaap, E., Cong, B., Liu, J., Meller,

J., Elber, R., Alpert, K. B., and Tanksley, S. D. 2000. fw2. 2: a quantitative trait locus key

to the evolution of tomato fruit size. Science 289:85-88.

Fulton, T. M., Beck-Bunn, T., Emmatty, D., Eshed, Y., Lopez, J., Petiard, V., Uhlig, J., Zamir,

D., and Tanksley, S. D. 1997. QTL analysis of an advanced backcross of Lycopersicon

peruvianum to the cultivated tomato and comparisons with QTLs found in other wild

species. Theoretical and Applied Genetics 95:881-894.

Gardner, R. G., and Panthee, D.R. 2010. 'Plum Regal' Fresh-market plum tomato hybrid and its

parents, NC 25P and NC 30P. HortScience 45:824-825.

167

Gonzalo, M. J., and Van Der Knaap, E. 2008. A comparative analysis into the genetic bases of

morphology in tomato varieties exhibiting elongated fruit shape. Theoretical and Applied

Genetics 116:647-656.

Grandillo, S., Ku, H. M., and Tanksley, S. D. 1996. Characterization of fs8. 1, a major QTL

influencing fruit shape in tomato. Molecular Breeding 2:251-260.

Huang, Z., Van Houten, J., Gonzalez, G., Xiao, H., and van der Knaap, E. 2013. Genome-wide

identification, phylogeny and expression analysis of SUN, OFP and YABBY gene family

in tomato. Molecular genetics and genomics 288:111-129.

Ivors, K. 2010. Commercial production of staked tomatoes in the Southeast. North Carolina State

Univ. Coop. Ext., Raleigh, NC.

Kabelka, E., Franchino, B., and Francis, D. 2002. Two loci from Lycopersicon hirsutum LA407

confer resistance to strains of Clavibacter michiganensis subsp. michiganensis.

Phytopathology 92:504-10.

Kosambi, D. D. 1944. The estimation of map distances from recombination values. Annals of

Eugenics. 12:172-175.

Ku, H. M., Grandillo, S., and Tanksley, S. D. 2000. fs8. 1, a major QTL, sets the pattern of

tomato carpel shape well before anthesis. Theoretical and Applied Genetics 101:873-878.

Lecomte, L., Duffé, P., Buret, M., Servin, B., and Causse, M. 2004. Marker-assisted

introgression of five QTLs controlling fruit quality traits into three tomato lines revealed

interactions between QTLs and genetic backgrounds. Theoretical and Applied Genetics

109:658-68.

Levin, I., and Schaffer, A. A. 2013. Molecular mapping of complex traits in tomato. In Genetics,

Genomics, and Breeding of Tomato (eds Liedl, B. E. et al.) Ch. 5.

168

Lin, T., Zhu, G., Zhang, J., Xu, X., Yu, Q., Zheng, Z., Zhang, Z., Lun, Y., Li, S., and Wang, X.

2014. Genomic analyses provide insights into the history of tomato breeding. Nature

Genetics 46:1220.

Liu, Y. S., Gur, A., Ronen, G., Causse, M., Damidaux, R., Buret, M., Hirschberg, J., and Zamir,

D. 2003. There is more to tomato fruit colour than candidate carotenoid genes. Plant

Biotechnology Journal 1:195-207.

Liu, J., Van Eck, J., Cong, B., and Tanksley, S. D. 2002. A new class of regulatory genes

underlying the cause of pear-shaped tomato fruit. Proceedings of the National Academy

of Sciences 99:13302-13306.

Liu, X., Geng, X., Zhang, H., Shen, H., and Yang, W. 2017. Association and genetic

identification of loci for four fruit traits in tomato using InDel markers. Frontiers in Plant

Science 8:1269.

Merk, H. L., Ashrafi, H., and Foolad, M. R. 2012. Selective genotyping to identify late blight

resistance genes in an accession of the tomato wild species Solanum pimpinellifolium.

Euphytica. 187:63–75.

