Molecular Characterization of Spinach (Spinacia

MOLECULAR CHARACTERIZATION OF SPINACH (SPINACIA

OLERACEA) MICROBIAL COMMUNITY STRUCTURE AND ITS

INTERACTION WITH ESCHERICHIA COLI O157:H7 IN MODIFIED

ATMOSPHERE CONDITIONS

Gabriela Lopez-Velasco

Dissertation submitted to the faculty of the Virginia Polytechnic Institute and State University in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

Food Science and Technology

Monica A. Ponder, Committee Chair.

Renee R. Boyer, Committee Member.

Joseph O. Falkinham III, Committee Member.

Gregory Welbaum, Committee Member.

Robert C. Williams, Committee Member.

March 31, 2010

Blacksburg, VA

Keywords: spinach, microbial diversity, bacterial interaction, microbial community.

OLERACEA) MICROBIAL COMMUNITY STRUCTURE AND ITS

INTERACTION WITH ESCHERICHIA COLI O157:H7 IN MODIFIED

ATMOSPHERE CONDITIONS

Gabriela Lopez-Velasco

ABSTRACT

Leafy greens like lettuce and spinach are a common vehicle for foodborne illness in United

States. It is unknown if native plant epiphytic bacteria may play a role in the establishment of enteric pathogens on leaf surfaces. The objective of this study was to characterize the bacterial communities of fresh and packaged spinach leaves and to explore interactions with E. coli O157:H7. We assessed the bacterial diversity present on the spinach leaf surfaces and how parameters such as spinach cultivar, field conditions, post-harvest operations and the presence of E. coli O157:H7 affected its diversity.

Differences in bacterial population size and species richness were associated with differences in plant topography; flat leaves had smaller bacterial populations than savoy leaves, which correlated with larger number of stomata and trichomes in savoy leaves. During spinach growing season shifts in environmental conditions affected richness and population size of the spinach bacterial community.

Decreases in the overall soil and ambient temperature and increased rainfall decreased richness and bacterial population size.

Fresh spinach richness and composition assessed by parallel pyrosequencing of 16S rRNA elucidated 600 operational taxonomic units, with 11 different bacterial phyla. During postharvest operations diversity indexes and evenness tended to decrease, likely attributed to storage at low temperature and time of storage (4°C and 10°C), that promoted the dominance of -Proteobacteria. Bacteria isolated from fresh spinach elicited growth inhibition of E. coli O157:H7 in vitro, which was associated with nutrient competition. In contrast growth enhancement produced by epiphytes was associated to low correlations in carbon source utilization and the ability of E. coli O157:H7 to rapidly utilize carbon resources. In packaged spinach, E. coli O157:H7 altered the composition of the bacterial community and its growth was promoted on packaged spinach when a disinfection and temperature abuse occurred, removal of the epiphytic bacteria resulted in significant increases in numbers of E. coli

O157:H7 at 10°C and was associated with increased expression of E. coli O157:H7 virulence and stress response genes. The large diversity present on the surface of spinach leaves significantly impacted the ecology of enteric pathogens like E. coli O157:H7 on the phyllosphere.

iii

ACKNOWLEDGEMENTS

I would like to thank my advisor, Dr. Monica Ponder for all your guidance and support. I learned a lot from you and with you. Thank you for sharing your experience with me and for giving me the opportunity of being your student.

To my committee: Dr. Joe, Dr. Greg, Dr. Renee and Dr. Rob. Thank you for always being available to talk and discuss my project, for all your support and guidance. I really enjoyed all the moments we shared.

To my lab mates: Heather Totty, Heather Tydings, Phyllis Carder, Peter Ziegler, Twyla Smith and

Megan Shepherd. It was a great experience to work with each one of you. Thank you for your friendship and support.

To the National Council for Science and Technology (CONACyT) that provided financial support for my postgraduate education.

To all students, staff and professors from the Department of Food Science and Technology and the

Fralin Life Institute, thank you for the great environment that made my graduate studies a great experience. I would like to thank to Dr. Marcy, Dr. Sumner, Jennifer, Julie, Joell, Trina, Kim and Walter for all your help and support.

To all my friends in Blacksburg, Umesh, Elaine, Tyler, Roberto, Seba, Heather, Gabi, Laurita,

Agustin, Naiara, Hedie, Marianne, Chucho, Zara, Jose, David, Andrea, Fulvia and Gulden. This would not have been possible without your friendship. I love you very much.

To my family in Mexico, Alma, Gonzo, Jorge, Miri, Margarita, Rafael, for all your love and support.

DEDICATORY

To Jose Manuel, I do not have words to thank you all what we have lived together here. Thank you for being always so supporting, for sharing with me this great experience, for your patience and for your love.

I love you very much.

ATTRIBUTIONS

Dr. Monica Ponder was the principal advisor in the development of this research project, providing advice, guidance and resources.

Dr. Joseph Falkinham III provided important input and guidance in the development of techniques to screen bacterial interactions with E. coli O157:H7 and spinach phyllosphere as well and provided his expertise in microbial ecology.

Dr. Gregory Welbaum provided his expertise in horticultural crops and assistance in spinach production and analysis of spinach anatomical features.

Dr. Renee R. Boyer and Dr. Robert C. Williams provided guidance and assistance in the experiments associated with bagged spinach and the behavior of E. coli O157:H7 and provided their expertise in food safety.

Brinkley Benson actively participated in the production of spinach during Fall of 2007 and Spring of 2008.

Dr. Glenda Gillaspy provided her expertise and assistance in the utilization of DIC microscopy.

Dr. Jose Herrera-Alonso and MS. Roberto Franco collaborated with spinach harvest and irrigation operations during the spinach production.

MS. Gabriela Marquez, MS. Sebastian Arriola and MS. Ciro Velasco provided advise in statistical analysis.

Mss. Heather Tydings was involved in the collection and screening of isolates from spinach during the summer of 2008 as undergraduate intern. She actively participated in the media preparation and packaging of spinach.

Dr. Mane Shrinivasrao, Dr. Monica Ponder and Dr. Clive Evans provided their expertise in the analysis of sequencing obtained from 454-pyrosequencing of 16S rRNA microbiomes gene amplicons from spinach, by providing guidance and assistance in metagenomic analysis.

vii

Table of contents Page

Abstract ii

Acknowledgements iv

Dedicatory v

Attributions vi

Table of contents viii

List of tables xiii

List of figures xvi

Chapter 1. Introduction and Justification 1

References 2

Chapter 2. Literature review 4

Produce as part of a healthy diet 4

Trends in produce consumption 4

Produce and food safety 5

Outbreaks associated with the consumption of produce items 5

E. coli O157:H7 outbreaks 9

Consequences of foodborne illnesses 9

The reality of food safety 10

E. coli O157:H7 as human pathogen 11

Mechanisms of pathogenicity of EHEC 11

Emergence of recently recognized foodborne pathogens 13

Changes in transmission vehicles associated with microbial adaptation 13

Societal factors associated with emergence of foodborne pathogens 14

Globalization of the food supply 14

viii

Factors that contribute to the increased emergence of E. coli O157:H7 as foodborne 15

pathogen

Contamination of produce with Escherichia coli O157:H7during pre-harvest operations 16

Animal Manure 18

Irrigation Water 19

Contaminated soils 20

Wild life and other sources of produce contamination 21

Ecology of foodborne pathogens on plant surfaces 21

Persistence of foodborne pathogens on edible plants surfaces 22

Establishment of foodborne pathogens on edible plants 23

Plant surface as environment for bacterial growth 23

Factors that contribute to plant colonization and fitness of the phylloepiphytic bacteria 25

and enteric pathogens

26 Survival strategies of phylloepiphytic bacteria on leaf surfaces under environmental

stress conditions

28 Overlapping mechanisms for survival and establishment of enteric pathogens on plant

surfaces

Interactions of enteric pathogens with the phyllosphere members 30

34 Biodiversity of the phyllosphere

Study of microbial communities 34

38 Fingerprinting techniques to study microbial communities: Denaturant gradient gel

electrophoresis

Phyla specific amplification through real-time PCR 39

Pyrosequencing 40

Studies of diversity on the phyllosphere 41

Studies of population dynamics on the phyllosphere 46

Spinach as vegetable crop 47

Botany and production requirements 47

Harvest and post-harvest of spinach 49

49 Spinach cultivars

Spinach production in United States 50

Trends of spinach consumption 51

Produce post-harvest and implications in food safety 51

Strategies to reduce microbial load and microbial growth on post-harvested produce 55

Preventive strategies to minimize the risk of produce contamination with foodborne pathogens 60

References 62

Chapter 3. Phylloepiphytic microbial community structure on spinach (Spinacia oleracea) 91 leaves at harvest as affected by cultivar and environment

Abstract 91

Introduction 93

Materials and Methods 94

Results 99

Discussion 101

Acknowledgements 104

References 105

Chapter 4. Growth of E. coli O157:H7 is influenced by interactions with spinach 119 phylloepiphytic bacteria

Abstract 119

Introduction 120

Materials and Methods 121

Results 125

Discussion 127

Acknowledgements 131

References 132

Chapter 5. Alterations of the phylloepiphytic bacterial community associated with 147 interactions of E. coli O157:H7 during storage of packaged spinach at refrigeration temperatures

Abstract 148

Introduction 149

Materials and Methods 151

Results and Discussion 158

Conclusion 168

Acknowledgements 169

References 170

Chapter 6. Pyrosequencing of 16S rRNA gene amplicons for characterization of the 187 phylloepiphytic bacteria of minimally processed spinach stored at refrigeration temperatures

Abstract 187

Introduction 188

Materials and methods 190

Results 193

Discussion 194

Conclusion 199

Acknowledgements 200

References 201

Chapter 7. Conclusion and future directions 217

Appendix

Appendix A. Comparison of DNA isolation methodologies for microbes associated 219

with the surface of spinach leaves

Appendix B. Differences in population size and microbial community composition 223

among spinach cultivars

Appendix C. Pyrosequencing of 16S rRNA microbiomes of spinach at various stages 228

of plant growth (from seed to young leaves)

xii

List of tables

Title Page

Chapter 2. Literature Review

Table 1. Produce associated outbreaks that occurred from 2000 to 2008 6

Table 2. Microbial interactions that can occur within a microbial community 30

32 Table 3. Bacterial species isolated from produce items that elicit antagonism

towards foodborne pathogens

43 Table 4. Representative phyla and genera of bacteria identified through 16S rRNA

analysis on the phyllosphere

Table 5. Summary of post-harvest procedures utilized on fresh produce to extend 56

shelf-life

Table 6. Fate of foodborne pathogens after application of technologies to extend 58

produce shelf life

Chapter 3. Phylloepiphytic microbial community structure on spinach (Spinacia oleracea) leaves at harvest as affected by cultivar and environment

Table 1. Environmental conditions registered in Blacksburg, VA from the time of 109

seeding to the last harvest

Table 2. Number of culturable epiphytic bacteria on baby spinach leaves for the 110

three spinach cultivars harvested at three different time points as determined by

plating on R2A media

Table 3. Morphological features of spinach cultivars 111

Chapter 4. Growth of E. coli O157:H7 is influenced by interactions with spinach phylloepiphytic bacteria

xiii

Table 1. Identified microorganisms that elicited in vitro interactions with E. coli 136

O157:H7

Table 2. Growth rate and yield of E. coli O157:H7 grown on spent media utilized 137

by phylloepiphytic spinach isolates.

Table 3. Pearson’s correlation between E. coli O157:H7 and phylloepiphytic 138

isolates in carbon source utilization

Table 4. Co-culture of E. coli O157:H7 with bacterial isolates from spinach 139

phyllosphere

140 Supplementary Table 1. Values of optical density of different carbon sources

utilized by E. coli O157:H7 and phylloepiphytic bacteria from BIOLOG® plates.

Chapter 5. Alterations of the phylloepiphytic bacterial community associated with interactions of E. coli O157:H7 during storage of packaged spinach at refrigeration temperatures

Table 1. E. coli O157:H7 specific primers for real time PCR using SYBR green 176

chemistry designed for this study

Table 2. Total culturable epiphytic bacteria on rinsed or commercial non- 177

inoculated spinach packaged under simulated retail conditions and, stored at 4oC

and 10oC for 15 days. (Enumeration on R2A media at 25oC incubated for 48h)

Table 3. Identities of bacteria retrieved from 16SrDNA DGGE bands of spinach 178

stored at refrigerated temperatures

Table 4. Populations of culturable E. coli O157:H7 artificially inoculated on 179

spinach, comparing samples that were rinsed or from commercial source, and

stored at 4oC and 10oC. (Enumeration on sorbitol MacConkey agar, incubated at

37oC for48 h)

Table 5. Normalized expression of E. coli O157:H7 espA, csgA, sodB, rpoS, rpoN 180

xiv

and stxA genes when inoculated onto spinach and stored at 4oC and 10oC over

a10-day period assessed by qRT-PCR

Supplementary Table 1. Growth of epiphytic microbiota on bagged rinsed or 181

commercial inoculated spinach, stored at 4oC and 10oC for 15 days. (Enumeration

on R2A media at 25oC incubated for 48h)

Chapter 6. Pyrosequencing of 16S rRNA gene amplicons for characterization of the phylloepiphytic bacteria of minimally processed spinach stored at refrigeration temperatures

Table 1. Number of reads, unique sequences and trimmed reads for spinach 208

microbiomes

Table 2. Description of fresh spinach microbiome 209

Table 3. Richness and diversity estimators that predict the number of species in 210

fresh and packaged spinach at 3% of dissimilarity

Table 4. Comparison of spinach microbiomes during storage at 4 and 10°C for up 211

to 15 days

Appendix C. Pyrosequencing of 16S rRNA microbiomes of spinach at various stages of plant growth (from seed to young leaves)

Table C1. Characteristics of the sequenced data 228

Table C2. Richness and diversity estimators for seed, cotyledons and young 229

spinach leaves microbiomes

Table C3. Microbiomes of seeds, cotyledon and young leaves from spinach, 230

cultivar Menorca grown in growth chambers. (Results are represented as relative

abundance of the total classified bacteria)

List of figures

Title Page

Chapter 2. Literature review

Figure 1. Routes of produce contamination with E. coli O157:H7 17

Figure 2. General flowchart for characterization of microbial communities from 37

environmental samples

Figure 3. Principle of DNA fragment separation by PCR-DGGE analysis 39

Chapter 3. Phylloepiphytic microbial community structure on spinach (Spinacia oleracea) leaves at harvest as affected by cultivar and environment

Figure 1. ESEM microscopy of spinach stomata for cultivars (A) Monza, (B) 112

Menorca and (C) Unipak

Figure 2. Analysis of the structure of epiphytic microbial community by DGGE 113

separation of 16S rRNA fragments with 341-f and 907r primers. Lanes: (1) ‘Monza’ –

October-, (2) ‘Menorca’ –October-, (3) ‘Unipak’ –October-, (4) ‘Monza” –November-, (5)

‘Menorca’ –November-, (6) ‘Unipak’ –November-, (7) ‘Monza” –December-, (8)

‘Menorca’ –December- and (9) ‘Unipak’ (December). Band identification: closest match to

NCBI database -accession number and % of similarity-(1) Uncultured proteobacterium

clone -EF602176- 97% - and Uncultured Oxalobacteraceae -FJ037285-97%-. (2)

Kineococcus sp. –FJ214365- 96%-, Uncultured beta proteobacterium -EFO74569- 98%-,

Terrabacter sp. -FJ772036- 99%- and Oryzihumus sp. -GQ355280- 99%-. (3) Uncultured

bacterium clone -FJ562178- 99%- and Uncultured actinobacterium clone -EF651015- 89%.

(4) Uncultured Baceroidetes bacterium -AM116743- 96%- and Flavobacterium sp. -

AY599661- 98%-. (5) Chryseobacterium sp. -EF471218-86%- and Uncultured

Proteobacterium clone -EF019491- 88%- (6) Spinacia oleracea chloroplast -AJ400848-

98%- (7) Uncultured soil bacterium clone -DQ297980- 97%-. (8) Uncultured alpha

xvi

Proteobacteria -CU926667- 95%-

Figure 3. UPGMA dendrogram constructed based on the similarity between 114

DGGE profiles for the three spinach cultivars and the three times of collection

Figure 4A. Abundance of members of five bacterial phyla in epiphytic 115

communities isolated from ‘Monza’ at three different harvest periods, determined

by qPCR. (n=3). Error bars represent standard deviations of the mean. *Tukey

comparison. P<0.05

Figure 4B. Abundance of members of five bacterial phyla in epiphytic 116

communities isolated from ‘Menorca’ at three different harvest periods, determined

by qPCR. (n=3). Error bars represent standard deviations of the mean. *Tukey

comparison. P<0.05

Figure 4C. Abundance of members of five bacterial phyla in epiphytic 117

communities isolated from ‘Unipak’ at three different harvest periods, determined

by qPCR. (n=3). Error bars represent standard deviations of the mean. *Tukey

comparison. P<0.05

Figure 5. Simpson’s diversity indices (D) calculated based on the abundance of 118

members of five phyla in communities isolated from three spinach cultivars

harvested at three different collection times. (Simpson’s diversity index was

represented as 1-D to facilitate the analysis). Independent replicates were used

(n=3). Error bars represent standard deviation of the mean.

Chapter 4. Growth of E. coli O157:H7 is influenced by interactions with spinach phylloepiphytic bacteria

Figure 1. Formation of growth inhibition zones against E. coli O157:H7 by 146

Acinetobacter spp. and Pseudomonas spp.

Chapter 5. Alterations of the phylloepiphytic bacterial community associated with

xvii interactions of E. coli O157:H7 during storage of packaged spinach at refrigeration temperatures

Figure 1A. Percentage of head space pressures of O2 of packaged spinach stored at 182

4 and 10 °C during 15 days. Values represent the means of 10 bags and error bars

represent the standard deviation from the mean.

Figure 1B. Percentage of head space pressures of CO2 of packaged spinach stored 183

at 4 and 10 °C during 15 days. Values represent the means of 10 bags and error

bars represent the standard deviation from the mean

Figure 2. DGGE of PCR-amplified 16S rDNA fragments from non-inoculated 184

packaged spinach. A. Disinfected and stored at 4oC, B. Disinfected and stored at

10oC, C. rinsed and stored at 4°C, D. Rinsed and stored at 10°C. (d0 –day zero-, d5

–day 5, d10 –day 10, d15 –day 15)

Figure 3. UPGMA clustering of 16S rDNA DGGE profiles from inoculated and 185

non-inoculated bagged spinach stored at 4oC or 10oC for a period of 15 days. (R)

Rinsed spinach, (D) Disinfected spinach, (I) inoculated spinach with E. coli

O157:H7, (d0 –day zero-, d5 –day 5-, d10 –day 10-, d15 –day 15-)

Figure 4. DGGE of PCR-amplified 16S rDNA fragments from packaged spinach 186

inoculated with E. coli O157:H7. A. Disinfected and stored at 4oC, B. Disinfected

and stored at 10oC, C. rinsed and stored at 4°C, D. Rinsed and stored at 10°C. (d0 –

day zero-, d5 –day 5-, d10 –day 10-, d15 –day 15-)

Chapter 6. Pyrosequencing of 16S rRNA gene amplicons for characterization of the phylloepiphytic bacteria of minimally processed spinach stored at refrigeration temperatures

Figure 1A-B. Relative phylum/class abundance of classified sequences on fresh 213 and packaged spinach microbiomes

xviii

Figure 2A-B. Rarefraction curves for fresh and packaged spinach indicating the 214

observed number of OTU’s within the 16S rRNA microbiomes derived from

spinach leaves. Curves were calculated with RDP using pyrosequencing tools

Figure 2C-D. Rarefraction curves for packaged spinach indicating the observed 215

number of OTU’s within the 16S rRNA microbiomes derived from spinach leaves.

Curves were calculated with RDP using pyrosequencing tools

Figure 3. Comparison of the four spinach microbiomes. UPGMA utilizing 216

Jaccard’s index at 3% of dissimilarity

Appendix A. Comparison of DNA isolation methodologies for microbes associated with the surface of spinach leaves

Figure A1. DGGE of PCR-amplified 16S rDNA fragments utilizing DNA and 222

RNA converted to cDNA obtained from two different extraction protocols

Appendix B. Differences in population size and microbial community composition among spinach cultivars

Figure B1. Microbial populations of spinach microbiota. Comparison among three 225

spinach cultivars. Psichrotrophic bacteria was determined after incubation of plates

at 4°C for 16 days, molds and yeasts as well as bacterial populations determined in

R2A were determined after incubation of plates at 25°C for 16 days. Proc mix from

SAS software (Statistical Analysis System 9.2 SAS Institute, Cary, NC) was

utilized to determine lsmeans, standard error and comparisons among means for

each treatment, means were separated using Tukey’s multiple comparison.

Significant difference was determined when p-values <0.05 and are indicated with

different letters. All data was previously evaluated for normality and homogeneity

of variance using the proc univariate function of SAS

Figure B2. Comparison of structure of the microbial community among spinach 226

xix

cultivars assessed by DGGE. A. UPGMA clustering of 16S rDNA DGGE profiles.

B. DGGE of PCR-amplified 16S rDNA fragments. Lanes: 1 Monza, 2 Menorca, 3

Unipak (using DNA as template for PCR reaction) 4 Monza, 5 Menorca, 6 Unipak

(using cDNA as template for PCR reaction)

Appendix 3. Pyrosequencing of 16S rRNA microbiomes of spinach at various stages of plant growth (from seed to young leaves)

Figure C1. Relative phylum/class abundance of classified sequences from seed, 234

cotyledon and young spinach leaves

Figure C2. Relative phylum/class abundance of classified sequences from seed, 235

cotyledon and young spinach leaves

Figure C3. Rarefraction curves for spinach seeds 236

Figure C4. Rarefraction curves for spinach cotyledons 237

Figure C5. Rarefraction curves for young spinach leaves 238

CHAPTER 1

INTRODUCTION AND JUSTIFICATION

The demand for fresh and minimally processed vegetables, has led to an increase in the quantity and variety of products that are available to consumers (8). Simultaneously, an increase in outbreaks associated with serotypes of Salmonella enterica, Escherichia coli O157:H7 and Listeria monocytogenes have been reported, particularly in spinach and lettuce (7). This is an important health risk for consumers and could negatively impact economics of the agricultural sector (1). E. coli O157:H7 and S. enterica are enteric pathogens which are known to infect a broad range of hosts, including humans. Livestock, in particular, cattle are considered to be natural reservoirs for these pathogens and infected cattle can shed ~ 50,000 CFU E. coli

O157:H7 per gram of feces (4). The leaves of plants (produce) become exposed (pre harvest) to these organisms when they come in contact with contaminated manure or irrigation waters (4). It is important to understand how these microorganisms are able to survive and establish themselves on leaf surfaces. Native bacteria found on leaf surfaces (phyllosphere) are well adapted to the conditions specific to the microenvironment (5). This environment provides low nutrient and water availability, high exposure to UV radiation and shifts in environmental conditions. The interactions with the plant and these microorganisms as well as microbe-microbe interactions are pivotal for establishment of pathogenic bacteria (5, 6). It is hypothesized that enteric pathogens can also interact with both the leaves and other phyllosphere members, which could contribute to their establishment on edible plants (2). In general, little knowledge about bacterial diversity on the phyllosphere has been developed. The first step towards understanding microbe-microbe interactions is to characterize the microbial community present on the plant surfaces. Spinach has demonstrated its ability to support enteric bacteria growth as showed with a multistate outbreak of E. coli

O157:H7 and therefore makes an excellent model for these studies (3).

The study of microbial communities on the spinach phyllosphere may provide information regarding the structural and functional properties of this community to further understand the impact of microbial

1 interactions that occur in the phyllosphere. In this study we addressed four main points: (i) the characterization of the microbial community on the leaf surfaces of fresh spinach to determine microbial diversity in the phyllosphere, (ii) The influence of spinach growing conditions and spinach cultivars on the structure of the microbial community of fresh spinach leaves, (iii) The interactions of a foodborne pathogen;

E. coli O157:H7 with spinach phyllosphere and (iv) The effect of packaging and storage at refrigeration temperatures on the physiology and survival of E. coli O157:H7 and its interaction with the microbial community of spinach. These objectives will provide a broader outlook of the spinach microbial diversity and its structure during cultivation and exposure to environmental conditions as well as its interaction with E. coli157:H7 during pre- and post- harvest operations.

References

1. Altekruse, S. F., M. L. Cohen, and D. L. Swerdlow. 1997. Emerging foodborne diseases.

Emerg Infect Dis 3:285-93.

2. Brandl, M. T. 2006. Fitness of human enteric pathogens on plants and implications for food

safety. Annu Rev Phytopathol 44:367-92.

3. CDC. 2006. Multistate outbreak of E. coli O157 infections. November-December 2006.

www.cdc.gov/ecoli/2006/december/121006.htm. [Online.]

4. Franz, E., and A. H. van Bruggen. 2008. Ecology of E. coli O157:H7 and Salmonella enterica

in the primary vegetable production chain. Crit Rev Microbiol 34:143-61.

5. Lindow, S. E., and M. T. Brandl. 2003. Microbiology of the phyllosphere. Appl Environ

Microbiol 69:1875-83.

6. Lindow, S. E., and J. H. Leveau. 2002. Phyllosphere microbiology. Curr Opin Biotechnol

13:238-43.

7. Sivapalasingam, S., C. R. Friedman, L. Cohen, and R. V. Tauxe. 2004. Fresh produce: a

growing cause of outbreaks of foodborne illness in the United States, 1973 through 1997. J Food

Prot 67:2342-53.

8. USDA. 2002. Profiling food consumption in America. In USDA (ed.), The fact book 2001-2002.

CHAPTER 2

LITERATURE REVIEW

PRODUCE AS PART OF A HEALTHY DIET

Changing dietary behaviors to prevent chronic diseases like coronary heart illnesses, some types of cancer, stroke, non-insulin dependent diabetes, hypertension and osteoporosis has been an important research area in the last few years. These diseases are mostly the result of a poor diet and a sedentary life style. Dietary changes, including an increase in the consumption of fruits and vegetables, have consistently demonstrated prevention of these illnesses (39). Fruits and vegetables provide vitamins, minerals, antioxidants and other active compounds that play a role in disease prevention. Moreover, the consumption of fresh fruits and vegetables is an important part of programs focused on prevention of obesity around the world (www.who.int). The Healthy People 2010 initiative proposes to increase

Americans fruit and vegetable consumption to prevent chronic diseases; proposing at least 2 daily servings of fruit and at least 3 daily servings of vegetables including at least one serving of dark green or orange vegetable (146).

In response to this initiative the U.S Department of Agriculture (USDA) with the U.S Department of Health and Human Services (DHHS) in 2005 issued the dietary guidelines for Americans to lower the risk of chronic diseases and to promote health (266). These recommendations increase consumption of fruits and vegetables between 100 and 200% (263).

TRENDS IN PRODUCE CONSUMPTION

In 2000 Americans consumed 20% more fruits and vegetable than they did during the 1970s. The amount of fruits and vegetables that were consumed fresh, as opposed to frozen or canned increased 28% and 23%, respectively, compared to the 1970s. (264). From 1981 to 2000 the U.S per capita consumption

4 of fresh fruits and vegetables increased by 36% (U.S Census bureau and USDA Economic Research

Service).

The increase in fresh fruit and vegetable consumption has been result of several factors that have changed consumption patterns. The first one has certainly being advocacy to consumers to make healthier food choices, including fresh produce on their diets. Introduction of pre-cut and ready to eat produce has increased convenience, encouraging the consumption of these products (264). Other factors include the increased availability of produce items that were considered seasonal that are now available on a year round basis and higher diversity in produce items. In 1987 about 173 produce items were available in a typical grocery store while in 2007 it was possible to find about 558 different produce items (78).

PRODUCE AND FOOD SAFETY

Although an increase in fruit and vegetable consumption is desirable and it is occurring around the world, increases in foodborne illnesses have resulted from produce consumption. From 1973 through

1997 the number of outbreaks increased from two outbreaks per year in the 1970s to 16 per year in the

1990s (237). Production of fresh fruits and vegetables occurs in a natural environment, then they are exposed to contamination with human pathogens (168). These pathogens cannot be inactivated or eliminated on fresh produce without altering the sensory and flavor components of the produce (52).

Outbreaks associated with the consumption of produce items. In the past 10 years foodborne outbreaks associated with the consumption of fresh produce have occurred. Salmonella sp. and

Escherichia coli O157:H7 have caused most of the produce consumption associated outbreaks since 2000 with lettuce, spinach, tomato and alfalfa sprouts the most common produce involved (Table 1)

Table 1. Produce associated outbreaks that occurred from 2000 to 2008A.

Microorganism involved State Date Cases Vehicle

Year 2000

S. enterica Thompson multistate Nov 43 Tomato

Year 2001

S. enterica Newport CA May 8 Cilantro

E. coli O157:H7 TX Nov 20 Lettuce

Year 2002

E. coli O157:H7 CA July 5 Alfalfa sprouts

S. enterica Newport CO July 13 Cilantro

S. enterica Newport multistate July 510 Tomato

E. coli O157:H7 IL Nov 13 Lettuce

Year 2003

S. enterica Saintpaul multistate Feb 16 Alfalfa sprouts

E. coli O157:H7 CA Sep 51 Lettuce

E. coli O157:H7 CA Oct 16 Spinach

S. enterica Chester multistate Nov 26 Alfalfa sprouts

S. enterica Enteriditis CA Nov 14 Lettuce

S. enterica Saintpaul multistate Nov 33 Iceberg lettuce/tomato

Year 2004

E. coli O157:H7 GA Apr 2 Alfalfa sprouts

S. enterica Bovismorbificans multistate Apr 35 Alfalfa sprouts

S. enterica Braenderup multistate June 137 Roma tomates

S. enterica Newport multistate July 97 Iceberg lettuce

S. enterica Javiana FL Sep 24 Green salad

E. coli O157:H7 NJ Nov 6 Lettuce

Year 2005

E. coli O157:H7 multistate Sep 12 Pre-packaged lettuce

E. coli O157:H7 multistate Oct 34 Pre-packaged lettuce

Year 2006

Shigella sonnei OR Jan 35 Lettuce based salad

S. enterica Berta multistate Jan 16 Tomato

S. typhimurium MD June 18 Lettuce/tomato

S. enterica Newport multistate June 115 Tomato

E. coli O121 UT July 3 Lettuce based salad

E. coli O157:H7 multistate Aug 205 Spinach

E. coli O157:H7 NM Aug 5 Spinach

E. coli O157:H7 multistate Nov 77 Lettuce

E. coli O157:H7 multistate Nov 65 Lettuce

E. coli O157:H7 NY Nov 20 Lettuce

E. coli O157:H7 OR Nov 3 Vegetable based salad

Year 2007

E. coli O157:H7 FL Jan 2 Cesar salad

S. typhimurium ME Feb 76 Lettuce/spinach

E. coli O157:H7 AL June 26 Lettuce based salad

S. enterica Newport DC June 46 Tomato

S. enterica Newport NY July 10 Tomato

Shigella sonnei CA July 72 Lettuce based salad

S. typhimurium MN Oct 23 Tomato

Year 2008

S. enterica Saintpaul multistate Nov 1442 Jalapeño peppersB

Year 2009

S. enterica Saintpaul multistate May 228 Alfalfa sproutsC

A. Data collected from: Bacterial foodborne and diarrheal disease. National case surveillance annual reports, years 2000-2007 B. MMWR. (2008) Outbreak of Salmonella serotype Saintpaul infections associated with multiple raw produce items. 57(34):929-934. C. www.cdc.gov

From 2005 to 2006, four multistate outbreaks of Salmonella associated with the consumption of tomatoes occurred in the U.S., which resulted in 459 cases of salmonellosis in 21 states. Tomatoes

7 involved in the outbreak were either whole or pre-cut from tomatoes supplied to restaurants from fields of

Florida, Ohio and Virginia (1).

One of the outbreaks occurred in July to November of 2005 and implicated tomatoes grown on the eastern shore of Virginia. S. enterica serovar Newport was isolated from pond water used for irrigating nearby tomato fields (1), the same strain had been previously involved in a Salmonella outbreak in 2002 (272). This outbreak caused 72 cases of salmonellosis in 16 states in the U.S. Later in July of

2006, 115 cases of Salmonella enterica serovar Newport occurred in 19 states due to the consumption of raw tomatoes however the source of tomatoes was never determined.

During November of 2005, 82 cases of S. enterica serovar Braenderup occurred in eight states.

The cases were linked to consumption of pre-packaged tomatoes grown in one field in Florida. The investigation indicated the likelihood of animal reservoirs, as known animal vectors of Salmonella sp. were present in drainage ditches close to fields. This investigation also showed that other serotypes besides Braenderup were also present in the area of investigation (1).

In 2006, three multistate outbreaks of E. coli O157:H7 infections associated with produce occurred, one associated with bagged fresh spinach accounted for 36% of the cases of foodborne illness that year, and two associated with lettuce in two fast-food chains that accounted for 16% of the cases of foodborne illness (4). During September of 2006, an outbreak of E. coli O157:H7 linked to the consumption of fresh, bagged spinach affected 26 states and Canada, with 205 cases of illness and 3 deaths (59). The strain of E. coli O157:H7 was traced to bagged spinach from California, particularly to one production date at one processing plant and fields located on 4 different ranches in central California

(129). Environmental investigations detected E. coli O157:H7 in cattle feces, surface water, soil, well and irrigation water and on feral swine (129).

In December 2006, an E. coli O157:H7 outbreak linked to the consumption of Mexican-style fast food caused 71 illnesses in 5 states in the U.S. The E. coli O157:H7 strain was traced to lettuce from a single supplier in New Jersey (58). A few days later a different outbreak, involving a Mexican-style fast

8 food restaurant, was linked to the consumption of lettuce from a supplier in Minneapolis causing 77 illnesses in two states (58).

One of the largest outbreaks occurred in 2008, 1442 cases of S. enterica serovar Saintpaul, were linked to the consumption of produce items associated with the preparation of guacamole; avocado, raw

Roma tomatoes, raw red onions, raw Serrano peppers and raw cilantro. The outbreak expanded to 43 states and the District of Columbia, and Canada. Trace back and environmental investigations found that jalapeno peppers from Mexico were the main vehicle of S. enterica serovar Saintpaul. The peppers were grown on a farm in Tamaulipas, Mexico, which also grew Roma tomatoes (2).

E. coli O157:H7 outbreaks. E. coli O157:H7 was first recognized as a foodborne pathogen in

1982 during an outbreak associated with the consumption of hamburger from a fast food chain (215).

Since then to 2002, 350 outbreaks of E. coli O157:H7 were reported, 52% of them were foodborne, 21% of unknown route, 14% person to person, 9% waterborne (recreational and drinking water) and 3% due to animal contact. From the foodborne outbreaks 41% was associated with ground beef, 21% produce, 6% other beef, 5% other food and 4% dairy products (210).

Consequences of foodborne illnesses. Foodborne outbreaks affect several sectors of society; food industry, government, health institutions, national labs, and especially human population can be severely affected when an outbreak occurs. The main impact regards public health issues. In the United

States foodborne illnesses affect 6 to 80 million people each year, causing ~ 9,000 deaths and costing billions of dollars (13). In 2008, a total of 18,499 laboratory-confirmed cases of infection from a foodborne pathogens were identified (5). The cost of disease for foodborne related illnesses in 2008 was over 2 billion dollars for salmonellosis and ~500 million for E. coli O157:H7 illnesses alone (75).

Along with the high economic costs associated with foodborne illness many of the foodborne pathogens may result in chronic sequelae or disability (13). Miscarriages and meningitis are caused by

Listeria monocytogenes infections. Toxoplasmosis is the cause of congenital malformation. Infections

9 with Campylobacter jejuni produce mild gastrointestinal symptoms, however some infections may result in the more serious flaccid paralyisis from Guillain-Barre syndrome. Salmonella sp. infections can lead to invasive disease and reactive arthritis (13). One of the most dangerous foodborne infections is caused by

E. coli O157:H7, infection can progress and cause hemolytic uremic syndrome, which is the most common cause of acute kidney failure in children, with devastating results including death (40).

Foodborne outbreaks severely affect the food industry. During the outbreaks occurring in fresh bagged spinach in 2006, a voluntarily recall was initiated by Dole Fresh Vegetable Inc. Although the contaminated spinach had been concentrated in 42,000 bags processed during a single shift, the recall was larger because at the time of the outbreak it was unknown how many products from different processors were involved, within seven days, 5 more companies recalled products (277).

Initial epidemiological investigations into the 2008 S. enterica serovar Saintpaul outbreak suggested that tomatoes were the cause of the outbreak. However, after recalls failed to reduce the number of cases additional investigation revealed that the main vector was jalapeno peppers (2). This recall resulted in a loss of more than 200 million dollars to the US. tomato industry. In addition to the high costs of associated with the recall and disposal of contaminated products, additional economic costs were incurred due to loss in sales due to decreased consumer confidence and in some cases legal action.

The reality of food safety. Food safety is a priority of Healthy People 2010 initiative along with a desired increase in the consumption of fresh fruits and vegetables, as part of health promotion and disease prevention (www.healthypeople.gov). As of 2008, none of the Healthy People 2010 targets for reduction of foodborne pathogens had been addressed (5).

To address these goals, strategies to prevent or reduce contamination must be implemented in all parts of the food production chain (from farm to table). This will ensure that consumers have a safe food supply, which in turn will result in a healthier population and improved public health. Important fields of study involve the understanding of pre- and post- harvest contamination of produce with foodborne

10 pathogens, mechanisms for survival and establishment of these bacteria , developing intervention technologies that reduce or impair the growth and establishment of pathogenic bacteria on produce, and advancement of public policies that encourage or mandate food safety interventions at all stages of produce production.

ESCHERICHIA COLI O157:H7 AS HUMAN PATHOGEN

A number of serotypes comprise the Enterohemorrhagic E. coli (EHEC) complex including

O16:H11, O103:H2, O111:NM, O113:H23 and O157:H7. The predominant EHEC type associated with food-borne outbreaks worldwide is O157:H7; although the other serotypes are also associated with cases of human illness (186). It is a zoonotic pathogen with a very low infectious dose causes illnesses ranging from mild diarrhea to bloody diarrhea and may progress to hemorrhagic colitis and hemolytic uremic syndrome (HUS) in children and thrombotic thrombocytopenic purpura in adults , leading to death (27)

(186).

Mechanisms of pathogenicity of EHEC. EHEC are very sophisticated pathogens that display several virulence-associated characteristics. As a gastrointestinal pathogen, EHEC need to pass the gastric stomach barrier which has a pH value as low as 1.5-2.5. Mechanisms of adaptation to this environment are present in several foodborne human pathogens including Salmonella enterica, E. coli, Vibrio cholera and Helicobacter pylori (84). E. coli O157:H7 has three acid resistance mechanisms; AR-1, AR-2 and

AR-3 which allow these bacteria to successfully survive under low pH conditions. These mechanisms are pairs of amino acid decarboxylases and antiporters. AR-1 is dependent on the stationary phase alternative factor sigma S (RpoS) and cAMP receptor protein (CRP) (57). AR-2 and AR-3 are glutamate and arginine dependent mechanisms respectively (84).

After EHEC survives the gastric challenge, they can adhere to epithelial cells subverting cytoskeletal processes. This is an attaching/effacing lesion (A/E lesion), a local effacement of the brush border microvillli and intimate attachment of the bacterium to the membrane of the host cell (86, 140).

The ability to form A/E lesions is encoded in the locus of enterocyte effacement (LEE) pathogenicity island. The regulation of LEE gene expression is dependent on environmental conditions and quorum sensing (15). LEE encodes structural components of a type III secretion system (TTSS), and effector proteins (93). TTSS is formed by Esp proteins, EspA is an important adhesion factor that establishes a link between the bacterium and the host epithelial cell and enables protein translocation. Through the

TTSS, intimin is translocated, then EspA filaments are excised allowing attachment through intimin and

Tir which is a receptor encoded in the host cell (87, 88). The effector proteins translocated through the

TTSS in the host cells are responsible for the disruption of intestinal barrier, alteration of the tight junctions, release of toxic proteins into the cytosol that cleave caspases that cause mitochondrial death and cytoskeleton rearrangement (93). This last effect allows pedestal formations that occur at the site of bacterial adhesion and provide a physical barrier for the host immune response. The loss of microvilli due to damage in the brush border cells induce persistent diarrhea (93).

The development of HUS is due to the production by E. coli O157:H7 of Shiga-like toxin (SLT).

SLT is a 70KDa holotoxin consists of five B subunits responsible for binding of the toxin to the cell surface glycosphingolipid receptor globotriaosylceramide (Gb3), and one A subunit, which inactivates ribosome to inhibit protein synthesis (193). Isolated E. coli O157:H7 from patients that developed HUS usually produce SLT-2 but frequently produces both SLT-1 and SLT-2 (163). SLT-2 has been demonstrated to have a preferential interaction with human renal microvascular endothelial cells and it is more potent as citotoxic agent than SLT-1(163).

Shiga toxins 1 and 2, are two sequentially and antigenetically distinct toxin groups transcribed from stxAB and stxAB2 operons, respectively (201). There is an increase risk for HUS development when both stx2 and eae are present in the infecting strain (76). The Sakai strain, that caused an foodborne outbreak in Japan during 1996 caused by radish sprout consumption, caused 8,000 infections, the majority of them children and the rate of HUS was 1.2% (172). In 1982, the EDL933 strain involved 44 people but no cases of HUS were present (215). During the spinach outbreak in 2006, 29% of the patients developed

HUS and the isolated strain carried SLT-2 (99), this information indicates that the clinical outcome is strain-specific.

EMERGENCE OF RECENTLY RECOGNIZED FOODBORNE PATHOGENS

Foodborne diseases are a widespread public health issue worldwide. In the last 25 years foodborne agents have appeared in the population by either extending to new vehicles of transmission, increase their incidence or have been recognized as new foodborne pathogens (271). In the United States,

Campylobacter jejuni, C. fetus sbp. Fetus, Cryptosporidium cayetanensis, E. coli O157:H7, Listeria monocytogenes, Norwalk-like viruses, Nitzschia pungens, Salmonella enterica, Vibrio cholerae O1, V. vulnificuls, V. parahaemolyticus, Yersinia enterocolitica have been associated with increasing numbers of foodborne illness in the last 30 years (252). A shared characteristic of these pathogens is that all have an animal reservoir and then are considered zoonotic pathogens that can be rapidly spread globally (252).

Several factors are involved in the emergence of these microorganisms as sources of foodborne disease (271) including: ability of microbial adaptation, economic and social development, changes in dietary habits, demographic changes, globalization of the food supply, changes in animal production, and new vehicles of transmission

Changes in transmission vehicles associated with microbial adaptation. Microorganisms have the ability to adapt to new environments through acquisition of new genes, which is a key process in the emergence of pathogens. For example E. coli O157:H7 has evolved from the EPEC-like ancestor by acquiring LEE and then Shiga toxin genes, making this pathogen highly virulent (199). S. enterica has also acquired phages, particularly PT 4, that enable it to resist acidity, heat and peroxide (114). S. enterica

PT 4 has became predominant in humans and poultry in Europe with increases in human illnesses and in

United States has increased 5 fold in Salmonella infections (200).

Food products such as meat and eggs were commonly associated with E. coli O157:H7 and

Salmonella, respectively. Fresh fruits and vegetables have become a major transmission source of these microorganisms, suggesting some type of adaptation of these pathogens to new environments (13).

Societal factors associated with emergence of foodborne pathogens. Social and economic development have lead to the introduction of new foods, now the food chain has become longer and complex, increasing opportunities for contamination. At the same time, dietary habits have changed, nowadays consumers prefer pre-packaged ready to eat produce, reducing food processing and offering healthier fresh products but reducing the opportunity to reduce or eliminate the risk of foodborne disease

(271). Also the percentage of people eating away from home has increased (165). Fast-food restaurants and salad bars have become more common in society. Outbreaks outside the home are 80% of reported outbreaks in the United States (13).

Due to demographic changes in developed countries, the proportion of the population with higher susceptibility to severe foodborne infections has increased (13). The increase in population with HIV and larger life expectancies due to advances in medical technology has increased the susceptible population to foodborne associated illnesses. HIV patients are more likely to gastrointestinal disease caused by S. enterica and L. monocytogenes than the general population (12).

Globalization of the food supply. The food industry has become globalized and food products are distributed across a greater geographic range than in previous decades. Food products are sent to central processing facilities and then distributed from large centralized food processing facilities to large areas. This increases the likelihood of cross contaminating large amounts of products that are widely dispersed, increasing the risk for large, dispersed outbreaks (13, 271), as evidenced by the increase in multistate outbreaks from a centralized source of food that have occurred in the last 10 years (Table 1).

Food trade has also impacted the emergence of foodborne pathogens and globalization has facilitated food trade leading to an increased number of cases originating from products grown in areas where different

14 pathogens are endemic (271). The classic example is the 1996 outbreak of Cyclospora cayetanensis, resulting in 1465 illnesses in 20 states in US and Canada, which has been eliminated from developed nations due to proper sewage processing and treatment. The vehicle for this outbreak was determined to be raspberries from Guatemala washed with contaminated water (106).

Factors that contribute to the increased emergence of E. coli O157:H7 as foodborne pathogen. E. coli O157:H7 is a commensal member of the gastrointestinal microbiota of animals. It is believed that cattle are the main reservoir, though E. coli O157:H7 has been isolated from sheep, pigs, horses, dogs, chickens and wildlife such as wild pigs and deer (32, 60, 107, 108, 142, 229). Healthy cattle and sheep harbor E. coli O157:H7 in their gastrointestinal tract and shed the bacteria in their feces (60,

102). The prevalence of this bacterium in a herd ranges between 0.3 to 6.1% for cattle and between 0.9 to

31% for sheep, with an average length of time that feces from animal remain culture positive of about 30 days for cattle and 50 days for sheep (226, 265).

The emergence and increased prevalence of E. coli O157:H7 since 1982 is likely associated with changes in animal production (89), particularly in animal nutrition and animal management. In order to increase retention of nitrogen and carbon in cattle digestive systems, improve immune response and milk production and immune response, the use of ionophores have been approved for animal feeding (51).

Ionophores are classified as antibiotics, affecting Gram-positive bacteria, presumably reducing competition and promoting the growth of Gram-negative bacteria, including E. coli species, which have better physiological strategies to deal with ionophores (29, 120).

The introduction of high carbohydrate diet for animal feeding may increase the prevalence of E. coli O157:H7 (270). E. coli species do not have the ability to degrade polysaccharides resulting from the breakdown of grasses; however E. coli has the ability to utilize the soluble carbohydrates produced from grain starch degradation by rumen (223). Cattle fed diets high in grain production more fatty acids and

15 have a reduced colonic pH compared to grass fed cattle. Cattle fed corn, had 1000-fold more E. coli, which also were less sensitive to acid shock (223).

Intensive livestock farming at a relative high density can also increase the prevalence of E. coli

O157:H7 among cattle and calves, especially when housed in the same lot. Increased stock density of cattle exposes cattle to contamination with E. coli O157:H7 through drinking water and livestock feeds which play a key role in the exposure of cattle to E. coli O157:H7 (Fig. 1) (69). In one study, Holstein

Friesian steers shedding E. coli O157:H7 were introduced among healthy steers in a pen. Within 24 h, drinking water, pen barriers and animal hides were positive for the pathogen and after 72 h, 66% of the healthy animals were contaminated suggesting transmission of E. coli O157:H7 between animals and through ingestion of the pathogen from contaminated water (170).

CONTAMINATION OF PRODUCE WITH E. COLI O157:H7 DURING PRE-HARVEST

OPERATIONS

Produce contamination with E. coli O157:H7 can occur during pre and post-harvest operations during crop production (Fig 1). The main sources of leafy green contamination are due to contaminated irrigation water and animal manure (89) although other causes such as wild life (129) can also be cause of produce contamination. Non-point source produce contamination from manure can come from pastured animals, roaming wild animals or manure intentionally spread on fields as fertilizer. Point sources of manure contamination can include: animal feedlots and housing facilities and manure storage areas like lagoons. Some point sources can lead to non-point sources of contamination by spreading manure to fields and water supplies such as runoff or leaching (91).

Figure 1. Routes of produce contamination with E. coli O157:H7

Animal manure. Animal manure is used as an organic fertilizer to provide nutrient to the crop and for soil amendment (89). Contamination can occur from application of poorly composted or raw manure or improper application time prior to harvest is major risk for produce contamination with human pathogens (34). E. coli O157:H7 survival in non treated raw manure ranged from less than 7 days at 33°C to 159 days at 7°C (232), supporting the high risk of raw manure application on agricultural fields.

Composting allows conversion of agricultural waste products into organic fertilizer, soil amendments or potting media and can also provide the reduction or elimination of bacterial pathogens parasites, fly larvae and weeds(135). During composting there is an increase from 25 to 58°C due to bacteria and yeast metabolism, and then a gradual temperature decreased. Four phases can be identified during the process: mesophilic, thermophilic, cooling and maturing (121). As demonstrated by denaturant gradient gel electrophoresis microbial succession occurs during composting due to changes in environmental conditions. Early in the process a decrease in mesophilic bacteria occurs to low or undetectable levels of enterobacteria with and increased population of lactic acid bacteria, during the thermophilic phase, Gram positive bacteria such as Bacillus sp. become more dominant and after the cooling phase bacterial population complexity increases more than in previous phases (121). During the thermophilic phase inactivation of human pathogens might occurs, but if the composting process fails to achieve the proper temperature or duration for inactivation then the compost might become a risk for contamination.

E. coli O157:H7 survived for 36 days during composting in bioreactors at an external temperature of 21°C; however the bacterium can be inactivated to undetectable levels after 14 days when the external temperature of the bioreactor is 50°C. The size of the E. coli O157:H7 populations influence their survival and time required for inactivation during composting; large microbial populations (> 7 log

CFU/g) require compositing for 1 week minimum, ensuring 50°C internal temperatures in order to reduce

18 the microbial population to undetectable limits (130). The survival of E. coli O157:H7 in compost has been reported to be influenced by the background microflora and moisture content (137). In autoclaved compost the E. coli O157:H7 population increased 3 fold; though samples containing background microflora, showed a 6.5 log/g reduction in O157:H7 populations, indicating the influence of the background microflora in the re-growth of E. coli O157:H7 if contamination occurs (137).

Irrigation water. Agriculture is a major user of ground and surface water in the United States, accounting for 80 percent of the Nation's consumptive water use and over 90 percent in many Western

States. Livestock and poultry manure, can degrade water quality if manure is over applied to land and enter water resources through runoff or leaching. A shift in the livestock and poultry industry over the last few years towards fewer and larger operations has prompted public concern about the microbiological safety of irrigation water (213). For example in California, which is the main source of leafy greens in U.

S, an intensive integrated agricultural network exists with dairies located in close proximity to extensive fruit and vegetable production fields (212). These dairy operations commonly housing 900 or more cows are common and are designed as free-stall operations (212). One of the main disadvantages of this type of production is water contamination, which is becoming more common in rural areas of US as result of large animal operations (116).

If water sources used for irrigation are contaminated an increased risk of contaminating the irrigated crop exists. The irrigation method affects the risk of produce contamination depending on its contact with the edible parts. Irrigation methods such as sprinklers or sprays, where the water is applied to the edible parts offer the most risk for widely dispersed crop contamination (136). Intermittent spray application of inoculated irrigation water containing 2 log10 CFU of E. coli O157:H7 for 14 days resulted in recovery of the bacterium for at least 30 days. A one log increase after 30 days occurred when 4 log10

CFU of O157:H7 cells were applied to lettuce plants either by a single or intermittent application of contaminated irrigation water. This suggests that growth of E. coli O157:H7 occurs on the lettuce leave and demonstrates the potential for produce contamination by irrigation water (241). Other study showed

19 that 90% of lettuce plants that were spray irrigated with 7 log CFU/mL of E. coli O157 remained contaminated for 20 days on the leaf surface (242).

Localized irrigation techniques like drip-and-trickle irrigation, have lower risk of crop contamination because water is directly applied to the soil surface and not come into contact with edible plant tissues. This method of irrigation is more expensive and generally not used for leafy greens (25).

Leaching of pathogens in soils with the possibility of contamination of well water is one mechanism of contamination of irrigation water (91).

