comparison of salmonella enterica serovar bovismorbificans 2011 hummus outbreak strains with...

Accepted Manuscript

Comparison of Salmonella enterica serovar Bovismorbificans 2011 Hummus outbreakstrains with non-outbreak strains

Morris Blaylock, Reginald Blackwell, Sosina Merid, Scott Jackson, Michael Kotewicz,Gopal Gopinath, Sherry L. Ayers, Jason Abbott, Jonathan Sabo, Laura Ewing,Jayanthi Gangiredla, Solomon Gebru, Isha Patel, Brooke Jones, Kim Dudley, KarenJarvis, Darcy E. Hanes, Alpha A. Diallo, Junia Jean-Gilles Beaubrun, PhD, FDA/CFSAN/OARSA/DVA

PII: S0740-0020(14)00041-0

DOI: 10.1016/j.fm.2014.02.016

Reference: YFMIC 2114

To appear in: Food Microbiology

Received Date: 19 June 2013

Revised Date: 4 February 2014

Accepted Date: 18 February 2014

Please cite this article as: Blaylock, M., Blackwell, R., Merid, S., Jackson, S., Kotewicz, M., Gopinath,G., Ayers, S.L., Abbott, J., Sabo, J., Ewing, L., Gangiredla, J., Gebru, S., Patel, I., Jones, B., Dudley,K., Jarvis, K., Hanes, D.E., Diallo, A.A., Jean-Gilles Beaubrun, J., Comparison of Salmonella entericaserovar Bovismorbificans 2011 Hummus outbreak strains with non-outbreak strains, Food Microbiology(2014), doi: 10.1016/j.fm.2014.02.016.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service toour customers we are providing this early version of the manuscript. The manuscript will undergocopyediting, typesetting, and review of the resulting proof before it is published in its final form. Pleasenote that during the production process errors may be discovered which could affect the content, and alllegal disclaimers that apply to the journal pertain.

http://dx.doi.org/10.1016/j.fm.2014.02.016

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

1

Title: Comparison of Salmonella enterica serovar Bovismorbificans 2011 Hummus 1

outbreak strains with non-outbreak strains. 2

3

Keywords: Salmonella, Bovismorbificans, hummus 4

5

Running Title: Molecular characterization of Salmonella enterica serovar Bovismorbificans 6

isolates. 7

8

Authors: Morris Blaylock1, Reginald Blackwell1, Sosina Merid1, Scott Jackson2, Michael 9

Kotewicz2, Gopal Gopinath2, Sherry L. Ayers3, Jason Abbott3, Jonathan Sabo3, Laura Ewing2, 10

Jayanthi Gangiredla2, Solomon Gebru2, Isha Patel2, Brooke Jones2, Kim Dudley2, Karen Jarvis2, 11

Darcy E. Hanes2, Alpha A. Diallo1, *Junia Jean-Gilles Beaubrun2. 12

13

*Corresponding author: Junia Jean-Gilles Beaubrun, PhD, FDA/CFSAN/OARSA/DVA, 8301 14

Muirkirk Rd, Rm 3404, Laurel MD 20708, 240-402-3578 (Lab), 301-785-4698 (cell), 301-210-15

5370 (Fax), [email protected]. 16

17

District of Columbia Public Health Laboratory, Department of Forensic Sciences, 401 E. Street 18

SW Washington, DC 200241, U.S. Food and Drug Administration, Laurel, MD 207082, U.S. 19

Food and Drug Administration, Center for Veterinary Medicine, Laurel MD 207083. 20

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

2

ABSTRACT 21

Eleven Salmonella enterica serovar Bovismorbificans isolates obtained from the U.S. District of 22

Columbia during a 2011 hummus-associated foodborne outbreak were compared to 12 non-23

outbreak isolates. All isolates from the outbreak demonstrated a single PFGE pattern that was 24

distinctly different from other isolates of S. Bovismorbificans as recorded in the PulseNet 25

Database. Results from molecular analyses of the hummus-associated S. Bovismorbificans 26

isolates indicate that the isolates from the outbreak were unique and have acquired an 80-90 kb 27

plasmid. The impact of this study is that the information gained will add and expand our 28

knowledge of diversity of the S. Bovismorbificans serovar. 29

INTRODUCTION 30

Outbreaks associated with a variety of Salmonella enterica serovars and food matrices 31

continue to increase each year, making identification, differentiation and characterization of the 32

various serotypes in foodborne outbreak a high priority for the field of food safety. Every year 33

approximately 40,000 cases of salmonellosis are reported in the United States. There are over 34

2500 different serovars of Salmonella enterica, and some of the most recent reported cases of 35

Salmonella enterica associated with foodborne outbreaks include: S. Typhimurium in ground 36

beef, S. Heidelberg in Kosher broiled chicken livers and ground turkey, S. Enteritidis in Turkish 37

pine nuts, eggs, alfalfa sprouts and spicy sprouts, S. Agona in fresh imported papayas, S. Hadar 38

in turkey burgers, S. Panama in Cantaloupe, S. Bareilly in spicy tuna, S. Braenderup in mangoes, 39

and S. Bredeney in Peanut Butter (20). 40

Human infections with S. enterica serovar Bovismorbificans are relatively infrequent in 41

the United States. Prior to 2011, the last major outbreak associated with S. Bovismorbificans 42