Muños, S., Ranc, N., Botton, E., Bérard, A., Rolland, S., Duffé, P., Carretero, Y., Le Paslier,

M.C., Delalande, C., Bouzayen, M., and Brunel, D. 2011. Increase in tomato locule

number is controlled by two SNPs located near WUSCHEL. Plant Physiology 156:2244-

2254.

Paran, I., and van der Knaap, E. 2007. Genetic and molecular regulation of fruit and plant

domestication traits in tomato and pepper. Journal of Experimental Botany 58:3841-3852.

Pascual, L., Desplat, N., Huang, B. E., Desgroux, A., Bruguier, L., Bouchet, J. P., Le, Q. H.,

Chauchard, B., Verschave, P., and Causse, M., 2015. Potential of a tomato MAGIC

169

population to decipher the genetic control of quantitative traits and detect causal variants

in the resequencing era. Plant Biotechnology Journal 13:565-577.

Piombino, P., Sinesio, F., Moneta, E., Cammareri, M., Genovese, A., Lisanti, M. T., Mogno, M.

R., Peparaio, M., Termolino, and P., Moio, L. 2013. Investigating physicochemical,

volatile and sensory parameters playing a positive or a negative role on tomato liking.

Food Research International 50:409-19.

Rodríguez, G. R., Kim, H. J., and van der Knaap, E. 2013. Mapping of two suppressors of

OVATE (sov) loci in tomato. Heredity 111:256.

Rodriguez, G. R., Moyseenko, J. B., Robbins, M. D., Morejon, N. H., Francis, D. M., and van

der Knaap, E. 2010a. Tomato analyzer: A useful software application to collect accurate

and detailed morphological and colorimetric data from two-dimensional objects. Journal

of Visualized Experiments 1856. doi(37):10.3791/1856.

Rodríguez, G. R., Muños, S., Anderson, C., Sim, S. C., Michel, A., Causse, M., Gardener, B. M.,

Francis, D., and van der Knaap, E. 2011. Distribution of SUN, OVATE, LC and FAS in

the tomato germplasm and the relationship to fruit shape diversity. Plant Physiology pp-

110.

Rodriguez, G. R., Strecker, J., Brewer, M. T., Gonzalo, M. J., Anderson, C., Lang, L., Sullivan,

D., Wagner, E., Strecker, B., Drushal, R., et al. 2010b. Tomato analyzer user manual

version 3. 3rd ed. Ohio State University.

Ronen, G., Carmel-Goren, L., Zamir, D., and Hirschberg, J. 2000. An alternative pathway to

beta-carotene formation in plant chromoplasts discovered by map-based cloning of beta

and old-gold color mutations in tomato. Proceedings of the National Academy of

Sciences of the United States of America 97:11102-7

170

Saliba-Colombani, V., Causse, M., Langlois, D., Philouze, J., and Buret, M., 2001. Genetic

analysis of organoleptic quality in fresh market tomato. 1. Mapping QTLs for physical

and chemical traits. Theoretical and Applied Genetics 102:259-272.

Sim, S., Durstewitz, G., Plieske, J., Wieseke, R., Ganal, M. W., Van Deynze, A., Hamilton, J. P.,

Buell, C. R., Causse, M., and Wijeratne, S. 2012. Development of a large SNP

genotyping array and generation of high-density genetic maps in tomato. PLoS One

7:e40563

Stam, P. 1993. Construction of integrated genetic linkage maps by means of a new computer

package: Join Map. The Plant Journal. 3:739-744. van der Knaap, E., Chakrabarti, M., Chu, Y., Clevenger, J. P., Illa-Berenguer, E., Huang, Z.,

Keyhaninejad, N., Mu, Q., Sun, L., Wang, Y., and Wu, S. 2014. What lies beyond the

eye: the molecular mechanisms regulating tomato fruit weight and shape. Frontiers in

Plant Science 5:227. doi:10.3389/fpls.2014.00227.

Van Ooijen, J. W. 2006. Joinmap 4.0, Software for the calculation of genetic linkage maps in

experimental populations. Plant Research International, Wageningen, Netherlands.