Contaminated soils. Soil is generally contaminated directly by the application of animal waste or indirectly through runoff which can then percolate into the soil strata. Rainfall can spread pathogenic microorganisms into soil by runoff from stored or unincorporated manure or by leaching through the soil

(91). E. coli O157:H7 applied in contaminated manure have been isolated from soils deep beneath the surface. Furthermore, this pathogen and other coliforms are capable of a high growth rate in soil (91) and leaching can be influenced by soil type and tillage practice (tilled or no-till) as well as rainfall. Ammonia and nitrate concentrations in soil positively correlated with leached levels of E. coli O157:H7. Tillage practices did not prevent vertical transport of bacteria in soil (91). E. coli O157:H7 persisted in soils for more than 5 months after application of contaminated compost or irrigation water (122). Populations of E. coli O157:H7 on lettuce leaves planted on contaminated soil, were present after 36 days after planting with a 4 fold increase observed from day 7 to 36 (167).

The movement and dispersion of pathogenic bacterium can be related with vertical leaching due to rainfall and soil composition (91). Other studies have suggested that earthworms such as Lumbricus terrestris may also aid vertical movement of E. coli O157:H7 in soils by carrying the bacterium on their epidermis and to a lesser extent from earthworm intestine excretions (16, 281). Once soil is contaminated it may contaminate produce.

Experiments with soil microcosms inoculated with E. coli O157:H7 have showed that the persistence of this bacterium in the phyllosphere, rhizosphere and non-rhizosphere soils occurs for long periods. E. coli O157:H7 persisted in rye roots for 47-96 days and on alfalfa roots 92 days (92). In the case of lettuce leaves, the bacterium persisted over 45 days on the leaf surface, non-rhizosphere soil and in the rhizosphere (116). The long persistence of this bacterium in the agricultural environment might play an important role in the recontamination cycle of produce.

Wild life and other sources of produce contamination. Although cattle are believed to be the primary reservoir of E. coli O157:H7, other species can also asymptomatically carry this bacterium and can shed the bacterium in fecal material potentially contaminating irrigation water, soil and crops. E. coli

O157:H7 has been isolated from several species such as: deer, dogs, ducks, kangaroos, and wild pigs (7).

Field contamination with E. coli O157:H7 may have occurred due to presence of feral swine feces. These swine contained the same strain as cattle from a nearby dairy, which they were likely to have passed through to access the spinach fields (129). Wild deer are presumed to be associated with contamination of apples used to make unpasteurized apple cider. The apples were obtained from the ground where wild deer were feeding (3, 98, 247).

In addition to mammals, which carry E. coli O157:H7 in their intestines, other wildlife may transmit the pathogen through cross contamination events. Insects like lesser mealworm (Alphitobius diaperinus), fruit flies (Drosophila melanogaster) and house fly (Musca domestica) have also shown to be able to transmit E. coli O1578:H7 to fruits such as apples (72).

ECOLOGY OF FOODBORNE PATHOGENS ON PLANT SURFACES

E. coli O157:H7 is hosted by animals and it is well known to be an enteric pathogen. The intestinal environment where it is usually hosted is replete in nutrients, it is anaerobic and usually conditions such as humidity and temperature are homogeneous (42). When the host becomes a plant surface, the conditions experienced by E. coli O157:H7 as well as other enteric pathogens are quite

21 different. Harsh physicochemical conditions can be experienced; particularly environmental conditions

(humidity and temperature) fluctuate rapidly and widely during the day (109). Unlike the cattle or human intestine, there is low availability of nutrients (157) and exposure to UV radiation (95) which makes the plant a harsh environment to host enteric pathogens (42). Enteric pathogens seem to overcome that harsh environment and be able to establish on plant surfaces, which signifies a major risk for humans.

Persistence of foodborne pathogens on edible plants surfaces. The persistence

Enterobacteriaceae human pathogens, on the surface and sometimes inside edible plants have been reported in several studies (19), indicating that pathogenic bacteria such as E. coli and Salmonella sp. can survive outside the animal host in the agricultural environment, and potentially survive stresses encountered in vegetable (70).

E. coli O157:H7 has been shown to persist on greenhouse and field produced leafy greens. E. coli O157:H7 inoculated on the adaxial and abaxial side of iceberg lettuce was isolated after 25 days from the abaxial leaf surface which was held at 25°C during the day and at 7°C during the night (294). In addition, E. coli O157:H7 associated with contaminated manure has been shown to persist on edible portions of iceberg lettuce and parsley for up to 177 days in the field (122, 242). E. coli O157:H7 is also able to multiply on romaine leaves, increasing the population in a range of 16- to 100-fold from the initial inoculum (4 log10 CFU/g) on plants grown in growth chambers at high humidity and at 28 °C (44). The increase in population size is dependent on the amount of nutrients available on the leaf surface, which is a function of leaf age. Population sizes were 10-fold higher on the inner leaves (young plants) of romaine lettuce compared to the middle leaves (old plants) (44). These studies demonstrated E. coli O157:H7 growth on the surface of leafy greens, and the establishment of a niche environment suitable for survival on the leaf surface.

The stage of plant development may also influence the potential for growth and persistence of foodborne pathogens on produce. Increases of 3.5 – 4.5 log CFU/g E. coli O157:H7 and Salmonella

22 enterica serotype Typhimurium have been reported on the edible sprouts of carrot, cress, lettuce, radish, spinach and tomato whose seeds were inoculated with 2 log CFU/ml suspensions (124). E. coli O157:H7 was able to internalize in the seedlings but was not recovered within the tissues of mature plants and was isolated from the roots of growing plants (124).

ESTABLISHMENT OF FOODBORNE PATHOGENS ON EDIBLE PLANTS

Various factors are involved in the establishment of bacteria on plant surfaces. Some of those factors might be inherent to the plant itself which can either impair or promote bacterial growth but external factors, like the environment, also have a great influence in the establishment and survival of microorganisms including foodborne pathogens. Finally, bacterial adaptation to the plant environment plays an important role in their survival. Insights to understand how enteric pathogens survive, establish and colonize edible plants can be found by studying the plant as an environment to support bacterial growth and by analyzing mechanisms utilized by indigenous bacteria that are already adapted to the plant surface.

Plant surface as environment for bacterial growth. The leaf surface is known as phylloplane, which is composed of plant epidermal cell walls and cuticle, which provide structural rigidity (234). The phylloplane is covered by a cuticle which is an abiotic covering over the epidermal cells which is normally the first contact for microorganisms (234). The cuticle is formed by polysaccharides, cutin and waxes, which normally functions as a barrier for water loss but also works as a substrate for microbial adhesion and as barrier for epidermal penetration of plant pathogens (182, 289). The integrity of the cuticle is also pivotal to prevent enteric pathogens to edible plants.

Interspersed among epidermal cells are guard cells which surround stomata and regulate gas exchange with the atmosphere. Other specialized cells are lacticifers, lenticels, oil glands and salt glands surface resin ducts that can release various substances from the plant interior, many of these excretion cells can contain antimicrobial compounds (202). Other structures in the phylloplane include hydathodes,

23 and trichomes. Hydathodes are specialized structures connected to the plant xylem. They exude guttation fluid from the xylem to the leaf surface under humid conditions and can secrete certain antimicrobial compounds (123, 234). Trichomes are stalked multicellular structures, which have a hairy appearance at low magnification and they can either be simple or glandular secreting trichomes (77). Simple trichomes are found on plant surfaces and can entrap fungal spores, bacteria, insects and affect leaf temperature

(234, 274). Glandular secreting trichomes produce leaf exudates that consist of terpenoids, phenylpropanoids, minerals, salts, sugars, water and volatiles (77, 203, 279).

The phylloplane is naturally colonized by a variety of bacteria, yeasts and fungi, forming what is known as the phyllosphere (155), particularly microorganisms able to colonize surface plants are known as phylloepiphytic microorganisms (26). Fungi are considered transient inhabitants of the leaf surfaces, in contrast, yeast and bacteria are able to colonize the phylloplane more actively, however bacteria is the most abundant member of the phyllosphere (18). The sources of microorganisms on the phyllosphere are the atmosphere, insect, seed or animal-borne sources (280). Tree buds, seeds of annual plants and debris from previous crops, are the most important sources of bacteria for new plants and they are also already adapted to the phyllosphere (164).

The colonization of the phylloplane by bacteria is not uniform across the leaf surfaces but is likely localized in specific sites (26). Scanning electron microscopy (SEM) has shown that the most common sites of bacterial colonization include: the trichomes, stomata, along the veins, within depressions in the cuticle, beneath the cuticle, and near hydathodes (37, 141, 178, 287). Larger bacterial populations are isolated on the abaxial side of the leaves (lower side) than on the adaxial side (upper side), presumably the result of higher stomata, trichome densites and a thinner cuticular layer in the abaxial side (26, 64, 141, 250). GFP tagged Salmonella enterica var Thompson has been visualized on the intercellular margins on the leaf, areas near stomatal openings and on the abaxial side of spinach and cress

(276). Similar patterns of colonization occurred on the surfaces of Arabidopsis thaliana, where L.

24 monocytogenes expressing GFP was mostly localized in the intercellular spaces and near the stomata

(173). E. coli O157:H7 has been visualized near opened stomata on spinach leaves (175).

Factors that contribute to plant colonization and fitness of the phylloepiphytic bacteria and enteric pathogens. The major limiting factor for bacterial colonization is the availability of carbon followed by nitrogen (283). Exogenous nutrient sources include pollen, dust, microbial debris and plant sap released from wounds made by insects feeding or frost damage (151). However healthy plants can passively leak small amounts of metabolites like carbohydrates, amino acids and organic acids to the leaf surface (160, 171, 258).

A typical uncolonized bean leaf secretes 0.2 to 10 g of surface sugar which is enough to support

7 to 8 log cells/leaf (171). Using bacterial bio-sensors, (Erwinia herbicola harboring sucrose- and fructose-responsive scrY promoter fused to gfp exhibiting sugar-dependent GFP fluorescence) it was shown that there is a high level of heterogeneity in sucrose and fructose availability on leaf surfaces, the amounts allow short and transient adaptation to the environment for further multiplication and colonization. Most areas of the leaf harbor small amounts of nutrients (0.15 pg of fructose) while other can harbor nutrients in relative abundance (4.6 pg of fructose) (151), which might explain the observed heterogeneous patterns of leaf bacterial colonization.

A recent study showed that leaf age was important for the establishment of pathogenic E. coli and

S. enterica in lettuce, finding that exudates of young leaves were about three times richer in total nitrogen and carbon than middle age leaf exudates, concluding that growth of enteric pathogens is higher on young lettuce leaves due to larger nutrient availability (44). Several phyllosphere bacteria produce the plant growth regulator indole-3-aceitic acid (IAA), which promotes cell wall loosening at low concentration, and stimulates the release of saccharides from the plant cell wall (47, 94, 96, 156). The ability to produce increases the epiphytic fitness of common phyllosphere bacteria, Pantoea agglomerans, two fold

25 compared to an IAA mutant (45). IAA production was also associated with increased fitness during period of drought stress on plant surfaces (166).

Nutrient limitation and water availability are factors that challenge bacterial growth on leaf surfaces (155). On plants, water stimulates the leaching of nutrients, then structures that are more likely to retain water, such as veins, trichomes and hydathodes are expected to harbor more nutrients and support larger bacterial populations (160). Measurement of cuticle transpiration on Hedera helix and

Prunus laurocerasus, using radiolabeled water, showed that known epiphytes like Pseudomonas graminis, Xanthomonas campestris and Corynebacterium fascians¸ had the ability to alter leaf surface permeability to increase wettability and gain epiphytic fitness by producing biosurfactants reducing surface tension (230). The hydrophobic nature of the cuticle, and bio-surfactant production, increases wettability and allows the solubilization and diffusion of nutrients for utilization by epiphytic bacteria

(50, 115, 187).

Survival strategies of phylloepiphytic bacteria on leaf surfaces under environmental stress conditions. Desiccation stress, direct exposure to UV light and fluctuations in temperature are environmental stresses that constantly challenge the survival and fitness of bacteria that colonize the leaf surface (95). Phylloepiphytic bacteria are able to survive these stress conditions by two major strategies:

1. Tolerance strategies to survive under direct exposure to environmental stresses and 2. Avoidance strategy through sites that are protected from those stresses (95).

UV tolerance is a trait elicited by several epiphytic bacteria (154). The lethality of solar UV exposure is mostly attributed to intracellular generation of reactive oxygen species (ROS) like H2O2, superoxide anion, and singlet oxygen and not necessarily to direct DNA damage (206). Successful phylloepiphytic bacteria, have to be efficient in UV-induced DNA damage repair mechanisms or preferentially colonize sites protected from UV, such as the interior of a leaf (often phytopathogens) or at the base of structures such as trichomes (often called saprophytes) (125, 126).

Epiphytic bacteria are able to tolerate UV associated stresses by through expression of genes associated with increased pigmentation and DNA repair (95). The ability to produce a carotenoid pigment provides Clavibacter michiganensis increased ability to survive on peanut leaves compared to a pigment deficient mutant (125). The recA gene involved in DNA repair confers UV tolerance on a number of epiphytic bacteria (154). Elimination of recA in Pseudomonas syringae significantly increased the susceptibility to UV light in culture (282). The rulAB genes which also participate in DNA repair enhanced the survival of P. syringae on bean leaves grown in the field (249). Homologues of rulAB are present in enteric pathogens, particularly E. coli and S. enterica (190, 285), and may contribute to their survival on plant surfaces. However whether this mechanism is utilized by enteric pathogens for UV tolerance has not been determined.

Bacterial aggregates often composed of large mixed bacterial species occurred on the leaf surfaces (191) and they have important implications in the ability of phylloepiphytic bacteria to survive and colonize the leaf surfaces (42, 155). Aggregation provides a means to modify the plant habitat (155) through secretion of extracellular polysaccharide (EPS). EPS protects members of the bacterial aggregate from desiccation stress and reactive oxygen species compared to solitary cells (68, 155, 177). Cell aggregation can also result in bacteria-bacteria communication through cell density-dependent signaling

(quorum sensing) which may lead to expression of genes that enhance epiphytic fitness (155, 157). In vitro assays have shown that immigrant bacteria such as P. syringae are two times more likely to survive when they land on a aggregate of the plant epiphyte Pantoea agglomerans than when landed on uncolonized parts of the leaf (176).

Bacterial aggregation on leaf surfaces has also been observed for enteric pathogens; large aggregates of S. enterica have been reported on spinach surfaces (276), cilantro (46) and parsley (145).

Although the effect of quorum sensing on enteric pathogens associated with leaf surfaces has not been well studied. Pathogenic E. coli species and Salmonella enterica do not synthesize cell signaling molecules such as acyl homoserine lactones (AHL), however they do produce receiver molecules (luxR

27 homologues) that respond to AHL quorum sensing signaling (8). This could make possible the communication with members of the phyllosphere which might also have a role in the fitness and survival of enteric pathogens on the leaf surface.

Overlapping mechanisms for survival and establishment of enteric pathogens on plant surfaces. E. coli O157:H7 is well adapted to its animal hosts, and in most of the cases, without causing host sickness. As previously described, these bacteria are also able to colonize and multiply on plant surfaces, despite the harsh environmental conditions on the phylloplane. This strongly suggests that enteric pathogens are well adapted to the changing conditions that occur outside the animal host.

It has been suggested that life outside the animal host for enteric human pathogens could be part of their life cycle, in which they use plants as an alternative host to survive in the environment and then as a vehicle to re-colonize animal hosts once ingested (259). In this process bacteria shed in animal feces come in contact with grasses that are eaten by the animal host where colonization occurs (259). It is suggested that survival strategies utilized by enteric pathogens to establishment and colonization the human and animal intestines may overlap with those used to survive and colonize plants (42).

Mechanisms of survival and establishment of enteric pathogens on surface plants have not been extensively studied. Knowledge of these mechanisms can certainly provide better means to control these microorganisms pre- and post- harvest.

Low water availability is a main limitation of bacterial survival on plant surfaces (155). Food borne pathogens, particularly S. enterica, are highly tolerant of long-term desiccation in non-host environments. S. enterica ser. Thompson inoculated onto dry almonds surfaces can be cultured after 171 and 550 days when stored at 35°C and -20 to 4°C respectively (260, 261). S. enterica serovar Thompson survived dry conditions (60% relative humidity) on cilantro plants in comparable numbers to epiphytes;

P. agglomerans and P. chlororaphis. Furthermore, S. enterica was able to recover to a maximum population size on cilantro surfaces when returned to high humidity conditions (46). This tolerance to

28 low water availability provides one strategy for S. enterica to persist on the plant surfaces despite the changing environmental conditions (42).

Bacteria respond to starvation conditions by sustaining metabolic activity and maintaining cell viability. Low nutrient conditions, similar to nutrient levels on leaf surfaces, trigger the starvation response through expression of rpoS. (42). RpoS (sigma factor) is a master regulator of general stress responses, whose activation occurs under a variety of stress conditions and during stationary phase in many bacteria (105). In Pseudomonas sp. RpoS enhances colonization and fitness on roots and soils

(174, 248). RpoS is a regulator that controls epiphytic fitness, quorum sensing signals and plant interactions in the plant pathogen P. syringae pv tomato strain DC3000 (61). P. fluorescens Pf-5, a rhizosphere-inhabiting bacterium, required rpoS expression for osmotic, dehydration and oxidative stresses, and this gene plays a role in tolerance of Pf-5 to freezing and starvation (248). In E. coli and

Salmonella sp., RpoS provides cross protection against a number of stress conditions including oxidative and osmotic stress and starvation, allowing alternative utilization of carbon and energy sources due to nutrient limitation (192).

Quorum sensing could play a role in the survival of enteric pathogens on plant surfaces. A number of epiphytic and/or plant pathogenic bacteria produce quorum sensing signaling, molecules, particularly HSL, that could be received by closely related enteric pathogens and aid their survival on plants (42). For example in E. coli O157:H7, quorum sensing signaling molecules assist in regulation of type III secretion system (243) which provides the ability to activate virulence genes and colonize the host

(275). Among the proteins that form the type III secretion systems is EspA, which plays a role during infection of mammalian cells mediating attachment in early stages, and translocation of effector proteins later in the infection process (93). E. coli O157:H7 cells produce filaments, visualized by scanning electron microscopy, that allow adherence to spinach leaf surfaces. SEM analyses of espA mutants do not show filament production, do not cause human disease, and presumably do not adhere to the spinach leaf surfaces (233).

Attachment is a pre-requisite for establishment of bacteria on plant surfaces (42). Then by using mechanisms of virulence to invade the mammalian host, enteric pathogens could also mediate its survival and establishment on plant surfaces. E. coli O157:H7 and S. enterica produce other non-specific mammalian cell attachment mechanisms including capsules (128), curli (41) and aggregative fimbriae to attach to leaf surfaces with a rpoS dependent mechanism (22).

INTERACTIONS OF ENTERIC PATHOGENS WITH THE PHYLLOSPHERE MEMBERS

Microbial interactions within a microbial community (like the phyllosphere) composed by a mix of different microorganisms, can occur via multiple mechanisms through physical contact or via signaling molecules (235). Alternatively, there are indirect interactions that occur when changes in the physicochemical properties of the environment can be induced by other microorganism in response to other strain (90). The effects of such interactions on the fitness of other microorganisms can be positive, neutral or negative (235) (Table 2). Pathogenic bacteria can adapt to the phyllosphere environment, but they may be outcompeted by epiphytic bacteria or gain epiphytic fitness by interactions with other phyllosphere members (66). The mechanisms that dictate interactions among microorganisms are responsible for the properties of the microbial community as a whole (158).

Table 2. Microbial interactions that can occur within a microbial community.

INTERACTION DESCRIPTION

Positive interactions Symbiosis: All members of the association derive benefit from one another

Neutral interactions Commensalism: One organism derives benefit from the other and the other is neither

harmed nor benefit from the association

Negative interactions Antagonism: Occur via competition for a common resource (i.e carbon sources),

predation of one organism upon another, or production of harmful metabolism that

directly antagonize other microorganism.

When enteric pathogens land on the leaf surfaces, they have to find an ecological niche which already has highly adapted bacteria to this habitat (42), which will compete for resources with other members of the phyllosphere and also with enteric pathogens. In vitro studies have identified several microorganisms, including epiphytic bacteria, with the ability to antagonize the growth of several foodborne pathogens (Table 3). Mechanisms of inhibition likely occur by outcompeting for available nutrients or by the presence of a diffusible factor that limits the growth of foodborne pathogens (66).

There is special interest in these microorganisms due to the potential use as a bio-control agent to prevent enteric pathogens from establishment on the leaf surfaces (104). Bio-control has been demonstrated as strategy for controlling plant pathogens (228, 292). The application of bacterial antagonist towards foodborne pathogens could certainly be part of an integrated strategy to prevent the growth of these microorganisms pre-harvest and post-harvest.

Table 3. Bacterial species isolated from produce items that elicit antagonism towards foodborne pathogens.

Antagonist Source of isolation Pathogen inhibited Reference Green cabbage E. coli O157:H7 (231) P. fluorescens Celery L. monocytogenes Green onions S. enterica var Montevideo Green lettuce S. aureus Iceberg lettuce Iceberg lettuce E. coli O157:H7 (153, 231) P. aeruginosas Alfalfa sprouts E. coli Baby carrot L. monocytogenes Romaine lettuce Diced carrots L. monocytogenes (231) Aeromonas hydrophila Green onions S. enterica var Montevideo Purple cabbage S. aureus Diced green lettuce E. coli O157:H7 (231) Aeromonas salmonicida L. monocytogenes S. enterica var Montevideo S. aureus Alfalfa sprouts L. monocytogenes (153) Bacillus pumilus Baby carrot Romaine lettuce Alfalfa sprouts L. monocytogenes (153) B. mojavensis Baby carrots E. coli Romaine lettuce Alfalfa sprouts L. monocytogenes B. megaterium Baby carrots E. coli Romaine lettuce S. enterica var Chester Arabidopsis thaliana E. coli O157:H7 (66) Enterobacter asburiae Lettuce Lettuce L. monocytogenes (11) Lactic acid bacteria Lettuce L. monocytogenes (257) Leuconostoc spp. Golden delicious apples E. coli Lactobacillus plantarum S. enterica var Weisella spp. typhimurium Lactococcus lactis Fresh cut lettuce E. coli O157:H7 (131) Enterobacter sp. Spinach Fresh cut lettuce E. coli O157:H7 (131) Kleibsella sp. Spinach Fresh cut lettuce E. coli O157:H7 (131) Pantoea sp. Spinach Fresh cut lettuce E. coli O157:H7 (131) Bukholderia sp. Spinach Fresh cut lettuce E. coli O157:H7 (131) Kluyvera sp. Spinach

Mechanisms of inhibition likely occur by outcompeting for available nutrients or by the presence of a diffusible factor (antibiotics) that limits the growth of foodborne pathogens (66).

There is special interest in these microorganisms due to the potential use as a bio-control agent to prevent enteric pathogens from establishment on the leaf surfaces (104). Bio-control has been demonstrated as a strategy for controlling plant pathogens (228, 292). The application of bacterial antagonist towards foodborne pathogens could certainly be part of an integrated strategy to prevent the growth of these microorganisms pre-harvest and post-harvest.

In contrast, if organisms use different carbon and energy sources positive interactions could favor the establishment of enteric pathogens on plant surfaces, allowing co-existence on the phyllosphere and possibly enhancing epiphytic fitness of other populations (19), including enteric pathogens. Using epiphytes isolated from Arabidopsis thaliana and lettuce, it was possible to demonstrate that Wausteria paucula enhanced growth of E coli O157:H7 by approximately 5- fold when co-cultured in sterile lettuce plant exudates. Co-culture of Enterobacter asburiae in sterile lettuce exudates limited growth of E. coli

O157:H7 by approximately 3-logs (66). This antagonism is part due to competition for carbon sources present in the lettuce exudates, E. asburiae and E. coli O157:H7 shared utilization of 25 out of 27 carbon sources. In contrast, W. paucula failed to utilize effectively most of the same substrates (66), and may increase growth of E. coli O157:H7 and possibly aid in its establishment. W. paucula modifies the lettuce tissue, releasing plant exudates that can then be used to supply a suitable niche for E. coli O157:H7 (66).

On the phylloplane, food borne pathogens will interact with bacteria, fungi, and protozoa.

Protozoa, are common members of the native microbiota on fresh vegetables (221), and have the ability to increase survival of enteric bacteria (48, 97). S. enterica and E. coli O157:H7 are both ingested and expelled undigested in protective vesicles by the protist Tetrahymena pyriformis on plant surfaces (48). E. coli O157:H7 and L. monocytogenes can also resist digestion by Glaucoma sp. isolated from spinach and lettuce (97).

Positive interactions with the phyllosphere inhabitants may aid in survival and even protection of enteric pathogens on edible plants and the study of these types of associations may help to understand the ecology of enteric pathogens on the leaf surfaces.

BIODIVERSITY OF THE PHYLLOSPHERE

Due to the possible interaction of the phyllosphere members with enteric pathogens it is important to characterize the members of the phyllosphere, identify mechanisms of survival in the extreme conditions and understand interactions with foodborne pathogens.

A group of microbial individuals that belong to the same species and close to each other in space and time constitute a population (204). Microbial communities are an assemblage of microbial populations, which occupy a common space at the same time (158). They comprise an interwoven matrix of biological diversity often modified by physical and chemical variations over space and time (158). The study of the microbial communities, including the phyllosphere, can provide guidance to strategies to manipulate those communities on plants or animals to suppress pathogens, to maintain community integrity when chemicals are applied to the environment or to successfully introduce beneficial microorganism like bio-control agents in agriculture or as probiotic in veterinary and human medicine

(158).

Study of microbial communities. To study microbial communities it is necessary to study their properties, which are often divided in two main categories: structural and functional. Structural properties describe how the community varies with changes in environmental conditions and its composition.

Functional properties define the community’s behavior in terms of nutrient utilization, environment modification, intra and inter species interaction and how responds to perturbations in the environment

(158, 161).

Characterizing structural properties of a microbial community, requires knowing which members are present, how many of them and the abundance of those members in a defined community, which is often called community composition, richness and evenness, respectively (158). These three parameters are used to calculate diversity indices, which take into account richness and evenness of distribution of the different species present in the community (139). Diversity is an attribute of the community, helpful to compare communities or to compare the same community under different environmental conditions,

34 which often helps to define robustness which is another structural attribute that define the ability of the community to maintain its structural and functional integrity when facing environmental perturbations

(158).

Determining richness implies the characterization of the community and its composition by identifying the different microbial phylotypes within a community sample. To date two main approaches can be used to characterize the microbial communities; those based on culturing and those that are often called non-culture dependent.

Culture dependent techniques involve isolating a microorganism from particular environment

(soil, plants, water, fecal material etc.) that can be grown under certain conditions of nutrients, temperature, and water activity. Culture conditions typically attempt to mimic the conditions in which the microorganisms live. Once the microorganism is cultured, a pure culture of a single microorganism can be further identified by either physiological and/or metabolic traits or using sequencing of its genetic material (158). About two decades ago, several researchers noticed that microscopic counts exceeded viable-cells counts by several orders of magnitude, which has been known as “the great plate count anomaly” (14, 246, 256). This fact has made culture dependent techniques unsuitable to completely describe all members of the community present. The main reasons why “the great plate count anomaly occurs” include: cultivation conditions are not suitable (14), there are unknown species that haven never been cultured due to lack of adequate methodologies and microorganisms and particularly bacteria, can enter in what is known viable but non-culturable state (VBNC) (196, 219). Today is well known that the percentage of culturable bacteria range between 0.1 to 1% (14, 218), which has lead to the development of alternative techniques to increase the number of identified microorganisms within a community (139).

However it is important to point out that culture dependent techniques define functional properties of the microbial communities because they provide morphological and physiological data to infer the function of the different species in a community (158). At the same time, data from culture-dependent experiments allows testing of potential functions from new microbes that are identified without cultivation (161).

Culture independent techniques are based on the analysis of the structural molecules present in a microorganism including: proteins, fatty acids and nucleic acids (139) often called molecular techniques, or by direct observation of the microorganisms through the use of microscopic techniques (38). Among the most popular culture-independent techniques are those that involve the analysis of nucleic acids; DNA and RNA.

Molecular techniques based on nucleic acids began around 1980s (244, 245). In 1986, Pace and collaborators described the use of 16S rRNA molecule as a strategy to study microbial ecology (14, 197).

16S rRNA molecule is a central component of the small subunit of ribosomes, which are involved in the translation of mRNA to proteins (284). Its application in microbial ecology allows identification of prokaryotes and prediction of phylogenetic relationships (197, 284). 16S rRNA genes have characteristics that allow this molecule to be used as molecular marker. It has an average length of 1,500 nucleotides which provides sufficient information for reliable phylogenetic analysis, and contains conserved regions among prokaryotes allowing design of universal molecular hybridization tools useful for amplification and variable regions to differentiate among different microorganisms. It occurs and it has homologous function in all bacteria (284).

A general approach to study a microbial community consists of the isolation of microbial nucleic acids from an environmental sample (Fig. 2). The application of 16S rRNA genes for characterization of microbial communities is often coupled to the polymerase chain reaction (PCR). PCR allows amplification of the 16S rRNA gene, resulting in a mixture of amplicons with different sequences

(metagenome); the amplicons need to be further separated to identify individual sequences through comparison with databases (14, 179).

There are several techniques available for the

separation and analysis of the generated amplicons, which can

determine different properties of the microbial communities.

One of the first techniques developed in the 1980s

was the construction of clone libraries (278). 16S rRNA

amplicons are inserted in a plasmid (cloning) which will be

utilized for the transformation of bacteria, often E. coli

transformants. Each transformed cell will contain a single

fragment in the plasmid, which can be isolated and sequenced.

The main disadvantage is that the screening of the clones

cannot be exhaustive and the richness of the community is

underestimated (291). Clone libraries as well as all the

techniques based in 16S rRNA and PCR, have disadvantages

in two main points: the nucleic acid isolation and the PCR.

Microbial cell lysis is a fundamental step in the

isolation of DNA. The efficiency of the lysis conditions can

vary from microorganism to microorganism due to differences Figure 2. General flowchart for characterization of microbial in the cell wall and cellular membrane (195). Different lysis communities from environmental samples. conditions have shown to alter the description of the microbial community (56) due to different lysis efficiency. In addition during the isolation of nucleic acids, several PCR inhibitors, including protein, humic acids and tannin compounds can decrease the efficiency of the PCR reaction (56). Incomplete detachment of microbial cells from their matrix (i.e plant surface aggregates) can also limit the complete lysis of the entire community (181).

Disadvantages related with the PCR reaction include different affinities of the designed oligos to the DNA template, different copy numbers of the target genes, and preferential amplification towards the more abundant members of the community as well as different amplification related with the GC content of the template, resulting in so called PCR bias (139, 179), which can prevent the amplification of other rRNA resulting in the underestimation of the microbial richness.

Fingerprinting techniques to study microbial communities: Denaturant gradient gel electrophoresis. Other techniques have been developed to allow rapid, simultaneous and reproducible analysis of microbial communities. These techniques include the development of a fingerprint of the community, which besides providing diversity parameters of the community they also provide information regarding spatial and temporal variability and dynamics of the community in response to changes in environmental conditions. In addition, these techniques allow comparison among entire communities (239). Denaturant gradient gel electrophoresis (DGGE); (183), temperature gradient gel electrophoresis (TGGE); (184), single strand conformation polymorphism (SSCP); (147) and terminal restriction fragment length polymorphism (T-RFLP); (255) are fingerprinting techniques which are based in the 16S rRNA and PCR analysis. Fingerprints utilizing functional genes can be amplified from the community, providing the opportunity to study functional properties of the communities (239).

PCR-DGGE is a fingerprinting method that separates nucleic acid fragments based on their sequence composition (Fig. 3). It has been utilized to analyze microbial communities from different environments including the phyllosphere (169, 288). DNA fragments produced by amplification are separated through electrophoresis, which uses a linear gradient of a denaturant solution of urea and formamide. Differential migration occurs due to differences in the content of guanine and cytosine (GC);

DNA fragments with higher GC content require a greater gradient to denature the double strand DNA resulting in a higher number of hydrogen bonds between GC than adenine and thiamine (AT). The migration fragments stop when the double strand is completely opened. To produce sharp bands, fragment

38 migration is stabilized by the addition of a GC clamp to the end of one PCR primer, to prevent complete strand separation (185).

Figure 3. Principle of DNA fragment separation by PCR-DGGE analysis.

The resulting fingerprint is a profile that represents the community structure, an approximation of the numbers of populations is represented by each band and their relative abundance is represented by the band intensity within the amplified community (185). To identify individual bands and assess richness, once bands are separated, they can be cut and separated DNA fragments can be diffused from the gel and sequenced for further band identification (103). Each band might not necessarily be represented by a single sequence because co-migration of bands can occur, which can underestimate the richness and overestimate the abundance of the community. Fragments can be cloned to separate the bands that co- migrated (103, 184).

Phyla specific amplification through real-time PCR. Other PCR-based microbial community analysis is entirely based in the quantification of a particular group of microorganisms, usually at the phylum level, using target specific primers and real-time PCR (RT-PCR) amplification. In conventional

PCR, the amplicons can be detected by end-point analysis, where DNA is detected on an agarose gel after the reaction has finished. In contrast, real-time PCR allows the product to be detected and measured as the reaction progresses in “real-time” (35). Real time detection is possible by including a fluorescent molecule (DNA binding dyes) in the reaction. Fluorescence is detected and its measurement reflects the amount of amplified product in each cycle. Different from conventional PCR is that RT-PCR allows determination of the starting template copy number with high sensitivity making this techniques quantitative (qRT-PCR) (35).

qRT-PCR can be utilized to study microbial communities targeting phylum and class specific

PCR primers for rapid screening of PCR amplicons or metagenomic clone libraries for ribosomal sequences of a particular taxonomic group. The abundance of the targeted group can be monitored through qRT-PCR (36, 80). Phyla specific primers have been designed to target: Proteobacteria, -

Proteobacteria, -Proteobacteria, Firmicutes, Actinobacteria, Bacteroidetes and Basidiomycota (36, 80) in microbial communities from soil (36, 80) and other communities. The derived information can contribute to the description of microbial diversity in terms of its abundance, robustness, and in more general terms, richness. Specific probes targeting selected genera or even species can also be designed for the same purpose.

Pyrosequencing. The determination of richness and composition of the microbial community is one of the main steps to describe microbial diversity (158). PCR-DGGE analyses as well as qRT-PCR provide important information to study structural properties of the community; however the identification of individual species and the richness assessment is often incomplete. The construction of clone libraries provides good information about these two parameters (14), however, the screening of clone libraries is expensive, time consuming and often incomplete. Pyrosequencing is a recently developed technology in which more than 300,000 sequences can be simultaneously determined without cloning. A highly variable region of 16S rRNA gene is amplified using primers that target adjacent conserved regions, followed by direct sequencing of individual PCR products (159, 217). The methodology has been used to describe

40 microbial communities from deep sea environments (112, 240), intestinal tract (17, 73), soils (216, 262) and food (113). The main disadvantage of this methodology is that the amplified sequences range between 100 to 300 nt, which might be unsuitable for performing phylogenetic based community analysis

(159). However different algorithms have overcome this bias (159), which allows phylogenetic classification with the sequences with high fidelity.

It is important to point out that different methodologies provide different information about the structural and functional properties of microbial communities. It is then pivotal to understand the importance of the application of more than one techniques and also the use of culture- and non-culture- dependent techniques, polyphasic analysis, to successfully and accurately describe microbial communities.

Studies of diversity on the phyllosphere. Interactions of phyllosphere bacteria can significantly affect the fitness of plants in the environment, promote plant growth and suppress or stimulate colonization of plant and human pathogens (155, 280).

The inhabitants of the phyllosphere include Achaea, filamentous fungi and yeast however bacteria are considered to be the dominant microbial inhabitants (280). Culturable bacteria number between 102 to 1012 cells per gram of leaf (254). The global surface area occupied by the phyllosphere is estimated to total over 4 x 108 Km2, supporting an estimated bacterial population of 1026 cells (180). Most of the studies on the population dynamics and diversity in the phyllosphere have been made through culture dependent techniques (110). These studies have evidenced more than 85 different species in 37 genera in the phyllospheres of rye, olive, sugar beet and wheat (109, 254).

In 2001, a study of the phyllospheres of different plant species using DGGE (non-culture dependent) revealed that the most dominant phyllosphere organism were not detected by conventional culture-dependent methods (288). In this study 5 out of 17 bands cut from DGGE gels had less than 90% similarity to database entries, suggesting that plants may support novel bacteria (288). In 2006 a different study on 20,000 vascular plants inhabiting the Brazilian Atlantic forest by 16S rRNA analysis, revealed

41 that about 97% of the bacteria were from unknown species and the phyllosphere of any one of the plants carries at least 95 to 671 bacterial species (143). Non-culture independent approaches to study phyllosphere communities have changed the understanding of the structure and diversity of the phyllosphere, and at the same time they have contributed new insights into the complexity of the phyllosphere and their interactions with plants and the environment (280).

Studies in the phyllosphere of different crops using 16S r RNA gene clones libraries, have shown that Proteobacteria is the dominant group found on leaves, with -proteobacteria being the most abundant members, of significant importance in terms of abundance are also Bacteroidetes,

Actinobacteria and Firmicutes and in less abundance Cyanobacteria and Acidobacteria (118, 134, 143,

211). Different bacterial species have been identified from the phyllosphere of agricultural and horticultural crops (Table 4).

Table 4. Representative phyla and genera of bacteria identified through 16S rRNA analysis on the phyllosphere

Phylum/Family/Genera Phyllosphere Technique Reference utilized for analysis

Proteobacteria

-proteobacteria

Sphingomonas sp. Citrus Valencia leaves DGGE (134, 162, 288) Alfalfa sprouts Clone libraries Maize leaves

Bradyrhizobium Alfalfa sprouts Clone libraries (162)

Methylophilaceae Alfalfa sprouts Clone libraries (162)

Acetobacter Alfalfa sprouts Clone libraries (162)

Rhizobium sp. Thlaspi geosingense T-RFLP (118)

-proteobacteria

Comamonadaceae Alfalfa sprouts Clone libraries (134, 162) Maize leaves

Oxalobacteraceae Alfalfa sprouts Clone libraries (162)

Methylophilaceae Alfalfa sprouts Clone libraries (162)

Caulobacter sp. Maize leaves Clone libraries (134)

Roseateles sp. Maize leaves Clones (134) libraries

-proteobacteria

Moraxellaceae Alfalfa sprouts Clone libraries (162)

Acinetobacter sp. Citrus Valencia leaves DGGE (134, 288) Spinach Clone libraries Maize leaves

Marinobacter sp. Citrus Valencia leaves DGGE (288)

Enterobacteriaceae Alfalfa sprouts Clone libraries (134, 162, 209) Spinach DGGE Maize leaves Lettuce

Enterobacter agglomerans Citrus Valencia leaves DGGE (288)

Pseudomonas sp. Alfalfa sprouts Clone libraries (162, 209, 296) Spinach DGGE Lettuce

P. rhodesiae Spinach DGGE (209)

P. fluorescens Thlaspi geosingense T-RFLP (118)

Erwinia sp. Spinach DGGE (209)

Pantoea sp. Spinach DGGE (134, 209, 296) Maize leaves Clone libraries Lettuce

Xanthomonadaceae Alfalfa sprouts Clone libraries (162)

Buchnera Maize leaves Clone libraries (134)

Moraxella sp. Lettuce Clone libraries (296)

Providencia sp. Lettuce DGGE (296)

-proteobacteria

Desulforomonas sp. Citrus Valencia leaves DGGE (288)

Bacteroidetes

Sphingobacteriaceae Alfalfa sprouts Clone libraries (162)

Flavobacteriaceae Alfalfa sprouts Clone libraries (118, 134, 162) Maize leaves T-RFLP Thlaspi geosingense

Actinobacteria

Curtobacterium sp. Thlaspi geosingense T-RFLP (118) Rubrobacteridae Thlaspi geosingense T-RFLP (118)

Norcardioides Thlaspi geosingense T-RFLP (118) Frankineae Lettuce Clone libraries (296) Microbacterium sp. Lettuce Clone libraries (296) Kineococcus sp. Lettuce Clone libraries (296)

Firmicutes

Bacillus sp. Thlaspi geosingense T-RFLP (118, 296) Lettuce Clone libraries

Bacillus pumilus Citrus Valencia leaves DGGE (288)

Paenibacillus sp. Maize leaves Clone libraries (134)

Clostridium bifermentas Citrus Valencia leaves DGGE (288)

Weisella hellenica Maize leaves Clone libraries (134)

Exiguobacterium sp. Maize leaves Clone libraries (134)

Enterococcus Lettuce Clone libraries (296)

Lactobacillus sp. Lettuce Clone libraries (296)

Leuconostoc sp. Lettuce Clone libraries (296)

Planococcus sp. Lettuce DGGE (296)

Deinococcus-Thermus

Deinococcus sp. Maize leaves Clone libraries (134)

Verrucomicrobia

Verrucomicrobiae Thlaspi geosingense T-RFLP (118)

Cyanobacteria

Oscillatoria sp. Lettuce Clone libraries (296)

These studies indicate a high diversity in the phyllosphere, including a great abundance of unknown microorganisms never cultured or previously identified. Research to increase the identification of the members of the phyllosphere can enrich the knowledge about these ecosystems, which will

45 contribute to increase our knowledge and understanding of the phyllosphere diversity. Even more, the ability to culture these unknown bacteria will significantly impact the knowledge of the functional properties of the phyllosphere.

Studies of population dynamics on the phyllosphere. The diversity in the phyllosphere is affected by several factors such as plant genotype and agricultural practices as well as environmental factors (110, 280), directly affect population sizes of phyllosphere communities. Plant species influence diversity and population size (138), which could be the result of leaf physico-chemical properties (280).

Comparison among phyllosphere DGGE fingerprints of orange, cotton, corn sugar beet and green bean leaves, showed that community structure is similar within the same species but varies among different species (288) (also reviewed in Whipps et al. 2008). The population size also differs among different plant species (280). This could be a result of specific plant structure (138); upright, vertical plants and succulent herbaceous plants have higher bacterial populations than waxy broad-leaved and bushy plants that grow closer to the ground (138, 194). Variations among savoy cultivars of spinach had higher populations of inoculated E. coli O157:H7 than flat leaf cultivars, which could be the result of increased sites of colonization and surface area on savoy cultivars (175).

Changes in diversity and population dynamics attributed to seasonality have also been reported.

Lower bacterial diversity was observed on olive and sugar beet leaves during summer and winter seasons than in autumn when conditions were cooler and rainy (74, 254). Phyllosphere of lettuce grown under conventional and organic conditions during spring and autumn seasons had distinguished changes in diversity and microbial populations mostly attributed to seasonal changes rather than to agricultural methods (296). Seasonality differences could result from environmental changes as well as plant age

(254). Studies of the phyllosphere of sugar beet and endive leaves made throughout the growing season, showed larger microbial populations on senescing primary leaves while mature leaves maintained a stable population size. In addition larger bacteria diversity and population size were observed on young and inner leaves than on mature and outer leaves (127, 194, 254).

Environmental conditions to which the phyllosphere is exposed can sharply change during the day (42). These diel changes can significantly affect phyllosphere population size (111). P. syringae population sizes on snap bean leaves were determined at 2 h intervals during three 26 h periods. Overall, population sizes differed during each 26 h period; one population increased continuously despite the intense solar radiation, low humidity and water availability. The other two populations, diel variations and short-term changes in weather conditions did not correlate with the population sizes (111). These results likely suggest that sharp changes in environmental conditions have different effect on bacterial populations, which could be a result of microsite formation on plant surfaces. But given the heterogeneity of the leaf surface physicochemical conditions might not necessarily correlate with the population size

(42). Phyllospheres of orchardgrass and cucumber had larger population sizes when incubated under cool, moist conditions than when incubated under dry conditions (138).

Diversity and population size of the phyllosphere are significantly impacted by environmental conditions, how these changes could become an opportunity for either suppression or enhanced growth of pathogenic bacteria needs to be elucidated. The study of the structural and functional properties of the phyllosphere of edible plants will improve our understanding of human pathogen survival on crop phyllospheres.

SPINACH AS A VEGETABLE CROP

Botany and production requirements. Spinach (Spinacia oleracea) belongs to the family

Chenopodiaceae. This family contains about 75 genera; most of them are weedy plants, but also the important field crops: table beet (Beta vulgaris var. crassa), swiss chard (Beta vulgaris var. cicla), spinach beet (Beta vulgaris var. orientalis) and orach (Atriplex hortensis). Spinach is classified as a dioecious plant although different sexual types occur, including monoecious and hermaphrodicitc types which are result of genetic variation and environmental influences (220). This family is characterized for having succulent tissues, tolerance to salinity, low tolerance of acidity and small inconspicuous flowers that are wind pollinated (220). Spinach is a native of Asia (likely origin in the Persian region) and it has

47 been cultivated in China since 7th century. Its use in Europe dates as early as the mid 13th century, with seed accompanying colonist to the New World (268).

Spinach growth occurs under cool conditions that range from 18 to 20°C but its growth is most rapid at 10 to 15 °C (150). Spinach withstands temperatures of -9 to -6°C without injury better than most vegetable crops, however freezing is more harmful to small seedlings and immature plants. At maturity the crop can tolerate subfreezing temperatures for weeks (150). Growth temperature can affect leaf quality; low temperatures increase leaf thickness but decrease size and smoothness. For seed germination, optimum temperatures ranged between 5 and 15°C. However, emergence time tends to increase with decreasing temperature. High temperatures can be detrimental for vegetative development of the spinach.

High temperatures and lengthening days favor spinach bolting and once flowering occurs, vegetative growth almost stops reducing market value (220). Bolting can be prevented by using slow bolting or bolting resistant cultivars, and by minimizing physiological stress such as increased day length, high temperatures, close plant spacing and inadequate fertility and irrigation (150). Spring planted spinach is subjected to increasing temperature and day length while spinach planted for fall is exposed to declining temperature and decreasing day length (253).

Spinach can be grown in a wide variety of soil types, but those with high moisture holding capacity and good drainage are often preferred, as well as sandy loams and mucks. Nitrogen fertilization of soils is particularly important, especially during winter production, because of little nitrification at low soil temperatures (150).

In California, sprinkler irrigation is used for seed germination and to prevent soil crusting. Once a uniform stand is established, most producers switch to furrow irrigation. Some growers use sprinkle irrigation during the entire season, however this practice favors infection and spread of spotting diseases

(150). The moisture requirements are low because of the low transpiration rate of the spinach during the cool season (220). Between emergence and harvest one to three irrigations are usually required, totaling 6

48 to 12 acre-inches of water for a fresh market crop and 18 to 24 acre-inches of water for a processing crop

(150).

Harvest and post-harvest of spinach. Spinach achieves a marketable size from of 30 to 80 days depending on the season and temperature, and it can be as long as to 150 days for overwinter production.

Plants ready for harvest should have 5 to 8 fully developed leaves, although a single seedling can produce about 25 leaves. Delayed harvest might increase plant weight but significantly reduce the crop quality

(220). For fresh market, plants are pulled by hand or undercut below the stem and tied in bunches directly in the field. Harvesting later in the day, when plants are less turgid, is a common practice to minimize leaf damage. For processing markets spinach leaves are mechanically harvested cutting about 10-15 cm above the growing point, which allows re-growth for a second harvest three to four weeks later. Spinach can be processed by either canning or freezing (150).

Harvested spinach is highly perishable due to the large surface-to-weight ratio and the high respiration rate after harvesting, so rapid cooling is essential to prevent wilting and weight loss (220).

Hydrocooling is often used to remove heat field or spinach can be vacuum cooled. For short term cooling, ice can be used (251). Fresh bunched spinach stored at 0°C at 95-100% humidity with light misting will typically have a shelf life of 10 to 14 days.

Spinach cultivars. Spinach cultivars are classified based on the leaf type (150):

Savoy (curly or wrinkled)

Semi-savoy

Smooth (flat)

Most commercial cultivars are smooth-seeded to facilitate handling and accurate planting.

Commercial spinach cultivars are hybrids with disease and bolting resistant. Foliar disease caused by

Peronospora farinosa f. sp. spinaciae is the most destructive spinach disease worldwide. Single-gene resistance is available (67).

Savoy leaf types are primarily used for fresh market while smooth types for processing because their wrinkle –free leaves are easier to wash clean (150). Semi-savoy cultivars are grown for either market

(267). In the past the use of smooth cultivars in the fresh market was limited because bunched and bulk product quality declined during shipping; flat leaves tend to compact together decreasing the airflow.

However due to strong growth in packaged fresh spinach sales, along with the ease washing of flat cultivars, the use of smooth cultivars in the fresh market has increased in the past years (267).

Spinach production in United States. The United States is the world’s second-largest producer of spinach, with 3% of the world production, following China which accounts for 85% (267). In United

States, the majority of spinach is produced in California (73%) followed by Texas (15%), Arizona (10%) and New Jersey (3%), with 12 other states that report at least 100 acres (267).

California produces spinach in four areas: The southern coast (Santa Barbara and Ventura

Counties); the central coast (Monterey, San Benito, Santa Clara, And Santa Cruz counties); the southern desert valleys (Imperial and Riverside counties) and the central San Joaquin Valley (Stanislaus and Tulare counties). Half of California’s spinach acreage and production is in Monterey County. The Southern coast and San Joaquin Valley each produces about one fourth of the acreage (150).

Spinach is produced round year in California; Coachella Valley planting occurs from October through December for harvest from November through March. In San Joaquin Valley planting starts in late October through January for harvest from February through April (150).

The farm value of US spinach crop for fresh and processing market was valued in $175 million for 2004-2006, with fresh market spinach accounting for 94%. California accounts for three-fourths of the value for fresh and processing crops (268).

About 96% of the spinach consumed in US is produced domestically, although imports (largely from Mexico) totaled 23 million pounds in 2006 compared with 3 million in 1996. During the last 20

50 years exports have raised by 70% to 47 million pounds in 2006 with Canada being the major importer

(268).

Trends of spinach consumption. Spinach is consumed fresh or processed either frozen or canned. Among processed spinach, frozen products accounted for 83% of the production (267). Canned production has significantly decreased since 1940. The use of spinach in processed forms has trended lower as consumer demand has shifted toward fresh-market, bagged spinach in the past three decades

(268).

Per capita fresh spinach consumption by 1972 was 0.29 pound per person, by mid-1970s the increasing popularity of salad bars along with widespread interest in nutritional benefits of this crop help raise the per capita consumption reported in 1.6 pounds per person in 2003 (267).

Triple-washed cello-bagged spinach and baby spinach are two of the faster growing segments of the packaged salad industry, accounting for around one-tenth of supermarket sales in the 2 billion fresh- cut salad industry in 2002 (268). Similar trends have been observed for romaine lettuce and bagged salad mixes (269).

PRODUCE POST-HARVEST AND IMPLICATIONS IN FOOD SAFETY

After harvest physiological changes on leaf tissues occurs as a result of the plant metabolism as well as activities of indigenous microbiota. These changes have important effects on the sensorial and microbiological quality (214). The quality of fresh or minimally processed vegetable product is determined by the appearance, texture and flavor (133). Besides these aspects, consumers are also interested in the nutritional quality and safety (214).

Harvest leads to mechanical injury of plant tissues, which will initiate physiological changes associated with the plant homeostasis trying to repair wounds and prevent infection of opportunistic microorganisms (70). When injury occurs, disruption of natural barriers to microorganisms, including human pathogens occurs, which pose a health risk in produce consumption. E. coli O157:H7 inoculated

51 on lettuce leaves that were mechanically bruised and cut leaves, showed a population increase 4- and 11- fold, respectively, within a period of 4 hours, implying that plant lesions promote the rapid multiplication of this microorganism (43). Using confocal microscopy, the same study revealed that the population of E. coli O157:H7 was localized internally and externally (43). S. enterica and Shigella sonnei, also grow more rapidly and achieved larger population size on cut leaves of cilantro and parsley (53, 286). These studies strongly suggest that disruption of plant natural barriers promote growth of pathogenic microorganism, most likely by increasing the availability of nutrients leached from the plant. Therefore, harvesting methods that minimize disruption of plant tissues could be a critical control point to prevent or reduce the growth of human foodborne pathogens (43).

Post-harvest vegetables are living plants which continue respiring as long as nutrients and gases are available. During harvest, the plant leaf is detached from water and nutrient sources which will limit their physiology and transpiration (227). After harvest the respiration rate increases limiting shelf life

(227). Spoilage attributed to the produce’s native microbiota which will contribute to tissue deterioration as well (214). Produce is often further processed before marketed, although the process in most of the cases is minimal, it does promote faster physiological deterioration, biochemical changes and spoilage

(214). Minimal processing of produce often includes several washing steps, peeling, slicing, dicing or shredding (144), which can disrupt plant tissue.