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

3

occurred in 2004 with 35 confirmed cases associated with the consumption of alfalfa sprouts 43

(20). From 1999 to 2009, only 758 illnesses associated with S. Bovismorbificans were reported, 44

compared to 1000 confirmed cases associated with S. Enteritidis in 1999 alone (Foodborne 45

Outbreak Online Database, CDC, www.cdc.gov/salmonella/outbreaks.html). The majority of 46

foodborne outbreaks associated with S. Bovismorbificans has occurred in Europe and have been 47

traced to pork products, lettuce and sprouts (18). In Finland in 1994, 201 cases were reported 48

from a large sprout-associated S. Bovismorbificans outbreak (16). In Sweden in 1994, there was 49

a large sprout-associated, S. Bovismorbificans outbreak (15) and a second nationwide outbreak 50

occurred in 2009 with 42 clinical isolates identified and was also associated with ready-to-eat 51

alfalfa sprouts. The sprouts samples were traced back to a domestic producer, but the seeds 52

originated in Italy (18). Over a 13 week period between November 2004 and March 2005, 525 53

cases of laboratory confirmed S. Bovismorbificans associated with raw pork were reported to 54

The Robert Koch Institute (18). 55

From August through November of 2011, sesame seed paste (tahini) and humus 56

containing a rare serotype of Salmonella caused illness in 23 people in 7 states, including the 57

District of Columbia (23). Reported cases were largely concentrated in the Mid-Atlantic region, 58

with eight in Washington, D.C., seven in Maryland, three in Virginia and one each in Delaware 59

and New Jersey. Three cases were also reported outside this region – one in California, one in 60

Michigan and one in New Hampshire (23). In this study, eleven S. Bovismorbificans isolates 61

from clinical and hummus food samples were obtained from the DC Public Health Laboratory 62

(DC_PHL) (Table 1). PFGE and conventional serotyping were used to identify the causal agent 63

as a unique strain of S. Bovismorbificans. Hummus has not previously been associated as a 64

source of Salmonella contamination. The DC_PHL and the FDA Office of Applied Research and 65

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

4

Safety Assessment (OARSA) collaborated to characterize the outbreak strains. In addition to the 66

original PFGE and the Kauffman White serotyping analyses, molecular serotyping using PCR, 67

Optical Mapping, Multiple-locus variable-number tandem-repeats analysis (MLVA), antibiotic 68

susceptibility testing, molecular fingerprinting using the Diversilab system, and plasmid analyses 69

were used to characterize these outbreak strains. 70

71

MATERIALS AND METHODS 72

Strains. A total of 23 isolates were examined in this study (Table 1). DC_PHL identified 11 S. 73

Bovismorbificans isolates from clinical and food samples (hummus) associated with the 2011 74

Washington DC outbreak. Two additional isolates associated with the outbreak were obtained 75

from the Michigan and Delaware Departments of Health, Division of Public Health laboratories. 76

Ten additional S. Bovismorbificans isolates, not associated with the outbreak, were also 77

examined; and these isolates were obtained from the FDA Center for Food Safety and Applied 78

Nutrition (CFSAN) and the Center for Veterinary Medicine (CVM) Salmonella culture 79

collections. Two outlier strains of Salmonella: S. Typhimurium and S. Newport were also 80

included in the studies (Table 1). 81

Molecular Serotyping (PCR Analysis). To evaluate the molecular serotyping PCR method’s 82

usefulness during an outbreak, all isolates used in this study were serotyped using the 83

conventional serotyping method and a PCR serotyping method described by Jean-Gilles 84

Beaubrun et al., (12,14). Genomic DNA was isolated using the NucliSENS EasyMag instrument 85

(BioMerieux, Inc. Hazelwood, MO) according to the manufacturer’s instructions. PCR products 86

were visualized using the Agilent 2100 Bio-analyzer (Agilent Technologies, Waldbronn, 87

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

5

Germany), and the DNA 1000 Reagents kit (Agilent Technologies) following the manufacturer’s 88

protocol. 89

Pulsed Field Gel Electrophoresis. All isolates of S. Bovismorbificans were analyzed for genetic 90

relatedness using pulsed-field gel electrophoresis (PFGE) with XbaI and BlnI according to the 91

U.S. CDC PulseNet protocol (21). Electrophoresis was performed with a CHEF-DR system 92

(Bio-Rad Laboratories, Hercules, California) using 1% SeaKem agarose in 0.5× Tris-borate-93

EDTA at 180 V. Running conditions consisted of one phase from 2.2 s to 63.8 s at a run time of 94

18 h. PFGE profiles were analyzed using Bionumerics software v3.5 (Applied Maths, Sint-95

Martens-Latem, Belgium) (3). 96

Antimicrobial Susceptibility Testing. All of the S. Bovismorbificans isolates were tested for 97

susceptibility to standardized panels of antimicrobial drugs by the broth microdilution method 98

(Sensititre® panel type: CMV1AGNF, Trek Diagnostic Systems, Westlake, OH) and the analysis 99

was conducted according to the rigorously standardized Clinical and Laboratory Standards 100