Visa, S., Cao, C., Gardener, B.M., and van der Knaap, E. 2014. Modeling of tomato fruits into

nine shape categories using elliptic fourier shape modeling and Bayesian classification of

contour morphometric data. Euphytica 200:429-439.

Wang, L., Li, J., Zhao, J., and He, C., 2015. Evolutionary developmental genetics of fruit

morphological variation within the Solanaceae. Frontiers in Plant Science 6:248.

Würschum, T. 2012. Mapping QTL for agronomic traits in breeding populations. Theoretical and

Applied Genetics 125:201-210.

171

Xiao, H., Jiang, N., Schaffner, E., Stockinger, E. J., and van der Knaap, E., 2008. A

retrotransposon-mediated gene duplication underlies morphological variation of tomato

fruit. Science 319:1527-1530.

172

CHAPTER 5: FUTURE DIRECTIONS FOR THE CURRENT RESEARCH ON

BACTERIAL SPOT DISEASE RESISTANCE IN TOMATO

Bacterial spot disease has remained as an unsolved problem for many years in tomato fields. The widespread existence of copper resistant bacterial spot pathogen in tomato production fields of NC necessitated the introgression of host resistance against this disease (Chapter 2).

Despite several breeding efforts, there is no commercial resistant tomato cultivar available yet.

The emergence of new race before the deployment of resistance against the current race has complicated the process to achieve durable host resistance. The race T4 has persisted in

Southeast US for 20 years and is also predominant in the tomato production regions of North

Carolina (Chapter 2). So far, one resistant gene RxopJ4 has been identified to this race in

Solanum pennelli LA716, but the resistant gene is linked to several negative traits (Sharlach et al.

2013). In the present study, we identified and validated three major quantitative trait loci (QTL) on chromosome 1, 4, and 6 controlling bacterial spot resistance in tomato against race T4

(Chapter 3). Among them, QTL on chromosome 6 explained the substantial percentage of phenotypic variance, i.e. upto 26% in NC 10204 and 23% in NC 13666 population. Since, the donor source of this QTL in NC 10204 is NC 30P, a released breeding line, this QTL will be useful in the breeding program without much concern of linkage drag. The QTL on chromosome

1 also explained up to 23% of phenotypic variance in NC 10204 population, however, the positions of QTL were different in different environments (Chapter 3). Therefore, it is necessary to verify if there are different QTL on chromomse 1 or the QTL is heavily affected by the environment for the successful utilization of this QTL in the breeding program.

173

What are the Next Steps? The first step would be to confirm the QTL effect on chromomse 6 as this QTL has explained the largest percentage of the phenotypic variance observed in NC 10204 population.

Although the QTL on chromosome 6 detected in this study was located near RxopJ4 gene

(Chapter 3), there is also a possibility that the QTL is associated with the plant growth habit.

Since NC 10204 was segregating for the plant growth habit, indivual plant with indeteminate growth habit might appear resistant compared to determinate growth habit. A study has reported that the bacterial spot disease resistance against race T1 was affected by the plant growth habit

(Yang et al., 2005).

The self-pruning (sp) gene is also located on chromosome 6 at the position of ~46 Mb according to the gene information available on Sol Genomics Network (SGN) (Fernandez-Pozo et al. 2015), which is close to the position of the QTL for bacterial spot disease resistance in our study i.e. 37Mb to 42Mb (Chapter 3). In the previous study, the QTL controlling growth habit was mapped to chromomse 6 in NC 10204 population and was linked with the marker solcap_snp_sl_57352 (McNellie, 2015). The same marker was linked with the bacterial spot disease resistance in this study in MHCREC in 2017 (Chapter 3). Therefore, the possibility that the bacterial spot disease resistance to be asscociated with the indeterminate growth habit cannot be overlooked. This necessitates to confirm further if the QTL on chromosome 6 detected in this study is segregating for the bacterial spot disease resistance or the growth habit or both. For this, a subset of most resistant and most susceptible lines can be screened with the marker associated with the sp gene to test if the bacterial spot resistance is associated with the sp gene. Once, the effect of QTL on chromosome 6 is confirmed, then the position of this QTL can be narrowed down and then introgress this resistance into other breeding lines.