In order to prevent produce deterioration, modified atmosphere packaging (MAP) is often used in combination with storage at low temperatures to control growth of spoilage microorganisms and to minimize produce’s respiration rate (214). MAP is a common technique used for several produce items including baby spinach and ready to eat salads mixes. MAP is a preservation technique in which gas composition is altered from that of air (9). In most of the cases, low levels of O2 and high levels of CO2 reduce produce respiration rate, delay senescence and extend storage life of fresh produce (225). The modified atmosphere can be achieved passively or actively. In passive modification, the package is sealed with normal air and the change in the atmosphere occurs by produce respiration and film gas permeability

(238) as occurs with bagged baby spinach (208). Active modification implies flushing the packages with gas mixtures before closed, mixtures can include CO2, N2 and O2 and in some cases noble gases and CO are also used (227).

Temperature control is another important factor for maintaining the quality and shelf life of produce after harvest (33). Fluctuations in temperature occur from harvest to the point of display for sale

(83). After harvest a rapid reduction in temperature is needed to reduce produce damage (usually about to about 5°C), but the storage of the recently harvested produce might range from 5 to 8°C for a period of

24h after harvest. Produce is exposed to temperatures above 10°C during transportation, processing

(including packaging) and typically at the point of sale, sometimes as high as 15°C (33). Low temperatures reduce plant metabolism, extending shelf life, and controlling bacterial growth, therefore minimal fluctuations during transportation and retail display can alter the microbial population size in packaged produce.

Microbiota on plants in the field will remain associated with vegetables after harvest, during transportation, further processing and storage (117). Additional microbiota could be added by using contaminated harvest tools, during hydrocooling, transfer by farm workers, work surfaces and during washing. All of these could also be an entrance point for pathogenic microorganism (post-harvest contamination) (117). The initial microbial population differs from crop to crop but it ranges from 3 to 7 log CFU/g (188). Bacteria significantly contribute to post-harvest microbial spoilage, much of these are bacteria that cause soft rots, spots, blights and wilts. The cultured bacteria most often associated for spoilage include:: Erwinia carotovora¸ Pseudomonas fluorescens, Pseudomonas sp., Corynebacterium sp., Xantomonas sp. and lactic acid bacteria (117).

The microbial community of packaged salads stored at 4 and 10°C has been shown, using culture-independent techniques, to be dominated by members of the Enterobacteriaceae, with

Pseudomonas sp. as the dominant species (222). Storage of spinach under MAP conditions at 5°C resulted in a 5 to 8 log CFU/g increase of the culturable mesophilic and psychrotrophic populations in a

53 period of 8 days (65). Storage at 10°C, for 12 days resulted in even larger population increases of mesophilic, psychrotrophic bacteria, with Pseudomonas sp. reaching 10 log CFU/g (20). MAP packing often results in limited reduction of bacterial populations, which is why MAP packaging is coupled with storage at temperatures below 5°C to control microbial growth (293). Populations of two common epiphytes, Pseudomonas sp. and Enterobacteriaceae, were not significantly reduced during typical MAP storage conditions (1.5 to 21% of O2 ;5 and 20% of CO2 at 8°C (293). Reductions were seen in 20% CO2, however at this high composition of CO2 the plant cells are physiologically damaged, affecting the sensory quality of the product (208).

Of greater concern is the fate of pathogens on MAP produce. As seen in studies of the native bacteria, E. coli O157:H7 populations are not reduced by MAP alone, however when combined with storage at 8°C or below no growth occurs and viability is often reduced (63, 70). Populations of S. enterica and E. coli O157:H7 remain viable after several days of storage (6, 71, 85). S. enterica var.

Enteritidis and E. coli O157:H7 inoculated at 6 log CFU/g onto basil, chives, cilantro, parsley and rosemary were only reduced by 1- fold after storage for 19 days at 4°C, during which the product showed no signs of spoilage. In another study, MAP packaged shredded lettuce inoculated with 4 log CFU/g of E. coli O157:H7 stored at 4 and 8°C for 12 days, showed that at 4°C the E. coli populations did not increase significantly, however an increase of 2 log CFU/g was detected in spinach stored at 8°C (85), indicating the importance of avoiding temperature abuse to prevent the growth of pathogenic bacteria and as discussed before to control the growth of native microbiota.

In fresh as well as packaged spinach, native microbiota and human pathogens coexist and might share the same niche. It has been suggested that techniques that aim to increase shelf life by controlling growth of spoilage microorganisms, might enhance growth of pathogens due to a reduction of competing microbiota (208). The microbial interactions that occur post-processing and during storage in produce might play a major role in the survival of pathogenic bacteria (33), but this field has not been explored.

A field of interest has been the effect of minimally processing and storage of vegetables on the physiology of E. coli O157:H7. During storage, low temperatures and changes in atmosphere composition could be stress factors that impact the fate of E. coli O157:H7. A recent study made on romaine lettuce stored at 4 and 15°C for a period of 9 days, analyzed the expression of different stress and virulence genes. At 4°C up-regulation of sodB and rpoS (oxidative stress and general stress response genes, respectively) occurred during the first days of storage, and stx genes (shiga-like-toxin) tend to up- regulation during the storage (55). Another study made with fresh-cut lettuce in MAP stored at 10 and

15°C inoculated with E. coli O157:H7 defective rpoS, showed that storage temperature greater than 10°C enhanced resistance to gastric acid challenge. This effect was associated with the utilization of alternative acid-resistant pathways activated under low atmospheric conditions in combination with temperature of storage (62). These results suggest that initial adaptation to changing conditions due to storage might occur, allowing adaptation of this bacterium to new conditions by triggering general stress responses. But even more important is that virulence can be enhanced during storage, which might have important consequences for consumer safety.

Strategies to reduce microbial load and microbial growth on post-harvested produce.

Besides low temperature storage and MAP, other technologies have been developed to use in conjugation with MAP to reduce the initial bacterial numbers, control microbial growth, in order to minimize the risk due to the presence of human pathogens and to extend the shelf-life (Table 5

Table 5. Summary of post-harvest procedures utilized on fresh produce to extend shelf-life (214).

Treatment Description

Chlorine Liquid chlorine and hypochlorite are used as disinfectants at levels of 50-200 ppm of free chlorine with contact times of less than 5 min. The objective of this treatment is to decontaminate vegetables and it attempt to eliminate pathogens, using its oxidation capacity, however its effectiveness is questionable.

Chlorine dioxide It has higher oxidation capacity than chlorine (28) and it does not react with nitrogen-containing compounds to form dangerous chloramines (214). Its activity against pathogens on produce has showed to be effective with E. coli O157:H7 (236), L. monocytogenes and S. enterica var. Typhimurium (149), when used in combination with other treatments.

Organic acids Lactic acid, citric acid, acetic acid, ascorbic acid and tartaric acid, have strong antimicrobial activity against psychrophilic and mesophilic microorganisms (23). They reduce pH in the environment, disrupting membrane transport and permeability, and causing anion accumulation and reduction of internal cellular pH. It has acted successfully against vegetable microbiota (207). Various organic acids are generally recognized as safe (GRAS) which facilitates their application in food matrices.

Hydrogen Possesses bactericidal and inhibitory activity due to its oxidant activity and to its peroxide capacity to generate cytotoxic hydroxyl radicals (132). Reports have described its ability to reduce vegetable microbiota on several vegetables products (30), however it might cause browning in a few items like shredded lettuce.

Calcium-based Calcium besides its antibacterial properties by uncoupling microbial transport solutions processes (224), it helps to maintain vegetable cell wall integrity by interaction with pectin to form calcium pectate and reduction of chlorophyll and protein loss (21). Calcium has been showed to be effective in reducing and keeping microbiota populations (214).

Ozone It has strong antimicrobial activity as well as high reactivity penetrability and spontaneous decomposition to a non-toxic product (100). It is effective to extend shelf life of fresh non-cut vegetables by reducing microbiota populations and by oxidation of ethylene (31). Studies with Shigella sonnei demonstrated inactivation of this microorganism after ozone exposure (24). However is is highly corrosive and the initial capital cost is too high (214).

Electrolysed water It is electrolysed oxidizing water, generated by electrolysis of sodium chloride to produce a basic aqueous solution at the cathode and an electrolysed acidic solution at the anode. Acidic electrolysed water has a strong bactericidal effect against pathogens and spoilage microorganisms due to a high oxidation reduction potential (23). Inactivation of several human pathogens E. coli and other indigenous vegetable microbiota by electrolysed water has shown to be effective (214).

Treatment Description

Irradiation It uses low-dose gamma irradiation to reduce bacteria, parasitic and protozoan pathogens, without altering sensorial quality, and it has been approved by the FDA for use in fruit and vegetables at a maximum level of 1 KGy (214).

Ultraviolet light It acts as an antimicrobial agent due to direct DNA damage, it also induces synthesis of anthocyanins and stilbenoids (54) which are health beneficial. On lettuce UV light reduce deterioration and reduce microbiota populations (10).

High pressure Food is subjected to 3000-8000 bars of pressure by which microorganisms and enzymes are inactivated without altering nutritional and sensorial quality of vegetables. However this technology can affect the integrity of porous products and it disrupts food tissues which limits its application in fresh vegetables (198).

The effectiveness of the treatments previously described has been evaluated in combination with

MAP and/or refrigeration, for elimination of foodborne pathogens and control of indigenous microbiota

(Table 6).

Table 6. Fate of foodborne pathogens after application of technologies to extend produce shelf life.

Treatment Microorganism Conditions Result Reference targeted

Chlorination E. coli O157:H7 Inoculated cut romaine lettuce and baby Less than 1-log reduction for romaine lettuce (189) spinach with 7 log CFU/g of cells forced to Non significant reduction was observed for spinach internalization of tissue. Produce were leaves. subjected to three washes of 300 to 600 ppm of sodium hypochlorite .

Chlorination, E. coli O157:H7 Spinach inoculated with 5-6 log CFU/g of Decreased levels of E. coliO157:H7 by 2.6 and 1.1 (148) carbon cells, was washed with 100ppm of chlorine log CFU/g with chlorine dioxide and sodium dioxide and dioxide or 100 ppm of sodium hypochlorite hypochlorite treatments, respectively. MAP for 5 min, bagged under MAP and stored at 7°C.

Carbon E. coli K12 Inoculated fresh spinach leaves with 7 log Populations of E. coli K12 were reduced to non- (295) dioxide and CFU/g of cells was exposed to 40 min to detectable levels. high dense phase carbon dioxide (DPCD) with pressure high pressure (5MPa).

Ozone E. coli O157:H7 Baby spinach inoculated with 7log CFU/g of After 30 min of ozone treatment, decreased E. coli (273) cells, treated with ozone gas in combination O157:H7 populations by 4.1 to 5 log CFU/g after of vacuum cooling to mimic agricultural three days of storage at 4°C. practices.

Gamma E. coli O157:H7 Inoculated cut romaine lettuce and baby 4 log reduction and 3 log reduction for romaine (189) irradiation spinach with 7 log CFU/g of cells forced to lettuce and spinach respectively, at minimum internalization of vegetable tissues. irradiation level of 75 KGy Application of 0.25 to 1.5 KGy of gamma irradiation was applied.

Gamma E. coli O157:H7 Lettuce inoculated with 7 log CFU/g of cells Chlorination alone reduced the E. coli population by (82) irradiation and indigenous was chlorinated with 200 ppm and irradiated 1-2 logs while irradiation plus chlorination produced and microbiota with different doses of gamma irradiation a 5.4 log reduction of E. coli O157:H7. chlorination (0.15 to 0.55KGy. Irradiation at 0.55KGy reduced standard plate, yeast and mold counts by 2 logs.

Treatment Microorganism Conditions Result Reference targeted

Gamma E. coli Cucumber, spinach and seasoned burdock Exposure to 1KGy of irradiation, reduced by 4-log (290) Irradiation were exposed to 30KGy of irradiation at 12°C populations of S. aureus while exposure to2KGy S. enterica var to first eliminate indigenous microbiota and reduced population of L. ivannovii, E. coli and S. Typhimurium after inoculated with 7 log CFU/g of cells and enterica by 4 log. L. ivanovii exposed to gamma irradiation. Exposure to 3KGy irradiation reduced all the Staphylococcus pathogen below the limit of detection. aureus

UV ligth Hepatitis A virus Strawberries, green onions and lettuce were UV light on lettuce, green onion and strawberries (81) spot inoculated with 107-109 tissue culture showed inactivation of 4.5 to 4.6 log, 2.5 to 5.6 log infective doses per mL (TCID) of each virus and 1.9 to 2.6 log of TCID/mL, respectively. and exposed to UV light (<240mW s/cm2).

UV light and S. enterica Romaine lettuce and baby spinach inoculated On spinach, S. enterica and E. coli surface (101) hydrogen on their surfaces with 5 log CFU/g of each populations were reduced by 3 and 2 log, E. coliO157:H7 peroxide bacterium and were sprayed with 1.5% H2O2 respectively. under constant UV illumination (37.8 On romaine lettuce S. enterica and E. coli surface mJ/cm2) and stored at 4°C for four hours. populations were reduced by approximately 2 log.

The results of these studies certainly suggest that some technologies could be effective in eliminating human pathogens from contaminated produce; however the levels of inoculation tend to be higher than the real levels of human pathogens on produce. Even when significant reduction of pathogen population occurs, the low infective dose of many pathogens could make these technologies ineffective for the purpose of reducing numbers of human illnesses associated with eating contaminated produce.

Therefore preventive strategies that impair the entrance of human pathogens at the beginning of the food supply chain are necessary to control human foodborne pathogens.

PREVENTIVE STRATEGIES TO MINIMIZE THE RISK OF PRODUCE CONTAMINATION

WITH FOODBORNE PATHOGENS

There is a great concern about produce food safety due to the increased number of outbreaks associated with produce vehicles (domestic and imported) (237). Prevention of produce contamination from enteric pathogens has been encouraged by many government agencies in the United States in an effort to offer consumers a safe food supply.

In response to the high increased number of foodborne outbreaks, by 1997, (then) President Bill

Clinton announced a Food Safety initiative to improve the safety of the United States food supply. In May

1997, the Department of Health and Human Services and the U.S Department of Agriculture (USDA) and the Environmental Protection Agency (EPA) issued a report that identified produce as an area of concern.

As result, in October 1998, the USDA and the US Food and Drug Administration (FDA) developed and issued a guidance document for the fruit and vegetable industry titled “Guide to minimize microbial food safety hazards for fresh fruits and vegetables” (205).

The guide describes good agricultural practices (GAPs) and good management practices (GMPs) to minimize the risk of microbial contamination in fresh produce; it addresses operations related with growing, harvesting, packing and transporting fresh produce. The guide might be implemented by individual growers, packers and shippers (79). It identifies the critical points that need to be considered

60 during agricultural production; water, manure, processing, workers health, keeping records and trace back procedures (79).

The guide cannot be enforced by law and it only focuses on microbial hazards and risk reduction but not risk elimination (79). GAPs are generalized rules that might not apply to all agricultural facilities, necessitating development of commodity specific food safety guidelines, which consider different crops, growth environment, water resources and geography (34). A few examples are:

California leafy green handler marketing agreement (Encourage implementation of

GMPs and GAPS through certification) (152).

Florida Department of Agriculture, (packing of fresh tomatoes) (49).

Commodity specific food safety guideless for the production and harvest of lettuce and

leafy greens (119).

In addition since 2007, FDA is implementing a Food Protection Plan that addresses both food safety and food defense for domestic and imported products (78) including produce. This plan is focused on risk over a product’s life cycle from production to consumption, and it provides three elements of protection: 1. Prevent foodborne contamination, 2. Intervene at critical points in the food supply chain and

3. Respond rapidly to minimize the harm (78).

Joint efforts by U.S. government and produce industry are focused in the improvement of the microbiological quality of produce items, which hopefully will have an impact in human health by preventing both, illnesses caused by foodborne pathogens and chronic diseases due to the consumption of safe and healthy fresh fruits and vegetables.

REFERENCES

1. 2007. Multistate outbreaks of Salmonella infections associated with raw tomatoes eaten in

restaurants--United States, 2005-2006. MMWR Morb Mortal Wkly Rep 56:909-11.

2. 2008. Outbreak of Salmonella serotype Saintpaul infections associated with multiple raw produce

items--United States, 2008. MMWR Morb Mortal Wkly Rep 57:929-34.

3. 1997. Outbreaks of Escherichia coli O157:H7 infection and cryptosporidiosis associated with

drinking unpasteurized apple cider--Connecticut and New York, October 1996. MMWR Morb

Mortal Wkly Rep 46:4-8.

4. 2007. Preliminary FoodNet data on the incidence of infection with pathogens transmitted

commonly through food--10 states, 2006. MMWR Morb Mortal Wkly Rep 56:336-9.

5. 2008. Preliminary FoodNet data on the incidence of infection with pathogens transmitted

commonly through food--10 states, 2007. MMWR Morb Mortal Wkly Rep 57:366-70.

6. Abdul-Raouf, U. M., L. R. Beuchat, and M. S. Ammar. 1993. Survival and growth of

Escherichia coli O157:H7 on salad vegetables. Appl Environ Microbiol 59:1999-2006.

7. Ahmed, W., J. Tucker, K. A. Bettelheim, R. Neller, and M. Katouli. 2007. Detection of

virulence genes in Escherichia coli of an existing metabolic fingerprint database to predict the

sources of pathogenic E. coli in surface waters. Water Res 41:3785-91.

8. Ahmer, B. M. 2004. Cell-to-cell signalling in Escherichia coli and Salmonella enterica. Mol

Microbiol 52:933-45.

9. Al-Ati, T., and J. H. Hotchkiss. 2002. Application of packaging and modified atmosphere to

fresh-cut fruits and vegetables. In O. Lamikanra (ed.), Fresh-cut fruits and vegetables. Science,

technology and market. CRC Press, Boca Raton, FL.

10. Allende, A., and F. Artes. 2003. UV-C radiation as a novel technique for keeping quality of

fresh processed 'Lollo Rosso' lettuce. Food Res Int 36:739-746.

11. Allende, A., B. Martinez, V. Selma, M. I. Gil, J. E. Suarez, and A. Rodriguez. 2007. Growth

and bacteriocin production by lactic acid bacteria in vegetable broth and their effectiveness at

reducing Listeria monocytogenes in vitro and in fresh-cut lettuce. Food Microbiol 24:759-66.

12. Altekruse, S. F., Hyman, K. Klontz, B. Timbo, and L. Tollefson. 1994. Foodborne bacterial

infections in individuals with the human immunodeficiency virus. South Med J 87:169-73.

13. Altekruse, S. F., M. L. Cohen, and D. L. Swerdlow. 1997. Emerging foodborne diseases.

Emerg Infect Dis 3:285-93.

14. Amann, R. I., W. Ludwig, and K. H. Schleifer. 1995. Phylogenetic identification and in situ

detection of individual microbial cells without cultivation. Microbiol Rev 59:143-69.

15. Anand, S. K., and M. W. Griffiths. 2003. Quorum sensing and expression of virulence in

Escherichia coli O157:H7. Int J Food Microbiol 85:1-9.

16. Anderson, G. L., S. J. Kenney, P. D. Millner, L. R. Beuchat, and P. L. Williams. 2006.

Shedding of foodborne pathogens by Caenorhabditis elegans in compost-amended and

unamended soil. Food Microbiol 23:146-53.

17. Andersson, A. F., M. Lindberg, H. Jakobsson, F. Backhed, P. Nyren, and L. Engstrand.

2008. Comparative analysis of human gut microbiota by barcoded pyrosequencing. PLoS One

3:e2836.

18. Andrews, J. H., and R. F. Harris. 2000. The Ecology and biogeography of microorganisms on

plant surfaces. Annu Rev Phytopathol 38:145-180.

19. Aruscavage, D., K. Lee, S. Miller, and J. LeJeune. 2006. Interaction affecting the proliferation

and control of human pathogens on edible plants. J Food Sci 71:89-99.

20. Babic, I., S. Roy, A. E. Watada, and W. P. Wergin. 1996. Changes in microbial populations on

fresh cut spinach. Int J Food Microbiol 31:107-19.

21. Baker, R. A. 1993. Firmness of canned grapefruit sections improved with calcium lactate. J Food

Sci 58:1107-1110.

22. Barak, J. D., L. Gorski, P. Naraghi-Arani, and A. O. Charkowski. 2005. Salmonella enterica

virulence genes are required for bacterial attachment to plant tissue. Appl Environ Microbiol

71:5685-91.

23. Bari, M. L., Y. Sabina, S. Isobe, T. Uemura, and K. Isshiki. 2003. Effectiveness of

electrolyzed acidic water in killing Escherichia coli O157:H7, Salmonella enteritidis, and Listeria

monocytogenes on the surfaces of tomatoes. J Food Prot 66:542-8.

24. Barry-Ryan, C., and D. O'Beirne. 1999. Ascorbic acid retention in shredded iceberg lettuce as

affected by minimal processing. J Food Sci 64:498-500.

25. Bastos, R. K., and M. D. D. 1995. The bacterial quality of salad crops drip and furrow irrigated

with waste stabilization pond effluent: an evaluation of WHO guidelines. Water Sci Technol

12:425-430.

26. Beattie, G. A., and S. E. Lindow. 1999. Bacterial colonization of leaves: a spectrum of

strategies. Phytopathology 89:353-9.

27. Bell, B. P., M. Goldoft, P. M. Griffin, M. A. Davis, D. C. Gordon, P. I. Tarr, C. A. Bartleson,

J. H. Lewis, T. J. Barrett, J. G. Wells, and et al. 1994. A multistate outbreak of Escherichia

coli O157:H7-associated bloody diarrhea and hemolytic uremic syndrome from hamburgers. The

Washington experience. JAMA 272:1349-53.

28. Benarde, M. A., B. M. Israel, V. P. Olivieri, and M. L. Granstrom. 1965. Efficiency of

chlorine dioxide as a bactericide. Appl Microbiol 13:776-80.

29. Bergen, W. G., and D. B. Bates. 1984. Ionophores: their effect on production efficiency and

mode of action. J Anim Sci 58:1465-83.

30. Beuchat, L. R. 1999. Survival of enterohemorrhagic Escherichia coli O157:H7 in bovine feces

applied to lettuce and the effectiveness of chlorinated water as a disinfectant. J Food Prot 62:845-

31. Beuchat, L. R., B. V. Nail, B. B. Adler, and M. R. Clavero. 1998. Efficacy of spray application

of chlorinated water in killing pathogenic bacteria on raw apples, tomatoes, and lettuce. J Food

Prot 61:1305-11.

32. Beutin, L., D. Geier, H. Steinruck, S. Zimmermann, and F. Scheutz. 1993. Prevalence and

some properties of verotoxin (Shiga-like toxin)-producing Escherichia coli in seven different

species of healthy domestic animals. J Clin Microbiol 31:2483-8.

33. Bhagwat, A. A. 2006. Microbiological safety of fresh-cut produce: Where are we now?, p. 121-

165. In K. R. Matthews (ed.), Microbiology of fresh produce. ASM Press, Washington, DC.

34. Bihn, E. A., and R. B. Gravani. 2006. Role of good agricultural practices in fruit and vegetable

safety. In K. R. Matthews (ed.), Microbiology of fresh produce. ASM Press, Washington DC.

35. BIORAD. Real-time PCR application guide. BIORAD Laboratories Inc. USA 5279.

36. Blackwood, C. B., A. Oaks, and J. S. Buyer. 2005. Phylum- and class-specific PCR primers for

general microbial community analysis. Appl Environ Microbiol 71:6193-8.

37. Borges-Baldotto, L., and F. Lopes-Olivares. 2008. Phylloepiphytic interaction between bacteria

and different plant species in atropical agricultural system. Can J Microbiol 54:918-931.

38. Bottari, B., D. Ercolini, M. Gatti, and E. Neviani. 2006. Application of FISH technology for

microbiological analysis: current state and prospects. Appl Microbiol Biotechnol 73:485-94.

39. Bowen, D. J., and S. A. Beresford. 2002. Dietary interventions to prevent disease. Annu Rev

Public Health 23:255-86.

40. Boyce, T. G., D. L. Swerdlow, and P. M. Griffin. 1995. Escherichia coli O157:H7 and the

hemolytic-uremic syndrome. N Engl J Med 333:364-8.

41. Boyer, R. R., S. S. Sumner, R. C. Williams, M. D. Pierson, D. L. Popham, and K. E. Kniel.

2007. Influence of curli expression by Escherichia coli 0157:H7 on the cell's overall

hydrophobicity, charge, and ability to attach to lettuce. J Food Prot 70:1339-45.

42. Brandl, M. T. 2006. Fitness of human enteric pathogens on plants and implications for food

safety. Annu Rev Phytopathol 44:367-92.

43. Brandl, M. T. 2008. Plant lesions promote the rapid multiplication of Escherichia coli O157:H7

on postharvest lettuce. Appl Environ Microbiol 74:5285-9.

44. Brandl, M. T., and R. Amundson. 2008. Leaf age as a risk factor in the contamination of lettuce

with Escherichia coli O157:H7 and Salmonella enterica. Appl Environ Microbiol.

45. Brandl, M. T., and S. E. Lindow. 1998. Contribution of indole-3-acetic acid production to the

epiphytic fitness of Erwinia herbicola. Appl Environ Microbiol 64:3256-63.

46. Brandl, M. T., and R. E. Mandrell. 2002. Fitness of Salmonella enterica serovar Thompson in

the cilantro phyllosphere. Appl Environ Microbiol 68:3614-21.

47. Brandl, M. T., B. Quinones, and S. E. Lindow. 2001. Heterogeneous transcription of an

indoleacetic acid biosynthetic gene in Erwinia herbicola on plant surfaces. Proc Natl Acad Sci U

S A 98:3454-9.

48. Brandl, M. T., B. M. Rosenthal, A. F. Haxo, and S. G. Berk. 2005. Enhanced survival of

Salmonella enterica in vesicles released by a soilborne Tetrahymena species. Appl Environ

Microbiol 71:1562-9.

49. Bronson, C. H. Division of fruit and Vegetables: Tomatoes. Florida Department of Agriculture

and Consumer Services.

50. Bunster, L., N. J. Fokkema, and B. Schippers. 1989. Effect of Surface-Active Pseudomonas

spp. on Leaf Wettability. Appl Environ Microbiol 55:1340-1345.

51. Callaway, T. R., T. S. Edrington, J. L. Rychlik, K. J. Genovese, T. L. Poole, Y. S. Jung, K.

M. Bischoff, R. C. Anderson, and D. J. Nisbet. 2003. Ionophores: their use as ruminant growth

promotants and impact on food safety. Curr Issues Intest Microbiol 4:43-51.

52. Calvin, L. 2007. Outbreak linked to spinach forces reassessment of food safety practices.

AMBER WAVES 5:24-31.

53. Campbell, J. V., J. Mohle-Boetani, R. Reporter, S. Abbott, J. Farrar, M. Brandl, R.

Mandrell, and S. B. Werner. 2001. An outbreak of Salmonella serotype Thompson associated

with fresh cilantro. J Infect Dis 183:984-7.

54. Cantos, E., J. C. Espin, and F. A. Tomas-Barberan. 2001. Effect of wounding on phenolic

enzymes in six minimally processed lettuce cultivars upon storage. J Agric Food Chem 49:322-

30.

55. Carey, C. M., M. Kostrzynska, and S. Thompson. 2009. Escherichia coli O157:H7 stress and

virulence gene expression on Romaine lettuce using comparative real-time PCR. J Microbiol

Methods 77:235-42.

56. Carrigg, C., O. Rice, S. Kavanagh, G. Collins, and V. O'Flaherty. 2007. DNA extraction

method affects microbial community profiles from soils and sediment. Appl Microbiol

Biotechnol 77:955-64.

57. Castanie-Cornet, M. P., T. A. Penfound, D. Smith, J. F. Elliott, and J. W. Foster. 1999.

Control of acid resistance in Escherichia coli. J Bacteriol 181:3525-35.

58. CDC 2006, posting date. Multistate outbreak of E. coli O157 infections. November-December

2006. www.cdc.gov/ecoli/2006/december/121006.htm. [Online.]

59. CDC 2006, posting date. Ongoing mulstistate outbreak of Escherichia coli serotype O157:H7

infections associated with consumption of fresh spinach. United States. www.cdc.gov. [Online.]

60. Chapman, P. A., C. A. Siddons, D. J. Wright, P. Norman, J. Fox, and E. Crick. 1993. Cattle

as a possible source of verocytotoxin-producing Escherichia coli O157 infections in man.

Epidemiol Infect 111:439-47.

61. Chatterjee, A., Y. Cui, H. Hasegawa, and A. K. Chatterjee. 2007. PsrA, the Pseudomonas

sigma regulator, controls regulators of epiphytic fitness, quorum-sensing signals, and plant

interactions in Pseudomonas syringae pv. tomato strain DC3000. Appl Environ Microbiol

73:3684-94.

62. Chua, D., K. Goh, R. Saftner, and A. Bhagwat. 2008. Fresh-cut lettuce in modified atmosphere

packages stored at improper temperatures supports enterohemorrhagic E. coli isolates to survive

gastric acid challenge. J Food Sci 78:148-153.

63. Chung, C. H., W. Bang, and M. A. Drake. 2006. Stress response of Escherichia coli Comp Rev

Food Sci Food Saf 5:52-83.

64. Cimmino, A., G. Marchi, G. Surico, A. Hanuszkiewicz, A. Evidente, and O. Holst. 2008. The

structure of the O-specific polysaccharide of the lipopolysaccharide from Pantoea agglomerans

strain FL1. Carbohydr Res 343:392-6.

65. Conte, A., G. Conversa, C. Scrocco, I. Brescia, J. Laverse, A. Elia, and M. DelNobile. 2008.

Influence of growing periods on the quality of baby spinach leaves at harvest and during storage

as minimally processed produce. Postharvest Biol and Technol 50:190-196.

66. Cooley, M. B., D. Chao, and R. E. Mandrell. 2006. Escherichia coli O157:H7 survival and

growth on lettuce is altered by the presence of epiphytic bacteria. J Food Prot 69:2329-35.

67. Correll, J. C., T. E. Morelock, M. C. Black, S. T. Koike, L. P. Brandenberger, and F. J.

Dainello. 1994. Economically important diseases of spinach. Plant Disease 78:653-660.

68. Costerton, J. W., Z. Lewandowski, D. DeBeer, D. Caldwell, D. Korber, and G. James. 1994.

Biofilms, the customized microniche. J Bacteriol 176:2137-42.

69. Crump, J. A., P. M. Griffin, and F. J. Angulo. 2002. Bacterial contamination of animal feed

and its relationship to human foodborne illness. Clin Infect Dis 35:859-65.

70. Delaquis, P., S. Bach, and L. D. Dinu. 2007. Behavior of Escherichia coli O157:H7 in leafy

vegetables. J Food Prot 70:1966-74.

71. Delaquis, P., S. Stewart, S. Cazaux, and P. Toivonen. 2002. Survival and growth of Listeria

monocytogenes and Escherichia coli O157:H7 in ready-to-eat iceberg lettuce washed in warm

chlorinated water. J Food Prot 65:459-64.

72. Dingman, D. W. 2000. Growth of Escherichia coli O157:H7 in bruised apple (Malus domestica)

tissue as influenced by cultivar, date of harvest, and source. Appl Environ Microbiol 66:1077-83.

73. Dowd, S. E., T. R. Callaway, R. D. Wolcott, Y. Sun, T. McKeehan, R. G. Hagevoort, and T.

S. Edrington. 2008. Evaluation of the bacterial diversity in the feces of cattle using 16S rDNA

bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP). BMC Microbiol 8:125.

74. Ercolani, G. L. 1983. Variability among isolates of Pseudomonas syringae pv. savastanoi from

the phylloplane of the olive. J Gen Microbiol 129:901-16.

75. ERS/USDA. 2008. Data foodborne illness cost.[online]

76. Ethelberg, S., K. E. Olsen, F. Scheutz, C. Jensen, P. Schiellerup, J. Enberg, A. M. Petersen,

B. Olesen, P. Gerner-Smidt, and K. Molbak. 2004. Virulence factors for hemolytic uremic

syndrome, Denmark. Emerg Infect Dis 10:842-7.

77. Fahn, A. 1997. Epidermis, p. 152-184. In A. Fahn (ed.), Plant Anatomy. Aditya Books, New

Delhi.

78. FDA 2007, posting date. Food Protection plan: An integrated strategy for protecting the nation's

food supply. www.fda.gov. [Online.]

79. FDA, USDA, and CFSAN. 1998. Guide to minimize microbial food safety hazards for fresh

fruits and vegetables.

80. Fierer, N., J. A. Jackson, R. Vilgalys, and R. B. Jackson. 2005. Assessment of soil microbial

community structure by use of taxon-specific quantitative PCR assays. Appl Environ Microbiol

71:4117-20.

81. Fino, V. R., and K. E. Kniel. 2008. UV light inactivation of hepatitis A virus, Aichi virus, and

feline calicivirus on strawberries, green onions, and lettuce. J Food Prot 71:908-13.

82. Foley, D., M. Euper, F. Caporaso, and A. Prakash. 2004. Irradiation and chlorination

effectively reduces Escherichia coli O157:H7 inoculated on cilantro (Coriandrum sativum)

without negatively affecting quality. J Food Prot 67:2092-8.

83. Fonseca, J. M. 2006. Postharvest handling and processing: sources of microorgnisms and impact

of sanitizing procedures. In K. R. Matthews (ed.), Microbiology of fresh produce. ASM Press,

Washington D.C.

84. Foster, J. W. 2004. Escherichia coli acid resistance: tales of an amateur acidophile. Nat Rev

Microbiol 2:898-907.

85. Francis, G. A., and D. O'Beirne. 2001. Effects of vegetable type, package atmosphere and

storage temperature on growth and survival of Escherichia coli O157:H7 and Listeria

monocytogenes. J Ind Microbiol Biotechnol 27:111-6.

86. Frankel, G., D. C. Candy, P. Everest, and G. Dougan. 1994. Characterization of the C-terminal

domains of intimin-like proteins of enteropathogenic and enterohemorrhagic Escherichia coli,

Citrobacter freundii, and Hafnia alvei. Infect Immun 62:1835-42.

87. Frankel, G., A. D. Phillips, I. Rosenshine, G. Dougan, J. B. Kaper, and S. Knutton. 1998.

Enteropathogenic and enterohaemorrhagic Escherichia coli: more subversive elements. Mol

Microbiol 30:911-21.

88. Frankel, G., A. D. Phillips, L. R. Trabulsi, S. Knutton, G. Dougan, and S. Matthews. 2001.

Intimin and the host cell--is it bound to end in Tir(s)? Trends Microbiol 9:214-8.

89. Franz, E., and A. H. van Bruggen. 2008. Ecology of E. coli O157:H7 and Salmonella enterica

in the primary vegetable production chain. Crit Rev Microbiol 34:143-61.

90. Fredrickson, A. G. 1977. Behavior of mixed cultures of microorganisms. Annu Rev Microbiol

31:63-87.

91. Gagliardi, J. V., and J. S. Karns. 2000. Leaching of Escherichia coli O157:H7 in diverse soils

under various agricultural management practices. Appl Environ Microbiol 66:877-83.

92. Gagliardi, J. V., and J. S. Karns. 2002. Persistence of Escherichia coli O157:H7 in soil and on

plant roots. Environ Microbiol 4:89-96.

93. Garmendia, J., G. Frankel, and V. F. Crepin. 2005. Enteropathogenic and enterohemorrhagic

Escherichia coli infections: translocation, translocation, translocation. Infect Immun 73:2573-85.

94. Glickmann, E., L. Gardan, S. Jacquet, S. Hussain, M. Elasri, A. Petit, and Y. Dessaux. 1998.

Auxin production is a common feature of most pathovars of Pseudomonas syringae. Mol Plant

Microbe Interact 11:156-62.

95. Gnanamanickam, S., and J. Immanuel. 2006. Epiphytic baceria, their ecology and functions, p.

131-154. In S. Gnanamanickam (ed.), Plant-Associated bacteria. Springer, Dordrecht, The

Netherlands.

96. Goldbert, R. 1980. Cell wall polysaccharidase activities and growth processes: a possible

relationship. Physiol Plant 50:261-264.

97. Gourabathini, P., M. T. Brandl, K. S. Redding, J. H. Gunderson, and S. G. Berk. 2008.

Interactions between food-borne pathogens and protozoa isolated from lettuce and spinach. Appl

Environ Microbiol 74:2518-25.

98. Goverd, K. A., F. W. Beech, R. P. Hobbs, and R. Shannon. 1979. The occurrence and survival

of coliforms and salmonellas in apple juice and cider. J Appl Bacteriol 46:521-30.

99. Grant, J., A. M. Wendelboe, A. Wendel, B. Jepson, P. Torres, C. Smelser, and R. T. Rolfs.

2008. Spinach-associated Escherichia coli O157:H7 outbreak, Utah and New Mexico, 2006.

Emerg Infect Dis 14:1633-6.

100. Grass, M. L., L. Vidal, N. Betoret, A. Chiralt, and P. Fito. 2003. Calcium fortification of

vegetables by vacuum impregnation interactions with cellular matrix. J Food Engineer 56:279-

284.

101. Hadjok, C., G. S. Mittal, and K. Warriner. 2008. Inactivation of human pathogens and

spoilage bacteria on the surface and internalized within fresh produce by using a combination of

ultraviolet light and hydrogen peroxide. J Appl Microbiol 104:1014-24.

102. Hancock, D. D., T. E. Besser, M. L. Kinsel, P. I. Tarr, D. H. Rice, and M. G. Paros. 1994.

The prevalence of Escherichia coli O157.H7 in dairy and beef cattle in Washington State.

Epidemiol Infect 113:199-207.

103. Handschur, M., G. Pinar, B. Gallist, W. Lubitz, and A. G. Haslberger. 2005. Culture free

DGGE and cloning based monitoring of changes in bacterial communities of salad due to

processing. Food Chem Toxicol 43:1595-605.

104. Heaton, J. C., and K. Jones. 2008. Microbial contamination of fruit and vegetables and the

behaviour of enteropathogens in the phyllosphere: a review. J Appl Microbiol 104:613-26.

105. Hengge-Aronis, R. 1996. Back to log phase: sigma S as a global regulator in the osmotic control

of gene expression in Escherichia coli. Mol Microbiol 21:887-93.

106. Herwaldt, B. L., and M. L. Ackers. 1997. An outbreak in 1996 of cyclosporiasis associated with

imported raspberries. The Cyclospora Working Group. N Engl J Med 336:1548-56.

107. Heuvelink, A. E., F. L. van den Biggelaar, E. de Boer, R. G. Herbes, W. J. Melchers, J. H.

Huis in 't Veld, and L. A. Monnens. 1998. Isolation and characterization of verocytotoxin-

producing Escherichia coli O157 strains from Dutch cattle and sheep. J Clin Microbiol 36:878-

82.

108. Heuvelink, A. E., F. L. van den Biggelaar, J. Zwartkruis-Nahuis, R. G. Herbes, R. Huyben,

N. Nagelkerke, W. J. Melchers, L. A. Monnens, and E. de Boer. 1998. Occurrence of

verocytotoxin-producing Escherichia coli O157 on Dutch dairy farms. J Clin Microbiol 36:3480-

109. Hirano, S. S., and C. D. Upper. 2000. Bacteria in the leaf ecosystem with emphasis on

Pseudomonas syringae a pathogen, ice nucleus, and epiphyte. Microbiol Mol Biol Rev 64:624-

53.

110. Hirano, S. S., and C. D. Upper. 1991. Bacterial community dynamics, p. 271-313. In J. H.

Andrews and S. S. Hirano (ed.), Microbial ecology of leaves. Springer-Verlag, New York.

111. Hirano, S. S., and C. D. Upper. 1989. Diel Variation in Population Size and Ice Nucleation

Activity of Pseudomonas syringae on Snap Bean Leaflets. Appl Environ Microbiol 55:623-630.

112. Huber, J. A., D. B. Mark Welch, H. G. Morrison, S. M. Huse, P. R. Neal, D. A. Butterfield,

and M. L. Sogin. 2007. Microbial population structures in the deep marine biosphere. Science

318:97-100.

113. Humblot, C., and J. P. Guyot. 2009. Pyrosequencing of tagged 16S rRNA gene amplicons for

rapid deciphering of the microbiomes of fermented foods such as pearl millet slurries. Appl

Environ Microbiol 75:4354-61.

114. Humphrey, T. J., E. Slater, K. McAlpine, R. J. Rowbury, and R. J. Gilbert. 1995. Salmonella

enteritidis phage type 4 isolates more tolerant of heat, acid, or hydrogen peroxide also survive

longer on surfaces. Appl Environ Microbiol 61:3161-4.

115. Hutchison, M. L., and K. Johnstone. 1993. Evidence for the involvement of the surface active

properties of the extracellular toxin tolaasin in the manifestation of brown blotch disease

symptoms by Pseudomonas tolaasi on Agaricus bisporus. Physiol Mol Plant Pathol 42:373-384.

116. Ibekwe, A. M., P. M. Watt, P. J. Shouse, and C. M. Grieve. 2004. Fate of Escherichia coli

O157:H7 in irrigation water on soils and plants as validated by culture method and real-time

PCR. Can J Microbiol 50:1007-14.

117. ICMSF. 2005. Vegetables and vegetable products, p. 277-325. In ICMSF (ed.), Microbial

ecology of food commodities, 2nd. ed. Kluwer Academic/Plenum Publishers, New York.

118. Idris, R., R. Trifonova, M. Puschenreiter, W. W. Wenzel, and A. Sessitsch. 2004. Bacterial

communities associated with flowering plants of the Ni hyperaccumulator Thlaspi goesingense.

Appl Environ Microbiol 70:2667-77.

119. IFPA, PMA, UFFVA, and WG. Commodity specific food safety guidelines for the lettuce and

leafy greens supply chain. [online]

120. Ipharraguerre, I. R., and J. H. Clark. 2003. Usefulness of ionophores for lactating dairy cows:

A review. Animal Feed Sci and Tech 106:39-57.

121. Ishii, K., M. Fukui, and S. Takii. 2000. Microbial succession during a composting process as

evaluated by denaturing gradient gel electrophoresis analysis. J Appl Microbiol 89:768-77.

122. Islam, M., M. P. Doyle, S. C. Phatak, P. Millner, and X. Jiang. 2004. Persistence of

enterohemorrhagic Escherichia coli O157:H7 in soil and on leaf lettuce and parsley grown in

fields treated with contaminated manure composts or irrigation water. J Food Prot 67:1365-70.

123. Ivanoff, S. S. 1963. Guttation injuries of plants. Bot Rev 29:202-229.

124. Jablasone, J., K. Warriner, and M. Griffiths. 2005. Interactions of Escherichia coli O157:H7,

Salmonella typhimurium and Listeria monocytogenes plants cultivated in a gnotobiotic system.

Int J Food Microbiol 99:7-18.

125. Jacobs, J. L., T. L. Carroll, and G. W. Sundin. 2005. The role of pigmentation, ultraviolet

radiation tolerance, and leaf colonization strategies in the epiphytic survival of phyllosphere

bacteria. Microb Ecol 49:104-13.

126. Jacobs, J. L., and G. W. Sundin. 2001. Effect of solar UV-B radiation on a phyllosphere

bacterial community. Appl Environ Microbiol 67:5488-96.

127. Jacques, M., L. L. Kinkel, and C. E. Morris. 1995. Population sizes, immigration, and growth

of epiphytic bacteria on leaves of different ages and positions of field-grown endive (Cichorium

endivia var. latifolia). Appl Environ Microbiol 61:899-906.

128. Jain, S., and J. Chen. 2007. Attachment and biofilm formation by various serotypes of

Salmonella as influenced by cellulose production and thin aggregative fimbriae biosynthesis. J

Food Prot 70:2473-9.

129. Jay, M. T., M. Cooley, D. Carychao, G. W. Wiscomb, R. A. Sweitzer, L. Crawford-Miksza,

J. A. Farrar, D. K. Lau, J. O'Connell, A. Millington, R. V. Asmundson, E. R. Atwill, and R.

E. Mandrell. 2007. Escherichia coli O157:H7 in feral swine near spinach fields and cattle,

central California coast. Emerg Infect Dis 13:1908-11.

130. Jiang, X., J. Morgan, and M. P. Doyle. 2003. Fate of Escherichia coli O157:H7 during

composting of bovine manure in a laboratory-scale bioreactor. J Food Prot 66:25-30.

131. Johnston, M. A., M. A. Harrison, and R. A. Morrow. 2009. Microbial antagonists of

Escherichia coli O157:H7 on fresh-cut lettuce and spinach. J Food Prot 72:1569-1575.

132. Juven, B. J., and M. D. Pierson. 1996. Antibacterial effects of hydrgen peroxide and methods

for its detection and quantitation. J Food Prot 59:1233-1241.

133. Kader, A. A. 2002. Quality parameters of fresh-cut fruit and vegetable products. In O. Lamikanra

(ed.), Fresh-cut fruits and vegetables. Science, echnology adn arket. CRC Press, Boca Raton, FL.

134. Kadivar, H., and A. E. Stapleton. 2003. Ultraviolet radiation alters maize phyllosphere bacterial

diversity. Microb Ecol 45:353-61.

135. Kashmanian, R. M., and R. F. Rynk. 1996. Agricultural composting in the United States:

trends and driving forces. J Soil Water Conserve 51:194-201.

136. Keraita, B., F. Konradsen, P. Drechsel, and R. Abaidoo. 2007. Effect of low-cost irrigation

methods on microbial contamination of lettuce irrigated with untreated wastewater. Trop Med Int

Health 12:15-22.

137. Kim, J., F. Luo, and X. Jiang. 2009. Factors impacting the regrowth of Escherichia coli

O157:H7 in dairy manure compost. J Food Prot 72:1576-84.

138. Kinkel, L. L., M. Wilson, and S. E. Lindow. 2000. Plant Species and Plant Incubation

Conditions Influence Variability in Epiphytic Bacterial Population Size. Microb Ecol 39:1-11.

139. Kirk, J. L., L. A. Beaudette, M. Hart, P. Moutoglis, J. N. Klironomos, H. Lee, and J. T.

Trevors. 2004. Methods of studying soil microbial diversity. J Microbiol Methods 58:169-88.

140. Knutton, S., D. R. Lloyd, and A. S. McNeish. 1987. Adhesion of enteropathogenic Escherichia

coli to human intestinal enterocytes and cultured human intestinal mucosa. Infect Immun 55:69-

77.

141. Krimm, U., D. Abanda-Nkpwatt, W. Schwab, and L. Schreiber. 2005. Epiphytic

microorganisms on strawberry plants (Fragaria ananassa cv. Elsanta): identification of bacterial

isolates and analysis of their interaction with leaf surfaces. FEMS Microbiol Ecol 53:483-92.

142. Kudva, I. T., K. Blanch, and C. J. Hovde. 1998. Analysis of Escherichia coli O157:H7 survival

in ovine or bovine manure and manure slurry. Appl Environ Microbiol 64:3166-74.

143. Lambais, M. R., D. E. Crowley, J. C. Cury, R. C. Bull, and R. R. Rodrigues. 2006. Bacterial

diversity in tree canopies of the Atlantic forest. Science 312:1917.

144. Lamikanra, O. 2002. Preface. In O. Lamikanra (ed.), Fresh-cut fruits and vegetables. Science,

technology and market. CRC Press, Boca Raton, FL.

145. Lapidot, A., and S. Yaron. 2009. Transfer of Salmonella enterica serovar Typhimurium from

contaminated irrigation water to parsley is dependent on curli and cellulose, the biofilm matrix

components. J Food Prot 72:618-23.

146. Lawrence, J., and E. L. Dyckman 2002. Fruits and vegetables enhanced federal efforts to

increase consumption could yield health benefits for Americans.

147. Lee, D. H., Y. G. Zo, and S. J. Kim. 1996. Nonradioactive method to study genetic profiles of

natural bacterial communities by PCR-single-strand-conformation polymorphism. Appl Environ

Microbiol 62:3112-20.

148. Lee, S. Y., and S. Y. Baek. 2008. Effect of chemical sanitizer combined with modified

atmosphere packaging on inhibiting Escherichia coli O157:H7 in commercial spinach. Food

Microbiol 25:582-7.

149. Lee, S. Y., M. Costello, and D. H. Kang. 2004. Efficacy of chlorine dioxide gas as a sanitizer of

lettuce leaves. J Food Prot 67:1371-6.

150. LeStrange, M., S. Koike, J. Valencia, and W. Chaney. Spinach production in California.

University of California. Report.

151. Leveau, J. H., and S. E. Lindow. 2001. Appetite of an epiphyte: quantitative monitoring of

bacterial sugar consumption in the phyllosphere. Proc Natl Acad Sci U S A 98:3446-53.

152. LGMA, posting date. California leafy green products. [Online.]

153. Liao, C. H., and W. F. Fett. 2001. Analysis of native microflora and selection of strains

antagonistic to human pathogens on fresh produce. J Food Prot 64:1110-5.

154. Lindow, S. E. 1991. Determinants of epiphytic fitness in bacteria, p. 295-310. In J. H. Andrews

and S. S. Hirano (ed.), Microbial ecology of leaves. Springer Verlag, New York.

155. Lindow, S. E., and M. T. Brandl. 2003. Microbiology of the phyllosphere. Appl Environ

Microbiol 69:1875-83.

156. Lindow, S. E., C. Desurmont, R. Elkins, G. McGourty, E. Clark, and M. T. Brandl. 1998.

Occurrence of indole-3-acetic Acid-producing bacteria on pear trees and their association with

fruit russet. Phytopathology 88:1149-57.

157. Lindow, S. E., and J. H. Leveau. 2002. Phyllosphere microbiology. Curr Opin Biotechnol

13:238-43.

158. Little, A. E., C. J. Robinson, S. B. Peterson, K. F. Raffa, and J. Handelsman. 2008. Rules of

engagement: interspecies interactions that regulate microbial communities. Annu Rev Microbiol

62:375-401.

159. Liu, Z., C. Lozupone, M. Hamady, F. D. Bushman, and R. Knight. 2007. Short

pyrosequencing reads suffice for accurate microbial community analysis. Nucleic Acids Res

35:120.

160. Long, W. G., D. V. Sweet, and H. B. Tukey. 1956. Loss of nutrients from plant foliage by

leaching as indicated by radioisotopes. Science 123:1039-1040.

161. Lopez-Garcia, P., and D. Moreira. 2008. Tracking microbial biodiversity through molecular

and genomic ecology. Res Microbiol 159:67-73.

162. Loui, C., G. Grigoryan, H. Huang, L. W. Riley, and S. Lu. 2008. Bacterial communities

associated with retail alfalfa sprouts. J Food Prot 71:200-4.

163. Louise, C. B., and T. G. Obrig. 1995. Specific interaction of Escherichia coli O157:H7-derived

Shiga-like toxin II with human renal endothelial cells. J Infect Dis 172:1397-401.

164. Manceau, C. R., and M. N. Kasempour. 2001. Endophytic versus epiphytic colonization of

plants: what comes first?, p. 115-123. In S. E. Lindow, I. Hecht-Poinar, and J. Elliot (ed.),

Phyllosphere microbiology. APS Press, St Paul, SA.

165. Manchester, A., and A. Clauson. 1995. 1994 spending for food away from home outpaces food

at home. Food Review 18:12-15.

166. Manulis, S., A. Haviv-Chesner, M. T. Brandl, S. E. Lindow, and I. Barash. 1998. Differential

involvement of indole-3-acetic acid biosynthetic pathways in pathogenicity and epiphytic fitness

of Erwinia herbicola pv. gypsophilae. Mol Plant Microbe Interact 11:634-42.

167. Mark Ibekwe, A., C. M. Grieve, S. K. Papiernik, and C. H. Yang. 2009. Persistence of

Escherichia coli O157:H7 on the rhizosphere and phyllosphere of lettuce. Lett Appl Microbiol

49:784-90.

168. Matthews, K. R. 2006. Microorganisms associated with fruits and vegetables. In K. R. Matthews

(ed.), Microbiology of Fresh Produce. ASM Press, Washington DC.