Institute (CLSI) protocols (6, 7, 8) ). CLSI approved interpretive criteria were used when 101

available; otherwise provisional National Antimicrobial Resistance Monitoring System 102

(NARMS) breakpoints were used (6, 7, 8). 103

Optical mapping. Genome mapping was conducted using Optical mapping. High molecular 104

weight DNA was prepared from bacterial colonies following the manufacturer’s procedures 105

(OpGen, Gaithersburg, MD). DNA immobilized on treated glass slides was digested with the 106

BamHI restriction enzyme, and stained with fluorescent dye. Contiguous immobilized 107

restriction fragments were sized and mapped across molecules ranging from 100,000 to 108

1,000,000 base pairs (bp). Images from a maximum of 50,000 molecules were collected and 109

assembled on the Argus mapping station, short DNA sequence contigs were assembled into a 110

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

6

complete circular chromosome. Optical maps were aligned and compared to reference 111

sequenced genomes (13). 112

Multiple-locus variable-number tandem-repeats analysis (MLVA). MLVA analysis was 113

conducted with the outbreak S. Bovismorbificans isolates using multiplex PCR and capillary 114

electrophoresis (QIAxcel System, Gaithersburg, MD). DNA templates were prepared from broth 115

cultures of LB grown at 37°C for 4 hrs. Cells were harvested and DNA was extracted using the 116

Wizard Genomic DNA Purification System (Promega, Madison, WI). Total DNA was 117

resuspended in 0.3 ml Tris-EDTA (Promega) buffer to give a final concentration 350 ng/ul to 118

550 ng/ul. S Enteritidis strain P125109 (Gene Bank AM933172.1) genome sequence was used to 119

search for variable number tandem repeats (VNTR) using the Tandem Repeat Finder (2, 19). 120

Five of the VNTR regions were selected and primers were designed using Primer3 (v. 0.4.0) 121

software (Primer3source forge.net) and designated as SE1, SE2 SE3, SE4 and SE5 (Table 2). 122

Two multiplex reactions for the five targeted loci (SE1-SE5) were used to generate MLVA 123

profiles with high diversity capacity using the QIAxcel bio-analyzer (QIAxcel System ). 124

PCR amplifications were performed using the Type-it Microsatellite PCR Kit (Qiagen 125

Inc., Balencia, CA) using a MJ Research PTC-200 Thermo Cycler (MJ Research, Ramsey, MN) 126

and 12.5 µl 2x Type-it Multiplex PCR Master Mix, 12.5 µl primer mix (2 µM of each primer), 6 127

µl RNase-free water, 5 µl 5x Q-Solution and 1.0 µl Template DNA. The PCR conditions were, 128

95 °C for 5 min; 35 cycles of 95°C for 30s, 60°C for 90s, 72°C for 30s; final extension at 68°C 129

for 10 min. PCR products were run on the QIAxcel System and the molecular sizes were 130

calculated using the Biocalculator software (eGene,Inc., Irvine CA) (19). 131

REP-PCR. The BioMerieux’s Diversilab® system is a Repetitive Extragenic Palindromic 132

(REP)-PCR based DNA fingerprinting analysis system (BioMerieux, Durham, NC). This method 133

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

7

was evaluated for its ability to differentiate or subtype the outbreak and non-outbreak isolates. 134

Cultures were grown for 18-24 h at 37±2°C on sheep Blood Agar (SBA). DNA was isolated 135

using Roche MagNa Pure Compact (Roche, Indianapolis, IN). Final DNA concentrations were 136

adjusted to 75 +/- 25 ng/µl. Data analysis was performed using BioMerieux’s Diversilab® 137

software. 138

Salmonella virulence plasmid analysis. The presence of the Salmonella virulence plasmid was 139

also assessed among the isolates. Plasmids were isolated using the Qiagen Plasmid Midi Kit 140

(Qiagen). Strains containing plasmid within the range of known Salmonella virulence plasmids 141

were evaluated for the presence of plasmid-associated virulence genes based on primers reported 142

by Carattoli et al., (4) and primers designed for this study that are specific to the origin of 143

replication genes such as qrmD, Orf4, Orf5, spvB and pef (Table 2) (4, 11). 144

145

RESULTS AND DISCUSSION 146

Molecular Serotyping: The Agilent Bio-analyzer 2100 provided a digital gel image depicting 147

the various sized amplicons which is used to generate the serotype-specific banding patterns 148

corresponding to each known serotype. Molecular serotyping was conducted and the results 149

confirmed that all 21 isolates used in this study are of the S. Bovismorbificans pattern BCEG as 150

reported by Jean-Gilles Beaubrun et al., (12) and confirming the results of the traditional 151

serotyping method. 152

PFGE: The PFGE banding pattern was identical for all S. Bovismorbificans isolates associated 153

with the hummus outbreak, and this pattern was distinguishable from the PFGE patterns obtained 154

from the non-outbreak isolates. There are five distinct clusters showing different banding 155

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

8

patterns (Figure 1). The S. Bovismorbificans isolates cultured directly from the hummus have a 156

PFGE pattern which is indistinguishable from the patterns of the outbreak clinical strains (Figure 157