Next, the QTL on chromosome 1 were detected in different positions in different environment. Therefore, the QTL on chromosome 1 need further verification if there are two 174

different QTL or the environment is influencing the trait. For this, we need to phenotype the population again at least in one environment and conduct a QTL analysis.

After we confim the QTL effects on chromosome 1 and 6, these QTL can be stacked wih other major resistant loci on chromosomes 3 and 11 against race T4 identified by Hutton et al.

(2010) to obtain durable resistance. The durability of the host resistance depends on the evolutionary potential of the pathogen, their mutation rate and ability to transfer genetic materials, selection pressure and relative fitness of the virulent strains (Stall et al., 2009). The stacking of multiple resistance loci will require pathogens to mutate in multiple loci, which will hold the resistance in the host for more extended period.

Another area for the future research would be to study the effectors present in the

Xanthomonas strains in NC. In our study, we studied the race profile of Xanthomonas strains in

NC through their response on tomato and pepper differentials (Chapter 2), but not the diversity of the effector repetoires present in them. Therefore, studying the effectors present in the

Xanthomonas strains in the future will help to design strategies to develop disease resistance targeting the core effectors (Schwartz et al., 2015). For instance, a type III secretion factor belonging to XopJ effector family, avrBsT was detected among X. perforans and some strains of

X. euvesicatoria and X. vesicatoria, which was found to have a role in the pathogen fitness under the field condition (Abrahamian et al., 2018). This suggests, avrBsT might be a suitable candidate to target X. perforans in the regions where this species is predominant. Therefore, breeding approaches targeting evolutionary conserved bacterial effector in combination with resistant gene pyramiding could provide durable resistance against bacterial spot pathogens.

In recent years, due to the technological advancements, the repertoire of core effectors of the pathogens and Nucleotide-Binding Domain Leucine-Rich Repeat (NLR)’ome of the tomato plant are available. This has opened the pathway to identify R genes targeting the core effectors

175

of the pathogens. A pangenome of tomato will be available in the near future and will explore the genetic variation in the available germplasm and add more R genes in the repertoire of NLRome of tomato. This will assist in identifying multiple NLRs to recognize core effectors of

Xanthomonas causing bacterial spot in tomato. Advancement in the computational tools will facilitate the identification of candidate effectors, while resistance gene enrichment sequencing

(RenSeq) will identify multiple NLRs that recognizes core pathogen effectors to achieve disease resistance (Zhang and Coaker, 2017). Both core effectors and NLRs can be transiently expressed in the tomato plant, and the HR phenotype indicates the recognition of effector genes by NLR.

Development of high throughput functional screening system will further accelerate the identification of promising NLRs to recognize the core effectors and inform R gene deployment before the pathogen evolution (Zhang and Coaker, 2017).

The multiple NLRs effective against the core effector of Xanthomonas can be stacked into single breeding lines and cultivars in different combinations to increase the durability of the resistance. This will require pathogens to undergo significant evolutionary hurdles to overcome such resistance. In addition, other sources of resistance such as quantitative, recessive, and immune receptors can be integrated to implement a multi-tiered strategy for durable disease resistance (Zhang and Coaker, 2017). In future, tomato breeding programs should aim to identify and utilize tomato NLRs recognizing the core effectors of Xanthomonas to achieve the effective and durable resistance against bacterial spot disease in tomato.

Similarly, comparative genomics has enabled capacity to compare the genomes of closely related species and study the conserved genes among different genus within the same family.

Additionally, emerging technology has enabled precise genome editing of the plant genome to insert or mutate the gene of interest. For instance, genome editing technology using clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9 would be ideal to integrate

176

NLRs from wild species into a cultivated tomato (Zhang and Coaker, 2017). Likewise, tomato breeders can use the CRISPR/Cas9 system to edit the Bs2 syntenic gene in tomato to make it function as the Bs2 gene, so that tomato lines with Bs2 gene could be commercialized, unlike the trangenic Bs2 tomato lines.