169. McCaig, A. E., L. A. Glover, and J. I. Prosser. 2001. Numerical analysis of grassland bacterial

community structure under different land management regimens by using 16S ribosomal DNA

sequence data and denaturing gradient gel electrophoresis banding patterns. Appl Environ

Microbiol 67:4554-9.

170. McGee, P., L. Scott, J. J. Sheridan, B. Earley, and N. Leonard. 2004. Horizontal transmission

of Escherichia coli O157:H7 during cattle housing. J Food Prot 67:2651-6.

171. Mercier, J., and S. E. Lindow. 2000. Role of leaf surface sugars in colonization of plants by

bacterial epiphytes. Appl Environ Microbiol 66:369-74.

172. Michino, H., K. Araki, S. Minami, S. Takaya, N. Sakai, M. Miyazaki, A. Ono, and H.

Yanagawa. 1999. Massive outbreak of Escherichia coli O157:H7 infection in schoolchildren in

Sakai City, Japan, associated with consumption of white radish sprouts. Am J Epidemiol

150:787-96.

173. Milillo, S. R., J. M. Badamo, K. J. Boor, and M. Wiedmann. 2008. Growth and persistence of

Listeria monocytogenes isolates on the plant model Arabidopsis thaliana. Food Microbiol

25:698-704.

174. Miller, C. D., Y. C. Kim, and A. J. Anderson. 2001. Competitiveness in root colonization by

Pseudomonas putida requires the rpoS gene. Can J Microbiol 47:41-8.

175. Mitra, R., E. Cuesta-Alonso, A. C. Wayadande, J. Talley, S. Gilliland, and J. Fletcher. 2009.

Effect of route of introduction and host cultivar on the colonization, internalization, and

movement of the human pathogen Escherichia coli O157:H7 in spinach. J Food Prot 72:1521-

1530.

176. Monier, J. M., and S. E. Lindow. 2005. Aggregates of resident bacteria facilitate survival of

immigrant bacteria on leaf surfaces. Microb Ecol 49:343-52.

177. Monier, J. M., and S. E. Lindow. 2003. Differential survival of solitary and aggregated bacterial

cells promotes aggregate formation on leaf surfaces. Proc Natl Acad Sci U S A 100:15977-82.

178. Monier, J. M., and S. E. Lindow. 2004. Frequency, size, and localization of bacterial aggregates

on bean leaf surfaces. Appl Environ Microbiol 70:346-55.

179. Morales, S. E., T. F. Cosart, J. V. Johnson, and W. E. Holben. 2009. Extensive phylogenetic

analysis of a soil bacterial community illustrates extreme taxon evenness and the effects of

amplicon length, degree of coverage, and DNA fractionation on classification and ecological

parameters. Appl Environ Microbiol 75:668-75.

180. Morris, C. E., and L. L. Kinkel. 2002. Fifty years of phyllosphere microbiology: significant

contributions to research in related fields, p. 365-375. In S. E. Lindow and I. Hecht-Poinar (ed.),

Phyllosphere microbiology. APS Press, Saint Paul, MN.

181. Morris, C. E., J. M. Monier, and M. A. Jacques. 1998. A technique to quantify the population

size and composition of the biofilm component in communities of bacteria in the phyllosphere.

Appl Environ Microbiol 64:4789-95.

182. Muller, C., and M. Riederer. 2005. Plant surface properties in chemical ecology. J Chem Ecol

31:2621-51.

183. Muyzer, G., E. C. de Waal, and A. G. Uitterlinden. 1993. Profiling of complex microbial

populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-

amplified genes coding for 16S rRNA. Appl Environ Microbiol 59:695-700.

184. Muyzer, G., and K. Smalla. 1998. Application of denaturing gradient gel electrophoresis

(DGGE) and temperature gradient gel electrophoresis (TGGE) in microbial ecology. Antonie Van

Leeuwenhoek 73:127-41.

185. Nakatsu, C. H., V. Torsvik, and L. Ovreas. 2000. Soil community analysis using DGGE of 16S

rDNA polymerase chain reaction products. Soil Sci Soc Am J 64:1382-1388.

186. Nataro, J. P., and J. B. Kaper. 1998. Diarrheagenic Escherichia coli. Clin Microbiol Rev

11:142-201.

187. Neu, T. R., T. Nartner, and K. Poralla. 1990. Surface active properties of viscosin: a

peptidolipid antibiotic. Appl Microbiol Biotechnol 32:518-520.

188. Nguyen-the, C., and F. Carlin. 1994. The microbiology of minimally processed fresh fruits and

vegetables. Crit Rev Food Sci Nutr 34:371-401.

189. Niemira, B. A. 2007. Relative efficacy of sodium hypochlorite wash versus irradiation to

inactivate Escherichia coli O157:H7 internalized in leaves of Romaine lettuce and baby spinach.

J Food Prot 70:2526-32.

190. Nohmi, T., A. Hakura, Y. Nakai, M. Watanabe, S. Y. Murayama, and T. Sofuni. 1991.

Salmonella typhimurium has two homologous but different umuDC operons: cloning of a new

umuDC-like operon (samAB) present in a 60-megadalton cryptic plasmid of S. typhimurium. J

Bacteriol 173:1051-63.

191. Normander, B., B. B. Christensen, S. Molin, and N. Kroer. 1998. Effect of bacterial

distribution and activity on conjugal gene transfer on the phylloplane of the bush bean (Phaseolus

vulgaris). Appl Environ Microbiol 64:1902-9.

192. Nystrom, T. 2004. Growth versus maintenance: a trade-off dictated by RNA polymerase

availability and sigma factor competition? Mol Microbiol 54:855-62.

193. O'Brien, A. D., V. L. Tesh, A. Donohue-Rolfe, M. P. Jackson, S. Olsnes, K. Sandvig, A. A.

Lindberg, and G. T. Keusch. 1992. Shiga toxin: biochemistry, genetics, mode of action, and

role in pathogenesis. Curr Top Microbiol Immunol 180:65-94.

194. O'Brien R, D., and S. E. Lindow. 1988. Effect of Plant Species and Environmental Conditions

on Ice Nucleation Activity of Pseudomonas syringae on Leaves. Appl Environ Microbiol

54:2281-2286.

195. O'Donnell, A. G., and H. E. Gorres. 1999. 16S rDNA methods in soil microbiology. Curr Opin

Biotechnol 10:225-9.

196. Oliver, J. D., L. Nilsson, and S. Kjelleberg. 1991. Formation of nonculturable Vibrio vulnificus

cells and its relationship to the starvation state. Appl Environ Microbiol 57:2640-4.

197. Pace, N. R. 1997. A molecular view of microbial diversity and the biosphere. Science 276:734-

40.

198. Palou, E., A. Lopez-Malo, G. V. Barbosa-Canovas, and J. Welti-Chanes. 2000. High

hydrostatic pressure and minimal processing, p. 205-222. In S. M. Alzamora, M. S. Tapia, and A.

Lopez-Malo (ed.), Minimally processed fruits and vegetables. Aspen, Maryland.

199. Park, S., R. W. Worobo, and R. A. Durst. 1999. Escherichia coli O157:H7 as an emerging

foodborne pathogen: a literature review. Crit Rev Food Sci Nutr 39:481-502.

200. Passaro, D. J., R. Reporter, L. Mascola, L. Kilman, G. B. Malcolm, H. Rolka, S. B. Werner,

and D. J. Vugia. 1996. Epidemic Salmonella enteritidis infection in Los Angeles County,

California. The predominance of phage type 4. West J Med 165:126-30.

201. Persson, S., K. E. Olsen, S. Ethelberg, and F. Scheutz. 2007. Subtyping method for

Escherichia coli shiga toxin (verocytotoxin) 2 variants and correlations to clinical manifestations.

J Clin Microbiol 45:2020-4.

202. Phillips, M. A., and R. B. Croteau. 1999. Resin-based defenses in conifers. Trends Plant Sci

4:184-190.

203. Pichersky, E., J. P. Noel, and N. Dudareva. 2006. Biosynthesis of plant volatiles: nature's

diversity and ingenuity. Science 311:808-11.

204. Pierre, D. 2004. Interactions between microorganisms. In D. Pierre (ed.), Microbial ecology of

the soil and plant growth. Science Publisher Inc., NH.

205. PMA, U. W. G., International Fresh-cut produce association. Good agricultural practices.

[Online.]

206. Pourzand, C., and R. M. Tyrell. 1999. Apoptosis, the role of oxidative stress and the example of

solar UV radiation. Photochem Photobiol 70:380-390.

207. Priepke, P. E., L. S. Wei, and A. I. Nelso. 1976. Refrigerated storage of prepackaged salad

vegetables. J Food Sci 41:379-382.

208. Ragaert, P., F. Devlieghere, and J. Debevere. 2007. Role of microbiological and physiological

spoilage mechanisms during storage of minimally processed vegetables. Postharvest Biol and

Technol 44:185-194.

209. Randazzo, C. L., G. O. Scifo, F. Tomaselli, and C. Caggia. 2009. Polyphasic characterization

of bacterial community in fresh cut salads. Int J Food Microbiol 128:484-90.

210. Rangel, J. M., P. H. Sparling, C. Crowe, P. M. Griffin, and D. L. Swerdlow. 2005.

Epidemiology of Escherichia coli O157:H7 outbreaks, United States, 1982-2002. Emerg Infect

Dis 11:603-9.

211. Rasche, F., E. Marco-Noales, H. Velvis, L. S. van Overbeek, M. M. Lopez, J. D. van Elsas,

and A. Sessitsch. 2006. Structural characteristics and plant-beneficial effects of bacterial

colonizing the shoots of field grown conventional and genetically modified T4-lysozyme

producing potatoes. Plant Soil 289:123-140.

212. Ravva, S. V., C. Z. Sarreal, B. Duffy, and L. H. Stanker. 2006. Survival of Escherichia coli

O157:H7 in wastewater from dairy lagoons. J Appl Microbiol 101:891-902.

213. Ribaudo, M., J. Kaplan, L. Christensen, N. Gollehon, N. Johansoon, V. Breneman, M.

Aillery, J. Agapoff, and M. Petters. 2003. Manure management for water quality. Costs to

animal feeding operations of applying manure nutrients to land.

214. Rico, D., A. B. Martin-Diana, J. M. Barat, and C. Barry-Ryan. 2007. Extending and

measuring the quality of fresh-fruit and vegetables: a review. Trends in Food Sci and Tech

18:337-386.

215. Riley, L. W., R. S. Remis, S. D. Helgerson, H. B. McGee, J. G. Wells, B. R. Davis, R. J.

Hebert, E. S. Olcott, L. M. Johnson, N. T. Hargrett, P. A. Blake, and M. L. Cohen. 1983.

Hemorrhagic colitis associated with a rare Escherichia coli serotype. N Engl J Med 308:681-5.

216. Roesch, L. F., R. R. Fulthorpe, A. Riva, G. Casella, A. K. Hadwin, A. D. Kent, S. H. Daroub,

F. A. Camargo, W. G. Farmerie, and E. W. Triplett. 2007. Pyrosequencing enumerates and

contrasts soil microbial diversity. Isme J 1:283-90.

217. Ronaghi, M., and E. Elahi. 2002. Pyrosequencing for microbial typing. J Chromatogr B Analyt

Technol Biomed Life Sci 782:67-72.

218. Rondon, M. R., P. R. August, A. D. Bettermann, S. F. Brady, T. H. Grossman, M. R. Liles,

K. A. Loiacono, B. A. Lynch, I. A. MacNeil, C. Minor, C. L. Tiong, M. Gilman, M. S.

Osburne, J. Clardy, J. Handelsman, and R. M. Goodman. 2000. Cloning the soil

metagenome: a strategy for accessing the genetic and functional diversity of uncultured

microorganisms. Appl Environ Microbiol 66:2541-7.

219. Roszak, D. B., D. J. Grimes, and R. R. Colwell. 1984. Viable but nonrecoverable stage of

Salmonella enteritidis in aquatic systems. Can J Microbiol 30:334-8.

220. Rubatzky, V., and M. Yamaguchi. 1997. Spinach, table beets, and other vegetable chenopods,

p. 457-463. In V. Rubatzky and M. Yamaguchi (ed.), World Vegetables, 2nd. ed. International

Thomson Publishing, NY.

221. Rude, R. A., G. J. Jackson, J. W. Bier, T. K. Sawyer, and N. G. Risty. 1984. Survey of fresh

vegetables for nematodes, amoebae, and Salmonella. J Assoc Off Anal Chem 67:613-5.

222. Rudi, K., S. L. Flateland, J. F. Hanssen, G. Bengtsson, and H. Nissen. 2002. Development and

evaluation of a 16S ribosomal DNA array-based approach for describing complex microbial

communities in ready-to-eat vegetable salads packed in a modified atmosphere. Appl Environ

Microbiol 68:1146-56.

223. Russell, J. B., F. Diez-Gonzalez, and G. N. Jarvis. 2000. Potential effect of cattle diets on the

transmission of pathogenic Escherichia coli to humans. Microbes Infect 2:45-53.

224. Saftner, R. A., J. Baj, J. A. Abbott, and Y. S. Lee. 2003. Sanitary dips with calcium propionate,

calcium chloride, or calcium amino acid chelates maintain quality and shelf stability of fresh-cut

honeydew chunks. Postharvest Biol and Technol 29:257-269.

225. Saltveit, M. E. 1993. A summary of CA and MA requirements and recommendations for the

storage of harvested vegetables. Cornell University.

226. Samadpour, M., J. E. Ongerth, J. Liston, N. Tran, D. Nguyen, T. S. Whittam, R. A. Wilson,

and P. I. Tarr. 1994. Occurrence of Shiga-like toxin-producing Escherichia coli in retail fresh

seafood, beef, lamb, pork, and poultry from grocery stores in Seattle, Washington. Appl Environ

Microbiol 60:1038-40.

227. Sandhya. 2009. Modified atmosphere packaging of fresh produce: Current status and future

needs. LWT-Food Sci and Technol 1-12.

228. Scherwinski, K., R. Grosch, and G. Berg. 2008. Effect of bacterial antagonists on lettuce:

active biocontrol of Rhizoctonia solani and negligible, short-term effects on nontarget

microorganisms. FEMS Microbiol Ecol 64:106-16.

229. Schoeni, J. L., and M. P. Doyle. 1994. Variable colonization of chickens perorally inoculated

with Escherichia coli O157:H7 and subsequent contamination of eggs. Appl Environ Microbiol

60:2958-62.

230. Schreiber, L., U. Krimm, D. Knoll, M. Sayed, G. Auling, and R. M. Kroppenstedt. 2005.

Plant-microbe interactions: identification of epiphytic bacteria and their ability to alter leaf

surface permeability. New Phytol 166:589-94.

231. Schuenzel, K. M., and M. A. Harrison. 2002. Microbial antagonists of foodborne pathogens on

fresh, minimally processed vegetables. J Food Prot 65:1909-15.

232. Semenov, A. V., A. H. van Bruggen, L. van Overbeek, A. J. Termorshuizen, and A. M.

Semenov. 2007. Influence of temperature fluctuations on Escherichia coli O157:H7 and

Salmonella enterica serovar Typhimurium in cow manure. FEMS Microbiol Ecol 60:419-28.

233. Shaw, R. K., C. N. Berger, B. Feys, S. Knutton, M. J. Pallen, and G. Frankel. 2008.

Enterohemorrhagic Escherichia coli exploits EspA filaments for attachment to salad leaves. Appl

Environ Microbiol 74:2908-14.

234. Shepherd, R. W., and G. J. Wagner. 2007. Phylloplane proteins: emerging defenses at the

aerial frontline? Trends Plant Sci 12:51-6.

235. Sieuwerts, S., F. A. de Bok, J. Hugenholtz, and J. E. van Hylckama Vlieg. 2008. Unraveling

microbial interactions in food fermentations: from classical to genomics approaches. Appl

Environ Microbiol 74:4997-5007.

236. Singh, N., R. K. Singh, A. K. Bhunia, and R. L. Stroshine. 2002. Efficacy of chlorine dioxide,

ozone, and thyme essential oil or a sequential washing in killing Eschericha coli O157:H7 on

lettuce and baby carrots. LWT Food Sci and Technolo 35:720-729.

237. Sivapalasingam, S., C. R. Friedman, L. Cohen, and R. V. Tauxe. 2004. Fresh produce: a

growing cause of outbreaks of foodborne illness in the United States, 1973 through 1997. J Food

Prot 67:2342-53.

238. Sivertsvik, M., J. T. Rosnes, and H. Bergslien. 2002. Modified atmosphere packaging In T.

Ohlsson and N. Bengtsson (ed.), Minimal processing technologies in the food industry.

Woodhead, Cambridge, UK.

239. Smalla, K., M. Oros-Sichler, A. Milling, H. Heuer, S. Baumgarte, R. Becker, G. Neuber, S.

Kropf, A. Ulrich, and C. C. Tebbe. 2007. Bacterial diversity of soils assessed by DGGE, T-

RFLP and SSCP fingerprints of PCR-amplified 16S rRNA gene fragments: do the different

methods provide similar results? J Microbiol Methods 69:470-9.

240. Sogin, M. L., H. G. Morrison, J. A. Huber, D. Mark Welch, S. M. Huse, P. R. Neal, J. M.

Arrieta, and G. J. Herndl. 2006. Microbial diversity in the deep sea and the underexplored "rare

biosphere". Proc Natl Acad Sci U S A 103:12115-20.

241. Solomon, E. B., H. J. Pang, and K. R. Matthews. 2003. Persistence of Escherichia coli

O157:H7 on lettuce plants following spray irrigation with contaminated water. J Food Prot

66:2198-202.

242. Solomon, E. B., C. J. Potenski, and K. R. Matthews. 2002. Effect of irrigation method on

transmission to and persistence of Escherichia coli O157:H7 on lettuce. J Food Prot 65:673-6.

243. Sperandio, V., A. G. Torres, J. A. Giron, and J. B. Kaper. 2001. Quorum sensing is a global

regulatory mechanism in enterohemorrhagic Escherichia coli O157:H7. J Bacteriol 183:5187-97.

244. Stahl, D. A., D. J. Lane, G. J. Olsen, and N. R. Pace. 1984. Analysis of Hydrothermal Vent-

Associated Symbionts by Ribosomal RNA Sequences. Science 224:409-411.

245. Stahl, D. A., D. J. Lane, G. J. Olsen, and N. R. Pace. 1985. Characterization of a Yellowstone

hot spring microbial community by 5S rRNA sequences. Appl Environ Microbiol 49:1379-84.

246. Staley, J. T., and A. Konopka. 1985. Measurement of in situ activities of nonphotosynthetic

microorganisms in aquatic and terrestrial habitats. Annu Rev Microbiol 39:321-46.

247. Steele, B. T., N. Murphy, G. S. Arbus, and C. P. Rance. 1982. An outbreak of hemolytic

uremic syndrome associated with ingestion of fresh apple juice. J Pediatr 101:963-5.

248. Stockwell, V. O., and J. E. Loper. 2005. The sigma factor RpoS is required for stress tolerance

and environmental fitness of Pseudomonas fluorescens Pf-5. Microbiology 151:3001-9.

249. Sundin, G. W., and J. Murillo. 1999. Functional analysis of the Pseudomonas syringae rulAB

determinant in tolerance to ultraviolet B (290-320 nm) radiation and distribution of rulAB among

P. syringae pathovars. Environ Microbiol 1:75-87.

250. Surico, G. 1993. Scanning electron microscopy of olive and oleander leaves colonized by

Pseudomonas syringae subsp. savastanoi. J Phytopathol 138:31-40.

251. Suslow, T. Postharvest handling for organic crops. University of California.

252. Tauxe, R. V. 1997. Emerging foodborne diseases: an evolving public health challenge. Emerg

Infect Dis 3:425-34.

253. Thompson, H. C., and W. C. Kelly. 1957. Potherbs or greens, p. 214-223. In H. C. Thompson

and W. C. Kelly (ed.), Vegetable crops, 5th ed. McGraw-Hill, New York.

254. Thompson, I. P., M. J. Bailey, J. S. Fenlon, T. R. Fermor, A. K. Lillley, J. M. Lynch, P. J.

McCormack, M. P. McQuilken, and K. J. Purdy. 1993. Quantitative and qualitative seasonal

changes in the microbial community from the phyllosphere of sugar beet. Plant Soil 150:177-191.

255. Tiedje, J. M., S. Asuming-Brempong, K. Nusslein, T. L. Marsh, and S. J. Flynn. 1999.

Opening the black box of soil microbial diversity. Appl Soil Ecol 13:109-122.

256. Torsvik, V., J. Goksoyr, and F. L. Daae. 1990. High diversity in DNA of soil bacteria. Appl

Environ Microbiol 56:782-7.

257. Trias, R., L. Baneras, E. Badosa, and E. Montesinos. 2008. Bioprotection of Golden Delicious

apples and Iceberg lettuce against foodborne bacterial pathogens by lactic acid bacteria. Int J

Food Microbiol 123:50-60.

258. Tukey, H. B. 1970. The leaching of substances from plants, Department of Floriculture and

Ornamental Horticulture. Cornell University, Ithaca.

259. Tyler, H. L., and E. W. Triplett. 2008. Plants as a habitat for beneficial and/or human

pathogenic bacteria. Annu Rev Phytopathol 46:53-73.

260. Uesugi, A. R., M. D. Danyluk, and L. J. Harris. 2006. Survival of Salmonella enteritidis phage

type 30 on inoculated almonds stored at -20, 4, 23, and 35 degrees C. J Food Prot 69:1851-7.

261. Uesugi, A. R., and L. J. Harris. 2006. Growth of Salmonella enteritidis phage type 30 in

almond hull and shell slurries and survival in drying almond hulls. J Food Prot 69:712-8.

262. Urich, T., A. Lanzen, J. Qi, D. H. Huson, C. Schleper, and S. C. Schuster. 2008.

Simultaneous assessment of soil microbial community structure and function through analysis of

the meta-transcriptome. PLoS One 3:2527.

263. USDA. Food guide in comparison to national health and nutrition examination. Survey 2001-

2002 consumption data.

264. USDA. 2002. Profiling food consumption in America. In USDA (ed.), The fact book 2001-2002.

265. USDA. 1997. An update: Escherichia coli O157:H7 in humans and cattle.

266. USDA and HHS. 2005. Dietary guidelines for Americans.

267. USDA/ERS. 2004. Factors affecting spinach consumption in the United States.

268. USDA/ERS. 2007. Fresh market spinach: Background information and statistics.

269. USDA/ERS. 2001. Lettuce: In & out of the bag.

270. Vanselow, B. A., D. O. Krause, and C. S. McSweeney. 2005. The shiga toxin-producing

Escherichia coli, their ruminant host, and potential on-farm interventions: A review. Australian J

Agric Res 56:219-244.

271. Venter, T. V. 2008, posting date. Emerging food-borne diseases: A global responsibility. FAO.

[Online.]

272. Voetsch, A. C., T. J. Van Gilder, F. J. Angulo, M. M. Farley, S. Shallow, R. Marcus, P. R.

Cieslak, V. C. Deneen, and R. V. Tauxe. 2004. FoodNet estimate of the burden of illness

caused by nontyphoidal Salmonella infections in the United States. Clin Infect Dis 38 Suppl

3:S127-34.

273. Vurma, M., R. B. Pandit, S. K. Sastry, and A. E. Yousef. 2009. Inactivation of Escherichia

coli O157:H7 and natural microbiota on spinach leaves using gaseous ozone during vacuum

cooling and simulated transportation. J Food Prot 72:1538-1546.

274. Wagner, G. J. 2004. New approaches for studying and exploiting and old protuberance, the plant

trichome. Ann Bot 93:3-11.

275. Walters, M., and V. Sperandio. 2006. Quorum sensing in Escherichia coli and Salmonella. Int J

Med Microbiol 296:125-31.

276. Warner, J. C., S. D. Rothwell, and C. W. Keevil. 2008. Use of episcopic differential

interference contrast microscopy to identify bacterial biofilms on salad leaves and track

colonization by Salmonella Thompson. Environ Microbiol 10:918-25.

277. Weise, E., and J. Schmit. 2007. Spinach recall: 5 faces, 5 agonizing deaths, 1 year later. USA

Today.

278. Weller, R., and D. M. Ward. 1989. Selective Recovery of 16S rRNA Sequences from Natural

Microbial Communities in the Form of cDNA. Appl Environ Microbiol 55:1818-1822.

279. Werker, E. 2000. Trichome diversity and development. Adv Bot Res 31:1-35.

280. Whipps, J. M., P. Hand, D. Pink, and G. D. Bending. 2008. Phyllosphere microbiology with

special reference to diversity and plant genotype. J Appl Microbiol 105:1744-55.

281. Williams, A. P., P. Roberts, L. M. Avery, K. Killham, and D. L. Jones. 2006. Earthworms as

vectors of Escherichia coli O157:H7 in soil and vermicomposts. FEMS Microbiol Ecol 58:54-64.

282. Willis, D. K., E. H. Hrabak, S. E. Lindow, and N. J. Panopoulos. 1988. Construction ad

characterization of Pseudomonas syringae recA mutant strains. Molec Plant-Microbe interactions

1:80-86.

283. Wilson, M., and S. E. Lindow. 1994. Coexistence among epiphytic bacterial populations

mediated through nutritional resource partitioning. Appl Environ Microbiol 60:4468-4477.

284. Woese, C. R. 1987. Bacterial evolution. Microbiol Rev 51:221-71.

285. Woodgate, R., and S. G. Sedgwick. 1992. Mutagenesis induced by bacterial UmuDC proteins

and their plasmid homologues. Mol Microbiol 6:2213-8.

286. Wu, F. M., M. P. Doyle, L. R. Beuchat, J. G. Wells, E. D. Mintz, and B. Swaminathan. 2000.

Fate of Shigella sonnei on parsley and methods of disinfection. J Food Prot 63:568-72.

287. Yadav, R. K., K. Karamanoli, and D. Vokou. 2005. Bacterial colonization of the phyllosphere

of mediterranean perennial species as influenced by leaf structural and chemical features. Microb

Ecol 50:185-96.

288. Yang, C. H., D. E. Crowley, J. Borneman, and N. T. Keen. 2001. Microbial phyllosphere

populations are more complex than previously realized. Proc Natl Acad Sci U S A 98:3889-94.

289. Yaphremov, A., and L. Schreiber. 2005. The dark side of the cell wall:molecular genetics of

plant cuticle. Plant Biosys 139:74-79.

290. Young Lee, N., C. Jo, D. Hwa Shin, W. Geun Kim, and M. Woo Byun. 2006. Effect of

gamma-irradiation on pathogens inoculated into ready-to-use vegetables. Food Microbiol 23:649-

56.

291. Youssef, N. H., and M. S. Elshahed. 2008. Species richness in soil bacterial communities: a

proposed approach to overcome sample size bias. J Microbiol Methods 75:86-91.

292. Zachow, C., R. Tilcher, and G. Berg. 2008. Sugar beet-associated bacterial and fungal

communities show a high indigenous antagonistic potential against plant pathogens. Microb Ecol

55:119-29.

293. Zagory, D. 1999. Effects of post-processing handling and packaging on microbial populations.

Postharvest Biol. and Technol. 15:313-321.

294. Zhang, G., L. Ma, L. R. Beuchat, M. C. Erickson, V. H. Phelan, and M. P. Doyle. 2009. Lack

of internalization of Escherichia coli O157:H7 in lettuce (Lactuca sativa L.) after leaf surface and

soil inoculation. J Food Prot 72:2028-37.

295. Zhong, Q., D. G. Black, P. M. Davidson, and D. A. Golden. 2008. Nonthermal inactivation of

Escherichia coli K-12 on spinach leaves, using dense phase carbon dioxide. J Food Prot 71:1015-

296. Zwielehner, J., M. Handschur, A. Michaelsen, S. Irez, M. Demel, E. B. Denner, and A. G.

Haslberger. 2008. DGGE and real-time PCR analysis of lactic acid bacteria in bacterial

communities of the phyllosphere of lettuce. Mol Nutr Food Res 52:614-23.

CHAPTER 3

PHYLLOEPIPHYTIC MICROBIAL COMMUNITY STRUCTURE ON

SPINACH (SPINACIA OLERACEA) LEAVES AT HARVEST AT

AFFECTED BY CULTIVAR AND ENVIRONMENT

ABSTRACT

Little is known about how phylloepiphytic microbial community structure changes on spinach leaves during production on the field and with the environment. The effect of changing environmental conditions prior to harvest and leaf surface topography on the diversity of phylloepiphytic bacteria from spinach cultivars was analyzed. Organically grown savoy, semisavoy and flat leaf cultivars of spinach were collected at three different periods during the fall growing season. Leaf surface topography had an effect on diversity and number of culturable bacteria on the phylloepiphytic community of spinach.

Savoy cultivars, which had larger surface area and more stomata and glandular trichomes where bacterial aggregates were observed, featured more diverse communities with increased richness and larger culturable bacterial populations compared to flat-leaved cultivars. Bacterial community richness was compared using denaturant gradient gel electrophoresis (DGGE), while abundance was quantified using primers specific for the 16s rRNA of major phyla. Environmental conditions immediately before harvest also affected bacterial community structure. The most diverse communities, both in richness and abundance, were observed during the first collection period, immediately following a period of rapid spinach growth. During the second collection period the communities were exposed to lower air and soil temperatures and decreased precipitation resulting in a significantly reduced bacterial population size and bacterial community richness. Bacterial population size increased during the third collection period; however, diversity was significantly lower than the first collection period, indicating that only a fraction of the bacterial community adapted and grew following changes in environmental conditions. Decreased richness and diversity indices corresponded with changes in the abundance of members of the phyla -

Proteobacteria and Actinobacteria. Cloning and sequencing of the most prominent bands from DGGE identified bacterial genera previously reported from phyllosphere communities; however, the majority of bands belonged to uncultured bacteria. This study described the effect of the plant characteristics and environmental conditions that affect spinach microbiota population size and diversity, which have implications in the survival of food and plant bacterial pathogens.

Key words: spinach, cultivar, collection period, environmental conditions, epiphytic bacteria, DGGE profile, richness, abundance, phyllosphere.

INTRODUCTION

Phyllosphere microbial communities are diverse, consisting of bacteria, fungi, algae, protozoa and infrequently nematodes. To successfully colonize the phyllosphere, epiphytic microorganisms must survive harsh conditions including nutrient limitation, low water availability, and UV irradiation (7) all of which affect microbial fitness and diversity (30). Recent culture-independent analyses of leaf washings of several plants (corn, green beans, sugar beet, Valencia oranges and cotton) indicated that leaf surface microbial communities are more complex than had been reported by culture-based methods (32). The study by Yang et al. revealed that the majority of 16S rDNA sequences recovered from the leaf washings of various plant species were from bacteria not previously described on leaf surfaces, with some sequences representing novel species. The membership of these phyllosphere communities is dominated by members of a few phylogenetic groups, chiefly -Proteobacteria, - Proteobacteria and

Bacteroidetes. -Proteobacteria and Firmicutes also form a large part of the bacterial community, while

Acidobacteria, Actinobacteria and Cyanobacteria occur infrequently (29).

Epiphytic population sizes and diversity are influenced by plant species (15), plant cultivar (24,

27), and stage of growth (11). Exposure to environmental conditions as well as different characteristics of the plants can modify the phyllosphere diversity and abundance (14). Culturable bacteria increased in overall numbers and types of colony morphologies observed on leaves of sugar beets collected during the autumn months compared to leaves collected during mid summer and winter months (28). Even small changes in climatic conditions influenced the microbiological composition of spinach (2). Fluctuations in bacterial diversity can also be attributed to differences among plant species and cultivars. Using denaturing gradient gel electrophoresis to address microbiota structure of different plant species, Yang et al. (2001) demonstrated that there was greater similarity in the structure of the microbiome within the same species (32). In relation to plant cultivar, differences in leaf surface topography and nutrients were important factors that affected the microbial community of the phyllosphere (9). Bacteria preferentially establish on epidermal cell wall junctions, glandular and non-glandular trichomes, veins, stomata and

93 epidermal cell wall surfaces (1). A study of the phyllosphere of Mediterranean perennial species, demonstrated that the population size of epiphytic bacteria is positively correlated with trichome densities

(31), nutrient-content, water availability and carotenoid composition (27, 31). Spinach cultivars that differed in surface characteristics (flat and savoy) exhibited different capacities for attachment of

Escherichia coli O157:H7 (19). It is possible that changes in leaf blade topography might be related to changes on the leaf surface, influencing the microbial diversity and population size on plants.

In recent years, spinach consumption was implicated in two E. coli O157:H7 outbreaks (12).

Survival and control of E. coli O157:H7 is being studied using a variety of approaches (3). Study of the phyllosphere of spinach leaves would improve the understanding of how pathogenic microorganisms are able to populate edible plants during pre- and postharvest handling since possible interactions with the members of bacterial populations on plants might play a role in their survival. The purpose of this work was to determine how the bacterial diversity of the spinach leaf epiphytic community changes in relation to environmental conditions on savoy, semi-savoy and flat-leaf spinach cultivars. Characterization of the membership of the epiphytic microbial community and changes in its structure will provide a better understanding of the ecology of the spinach phyllosphere that could be applied to elucidate mechanisms of survival for plant or food pathogens.

MATERIALS AND METHODS

Spinach production and harvest

Three spinach (Spinacia oleracea) cultivars „Monza‟ (flat), „Menorca‟ (savoy) and „Unipak‟

(semi-savoy) (Seedway LLC. USA), were seeded in September 19, 2007 at the Virginia Tech Kentland

Research Farm. One plot of 6 x 4 m was equally subdivided into four subplots approximately 0.9 x 6 m.

Each subplot contained four rows on raised beds each seeded with a different cultivar based in a complete randomized block design. Organic agricultural practices specified for the USDA National Organic

Program were followed. Nutrients were applied using composted vegetable waste and no pesticides were

94 used. Drip irrigation was applied every two days from seeding until irrigation was terminated on

November 21. Two hundred grams of spinach leaves from each cultivar were harvested when 7-10 cm long measured from petiole end to tip (“baby” leaf stage) from plots randomly selected using a round plastic hoop (covering approximately 0.30 m2), the sampling was performed at the center of the plot and the leaves of each cultivar were pooled. The leaves were harvested with scissors at the base of the petioles and stored in new self-sealing sterile plastic bags. Scissors were wiped with ethanol before use and between harvests of different cultivars. Samples were immediately taken to a laboratory, stored at approximately 4°C, and processed within 3 h. Leaves were harvested 36 days after seeding on October

25, 2007 (October harvest) and plants were allowed to re-grow for two subsequent harvests; after 58 days on November 17, 2007 (November harvest) and after 78 days on December 7, 2007 (December harvest).

For the three harvest periods, leaves were of similar physiological age. Leaves collected during

November and December harvest were sampled in relation to major microbial disturbances such as shifts in temperature and reduced rainfall.

Environmental conditions

Environmental conditions (maximum and minimum daily air and soil temperatures and precipitation) were recorded at a weather station (http://www.vaes.vt.edu/colleges/kentland/weather/) approximately 500 m from the field plot at the Virginia Tech Kentland Research Farm (Table 1). The accumulated growing degree units (GDUs), which are the number of temperature degrees or heat units above 4.7°C during the grow season, were calculated as: [(daily maximum temperature + daily minimum temperature)/2]-Base temperature. The base temperature used was 4.7 °C, the minimum air temperature at which crop development is expected to cease (33).

Microbial counts and DNA isolation from epiphytic community on spinach leaves

For each cultivar, 10 g of spinach leaves were placed in a Filtra-Bag® (FisherSci, Canada) with

90 mL of 1% (wt/vol) peptone water (Sigma-Aldrich Co., USA) and processed in a Pulsifier®

(Microbiology International, Frederick, MD) for 5 min. The pulsifier removed bacteria attached to the surface of the spinach leaves, without disrupting the tissue, and reduced the chlorophyll and spinach DNA in the sample as compared to commercial stomacher (6). One milliliter of the resulting suspension was serially diluted and plated onto R2A medium (Difco, MI). The plates were incubated at 25°C for 16 days.

The rest of the bacterial suspension was utilized to collect cells by centrifugation at 4,000 rpm/20min/4oC washed with 1X PBS and resuspended in 100 L of 1X TE buffer. The cells were lysed by incubation with 300 g of lysozyme (Fisher, Fair Lawn NJ), 10 units of mutanolysin (Fisher) and 25g of achromopeptidase (Sigma-Aldrich Co., St Louis MO) at 37oC for 30 min. Lysates were treated with 25 units of proteinase K (Fisher) and incubated at 65oC for

30 min. DNA was extracted from the lysed cells using the ZR soil microbe DNA kitTM (Zymo

Research Co., Orange, CA) per manufacturer‟s instructions. Microbial counts and DNA isolation were made in triplicate, each with 10 g of leaves per cultivar per harvest period.

Measurement of plant anatomical features

The leaf area for each cultivar at the "baby" leaf stage was measured using a leaf area meter

(model LI3000 LI-COR Biosciences, Lincoln, NE). Ten leaves from each cultivar were randomly selected and each leaf measurement was repeated three times, additionally length, thickness and width was measured to standardize the surface area value of each leaf. Five individual leaves of each cultivar were collected to determine the stomata density using differential interference contrast (DIC) microscopy

(Axiophot compound microscope Zeiss, Thornwood, NY) with 200x magnification. The numbers of stomata were manually counted on both the abaxial and adaxial surfaces of each leaf in five fields of approximately 0.051 mm2 each. Leaf surfaces were covered with clear nitrocellulose cotton solution (nail polish) that was removed along with the cuticle after drying with clear tape (Scotch, 3M, Minneapolis,

MN) to obtain stomata imprints for counting. Stomata density was converted to stomata per cm2. The density of glandular trichomes was assessed using a dissection microscope (Cambridge Instruments). Five

96 leaves of each cultivar were selected and trichomes were counted in 1 cm2 on abaxial and adaxial surfaces of each leaf, three fields of 1 cm2 were examined for each leaf.

Microbiota surrounding stomata were observed on gold-coated abaxial surface of mature leaves using environmental scanning electronic microscopy (FEI quanta 600 FEG) under high vacuum

(approximately 7 x 10-5 Torr).

DGGE analysis of PCR amplified 16S rRNA gene fragments

Product amplification. The 16S rRNA gene was amplified from the total epiphytic microbiota

DNA (50 ng/L) to generate a 566 bp fragment using the primers 341-f (5‟-CCT ACG GGA GGC AGC

AG-3‟) and 907r (5‟-CCG TCA ATT CMT TTG AGT TT-3‟) (21). The forward primer was modified to add a 40 nucleotide GC clamp at the 5‟ end (5‟-CGC CCG CCG CGC GCG GCG GGC GGG GCG GGG

GCA CGG GGG G-3‟) (21, 23). Each 25 L reaction contained 1.5mM of MgCl2, 50mM of KCl, 0.2mM of each dinucleotide, 1% of DMSO (dimethylsulfoxide), 25mM of Tris-HCl (pH 8), 1U/L of HotStart-IT

FideliTaq DNA polymerase (USB 71156, Cleveland, OH, USA), 0.5 M of each primer, and 50 ng of

DNA. The size and intensity of PCR products were confirmed using 0.9% agarose gels (Fisher-Scientific,

Atlanta, GA).

DGGE conditions. The PCR products were run on a 8% polyacrylamide gel in a 30-60% denaturant gradient of urea and formamide (100% denaturant corresponds to 7M urea plus 40% (vol/vol) of deionized formamide) using the Bio-Rad DCodeTM Universal Detection System (Bio-Rad, Hercules,

CA). Twenty-two microliters of PCR products were separated at constant 85V and temperature of 60oC for 17 h. The DNA bands were visualized by staining with ethidium bromide (5ug/mL) and photographed using the Molecular Imager ® GelDocTM XR (Bio-Rad). The bands were analyzed using

Quantity One®-1D analysis software (Bio-Rad) and the DGGE profiles were analyzed by clustering using the unweighted pair group method with mathematical averages (UPGMA; Dice coefficient of similarity) for the 3 cultivars at each harvest time.

Band identification. Bands consistently present in all three cultivars and at all harvest times were sliced and suspended in 100 L of water, and incubated at 4oC for 24 h to allow passive DNA diffusion. Diffused DNA was re-amplified using 341F and 907R primers as described above. The PCR products were purified from a 1 % (w/v) agarose gel and cloned into pCR4-TOPO® VECTOR

(Invitrogen) to transform One Shot® Match1 ™ chemically competent E. coli (Invitrogen). Plasmids were isolated using Quick lyse miniprep (Qiagen) and inserts were sequenced. Sequences were compared to the

NCBI database using BLASTN (www.ncbi.nlm.nih.gov) and identities were assigned based on similarity to known taxa. A total of 5 clones per band were analyzed.

Abundance of representative phyla of spinach phyllosphere. Relative abundance of dominant phyla was assessed by real time amplification of phyla specific regions of the 16S RNA gene as described by Fierer et al. 2007 (5). Phylum-specific abundance was determined targeting the following phylogenetic groups: Proteobacteria, -Proteobacteria, Firmicutes, Actinobacteria and Bacteroidetes. Briefly, each 25 L reaction contained a respective amount of DNA template, 12.5 L of HotSart-

TM IT SYBR®Green qPCR Master Mix 2X which contains 5 mM and 0.4 mM of MgCl2 and nucleotides respectively (USB® 75770 Cleveland, OH, USA), 10nM of fluorescein as passive reference dye (USB®

75767 Cleveland, OH, USA) and 0.5 M of forward and reverse primers. PCR conditions were denaturation at 95oC for 2min, followed by 40 cycles of denaturation at 95oC for 30 s, 30 s at the annealing temperature, and 72oC for 1 min. Annealing temperatures were: , 55°C for all Bacteria, 60oC for α-Proteobacteria, β-Proteobacteria, Actinobacteria, Firmicutes (5). Amplification was carried out with an iQTM Optical system Real-time PCR detection system (Bio-Rad).

A triplicate of each DNA sample from the three spinach cultivars and from the three times of collection were adjusted to 50 ng/L and real-time amplification was performed

Statistical analysis

Bacteria enumeration was normalized by conversion to log10 CFU/g of spinach. Statistical comparison among treatments was conducted using Proc mix and means were separated using Tukey’s multiple comparison using SAS (Statistical Analysis System 9.2, SAS Institute, Cary, NC). Statistical significance was determined when p< 0.05.

RESULTS

Environmental conditions. From planting to first harvest, temperatures were warmest (Table 1).

During the first harvest period, spinach plants were exposed to the most days over 25oC, and temperatures never exceeded 25oC after first harvest. All spinach cultivars were harvested as immature "baby" spinach after accumulating about 1000 GDUs (Table 1). Only 190 GDUs were accumulated in November and 85 in December so little crop development occurred after first harvest. Subfreezing temperatures were not reached until after first harvest. Average minimum temperatures fell from 10oC during the first harvest period to below zero for the remaining two periods. Precipitation for both September/October and

November were very similar, 11.5 cm for both periods, while December was much drier (Table 1). Soil temperatures declined gradually during each period, but remained above freezing.

Aerobic total microbial counts. Significantly larger microbial counts (p<0.05) were recorded for savoy „Menorca‟ and „Unipak‟ than for the flat-leaf cultivar Monza at all collection periods. The total numbers of aerobic culturable bacteria were significantly larger during the months of October and

December (Table 2) (p<0.05) than those recorded in November (Table 2). After the population decline in

November, the microbial counts recovered in December on „Monza‟ but not on the other two cultivars.

Spinach anatomical features. The mean leaf surface area of cultivars differed by approximately

0.2 cm2 each. The surface area of the savoy „Menorca‟ was largest, while the flat „Monza‟ had the smallest mean leaf surface area (Table 3). Stomata densities on abaxial versus adaxial leaf surfaces were

99 significantly different for each cultivar; however, only „Monza‟ possessed fewer stomata on the lower surface compared to the other cultivars (Table 3). Glandular trichomes were observed only on leaves of

„Menorca‟ and „Unipak‟ (Table 3).

Microbial aggregates were observed surrounding stomata on all three cultivars (Fig. 1), but only on ‘Unipak’ aggregates were directly observed on guard cells. The largest microbial density was observed on ‘Menorca’ leaves.

DGGE analysis. Differences in number and position of bands were noted among the harvest periods and among the three cultivars (Fig. 2). Seven bands were present in electropherograms of samples harvested in October and November for all three cultivars (Fig. 2). With exception of „Unipak‟ (Fig. 2, lane 9), another set of bands was present in samples from all three cultivars harvested during the month of

December (Fig. 2). Some bands showed variations in intensity but remained present during all harvest periods. The greatest species richness, as represented by the number of bands, was observed for

„Menorca‟ during October (Fig. 2 lane 2). A few bands from this set (Fig. 2) were present during

December for „Monza‟ and „Menorca‟, but absent from all cultivars in November.

Bands from DGGE (Fig. 2) were identified as belonging to the phyla α-Proteobacteria,

Bacteroidetes and Actinobacteria. The majority of sequences were closely related to sequences that belonged to non-culturable bacteria. Some bands were composed of more than one microorganism. Band

6 represents spinach chloroplasts and thus was not evaluated for changes in diversity or richness.

Dendrograms were constructed to compare community profiles (Fig. 3), showed more similarity between communities from spinach harvested in October and December compared to communities from spinach harvested in November.

Determination of specific phyla abundance and diversity indices. The abundance of all bacteria was determined by qRT-PCR using universal 16S rRNA primers. Community structure at the phylum level was determined using primers designed to specific regions of the 16S rDNA targeting all

100 bacteria, Actinobacteria, Firmicutes, α-Protoebacteria, and β-Proteobacteria (Fig. 4). The total number of bacterial 16S rDNA sequences detected in October and December were significantly higher (p<0.05) than bacterial sequences detected in November for all three cultivars. The most abundant 16srDNA sequences belonged to α-Proteobacteria and Firmicutes (Fig. 3). Members of the β-Proteobacteria were the least abundant. The total number of bacterial sequences for α-Proteobacteria and Bacteroidetes were significantly less abundant in November for all three cultivars.

Simpson‟s diversity index, calculated at each harvest time for all cultivars using the abundance of individual groups, were significantly lower (p<0.05) for samples collected during November than for

October and December (Fig. 5). The diversity indices for „Monza‟ were not significantly different in samples from October and November (Fig. 5). However, samples collected in December showed a significant increase in their diversity index. The largest diversity index during October was for

„Menorca‟, yet during November this cultivar had the lowest diversity index. „Unipak‟ had a larger diversity index in October, but declined in November and December without significant differences between these months (p>0.05). Real-time PCR showed no significant differences in phylum composition among cultivars collected during the same harvest periods, consistent with the DGGE patterns obtained for all three cultivars.

DISCUSSION

This study analyzed changes in the structure of the epiphytic microbial communities present on spinach leaves using culture and non-culture dependent techniques. We selected three spinach cultivars;

Monza, Menorca and Unipak, with differences in leaf surface area (Table 2). Spinach is a cool season crop with optimal growing temperatures between 15 to 18oC, a minimum growth temperature of 4.7°C

(26) and a maximum growth temperature of 24°C. Spinach can withstand minimum temperatures between

-9 to -6oC. All cultivars developed to the immature "baby" stage, like those typically harvested for modified atmosphere packaged salads, after accumulating approximately 1000 GDUs from planting to

101 first harvest in October. Fewer degree days were accumulated in November or December, so spinach plants largely remained in a non-growing vegetative state during the final two harvest periods (Table 1).

Only “baby” spinach leaves were collected to reduce bias due to leaf age, although little additional growth occurred after the first harvest due to the lack of GDUs accumulation. By harvesting leaves from the same stages of development, it was possible to determine that changes in the microbial community structure were due to environmental conditions, particularly temperature and rainfall, to which the spinach cultivars were exposed. Microbial diversity patterns differed between immature and mature plants, with increased species richness seen in younger leaves of olive and sugar beet leaves (4, 28).

Variations in the numbers of bacteria with leaf blade topography have been previously reported (29).

Broad leaf plant species, such as cucumber, lettuce and beans typically have larger aerobic microbial counts than grasses (15). Cultivar differences affect the ability of Pseudomonas syringae pv. syringae to colonize snap beans (8), and for Salmonella enterica to colonize lettuce (16). We selected three spinach cultivars; Monza, Menorca and Unipak, that differed in leaf surface area (Table 2). In this study, the savoy „Menorca‟ and semi-savoy „Unipak‟ had significantly higher diversity indices and larger numbers of culturable bacteria, than the flat leaf „Monza‟ during the first collection period. This suggests that leaf blade topography influenced the number of microorganisms harbored on the plant surface. Changes attributed to cultivar may be related to differences in total surface area that serves as niches for microbial establishment and development. In this study significantly larger mean leaf area was measured for

„Menorca‟. Savoy cultivars such as Menorca and Unipak have leaf blades that are curly between leaf veins, increasing the leaf surface area available for colonization. „Menorca‟ and „Unipak‟ had more anatomical leaf features such as glandular trichomes and stomata (Table 3, Fig. 1), where microbial aggregates and bacteria were observed (24). Veins, hooked trichomes and glandular trichomes, have preferential distribution associated with bacterial aggregates on leaf surfaces of beans (20) and strawberry leaves (17). Foodborne pathogens such as Listeria monocytogenes and E. coli O157:H7 associated with leaves of Arabidopsis thaliana and spinach, respectively, also have shown to be associated with these plant structures (18, 19).

102

DGGE profiles of the three cultivars are most similar during the month of November (Fig. 3).

However, the numbers of bacteria in different phyla were significantly different among cultivars harvested in November and December (Fig. 4), indicating that the changes associated with spinach cultivars are due to differences in abundance but not in richness.

In this study, decreases in culturable bacterial populations, species richness and diversity indices from the October to November harvest, were also associated with decreases in temperature and rainfall, and exposure of the crop to suboptimal and freezing temperatures (Table 1).

Overall, the microbial community of spinach is conserved and stable, as indicated by the presence of the same bands for all cultivars and at all harvest periods (Fig. 2). These bands likely represent bacteria that are robust, even in stressful conditions like low temperature. Actinobacteria and Firmicutes populations were maintained without significant variations during the season for the three spinach cultivars. This indicates that members of these two phyla maintain fitness regardless of changes in environmental conditions and therefore contribute to the overall stability of the community. In contrast numbers of bacteria belonging to the phylogenetic group -Proteobacteria (Fig. 4) fluctuated greatly between October and December harvests and were the least stable members of the community. The majority of the members of the epiphytic community belonged to the phylogenetic group -

Proteobacteria and Bacteroidetes. The same phylogenetic groups are dominant on other plant leaf surfaces including Thlaspi goesingense, Zea mays and Capsicum annuum (10, 13, 24, 29).

The number of culturable aerobic bacteria (Table 2), the DGGE profiles (Fig. 2) , results of phylum specific abundance (Fig. 4) and Simpson‟s diversity indices (Fig. 5) revealed that there was a clear shift in abundance, richness and diversity between the October and November harvests. The

December harvest had greater numbers of bacteria and was more diverse (larger Simpson‟s diversity index). It is likely that part of the epiphytic microbial community adapted to the decrease in temperature and rainfall then, increased in abundance, illustrating the ability of epiphytic microorganisms to overcome

103 stressful conditions. These changes could also be attributed to bacterial succession resulting from shifts in temperature. It was anticipated that lower air temperatures would cause a reduction in the number of bacteria; however, similar studies have showed that it is not possible to predict changes in bacterial community composition based on changes in meteorological conditions (25).

The microbial community structure of the spinach phyllosphere is impacted by the cultivar and environmental conditions during spinach development. Community members vary in fitness to decreased environmental temperatures and rainfall but after a period of adaptation, the phyllosphere community successfully recovers its abundance but not its richness.

ACKNOWLEDGEMENTS

We acknowledge the assistance of Brinkley Benson for his contribution in the cultivation of spinach during the fall season of 2007. To Glenda Gillaspy for her expertise and assistance in the utilization of DIC microscopy. ESEM was carried out using instruments in the Nanoscale

Characterization and Fabrication Laboratory, a Virginia Tech facility operated by the Institute for Critical Technology and Applied Science.

104

REFERENCES

1. Borges-Baldotto, L., and F. Lopes-Olivares. 2008. Phylloepiphytic interaction between bacteria

and different plant species in atropical agricultural system Can J Microbiol 54:918-931.

2. Conte, A., G. Conversa, C. Scrocco, I. Brescia, J. Laverse, A. Elia, and M. DelNobile. 2008.

Influence of growing periods on the quality of baby spinach leaves at harvest and during storage

as minimally processed produce. Postharvest Biol. and Technol. 50:190-196.