1). As demonstrated in Figure 1 using XbaI, the hummus outbreak isolates contained several 158

fragments at approximately 880 kb, 523 kb, and 104-66 kb that are clearly distinguishable from 159

the fragments observed in the non-outbreak isolates (Figure 1). Concurrently, using BlnI, there is 160

a 336-291kb variable region in the outbreak isolates that is not observed in the non-outbreak 161

isolates. The results also show that SAL185 and SAL644 seem to be more closely related to 162

each other, but certainly distinguishable from the outbreak isolates. PFGE patterns demonstrate 163

that the outbreak isolates from DC_PHL distinctly cluster together compared to the other isolates 164

of S. Bovismorbificans. Isolates obtained from the Michigan State laboratory form two distinct 165

clusters (Figure 1). One isolate from Michigan (SAL 682), which was identified as an outbreak 166

isolate, matches the PFGE pattern obtained from the DC_PHL isolates. However, SAL676 did 167

not match the PFGE pattern although it was originally reported to be part of the outbreak. 168

Antimicrobial Susceptibility Testing: The antimicrobial drug (AMD) susceptibility assay 169

demonstrates that there is no resistance in the isolates and the results are represented in Figure 1. 170

From the 21 isolates tested, it is clear that 12 of the isolates are related to the same hummus 171

outbreak and 9 isolates are completely unrelated and all of the isolates of S. Bovismorbificans 172

were susceptible to the AMDs found in the antimicrobial panel. The presence of various 173

antibiotic resistant genes was examined using PCR in addition to the AMD test that was 174

conducted. For example the qrmD gene targeted on the plasmid analysis did not suggest the 175

presence of the quinolone antibiotic resistant genes. However, the consistent finding was that the 176

isolates from the hummus outbreak were certainly different as compared to the non-outbreak 177

isolates used in this study. 178

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

9

Optical Mapping: Optical mapping revealed that the outbreak strains possessed three unique 179

gene regions including 1) the evolutionary distances between outbreak and non-outbreak isolates, 180

2) the chromosome structure of the outbreak strain, which is usually compared to the most 181

similar sequenced Salmonella reference isolates from GenBank, 3) and lastly, a list of specific 182

chromosomal markers unique for the outbreak strain. Optical mapping of S. Bovismorbificans 183

outbreak isolate SAL610 (SL1490) shows that it is most closely related to another non-outbreak 184

Bovismorbificans isolate SAL185 (SL0062) with a 3.8% difference (Table 3). The next closest 185

serovars are S. Typhimurium LT2 and S. Saintpaul SARA23, which were 5.3% and 5.4% 186

different. Optical mapping also revealed 11 major regions (A-K), excluding plasmids, that are 187

different relative to S. Typhimurium LT2 (4,857,372 kbp), and that the 4.740 Mbp chromosome 188

of S. Bovismorbificans is shorter by 117 kbp (Figure 2a and Table 3). OM results also showed 189

that the plasmid is unique to S. Bovismorbificans involved in the hummus associated outbreak. 190

Four genomic regions of interest , A, B, C, and D are simple insertion/deletion events; 191

region E is a variant Gifsy-2 prophage, and regions H – K are variable syntenic regions (VSRs). 192

Additionally genomic regions of interest H and I appear to be composites as they contained both 193

a VSR and variant prophages. VSRs are highly similar segments of the chromosome between 194

strains containing enough single nucleotide diversity to create and eliminate restriction sites 195

within sets of syntenic genes. A level of 1-3% nucleotide divergence creates enough change in 196

restriction fragment patterns that portions of otherwise aligned chromosomes appear as 197

unaligned sections of similar length. Upon closer examination, patterns of restriction site loss 198

and gain can be seen within the unaligned segments (Figure 2b). 199

200

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

10

MLVA analysis. Multi-locus variable-number tandem repeat analysis (MLVA) consisted of 201

PCR-based amplification of five chromosomal loci and accurate sizing of the resulting sets of 202

PCR amplicons by electrophoresis. MLVA examines specific tandem repeats (TRs) found at 203

different genetic loci within the genome. The data from the MLVA analysis demonstrates clear 204

and distinguishable differences between the isolates from the hummus outbreak and the non-205

outbreak isolates (Figure 3a and 3b). Two multiplexes were used for the MLVA analysis of the 206

21 isolates of S. Bovismorbificans. The results show eight different MLVA banding patterns 207

amongst the isolates for the first multiplex reaction, although the isolates from the Hummus 208

outbreak cluster together with one consistent pattern (Figure 3a). The second multiplex 209

contained two targets, which seem to be present in 16 of the 21 isolates (Figures 3b). 210

REP-PCR: The results from the Diversilab®’s REP-PCR analysis show two main clusters, 211

which clearly separate the outbreaks isolates from the non-outbreak isolates (Figure 4). In 212

addition, the results from the REP-PCR analysis suggest that SAL185 and SAL644, which are in 213

house controls, clustered with the outbreak related isolates, which supports the hypothesis that 214

the repetitive region of SAL185 and SAL 644 may be more closely related to the hummus-215

associated outbreak isolates than some of the other S. Bovismorbificans isolates used in this 216

study. SAL185 and SAL644 contained a pattern of REP sequences that are ~90 (± 5) percent 217

similar to those of the hummus-associated outbreak isolates. Based on the grouping of the 23 218

isolates presented in Figures 1, 3a, 3b, and 4using PFGE, MLVA, and REP-PCR, the results 219

demonstrated that there were genetic differences between the isolates from the outbreak as 220

compared to the non-outbreak isolates. 221

Plasmid analysis. The size of the Salmonella virulence plasmid is serovar dependent, ranging 222

from ∼50 to 100 kb (1). Optical mapping suggest the presence of an 80 kb plasmid. The 223