Similarly, the CRISPR/Cas9 system has also potential to integrate bs5, bs6, Bs3 and Bs7 genes from pepper into tomato once these genes are cloned. CRISPR/Cas9 holds promise in effectors-based breeding to stack multiple NLRs in a single breeding line. Pan et al. (2016) have demonstrated the efficiency of the CRISPR/Cas9 system to generate stable and heritable mutations in tomato plants. In this study, the CRISPR/Cas9 induced mutagenesis in tomato plants was highly specific without any off-targets (Pan et al., 2016), suggesting the potential use of CRISPR technique in tomatoes to develop disease resistance tomato lines. With all these new technologies available, we can expect a tomato breeding program in the future will succeed to develop bacterial spot disease resistant tomato cultivar.

177

REFERENCES

Abrahamian, P., Timilsina, S., Minsavage, G. V., KC, S., Goss, E. M., Jones, J. B., and Vallad,

G. E. (2018). The type III effector AvrBsT enhances Xanthomonas perforans fitness in

field-grown tomato. Phytopathology. https://doi.org/10.1094/PHYTO-02-18-0052-R.

Fernandez-Pozo, N., Menda, N., Edwards, J. D., Saha, S., Tecle, I. Y., Strickler, S. R.,

Bombarely, A., Fisher-York, T., Pujar, A., Foerster, H., Yan, A. (2015) The Sol

Genomics Network (SGN)—from genotype to phenotype to breeding. Nucleic Acids

Research 43:D1036–D1041. pmid:25428362.

Hutton, S. F., Scott, J. W., Yang, W., Sim, S. C., Francis, D. M., and Jones, J. B. (2010).

Identification of QTL associated with resistance to bacterial spot race T4 in tomato.

Theoretical and Applied Genetics 121:1275-1287.

McNellie, J.P. 2015. Mapping QTLs for Fruit Quality and Horticultural Traits in Fresh Market

Tomato (Solanum lycopersicum) Using Intra-specific SNP Markers. Master’s Thesis.

North Carolina State University

Pan, C., Ye, L., Qin, L., Liu, X., He, Y., Wang, J., Chen, L., and Lu, G. (2016). CRISPR/Cas9-

mediated efficient and heritable targeted mutagenesis in tomato plants in the first and

later generations. Scientific Reports 6:24765.

Schwartz, A. R., Potnis, N., Timilsina, S., Wilson, M., Patané, J., Martins, J. Jr., Minsavage, G.

V., Dahlbeck, D., Akhunova, A., Almeida, N., Vallad, G. E., Barak, J. D., White, F. F.,

Miller, S. A., Ritchie, D., Goss, E., Bart, R. S., Setubal, J. C., Jones, J. B, and Staskawicz,

B. J. (2015). Phylogenomics of Xanthomonas field strains infecting pepper and tomato

reveals diversity in effector repertoires and identifies determinants of host specificity.

Frontiers in Microbiology 6:535.

178

Sharlach, M., Dahlbeck, D., Liu, L., Chiu, J., Jiménez-Gómez, J. M., Kimura, S., Koenig, D.,

Maloof, J. N., Sinha, N., Minsavage, G. V., and Jones, J. B. (2013). Fine genetic mapping

of RXopJ4, a bacterial spot disease resistance locus from Solanum pennellii LA716.

Theoretical and Applied Genetics 126:601-609.

Stall, R. E., Jones, J. B., and Minsavage, G. V. (2009). Durability of resistance in tomato and

pepper to xanthomonads causing bacterial spot. Annual Review of Phytopathology

47:265-284.

Yang, W., Sacks, E. J., Lewis Ivey, M. L., Miller, S. A., and Francis, D. M. (2005). Resistance in

Lycopersicon esculentum intraspecific crosses to race T1 strains of Xanthomonas

campestris pv. vesicatoria causing bacterial spot of tomato. Phytopathology 95:519-527.

Zhang, M., and Coaker, G. (2017). Harnessing effector-triggered immunity for durable disease

resistance. Phytopathology, 107:912-919.

179