3. Delaquis, P., S. Bach, and L. D. Dinu. 2007. Behavior of Escherichia coli O157:H7 in leafy

vegetables. J Food Prot 70:1966-74.

4. Ercolani, G. L. 1991. Distribution of epiphytic bacteria on olive leaves and the influence of leaf

age and sampling time. Microb Ecol 21:35-48.

5. Fierer, N., J. A. Jackson, R. Vilgalys, and R. B. Jackson. 2005. Assessment of soil microbial

community structure by use of taxon-specific quantitative PCR assays. Appl Environ Microbiol

71:4117-20.

6. Fung, D., A. N. Sharpe, B. Hart, and Y. Liu. 1998. The pulsifier: A new instrument for

preparing food suspensions for microbiological analysis. J Rapid Meth and automation in

Microbiol 6:43-49.

7. Gnanamanickam, S., and J. Immanuel. 2006. Epiphytic bacteria, their ecology and functions,

p. 131-154. In S. Gnanamanickam (ed.), Plant-Associated bacteria. Springer, Dordrecht, The

Netherlands.

8. Hirano, S. S., L. S. Baker, and C. D. Upper. 1996. Raindrop Momentum Triggers Growth of

Leaf-Associated Populations of Pseudomonas syringae on Field-Grown Snap Bean Plants. Appl

Environ Microbiol 62:2560-2566.

9. Hirano, S. S., E. V. Nordheim, D. C. Arny, and C. D. Upper. 1982. Lognormal Distribution of

Epiphytic Bacterial Populations on Leaf Surfaces. Appl Environ Microbiol 44:695-700.

105

10. Idris, R., R. Trifonova, M. Puschenreiter, W. W. Wenzel, and A. Sessitsch. 2004. Bacterial

communities associated with flowering plants of the Ni hyperaccumulator Thlaspi goesingense.

Appl Environ Microbiol 70:2667-77.

11. Jacques, M., L. L. Kinkel, and C. E. Morris. 1995. Population Sizes, Immigration, and Growth

of Epiphytic Bacteria on Leaves of Different Ages and Positions of Field-Grown Endive

(Cichorium endivia var. latifolia). Appl Environ Microbiol 61:899-906.

12. Jay, M. T., M. Cooley, D. Carychao, G. W. Wiscomb, R. A. Sweitzer, L. Crawford-Miksza,

J. A. Farrar, D. K. Lau, J. O'Connell, A. Millington, R. V. Asmundson, E. R. Atwill, and R.

E. Mandrell. 2007. Escherichia coli O157:H7 in feral swine near spinach fields and cattle,

central California coast. Emerg Infect Dis 13:1908-11.

13. Kadivar, H., and A. E. Stapleton. 2003. Ultraviolet radiation alters maize phyllosphere bacterial

diversity. Microb Ecol 45:353-61.

14. Kinkel, L. L. 1997. Microbial population dynamics on leaves. Annu Rev Phytopathol 35:327-47.

15. Kinkel, L. L., M. Wilson, and S. E. Lindow. 2000. Plant Species and Plant Incubation

Conditions Influence Variability in Epiphytic Bacterial Population Size. Microb Ecol 39:1-11.

16. Klerks, M., E. Franz, G.-P. M, C. Zijlstra, and A. Bruggen. 2007. Differential interaction of

Salmonella enterica serovars with lettuce cultivars and plant-microbe factors influencing

colonization efficiency. Isme J 1:620-631.

17. Krimm, U., D. Abanda-Nkpwatt, W. Schwab, and L. Schreiber. 2005. Epiphytic

microorganisms on strawberry plants (Fragaria ananassa cv. Elsanta): identification of bacterial

isolates and analysis of their interaction with leaf surfaces. FEMS Microbiol Ecol 53:483-92.

18. Milillo, S. R., J. M. Badamo, K. J. Boor, and M. Wiedmann. 2008. Growth and persistence of

Listeria monocytogenes isolates on the plant model Arabidopsis thaliana. Food Microbiol

25:698-704.

19. Mitra, R., E. Cuesta-Alonso, A. C. Wayadande, J. Talley, S. Gilliland, and J. Fletcher. 2009.

Effect of route of introduction and host cultivar on the colonization, internalization, and

106

movement of the human pathogen Escherichia coli O157:H7 in spinach. J Food Prot 72:1521-

1530.

20. Monier, J. M., and S. E. Lindow. 2004. Frequency, size, and localization of bacterial aggregates

on bean leaf surfaces. Appl Environ Microbiol 70:346-55.

21. Muyzer, G., T. Brinkhoff, U. Nubel, C. Santegoeds, H. Schafer, and C. Wawer. 2004.

Denaturing gradient gel electrophoresis (DGGE) in microbial ecology, p. 743-770. In K. A.

Publishers (ed.), Molecular Microbial Ecology Manual, Second ed, vol. 13. Kluwer Academic

Publishers, Netherlands.

22. Muyzer, G., E. C. de Waal, and A. G. Uitterlinden. 1993. Profiling of complex microbial

populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-

amplified genes coding for 16S rRNA. Appl Environ Microbiol 59:695-700.

23. Muyzer, G., and K. Smalla. 1998. Application of denaturing gradient gel electrophoresis

(DGGE) and temperature gradient gel electrophoresis (TGGE) in microbial ecology. Antonie Van

Leeuwenhoek 73:127-41.

24. Rasche, F., R. Trondl, C. Naglreiter, T. G. Reichenauer, and A. Sessitsch. 2006. Chilling and

cultivar type affect the diversity of bacterial endophytes colonizing sweet pepper (Capsicum

anuum L.). Can J Microbiol 52:1036-45.

25. Redford, A. J., and N. Fierer. 2009. Bacterial succession on the leaf surface: a novel system for

studying successional dynamics. Microb Ecol 58:189-98.

26. Rubatzky, V., and M. Yamaguchi. 1997. Spinach, table beets, and other vegetable chenopods,

p. 457-463. In V. Rubatzky and M. Yamaguchi (ed.), World Vegetables, 2nd. ed. International

Thompson Publishing, NY.

27. Ruppel, S., A. Krumbein, and M. Schreiner. 2008. Composition of the phyllospheric microbial

populations on vegetable plants with different glucosinolate and carotenoid compositions. Microb

Ecol 56:364-72.

107

28. Thompson, I. P., M. J. Bailey, J. S. Fenlon, T. R. Fermor, A. K. Lillley, J. M. Lynch, P. J.

McCormack, M. P. McQuilken, and K. J. Purdy. 1993. Quantitative and qualitative seasonal

changes in the microbial community from the phyllosphere of sugar beet. Plant Soil 150:177-191.

29. Whipps, J. M., P. Hand, D. Pink, and G. D. Bending. 2008. Phyllosphere microbiology with

special reference to diversity and plant genotype. J Appl Microbiol 105:1744-55.

30. Wilson, M., and S. E. Lindow. 1994. Inoculum Density-Dependent Mortality and Colonization

of the Phyllosphere by Pseudomonas syringae. Appl Environ Microbiol 60:2232-2237.

31. Yadav, R. K., K. Karamanoli, and D. Vokou. 2005. Bacterial colonization of the phyllosphere

of mediterranean perennial species as influenced by leaf structural and chemical features. Microb

Ecol 50:185-96.

32. Yang, C. H., D. E. Crowley, J. Borneman, and N. T. Keen. 2001. Microbial phyllosphere

populations are more complex than previously realized. Proc Natl Acad Sci U S A 98:3889-94.

33. Yang, S., J. Logan, and D. L. Coffey. 1995. Mathematical formulae for calculating the base

temperature for growing degree days. Agric Forest Meteorol 74:61-74.

108

Table 1. Environmental conditions registered in Blacksburg, VA from the time of seeding to the last harvest¶.

From day of seeding From first harvest to From second harvest (September 19, 2007) second harvest to third harvest to first harvest (November 17, 2007). (December 7, 2007). (October 25, 2007).

Collection time October November December

Accumulated GDUs§ 995 190 86

Average of maximum o air temperatures ( C) 16.7 + 3.7 6.8 + 3.8 4.7 + 4.4

Average of minimum

air temperatures (oC) 9.8 + 4.5 -0.2 + 4.8 -2.0 + 4.6

Average of soil 17.4 + 1.9 10.4 + 2.5 7.1 + 2.0 temperatures (°C)

Total precipitation (cm) 11.6 11.5 1.8

Number of days at 25oC or above 13 0 0

Number of days at 0oC 0 0 4 or below

¶Climatological data recorded by a weather station approximately 500 m from the spinach test plot at the

Virginia Tech Kentland Research Farm near Blacksburg, VA over the period from seeding to harvest

§Accumulated GDUs are the number of temperature degrees or heat units above 5°C accumulated during the growing season; 5 °C is the base temperature to which plant growth is expected.

109

Table 2. Number of culturable epiphytic bacteria on baby spinach leaves for the three spinach cultivars harvested at three different time points as determined by plating on R2A media§.

Aerobic bacterial

population

¶ Cultivar log of CFU/ g spinach

Monza

October Harvest 5.93 + 0.0331

November Harvest 5.02 + 0.0362

December Harvest 5.84 + 0.0601

Menorca

October Harvest 6.47 + 0.0481

November Harvest 4.58 + 0.0652

December Harvest 5.14 + 0.0372

Unipak

October Harvest 7.47 + 0.0951

November Harvest 4.13 + 0.0383

December Harvest 5.49 + 1.0902

¶ Results are represented by the mean of three independent replicates + standard deviation. Different numbers denote significance difference (p<0.05) within each cultivar and incubation time.

110

Table 3. Morphological features of spinach cultivars£.

Leaf area Glandular trichomes Stomata

Normalized area (leaf (no. of glandular (no. of stomata/cm2) area/leaf length) trichomes/cm2)

Abaxial Adaxial Abaxial Adaxial Cultivar Entire leaf surface surface surface surface

2.63 x 104 + 3.69 x 104 + Monza 2.82 + 0.37b N/D N/D 6370a 3368a

3.28 x 104 + 4.79 x 104 + Menorca 3.27 + 0.12a 9.15 + 1.50b 31.3 + 6.79a 5590a 8207b

2.63 x 104 + 5.35 x 104 + Unipak 3.02 + 0.11b 3.00 + 2.12b 18.66 + 9.81a 3069a 1375b

£ Data represent means + standard deviations of each measurement of 5 spinach leaves.

Mean with different letter within a column are significantly different (p<0.05)

(N/D) glandular trichomes were not detected for „Monza‟

111

Figure 1. ESEM microscopy of spinach stomata for cultivars (A) Monza, (B) Menorca and (C) Unipak.

112

Figure 2. Analysis of the structure of epiphytic microbial community by DGGE separation of 16S rRNA fragments with 341-f and 907r primers. Lanes: (1) „Monza‟ –October-, (2) „Menorca‟ –October-, (3) „Unipak‟ –October-, (4)

„Monza” –November-, (5) „Menorca‟ –November-, (6) „Unipak‟ –November-, (7) „Monza” –December-, (8)

„Menorca‟ –December- and (9) „Unipak‟ (December). Band identification: closest match to NCBI database - accession number and % of similarity-(1) Uncultured proteobacterium clone -EF602176- 97% - and Uncultured

Oxalobacteraceae -FJ037285-97%-. (2) Kineococcus sp. –FJ214365- 96%-, Uncultured beta proteobacterium -

EFO74569- 98%-, Terrabacter sp. -FJ772036- 99%- and Oryzihumus sp. -GQ355280- 99%-. (3) Uncultured bacterium clone -FJ562178- 99%- and Uncultured actinobacterium clone -EF651015- 89%-. (4) Uncultured

Baceroidetes bacterium -AM116743- 96%- and Flavobacterium sp. -AY599661- 98%-. (5) Chryseobacterium sp. -

EF471218-86%- and Uncultured Proteobacterium clone -EF019491- 88%- (6) Spinacia oleracea chloroplast -

AJ400848- 98%- (7) Uncultured soil bacterium clone -DQ297980- 97%-. (8) Uncultured alpha Proteobacteria -

CU926667- 95%-

113

Figure 3. UPGMA dendrogram constructed based on the similarity between DGGE profiles for the three spinach cultivars and the three times of collection. (Percentage of similarity is indicated at each node)

114

Figure 4A. Abundance of members of five bacterial phyla in epiphytic communities isolated from

„Monza‟ at three different harvest periods, determined by qPCR. (n=3). Error bars represent standard deviations of the mean. *Tukey comparison. P<0.05

115

Figure 4B. Abundance of members of five bacterial phyla in epiphytic communities isolated from

„Menorca‟ at three different harvest periods, determined by qPCR. (n=3). Error bars represent standard deviations of the mean. *Tukey comparison. P<0.05

116

Figure 4C. Abundance of members of five bacterial phyla in epiphytic communities isolated from

„Unipak‟ at three different harvest periods, determined by qPCR. (n=3). Error bars represent standard deviations of the mean. *Tukey comparison. P<0.05

117

Figure 5. Simpson‟s diversity indices (D) calculated based on the abundance of members of five phyla in communities isolated from three spinach cultivars harvested at three different collection times.

(Simpson‟s diversity index was represented as 1-D to facilitate the analysis). Independent replicates were used (n=3). Error bars represent standard deviation of the mean.

118

CHAPTER 4

GROWTH OF ESCHERICHIA COLI O157:H7 IS INFLUENCED BY

INTERACTIONS WITH SPINACH PHYLLOEPIPHYTIC BACTERIA

ABSTRACT

Fresh edible plants, like spinach, have emerged as a major vehicle for Escherichia coli O157:H7 and other human enteric pathogens. E. coli O157:H7 can survive and establish populations on the leaf surfaces which are composed of a diverse community of other bacteria. We hypothesize that these bacteria impact the fitness of E. coli O157:H7 on edible plants. In this study we identified bacterial isolates obtained from spinach leaf surfaces (phylloepiphytes) that exhibited positive or negative interactions with E. coli O157:H7 in vitro. Negative interactions, defined as growth inhibition of the pathogen, were observed with bacteria belonging to the genera Erwinia, Pseudomonas and other genera not previously described, including Stenotrophomonas spp. and Brevundimonas spp. Utilization of the same carbon sources is likely responsible for competitive exclusion of E. coli O157:H7. We also demonstrated relationships between E. coli O157:H7 with phylloepiphytes Rhizobium sp., Acidovorax sp.,

Flavobacterium sp., Sphingomonas sp., Kaistia sp., and Patulibacter sp. These bacteria showed different carbon source utilization profiles than E. coli O157:H7 and did not affect the establishment on E. coli

O157:H7 in co-culture experiments. Members of the microbial community influenced the growth in vitro of E. coli O157:H7, showing the importance of the spinach bacterial community on the fate of this bacterium.

Key words: spinach, bacterial interactions, competitive exclusion, positive and negative interaction, phylloepiphytic bacteria, Escherichia coli O157:H7.

119

INTRODUCTION

Fresh edible plants have become a major vehicle for enteric pathogens such as E. coli O157:H7 and Salmonella enterica (14). Plants can become contaminated at several stages of production via: direct interaction with livestock (2), or wildlife (12); manure tainted irrigation waters (6, 16, 33) or unsanitary packaging facilities (3). Once the produce is contaminated, these pathogens survive the minimal processing procedures (5). Understanding the factors that enable enteric pathogens to survive on and colonize the surface of plants is paramount to the development of strategies to mitigate the risks to consumers.

Plant surfaces are colonized by bacteria (phylloepiphytes), which inhabit what is termed the phyllosphere (22). When enteric pathogens land on the plant surfaces, they must co-exist in an ecological niche already populated by a diverse community of bacterial epiphytes adapted to conditions on the surface (4). The fate of these incomers is to be outcompeted by epiphytic bacteria or possibly gain fitness on the leaf surface by interaction with other phyllosphere members (9). Bacterial interactions within a microbial community could be beneficial (positive), neutral (no effect) or antagonistic (negative) in nature. Negative interactions can occur through competition, in which bacteria compete for nutrients and space, or by antibiosis, as a result of secreted compounds that can directly inhibit growth or kill potential competitors (23). In contrast, positive interactions involve bacterial associations that can benefit the establishment of other bacterial populations by commensalism or symbiosis (23). Additionally, there are indirect interactions that occur when changes in the physicochemical properties of the environment are altered by the metabolism of another member (11). Alternatively, the reduction of some carbon sources could make available molecules, which can be used by enteric pathogens to co-exist on the phyllosphere and enhance epiphytic fitness of pathogenic bacteria, thus a positive interaction (1). Waustera paucula, an epiphyte isolated from Arabidopsis thaliana and lettuce, enhanced growth of E. coli O157:H7 by 5- fold in sterile lettuce plant exudates (9). W. paucula failed to utilize most of the same substrates available in lettuce exudates that were utilized by E. coli O157:H7 (9). This likely indicates that if phylloepiphytes

120 differ in their utilization of nutrients, they may allow E. coli O157:H7 to establish within the same niche without competition. Positive interactions with the phylloepiphytes could provide a mechanism for survival and protection of enteric pathogens on edible plants and the study of these types of associations may help to understand the ecology of enteric pathogens on leaf surfaces.

In vitro and in vivo studies using bacteria isolated from the phyllosphere of different plant species have demonstrated the ability of phylloepiphytes to antagonize the growth of several foodborne pathogens

(9, 17, 21, 31). There is special interest in these bacteria as bio-control agents to prevent enteric pathogens from establishment on the leaf surfaces (14). Bio-control has been demonstrated as a strategy for controlling plant pathogens (32, 37). The application of bacterial antagonists towards enteric pathogens could be part of an integrated strategy to reduce or promote the growth of these microorganisms.

The objective of this work was to characterize the types of interactions between E. coli O157:H7 and spinach phylloepiphytic bacteria to identify endemic bacteria, which influence fitness of E. coli

O157:H7 on edible plants.

MATERIALS AND METHODS

Escherichia coli O157:H7 and Wausteria paucula cultures.

E. coli O157:H7 (CDC PulseNet pattern number EXHX01.0124) previously isolated from spinach leaves during the spinach outbreak of 2006 and E. coli O157:H7 gfp+/kan+ (transformed from strain H1730, a clinical isolate associated with a lettuce outbreak) were used in this study. The strains were activated in tryptic soy broth (TSB, Difco, Spark MD) at 37oC for 48h. Cultures were screened for purity on sorbitol MacConkey agar (SMAC, Difco) and an isolated colony was transferred to TSB,

o incubated at 37 C/120 rpm until an OD600 nm of 0.9-1.0 was achieved. The cells were collected by centrifugation (4000 rpm, 15 min, 4oC), washed twice with 1X phosphate buffered solution (PBS; 137 mM NaCl, 2.7 mM KCl, 100mM Na2HPO4, 2 mM KH2PO4) and resuspended in 1X PBS to achieve a final

121 concentration of 106 CFU/mL. Growth kinetics of both E. coli O157:H7 strains were compared by construction of growth curves using an automated growth curve analysis system (Bioscreen C™

Piscataway, NJ). No differences in growth rates and kinetic values were observed.

W. paucula strain RM5019 was utilized as a positive control as it previously showed to enhance the growth of E. coli O157:H7 (9). After its reception the strain was grown on TSB and stored at -80°C in

TSB supplemented with 30% glycerol stocks until its utilization.

Isolation of bacteria from spinach surfaces

Fresh spinach was grown and collected as previously described (25). Briefly, 10 grams of spinach leaves (replicated six times) were placed in a filtered bag with 90 mL of 1% (wt/vol) peptone water

(Sigma-Aldrich Co., USA) supplemented with 1% (vol/vol) of Tween-90 (PTW) (Fisher, Atlanta, GA) and processed in a Pulsifier® (Microbiology International, Frederick, MD) for 5 min. One milliliter of the suspension was serially diluted and plated onto R2A media (Difco, MI) at 25°C for up to 16 days.

Bacterial suspensions were replicated in triplicate for each batch of 10 g of spinach.

A power test to determine the sample size of the number of isolates, based on bacterial population of the spinach samples, was assessed using Statistical solutions ™ software, with an value of 0.05 and a power of 0.80. A total of 1512 colonies were randomly selected from R2A plates that contained well isolated colonies and freezer stocks of each isolated were prepared in TSB with 30% glycerol. All isolates were stored at -80°C for further screening for interactions with E. coli O157:H7.

Screening bacterial isolates for growth inhibition of E. coli O157:H7

The ability to inhibit growth of E. coli O157:H7 was determined for all bacterial spinach isolates

(n=1512) using an agar spot test as described by Fleming et al. (10). Briefly, 100 L of a solution of 6 log10 cells (stationary phase) of E. coli O157:H7 suspended in PBS 1X was spread onto R2A plates and allowed to dry for 30 min within a biological safety cabinet. Each spinach isolate was resuscitated onto

R2A media and allowed to grow for 48 h before spotting onto the lawn of E. coli O157:H7 using a sterile

122 toothpick. A maximum of six colonies were screened on a single R2A plate. Plates were incubated at

25°C and examined for inhibition halos after 24 and 48 h. Isolates that produced inhibition halos greater than 5 mm in diameter were considered as bacteria that elicited negative interactions towards E. coli

O157:H7.

Screening bacterial isolates for nutrient competition with E. coli O157:H7

Frozen cultures of bacteria isolated from spinach surfaces and the control of W. paucula were resuscitated by inoculation onto R2A media and incubated at 25°C for 48 h. An isolated colony was transferred to 1/10 TSB (one tenth strength) and incubated at 25°C/120 rpm until stationary phase was achieved. One milliliter of each culture were transferred to a sterile 1.5 mL tube and centrifuged at

14,000 rpm/4°C/2 min to collect cells. One milliliter of the spent media was collected and filter- sterilized using a sterile PVDF (0.20 m) syringe filter unit (Fisher-Scientific). An amount of 300 L of the filtered spent media was transferred to a 100 micro well plate (Bioscreen C™ Piscataway, NJ) and each well was inoculated with 1,000 cells of E. coli O157:H7 resuspended in 1X PBS. Cultures were incubated at 25°C for 48 h and optical density (420nm to 580nm) measurements made every 15 min until stationary phase was reached using an automated growth curve analysis system (Bioscreen C™

Piscataway, NJ). Each growth curve was repeated three times. Growth rates and final yield expressed in absorbance were determined (27). Growth rates and final yields of E. coli O157:H7 for each epiphyte filtrate were compared with growth of E. coli O157:H7 in fresh 1/10 TSB and the filtrates of spent media from growth of E. coli O157:H7 and W. paucula using t-test for mean separation. Isolates associated with growth rates and final yields significantly greater than or equal to those obtained in W. paucula spent media or fresh 1/10 TSB were considered as presumptive positive or neutral interactions) when p>0.05. Isolates whose filtrates resulted in significantly lower growth rates or final yields of E. coli

O157:H7 are considered presumptively as competitors for nutrients with E. coli O157:H7 (negative interaction) when p<0.05.

123

Preliminary positive interactions were considered for bacteria whose spent media resulted in an increased equal growth rate or yield, than that of E. coli O157:H7 grown on fresh 1/10 TSB. These bacteria were then co-cultured with a transformed E. coli O157:H7 (gfp+/kan+) to confirm microbe- microbe interactions in low nutrient media. Fifty milliliters of 1/10 TSB were inoculated with approximately 5 log10 cells of each selected microorganism. The cultures were preconditioned by incubation for 8 h at 25°C, to reach approximately a 2 log increase and after this period approximately10,000 cells of E. coli O157:H7 (gfp+/kan+) were added. E. coli O157:H7 populations were determined by plate counts on TSA supplemented with 50 g/mL of kanamycin after 24 h. Overall log increase of E. coli O157:H7 (gfp+/kan+) was calculated and compared to the population size of E. coli O157:H7 grown in single culture under the same conditions. Each co-culture was replicated three times. Growth of E. coli O157:H7 was compared with the growth in co-culture with the isolates using t- test for mean separation. Comparison among all the treatments was determined with proc mix function of Statistical Analysis System 9.2 (SAS Institute, Cary, NC) and a Tukey’s test comparison was applied for mean separation.

Identification of phylloepiphytes which interacted with E. coli O157:H7 in vitro

Bacteria that exhibited either positive or negative interactions with E. coli O157:H7 were identified based on similarity of its partial 16S rRNA gene sequence to sequences of known bacteria in the ribosomal database (8). DNA from each isolate was obtained from 1 mL of culture using the

Puregene® DNA purification kit (GENTRA systems, Minneapolis, MN, USA) per manufacturer’s instructions. The 16S rRNA gene was amplified from each isolate DNA (50 ng) to generate a fragment of approximately 1300 bp using the primers 27-f (5’-AGA GTT TGA TCC TGG CTC AG-3’) and 1392-r

(5’-ACG GGC GGT GTG TAC-3’). Each 25 L PCR reaction contained 1X of GoTaq® flexi buffer

(Promega), 1.5 mM of MgCl2, 0.2mM of each dinucleotide, 1% of DMSO (dimethylsulfoxide), 25mM of

Tris-HCl (pH 8), 1U of GoTaq® Hot Start polymerase (Promega) , 0.5 M of each primer, and template

DNA. The PCR protocol consisted of denaturation at 94oC for 10 minutes, followed by 30 cycles of 94oC

124 for 30 s, amplification at 57oC for 30 s, extension at 72oC for 1 min, and a final elongation step at 72oC for 10min. The size of PCR products were confirmed using a 0.9% agarose gel (Fisher-Scientific, Atlanta,

GA). Amplicons were cleaned using the Qiagen PCR purification kit and sequenced. Sequences were compared to sequences deposited within the National Center for Biotechnology Information (NCBI) using the basic local alignment search tool and within the Ribosomal Database Project (RDP) V10 and identities assigned based on similarity larger than 97% utilizing RDP Classifier (34).

Carbon source utilization of epiphytic bacteria

Carbon source utilization was determined using BIOLOG EcoPlateTM (Hayward, CA) for selected isolates that elicited presumptive positive, neutral or negative interactions with E. coli O157:H7 in the nutrient competition assays. Plates contained 31 carbon sources in individual wells and a negative control well repeated in duplicate on the plate. Cells were washed 6 times with 0.8% sterile saline solution and resuspended to T= 28%+/- 2% as recommended by manufacturer. Each well of the EcoPlatesTM were inoculated with 150 L of a culture and incubated at 25°C. Plates were read after 2 and 7 days for color production at a wavelength of 492 nm using a plate reader (Multiskan Ascent, Thermo Electron Co), when increase in absorbance was no longer detected. Carbon sources which showed an average absorbance change of 0.2 from the blank plate were considered as utilized. Two plates were prepared for each isolate. Patterns of carbon source utilization by the different isolates were compared by construction of a correlation matrix proc corr function of Statistical Analysis System 9.2 (SAS Institute, Cary, NC).

RESULTS

Identification of spinach epiphytic bacteria capable of inhibiting growth of E. coli O157:H7 in vitro.

Inhibition of E. coli O157:H7 growth was visualized by inoculation of the spinach isolates on a lawn of E. coli O157:H7 (Figure 1) and defined when a measurable clear zone surrounded the isolate.

Zones of inhibition of E. coli growth ranged between 5.0 and 17mm in diameter. A total of 1512 bacterial isolates were screened, and 69 colonies resulted in growth inhibition of E. coli O157:H7 (Table 1). The

125 greatest percentage of these isolates was classified as -Proteobacteria (83%), followed by Firmicutes

(7%), Bacteroidetes (5%), Actinobacteria (2%) and the rest and -Proteobacteria. From our randomly selected colonies the largest numbers of identified bacteria were Erwinia spp. and Pseudomonas spp.

Growth of E. coli O157:H7 in media previously utilized for growth of phylloepiphytic spinach isolates.

Growth rates and final yield of E. coli O157:H7 were compared for different media compositions:

1/10 TSB (control), media previously utilized to culture a single phylloepiphytic isolate (spent media), and media previously used to culture E. coli O157:H7. Prior growth of several different phylloepiphytic enhanced the growth of E. coli O157:H7 (Table 2).

Growth rates and yields of E. coli O157:H7 on 1/10 TSB media previously utilized to culture

Arthrobacter spp., Sphingomonas spp., Rhizobium spp., Microbacterium phyllospherae, Kaistia spp.,

Acidovorax konjaci, Curtobacterium herbarum, Flavobacterium spp., Brevibacillus brevis and the W. paucula control, were not significantly different (p>0.05) from the growth rate of E. coli O157:H7 on fresh 1/10 TSB. The yield of E. coli O157:H7 on Flavobacterium spp. spent media was significantly higher than the yield from the fresh 1/10 TSB control. These bacteria belonged to 12 genera with 22 bacterial isolates (Table 1), these bacteria were Actinobacteria (38%), Firmicutes (32%), -

Proteobacteria (13%), -Proteobacteria (13%), and Bacteroidetes (4%).

In contrast Brevibacillus brevis, Bacillus spp.,and Pseudomonas koreensis spent media resulted in significantly lower (p<0.05) growth rates of E. coli O157:H7 compared to the fresh 1/10 TSB control but these growth rates were not significantly different from the control of E. coli O157:H7 spent media.

The growth rate and yield of E. coli O157:H7 grown on E. persicina spent media was significantly lower compared to other isolates (Table 2).

126

Carbon source utilization by phylloepiphytic bacteria isolated from spinach.

Pearson’s correlation between carbon sources utilized by E. coli O157:H7 and select phylloepiphytic bacteria were calculated based on BIOLOG profiles (Table 3 and Supplementary table 1).

The greatest correlation (0.6) was calculated for E. coli O157:H7 and E. persicina which was also significant (p=0.0002). A positive correlation (between 0.3-0.4) was determined for Arthrobacter spp., C. herbarum, Flavobacterium spp. (B), Kaistia spp., M. phyllospherae, P. koreensis, Rhizobium spp. and

Rhodococcus spp. All the correlations were significant (p<0.05) with the exception of Flavobacterium spp. (B), P. koreensis. Lower correlations (0.2-0.1) were calculated for B. cereus, Sphingomonas spp., and the control W. paucula however none of those were statistically significant (p>0.05). Finally negative correlations were found between E. coli O157:H7 and Ac. konjaci, Bacillus spp., Br. Brevis and

Patulibacter spp.; however only the correlation with Bacillus spp. was significantly different (p<0.05).

Co-culture of E. coli O157:H7 with phylloepiphytic bacteria.

Selected phylloepiphytic bacteria whose spent media showed an increase in growth rate or yield of E. coli O157:H7 were co-cultured with E. coli O157:H7. After 24h the yield of E. coli O157:H7 were compared to a control of the pathogen alone (Table 4). The growth of E. coli O157:H7 in 1/10 TSB was not significantly different from its growth in co-culture with Ac. konjaci, C. herbarum, Flavobacterium spp. (B), Patulibacter spp., Rhizobium spp., Rhodococcus spp. and the control W. paucula. In contrast significantly lower numbers of E. coli O157:H7were obtained from co-culture with Arthrobacter spp., Br. brevis, M. phyllospherae and P. koreeensis.

DISCUSSION

Edible plants, particularly leafy greens, have become important vehicles for enteric pathogens, raising questions about how enteric bacterial pathogens can survive on edible plants, and how plants can become a habitat for these bacteria (4). When bacteria like E. coli O157:H7 land on edible plants, they will compete for resources in this niche with native members of the phyllosphere, which are well adapted

127 to that habitat (4). Although factors like nutrient limitation, low water activity, and harsh environmental conditions can challenge the establishment of enteric pathogens on the phyllosphere (22), they do survive and may colonize the leaf surface. It is likely that interactions with the native microorganisms have a role in the persistence and establishment of pathogenic bacteria. In this study, in vitro interactions of E. coli

O157:H7 with the phylloepiphytic bacteria isolated from spinach leaves were studied under conditions of low nutrients. The objectives were to identify bacteria that can potentially inhibit (negative interactions) or favor (positive interaction) the growth of E. coli O157:H7.

The phyllosphere microbial community is dominated by members of the phyla Proteobacteria, with -Proteobacteria the most abundant class of bacteria (15, 18, 28, 35) (Table 1). Therefore it is not unanticipated that large number of the isolates which inhibited the growth of E. coli O157:H7 in vitro belonged to this class. The closely related members of the -Proteobacteria are similar in their physiological responses and carbon sources metabolized, suggesting the likelihood of competitive exclusion of E. coli O157:H7 due to nutrient composition (20).

P. koreensis, E. persicina and Bacillus spp. were selected to determine the possible mechanisms that allow these bacteria to inhibit the growth of E. coli O157:H7. E. coli O157:H7 grown on spent media of P. koreensis and E. persicina showed significantly lower specific growth rates (p<0.05) and yields compared to E. coli O157:H7 grown on fresh media (Table 2). In the particular case of E. persicina, it reduced the specific growth and final yield of E. coli O157:H7. This effect was reversed when 80% of the media was replenished with fresh 1/10 TSB, indicating that growth reduction was due to exhaustion of nutrients and not due to soluble inhibitors (data not shown), which was also supported with a positive and significant correlation of nutrient utilization with E. coli O157:H7 (Table 4). Nutrient depletion has been previously reported as a mechanism used by P. fluorescens and Enterobacter asburiae, to prevent growth of E. coli O157:H7 (9, 26). When these bacteria were grown in co-culture with E. coli O157:H7, there was a significant reduction in its population, compared to E. coli O156:H7 grown alone (Table 4).

128

Significant reduction in specific growth rate and yield reduction of E. coli O157:H7 were calculated when media previously used to culture Bacillus spp. and Br. brevis were utilized (Table 2). The carbon source utilization showed a significant negative correlation or no correlation with carbon sources utilized by E. coli O157:H7 (Table 3). This suggests these phylloepiphytes likely produced diffusible molecules that can reduce the growth of E. coli O157:H7. Besides competitive exclusion, diffusion of antibiotic molecules can also be used to outcompete other bacteria; several Firmicutes like Bacillus sp., which was also isolated within the group of negative interactions (Table 1), and Lactobacillus sp. are able to produce bacteriocins that inhibit bacterial growth (7). Stenotrophomonas spp., ubiquitous on the phyllosphere, also produces antibiotic molecules (13). Diffusible molecules with inhibitory effects could have potential biotechnological application in the bio-control of enteric pathogens during pre- and post- harvest operations.

Other isolates that also elicited a negative interaction were Pantoea agglomerans, Aeromonas spp., and Flavobacterium sp. (A) (Table 1), which have been previously isolated from other produce

(green peppers, celery, lettuce, green onions and cabbage) (20, 31), and they have shown to suppress the growth of human pathogens E. coli O157:H7, Listeria monocytogenes and Salmonella enterica (31) or plant pathogens like the fungi Rhizoctonia solani (29, 37). Other isolates belonging to the genera

Stenotrophomonas spp. Hafnia sp., and Brevundimonas sp. had not been previously reported to elicit any antagonistic effect towards enteric pathogens.

Similar to intrinsic characteristics that provide enteric pathogens, the ability to survive and establish on the phyllosphere, it is likely that interactions with already established members of this community can provide epiphytic fitness (4). In the phyllosphere, several studies have demonstrated the importance of bacterial interactions in the establishment of bacterial populations. For example

Corynebacterium sp. a N2 fixator in the phyllosphere allows P. putida to obtain reduced nitrogen for its growth (32). Siderophores produced in the rhizosphere, can also enhance Pseudomonas sp. growth (24).

129

Other indirect effects rely in the activity of certain microorganisms to alter plant surfaces and increase permeability, which could be beneficial for other members of the phyllosphere (19, 30).

It was demonstrated that W. paucula, an epiphyte isolated from lettuce and Arabidopsis thaliana, had the ability to enhance the growth of E. coli O157:H7 on sterile lettuce plants, likely associated with lack of competition for nutrients (9). Increases in specific growth rate and yield of E. coli O157:H7 on spent media produced by phylloepiphytic bacteria were considered possible commensalism interactions

(Table 2), no significant difference were found in the growth rates and yields of E. coli O157:H7 grown on media previously utilized by phylloepiphytic bacteria and the control with fresh media. Nine genera of bacterial isolates from the spinach phyllosphere that elicited this effect were classified as Actinobacteria

(38%) (Table1). Unlike bacteria that elicited negative interactions, the majority of these bacteria belonged to different taxonomical classes. Studies on the bean phyllosphere demonstrated that the level of co- existence of epiphytic bacteria was impacted with the ability to utilize carbon sources not utilized by a competing strain (36), even more the level of co-existence was also related with the ecological similarity of the epiphytes (36). This was supported by the comparison of carbon source utilization by E. coli

O157:H7 and phylloepiphytic bacteria. Lower correlations between each epiphyte and E. coli O157:H7 were calculated than correlation with E. persicina which was demonstrated to be a competitor. When grown in co-culture with the phylloepiphytic isolates, the populations of E. coli O157:H7 were not significantly different to E. coli O157:H7 growing alone (Table 4), which supports the lack of competition with these phylloepiphytic bacteria, even when the media has been pre-conditioned with the phylloepiphytic isolate . Although W. paucula, has been reported to enhance the growth of E. coli

O157:H7 on lettuce surfaces, an increase in E .coli O157:H7 numbers were not seen in this experiment when co-cultured in vitro (9). The exceptions were Arthrobacter spp. and M. phyllospherae which produced a significantly lower increase in the population of E. coli O157:H7, however it was observed that the population of these bacteria during the 8 h of pre-conditioning rapidly increase from 105 to 107

(data not shown), which could outcompete better towards a lower concentration of E. coli O157:H7 in the

130 growth media. These epiphytes also showed a significant positive correlation in carbon sources utilized for E. coli O157:H7. Finally, it is important to point out that the rest of the screened isolates that did elicit neither an negative effect nor a growth rate increase associated with spent media on E. coli

O157:H7, they elicited a similar effect as that observed when E. coli O157:H7 was grown on their own spent media (Table 2), which also indicates that this bacterium can compete with a large part of the phyllosphere.

Enteric pathogens on the leaf surface do not interact with a single bacterial population, but rather with a diverse microbial community. How the sum of these individual interactions (positive, neutral or negative) affect the overall fate of enteric pathogens on edible plants remains to be elucidated. Both negative and positive interactions influenced the growth in vitro of E. coli O157:H7 showing the importance of the bacterial community in the establishment of this bacterium. If certainly, negative interaction may provide means of bio-control of enteric pathogens on edible plants, positive interactions can also provide a better understanding of the ability of E. coli O157:H7 to successfully compete in the phyllosphere.

ACKNOWLEDGEMENTS

We acknowledge the assistance of Brinkley Benson for his contribution in the cultivation of spinach. Michael Cooley provided W. paucula strain and E. coli O157:H7 strain isolated from the spinach outbreak in 2006 was provided by Center for Disease Control and Prevention. Fellowship for partial financial support of Gabriela Lopez-Velasco was provided by the National Council for Science and

Technology (CONACyT) from Mexico. Support for Heather Tydings was provided in part by a Fralin

Institute Summer Undergraduate Research Fellowship.

131

REFERENCES

1. Aruscavage, D., K. Lee, S. Miller, and J. LeJeune. 2006. Interaction affecting the proliferation

and control of human pathogens on edible plants. J Food Sci 71:89-99.

2. Beutin, L., D. Geier, H. Steinruck, S. Zimmermann, and F. Scheutz. 1993. Prevalence and

some properties of verotoxin (Shiga-like toxin)-producing Escherichia coli in seven different

species of healthy domestic animals. J Clin Microbiol 31:2483-8.

3. Bhagwat, A. A. 2006. Microbiological safety of fresh-cut produce: Where are we now?, p. 121-

165. In K. J. Matthews (ed.), Microbiology of fresh produce. ASM Press, Washington, DC.

4. Brandl, M. T. 2006. Fitness of human enteric pathogens on plants and implications for food

safety. Annu Rev Phytopathol 44:367-92.

5. Calvin, L. 2007. Outbreak linked to spinach forces reassessment of food safety practices.

AMBER WAVES 5:24-31.

6. Carey, C. M., M. Kostrzynska, and S. Thompson. 2009. Escherichia coli O157:H7 stress and

virulence gene expression on Romaine lettuce using comparative real-time PCR. J Microbiol

Methods 77:235-42.

7. Chen, H., and D. Hoover. 2003. Bacteriocins and their food applications. Comprehansive

reviews in food science and food safety 2:82-100.

8. Cole, J. R., Q. Wang, E. Cardenas, J. Fish, B. Chai, R. J. Farris, A. S. Kulam-Syed-

Mohideen, D. M. McGarrell, T. Marsh, G. M. Garrity, and J. M. Tiedje. 2009. The

Ribosomal Database Project: improved alignments and new tools for rRNA analysis. Nucleic

Acids Res 37:141-5.

9. Cooley, M. B., D. Chao, and R. E. Mandrell. 2006. Escherichia coli O157:H7 survival and

growth on lettuce is altered by the presence of epiphytic bacteria. J Food Prot 69:2329-35.

10. Fleming, H. P., J. L. Etchells, and R. N. Costilow. 1975. Microbial Inhibition by an Isolate of

Pediococcus from Cucumber Brines. Appl Microbiol 30:1040-1042.

132

11. Fredrickson, A. G. 1977. Behavior of mixed cultures of microorganisms. Annu Rev Microbiol

31:63-87.

12. Hancock, D. D., T. E. Besser, D. H. Rice, E. D. Ebel, D. E. Herriott, and L. V. Carpenter.

1998. Multiple sources of Escherichia coli O157 in feedlots and dairy farms in the northwestern

USA. Prev Vet Med 35:11-9.

13. Hayward, A. C., N. Fegan, M. Fegan, and G. R. Stirling. 2009. Stenotrophomonas and

Lysobacter: ubiquitous plant-associated gamma-proteobacteria of developing significance in

applied microbiology. J Appl Microbiol.

14. Heaton, J. C., and K. Jones. 2008. Microbial contamination of fruit and vegetables and the

behaviour of enteropathogens in the phyllosphere: a review. J Appl Microbiol 104:613-26.

15. Idris, R., R. Trifonova, M. Puschenreiter, W. W. Wenzel, and A. Sessitsch. 2004. Bacterial

communities associated with flowering plants of the Ni hyperaccumulator Thlaspi goesingense.

Appl Environ Microbiol 70:2667-77.

16. Islam, M., M. P. Doyle, S. C. Phatak, P. Millner, and X. Jiang. 2004. Persistence of

enterohemorrhagic Escherichia coli O157:H7 in soil and on leaf lettuce and parsley grown in

fields treated with contaminated manure composts or irrigation water. J Food Prot 67:1365-70.

17. Johnston, M. A., M. A. Harrison, and R. A. Morrow. 2009. Microbial antagonists of

Escherichia coli O157:H7 on fresh-cut lettuce and spinach. J Food Prot 72:1569-1575.

18. Kadivar, H., and A. E. Stapleton. 2003. Ultraviolet radiation alters maize phyllosphere bacterial

diversity. Microb Ecol 45:353-61.

19. Krimm, U., D. Abanda-Nkpwatt, W. Schwab, and L. Schreiber. 2005. Epiphytic

microorganisms on strawberry plants (Fragaria ananassa cv. Elsanta): identification of bacterial

isolates and analysis of their interaction with leaf surfaces. FEMS Microbiol Ecol 53:483-92.

20. Lamb, E. G., and J. F. Cahill, Jr. 2008. When competition does not matter: grassland diversity

and community composition. Am Nat 171:777-87.

133

21. Liao, C. H., and W. F. Fett. 2001. Analysis of native microflora and selection of strains

antagonistic to human pathogens on fresh produce. J Food Prot 64:1110-5.

22. Lindow, S. E., and M. T. Brandl. 2003. Microbiology of the phyllosphere. Appl Environ

Microbiol 69:1875-83.

23. Little, A. E., C. J. Robinson, S. B. Peterson, K. F. Raffa, and J. Handelsman. 2008. Rules of

engagement: interspecies interactions that regulate microbial communities. Annu Rev Microbiol

62:375-401.

24. Loper, J. E., and M. D. Henkels. 1999. Utilization of heterologous siderophores enhances levels

of iron available to Pseudomonas putida in the rhizosphere. Appl Environ Microbiol 65:5357-63.

25. Lopez-Velasco, G., M. Davis, R. R. Boyer, R. C. Williams, and M. A. Ponder. 2010.

Alterations of the phylloepiphytic bacterial community associated with interactions of

Escherichia coli O157:H7 during storage of packaged spinach at refrigeration temperatures. Food

Microbiol In Press.

26. McKellar, R. C. 2007. Role of nutrient limitation in the competition between Pseudomonas

fluorescens and Escherichia coli O157: H7. J Food Prot 70:1739-43.

27. Metris, A., S. M. George, and J. Baranyi. 2006. Use of optical density detection times to assess

the effect of acetic acid on single-cell kinetics. Appl Environ Microbiol 72:6674-9.

28. Rasche, F., R. Trondl, C. Naglreiter, T. G. Reichenauer, and A. Sessitsch. 2006. Chilling and

cultivar type affect the diversity of bacterial endophytes colonizing sweet pepper (Capsicum

anuum L.). Can J Microbiol 52:1036-45.

29. Scherwinski, K., R. Grosch, and G. Berg. 2008. Effect of bacterial antagonists on lettuce:

active biocontrol of Rhizoctonia solani and negligible, short-term effects on nontarget

microorganisms. FEMS Microbiol Ecol 64:106-16.

30. Schreiber, L., U. Krimm, D. Knoll, M. Sayed, G. Auling, and R. M. Kroppenstedt. 2005.

Plant-microbe interactions: identification of epiphytic bacteria and their ability to alter leaf

surface permeability. New Phytol 166:589-94.

134

31. Schuenzel, K. M., and M. A. Harrison. 2002. Microbial antagonists of foodborne pathogens on

fresh, minimally processed vegetables. J Food Prot 65:1909-15.

32. Sieuwerts, S., F. A. de Bok, J. Hugenholtz, and J. E. van Hylckama Vlieg. 2008. Unraveling

microbial interactions in food fermentations: from classical to genomics approaches. Appl

Environ Microbiol 74:4997-5007.

33. Solomon, E. B., S. Yaron, and K. R. Matthews. 2002. Transmission of Escherichia coli

O157:H7 from contaminated manure and irrigation water to lettuce plant tissue and its subsequent

internalization. Appl Environ Microbiol 68:397-400.

34. Wang, Q., G. M. Garrity, J. M. Tiedje, and J. R. Cole. 2007. Naive Bayesian classifier for

rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol

73:5261-7.

35. Whipps, J. M., P. Hand, D. Pink, and G. D. Bending. 2008. Phyllosphere microbiology with

special reference to diversity and plant genotype. J Appl Microbiol 105:1744-55.

36. Wilson, M., and S. E. Lindow. 1994. Coexistence among epiphytic bacterial populations

mediated through nutritional resource partitioning. Appl Environ Microbiol 60:4468-4477.

37. Zachow, C., R. Tilcher, and G. Berg. 2008. Sugar beet-associated bacterial and fungal

communities show a high indigenous antagonistic potential against plant pathogens. Microb Ecol

55:119-29.

135

Table 1. Identified microorganisms that elicited in vitro interactions with E. coli O157:H7.

Identified bacteria GeneBank % of Interaction3 (number of isolates)1 accession no. of similarity closest match2 Acidovorax konjaci (1) GQ260128 97 (+) Acinetobacter calcoaceticus (4) FJ867364 96 (-) Aeromonas encheleia (1) AJ458416 97 (-) Arthrobacter spp. (1) GQ332346 97 (+) Bacillus cereus (5) FJ763650 97 (-) Bacillus pumilus (3) FJ263042 98 (-) Bacillus spp.(1) EU781520 97 (-) Brevibacillus brevis (1) EU931557 96 (+) Brevundimonas vesicularis (1) FJ999941 98 (-) Curtobacterium herbarum (3) AM410692 97 (+) Erwinia persicina (24) AJ937838 98 (-) Erwinia rhapontici (2) EU340562 98 (-) Flavobacterium spp. (5) –A-4 EF601822 100 (-) Flavobacterium spp. –B-4 AM110987 98 (+) Frigobacterium spp. (1) AF157479 98 (-) Kaistia spp. (1) FJ006913 99 (+) Microbacterium oleivorans (2) EU71438 99 (-) Microbacterium phyllospherae (2) NR025405 97 (+) Paenibacillus spp. (1) FJ940900 100 (-) Pantoea agglomerans (5) EF050808 99 (-) Patulibacter spp.(1) AJ871305 100 (+) Pseudomonas fluorescens (2) CP000094 97 (-) Pseudomonas koreensis (8) GQ368179 96 (-) Pseudomonas rhodesiae (1) FJ462694 96 (-) Pseudomonas spp. (12) FN547413 100 (-) Rhizobium spp. (1) FJ006911 98 (+) Rhodococcus spp. (1) AY188941 97 (+) Sphingomonas spp. (1) AF395031 98 (+) Stenotrophomonas maltophilia (3) GU385870 97 (-) Unclassified -Proteobacteria (6) (-) 1 Bacterial species isolated from spinach and identified using partial 16S rRNA (approximately 1300 bp).

2 National Center for Biotechnology Information database as on March 3, 2010.

3 Bacteria selected for inhibiting E. coli O157:H7growth in lawn inoculation experiments are referred as (-). Bacteria whose spent media enhanced E. coli O157:H7 growth rate and/or yield are referred as (+).

4 Flavobacterium spp. sequences A and B were annealed and had a 95% similarity between them.

136

Table 2. Growth rate and yield of E. coli O157:H7 cultures grown in vitro at 25°C using media previously used to culture phylloepiphytic spinach bacterial isolates.

Spent media produced by Specific growth rate Final yield (OD)1,2 ( ) of E. coli O157:H7 (h-1)1 Control fresh 1/10 TSB 0.117 + 0.004A 0.603 + 0.002B

E. coli O157:H7 0.066 + 0.003C 0.322 + 0.006D

W. paucula 0.104 + 0.005A 0.553 + 0.043C

A B Acidovorax konjaci 0.110 + 0.002 0.648 + 0.021

A C Arthrobacter spp. 0.108 + 0.004 0.527 + 0.026

C C Bacillus spp. 0.063 + 0.004 0.528 + 0.007

B C Brevibacillus brevis 0.095 + 0.007 0.513 + 0.023

A C Curtobacterium herbarum. 0.117 + 0.004 0.550 + 0.022

D E Erwinia persicina 0.042 + 0.007 0.338 + 0.019

A A Flavobacterium spp.-B- 0.123 + 0.004 0.757+ 0.061

A C Kaistia spp. 0.129 + 0.004 0.566 + 0.048

Microbacterium phyllospherae 0.108 + 0.007A 0.510 + 0.012C

Patulibacter spp. 0.111 + 0.004A 0.447 + 0.045D

Pseudomonas koreensis 0.075 + 0.006C 0.469 + 0.009D

A B Rhizobium spp. 0.109 + 0.007 0.606 + 0.003

Rhodococcus spp. 0.109 + 0.006A 0.569 + 0.023C

Sphingomonas spp. 0.117 + 0.005A 0.618 + 0.025B

1 Results represent the mean + standard deviation (n=3). Different letters within the same column denote significant differences assessed by t-test with the control of fresh 1/10 TSB (p<0.05).

2 Measurement of optical density (OD) with the wide band (420 to 580) when stationary phase was reached.

137

Table 3. Pearson’s correlation between E. coli O157:H7 and phylloepiphytic isolates in carbon source utilization.

Bacteria Pearson’s p-value of correlation

correlation (r)1

E. coli O157:H7 1.00 -

W. paucula 0.163 0.379

Acidovorax konjaci -0.006 0.974

Arthrobacter spp. 0.384 0.033

Bacillus spp. -0.008 0.038

Bacillus cereus 0.121 0.997

Brevibacillus brevis -0.074 0.695

Curtobacterium herbarum 0.327 0.013

Erwinia persicina 0.615 0.0002

Flavobacterium spp.-B- 0.329 0.070

Kaistia spp. 0.356 0.049

Microbacterium phyllospherae 0.403 0.025

Patulibacter spp. -0.127 0.497

Pseudomonas koreensis 0.305 0.095

Rhizobium spp. 0.488 0.005

Rhodococcus spp. 0.418 0.019

Sphingomonas spp. 0.265 0.148

1 Pearson correlation was calculated for all isolates and E. coli O157:H7, the data presented represents only the correlation of the pattern of carbon source utilization by E. coli O157:H7 with each of the bacterial isolates. Correlation was considered significant when p<0.05.

138

Table 4. Co-culture of E. coli O157:H7 with bacterial isolates from spinach phyllosphere.