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

11

plasmid was extracted and the origin of replication (ORi), repA gene was confirmed to indicate 224

the presence of an Inc FIIs plasmid in the S. Bovismorbificans isolates. Using the primer target 225

reported by Carattoli et al. (4) and primers designed for this study, the results shown in Table 4 226

confirm the presence of an Inc FIIs plasmid in the outbreak isolates. In addition, the PCR 227

product obtained suggests that the plasmids identified in the hummus outbreak isolates may be 228

closely related to plasmids found in S. Typhimurium. PCR products were obtained using the 229

primers targeted for pefA, spvB and Inc FIIs. 230

The plasmid in the outbreak isolates was confirmed using primer-specific targets as 231

demonstrated in Table 2. In a reports published in Genomes Announcements (11,12) the isolates 232

used in this study was whole genome sequenced and the results showed that outbreak isolates 233

contained two plasmids, one of which demonstrates similarity to the 93 Kb pSLT2 IncF-type 234

plasmid of S. Typhimurium and an Integration and Conjugative Element (ICE) similar to ICESb1 235

found in Salmonella bongori (17). The presence of recurrent and common plasmids in 236

epidemiologically unrelated Salmonella isolates of different serotypes suggests the successful 237

spread of these genetic determinants amongst Salmonella spp. This is to say that the hummus-238

associated strain may have acquired new genetic information through horizontal gene transfer 239

(HGT) from other Salmonella spp., thereby increasing their virulence. These findings are 240

significant since they suggest an uptake of a plasmid in the outbreak genome that may have 241

increased the probability of the pathogenicity of the hummus associated strains. The approach 242

used in this study using the Optical mapping and plasmid analysis could be applied to monitor 243

the circulation of plasmids within strains from different environments or to follow the horizontal 244

transmission of virulence genes among Enterobacteriaceae and track emerging pathogens. 245

246

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

12

CONCLUSION 247

Salmonella is one of the main causes of gastroenteritis in humans. S. Bovismorbificans have 248

been linked to foodborne outbreaks associated with alfalfa sprouts, homemade cheese, pasta 249

salad, striped bass, lettuce and pork (20). It has not been previously related to sesame paste 250

(tahini), which are used to make hummus. A report in MMWR and Food Safety News about the 251

District of Columbia outbreak investigation described the difficulties associated with the 252

hummus-associated outbreak in terms of identifying the probable cause or source of the 253

contamination (22, 23). In this study, S. Bovismorbificans isolates from the hummus-associated 254

outbreak were compared to non-outbreak related isolates using a polyphasic approach which 255

included PCR molecular serotyping, PFGE, AMD, optical mapping, MLVA, Diversilab®, and 256

plasmid analysis. The results showed the presence of plasmids in the hummus-associated isolates 257

that are not present in the other non-outbreak isolates. 258

The various molecular approaches used in this study support one another in determining 259

that the outbreak strains were unique unto themselves as compared to other non-outbreak S. 260

Bovismorbificans strains. In this multifaceted approach to investigate the uniqueness of the S. 261

Bovismorbificans isolates related to the hummus outbreak, PFGE analysis clustered the isolates 262

that were closely related, Optical mapping determined the presence of the plasmid unique to the 263

of outbreak isolates, and the other molecular approaches such as MLVA, and Diversilab also 264

show the uniqueness the outbreak isolates. Lastly, WGS results demonstrated the presence of 265

two plasmids, one of which possesses significant similarity to the S. Typhimurium 93 Kb pSLT2 266

IncF-type plasmid and an ICE similar to ICESb1 found in Salmonella bongori. 267

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

13

The expanded scientific knowledge though the use of this polyphasic molecular 268

approach to characterizing these pathogens has provided useful insights for both risk assessment 269

and methods development for food safety. The goal is to create faster method detection, 270

identification and characterization of Salmonellae in the event of an outbreak and thus improve 271

the safety of the food supply. Using Optical mapping and plasmid characterization may help 272

develop new approaches for epidemiology analysis. 273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

14

ACKNOWLEGMENTS 291

We acknowledge and thank everyone from the Department of Health from District of 292

Columbia Public Health Laboratory (DC_PHL) for providing the isolates used in this study, and 293

FDA scientists from the Center of Food Safety and Applied Nutrition (CFSAN), Office of 294

Applied Research and Safety Assessment (OARSA), Center for Veterinary Medicine (CVM), in 295

Laurel MD who were involved in the collaborative study. Also thanks to the Delaware and 296

Michigan Public Health for assisting with providing extra strains, Mark Mammel and Chris 297

Elkins at FDA for your support. 298

299

300

301

302

303

304

305

306

307

308

309

310

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

15

REFERENCES 311

1. Amavisit, P., D. Lightfoot, G. F. Browning, and P. F. Markham. 2003. Variation between 312

Pathogenic Serovars within Salmonella Pathogenicity Islands. J. Bacteriol. 185:3624-313

3635. 314

2. Benson, G. 1999. Tandem repeats finder: a program to analyze DNA sequences. Nucleic 315

Acids Res. 27:573–580. 316

3. BioNumerics. Keistraat 120, B-9830 Sint-Martens-Latem, Belgium. www.applied-317

maths.com. 318

4. Carattoli, A., A. Bertini, L. Villa, V. Falbo, K. L. Hopkins, E. J. Threlfall. 2005. 319

Identification of plasmids by PCR-based replicon typing. J. of Microbiological Meth. 63: 320

219-228. 321

5. Cavaco, L.M., H. Hasman, S. Xia, and F.M. Aarestrup. 2009. qnrD, a novel gene 322

conferring transferable quinolone resistance in Salmonella enterica Serovar Kentucky 323

and Bovismorbificans strains of Human Origin. Antimicrob. Agents Chemother. 53:603-324

608. 325

6. Clinical and Laboratory Standards Institute. Performance Standards for Antimicrobial 326

Susceptibility Testing; Seventeenth Informational Supplement (M100-S17). 2007. 327

CLSI, Wayne, Pa. 328


Disk and Dilution Susceptibility Tests for Bacteria Isolated from Animals; Informational 330

Supplement (M31-S1). 2004. CLSI, Wayne, Pa. 331

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

16


Disk and Dilution Susceptibility Tests for Bacteria Isolated from Animals; Approved 333

Standard – Second Edition (M31-A2). 2002. CLSI, Wayne, Pa. 334

9. Gopal Gopinath, Junia Jean-Gilles Beaubrun, Chris Grim, Morris Blaylock, Reginald 335

Blackwell, Sosina Merid, Alpha Diallo, and Darcy Hanes. 2014. Whole Genome 336

Sequences of six Salmonella enterica Serovar Bovismorbificans Isolates Associated with 337

a 2011 Multistate Hummus Outbreak. Genome Announcements under review. 338

10. Gopal Gopinath, Junia Jean-Gilles Beaubrun, Chris Grim, and Darcy Hanes. 2014. Draft 339

Genome Sequences of nine Salmonella enterica Serovar Bovismorbificans isolates from 340

various sources. Genome Announcements under review. 341

11. Heithoff, D. M., W. R. Shimp, P. W. Lau, G. Badie, E. Y. Enioutina, R. A. Daynes, B. A. 342

Byrne, J. K. House, and M. J. Mahan. 2008. Human Salmonella Clinical isolates distinct 343

from those of animal origin. Appl. Environ. Microbiol. 74:1757-1766. 344

12. Jean-Gilles Beaubrun, J., C. Cheng, K. Chen, L. Ewing, H. Wang, M.C. Agpaoa, M. J. 345

Huang, E. Dickey, J.M. Du, D.M. Williams-Hill, B. Hamilton, S.A. Micallef, R.E. 346

Rosenberg-Goldstein, A. George, S.W. Joseph, A.R. Sapkota, A.P. Jacobson, B.D. Tall, 347

M.H. Kothary, and D.E. Hanes. 2012. The use of a PCR-Based Method for Identification 348

of Salmonella enterica serotypes Newport, Typhimurium, Javiana and Saint Paul from 349

various food matrices. Food Microbiology, 31: 199-209. 350

13. Kotewicz ML, Mammel MK, LeClerc JE, Cebula TA. 2008. Optical mapping and 454 351

sequencing of Escherichia coli O157:H7 isolates linked to the US 2006 spinach-352

associated outbreak. Microbiology 154:3518–3528. 353

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

17

14. Kim, S., J. G. Frye, J. Hu, P. J. Fedorka-Cray, R. Gautom, and D. S. Boyle. 2006. 354

Multiplex PCR-Based Method for Identification of Common Clinical Serotypes of 355

Salmonella enterica subsp. Enterica. J. Clin. Microbiol. 44:3608-3615. 356

15. Po¨nka¨, A., Y. Andersson, A. Siitonen, B. de Yong, M. Jahkola, O. Haikala, A. 357

Kuhmonen, and P. Pakkala. 1995. Salmonella in alfalfa sprouts. Lancet 345:462–463. 358

16. Puohiniemi, R., T. Heiskanen, and A. Siitonen. 1997. Molecular epidemiology of two 359

international sprout-borne Salmonella outbreaks. J. Clin. Microbiol. 35:2487-91. 360

17. Seth-Smith HM, Fookes MC, Okoro CK, Baker S, Harris SR, Scott P, Pickard D,Quail 361

MA, Churcher C, Sanders M, Harmse J, Dougan G, Parkhill J, Thomson NR.Structure, 362

diversity, and mobility of the Salmonella pathogenicity island 7 family of integrative and 363

conjugative elements within Enterobacteriaceae. J. Bacteriol. 2012 Mar;194(6):1494-504. 364