E. coli O157:H7 (log10 CFU/mL)1

Isolate tested Alone In co-culture2

E. coli O157:H7 11.80 + 0.06A --

W. paucula 11.65 + 0.18A

Acidovorax spp. 11.83 + 0.07A

Arthrobacter spp. 8.80 + 0.09D

Br. brevis 5.90 + 0.18E

C. herbarum 11.49 + 0.10A

Flavobacterium spp. 11.58 + 0.09A

M. phyllospherae 10.73 + 0.02B

Patulibacter spp. 11.16 + 0.06A

P. koreensis 7.63 + 0.21C

Rhizobium spp. 11.61 + 0.10A

Rhodococcus spp. 11.84 + 0.06A

1 Plate counts (mean + standard deviation) were performed after 24 h of incubation in 1/10 TSB on preconditioned media (8 h) with the phylloepiphytic isolates at 25°C.

2 Different letters denote significant difference (p<0.05) with respect to the control of E. coli O157:H7 grown on fresh 1/10 TSB.

139

Supplementary Table 1. Values of optical density of different carbon sources utilized by E. coli

O157:H7 and phylloepiphytic bacteria from BIOLOG ® plates.

Carbon source E. coli E. persicina P. koreensins Pseudomonas O157:H7 spp. Pyruvic acid methyl ester 249.8 (2.2) 178.9 (0.5) 132.4 (1.7) 205.3 (10.5) Tween 40 99.4 (7.4) 182.8 (2.0) 133.3 (6.8) 170.7 (7.3) Tween 80 103 (7.7) 178.6 (2.0) 121.4 (6.9) 142.8 (2.5) a-Cyclodextrin 76 (2.6) 85.8 (15.6) 58.3 (1.0) 83.0 (6.7) Glycogen 93.1 (0.35) 115.7 (11.5) 79.6 (4.1) 92.9 (10.7) D-Cellobiose 77.4 (2.0) 305.3 (2.7) 66.3 (0.7) 80.3 (7.3) a-D-Lactose 277.1 (1.0) 196.7 (1.0) 66.1 (1.0) 82.3 (7.7) b-Methyl-D- Glucoside 248.2 (9.3) 267.3 (6.25) 65.0 (0.3) 83.2 (2.2) D-Xylose 272.2 (1.0) 168.8 (1.8) 144.8 (15.4) 164.3 (17.8) i-Erythritol 65.4 (1.0) 114.1 (6.5) 63.1 (2.0) 87.4 (4.2) D-Mannitol 234.2 (3.2) 307.2 (3.8) 149.3 (4.3) 87.9 (4.6) N-Acetyl-D- Glucosamine 249.2 (2.3) 285.8 (9.9) 126.4 (3.2) 83.9 (1.6) D-Glucosaminic Acid 72.9 (0.5) 94.4 (2.6) 130.5 (0.6) 81.9 (4.8) Glucose-1-Phosphate 252.9 (4.3) 262.9 (0.9) 69.8 (4.0) 85.7 (6.1) D,L-a-Glycerol phosphate 247.1 (6.7) 229.6 (10.3) 103.9 (12.0) 81.3 (5.3) D-Galactonic Acid g- lactone 70.7 (1.0) 85.0 (1.9) 111.7 (5.6) 82.2 (5.8) D-galacturonic acid 241.8 (7.7) 270.1 (0.2) 69.0 (0.4) 84.2 (5.0) 2-Hydroxy benzoic acid 60 (1.2) 50.3 (1.1) 39.6 (5.4) 68.9 (1.2) g-Hydroxybutyric acid 73.2 (0.6) 104.0 (3.0) 86.0 (5.9) 79.7 (1.6) 4-Hydroxi benzoic acid 90.7 (2.8) 86.2 (9.9) 74 (6.7) 83.8 (11.4) Itaconic acid 73.2 (0.5) 77.3 (14.0) 53.5 (6.0) 270.0 (2.7) a-ketobutyric acid 208.3 (2.0) 85.2 (10.0) 60.6 (3.5) 130.2 (11.9) D-Malic acid 282.5 (1.2) 121.1 (1.0) 82.0 (10.6) 152.6 (9.0) L-Arginine 76.8 (0.5) 98.1 (1.0) 114.9 (4.5) 226.7 (2.8) L-Aspargine 210.7 (2.15) 305.8 (19.8) 162 (6.0) 289.6 (9.7) L-Phenylalanine 73.5 (0.2) 83.3 (5.7) 91.6 (12.8) 113.7 (3.8) L-Serine 250.5 (9.5) 274.4 (1.6) 160.1 (2.6) 313 (6.2) L-Threonine 245.7 (2.0) 102.6 (5.5) 90.7 (6.3) 157 (8.8) Glycil-L-glutamic acid 260.3 (4.4) 224 (1.0) 147.1 (5.8) 138.1 (14.0) Phenylethylamine 74 (0.2) 83.9 (15.8) 72.8 (32.2) 93.6 (13.5) Putrescine 83.6 (1.5) 86.2 (16.0) 115.3 (3.7) 242.5 (17.3)

Results are the mean of OD492 x 100 (standard deviation of 2 plates), a negative value (-) was considered when OD492 <0.2

140

Supplementary Table 1 (continuation). Values of optical density of different carbon sources utilized by

E. coli O157:H7 and phylloepiphytic bacteria from BIOLOG ® plates.

Carbon source E. coli B. cereus Bacillus spp. Br. brevis O157:H7 Pyruvic acid methyl ester 249.8 (2.2) 101.4 (6.0) 38.2 (1.2) 30.6 (1.3) Tween 40 99.4 (7.4) 79.3 (13.0)) 37.0 (3.7) 46.4 (6.2) Tween 80 103 (7.7) 78.3 (6.0) 35.9 (1.5) 31.7 (16) a-Cyclodextrin 76 (2.6) 80.1 (3.5) 36.1 (1.4) 29.6 (1.0) Glycogen 93.1 (0.35) 171.5 (6.1) 37.1 (1.1) 35.1 (1.0) D-Cellobiose 77.4 (2.0) 205.9 (7.5) 35.9 (1.0) 29.9 (1.0) a-D-Lactose 277.1 (1.0) 30.1 (24.7) 33.3 (0.4) 25.6 (0.2) b-Methyl-D- Glucoside 248.2 (9.3) 176.8 (3.6) 33.3 (0.4) 25.7 (0.5) D-Xylose 272.2 (1.0) 86.2 (1.9) 57.3 (2.2) 47.4 (1.3) i-Erythritol 65.4 (1.0) 33.1 (1.0) 34.2 (1.6) 26.3 (1.3) D-Mannitol 234.2 (3.2) 200.4 (1.1) 35.4 (1.6) 25.6 ( 0.7) N-Acetyl-D- Glucosamine 249.2 (2.3) 186.1 (7.2) 35 (3.6) 34.2 (0.7) D-Glucosaminic Acid 72.9 (0.5) 65.2 (1.0) 34.0 (1.4) 25.7 (0.6) Glucose-1-Phosphate 252.9 (4.3) 145.4 (3.2) 34.7 (1.6) 27.6 (1.5) D,L-a-Glycerol phosphate 247.1 (6.7) 152.4 (1.6) 34.6 (0.3) 25.3 (0.2) D-Galactonic Acid g- lactone 70.7 (1.0) 45.4 (1.1) 35.9 (0.1) 27.9 (0.1) D-galacturonic acid 241.8 (7.7) 160.8 (2.4) (-) 29.1 (1.2) 2-Hydroxy benzoic acid 60 (1.2) 26.8 (0.4) 32.6 (1.9) 22.2 (1.4) g-Hydroxybutyric acid 73.2 (0.6) 56.4 (2.2) 30.6 ( 1.7) 24.1 (1.7) 4-Hydroxi benzoic acid 90.7 (2.8) 68.9 (2.8) 41.6 (1.3) 36.2 (5.0) Itaconic acid 73.2 (0.5) 40.9 (0.3) 34.3 ( 0.1) 26.4 (0.0) a-ketobutyric acid 208.3 (2.0) 31.9 (0.3) 35.4 ( 1.2) 25.7 (0.2) D-Malic acid 282.5 (1.2) 45.8 (0.25) 33 (0.5) (-) L-Arginine 76.8 (0.5) 65.1 (13.4) 31.2 ( 2) 23.8 (0.1) L-Aspargine 210.7 (2.15) 171.5 (3.4) 34.6 ( 1.2) 26.9 (1.3) L-Phenylalanine 73.5 (0.2) 46.3 (0.1) 35.5 ( 1.8) 24.5 (0.6) L-Serine 250.5 (9.5) 122.3 (8.1) 39.9 ( 3.0) 30.4 (0.2) L-Threonine 245.7 (2.0) 47.5 (1.7) 36.5 ( 2.0) 26.8 (1.4) Glycil-L-glutamic acid 260.3 (4.4) 94.3 (4.4) 35.7 ( 1.2) 27.3 (0.6) Phenylethylamine 74 (0.2) 44.9 (1.4) 36.1 ( 1.0) 24.4 (0.2) Putrescine 83.6 (1.5) 42.7 (1.5) 34.8 (0.1) 25.7 (0.6)

Results are the mean of OD492 x 100 (standard deviation of 2 plates), a negative value (-) was considered when OD492 <0.2

141

Supplementary Table 1 (continuation). Values of optical density of different carbon sources utilized by

E. coli O157:H7 and phylloepiphytic bacteria from BIOLOG ® plates.

Carbon source E. coli W. paucula Flavobacterium Acidovorax O157:H7 spp. (B) spp. Pyruvic acid methyl ester 249.8 (2.2) 199.8 (6.0) 64.7 (1.1) 52.8 (9.5) Tween 40 99.4 (7.4) 145.8 (4.7) 33 (1.7) 88.6 (10.8) Tween 80 103 (7.7) 114.9 (2.5) (-) 76.9 (4.5) a-Cyclodextrin 76 (2.6) 56.6 (1.7) 63.1 (3.4) 24.4 (3.8) Glycogen 93.1 (0.35) 61.0 (1.2) 45.8 (0.5) 23.4 (2.2) D-Cellobiose 77.4 (2.0) 54.1 (2.8) 100.8 (10.2) (-) a-D-Lactose 277.1 (1.0) 53.7 (1.3) 63.6 (0.7) (-) b-Methyl-D- Glucoside 248.2 (9.3) 53.6 (4.0) 42.8 (0.2) (-) D-Xylose 272.2 (1.0) 83.0 (2.8) 56.3 (1.4) 33.4 (2.7) i-Erythritol 65.4 (1.0) 61.1 (1.3) 40.7 (0.2) (-) D-Mannitol 234.2 (3.2) 61.5 (1.1) 45.2 (4.0) (-) N-Acetyl-D- Glucosamine 249.2 (2.3) 57.7 (3.2) 124.6 (2.3) (-) D-Glucosaminic Acid 72.9 (0.5) 65.4 (3.1) 42.7 (3.3) 25.3 (2.0) Glucose-1-Phosphate 252.9 (4.3) 54.5 (3.4) 53.8 (3.0) (-) D,L-a-Glycerol phosphate 247.1 (6.7) 59.6 (2.2) 46.1 (6.0) 27.8 (0.1) D-Galactonic Acid g- lactone 70.7 (1.0) 56.1 (2.8) 40.8 (0.4) (-) D-galacturonic acid 241.8 (7.7) (-) 66.6 (1.4) (-) 2-Hydroxy benzoic acid 60 (1.2) 52.3 (3.2) 31.3 (0.7) (-) g-Hydroxybutyric acid 73.2 (0.6) 53.7 (3.5) 42.1 (0.1) 24.2 (1.9) 4-Hydroxi benzoic acid 90.7 (2.8) 140.6 (4.3) 47.5 (2.3) (-) Itaconic acid 73.2 (0.5) 87.8 (2.0) 42.8 (2.2) 52.4 (5.3) a-ketobutyric acid 208.3 (2.0) 79.7 (4.0) 51.2 (1.3) 33.6 (11.0) D-Malic acid 282.5 (1.2) 96.5 (6.5) 47.5 (5.5) 131.7 (3.6) L-Arginine 76.8 (0.5) 53.7 (0.1) 57.4 (1.8) 22.2 (4.8) L-Aspargine 210.7 (2.15) 147.7 (3.2) 59.7 (1.2) 104.1 (10.0) L-Phenylalanine 73.5 (0.2) 102.6 (0.4) 40.4 (1.7) 45.1 (7.9) L-Serine 250.5 (9.5) 161.1 (6.0) 54.1 (4.4) (-) L-Threonine 245.7 (2.0) 94.5 (10.1) 42.4 (3.5) 30.6 (2.3) Glycil-L-glutamic acid 260.3 (4.4) 81.6 (5.0) 65.6 (1.8) (-) Phenylethylamine 74 (0.2) 55.8 (2.8) 39.8 (0.2) (-) Putrescine 83.6 (1.5) 55.6 (0.4) 43.7 (2.4) (-)

Results are the mean of OD492 x 100 (standard deviation of 2 plates), a negative value (-) was considered when OD492 <0.2

142

Supplementary Table 1 (continuation). Values of optical density of different carbon sources utilized by

E. coli O157:H7 and phylloepiphytic bacteria from BIOLOG ® plates.

Carbon source E. coli C. herbarum Microbacterium Sphingomonas O157:H7 phyllospherae spp. Pyruvic acid methyl ester 249.8 (2.2) 59.5 (4.5) 39.8 (5.1) 49.0 (5.1) Tween 40 99.4 (7.4) 65.4 (4.5) 31.0 (6.8) 61.3 (7.0) Tween 80 103 (7.7) 51.3 (6.3) 30.7 (3.9) 27.1 (0.5) a-Cyclodextrin 76 (2.6) 50.8 (5.8) 23.1 (2.3) (-) Glycogen 93.1 (0.35) 50.0 (5.0) 31.3 (5.8) 25.4 (6.2) D-Cellobiose 77.4 (2.0) 51.1 (3.4) 32.7 (7.7) 40.5 (5.6) a-D-Lactose 277.1 (1.0) 57.6 (5.0) 28.0 (5.7) 26.1(1.3) b-Methyl-D- Glucoside 248.2 (9.3) 50.4 (3.8) 31.5 (3.1) 27.3 (0.4) D-Xylose 272.2 (1.0) 82.6 (0.1) 57.2 (5.1) 70.1 (2.5) i-Erythritol 65.4 (1.0) 49.8 (5.5) 24.1 (7.1) 26.8 (2.5) D-Mannitol 234.2 (3.2) 54.6 (2.6) 48.3 (20.0) 26.2 (2.0) N-Acetyl-D- Glucosamine 249.2 (2.3) 53.6 (5.7) 27.1 (6.0) 34.6 (1.1) D-Glucosaminic Acid 72.9 (0.5) 47.1 (3.7) 27.9 (5.5) (-) Glucose-1-Phosphate 252.9 (4.3) 54.2 (4.0) 27.8 ( 5.2) (-) D,L-a-Glycerol phosphate 247.1 (6.7) 45.9 (1.4) 28.9 (7.3) (-) D-Galactonic Acid g- lactone 70.7 (1.0) 49.1 (3.1) 24.5 (2.2) 27.1 (1.0) D-galacturonic acid 241.8 (7.7) 53.2 (0.4) 31.2 (4.1) 23.3 (0.3) 2-Hydroxy benzoic acid 60 (1.2) 44.1 (3.1) 28.3 ( 3.5) (-) g-Hydroxybutyric acid 73.2 (0.6) 48.7 (3.8) 28.2 ( 4.3) (-) 4-Hydroxi benzoic acid 90.7 (2.8) 59.6 (8.5) 37.0 (13) 27.9 (8.1) Itaconic acid 73.2 (0.5) 51.1 (3.5) 28.8 (3.4) (-) a-ketobutyric acid 208.3 (2.0) 57.3 (5.1) 37.6 (5.9) 24.9 (0.7) D-Malic acid 282.5 (1.2) 49.4 (3.3) 33.5 (5.9) (-) L-Arginine 76.8 (0.5) 48.6 (4.2) 31.2 (1.3) (-) L-Aspargine 210.7 (2.15) 52.1 (4.2) 34.8 (3.7) 40.4 (6.2) L-Phenylalanine 73.5 (0.2) 49.0 (3.4) 30.7 (4.7) 25.5 (1.5) L-Serine 250.5 (9.5) 49.9 (3.7 32.9 (3.6) 32.6 (1.1) L-Threonine 245.7 (2.0) 50.2 (2.4) 31.5 (3.8) 31.0 (3.5) Glycil-L-glutamic acid 260.3 (4.4) 48.4 (2.5) 33.7 (5.2) 44.1 (2.4) Phenylethylamine 74 (0.2) 47.3 (2.2) 26.7 (2.6) (-) Putrescine 83.6 (1.5) 49.7 (2.2) 38.0 (4.1) (-)

Results are the mean of OD492 x 100 (standard deviation of 2 plates), a negative value (-) was considered when OD492 <0.2

143

Supplementary Table 1 (continuation). Values of optical density of different carbon sources utilized by

E. coli O157:H7 and phylloepiphytic bacteria from BIOLOG ® plates.

Carbon source E. coli Rhizobium Kaistia spp. Patulibacter O157:H7 spp. spp. Pyruvic acid methyl ester 249.8 (2.2) 45.8 (2.0) 39.4 (0.2) 94.7 (34.6) Tween 40 99.4 (7.4) 31.1 (5.6) 29.6 (0.4) 158.8 (5.0) Tween 80 103 (7.7) 30.3 (3.2) 29.2 (1.2) 138.4 (12.3) a-Cyclodextrin 76 (2.6) (-) 29.2 (0.3) 31.8 (4.2) Glycogen 93.1 (0.35) (-) 28.6 (0.5) 35.6 (2.8) D-Cellobiose 77.4 (2.0) 68.7 (3.7) 66.8 (1.2) 40.3 (4.5) a-D-Lactose 277.1 (1.0) 62.1 (4.9) 48.8 (1.8) 29.6 (2.3) b-Methyl-D- Glucoside 248.2 (9.3) 24.3 (8.6) 55.2 (1.8) 35.2 (0.4) D-Xylose 272.2 (1.0) 80.5 (5.0) 86.9 (3.8) 58.3 (7.1) i-Erythritol 65.4 (1.0) (-) 30.1 (1.9) 29.7 (0.6) D-Mannitol 234.2 (3.2) 69.5 (5.9) 34.8 (0.8) 30.4 (3.4) N-Acetyl-D- Glucosamine 249.2 (2.3) 56.8 (2.3) 67.6 (1.3) 27.1 (4.4) D-Glucosaminic Acid 72.9 (0.5) 34.8 (3.0) 29.2 (0.2) 31.3 (3.5) Glucose-1-Phosphate 252.9 (4.3) 61.4 (2.1) 43.5 (1.1) 31.0 (11.3) D,L-a-Glycerol phosphate 247.1 (6.7) (-) 27.8 (0.5) 29.2 (5.2) D-Galactonic Acid g- lactone 70.7 (1.0) 73.2 (0.4) 41.5 (0.1) 31.3 (2.0) D-galacturonic acid 241.8 (7.7) 59.8 (5.9) 51.0 (1.0) 32.4 (4.1) 2-Hydroxy benzoic acid 60 (1.2) (-) 24.4 (0.8) 38.5 (2.1) g-Hydroxybutyric acid 73.2 (0.6) 28.9 (2.8) 25.9 (3.1) 44.8 (0.2) 4-Hydroxi benzoic acid 90.7 (2.8) 24.3 (3.5) 44.3 (1.2) 47.1 (4.6) Itaconic acid 73.2 (0.5) (-) 29.7 (0.6) 42.3 (1.7) a-ketobutyric acid 208.3 (2.0) (-) 26.4 (1.4) 41.8 (7.6) D-Malic acid 282.5 (1.2) 62.5 (2.1) 26.2 (0.4) 33.3 (10.2) L-Arginine 76.8 (0.5) 56.4 (6.8) 28.4 (0.1) 28.9 (6.0) L-Aspargine 210.7 (2.15) 84.0 (0.3) 29.0 (0) 43.1 (2.6) L-Phenylalanine 73.5 (0.2) (-) 27.4 (0.2) 43.7 (1.0) L-Serine 250.5 (9.5) 56.8 (8.6) 28.4 (0.1) 37.7 (0.4) L-Threonine 245.7 (2.0) 29.6 (8.7) 31.3 (2.7) 52.4 (9.5) Glycil-L-glutamic acid 260.3 (4.4) 41.5 (1.2) 29.9 (0.3) 39.1 (1.7) Phenylethylamine 74 (0.2) (-) 24.9 (0.05) 39.1 (1.6) Putrescine 83.6 (1.5) 22.5 (2.4) 32.1 (2.2) 60.1 (13)

Results are the mean of A492 x 100 (standard deviation of 2 plates), a negative value (-) was considered when A492 <0.2

144

Supplementary Table 1 (continuation). Values of optical density of different carbon sources utilized by

E. coli O157:H7 and phylloepiphytic bacteria from BIOLOG ® plates.

Carbon source E. coli Arthrobacter Rhodococcus O157:H7 spp. spp. Pyruvic acid methyl ester 249.8 (2.2) 68.6 (9.9) 314.7 (21.3) Tween 40 99.4 (7.4) 59.1(1.3) 208.7 (6.4) Tween 80 103 (7.7) 43.3 (5.2) 277.4 (23.1) a-Cyclodextrin 76 (2.6) 41.6 (7.5) 190.7 (5.3) Glycogen 93.1 (0.35) 38.9 (1.0) 271.6 (18.3) D-Cellobiose 77.4 (2.0) 36.8 (7.2) 243.3 (20.5) a-D-Lactose 277.1 (1.0) 41.1 (4.0) 254.8 (25.2) b-Methyl-D- Glucoside 248.2 (9.3) 33.8 (0.45) 331.6 (15.3) D-Xylose 272.2 (1.0) 77.8 (8.6) 277.9 (20.6) i-Erythritol 65.4 (1.0) 39.6 (6.4) 247.9 (20.1) D-Mannitol 234.2 (3.2) 57.2 (11.0) 274.0 (1.7) N-Acetyl-D- Glucosamine 249.2 (2.3) 34.7 (2.5) 275.5 (18.2) D-Glucosaminic Acid 72.9 (0.5) 34.1 (3.4) 311.3 (22.8) Glucose-1-Phosphate 252.9 (4.3) 41.5 (1.5) 251.8 (19.0) D,L-a-Glycerol phosphate 247.1 (6.7) 40.5 (1.0) 250.6 (17.2) D-Galactonic Acid g- lactone 70.7 (1.0) 28.8 (3.5) 239.7 (10.3) D-galacturonic acid 241.8 (7.7) 36.8 (2.3) 243.5 (12.8) 2-Hydroxy benzoic acid 60 (1.2) 36.7 (2.3) 123 (2.2) g-Hydroxybutyric acid 73.2 (0.6) 32.9 (1.9) 254.1 (10.6) 4-Hydroxi benzoic acid 90.7 (2.8) 48.4 (10.1) 287.0 (7.1) Itaconic acid 73.2 (0.5) 35.6 (0.3) 257.9 (15.7) a-ketobutyric acid 208.3 (2.0) 41.1 (2.5) 316.6 (21.7) D-Malic acid 282.5 (1.2) 42.7 (2.5) 253.7 (15.9) L-Arginine 76.8 (0.5) 30.8 (2.9) 261.4 (8.1) L-Aspargine 210.7 (2.15) 53.0 (7.9) 320.3 (11.4) L-Phenylalanine 73.5 (0.2) 42.4 (1.9) 260.4 (10.2) L-Serine 250.5 (9.5) 51.0 (3.5) 300.8 (13.4) L-Threonine 245.7 (2.0) 37.1 (0.1) 282.7 (13.1) Glycil-L-glutamic acid 260.3 (4.4) 42.1 (0.8) 279.0 (4.3) Phenylethylamine 74 (0.2) 38.2 (3.2) 230.8 (17.2) Putrescine 83.6 (1.5) 40.3 (0.1) 277.7 (30.1)

Results are the mean of OD492 x 100 (standard deviation of 2 plates), a negative value (-) was considered when OD492 <0.2

145

Figure 1. Formation of growth inhibition zones against E. coli O157:H7 by (A) Acinetobacter spp. and

(B) Pseudomonas spp.

146

CHAPTER 5

ALTERATIONS OF THE PHYLLOEPIPHYTIC BACTERIAL

COMMUNITY ASSOCIATED WITH INTERACTIONS OF ESCHERICHIA

COLI O157:H7 DURING STORAGE OF PACKAGED SPINACH AT

REFRIGERATION TEMPERATURES

Gabriela Lopez-Velasco1, Marjorie Davis1, Renee R. Boyer1, Robert C. Williams1, Monica A.

Ponder1, *

1 Virginia Polytechnic Institute and State University. Department of Food Science and

Technology. 22 Duck Pond Dr.. Blacksburg VA, 24061.

*Author for correspondence. Tel: +1 540-231-5031. Fax.: 540-231-9293E-mail: [email protected]

(License no. 2384780062658)

147

ABSTRACT

This study investigated the effects of packaging and storage temperature on the spinach phylloepiphytic bacterial community and fate of E. coli O157:H7. Freshly harvested spinach was rinsed and/or disinfected, packaged and stored under typical retail conditions (4°C) or under temperature abuse conditions (10°C) for a period of 15 days. The final population size of culturable epiphytic bacteria after 15 days of storage was not affected by the temperature of storage or the presence of E. coli O157:H7. However, analysis of the bacterial community using denaturing gradient gel electrophoresis of 16s rDNA revealed changes with time of storage and the presence of E. coli O157:H7. Excision and sequencing of prominent DGGE bands identified that the majority of sequences belonged to the phyla Actinobacteria, Bacteroidetes, Firmicutes and Alphaprotebacteria. After 10 days of storage at 4°C or 10°C the population became more dominated by psychrotrophic bacteria. Removal of the epiphytic bacteria resulted in significant increases in numbers of E coli O157:H7 at 10°C and was associated with decreased expression of E. coli O157:H7 virulence (stxA, curli, eaeA) and stress response (rpoS, sodB) genes. In conclusion, storage temperature and time of storage of packaged spinach affected the diversity of the epiphytic spinach microbiota which influenced the growth, establishment, physiology and potentially virulence of E. coli O157:H7.

Key words: Spinach, E. coli O157:H7, low temperature storage, DGGE, microbial diversity, phyllosphere bacterial community

148

INTRODUCTION

Enterohemorrhagic Escherichia coli (EHEC) particularly serotype O157:H7 is responsible for an estimated 79 thousand illnesses each year, though the majority of cases are unreported (33). In recent years, fresh minimally processed, ready to eat produce items, such as lettuce, alfalfa, spinach and sprouts, have become a major vehicle of this microorganism (46). In

2006, three multistate, fresh produce associated outbreaks were attributed to E. coli O157:H7

(19). The largest outbreak, 205 confirmed illnesses and 3 deaths in the United States and Canada, was attributed to consumption of fresh, bagged baby spinach (22). The contaminated product was traced to one production date at one processing plant on the central California coast and distributed across the USA and Canada (22). The outbreak was ultimately traced back to a pre- harvest source, however the post harvest processing exacerbated the scale of contamination by mixing contaminated produce with spinach from other fields that were packaged and distributed

(22). It is standard practice to wash spinach in a chlorinated solution to remove debris and reduce the bacterial load (41). Unfortunately, this washing step is only bacteriostatic, as populations of

E. coli O157:H7 recover to pre-wash numbers after 7 days of storage at 7oC + 2oC (27).

Modified atmosphere packaging (MAP) of leafy greens, such as spinach in films, which allow for accumulation of CO2 and diffusion of O2, reduces plant tissue respiration and extends shelf life (41). This creates low oxygen conditions within the package which decreases growth of spoilage organisms but may promote growth and attachment of facultative anaerobic pathogens such as E. coli O157:H7 (1, 16, 48). To further increase the shelf life, MAP packaged produce is stored at low temperatures, normally at 4oC (40). E. coli O157:H7 on iceberg lettuce can survive

149 extended periods of storage at 4oC, up to 12 days, and actually increase in numbers when stored at 8oC for 12 days (13, 16).

The role of packaging and storage at low temperature is known to reduce growth, though not eliminate human pathogens on inoculated produce. Little is known how these hurdles affect the normal phyllosphere associated epiphytic (phylloepiphytic) microbial communities, which are adapted to life on the plant leaf surface. An estimated 106 to 107 cells / cm2 bacteria, belonging to more than 85 different species, inhabit the plant leaf surface, despite the low availability of nutrients and exposure to fluctuating climatological conditions (4, 18). Epiphytic bacteria are important because they can suppress or stimulate the colonization and infection of tissues by plant pathogens (30, 50). It is possible that epiphytic bacteria may have a similar role in establishment or persistence of human pathogens. As this pre-harvest associated microbiota remain associated with raw vegetables during transportation, further processing and storage (21), it is important to examine the effect of packaging and low temperature storage on epiphytic bacteria, especially as they may affect the establishment of E. coli O157:H7. To date no studies have examined the effect of typical ready to eat processing and storage conditions on the community structure of spinach phylloepiphytic bacteria. Previous studies on the survival of E. coli O157:H7 have used commercially available spinach which is typically 5-6 days post-harvest before inoculation, preventing investigation of the role of the initial microbial community on the growth and establishment of E. coli O157:H7. In this study the postharvest phylloepiphytic community of fresh (harvested within two hours of packaging) spinach was evaluated under typical retail storage conditions (4°C) or under temperature abuse conditions (10°C) for a period and alterations of the community profile due to presence of E. coli O157:H7 were identified. In

150 contrast, epiphytic and E. coli O156:H7 survival was evaluated on previously commercial packaged spinach as a comparison.

MATERIALS AND METHODS

Plant material.

Fresh spinach. A flat cultivar (Monza, Seedway LLC. USA) was seeded in April, 2008 at the Virginia Tech Experimental Research Farm. One plot of approximately 8 m x 4 m was divided in four sections to form four raised beds. The soil was prepared prior to seeding by plowing and fertilized using 10-10-10 fertilizer (1:1:1 N, P, K) (Southern States Cooperative,

Richmond VA). After seeding, the plot was irrigated every other day with well water. Spinach was harvested at baby stage (7 to 10 cm in length from the stem to tip) approximately 30 days after seeding. To randomly collect the spinach a hula-hoop (0.32 m2 of area) was thrown into the field and spinach within the hoop was collected until approximately two Kg of spinach leaves were harvested. The selected spinach plants were cut from the stem using scissors, placed in sterile plastic bags and transported to the laboratory on ice where they were processed within two hours.

Commercial spinach. Ready to eat packaged spinach from a national supplier was purchased at a local supermarket in Blacksburg, VA. The commercial spinach packages were obtained from the supermarket prior to exhibition in shell for consumer purchase. Each package had a sell by date of at least 10 days from the day of purchase. Packages were stored at 4oC for no longer than 6 hours prior to initiation of experiments.

Escherichia coli O157:H7 strains and inoculum preparation.

151

Escherichia coli O157:H7 (CDC PulseNet pattern number EXHX01.0124) (provided by

CDC) previously isolated from spinach leaves during the spinach outbreak of 2006 was utilized to inoculate fresh spinach. Escherichia coli O157:H7 strain H1730 transformed for GFP expression and kanamycin resistance, a clinical isolate associated with a lettuce outbreak

(provided by Dr. Larry Beuchat, University of Georgia) was utilized to inoculate commercial spinach. The strains were activated in tryptic soy broth (TSB, Difco, Spark MD) at 37oC for 48h.

Cultures were screened for purity on sorbitol MacConkey agar (SMAC Difco) an isolated colony

o was transferred to TSB and incubated at 37 C until an OD600 nm of 0.9-1.0 was achieved. The cells were collected by centrifugation (4000 rpm, 15 min, 4oC), washed twice with 1X phosphate buffered solution (PBS; 137 mM NaCl, 2.7 mM KCl, 100mM Na2HPO4, 2 mM

KH2PO4) and resuspended in 1X PBS to achieve a final concentration of 6 log CFU/mL. Since experiments using fresh spinach and commercial spinach were performed independently using different strains, growth kinetics of both strains were compared by construction of growth curves using an automated growth curve analysis system (Bioscreen C™ Piscataway, NJ). No differences in growth rates and kinetic values were observed, making both experiments comparable (data not shown).

Spinach processing.

Fresh spinach was received in the laboratory and immediately washed with sterile water to remove soil particles. All spinach leaves were pooled and divided in two sets, one set was disinfected with 12.5% (v/v) sodium hypochlorite solution (6% v/v) for 10 minutes to result in a

4 fold reduction in the background microbiota (Disinfected group) and one set received no further rinsing (Rinsed group). Both spinach groups were dried using a household salad spinner.

Each group (rinsed and disinfected) was again divided in half and a set of disinfected and rinsed

152 spinach were placed in a single layer on a sterile cutting board within a laminar flow bio-safety cabinet to facilitate drying. Escherichia coli O157:H7 (CDC strain) was inoculated onto whole spinach leaves via spot inoculation method previously described by Lang et al. 2004 (26).

Briefly, 50 L of 6 log CFU/mL of E. coli O157:H7 was transferred onto the surface of each leaf at three spatially distinct locations (150 L total). These samples were designated as rinsed inoculated and disinfected inoculated treatments. The same procedure was used for the control samples except that 50 L of 1X PBS was used instead of the E. coli O157:H7. These samples were designated rinsed control and disinfected control treatments.

Leaves from each treatment were transferred aseptically into polyethylene PD 961 bags

(14 x 21 cm) (O2 transmission rate =450 cc/100 sq. in. and CO2 transmission = 1355 cc/100 sq. in., Cryovac, Duncan S.C.) to achieve a final weight of 10 g of spinach per package. The bags were sealed using an Ultravac ® 225 vacuum packaging machine under a 30% vacuum (Ultravac solutions, Kansas, MO). Packages of spinach were held at either 4oC or 10oC and sampled until noticeable spoilage occurred. Spoilage was defined as chlorosis and cell turgidity, development of sliminess and weeping of tissue fluid (40).

For commercial spinach, bags obtained from the supermarket were opened and divided in two sets (Commercial group). Both sets were processed as described above; one set was inoculated with E. coli O157:H7 (Strain H1730 gfp+/kan+) at a concentration of 6 log CFU/g and the other set was inoculated with 1X PBS. These samples were designated commercial inoculated and commercial control respectively. Spinach was re-packaged as previously described and held in the same conditions. Spinach bags were sampled until noticeable spoilage occurred as previously described.

153

Analysis of gaseous atmosphere inside spinach bags.

Daily (10 days for commercial spinach and 15 days for fresh spinach) samples of headspace gas within each package were withdrawn using a sterile needle and analyzed to determine the percentage of O2 and CO2 in the packages using an oxygen and carbon dioxide gas analyzer (PBI-Dansensor, PBI development, Denmark).

Enumeration of background microflora.

Five packages of spinach for each treatment were removed from storage at each sampling day. The leaves were aseptically transferred into Pulsifier® bags (Microbiology International,

Frederick, MD) containing 90 mL of 1% (wt/vol) sterile peptone water (Sigma-Aldrich Co., St

Louis MO) supplemented with 1% (vol/vol) of Tween-90 (PTW) (Fisher, Atlanta, GA) and samples were pulsified for 5 minutes to detach epiphytic microbiota. Ten milliliters of the suspension were serially diluted and plated onto R2A (Difco, Spark MD) to quantify the culturable epiphytic bacteria or SMAC to enumerate E. coli O157:H7. The plates were incubated for 48h at 25oC for R2A and 37°C for SMAC plates. Random sorbitol negative colonies from SMAC were selected and the presence of O157 and H7 antigens confirmed using the RIMTM latex agglutination test (Remel, Lenexa, KS) for confirmation of O157 and H7 antigens.

Nucleic acid isolation

Six packages per treatment were individually processed for nucleic acid extraction.

Bacterial cells were disassociated from the spinach leaves using the same procedure described for enumeration. Bacterial cells were collected from the bacterial suspension by centrifugation at

4,000 rpm/20min/4oC washed with 1X PBS and resuspended in 100 L of 1X TE buffer. The

154 cells were lysed by incubation with 300 g of lysozyme (Fisher, Fair Lawn NJ), 10 units of mutanolysin (Fisher) and 25 g of achromopeptidase (Sigma-Aldrich Co., St Louis MO) at 37oC for 30 min. Lysates were treated with 25units of proteinase K (Fisher) and incubated at 65oC for

30 min. DNA was extracted from the lysed cells of three packages using the ZR soil microbe

DNA kitTM (Zymo Research Co., Orange, CA) per manufacturer’s instructions. The lysates of the remaining three packages were used for RNA isolation using RNeasy Mini Kit (Qiagen, CA) per manufacturer’s instructions. All plastic ware and solutions were previously treated to remove endonucleases and avoid RNA degradation. RNA later (Ambion, TX) was added to the initial PTW solution to stabilize RNA. RNA quality was determined by gel electrophoresis in formaldehyde agarose gel.

DGGE analysis

Product amplification. The 16S rRNA gene was amplified from the total epiphytic microflora DNA (50 ng/ L) to generate a 566 bp fragment using the primers 341-f (5’-CCT

ACG GGA GGC AGC AG-3’) and 907-r (5’-CCG TCA ATT CMT TTG AGT TT-3’). The forward primer was modified to add a 40 nucleotide GC clamp at the 5’ end (5’-CGC CCG CCG

CGC GCG GCG GGC GGG GCG GGG GCA CGG GGG G-3’) (34). Each 25 L PCR reaction contained 1.5mM of MgCl2, 50mM of KCl, 0.2mM of each dinucleotide, 1% of DMSO

(dimethylsulfoxide), 25mM of Tris-HCl (pH 8), 1U/ L of HotStart-IT FideliTaq DNA polymerase (USB 71156, Cleveland, OH, USA), 0.5 M of each primer, and 50 ng of DNA. The

PCR protocol consisted of denaturation at 94oC for 5 minutes, followed by 19 cycles of 94oC for

1min, amplification at 64oC for1min (decreasing 1oC every second cycle –touchdown-) and elongation at 72oC for 3 min, followed by additional 9 cycles of denaturation at 94oC for 1min,

155 amplification at 55oC for 1min and elongation at 72oC for 3 min, finally 1 cycle of 94oC for

1min, amplification at 55oC for 1min, and a final elongation step at 72oC for 10min (34).The size and intensity of PCR products were confirmed using a 0.9% agarose gels (Fisher-Scientific,

Atlanta, GA).

DGGE conditions. The PCR products were run on a 8% polyacrylamide gel in a 30-60% denaturant gradient of urea and formamide (100% denaturant corresponds to 7M urea plus 40%

(vol/vol) of deionized formamide) using the Bio-Rad DCodeTM Universal Detection System

(Bio-Rad, Hercules, CA). Twenty two microliters of PCR products were separated at constant voltage of 85V and temperature of 60oC for 17 hours. The DNA bands were visualized by staining with ethidium bromide and photographed using the Molecular Imager ® GelDocTM XR

(Bio-Rad). The bands were analyzed using Gel Compare II V.6 (Applied Maths, Austin, TX) and the DGGE profiles were analyzed by clustering using the unweighted pair group method with mathematical averages (UPGMA; Dice coefficient of similarity) using the same software producing dendrograms for the different treatments. Gels were repeated three times.

Band identification. Bands which varied in intensity or presence between different treatments were excised from the DGGE gel, suspended in 100 L of water, and incubated at

4oC for 24 h to allow passive DNA diffusion. Diffused DNA was re-amplified using 341-f and

907-r primers as described above. The PCR products were purified from a 1 % (w/v) agarose gel and cloned into pCR4-TOPO® VECTOR (Invitrogen) to transform One Shot® Mach1 ™ chemically competent E. coli (Invitrogen). Plasmids were isolated using Quick lyse miniprep

(Qiagen) and inserts were sequenced. Sequences were compared to the NCBI database using

BLASTN (www.ncbi.nlm.nih.gov).

156

Gene expression analysis.

Stress response, pathogenicity and attachment related gene expression of E. coli O157:H7 in inoculated spinach was measured to evaluate the effect of the different treatments on their expression using quantitative real time PCR (RT-PCR).

Primer design. All primers were designed specific to E. coli O157:H7 (accession number NC_002655) using Beacon Designer 6.0 software (Premier Biosoft Int) (Table 1).

Primer efficiency was determined by preparation of standard curves using RNA of Escherichia coli O157:H7 (CDC strain) using a ten-fold dilution to achieve concentrations from 100 copies to

107 copies. Each concentration in the standard curve was replicated in triplicate in separate assays. Melting curve analysis of the PCR products was conducted following each assay to confirm that the fluorescence signal originated from specific PCR product. Primers specificity was confirmed by testing primers with DNA from non-inoculated spinach and from DNA obtained from pure cultures of previously isolated spinach epiphytes. All Ct values were in a range of 36 to 39, and then a Ct value of 35 or less in the inoculated samples was considered for analysis (data not shown).

Real time PCR. Total RNA was converted to cDNA and primer specific cDNA amplified using the One-Step quantitative reverse transcription PCR kit (USB, Cleveland OH) per manufacturer’s instructions. Each 25 L reaction contained 50 ng of total RNA, 0.4 L of M-

MLV RT (reverse transcriptase), 0.4 L of RNase inhibitor, 12.5 L of HotSart-

ITTMSYBR®Green qPCR Master Mix 2X (USB® 75770 Cleveland, OH, USA), 10nM of fluorescein as passive reference dye (USB® 75767 Cleveland, OH, USA) and 0.5 M of forward and reverse primers. PCR conditions consisted of one cycle of 10 min at 50oC for reverse transcription of RNA, one cycle of 2 min at 95oC for activation of HotStart-IT polymerase and

157 reverse transcriptase inactivation, followed by 40 cycles of denaturation at 95oC for 30 s, 30 s at

60oC for primer annealing and real time detection, and 72oC for 1 min. cDNA synthesis and amplification was carried out with an iQ5TM Optical system Real time PCR detection system

(Bio-Rad). Ct values were normalized using rpoB as a housekeeping gene and gene expression was calculated by relative quantification (ratio=2 Ct). Each reaction was replicated in duplicate for technical replicate and in duplicate for biological replication.

Statistical analysis

Bacterial enumeration experiments were repeated with three different spinach packages

-1 per treatment with duplicate plate counts. Data was normalized by conversion to log10 CFU/g of spinach. For gene expression, results were performed with RNA isolated from two different packages with a duplicate of each treatment measured. Comparison among treatments was carried using proc mixed and Tukey’s multiple comparisons of means in Statistical Analysis

System 9.2 (SAS Institute, Cary, NC). Statistical significance was established when p-values were lower than 0.05. All data was previously evaluated for normality and homogeneity of variance using the proc univariate function of SAS.

RESULTS AND DISCUSSION

Spinach quality during storage. Microbial numbers were monitored on freshly harvested spinach until leaves showed visible signs of spoilage including chlorosis and cell turgidity, development of sliminess and weeping of tissue fluid(15 or 23 days at 10°C, and 4°C respectively). In contrast, the shelf life of commercially obtained spinach was only 10 days from the day it was re-packaged and stored at 4 or 10°C. The atmosphere in all bags increased in CO2 concentration and decreased in O2 concentration during storage. The rate of CO2 increase at

158

4°C was lower than 10°C, regardless of the microbial load within the bags, indicating that the plant respiration rate is critical for maintaining shelf life. Therefore, incubation at temperatures below 10°C is imperative to maintain good sensorial characteristics of packaged spinach (Figure

1).

Effect of time and temperature of storage on spinach phylloepiphytic microbiota. Plate counts were performed to quantify changes in bacterial numbers on fresh spinach throughout storage under typical retail conditions for a period of 15 days. Samples subjected to the rinsing treatment provide the most direct comparison to spinach prepared commercially and sold ready to eat. The disinfected treatment was designed to remove the largest number of epiphytic microbiota, allowing for the determination of the effect of native microflora on the survival and potential for growth of E. coli O157:H7 and does not represent procedures commonly used for ready to eat spinach. In order to compare these results with those from other researchers (12, 13), commercial ready to eat spinach was obtained from a supermarket and monitored during the same time of storage. The three groups of spinach were stored at 4oC to represent standard conditions or 10oC representing temperature abuse conditions.

Initial microbial counts of leaves at day 0 were not significantly different (p>0.05) between the packages that had the same spinach source and treatment. Bacterial counts of fresh, rinsed spinach were approximately 4 logs at day zero (Table 2), similar to counts from other fresh plant tissues which ranged from 3 to 7 log CFU/g (38). In contrast, microbial counts on minimally processed commercial spinach were significantly higher (8-9-log). Minimal processing can lead to increased nutrient availability due to tissue damage, which may explain the increased microbial load on these leaves (5, 49). Additionally, these leaves were a minimum of 5-6 days older than the fresh leaves, however fresh spinach held for 5 days still showed

159 significantly smaller populations than the commercial spinach at day 0 (Table 2). In this experiment the large microbial populations (9 log) on the commercial product stored at 4°C on day 10 were more similar to the counts obtained at 10oC for the rinsed fresh spinach at day 15.

Temperature of storage during transportation and prior to retail display are unknown, however these counts suggest that the commercial packages may have been subjected to temperature abuse, which would significantly affect the quality of product and potentially its safety.

The total bacterial populations associated with the packaged spinach increased with time of storage regardless of temperature. Storage of spinach for short periods did not result in differences in microbial numbers, except for the disinfected spinach where the remaining bacteria were able to quickly re-establish populations on the surface. Prolonged storage resulted in significant increases of 5 fold at 10oC and 3.6 fold at 4oC, by day 15 for fresh spinach for all treatments (Table 2). The rate of proliferation (log increase) was lower for commercial spinach populations compared to rinsed spinach treatments, however the total bacterial populations of 9 logs were achieved from all packages of spinach (Table 2), suggesting that leaves have a carrying capacity that restricts the total population size. This carrying capacity is likely dictated by space and nutrient availability on the leaf surface. These proliferating microorganisms are termed as “psychrotrophic”, as they are capable of growth at the temperatures used in this study.

Previous studies have noted that psychrotrophic bacteria, particularly the species Pseudomonas sp., Erwinia sp., Rhanella sp., and lactic acid bacteria, tend to dominate the culturable populations on vegetables (42). Many of these bacteria have been reported as responsible for spoilage and deterioration of fresh vegetables (21), however some epiphytic isolates have been reported to inhibit the growth of foodborne pathogens (Salmonella sp. E. coli and Listeria monocytogenes) (28, 29, 44).

160

The effect of packaging, storage temperature and disinfection on the epiphytic bacterial community were assessed by amplifying a section of the highly conserved 16s rDNA molecule and comparing the resulting DGGE profiles (Figure 2). DGGE is a culture independent technique broadly used to study bacterial population dynamics in different environments, allowing the study of community structure and behavior (35-37). Changes in microbial diversity can be visualized by identification of different band patterns, indicating changes to species richness, changes in intensity of the bands indicate changes in abundance. For the DGGE profiles it is assumed that each band represents a particular group of microorganisms (37).

Disinfection of the spinach leaves prior to packaging resulted in the greatest reduction in species richness (Figure 2A and 2B). The species richness was not further affected by the time of storage, however changes to the intensity of bands were observed that correlates with an increase in culturable microorganisms (Table 2, Figure 2, lanes 3, 4, 11, 12 and 16). The most intense band was sequenced and identified as the chloroplast and it was not considered for diversity analysis. The progenitor of the chloroplast is believed to be a photosynthetic bacterium thus it is not unexpected that the primers to the slowly evolving 16s rDNA amplify it. Increases in species richness and abundance were observed for rinsed spinach stored at 4°C and 10oC with prolonged storage (Figure 2C and D). Cluster analysis indicated that similarity between populations was influenced by temperature and by the treatments (Figure 3). Cluster analysis revealed that bacterial communities on spinach stored at 4°C were 50 to 60% similar to communities on spinach stored at 10°C (Figure 3). The treatment further reduced similarity of the communities as communities of rinsed and disinfected treatments stored at the same temperature were only 30 to 40% similar (Figure 3).. The length of storage also impacted the community structure, as patterns from days 0 and 5 were more similar to each other than days 10 and 15.

161

Sequencing of bands excised from samples taken after 1 or 5 days of storage (Table 3 bands 1-11) identify these bacteria as genera previously isolated from soil, rhizosphere or the phyllosphere. However a large number of the sequences belonged to uncultured microorganisms which have been reported by other studies on the phyllospheres of rye, olive, sugar beet and wheat (53). Further species richness increases on days 10 and 15 and increased intensity of many bands coincided with an overall increase in cultured microbial populations (Table 2). The majority of excised bands from profiles of spinach stored 10- 15 days were identified as bacteria that have not been cultured. Of the culturable bacteria identified all have been characterized as psychrotrophic microorganisms (Bands 13, 14, 17 and 20) some of them commonly isolated from extreme low temperatures such as Frigoribacterium sp. (39) and Exiguobacterium sp. (23).

This likely indicates that a shift in population diversity may occur as response to low temperature exposure. Unlike other phyllosphere studies which are dominated by members of the

Gammaproteobacteria (25, 50), this work showed that most of the excised bands belonged to members of the phyla Actinobacteria, Bacteroidetes, Firmicutes and Alphaprotebacteria and to a lesser extent to Betaproteobacteria and Gammaproteobacteria. It is possible these differences may be due to band co-migration (one band was representing more than one microorganisms) and micro-heterogenicity of the 16s rDNA gene (the same microorganism in different bands), a phenomenon commonly reported with DGGE (37) (Table 3).

Effect of time and temperature storage conditions on E. coli O157:H7 and its effect on the spinach epiphytic microbiota. E. coli O157:H7 numbers were maintained on inoculated spinach leaves stored under experimental conditions. A slight decrease in E. coli O157: H7 numbers were observed after 15 days at 4°C, while storage at 10°C resulted in a small increase in

E. coli O157: H7 counts (Table 4), however the results between two temperatures were not

162 significantly different. E .coli O157:H7 populations on spinach leaves stored at 4°C, decreased between day 0 and day 5, however by day 15 the populations returned to initial levels, suggesting the adaptability of this microorganism to low temperatures by recovery from a viable but not culturable state. This minimal effect on survival of E. coli O157:H7 is similar to other studies where no change in population or a slight population decline was observed after 14 days of storage of iceberg lettuce stored at 5oC (13, 14).

To determine the effect of epiphytic populations on survival of E. coli O157:H7 disinfected spinach was inoculated and stored under the same experimental conditions. In contrast to studies on rinsed leaves the populations of E. coli O157:H7 significantly increased, 4- fold and 1 fold after 15 days of storage at 10oC or 4oC respectively (p<0.05). In addition, E. coli

O157:H7 became the dominant microorganism with respect to the epiphytic population on the disinfected leaves stored at 10°C (Table 4). DGGE analysis revealed increased species richness of inoculated spinach bacterial communities compared to non-inoculated groups regardless of temperature of storage (Figures 4A-D).

Addition of E. coli O157:H7 to the leaf surfaces also altered the epiphytic microbial community of spinach leaves. Samples which were inoculated formed separated clusters from non-inoculated samples, indicating differences among the bacterial communities due to the presence of E. coli O157:H7. At the same time, samples were separated in clusters dependent on treatment (rinsing or disinfection) and temperature storage. The presence of E. coli O157:H7 most greatly altered the community structure of disinfected samples that were stored at 4°C

(Figure 3). The decreased similarity with length of storage is supported by the increased percentage of the total community which were E. coli O157:H7 (Table 4). These results indicate that the spinach microbiota influences the growth of E. coli O157:H7 and/or vice versa. The

163 reduction in initial microbial numbers on disinfected leaves likely provides E. coli O157:H7 with additional nutrients and space to colonize without competition from large numbers of epiphytic microbiota. Competitive exclusion for either space or for nutrients has been observed in vitro for several common phyllosphere inhabitants, including Pseudomonas fluorescens which depletes growth media of nutrients inhibiting the growth of E. coli O157:H7, a phenomenon that was reversed by the replenishment of nutrient broth (32). Another common epiphyte, Enterobacter asburiae, decreases survival of E. coli O157:H7 10- to 30 fold on lettuce leaves, in the absence of additional carbon sources, indicating it is better adapted to utilize carbon sources from the leaf surface (12). Presence of E .coli O157:H7 corresponded to changes in species richness in a few cases (Figure 4, lines 7, 11, 12, 15 and 16). In these situations it is likely that a synergy developed between the new epiphyte and the pathogen possibly due the ability of this food-borne pathogen to enhance bacterial growth through the production of diffusible molecules. Examples where one microbe produces a substrate for growth or more frequently a siderophore to increase iron availability are numerous in the phyllosphere (31). It has been suggested that siderophores produced by Salmonella enterica and E. coli might be utilized by members of the phyllosphere

(8). It is important to consider the effect of washing on the numbers of epiphytic microflora, as removal of this microflora may provide pathogenic microorganisms a more suitable niche for growth and establishment with reduced competition, especially if contamination occurs during processing steps. To our knowledge this is the first study that demonstrates that E. coli O157:H7 can influence the dynamics of the epiphytic microbial communities, which suggest that important microbe-microbe interactions could occur in the phyllosphere during spinach storage.

Effect of packaging and temperature on the expression of E. coli O157:H7 virulence and stress response genes. Attachment is necessary for the establishment of bacteria on the plant

164 surface (6, 8). It has been suggested that the attachment of foodborne pathogens to plant surfaces may involve genes used for attachment to animal tissues (8). The study characterizes the effect of storage at 4°C and 10°C on the expression of two different attachment genes commonly expressed during animal virulence: espA and curli (csgA). Previous studies have indicated that failure to produce either espA or curli resulted in significant reductions in the numbers of E. coli

O157:H7 that could adhere to the surface of lettuce at room temperature (7, 45). In this study we found that expression of both espA and csgA increased with prolonged storage when stored at

10°C compared to 4°C (Table 5). Expression of espA was significantly up regulated with prolonged storage at 4°C; however the magnitude of expression was much smaller. In contrast, disinfection of the spinach prior to inoculation resulted in a down regulation of espA and csgA; this may suggest that signal molecules produced by other microbiota are necessary signals for their expression. Quorum sensing signals such as autoinducer 3 (AI-3), normally produced by human microbiota, have been demonstrated to affect the expression of the type III secretion system in E. coli O157:H7, from which EspA is a protein component (24), which plays an important role in capsule flagellar and fimbriae biosynthesis (15).

Expression of E. coli O157:H7 genes sodB, rpoS, and rpoN, associated with stress response were affected by the presence of background microbiota and temperature of storage

(Table 5). Inoculation onto disinfected spinach resulted in no change in expression of sodB at

10°C. Despite the up regulation 4°C from day 0 to day 5, sodB expression returned to the initial expression levels after 10 days of storage. In this study E. coli O157:H7 gene sodB was up- regulated with prolonged storage when inoculated onto spinach at 10oC or 4°C. The overall amount of expression was higher in samples stored at 10°C. In contrast, a recent study of E. coli

O157:H7 inoculated on Romaine lettuce leaves, showed that expression of sodB, which is also

165 related with oxidative stress, was down regulated during storage at 15oC when compared to 4°C

(9) which may suggest differences on the behavior of E. coli O157:H7 dependent of the type of plant tissue but also with storage temperature. Superoxide dismutase proteins play a major role against oxidative stress due to the formation of hydrogen peroxide related radicals providing protection to mutagenesis (43). However down regulation was observed on disinfected spinach, indicating an influence of the background microbiota. It is possible that background microbiota can produce metabolites including hydrogen peroxide as part of their normal metabolism that trigger the expression of sodB, however if the background microbiota is reduced in population, the stress response might be attenuated. The largest changes in gene expression seen in this study occurred with rpoS. Expression of rpoS, a master regulator of the general stress response, is highly affected by temperature storage and presence of background microbiota. Expression of rpoS increased with time of storage and temperature even though expression decreased at 10°C compared to 4°C on dinfected spinach. Downregulation was observed on disinfected spinach stored at 10°C which corresponds to previous studies made on romaine lettuce, where upregulation of rpoS occurred when lettuce was stored at 4°C while this gene tended to be downregulated during storage at 15°C (9). Expression of rpoS may prime the cells for better survival in stress conditions, such as low temperature, increased acidity and oxidation, allowing strains to persist until transmission to humans or another host can occur (2, 3). Studies in planktonic growth of E. coli O157:H7 did not showed changes in rpoS gene expression after exposure to 5°C (17), which might suggest differences in gene expression profiles on the leaf surface. However, in a different study, E. coli K12 growth at 37°C exposed to cold shock at 23°C showed preferential expression of RpoS controlled genes indicating that decreases in environmental conditions can trigger stress responses regulated through RpoS (51). Expression

166 of rpoS on E. coli O157:H7 in planktonic growth of cold shocked cells at 7.5°C was increased by about 6-fold also with an increase of 11-fold in the expression of stx1a (2). Studies of E. coli

O157:H7 on fresh lettuce packed under MAP conditions, have indicated that E. coli when stored at improper temperatures (>5°C), can survive a subsequent gastric acid challenge (10) a characteristic that may be dependent of RpoS (11). Given the high levels of expression of rpoS when rinsed spinach was stored at 10oC, it is very likely that survival and potentially the pathogenicity of E. coli O157:H7 may be enhanced at temperature abuse conditions.

The effect on pathogenicity of E. coli O157:H7 of sub lethal stress, such as that experienced during storage of ready to eat produce, has important implications for food safety.

Verocytotoxin expression has been only minimally studied on fresh produce. Prolonged storage of E .coli O157:H7 at 4°C has been shown to increase expression of stx1a in vitro (2) and on

Romaine lettuce stored at 4°C for 9 days (9). In this study, expression of stxA was observed at

4°C and 10°C. Genes encoding Stx are induced by an SOS response, which are also regulated by quorum sensing (47), which indicates that quorum sensing is an important regulatory system for virulence and then the presence of the microbiota influences the expression of stxA. Clearly the effect of the microbiota in conjunction with the storage temperatures may have an important influence in the pathogenicity of this microorganism. RpoN (sigma 54), is involved in quorum sensing signaling in different microorganisms (20). Although RpoN is not essential for cell growth, it is required for expression of genes involved in nitrogen fixation, utilization of alternative carbon sources and virulence (52). The rpoN up-regulation was found in the groups of bagged spinach that were rinsed, however, for disinfected spinach this gene was downregulated as many of the other genes analyzed in this study. This suggests that the effect of the background microbiota is important in the physiology of this microorganism during the storage

167 of bagged spinach due to signals produced through quorum sensing systems sensed by E. coli

O157:H7 which have been reported to regulate expression of its virulence and flagellar genes

(47). Then its interaction with the microbiota is an exciting topic that still is not been well explored and that might provide important insights about the strategies of E. coli O157:H7 to successfully establish on edible plants.

CONCLUSIONS

The epiphytic microbiota on edible plants adapts to different environmental conditions to gain fitness into the community (8). When this environment is changed due to artificial conditions as those that occurred during post-harvest i.e. washing, packaging and storage, members of the community are also altered. In this study we have shown that abundance and diversity of the spinach microbiota were altered during storage at low temperatures with selection towards psychrotrophic microorganisms indicating the adaptation of certain members of the community to these changes. Bacterial-bacterial interactions can be inferred from the alterations to microbial community structure that was observed in the presence of E. coli O157:H7.

Reductions in abundance and diversity of the spinach epiphytic microbiota combined with abused temperature storage (10°C) may allow E. coli O157:H7 to grow and establish on spinach surface, indicating the importance of the microbiota in the establishment and fate of food-borne pathogens. Furthermore, the necessity of competition with these other microflora may increase the pathogenicity and stress responses of E. coli O157:H7. Further research to understand specific interactions of the phyllosphere microbiota with food-borne, particularly the role of quorum sensing is needed to identify better strategies for controlling survival and establishment of pathogenic microorganisms on edible plants. We conclude that temperature and time of

168 storage of packaged spinach affected the diversity of the epiphytic spinach microbiota which influenced the growth, establishment and physiology of E. coli O157:H7.

ACKNOWLEDGEMENTS

E. coli O157:H7 (pattern number EXHX01.0124) strain was provided by the PulseNet laboratory of the Centers for Disease Control and Prevention. All bags were generously provided by Cryovac® Food Packaging Systems. Fellowship for partial financial support of

Gabriela Lopez-Velasco was provided by the National Council for Science and Technology

(CONACyT) from Mexico.

169

REFERENCES

1. Abdul-Raouf, U. M., L. R. Beuchat, and M. S. Ammar. 1993. Survival and growth of

Escherichia coli O157:H7 on salad vegetables. Appl Environ Microbiol 59:1999-2006.

2. Allen, K. J., D. Lepp, R. C. McKellar, and M. W. Griffiths. 2008. Examination of stress and

virulence gene expression in Escherichia coli O157:H7 using targeted microarray analysis.

Foodborne Pathog Dis 5:437-47.

3. Anand, S. K., and M. W. Griffiths. 2003. Quorum sensing and expression of virulence in

Escherichia coli O157:H7. Int J Food Microbiol 85:1-9.

4. Andrews, J. H., and R. F. Harris. 2000. The Ecology and biogeography of microorganisms on

plant surfaces. Annu Rev Phytopathol 38:145-180.

5. Aruscavage, D., K. Lee, S. Miller, and J. LeJeune. 2006. Interaction affecting the proliferation

and control of human pathogens on edible plants. J Food Sci 71:89-99.

6. Barak, J. D., L. C. Whitehand, and A. O. Charkowski. 2002. Differences in attachment of

Salmonella enterica serovars and Escherichia coli O157:H7 to alfalfa sprouts. Appl Environ

Microbiol 68.

7. Boyer, R. R., S. S. Sumner, R. C. Williams, M. D. Pierson, D. L. Popham, and K. E. Kniel.

2007. Influence of curli expression by Escherichia coli 0157:H7 on the cell's overall

hydrophobicity, charge, and ability to attach to lettuce. J Food Prot 70:1339-45.

8. Brandl, M. T. 2006. Fitness of human enteric pathogens on plants and implications for food

safety. Annu Rev Phytopathol 44:367-92.

9. Carey, C. M., M. Kostrzynska, and S. Thompson. 2009. Escherichia coli O157:H7 stress and

virulence gene expression on Romaine lettuce using comparative real-time PCR. J Microbiol

Methods 77:235-42.

170

10. Chua, D., K. Goh, R. Saftner, and A. Bhagwat. 2008. Fresh-cut lettuce in modified atmosphere

packages stored at improper temperatures supports enterohemorrhagic E. coli isolates to survive

gastric acid challenge. J Food Sci 78:148-153.

11. Coldewey, S. M., M. Hartmann, D. S. Schmidt, U. Engelking, S. N. Ukena, and F. Gunzer.

2007. Impact of the rpoS genotype for acid resistance patterns of pathogenic and probiotic

Escherichia coli. BMC Microbiol 7:21.

12. Cooley, M. B., D. Chao, and R. E. Mandrell. 2006. Escherichia coli O157:H7 survival and

growth on lettuce is altered by the presence of epiphytic bacteria. J Food Prot 69:2329-35.

13. Delaquis, P., S. Bach, and L. D. Dinu. 2007. Behavior of Escherichia coli O157:H7 in leafy

vegetables. J Food Prot 70:1966-74.

14. Delaquis, S., S. Stewart, S. Cazaux, and P. Toivonen. 2002. Survival and growth of Listeria

monocytogenes and Escherichia coli O157:H7 in ready-to-eat iceberg lettuce washed in warm

chlorinated water. J Food Prot 65:459-64.

15. DeLisa, M. P., C. F. Wu, L. Wang, J. J. Valdes, and W. E. Bentley. 2001. DNA microarray-

based identification of genes controlled by autoinducer 2-stimulated quorum sensing in

Escherichia coli. J Bacteriol 183:5239-47.

16. Francis, G. A., and D. O'Beirne. 2001. Effects of vegetable type, package atmosphere and

storage temperature on growth and survival of Escherichia coli O157:H7 and Listeria

monocytogenes. J Ind Microbiol Biotechnol 27:111-6.

17. Gawande, P. V., and M. W. Griffiths. 2005. Effects of environmental stresses on the activities

of the uspA, grpE and rpoS promoters of Escherichia coli O157:H7. Int J Food Microbiol 99:91-

18. Gnanamanickam, S., and J. Immanuel. 2006. Epiphytic bacteria, their ecology and functions,

p. 131-154. In S. Gnanamanickam (ed.), Plant-Associated bacteria. Springer, Dordrecht, The

Netherlands.

171

19. Grant, J., A. M. Wendelboe, A. Wendel, B. Jepson, P. Torres, C. Smelser, and R. T. Rolfs.

2008. Spinach-associated Escherichia coli O157:H7 outbreak, Utah and New Mexico, 2006.

Emerg Infect Dis 14:1633-6.

20. Heurlier, K., V. Denervaud, G. Pessi, C. Reimmann, and D. Haas. 2003. Negative control of

quorum sensing by RpoN (sigma54) in Pseudomonas aeruginosa PAO1. J Bacteriol 185:2227-

35.

21. ICMSF. 2005. Vegetables and vegetable products, p. 277-325. In ICMSF (ed.), Microbial

ecology of food commodities, 2nd. ed. Kluwer Academic/Plenum Publishers, New York.

22. Jay, M. T., M. Cooley, D. Carychao, G. W. Wiscomb, R. A. Sweitzer, L. Crawford-Miksza,

J. A. Farrar, D. K. Lau, J. O'Connell, A. Millington, R. V. Asmundson, E. R. Atwill, and R.

E. Mandrell. 2007. Escherichia coli O157:H7 in feral swine near spinach fields and cattle,

central California coast. Emerg Infect Dis 13:1908-11.

23. Kampfer, P., F. A. Rainey, M. A. Andersson, E. L. Nurmiaho Lassila, U. Ulrych, H. J.

Busse, N. Weiss, R. Mikkola, and M. Salkinoja-Salonen. 2000. Frigoribacterium faeni gen.

nov., sp. nov., a novel psychrophilic genus of the family Microbacteriaceae. Int J Syst Evol

Microbiol 50 Pt 1:355-63.

24. Kendall, M. M., D. A. Rasko, and V. Sperandio. 2007. Global effects of the cell-to-cell

signaling molecules autoinducer-2, autoinducer-3, and epinephrine in a luxS mutant of

enterohemorrhagic Escherichia coli. Infect Immun 75:4875-84.

25. Krimm, U., D. Abanda-Nkpwatt, W. Schwab, and L. Schreiber. 2005. Epiphytic

microorganisms on strawberry plants (Fragaria ananassa cv. Elsanta): identification of bacterial

isolates and analysis of their interaction with leaf surfaces. FEMS Microbiol Ecol 53:483-92.

26. Lang, M. M., L. J. Harris, and L. R. Beuchat. 2004. Survival and recovery of Escherichia coli

O157:H7, Salmonella, and Listeria monocytogenes on lettuce and parsley as affected by method

of inoculation, time between inoculation and analysis, and treatment with chlorinated water. J

Food Prot 67:1092-103.

172

27. Lee, S. Y., and S. Y. Baek. 2008. Effect of chemical sanitizer combined with modified

atmosphere packaging on inhibiting Escherichia coli O157:H7 in commercial spinach. Food

Microbiol 25:582-7.

28. Liao, C. H. 2007. Inhibition of foodborne pathogens by native microflora recovered from fresh

peeled baby carrot and propagated in cultures. J Food Sci 72:M134-9.

29. Liao, C. H., and W. F. Fett. 2001. Analysis of native microflora and selection of strains

antagonistic to human pathogens on fresh produce. J Food Prot 64:1110-5.

30. Lindow, S. E., and M. T. Brandl. 2003. Microbiology of the phyllosphere. Appl Environ

Microbiol 69:1875-83.

31. Loper, J. E., and M. D. Henkels. 1999. Utilization of heterologous siderophores enhances levels

of iron available to Pseudomonas putida in the rhizosphere. Appl Environ Microbiol 65:5357-63.

32. McKellar, R. C. 2007. Role of nutrient limitation in the competition between Pseudomonas

fluorescens and Escherichia coli O157: H7. J Food Prot 70:1739-43.

33. Mead, P. S., L. Slutsker, V. Dietz, L. F. McCaig, J. S. Bresee, C. Shapiro, P. M. Griffin, and

R. V. Tauxe. 1999. Food-related illness and death in the United States. Emerg Infect Dis 5:607-

25.

34. Muyzer, G., T. Brinkhoff, U. Nubel, C. Santegoeds, H. Schafer, and C. Wawer. 2004.

Denaturing gradient gel electrophoresis (DGGE) in microbial ecology, p. 743-770. In K. A.

Publishers (ed.), Molecular Microbial Ecology Manual, Second ed, vol. 13. Kluwer Academic

Publishers, Netherlands.

35. Muyzer, G., and K. Smalla. 1998. Application of denaturing gradient gel electrophoresis

(DGGE) and temperature gradient gel electrophoresis (TGGE) in microbial ecology. Antonie Van

Leeuwenhoek 73:127-41.

36. Nakatsu, C. H. 2007. Soil microbial community analysis using denaturant gradient gel

electrophoresis. Soil Sci Soc Am J 71:562-571.

173

37. Nakatsu, C. H., V. Torsvik, and L. Ovreas. 2000. Soil Community Analysis Using DGGE of

16S rDNA Polymerase Chain Reaction Products. Soil Sci Soc Am J 64:1382-1388.

38. Nguyen-the, C., and F. Carlin. 1994. The microbiology of minimally processed fresh fruits and

vegetables. Crit Rev Food Sci Nutr 34:371-401.

39. Ponder, M. A., S. J. Gilmour, P. W. Bergholz, C. A. Mindock, R. Hollingsworth, M. F.

Thomashow, and J. M. Tiedje. 2005. Characterization of potential stress responses in ancient

Siberian permafrost psychroactive bacteria. FEMS Microbiol Ecol 53:103-15.

40. Ragaert, P., F. Devlieghere, and J. Debevere. 2007. Role of microbiological and physiological

spoilage mechanisms during storage of minimally processed vegetables. Postharvest Biology and

Technology 44:185-194.

41. Rico, D., A. B. Martin-Diana, J. M. Barat, and C. Barry-Ryan. 2007. Extending and

measuring the quality of fresh-fruit and vegetables: a review. Trends in Food Sci and Tech

18:337-386.

42. Rudi, K., S. L. Flateland, J. F. Hanssen, G. Bengtsson, and H. Nissen. 2002. Development and

evaluation of a 16S ribosomal DNA array-based approach for describing complex microbial

communities in ready-to-eat vegetable salads packed in a modified atmosphere. Appl Environ

Microbiol 68:1146-56.

43. Schellhorn, H. E., and H. M. Hassan. 1988. Response of hydroperoxidase and superoxide

dismutase deficient mutants of Escherichia coli K-12 to oxidative stress. Can J Microbiol

34:1171-6.

44. Schuenzel, K. M., and M. A. Harrison. 2002. Microbial antagonists of foodborne pathogens on

fresh, minimally processed vegetables. J Food Prot 65:1909-15.

45. Shaw, R. K., C. N. Berger, B. Feys, S. Knutton, M. J. Pallen, and G. Frankel. 2008.

Enterohemorrhagic Escherichia coli exploits EspA filaments for attachment to salad leaves. Appl

Environ Microbiol 74:2908-14.

174

46. Sivapalasingam, S., C. R. Friedman, L. Cohen, and R. V. Tauxe. 2004. Fresh produce: a

growing cause of outbreaks of foodborne illness in the United States, 1973 through 1997. J Food

Prot 67:2342-53.

47. Sperandio, V., A. G. Torres, J. A. Giron, and J. B. Kaper. 2001. Quorum sensing is a global

regulatory mechanism in enterohemorrhagic Escherichia coli O157:H7. J Bacteriol 183:5187-97.

48. Takeuchi, K., A. N. Hassan, and J. F. Frank. 2001. Penetration of Escherichia coli O157:H7

into lettuce as influenced by modified atmosphere and temperature. J Food Prot 64:1820-3.

49. Valentin-Bon, I., A. Jacobson, S. R. Monday, and P. C. Feng. 2008. Microbiological quality of

bagged cut spinach and lettuce mixes. Appl Environ Microbiol 74:1240-2.

50. Whipps, J. M., P. Hand, D. Pink, and G. D. Bending. 2008. Phyllosphere microbiology with

special reference to diversity and plant genotype. J Appl Microbiol 105:1744-55.

51. White-Ziegler, C. A., S. Um, N. M. Perez, A. L. Berns, A. J. Malhowski, and S. Young. 2008.

Low temperature (23 degrees C) increases expression of biofilm-, cold-shock- and RpoS-

dependent genes in Escherichia coli K-12. Microbiology 154:148-66.

52. Yamano, Y., T. Nishikawa, and Y. Komatsu. 1998. Involvement of the RpoN protein in the

transcription of the oprE gene in Pseudomonas aeruginosa. FEMS Microbiol Lett 162:31-7.

53. Yang, C. H., D. E. Crowley, J. Borneman, and N. T. Keen. 2001. Microbial phyllosphere

populations are more complex than previously realized. Proc Natl Acad Sci U S A 98:3889-94.

175

Table 1. E. coli O157:H7 specific primers for real time PCR using SYBR green chemistry designed for this study.

Target gene Product length Description Primer sequence (5’-3’) (bp) rpoB 113 Beta subunit of RNA polymerase F: GAG CCG CCG ACT CGT GAA G R: CGC AGC AGA GAA CGG TTG AAC csgE 127 Cryptic curli major subunit F: TGG CGA CGA TTT AGT GCT TCT TC R: GCT GGA TCA CAA TAA CGG TCA ATC espA 178 Protein translocator filament F: GCT ATT ATT CAC TAC CGT TGT CAG R: GAA TGC GAA AGC CAA ACT TCC rpoN 110 Sigma (54) factor F: ACT GCC TGC TGA TCC AAC TCT C R: GCG GAA GTC GTG ATT GGC TAA C sodB 101 Superoxide dismutase F: GCT GAC CAC CGA TGC GAC TC R:AGA AGT GCT CAA GAT AGC CAG GAC rpoS 136 Sigma (38) factor F: CGT CAT CTT GCG TGG TAT CTT CC R: GTC AGC CGT ATG CTT CGT CTT AAC stxA 180 Shiga like toxin (A subunit) F: ATC AGG CAG GAC ACT ACT CAA CC R: CTG TGG CAA GAG CCGA TGT TAC G

176

Table 2. Total culturable epiphytic bacteria on rinsed or commercial non-inoculated spinach packaged under simulated retail conditions and, stored at 4oC and 10oC for 15 days. (Enumeration on R2A media at 25oC incubated for 48h)

Treatment on non Log fold increase Log fold increase Log CFU/g of spinach§ inoculated spinach day 1 to day10 day 1 to day 15 Storage day 0 5 10 15 Rinsed/4oC 4.53 + 0.21c,3 5.48 + 0.55c,4 6.51 + 0.24b,3 8.22 + 0.26a,1 1.81 3.69 Disinfected/4oC 1.40 + 0.10c,2 6.40 + 0.33b,3 6.85 + 0.08b,3 9.40 + 0.02a,2 5.45 8.00 Commercial/4oC 8.47 + 0.25b,1 8.57 + 0.14b,2 9.64 + 0.18a,1 N/D 1.17 N/D Rinsed/10oC 4.60 + 0.17c,3 4.75 + 0.43c,4 8.14 + 0.15b,2 9.67 + 0.10a,2 3.54 5.07 Disinfected/10oC 1.47 + 0.11d,2 4.10 + 0.22c,5 7.50 + 0.63b,3 9.25 + 0.10a,2 6.05 7.80 Commercial/10oC 9.18 + 0.08b,1 9.68 + 0.16b,1 10.20 + 0.11a,1 N/D 1.02 N/D

§Data represent means + standard deviations of three spinach bags and three measurements per bag.

Means with different letter within a row are significantly different (p<0.05)

Means with different number within a column are significantly different (p<0.05)

N/D Not determined due to spoilage at day 15.

177

Table 3. Identities of bacteria retrieved from 16SrDNA DGGE bands of spinach stored at refrigerated temperatures.

Database match Sequence Band gel Database match Sequence Band gel (GeneBank accession no. of closest match) similarity position (GeneBank accession no. of closest match) similarity position (%) (%) Chryseobacterium sp. (DQ279360) 99 1 Curtobacterium pusillum (DQ086779) 99 12 Flavobacterium sp. (AM412168) 98 1 Sphingobacterium sp. (FJ268960) 99 12 Clavibacter michiganensis (AJ310416) 93 1 Burkholderia sp. (FJ786047) 98 12 Uncultured proteobacterium clone (EF602176) 97 2 Sphingobacterium sp. (FJ268960) 99 13 Uncultured oxalobacteraceae (FJ037285) 97 2 Stenotrophomonas sp. maltophilia (FJ81185) 98 14 Uncultured baceroidetes bacterium 96 3 Bacillus sp. (AM293003) 99 14 (AM116743) Flavobacterium sp. (AY599661) 98 3 Uncultured bacterium (GQ099221) 97 14 Chryseobacterium sp. (EF471218) 86 4 Chryseobacterium sp. (EU07842) 98 14 Uncultured proteobacterium clone (EF019491) 88 4 Uncultured bacterium (EU772179) 95 15 Spinacia oleracea chloroplast (AJ400848) 96 5 Frigoribacterium faeri (AM410686) 99 15 Spinacia oleracea chloroplast (AJ400848) 98 6 Arthrobacter sp. (EU378900) 99 15 Flexibacteraceae bacterium (EU155017) 89 6 Sphingobacterium sp. (FJ378900) 99 15 Uncultured alpha Proteobacteria (CU926667) 95 7 Flavobacteriaceae bacterium (DQ173014) 99 15 Uncultured soil bacterium clone (DQ297980) 97 8 Sinorhizobium sp. (FJ687974) 91 16 Plantibacter flavus (EU939701) 99 9 Uncultured bacterium (FN421830) 96 16 Uncultured bacterium (FM874513) 99 9 Uncultured Deinococcus (FJ222450) 90 16 Uncultured actinobacterium (GQ144838) 98 9 Exiguobacterium sibiricum (CP001022) 99 17 Bradyrhizobium sp. (FJ959100) 99 9 Arthrobacter psychrolactophilus (AF134183) 92 18 Kineococcus sp. (YIM65377) 96 10 Uncultured bacterium (DQ125585) 97 18 Uncultured beta proteobacterium (EFO74569) 98 10 Arthrobacter sp. (D84563) 99 18 Terrabacter sp. (FJ772036) 99 10 Frigoribacterium sp. (AF157479) 99 19 Oryzihumus sp. (GQ355280) 99 10 Arthrobacter sp. (D84563) 99 19 Uncultured bacterium clone (FJ562178) 99 11 Uncultured bacterium (EU472696) 100 19 Uncultured actinobacterium clone (EF651015) 89 11 Bacterium B2 (GQ398340) 99 19 Stenotrophomonas sp. (AY689032) 99 20 Uncultured bacterium (AM697264) 99 20 Uncultured bacterium (GQ076113) 99 20

Band position in gel refers to figures 2 and 3

178

Table 4. Populations of culturable E. coli O157:H7 artificially inoculated on spinach, comparing samples that were rinsed or from commercial source, and stored at 4oC and 10oC. (Enumeration on sorbitol MacConkey agar, incubated at 37oC for48 h)

Treatment on Log fold increase day 1 to inoculated Log CFU/g of spinach¶ day 15 spinach Storage day 0 5 10 15 5.69 + 0.01a,1 4.31 + 0.32b,1 5.09 + 0.25a,2 5.50 + 0.08a,3 Rinsed /4oC -0.19 (19.5%) (0.56%) (0.14%) (0.31%) b,1 b,1 a,1 a,2 o 6.17 + 0.68 5.69 + 0.75 7.35 + 0.11 6.98 + 0.41 Disinfected 4 C (45.1%) (7.17%) (10.6%) (2.46%) 0.81 a,1 a,1 5.40 + 1.47 a,2 o 6.10 + 0.38 (0.12%) 5.90 + 0.96 ¥ Commercial/4 C (3.01%) (0.02%) N/D -0.20

a,1 a,1 a,1 a,2 o 5.70 + 0.07 5.14 + 0.28 6.83 + 0.35 6.45 + 0.15 Rinsed /10 C (11.88%) (2.18%) (2.33%) (1.04%) 0.76 b,1 b,1 b,2 a,1 o 5.49 + 0.30 4.55 + 0.58 6.02 + 0.16 9.43 + 0.14 * Disinfected/10 C (6.22%) (0.95%) (0.26%) (73.3%)* 3.94 a,1 a,1 a,1 o 6.10 + 0.10 6.10+ 0.32 6.60 + 0.40 ¥ Commercial/10 C (0.30%) (0.02%) (0.01%) N/D 0.50 ¶Data represent means + standard deviations of three spinach bags and three measurements per bag.

Means with different letter within a row are significantly different (p<0.05)

Means with different number within a column are significantly different (p<0.05)

¥Difference was calculated between days 1 and 10 of storage.

(*) Denotes significant difference (p<0.05) between day 1 and day 15 of storage

(%) Percentage of E. coli O157:H7 in the epiphytic population

N/D Not determined due to spoilage at day 15.

179

Table 5. Normalized expression of E. coli O157:H7 espA, csgA, sodB, rpoS, rpoN and stxA genes when inoculated onto spinach and stored at 4oC and 10oC over a10-day period assessed by qRT-PCR.

Treatment Rinsed/4oC espA csgA sodB rpoS rpoN stxA

Day 0 1.69 + 0.25b 0.71+ 0.05a 0.87+ 0.04b 1.54 + 0.16a 1.24 + 0.09a 1.07 + 0.15a Day 5 1.17 + 0.02b 0.77 + 0.28a 1.39 + 0.07a 1.24 + 0.01a 1.55 + 0.55a 7.14 + 0.17b Day 10 3.02 + 0.57a 2.13 + 0.54a 1.73 + 0.05a 2.35 + 0.73a 2.21 + 0.31a 1.94 + 0.17a Disinfected/4oC espA csgA sodB rpoS rpoN stxA

Day 0 0.41 + 0.04a 0.07 + 0.02a 0.28+ 0.14b 0.79 + 0.87b 0.40 + 0.24a 0.18 + 0.02a Day 5 1.01 + 0.00a 0.034 + 0.01a 1.01 + 0.00a 4.11 + 3.27a 0.26 + 0.17a 0.21 + 0.13a Day 10 0.24 + 0.04a 0.01 + 0.00a 0.24 + 0.04b 6.04 + 0.79a 0.42 + 0.26a 0.28 + 0.16a Rinsed/10oC espA csgA sodB rpoS rpoN stxA

Day 0 2.89 + 0.12b 2.07 + 0.06c 2.13+ 0.15b 2.86 + 0.41c 2.75 + 0.11b 2.54 + 0.02c Day 5 7.03 + 0.24a 5.28 + 0.25b 4.87 + 0.02a 5.87 + 0.25b 5.83 + 0.14a 5.88 + 0.25b Day 10 7.44 + 0.43a 7.01 + 1.12a 4.52 + 0.31a 10.49 + 0.77a 5.96 + 0.23a 10.5 + 0.77a Disinfected/10oC espA csgA sodB rpoS rpoN stxA

Day 0 0.56 + 0.59a 0.02 + 0.00a 0.33+ 0.02a 9.62 + 0.95a 0.17 + 0.02a 4.65 + 0.68a Day 5 0.07 + 0.06a 0.1 + 0.12a 0.20 + 0.08a 3.84 + 0.18b 0.09 + 0.07a 2.21 + 0.31b Day 10 0.05 + 0.02a 0.03 + 0.03a 0.60 + 0.18a 1.02 + 1.43c 0.28 + 0.19a 0.01 + 0.00c

§ Housekeeping gene rpoB was used to normalize the levels of mRNA of the studied genes. A value of one indicates expression levels equal to the housekeeping expression

(*) Data represent means + standard deviations of normalized mRNA gene expression levels from RNA isolated of two spinach bags and two measurements per bag.

Means with different letter within a column and treatment are significantly different (p<0.05).

180

Supplementary Table 1. Growth of epiphytic microbiota on bagged rinsed or commercial inoculated spinach, stored at 4oC and 10oC for 15 days.

(Enumeration on R2A media at 25oC incubated for 48h)

Treatment on Log fold increase day 1 to inoculated Log CFU/g of spinach§ day 15 spinach Storage day 0 5 10 15 o c,2 c,3 b,3 a,1 Rinsed /4 C 6.36 + 0.16 6.52 + 0.21 7.95 + 0.08 8.02 + 0.02 1.66

o c,1 a,2 Commercial/4 C 7.88 + 0.79 9.60 + 0.34 N/D 1.72¥ 8.63 + 0.07b,2 b b a a Disinfected/4°C 6.76 + 0.20 7.24 + 0.41 8.03 + 0.43 8.33 + 0.13 1.57 6.59 + 0.06b,2 6.79 + 0.19b,3 8.39 + 0.10a,3 8.50 + 0.17a,1 o Rinsed /10 C 1.91

8.57 + 0.04c,1 9.71+ 0.0.06b,1 10.45 + 0.05a,1 o N/D ¥ Commercial/10 C 1.88

6.79 + 0.17b 6.73 + 0.11b 8.65 + 0.39a 9.62 + 0.14a,* Disinfected/10°C 2.83

§Data represent means + standard deviations of three spinach bags and three measurements per bag.

Means with different letter within a row are significantly different (p<0.05)

Means with different number within a column are significantly different (p<0.05)

¥Difference was calculated between days 1 and 10 of storage.

^ Percentage of E. coli O157:H7 in the epiphytic population

181

Figure 1A. Percentage of head space pressures of O2 of packaged spinach stored at 4 and 10 °C during 15 days. Values represent the means of 10 bags and error bars represent the standard deviation from the mean.

182

Figure 1B. Percentage of head space pressures of CO2 of packaged spinach stored at 4 and 10 °C during

15 days. Values represent the means of 10 bags and error bars represent the standard deviation from the mean.

183

Figure 2. DGGE of PCR-amplified 16S rDNA fragments from non-inoculated packaged spinach. A.

Disinfected and stored at 4oC, B. Disinfected and stored at 10oC, C. rinsed and stored at 4°C, D. Rinsed and stored at 10°C. (d0 –day zero-, d5 –day 5, d10 –day 10, d15 –day 15).

184

Figure 3. UPGMA clustering of 16S rDNA DGGE profiles from inoculated and non-inoculated bagged spinach stored at 4oC or 10oC for a period of 15 days. (R) Rinsed spinach, (D) Disinfected spinach, (I) inoculated spinach with E. coli O157:H7, (d0 –day zero-, d5 –day 5-, d10 –day 10-, d15 –day 15-).

185

Figure 4. DGGE of PCR-amplified 16S rDNA fragments from packaged spinach inoculated with E. coli

O157:H7. A. Disinfected and stored at 4oC, B. Disinfected and stored at 10oC, C. rinsed and stored at

4°C, D. Rinsed and stored at 10°C. (d0 –day zero-, d5 –day 5-, d10 –day 10-, d15 –day 15-).

186

CHAPTER 6

PYROSEQUENCING OF 16S rRNA GENE AMPLICONS FOR

CHARACTERIZATION OF THE PHYLLOEPIPHYTIC BACTERIA OF

MINIMALLY PROCESSED SPINACH STORED AT REFRIGERATION

TEMPERATURES

ABSTRACT

Spinach was chosen as a model to examine alterations to the community structure and population dynamics of phylloepiphytic bacteria associated with packaging and refrigeration of ready to eat fresh produce using pyrosequencing of 16s rRNA gene amplicons. The fresh and packaged spinach phylloepiphytic communities were diverse harboring a wide variety of members. Sequences were assigned to 11 different phyla with the largest number of reads belonging to Proteobacteria followed by

Firmicutes, Actinobacteria and Acidobacteria. Other phyla represented in lower relative abundance were

Planctomycetes, Verrucomicrobia and TM7, which contained novel genera not previously described in the phyllosphere. Spinach packaging and storage affected the richness, diversity and evenness of the spinach bacterial community. Overall packaging and refrigeration resulted in a decrease in richness and diversity of the spinach phylloepiphytic community. Storage at both temperatures was associated with increases in members of the Enterobacteriaceae and the genera Pseudomonas spp. Refrigeration was associated with increases in the relative abundance of several families known to contain psychrotrophs, while the relative abundance of sequences from mesophilic bacteria decreased. The application of pyrosequencing allowed a broader description of the composition and diversity of fresh spinach and allowed in greater detail to identify the changes in bacterial composition associated with spinach packaging and time of storage.

Key words: fresh spinach, packaged spinach, pyrosequencing, microbiome, phylloepiphytic bacterial community, low temperature.

187

INTRODUCTION

The leaf surfaces of above ground plants, known as the phyllosphere, are inhabited by a community of bacteria, archaea, filamentous fungi and yeast. Bacteria are considered to be the dominant microbial inhabitants with culturable bacteria numbering between 102 to 1012 cells per gram of leaf (58).

The global surface area occupied by the phyllosphere is estimated to total over 4 x 108 Km2, supporting an estimated bacterial population of 1026 cells (44). Bacteria belonging to more than 85 different species in

37 genera have been cultured from the phyllospheres of rye, olive, sugar beet and wheat (21, 58). Recent studies using non-culture dependent techniques reveal that the diversity of the phyllosphere is much greater than described by using culture dependent analyses (39, 63). These phyllosphere bacteria have important roles in promoting plant growth and suppressing colonization of plant and human pathogens.

Recent studies have shown that the members of these microbial communities interact and even improve epiphytic fitness of human pathogens on the plant surface (42).

Recent outbreaks of human disease associated with fresh produce have highlighted the vulnerability of fresh market production to colonization by foodborne pathogens such as, Escherichia coli

O157:H7 and Salmonella enterica. In 2006, a multistate outbreak of E. coli O157:H7 was traced back to consumption of spinach grown in California, packaged and shipped across the United States and Canada

(5, 6). Packaging of leafy greens is a common technology utilized to increase shelf life of leafy greens due to a decreased plant respiration rate. In combination with storage at refrigeration temperatures, growth of spoilage associated microbes can also be controlled (52).

The role of packaging and storage at low temperature is known to reduce growth, though not eliminate human pathogens on inoculated produce (13). Little is known how these hurdles affect the phylloepiphytic microbial communities, which are adapted to life on the plant leaf surface. As these pre- harvest associated microbiota remain associated with raw vegetables during transportation, further

188 processing and storage (26), it is important to examine the effect of packaging and low temperature storage on epiphytic bacteria as they may offer a strategy to control foodborne pathogens.

Culture independent approaches to study phyllosphere communities have changed the understanding of the structure and diversity of the phyllosphere, and at the same time they have contributed to find new insights into the complexity of the phyllosphere and their interactions with plants and the environment (61). Pyrosequencing is a recently developed technology in which more than

300,000 sequences can be simultaneously determined without cloning. A highly variable region of 16S rRNA gene is amplified using primers that target adjacent conserved regions, followed by direct sequencing of individual PCR products (41, 54). The methodology has been used to describe microbial communities from deep sea environments (23, 57), intestinal tract (3, 15), soils (53, 59), fermented foods

(25) and the rhizosphere (29).

In this study, we investigated through 454-pyrosequencing of bacterial 16S rRNA gene amplicons, the composition and richness of spinach microbiomes obtained from fresh and packaged spinach stored at refrigeration temperatures (4 and 10°C). The purpose of this study was to characterize the bacteria composition on fresh spinach leaves and to evaluate changes to the bacterial community composition.

189

MATERIALS AND METHODS

DNA isolation.

Spinach production and the packaging of spinach have been previously described (42). Briefly, a flat cultivar of spinach (Monza, Seedway LLC. USA) was grown using conventional agricultural practices at the Virginia Tech Kentland farm. Leaves were harvested, rinsed with sterile water to remove soil and packaged within two hours after harvest. Ten grams of spinach were subjected to four different treatments: no treatment (fresh spinach), packaged and stored at 4°C for 1 day, packaged and stored at

4°C for 15 days and packaged and stored at 10°C for 15 days. DNA was isolated from fresh spinach and from three different packages for each treatment and extraction from each package and from fresh spinach was completed in triplicate. The leaves were aseptically transferred into Pulsifier® bags (Microbiology

International, Frederick, MD) containing 90 mL of 1% (wt/vol) sterile peptone water (Sigma-Aldrich

Co., St Louis MO) supplemented with 1% (vol/vol) of Tween-90 (PTW) (Fisher, Atlanta, GA) and samples were pulsified for 5 minutes to detach epiphytic microbiota. Bacterial cells were collected from the bacterial suspension by centrifugation at 4,000 rpm/20min/4oC washed with 1X PBS and resuspended in 100 L of 1X TE buffer. The cells were lysed by incubation with 300 g of lysozyme (Fisher, Fair

Lawn, NJ), 10 units of mutanolysin (Fisher) and 25 g of achromopeptidase (Sigma-Aldrich Co., St Louis

MO) at 37oC for 30 min. Lysates were treated with 25units of proteinase K (Fisher) and incubated at

65oC for 30 min. DNA was extracted from the lysed cells of each sample using the ZR soil microbe DNA kitTM (Zymo Research Co., Orange, CA) per manufacturer’s instructions.

The total number of 16S rRNA copies was assessed by real-time PCR using universal primers as described by Fierer et al. (16). Samples were adjusted to 5 to 6 log10 copies of the16S rRNA gene prior to amplification and further pyrosequencing of the 16S rRNA gene.

190

PCR amplicons library construction.

A 237 nucleotide sequence of the V4 region of the 16S rRNA gene (Escherichia coli positions

565_to_802) was amplified by PCR (9). The 16s rRNA gene was amplified using the forward primer 5’-

GCCTCCCTCGCGCCATCAGAYTGGGYDTAAAGNG-3’ where the underlined sequence is that of 454

Life Science primer A (Roche) and the italicized sequence is of the broad range V4 region primer.

Separate amplifications were performed with three reverse primers as described by Jesus et al., 2010 (29).

5’-GCCTTGCCAGCCCGCTCAGTACCRGGGTHTCTAATCC-3’,

5’-GCCTTGCCAGCCCGCTCAGCTACDSRGGTMTCTAATC-3’,

5’-GCCTTGCCAGCCCGCTCAGTACNVGGGTATCTAATCC-3’ where the underlined sequence is that of 454 Life Science primer B and the italicized sequence for the universal reverse primers

(http://pyro.cme.msu.edu/pyro/help.jsp). Reactions were performed in triplicate with each primer set. The amplicons PCR products were generated by amplifying 50 ng of genomic DNA. Each 25 L PCR reaction contained 1.5mM of MgCl2, 50mM of KCl, 0.2mM of each dinucleotide, 1% of DMSO

(dimethylsulfoxide), 25mM of Tris-HCl (pH 8), 1U/ L of HotStart-IT FideliTaq DNA polymerase (USB

71156, Cleveland, OH, USA), 0.5 M of each primer. The PCR protocol consisted of denaturation at

95oC for 5 minutes, followed by 30 cycles of 95oC for 30 s, amplification at 50oC for 45 s and elongation at 72oC for 1 min, followed by a final elongation step at 72oC for 7 min. Products of each primer set were cleaned using the Qiagen minElute PCR purification kit (Qiagen, Valencia CA) per manufacturer’s instructions. Products of these three primer sets were then pooled in equal concentrations to create amplicon libraries (33). The concentration and quality of the products were assessed using a DNA1000

LabChip on the Bioanalyser 2100 (Agilent, Palo Alto CA). Only sharp, distinct amplification products with a total yield of >200 ng were used for 454 pyrosequencing.

Four separate libraries were constructed for each of the following conditions: Fresh spinach, packaged spinach stored at 4°C for 1 day, packaged spinach stored for 15 days at 4°C and packaged spinach stored for 15 days at 10°C. The amplicons libraries were bound to beads with two DNA

191 molecules per bead. The beads were taken through the emPCR process using the Amplicon Primer A kit

(Roche). After breaking the emulsion, the DNA strands were enriched and prepped for deposition onto the PicoTiter Plate with an eight gasket format (454 Life Sciences) to separate each individual sample.

Pyrosequencing was done on a Genome Sequencer GS FLX system (LR 700 100 cycles, Roche, Base,

Switzerland).

Pre-processing of sequence reads.

The reads obtained from GS-FLX were preprocessed to identify sequencing errors. Sequences were eliminated if they did not contain an exact match with the forward primer at the proximal end, if they were less than 50 nt before reaching reverse primer on the distal end, and if they did not contain at least 12 bases of the reverse primer before the read ended. Reads were then grouped into clusters containing other sequences that were 99% similar over 99% of their lengths. A unique representative read sequence was selected randomly from each cluster (unique sequence) for further taxonomic classification and OTU assignment.

Taxonomic Assignment of sequence reads.

Metagenomes of fresh and packaged spinach were analyzed using the RDP pyrosequencing pipeline (http://pyro.cme.msu.edu) release 10 (9). Each unique sequence was aligned using the RDP pyrosequencing function, Aligner (45). Aligned data sets were clustered using the default parameters for the RDP Clustering function. The resulting clusters were utilized to calculate the Shannon index, Chao1 estimator and rarefraction curves using the pyrosequencing analysis tools from RDP at the level of 3% dissimilarity, approximately species level. The RDP sample abundance statistics tool was utilized to calculate the Jaccard’s index to compare the four spinach microbiomes (7). Aligned data sets from each microbiome were merged into a single cluster file, which was utilized to construct a distance matrix at 3% dissimilarity, that produce a dendrogram using unweighted pair group method with arithmetic mean

(UPGMA) . The RDP Classifier, a Bayesian rRNA classifying algorithm (60), was used to assign

192 phylogenetic groups, based on sequence similarity. Matches with a RDP confidence estimate below 60% were designated as unclassified bacteria.

RESULTS

Characteristics of the sequenced data. The average sequence length of the pyrosequencing data was 200 bp (range, 100-281). The total number of reads for each sample in each gasket ranged from

14,000 to 24,000 (Table 1) depending on the sample. Approximately 4,000 reads which did not meet the quality parameters were trimmed. Sequence reads that belonged to chloroplast sequences (3 to 15% of reads) were not considered for analysis as those were considered contamination from the spinach plant.

The majority of the unique sequences belonged to the domain Bacteria (99% of total sequences).

Sequences obtained from the fresh spinach samples and packaged spinach stored at 4°C for 1 day were assigned to the domain Bacteria but no further classification of the sequences could be made for 75% and

85% of the sequences respectively. For samples stored at 4 and 10°C for a period of 15 days, the percentage of classified bacteria increased to 60 and 90%, respectively.

Fresh spinach microbiome. Bacterial sequences classified for fresh spinach were assigned to 11 phyla (Fig. 1A-B). The majority of the sequences belonged to Proteobacteria (80%), particularly the classes of -Proteobacteria and -Proteobacteria. Bacteria belonging to the phyla Firmicutes,

Chlamydiae, Gemmatimonadetes, Verrucomicrobia, TM7, Deinococcus-Thermus, Planctomycetes,

Actinobacteria, Acidobacteria and Bacteroidetes were also identified (Fig. 1A-B). From all the classified sequences, 54% were attributed to a genus and 46% to a family. Within the identified phyla, 75 different genera were identified (Table 2). The most represented sequences (each representing more than 1% of the total classified bacteria) belonged to genera Spartobacteria genera incertae sedis, Deinococcus spp., Gp4,

Gp6, Methylobacterium spp., Rhizobium spp., Brevundimonas spp., Acinetobacter spp., Pseudomonas spp., Ralstonia spp., Sphingomonas spp., Naxibacter spp. and Massilia spp.. The most numerous sequences were identified as Pseudomonas spp. (9%).

193

Estimates of the operational taxonomic units (OTU’s) were assessed using of rarefraction curves

(Fig. 2A) and a non parametric method; Chao 1 (Table 3). Rarefraction curves reached saturation at a distance level of 20% (phylum level) but not at 3 or 5% of dissimilarity.

Changes in bacterial community composition during storage of packaged spinach at refrigeration temperatures. Storage at refrigeration temperatures affected the microbiome composition.

Microbiomes of fresh and packaged spinach held at refrigerated temperatures were compared using

Jaccard’s index. Fresh spinach and packaged spinach stored at 4°C for one day were most similar compared to the samples held at 4 or 10°C for 15 days, which formed separate clusters (Fig 3). Chao 1 estimator, Shannon’s index and rarefraction curves indicated a reduction in richness, evenness and diversity after spinach was packaged and held at 4°C or at 10°C for 15 days (Table 3 and Fig. 2A-D). An increase in Shannon’s index was calculated for packaged spinach held at 4°C for 15 days and was associated with decreased evenness.

Reductions in the richness, evenness and diversity of the community are attributed to reduced abundance of sequences belonging to all eleven phyla, with the exception of Proteobacteria, due to an increase in relative abundance of -Proteobacteria (Fig 1A). These reductions were most pronounced for spinach held at 10°C after 15 days of storage (Fig 1B), particularly the families of Oxalobacteraceae,

Rhizobiales, Acinetobacter and Sphingomonadaceae. Despite an overall reduction in relative abundance of sequences belonging to the families Enterobacteriaceae and Micrococcineae, the genera of

Pseudomonas spp., Stenotrophomonas spp., Pantoea spp., and Escherichia spp. (Table 4) increased in abundance in the microbiome of spinach held at 10°C for 15 days. The microbiome of spinach stored at

4°C for 15 days was dominated by sequences belonging to the family Enterobacteriaceae (52% of the total classified bacteria) and the genera Pseudomonas spp. (27 % of total classified bacteria), while the relative abundance of Deinococcus spp. and Sphingomonas spp. reduced (Table 4).

194

DISCUSSION

In this study, pyrosequencing of the 16S rRNA gene amplicons was utilized to investigate the composition of the phylloepiphytic bacterial community of fresh spinach and to identify the effects of typical retail storage conditions (4°C) or temperature abuse conditions (10°C) on the composition and diversity of the spinach bacterial community. Pyrosequencing of 16S rRNA genes allowed identification of 11different phyla present in spinach microbiome (Fig 1A, B), expanding our knowledge about the diversity and complexity of the phyllosphere. It is likely that the phyllosphere microbiome is even more diverse and includes bacterial divisions not yet described. This is indicated by the large numbers of sequences (75% of fresh and 85% packaged spinach held at 4°C for day1) that could not be classified further than the Bacteria domain. These sequences may represent bacterial sequences that have not been placed into a recognized phylum within the RDP database or represent bacterial divisions not yet described (3).

Pyrosequencing of the spinach microbiome produced ~20,000 reads; however the numbers of bacterial sequences generated were reduced by the amplification of chloroplasts from contaminating plant

DNA. Between 3-15% of the sequences were identified as chloroplast, despite the use of a pulsifier to dissociate epiphytic bacteria without disrupting the leaf tissue (18). The amplification of chloroplast sequences by 16s rDNA primers has been previously reported and may create bias within metagenomic libraries as they might contain a high proportion of plant genomic DNA (30).

Despite these limitations, pyrosequencing showed the spinach microbiome was highly diverse harboring a wide variety of phyla and bacteria genera (Table 2). The majority of sequences were identified at the phyla level due to the small size of the amplified fragment (~200bp) (62). This is similar to other culture independent methodologies such as DGGE which amplify 300-600bp fragments which limits the ability to discriminate beyond the genus taxonomic level (32, 42, 48, 63).

The spinach microbiome is dominated by members of the phyla of Proteobacteria.

Proteobacterial sequences are abundant in the phyllosphere of Thlapsi geosingenese (27), Trichilia

195 catigua (39), Zea mays (32), Capsicum annum (50) and Solanum tuberosum (49), suggesting that

Proteobacteria is the dominant phylum of above ground plants. The most abundant sequences in this study, Sphingomonas spp., Brevundimonas spp., Methylobacterium spp., Acinetobacter spp.,

Pseudomonas spp., Naxibacter spp.,and Massilia spp., have been previously detected through analysis of

16S rRNA gene in the phyllospheres of other crop plants (28, 32, 36, 63, 65). The composition of the spinach microbiome was similar to the phyllospheres of soybean, clover and Arabidopsis thaliana, which were recently characterized by pyrosequencing. However, the dominant class within the Proteobacteria phyla was -Proteobacteria (14), while -Proteobacteria was the most abundant class in this study.

Additionally more OTU’s (approximately 1000 OTU’s) were detected within the spinach phyllosphere as evidenced by rarefraction curves (Fig 2A, Table 2) compared to these phyllospheres (approximately 60

OTU’s). In the same study, proteomic analysis revealed that the most abundant proteins belonged to

Methylobacterium spp. Sphingomonas spp. and Pseudomonas spp., which suggested that these are highly active in the phyllosphere (14). Sequences identified as these bacteria were among the most abundant members of the fresh spinach phyllosphere along with some species of Deinococcus-Thermus, -

Proteobacteria and Acidobacteria (Table 2). The lower relative abundances of sequences belonging to the phyla Chlamydiae, Gemmatimonadetes, Verrucomicrobia, TM7, Planctomycetes, Actinobacteria,

Acidobacteria and Bacteroidetes were comparable to that reported in phyllospheres of other plants (28,

64). The similarities of composition of these different plants suggests that phyllosphere has a defined composition of bacterial species, and variations in the individual members can be attributed to specific plant species and other environmental factors that may determine the establishment of epiphytes on plant surfaces (28, 61). Additionally, the spinach microbiome contained novel sequences for genera within the phyla Planctomycetes, Verrucomicrobia and TM7 not previously described as associated with the phyllosphere (Table 2). These phyla have been described mostly through culture independent techniques

(12, 51), but recent studies suggest that these members although represented in lower abundance might have ecological importance within the bacterial community; Planctomycetes and the division TM7 have

196 been reported to be ubiquitous in different aquatic and terrestrial environment, with the ability to survive and grow under a wide range of conditions (24, 37). Verrucomicrobia is a prevalent member of the rhizosphere and it was recovered from soils, possessing characteristics broadly distributed among rhizosphere members; slow growth, aerobic and heterotrophic (12).