18. Rimhanen-Finne, R., R. T. Niskanen, T. Lienemann, T. Johansson, M. Sjo¨man, T. 365

Korhonen, S. Guedes, H. Kuronen, M. J. Virtanen, J. Ma¨ kinen, J. Jokinen, A. Siitonen 366

and M. Kuusi. 2011. Nationwide Outbreak of Salmonella Bovismorbificans Associated 367

with Sprouted Alfalfa Seeds in Finland, 2009. Zoonoses Public Health. 58: 589–596. 368

19. Rozen, S. and H. J. Skaletsky. 2000. Primer3 on the WWW for general users and for 369

biologist programmers. In: Krawetz S, Misener S (eds) Bioinformatics Methods and 370

Protocols: Methods in Molecular Biology. Humana Press, Totowa, NJ, pp 365-386 371

20. www.cdc.gov/salmonella/outbreaks.html 372

21. www.cdc.gov/pulsenet/protocols.html 373

22. www.cdc.gov/mmwr/preview/mmwrhtml/mm6146a3.htm. Multistate Outbreak of 374

Salmonella Serotype Bovismorbificans Infections Associated with Hummus 375

and Tahini — United States, 2011. 376

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

18

23. www.foodsafetynews.com/2012/11/government-reports-2011-salmonella-outbreak-377

linked-to-tahini. CDC: Salmonella from Tahini Sickened 23 Last Year. 378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

19

FIGURES 400

Figure 1: PFGE and AMD results of the Salmonella enterica serovar Bovismorbificans. 401

402

Figure 2a: An Optical Mapping image representing the eleven regions of differences (A-K) 403

between SAL185 (SL62), SAL610 (SL1490), and from reference strain LT2. 404

405

Figure 2b: An Optical Mapping image representing the variable syntenic regions of SAL185, 406

SAL609 and from reference strain LT2. 407

408

Figure 3a: QIAxcel System, Capillary electrophoresis sizing run for PCR products of Three TR 409

loci from isolates of Salmonella enterica serovar Bovismorbificans, hummus-associated outbreak 410

isolates (SAL609-SAL619 and SAL682), compared to the non-outbreak strains (SAL676-411

SAL683) and in house reference strain SAL185. 412

413

Figure 3b: QIAxcel System, Capillary electrophoresis sizing run for PCR products of two TR 414

loci from Isolates of Salmonella enterica serovar Bovismorbificans, hummus-associated 415

outbreak isolates (SAL609-SAL619 and SAL682), compared to the non-outbreak strains 416

(SAL676-SAL683) and in house reference strain SAL185. 417

418

Figure 4: A representation of the Rep-PCR results using Diversilab® software to demonstrate in 419

a dendogram, digital gel image the percent relatedness between hummus-associated outbreak 420

isolates (SAL609-SAL619 and SAL682), compared to the non-outbreak strains (SAL676-421

SAL683) and in house reference strain SAL185. 422

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

20

TABLES 423

Table 1. Summary of all the strains used in this study and the appropriate identification from 424

various laboratories. 425

426

Table 2. List of the primers used in this study for molecular characterization of Salmonella 427

serovar Bovismorbificans. 428

429

Table 3. . Summary of the Optical Mapping gene relatedness between Salmonella serovars 430

Bovismorbificans, Typhimurium, Saintpaul, Newport, Kentucky and Schwarzengrund. 431

432

Table 4. Results of the plasmids PCR analysis of Salmonella serovar Bovismorbificans from the 433

Hummus associated outbreak isolates compared to non-outbreak isolates. 434

435

436

437

438

439

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

ID DMB CVM DCPHL Source Serotypes Location Outbreak

SAL194 SARA2 LT2 Typhimurium

SAL609 SL1489 41673 11-399 Stool-RFBovismorbificans DC Hummus







SAL616 SL1496 41680 11-528 FoodBovismorbificans DC Hummus




SAL676 SL1542 41838 12-180 Clinical Bovismorbificans DE Hummus

SAL677 SL1543 41839 12-198 Clinical Bovismorbificans MI N/A





SAL682 SL1548 41844 12-203 Clinical Bovismorbificans MI Hummus


SAL185 SL0062 41672 Clinical Bovismorbificans MI N/A

SAL644 SL1541 17906 Human Bovismorbificans CVM N/A

SAL155 SL1541 Tomato Farm Newport VA N/A

Table 1. Summary of all the strains used in this study and the appropriate identification from various laboratories.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table 2: List of the primers used in this study for molecular characterization of Salmonella serovar Bovismorbificans.

Primer Primer sequence Target Size bp

Forward Reverse

Plasmid 5’-TATTGCACTGGGTGTTCTGGGC-3’ 5’-TTGTAAGCCACTGCGAAAGACGC-3’ Pef A 473 This study

Plasmid 5’-AGCATCTCTGCCTCTCCCTTA-3’ 5’-CAGTATACCGCTGCCT-3’ spvB 1491 This study

Plasmid 5’-AGTTTTCACGAGATCAATTTACG-3’ 5’-CCTAAACTCTCAACAAGCTGAAG-3’ qnrD 578 This study

Plasmid5’-TCACACTCCTGCTGTCGGCGGAGC-3’

5’-TCCGTTGTCAGCAGCTTTTCAAGG-3’ orf5 454 This study

Plasmid 5’-TCAGTCAGCGAAAAGATCTCGAAC-3’ 5’-TGCTGCGAGATATCATGCGTCAGAG-3’ orf4 732 This study

Plasmid 5’-CTGTCGTAAGCTGATGGC-3’ 5’-CTCTGCCACAAACTTCAGC-3’ repA 270 Carattoli et. al.