Comparison with metagenomes analyzed through pyrosequencing for other environments suggests that the richness of the spinach phyllosphere, estimated using total OTU’s and rarefraction curves, was lower than other environments, like soil, whose richness is estimated in a range of 3000 to

6000 OTU’s and is often dominated by Actinobacteria and -Proteobacteria (2, 53), marine environment often composed by -Proteobacteria and its richness estimated in more than 9000 OUT’s (19, 23). But its richness was higher than the microbiome of human gut whose richness has been estimated in about 300

OTU’s and is dominated by Firmicutes (3).

Members of the spinach phyllosphere microbiome have been shown to have functional importance within the phyllosphere and with their plant host. Certain Pseudomonas spp. and

Brevundimonas spp. species are well known as plant symbionts promoting plant growth (47) and nitrogen fixation (43). Species like Methylobacterium spp. and Sphingomonas spp. are often found associated with plants (22, 28), showing specific abilities to successfully colonize leaves as epiphytes, such as pigment production for UV protection and utilization of specific carbon sources present on the leaf surfaces (14,

34) which favor their fitness in the phyllosphere. For example species of these phyla like Flavobacterium spp. and Corynebacterium spp. have shown the ability to fix nitrogen in maize plants (20). Deinococcus spp. has shown ability to inhibit fungal plant diseases in rice leaves (64). Ralstonia spp. and Massilia spp. have shown ability to participate in quorum sensing event in the rhizosphere (11).

Changes associated with minimal processing of spinach in bacterial community composition were evaluated. Previous studies by this research group indicate that length and temperature of storage of packaged spinach was associated with changes to the spinach bacterial community (42). Application of

197

DGGE community analysis indicates that the overall species richness decreases with prolonged storage at

4 or 10°C. Examination of these communities using pyrosequencing allowed a more in depth identification of the members responsible for the observed changes to the richness and evenness of the community as a result of processing hurdles.

Overall packaging and refrigeration resulted in a decrease in richness, evenness and diversity of the spinach bacterial community. Membership of the spinach microbiome was the most similar to fresh samples and spinach held at 4°C (Fig 3) despite decreased richness and evenness. Substantial decreases in richness and evenness were seen when spinach was stored for 15 days at 4°C or 10°C (Table 3). These microbiomes are composed of larger numbers of sequences belonging to genera associated with psychrotrophic growth, including members of the family Enterobacteriaceae, Corynebacterineae ,

Micrococcineae and the genera Pseudomonas, Methylobacterium and Pantoea (Fig. 1A and Table 4).

Previous studies have noted that psychrotrophic bacteria, particularly Pseudomonas sp., Erwinia sp.,

Rhanella sp., and lactic acid bacteria, tend to dominate the culturable populations on vegetables (48, 55).

Many of these bacteria have been reported as responsible for spoilage and deterioration of fresh vegetables (26, 35, 46).

In contrast the abundance of sequences within the class of - and Proteobacteria,

Acidobacteria and Deinococcus-Thermus decreased with refrigeration as compared with fresh spinach, which suggests that these members are prone to negative selection under low temperatures. These results were most significant after 15 days of storage, suggesting that length of storage and the temperature of storage led to the decreased populations of these bacteria. This effect could also be related with increased dominance of Proteobacteria, which can outcompete members of these genera or due to the larger relative abundance of sequences belonging to these genera, these phyla were not detected.

Of particular concern was the increase in relative abundance for Escherichia spp. (23% of the total classified sequences) during storage at 10°C (Table 4). It is likely that the time and temperature of storage increase fitness of this genus, which might be a risk if species like E. coli O157:H7 land on edible

198 plants prior to packaging and they are not held at the correct temperature. Thus the cold chain for packaged and ready-to-eat produce plays an important role in the control of Escherichia spp. Studies on packaged ready-to-eat leafy greens have indicated that growth of E. coli O157:H7 below 8°C is unlikely to occur and viability is often reduced (13). However its population is still viable after several days of storage (1, 17). In packaged shredded lettuce inoculated with 4 log CFU/g of E. coli O157:H7 stored at 4 and 8°C for 12 days, showed that at 4°C the E. coli populations did not increase significantly, however an increase of 2 log CFU/g was detected in spinach stored at 8°C (17), indicating the importance to avoid temperature abuse to prevent the growth of pathogenic bacteria. Other risks could also be related to the increase of Stenotrophomonas spp. and Massilia spp., which contain species that can be opportunistic pathogens for immunocompromised humans (8, 38).

In the establishment of enteric pathogens like E. coli O157:H7, bacteria-bacteria interactions with members of the phyllosphere can influence their fate on edible plants (61). Studies with Pseudomonas spp. (56), Bacillus spp. (40), Enterobacter spp.(10) and Pantoea spp. (31) have shown that these genera contain species that elicit, in vitro, antagonism towards enteric pathogens like E. coli O157:H7 and S. enterica Some of these genera were detected in the spinach microbiomes analyzed in this study and some of them showed to be predominant members of the bacterial community. Then opportunities for bio- control using these members that are well adapted to the phyllosphere environment during pre- and post- harvest operations are certainly likely. In addition it is important to point out that in fresh as well as packaged spinach, native microbiota and human pathogens coexist and might share the same niche.

Previous work showed that E. coli O157:H7 is able to establish commensal interaction in vitro with members of the spinach phyllosphere (Chapter 4), due to differences in carbon source utilization which allow them co-existence. This suggests that microbial interactions that occur post-processing and during storage in produce might play a major role in the survival of pathogenic bacteria (4), especially if conditions favor their growth.

199

CONCLUSIONS

The application of pyrosequencing to describe composition and diversity of the phyllosphere on spinach leaves provided a broader outlook of the bacterial composition of this community complementing other phyllosphere studies that have used culture- and non-culture depending approaches. This study showed that the spinach microbiome is highly diverse and changes dynamically during storage at refrigeration temperatures. Further research is necessary to correlate the composition and changes in diversity with functionality of the phyllosphere community to better comprehend the microbial ecology that occur on the phyllosphere of edible plants.

ACKNOWLEDGEMENTS

This work was supported in part by the Virginia Bioinformatics Institute and Fralin Life Science

Institute Exploratory Grant. Fellowship for partial financial support of Gabriela Lopez-Velasco was provided by the National Council for Science and Technology (CONACyT) from Mexico.

200

REFERENCES

1. Abdul-Raouf, U. M., L. R. Beuchat, and M. S. Ammar. 1993. Survival and growth of

Escherichia coli O157:H7 on salad vegetables. Appl Environ Microbiol 59:1999-2006.

2. Acosta-Martinez, V., S. E. Dowd, Y. Sun, and V. Allen. 2008. Tag-encoded pyrosequencing

analysis of bacterial diversity in a single soil type as affected by management and land use. Soil

Biol Biochem 40:2762-2770.

3. Andersson, A. F., M. Lindberg, H. Jakobsson, F. Backhed, P. Nyren, and L. Engstrand.

2008. Comparative analysis of human gut microbiota by barcoded pyrosequencing. PLoS One

3:2836.

4. Bhagwat, A. A. 2006. Microbiological safety of fresh-cut produce: Where are we now?, p. 121-

165. In K. J. Matthews (ed.), Microbiology of fresh produce. ASM Press, Washington, DC.

5. CDC 2006, posting date. Multistate outbreak of E. coli O157 infections. November-December

2006. www.cdc.gov/ecoli/2006/december/121006.htm. [Online.]

6. CDC. 2006. Update of multistate outbreak of Escherichia coli O157:H7 infections from fresh

spinach. CDC website.

7. Chao, A., R. L. Chazdon, R. K. Colwell, and T. J. Shen. 2006. Abundance-based similarity

indices and their estimation when there are unseen species in samples. Biometrics 62:361-71.

8. Chatelut, M., J. L. Dournes, G. Chabanon, and N. Marty. 1995. Epidemiological typing of

Stenotrophomonas (Xanthomonas) maltophilia by PCR. J Clin Microbiol 33:912-4.

9. Cole, J. R., Q. Wang, E. Cardenas, J. Fish, B. Chai, R. J. Farris, A. S. Kulam-Syed-

Mohideen, D. M. McGarrell, T. Marsh, G. M. Garrity, and J. M. Tiedje. 2009. The

Ribosomal Database Project: improved alignments and new tools for rRNA analysis. Nucleic

Acids Res 37:141-5.

10. Cooley, M. B., D. Chao, and R. E. Mandrell. 2006. Escherichia coli O157:H7 survival and

growth on lettuce is altered by the presence of epiphytic bacteria. J Food Prot 69:2329-35.

201

11. d'Angelo-Picard, C., D. Faure, I. Penot, and Y. Dessaux. 2005. Diversity of N-acyl

homoserine lactone-producing and -degrading bacteria in soil and tobacco rhizosphere. Environ

Microbiol 7:1796-808.

12. da Rocha, U. N., L. van Overbeek, and J. D. van Elsas. 2009. Exploration of hitherto-

uncultured bacteria from the rhizosphere. FEMS Microbiol Ecol 69:313-28.

13. Delaquis, P., S. Bach, and L. D. Dinu. 2007. Behavior of Escherichia coli O157:H7 in leafy

vegetables. J Food Prot 70:1966-74.

14. Delmotte, N., C. Knief, S. Chaffron, G. Innerebner, B. Roschitzki, R. Schlapbach, C. von

Mering, and J. A. Vorholt. 2009. Community proteogenomics reveals insights into the

physiology of phyllosphere bacteria. Proc Natl Acad Sci U S A 106:16428-33.

15. Dowd, S. E., T. R. Callaway, R. D. Wolcott, Y. Sun, T. McKeehan, R. G. Hagevoort, and T.

S. Edrington. 2008. Evaluation of the bacterial diversity in the feces of cattle using 16S rDNA

bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP). BMC Microbiol 8:125.

16. Fierer, N., J. A. Jackson, R. Vilgalys, and R. B. Jackson. 2005. Assessment of soil microbial

community structure by use of taxon-specific quantitative PCR assays. Appl Environ Microbiol

71:4117-20.

17. Francis, G. A., and D. O'Beirne. 2001. Effects of vegetable type, package atmosphere and

storage temperature on growth and survival of Escherichia coli O157:H7 and Listeria

monocytogenes. J Ind Microbiol Biotechnol 27:111-6.

18. Fung, D., A. N. Sharpe, B. Hart, and Y. Liu. 1998. The pulsifier: A new instrument for

preparing food suspensions for microbiological analysis. J Rapid Meth and automation in

Microbiol 6:43-49.

19. Gilbert, J. A., D. Field, P. Swift, L. Newbold, A. Oliver, T. Smyth, P. J. Somerfield, S. Huse,

and I. Joint. 2009. The seasonal structure of microbial communities in the Western English

Channel. Environ Microbiol 11:3132-9.

202

20. Giri, S., and B. R. Pati. 2004. A comparative study on phyllosphere nitrogen fixation by newly

isolated Corynebacterium sp. & Flavobacterium sp. and their potentialities as biofertilizer. Acta

Microbiol Immunol Hung 51:47-56.

21. Hirano, S. S., and C. D. Upper. 2000. Bacteria in the leaf ecosystem with emphasis on

Pseudomonas syringae a pathogen, ice nucleus, and epiphyte. Microbiol Mol Biol Rev 64:624-

53.

22. Hirano, S. S., and C. D. Upper. 1991. Bacterial community diynamics, p. 271-313. In J. H.

Andrews and S. S. Hirano (ed.), Microbial ecology of leaves. Springer-Verlag, New York.

23. Huber, J. A., D. B. Mark Welch, H. G. Morrison, S. M. Huse, P. R. Neal, D. A. Butterfield,

and M. L. Sogin. 2007. Microbial population structures in the deep marine biosphere. Science

318:97-100.

24. Hugenholtz, P., G. W. Tyson, R. I. Webb, A. M. Wagner, and L. L. Blackall. 2001.

Investigation of candidate division TM7, a recently recognized major lineage of the domain

Bacteria with no known pure-culture representatives. Appl Environ Microbiol 67:411-9.

25. Humblot, C., and J. P. Guyot. 2009. Pyrosequencing of tagged 16S rRNA gene amplicons for

rapid deciphering of the microbiomes of fermented foods such as pearl millet slurries. Appl

Environ Microbiol 75:4354-61.

26. ICMSF. 2005. Vegetables and vegetable products, p. 277-325. In ICMSF (ed.), Microbial

ecology of food commodities, 2nd. ed. Kluwer Academic/Plenum Publishers, New York.

27. Idris, R., R. Trifonova, M. Puschenreiter, W. W. Wenzel, and A. Sessitsch. 2004. Bacterial

communities associated with flowering plants of the Ni hyperaccumulator Thlaspi goesingense.

Appl Environ Microbiol 70:2667-77.

28. Jackson, E. F., H. L. Echlin, and C. R. Jackson. 2006. Changes in the phyllosphere community

of the resurrection fern, Polypodium polypodioides, associated with rainfall and wetting. FEMS

Microbiol Ecol 58:236-46.

203

29. Jesus, E. d. C., E. Susilawati, S. L. Ssmith, Q. Wang, B. Chai, R. Farris, J. L. M. Rodrigues,

K. D. Thelen, and J. M. Tiedje. 2010. Bacterial communities in the rhizosphere of biofuel crops

grown on marginal lands as evaluated by 16S rRNA gene pyrosequences. Bioenerg Res 3:20-27.

30. Jiao, J. Y., H. X. Wang, Y. Zeng, and Y. M. Shen. 2006. Enrichment for microbes living in

association with plant tissues. J Appl Microbiol 100:830-7.

31. Johnston, M. A., M. A. Harrison, and R. A. Morrow. 2009. Microbial antagonists of

Escherichia coli O157:H7 on fresh-cut lettuce and spinach. J Food Prot 72:1569-1575.

32. Kadivar, H., and A. E. Stapleton. 2003. Ultraviolet radiation alters maize phyllosphere bacterial

diversity. Microb Ecol 45:353-61.

33. Kanagawa, T. 2003. Bias and artifacts in multitemplate polymerase chain reactions (PCR). J

Biosci Bioeng 96:317-23.

34. Knief, C., A. Ramette, L. Frances, C. Alonso-Blanco, and J. A. Vorholt. 2010. Site and plant

species are important determinants of the Methylobacterium community composition in the plant

phyllosphere. Isme J. In press.

35. Kraft, A. A. 1992. Spoilage of dairy products, vegetables, fruits and other foods., p. 113-120. In

A. A. Kraft (ed.), Psychrotrophic bacteria in foods. Disease and spoilage. CRC Press, Boca

Raton, FL.

36. Krimm, U., D. Abanda-Nkpwatt, W. Schwab, and L. Schreiber. 2005. Epiphytic

microorganisms on strawberry plants (Fragaria ananassa cv. Elsanta): identification of bacterial

isolates and analysis of their interaction with leaf surfaces. FEMS Microbiol Ecol 53:483-92.

37. Kulichevskaya, I. S., A. O. Ivanova, S. E. Belova, O. I. Baulina, P. L. Bodelier, W. I.

Rijpstra, J. S. Sinninghe Damste, G. A. Zavarzin, and S. N. Dedysh. 2007. Schlesneria

paludicola gen. nov., sp. nov., the first acidophilic member of the order Planctomycetales, from

Sphagnum-dominated boreal wetlands. Int J Syst Evol Microbiol 57:2680-7.

204

38. La Scola, B., R. J. Birtles, M. N. Mallet, and D. Raoult. 1998. Massilia timonae gen. nov., sp.

nov., isolated from blood of an immunocompromised patient with cerebellar lesions. J Clin

Microbiol 36:2847-52.

39. Lambais, M. R., D. E. Crowley, J. C. Cury, R. C. Bull, and R. R. Rodrigues. 2006. Bacterial

diversity in tree canopies of the Atlantic forest. Science 312:1917.

40. Liao, C. H. 2007. Inhibition of foodborne pathogens by native microflora recovered from fresh

peeled baby carrot and propagated in cultures. J Food Sci 72:M134-9.

41. Liu, Z., C. Lozupone, M. Hamady, F. D. Bushman, and R. Knight. 2007. Short

pyrosequencing reads suffice for accurate microbial community analysis. Nucleic Acids Res

35:e120.

42. Lopez-Velasco, G., M. Davis, R. R. Boyer, R. C. Williams, and M. A. Ponder. 2010.

Alterations of the phylloepiphytic bacterial community associated with interactions of

Escherichia coli O157:H7 during storage of packaged spinach at refrigeration temperatures. Food

Microbiol In Press.

43. Montanez, A., C. Abreu, P. R. Gill, G. Hardarson, and M. Sicardi. 2009. Biological nitrogen

fixation in maize (Zea mays L.) by 15N isotope-dilution and identification of associated

culturable diazotrophs. Biol Fertil Soils 45:253-263.

44. Morris, C. E., and L. L. Kinkel. 2002. Fifty years of phyllosphere microbiology: significant

contributions to research in related fields, p. 365-375. In S. E. Lindow and I. Hecht-Poinar (ed.),

Phyllosphere microbiology. APS Press, Saint Paul, MN.

45. Nawrocki, E. P., and S. R. Eddy. 2007. Query-dependent banding (QDB) for faster RNA

similarity searches. PLoS Comput Biol 3:56.

46. Nguyen-the, C., and F. Carlin. 1994. The microbiology of minimally processed fresh fruits and

vegetables. Crit Rev Food Sci Nutr 34:371-401.

47. Preston, G. M. 2004. Plant perceptions of plant growth-promoting Pseudomonas. Philos Trans R

Soc Lond B Biol Sci 359:907-18.

205

48. Randazzo, C. L., G. O. Scifo, F. Tomaselli, and C. Caggia. 2009. Polyphasic characterization

of bacterial community in fresh cut salads. Int J Food Microbiol 128:484-90.

49. Rasche, F., E. Marco-Noales, H. Velvis, L. S. van Overbeek, M. M. Lopez, J. D. van Elsas,

and A. Sessitsch. 2006. Structural characteristics and plant-beneficial effects of bacterial

colonizing the shoots of field grown conventional and genetically modified T4-lysozyme

producing potatoes. Plant Soil 289:123-140.

50. Rasche, F., R. Trondl, C. Naglreiter, T. G. Reichenauer, and A. Sessitsch. 2006. Chilling and

cultivar type affect the diversity of bacterial endophytes colonizing sweet pepper (Capsicum

anuum L.). Can J Microbiol 52:1036-45.

51. Rheims, H., C. Sproer, F. A. Rainey, and E. Stackebrandt. 1996. Molecular biological

evidence for the occurrence of uncultured members of the actinomycete line of descent in

different environments and geographical locations. Microbiology 142:2863-70.

52. Rico, D., A. B. Martin-Diana, J. M. Barat, and C. Barry-Ryan. 2007. Extending and

measuring the quality of fresh-fruit and vegetables: a review. Trends in Food Sci and Tech

18:337-386.

53. Roesch, L. F., R. R. Fulthorpe, A. Riva, G. Casella, A. K. Hadwin, A. D. Kent, S. H. Daroub,

F. A. Camargo, W. G. Farmerie, and E. W. Triplett. 2007. Pyrosequencing enumerates and

contrasts soil microbial diversity. Isme J 1:283-90.

54. Ronaghi, M., and E. Elahi. 2002. Pyrosequencing for microbial typing. J Chromatogr B Analyt

Technol Biomed Life Sci 782:67-72.

55. Rudi, K., S. L. Flateland, J. F. Hanssen, G. Bengtsson, and H. Nissen. 2002. Development and

evaluation of a 16S ribosomal DNA array-based approach for describing complex microbial

communities in ready-to-eat vegetable salads packed in a modified atmosphere. Appl Environ

Microbiol 68:1146-56.

56. Schuenzel, K. M., and M. A. Harrison. 2002. Microbial antagonists of foodborne pathogens on

fresh, minimally processed vegetables. J Food Prot 65:1909-15.

206

57. Sogin, M. L., H. G. Morrison, J. A. Huber, D. Mark Welch, S. M. Huse, P. R. Neal, J. M.

Arrieta, and G. J. Herndl. 2006. Microbial diversity in the deep sea and the underexplored "rare

biosphere". Proc Natl Acad Sci U S A 103:12115-20.

58. Thompson, I. P., M. J. Bailey, J. S. Fenlon, T. R. Fermor, A. K. Lillley, J. M. Lynch, P. J.

McCormack, M. P. McQuilken, and K. J. Purdy. 1993. Quantitative and qualitative seasonal

changes in the microbial community from the phyllosphere of sugar beet. Plant Soil 150:177-191.

59. Urich, T., A. Lanzen, J. Qi, D. H. Huson, C. Schleper, and S. C. Schuster. 2008.

Simultaneous assessment of soil microbial community structure and function through analysis of

the meta-transcriptome. PLoS One 3:2527.

60. Wang, Q., G. M. Garrity, J. M. Tiedje, and J. R. Cole. 2007. Naive Bayesian classifier for

rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol

73:5261-7.

61. Whipps, J. M., P. Hand, D. Pink, and G. D. Bending. 2008. Phyllosphere microbiology with

special reference to diversity and plant genotype. J Appl Microbiol 105:1744-55.

62. Wommack, K. E., J. Bhavsar, and J. Ravel. 2008. Metagenomics: read length matters. Appl

Environ Microbiol 74:1453-63.

63. Yang, C. H., D. E. Crowley, J. Borneman, and N. T. Keen. 2001. Microbial phyllosphere

populations are more complex than previously realized. Proc Natl Acad Sci U S A 98:3889-94.

64. Yang, J. H., H. X. Liu, G. M. Zhu, Y. L. Pan, L. P. Xu, and J. H. Guo. 2008. Diversity

analysis of antagonists from rice-associated bacteria and their application in biocontrol of rice

diseases. J Appl Microbiol 104:91-104.

65. Yutthammo, C., N. Thongthammachat, P. Pinphanichakarn, and E. Luepromchai. 2010.

Diversity and activity of PAH-Degrading bacteria in the phyllosphere of ornamental plants.

Microb Ecol. In Press.

207

Table 1. Number of reads, unique sequences and trimmed reads for spinach microbiomes.

Total reads Trimmed Unique reads

Sample

Fresh spinach 22,162 3,066 8,862

Packaged spinach

Stored 4°C/day1 17,757 4,064 5,911

Stored 4°C/day15 14,686 1,862 9,467

Stored 10°C/day15 24,294 3,915 17,404

208

Table 2. Taxonomic identification of sequences detected in the fresh spinach microbiome1.

Taxonomical Classification % Taxonomical Classification % Chlamydiae 0.05 Ferruginibacter spp. 0.05 Gemmatimonadetes -Proteobacteria Gemmatimonas 0.54 Rubellimicrobium spp. 0.09 Chloroflexi Rhodopila spp. 0.13 Sphaerobacterales 0.09 Roseomonas spp. 0.32 Novosphingobium spp. 0.27 Verrucomicrobia Sphingomonas spp. 6.43 Subdivision3 genera incertae sedis 0.05 Sphingosinicella spp. 0.45 Xiphinematobacter 0.05 Porphyrobacter spp. 0.09 Spartobacteria genera incertae sedis 1.21 Aurantimonas spp. 0.18 Unclassified Verrucomicrobia 0.05 Mesorhizobium spp. 0.05 TM7 Balneimonas spp. 0.05 TM7 genera incertae sedis 0.72 Bosea spp. 0.09 Deinococcus-Thermus Bradyrhizobium spp. 0.05 Truepera spp. 0.18 Labrys spp. 0.09 Deinococcus spp. 2.97 Methylobacterium spp. 1.21 Bacteria incertae sedis Beijerinckia spp. 0.14 Ktedonobacter spp. 0.05 Devosia spp. 0.18 Planctomycetes Hyphomicrobium spp. 0.09 Singulisphaera spp. 0.75 Rhizobium spp. 1.22 Actinobacteria Phenylobacterium spp. 0.13 Iluminobacter spp. 0.05 Brevundimonas spp. 2.79 Desulfosporosinus spp. 0.05 Unclassified Rhodospirillales 0.32 Kineosporiineae 0.05 Unclassified Sphingomonadaceae 2.38 Streptosporangineae 0.05 Unclassified Sphingomonadales 1.08 Micromonosporineae 0.14 Unclassified Bradyrhizobiaceae 0.13 Pseudonocardineae 0.09 Unclassified Methylobacteriaceae 0.22 Frankineae 0.36 Unclassified Beijerinckiaceae 0.05 Corynebacterineae 0.41 Unclassified Hyphomicrobiaceae 0.32 Propionibacterineae 2.61 Unclassified Rhizobiales 1.22 Micrococcineae 1.39 Unclassified -Proteobacteria 1.04 Unclassified Solirubrobacterales 0.14 Proteobacteria Unclassified Rubrobaceridae 0.05 Acinetobacter spp. 1.53 Unclassified Actinomycetales 0.14 Pseudomonas spp. 9.18 Unclassified Actinobacteria 0.18 Steroidobacter spp. 0.05 Firmicutes Pseudoxanthomonas spp. 0.05 Exiguobacterium spp. 0.05 Stenotrophomonas spp. 0.31 Bacillus spp. 0.27 Dokdonella spp. 0.05 Paenibacillus spp. 0.05 Luteimonas spp. 0.05 Unclassified Paenibacillaceae 0.05 Dyella spp. 0.27 Unclassified Bacillales 0.63 Lysobacter spp. 0.27 Serratia spp 0.13 Pantoea spp. 0.13 Bacteroidetes Unclassified Pseudomonaceae 0.40 Mucilaginibacter spp. 0.05 Unclassified Xanthomonadaceae 1.40 Pedobacter spp. 0.05 Unclassified Enterobacteriaceae 13.7 Dyadobacter spp. 0.05 Unclassified -Proteobacteria 1.03 Hymenobacter spp. 0.45 Segetibacter spp. 0.05

209

Table 2 continuation. Description of fresh spinach microbiome1

Taxonomical Classification % Taxonomical Classification % Acidobacteria -Proteobacteria Gp13 0.05 Nitrospira spp. 0.05 Gp1 0.31 Aquaspirillum spp. 0.05 Gp7 0.09 Aquabacterium spp. 0.05 Gp17 0.05 Methylibium spp. 0.05 Gp16 0.45 Burkholderia spp. 0.23 Gp4 2.02 Ralstonia spp. 1.22 Gp6 2.65 Polaromonas spp. 0.36 Gp3 0.36 Acidovorax spp. 0.09 -Proteobacteria Variovorax spp. 0.14 Bacteriovorax spp. 0.05 Undibacterium spp. 0.04 Geobacter spp. 0.05 Duganella spp. 1.03 Nannocystaceae 0.05 Janthinobacterium spp. 0.04 Polyangiaceae 0.36 Naxibacter spp. 3.73 Cystobacteraceae 0.09 Massilia spp. 7.38 Unclassified Desulfomonadales 0.05 Unclassified Alcaligenaceae 0.09 Unclassified Rhodobacteraceae 0.14 Unclassified Burkholderiaceae 0.09 Unclassified -Proteobacteria 0.14 Unclassified Comamonadaceae 2.30 Unclassified Oxalobacteraceae 8.55 Unclassified Burkholderiales 1.23 Unclassified -Proteobacteria 0.94 Unclassified Proteobacteria 2.21 1 The percentage of sequence reads was calculated based on the total number of classified reads according with the RDP Classifier tool.

210

Table 3. Richness and diversity estimators that predict the number of species in fresh and packaged spinach at 3% of dissimilarity.

Chao Shannon index Species (H’) evenness (E)1 Sample Lower limit Chao1 Upper limit Fresh spinach 1945 1,741 2,204 4.21 + 0.59 0.56

Packaged spinach Stored 4°C/day1 981.0 1,140 1,358 3.77 + 0.59 0.53

Stored 4°C/day15 840.0 942.0 1,083 3.90 + 0.61 0.56

Stored 947.0 1,068 1,233 3.51 + 0.54 0.50 10°C/day15 1 Evenness was calculated as E=H’/H max where H max=ln S being S the total number of species in the sample, estimated with Chao1.

211

Table 4. Comparison of spinach microbiomes during storage at 4 and 10°C for up to 15 days1.

Phyla Family and/or genus relative abundance (%)3 Fresh 4°C/day1 4°C/day15 10°C/day10 Actinobacteria Corynebacterineae3 0.41 0.80 3.20 0.10 Propionibacterineae3 2.61 1.61 1.02 0.06 Micrococcineae3 1.40 1.41 0.67 1.69 Acidobacteria Gp4 2.03 0.80 0.05 0.04 Gp6 2.66 1.20 0.16 0.05 Deinococcus- Thermus Deinococcus 2.97 3.92 1.85 0.06 Firmicutes Exiguobacterium spp. 0.05 0.20 0.00 0.21 Proteobaceria Alpha Sphingomonadaceae3 2.39 1.31 0.00 0.06 Sphingomonadales3 1.08 0.30 0.00 0.00 Rhizobiales3 1.22 2.11 0.20 0.04 Unclassified -Proteobacteria 1.04 0.60 0.13 0.02 Sphingomonas spp. 6.44 10.04 0.00 0.87 Methylobacterium spp. 1.22 0.50 3.49 0.18 Rhizobium spp. 1.22 1.00 0.13 0.10 Brevundimonas 2.79 1.00 0.67 0.11 Gamma Pseudomonadaceae3 0.41 0.10 0.31 0.41 Xanthomonadaceae3 1.40 1.20 0.40 0.37 Enterobacteriaceae3 13.69 13.55 52.10 37.10 Unclassified -Proteobacteria3 1.04 9.54 2.14 1.59 Acinetobacter spp. 1.53 0.40 0.11 0.04 Pseudomonas spp. 9.19 14.76 27.66 26.25 Stenotrophomonas spp. 0.32 0.20 0.24 1.28 Pantoea spp. 0.14 0.10 0.44 1.14 Escherichia spp. 0.00 0.00 0.05 23.07 Beta Comamonadaceae3 2.30 1.71 0.29 0.06 Oxalobacteraceae3 8.55 3.61 0.31 0.64 Burkholderiales3 1.26 0.30 0.09 0.04 Unclassified -Proteobacteria 0.95 0.60 0.04 0.02 Ralstonia spp. 1.22 0.80 0.04 0.01 Naxibacter spp. 3.74 1.20 0.42 0.04 Massilia spp. 7.38 2.31 0.22 3.20

1(Representative bacteria genera that showed major shifts in bacteria community composition during storage)

2 The percentage of relative abundance was calculated based on the total number of classified reads according with the RDP Classifier tool.

3Sequence reads classified only at family level.

212

Figure 1A-B. Relative phylum/class abundance of classified sequences on fresh and packaged spinach microbiomes.

213

A. Fresh spinach

1200 3% dissimilarity 5% dissimilarity 10% dissimilarity 1000 20% dissimilarity

800

600

400

200

Operational taxonomic units (OTU's)

0 2000 4000 6000 8000 10000 Number of sequences sampled

B. Packaged spinach stored at 4oC (day 1)

600 3% dissimilarity 5% dissimilarity 500 10% dissimilarity 20% dissmilarity

400

300

200

100

Operational units (OTU's) taxonomic 0

0 1000 2000 3000 4000 5000 6000 7000 Number of sequences sampled

Figure 2A-B. Rarefraction curves for fresh and packaged spinach indicating the observed number of

OTU’s within the 16S rRNA microbiomes derived from spinach leaves. Curves were calculated with

RDP using pyrosequencing tools.

214

C. Packaged spinach stored at 4oC (day 15)

3% dissimilarity 600 5% dissimilarity 10% dissimilarity 500 20% dissimilarity

400

300

200

100

Operational taxonomic units (OTU's) units taxonomic Operational

0 2000 4000 6000 8000 10000 12000

Number of sequences sampled

D. Packaged spinach stored at 10oC (day 15)

700 3% dissimilarity 5% dissimilarity 600 10% dissimilarity 20% dissmilarity 500

400

300

200

100

Operational taxonomic units (OTU's) Operational taxonomic 0

0 2000 4000 6000 8000100001200014000160001800020000

Number of sequences sampled

Figure 2C-D. Rarefraction curves for packaged spinach indicating the observed number of OTU’s within the 16S rRNA microbiomes derived from spinach leaves. Curves were calculated with RDP using pyrosequencing tools.

215

Figure 3. Comparison of the four spinach microbiomes. UPGMA utilizing Jaccard’s index at 3% of dissimilarity.

216

CHAPTER 7

CONCLUSIONS AND FUTURE DIRECTIONS

The purpose of this work was to study the bacterial diversity the spinach phyllosphere and the changes that occur during pre- and post-harvest operations as well as its interaction with the enteric pathogen E. coli O157:H7. Spinach bacterial community on the surface (phylloepiphytes) showed high bacterial diversity.

The abundance of phylloepiphytes was affected by spinach cultivar, which was mostly associated with differences that existed among cultivars related with surface area, and sites of microbial colonization such as trichomes and stomata, additionally effects in abundance and bacterial composition were associated with changes in environmental conditions during spinach production. Changes in population size and bacterial composition during spinach production altered the microbial diversity, how these changes might affect the establishment of enteric pathogens on edible plants needs to be determined.

After harvest, bacterial communities of spinach were characterized using non-culture dependent techniques, particularly pyrosequencing. The application of this technique allowed a broader description of the bacterial community. Pyrosequencing evidenced the presence of bacterial members that are ubiquitously distributed in the phyllosphere as well as in other environments. It would be important in terms of microbial ecology, to understand the role of members of the phyllosphere and to study particular populations of novel bacteria. Several members of the spinach bacterial community detected through pyrosequencing were also isolated from fresh spinach leaves. We demonstrated that epiphytic bacteria can elicit interactions with E. coli O157:H7 in vitro by either inhibiting its growth of through commensal and neutral interactions. It is paramount to support this data with in vivo studies to demonstrate that these interactions occur on leaf surfaces. These studies could contribute to understand how epiphytic bacteria

217 can contribute to the fitness of enteric pathogens on the leaves and also they can contribute to the design of strategies for bio-control of enteric pathogens.

Post-processing operations like packaging and storage of spinach leaves also affected the composition and population dynamics of the spinach bacterial communities. The presence of native associated bacteria significantly contributed to the physiology and pathogenicity of E. coli O157:H7 during storage time on packaged spinach leaves. The presence of this enteric pathogen on packaged spinach was able to alter the composition of the microbial community indicating active interaction of this enteric pathogen, which indicates the importance of the spinach bacterial community in the ecology of pathogenic bacteria. Through pyrosequencing we were able to distinguish which bacterial members had major shifts during storage, it would be important to show how specific changes in certain bacterial populations affect the survival of E. coli O157:H7. It would be important to expand our understanding of phyllosphere microbial ecology by examining the functionality of different members during packaging and storage at refrigeration temperatures to correlate their microbial function and its effect in the establishment of enteric pathogens.

218

APPENDIX A

COMPARISON OF DNA ISOLATION METHODOLOGIES FOR

MICROBES ASSOCIATED WITH THE SURFACE OF SPINACH LEAVES

DESCRIPTION

The objective of this experiment was to identify a DNA extraction protocol that would yield large amounts of microbial DNA, while minimizing the amount of plant DNA extracted from the leaf surfaces. Two methods were compared based on the final yield of DNA produced and the richness of the community as visualized by DGGE of the 16s rDNA.

Due to the tight associations of microbes with their plant hosts it is difficult to extract microbial nucleic acids that are not contaminated with plant nucleic acids. The presence of these plant nucleic acids can result in an incomplete or biased view of the microbial community. The main reason is that chloroplast nucleic acids are similar to the bacterial 16S rRNA, and are also amplified by these universal primers. This can result in preferential amplification of plant and select bacterial sequences at the expense of other bacterial members of the community, resulting in decreased species richness and abundance. Therefore, it is important to prevent or minimize disruption of plant cells, while detaching microbial cells.

In this experiment, we utilized two different methods to obtain bacteria nucleic acids from spinach leaves.

219

Method 1: Extraction from frozen, whole leaves

In the first method 5 spinach leaves were immersed in liquid nitrogen and the resulting frozen leaves ground to a powder using a mortar and pestle. The frozen powder (20mg) was immediately re-suspended in a lysis buffer from the ZR soil microbe DNA kitTM (Zymo

Research Co., Orange, CA) for DNA isolation. DNA extraction was performed using the Zymo soil microbe kit per manufacturer’s instructions. For RNA isolation 20mg of the frozen powder were resuspended in 600 mLof RLT buffer containing -mercaptoethanol (kit RNeasy Mini kit

Qiagen, CA) and 0.1 mL of 0.20mm silica beads. In both cases, samples were processed in a bead beater and kit’s manufacturer instructions were followed.

Method 2: Extraction from a solution obtained from pulsified spinach leaves

The second method involved isolating nucleic acids from bacteria that were disassociated from the leaf surface prior to extraction. Ten grams of spinach leaves were suspended in 1% peptone (Sigma-Aldrich Co., USA) water supplemented with 1% (vol/vol) of Tween-90 (PTW) within a sterile filter bag (Fisher). Samples were processed in a pulsifier (Microbiology

International, Frederick, MD) for 5 min to detach bacterial cells from the leaves. Previous studies have shown that the pulsifier minimizes tissue disruption in comparison with a commercial stomacher often used in microbiology to detach bacterial cells from food matrices

(1). Cell suspensions were processed for nucleic acid isolation as previously described in chapters 3 and 5.

RESULTS

Significant differences were determined for the total yield of nucleic acids obtained using the two methods. High yields (1 to 4 g) were extracted from the frozen, ground plant. In

220 contrast, a more modest amount of DNA (250 ng) was extracted from the pulsified suspensions.

To assess the relative abundance of bacterial and plant DNA, community analyses were performed using DGGE as described in chapters 3 and 5. DGGE profiles were compared from

DNA extracted from leaves of the same batch of spinach extracted using the two methodologies.

Despite the larger yield obtained by using frozen, ground leaves, a decrease in the number of bands on the DGGE gel indicates this method provided an incomplete picture of the bacterial community diversity (Fig 1). This is likely due to the presence of more plant DNA in the samples which outcompeted bacterial sequences for amplification during PCR.

221

Figure A1. DGGE of PCR-amplified 16S rDNA fragments utilizing DNA and RNA converted to cDNA obtained from two different extraction protocols.

References:

1. Kang, D., R. Dougherty, and D. Fung. 2001. Comparison of pulsifier and stomacher to

detach microorganisms fro lean meat tissues. J Rapid Meth and automation in Microbiol.

222

APPENDIX B

DIFFERENCES IN POPULATION SIZE AND MICROBIAL COMMUNITY

COMPOSITION AMONG SPINACH CULTIVARS

DESCRIPTION

The objective of this experiment was to determine if the cultivar of species was associated with differences in microbial population size and differences in composition of active and total members of the bacterial community. The leaves of different spinach cultivars have differences in topography, particularly the surface area and amount of ruffling of the leaves. These differences in total leaf surface area, expected to be larger in savoy than in flat cultivars, may affect population sizes of microbial communities. Spinach production and harvest, nucleic acid isolation and denaturant gradient gel electrophoresis were previously described in chapter 5. Ennumeration of bacterial populations was described in chapter 3.

RESULTS

Spinach leaves were colonized by large numbers of culturable bacteria and molds. (Figure 1). As seen in chapter 3, significant differences in total culturable mesophilic bacteria were observed between cultivars. The flat cultivar, Monza, had lower microbial counts than Menorca and Unipak (Figure 1). In contrast, no signifcant differences were observed in numbers of psychrotrophic bacteria or molds and yeast cultured from the three cultivars of spinach (Figure 1).

This experiment also adresses differences in the communites of active bacteria and those which are in metabolic stasis on the leaf surface. Bacteria which are active will be profiled in the cDNA samples, while inactive bacterial species will appear only in the total DNA profiles. The structure of the microbial community was assessed by DGGE (Fig 2A-B). A unique cluster was formed for profiles constructed with DNA and other with those constructed with cDNA (Fig 2A). Band patterns constructed

223 with total bacterial DNA had fewer bands compared to those constructed from cDNA. This was unexpected, as it was anticipated that an equal number or more bands would be visible in profiles generated from the total bacterial DNA compared to the cDNA. This increased number of bands are likely the result of expression of multiple different copies of the 16s rRNA gene on the leaf surface (2). The presence of microhetegeneity in 16S rRNA in organisms is often related to temporal changes and to ecological adaptations (1). These copies have non random modifications that are maintained to provide a fine-tuning of the ribosome function to optimize translation machinery performance and ultimately bacterial niche fitness (3). This can also be attributed to lower co-migration of bands and differences in amplification associated with the template utilized; cDNA or DNA. Comparison of DGGE profiles among different cultivars showed almost 90% similarity in the band patterns of DNA-DGGE prints as well as similar number of bands, suggesting that differences in the structure of the bacterial community on spinach leaves might not be related with species richness but rather with population size as showed with the enumeration of bacterial population (Fig 1).

224

Figure B1. Microbial populations of spinach microbiota. Comparison among three spinach cultivars.

Psichrotrophic bacteria was determined after incubation of plates at 4°C for 16 days, molds and yeasts as well as bacterial populations determined in R2A were determined after incubation of plates at 25°C for 16 days. Proc mix from SAS software (Statistical Analysis System 9.2 SAS Institute, Cary, NC) was utilized to determine lsmeans, standard error and comparisons among means for each treatment, means were separated using Tukey’s multiple comparison. Significant difference was determined when p-values <0.05 and are indicated with different letters. All data was previously evaluated for normality and homogeneity of variance using the proc univariate function of SAS.

225

Figure B2. Comparison of structure of the microbial community among spinach cultivars assessed by

DGGE. A. UPGMA clustering of 16S rDNA DGGE profiles. B. DGGE of PCR-amplified 16S rDNA fragments. Lanes: 1 Monza, 2 Menorca, 3 Unipak (using DNA as template for PCR reaction) 4 Monza, 5

Menorca, 6 Unipak (using cDNA as template for PCR reaction).

226

REFERENCES

1. Casamayor, E. O., C. Pedros-Alio, G. Muyzer, and R. Amann. 2002. Microheterogeneity in

16S ribosomal DNA-defined bacterial populations from a stratified planktonic environment is

related to temporal changes and to ecological adaptations. Appl Environ Microbiol 68:1706-14.

2. Dahllof, I., H. Baillie, and S. Kjelleberg. 2000. rpoB-based microbial community analysis

avoids limitations inherent in 16S rRNA gene intraspecies heterogeneity. Appl Environ Microbiol

66:3376-80.

3. Jensen, S., P. Frost, and V. L. Torsvik. 2009. The nonrandom microheterogeneity of 16S rRNA

genes in Vibrio splendidus may reflect adaptation to versatile lifestyles. FEMS Microbiol Lett

294:207-15.

227

APPENDIX C

PYROSEQUENCING OF 16S rRNA MICROBIOMES OF SPINACH AT

VARIOUS STAGES OF PLANT GROWHT (FROM SEED TO YOUNG

LEAVES)

DESCRIPTION

Pyrosequencing of the 16S rRNA gene was utilized to study the changes in bacterial community composition present on spinach during its stages of plant growth. Methods were exactly followed as described in chapter 6.

RESULTS

Table C1. Characteristics of the sequenced data.

Total reads Trimmed Unique % of classified reads1 sequences2

Seed 17,341 2,541 6,017 34.93

Cotyledons 35,503 10,303 6,677 15.08

Young leaves 34,807 5,849 13,750 69.26

(3-4days)

1 Reads that were trimmed due to low quality (size and sequence)

2Sequence reads that were classified according to the RDP database (Cutoff 60%)

228

Table C2. Richness and diversity estimators for seed, cotyledons and young spinach leaves microbiomes.

Chao Shannon index Species (H’) evenness (E)1 Sample Lower limit Chao1 Upper limit Seed 229 272 350 2.28 + 0.44 0.41 Cotyledon 226 271 353 2.39 + 0.46 0.42 Young leaves 745 863 1,032 3.15 + 0.51 0.46 (3-4 days) 1 Evenness was calculated as E=H’/H max where H max=ln S being S the total number of species in the sample, estimated with Chao1.

229

Table C3. Microbiomes of seeds, cotyledon and young leaves from spinach, cultivar Menorca grown in growth chambers. (Results are represented as relative abundance of the total classified bacteria).

Relative abundance % Bacterial member Spinach seed Cotyledon Young leaf (3-4 days) Deinococcus-Thermus Thermus spp. - - 0.05 Planctomycetes Planctomyces spp. - 0.26 0.08 Gemmata spp. - - 0.03 Singulisphaera spp. - - 0.06 Unclassified Planctomycetaceae - - 0.09 Bacteria_incertae_sedis Ktedonobacter - 0.26 0.09 Acidobacteria - - 0.00 Gp16 - - 0.02 Gp3 - - 0.03 Gp6 - - 0.06 TM7 TM7 genera incertae sedis - - 0.09 Chlamydiae Unclassified Parachlamydiaceae - - 0.03 OD1 OD1 genera incertae sedis - - 0.02 Verrumicrobia Luteolibacter spp. - - 0.02 Unclassified Opitutaceae - - 0.03 Unclassified Verrucomicrobiaceae - - 0.02 Firmicutes Enterococcus spp. 1.44 - 0.03 Exiguobacterium. 0.08 - - Bacillus spp. 0.72 - 0.02 Staphylococcus spp. 14.57 - 0.03 Alicyclobacillus spp. - 0.52 0.35 Paenibacillus spp. - 0.52 - Streptococcus spp. - - 0.21 Lysinbacillus spp. - - 0.02 Brevibacillus spp. - - 0.02 Cohnella spp. - - 0.03 Sulphobacillus spp. - - 0.06 Subdoligranulum spp. - - 0.02

230

Relative abundance % Bacterial member Spinach seed Cotyledon Young leaf (3-4 days) Unclassified Bacillaceae 1.52 - 0.03 Unclassified Paenibacillaceae - - 0.02 Unclassified Bacilli 1.04 0.26 - Unclassified Staphylococcaceae 2.24 - 0.02 Unclassified Clostridiaceae - 0.26 - Unclassified Bacillales 0.88 0.26 0.03 Unclassified Lactobacillales - - 0.02 Unclassified Clostridiales - - 0.02 Unclassified Clostridia - - 0.02 Unclassified Firmicutes - - 0.02 Chloroflexi Sphaerobacter spp. - - 0.06 Unclassified Thermomicrobia - - 0.03 Unclassified Chloroflexi - - 0.02 Actinobacteria Corynebacterium spp. 0.08 - - Sanguibacter spp. 0.08 - - Streptomyces spp. - 0.52 - Conexibacter spp. - - 0.05 Thermomonosporaceae - - 0.03 Acidothermaceae - - 0.02 Streptomycetaceae - - 0.03 Nocardiodaceae - - 0.23 Nocardiaceae - - 0.02 Mycobacteriaceae - - 0.02 Microbacteriaceae - - 0.05 Micrococcaceae - - 0.15 Cellulomonadaceae - - 0.02 Unclassified Micrococcineae 0.16 - - Unclassified Thermomonosporaceae - 0.26 - Unclassified Cellulomonadaceae - 0.26 - Unclassified Solirubrobacterales - - 0.05 Unclassified Actinomycetales - - 0.02 Unclassified Actinobacteria - - 0.08 Proteobacteria -Proteobacteria Byssovorax spp. - - 0.02 Unclassified Sorangiineae - - 0.03 -Proteobacteria

231

Relative abundance % Bacterial member Spinach seed Cotyledon Young leaf (3-4 days) Nitrosopira spp. - 0.26 0.00 Castellaniella spp. - 0.26 0.00 Cupriavidus spp. - 0.77 0.12 Massilia spp. - 0.26 0.24 Aquabacterium spp. - - 0.02 Burkholderia spp.. - - 0.02 Achromobacter spp. - - 0.24 Unclassified Alcaligenaceae 0.16 - 0.21 Unclassified Oxalobacteraceae - 6.44 0.23 Unclassified Burkholderiales - 0.52 - Unclassified Comamonadaceae - - 0.03 Unclassified -Proteobacteria - - 0.06 Proteobacteria Rhizobium spp. 3.68 - 1.93 Phenylobacterium spp. - 0.52 0.02 Hyphomicrobium spp. - 0.26 0.11 Azospirillum spp. - 0.26 - Caulobacter spp. - - 0.08 Brevundimonas spp. - - 0.02 Stella spp. - - 0.02 Azospirillum spp. - - 0.11 Sphingomonas spp. - - 0.02 Sphingosinicella spp. - - 0.05 Pseudochrobactrum spp. - - 0.11 Bosea spp. - - 0.20 Methylobacterium spp. - - 0.02 Blastochloris spp. - - 0.02 Devosia spp. - - 0.08 Kaistia spp. - - 0.02 Unclassified Hyphomicrobiaceae - 0.26 0.08 Unclassified Acetobacteraceae - - 0.02 Unclassified Rhodospirillales - - 0.05 Unclassified Erythrobacteraceae - - 0.03 Unclassified Sphingomonadaceae - - 0.02 Unclassified Methylocystaceae - - 0.02

232

Relative abundance %

Bacterial member Spinach seed Cotyledon Young leaf (3-4 days) Unclassified Sphingomonadales - 0.02 Unclassified -Proteobacteria - - 0.20 -Proteobacteria Pseudomonas spp. 12.73 0.26 14.45 Pantoea spp. 0.16 0.77 0.38 Dyella spp. - 0.26 - Rhodanobacter spp. - 0.26 0.03 Hydrocarboniphaga spp. - 0.26 0.00 Solimonas spp. - - 0.02 Lysobacter spp. - - 0.02 Dyella spp. - - 0.05 Stenotrophomonas spp. - - 1.99 Erwinia spp. - - 0.02 Battiauxella spp. - - 0.02 Psychrobacter spp. - - 0.02 Unclassified Pseudomonadaceae 0.40 0.26 1.10 Unclassified Enterobacteriaceae 53.48 81.70 73.81 Unclassified Xanthomonadaceae - 0.52 0.08 Unclassified Sinobacteraceae - 0.26 - Unclassified Pseudomonadales - - 0.03 Unclassified -Proteobacteria 1.12 0.52 0.56 Unclassified Proteobacteria 4.64 1.03 0.20

233

Figure C1. Relative phylum/class abundance of classified sequences from seed, cotyledon and young spinach leaves.

234

Figure C2. Relative phylum/class abundance of classified sequences from seed, cotyledon and young spinach leaves.

235

A. Spinach Seed

200 reads vs 3% dissimilarity 180 reads vs 5% dissimilarity reads vs 10% dissimilarity 160 reads vs 20% dissimilarity 140 120 100 80 60 40 20

Operational taxonomic units (OTU's) Operational taxonomic 0

0 1000 2000 3000 4000 5000 6000 7000

Number of sequences sampled

Figure C3. Rarefraction curves for spinach seeds indicating the observed number of OTU’s within the 16S rRNA microbiomes derived from spinach microbiomes.

236

B. Spinach cotyledons

180 reads vs 3% dissimilarity reads vs 5% dissimilarity 160 reads vs 10% dissimilarity reads vs 20% dissimilarity 140 120 100 80 60 40 20

Operational Taxonomic Units (OTU's) 0

0 2000 4000 6000 8000 Number of sequences sampled

Figure C4. Rarefraction curves for spinach cotyledons, indicating the observed number of OTU’s within the 16S rRNA microbiomes derived from spinach microbiomes.

237

C. Young spinach leaves (3-4 days) reads vs 3% dissimilarity 500 reads vs 5% dissimilarity reads vs 10% dissimilarity reads vs 20% dissimilarity 400

300

200

100

Operational Taxonomic Units (OTU's)

0 2000 4000 6000 8000 10000 12000 14000 16000 Number of sequences sampled

Figure C5. Rarefraction curves for young spinach leaves, indicating the observed number of OTU’s within the 16S rRNA microbiomes derived from spinach microbiomes.

238