MLVA 5’-AGGAAACACAGCCGCAAT-3’ 5’-AACCGGCTGTTCGATAACC-3’ SE1 151 This study

MLVA 5’-GCGCTAAACAAGCCGCTCAT-3’ 5’-TCGTTAACAACCTGCTGCTGT-3’ SE2 413 This study

MLVA 5’-ATGTTCTGGCGGACATGG-3’ 5’-GGCATCCTGATACGCTTTTG-3’ SE3 116 This study

MLVA 5’-TGGCTATTGGCGTTGAAAAT-3’ 5’-AAGGTAATAACAGAGTCATC-3’ SE4 488 This study

MLVA 5’-TCCTGCATACCAACGATCAC-3’ 5’-CTCTTGCGCCACCTCTTTAC-3’ SE5 112 This study

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table 3. Summary of the Optical Mapping gene relatedness between Salmonella serovars Bovismorbificans, Typhimurium, Saintpaul, Newport, Kentucky and Schwarzengrund.

runC Ncol SAL185/

SL0062

SAL610/S

L1490

SAL194/L

T2

Saintpaul

in silico

Saintpaul

O Map

Newport Kentucky Schwarzengrund

in silico

Serovar Bovismorbificans SL0062 [Ncol] 0.0% 3.8% 5.3% 5.4% 13.0% 14.7% 20.1% 36.3%

Serovar Bovismorbificans SL1490 [Ncol] 3.8% 0,0% 9.3% 7.8% 7.3% 14.5% 22.3% 30.0%

Serovar Typhimurium LT2 dwl [Ncol] in silico 5.3% 9.3% 0.0% 5.1% 5.1% 4.1% 7.2% 8.0%

Serovar Saintpaul SARA23 gcontig 111272 5.4% 7.8% 5.1% 0.0% 0.6% 3.2% 5.1% 11.5%

Serovar Saintpaul O Map SARA23 [Ncol] 13.0% 7.3% 5.1% 0.6% 0.0% 8.6% 7.9% 17.7%

Serovar Newport SL254 [Ncol] in silico 14.7% 14.5% 4.1% 3.2% 8.6% 0.0% 7.8% 9.3%

Serovar Kentucky CVM29188 [Ncol] in silico 20.1% 22.3% 7.2% 5.1% 7.9% 7.8% 0.0% 21.5%

Serovar Schwarzengrund CVM19633 in silico 36.3% 30.0% 8.0% 11.5% 17.7% 9.3% 21.5 0.0%

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Lab ID Serotypes qnrDa Orf4a Orf5a pefAa

500 bp

spvBa

1300 bp

FIIsb

270 bp

*SAL194 Typhimurium - - - + + +

SAL609 Bovismorbificans - - - + + +











*SAL676 Bovismorbificans - - - + - -

*SAL677 Bovismorbificans - - - + - +





SAL682 Bovismorbificans - - - + - +


*SAL185 Bovismorbificans - - - - - -

*SAL644 Bovismorbificans - - - - - -

Table 4. Results of the plasmids PCR analysis of Salmonella serovar Bovismorbificans from the Hummus associated outbreak

isolates compared to non-outbreak isolates.

* non-outbreak isolates, a is primers designed for this study, b Carattoli et al. (15)

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 1: PFGE and MIC results of the Salmonella enterica serovar Bovismorbificans.

880 523 325-280

104-66

336-291

104-72

kb

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 2a: An Optical Mapping image representing the eleven regions of differences (A-K) between SAL185 (SL62), SAL610 (SL1490), and from reference strain LT2.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 2b: An Optical Mapping image representing the variable syntenic regions of SAL185 (SL62), SAL610 (SL1490) and from reference strain LT2.

#fragments locus1 locus 2 locus 3 locus 4

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 3a: QIAxcel System, Capillary electrophoresis sizing run for PCR products of Three TR loci from isolates of Salmonella entericaserovar Bovismorbificans, hummus-associated outbreak isolates (SAL609-SAL619 and SAL682), compared to the non-outbreak strains (SAL676-SAL683) and in house reference strain SAL185.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 3b: QIAxcel System, Capillary electrophoresis sizing run for PCR products of two TR loci from Isolates of Salmonella entericaserovar Bovismorbificans, hummus-associated outbreak isolates (SAL609-SAL619 and SAL682), compared to the non-outbreak strains (SAL676-SAL683) and in house reference strain SAL185.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Rep-PCR

Figure 4: A representation of the Rep-PCR results using Diversilab® software to demonstrate in a dendogram, digital gel image the percent relatedness between hummus-associated outbreak isolates (SAL609-SAL619 and SAL682), compared to the non-outbreak strains(SAL676-SAL683) and in house reference strain SAL185.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Research Highlights

• Comparing Bovismorbificans outbreak versus non outbreak isolates.

• Identification of a plasmid in the outbreak isolates.

• Using multiple molecular approaches for characterization of plasmids

comparison of salmonella enterica serovar bovismorbificans 2011 hummus outbreak strains with...

Documents