Nontyphoidal Salmonella spp. are the leading bacterial

Transcription

1 FOODBORNE PATHOGENS AND DISEASE Volume XX, Number XX, 2018 ª Mary Ann Liebert, Inc. DOI: /fpd Original Article CRISPR Typing and Antibiotic Resistance Correlates with Polyphyletic Distribution in Human Isolates of Salmonella Kentucky Dorothy Vosik, 1 Deepanker Tewari, 2 Lisa Dettinger, 3 Nkuchia M. M ikanatha, 4,5 and Nikki W. Shariat 1 Abstract Although infrequently associated with reported salmonellosis in humans, Salmonella enterica, subsp. enterica serovar Kentucky (ser. Kentucky) is the most common nonclinical, nonhuman serovar reported in the United States. The goal of this study was to use Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) multi-virulence-locus sequence typing (MVLST) to subtype a collection of human clinical isolates of ser. Kentucky submitted to the Pennsylvania Department of Health and to determine the extent of antibiotic resistance in these strains. This analysis highlighted the polyphyletic nature of ser. Kentucky, and separated our isolates into two groups, Group I and Group II, which were equally represented in our collection. Furthermore, antimicrobial susceptibility testing on all isolates using a National Antimicrobial Resistance Monitoring System (NARMS) panel of antibiotics demonstrated that resistance profiles could be divided into two groups. Group I isolates were resistant to cephems and penicillins, whereas Group II isolates were resistant to quinolones, gentamicin, and sulfisoxazole. Collectively, 50% of isolates were resistant to three or more classes of antibiotics and 30% were resistant to five or more classes. The correlation of antibiotic resistance with the two different lineages may reflect adaptation within two distinct reservoirs of ser. Kentucky, with differential exposure to antimicrobials. Keywords: Salmonella Kentucky, antibiotic resistance, CRISPRs Introduction Nontyphoidal Salmonella spp. are the leading bacterial cause of foodborne illness in the United States with over 1 million cases annually (Scallan et al., 2011). Serovar Kentucky, is the most common nonhuman serovar reported, accounting for 15% of all laboratory-confirmed serovars (CDC, 2014). It is the most frequent serovar found in chickens, second found in retail chicken samples, and in the top 10 serovars found in cattle (NARMS, 2014). Serovar Kentucky is not commonly associated with human illness in the United States, as *100 cases are reported to the Centers for Disease Control and Prevention each year (CDC, 2017). However, it is frequently reported elsewhere in the world; in Europe it accounts for 1% of annual salmonellosis cases (European Food Safety Authority and European Center for Disease Prevention and Control, 2016) and its incidence in poultry has increased (reviewed in Antunes et al., 2016). A study showed that from 2003 to 2011, 13% of all Salmonella isolates obtained in Morocco were ser. Kentucky (Le Hello et al., 2013). The National Antimicrobial Resistance Monitoring System (NARMS) was established in 1996 among state health officials, the Centers for Disease Control and Prevention, the United States Food and Drug Administration, and the United States Department of Agriculture. One component of NARMS is performing antimicrobial susceptibility testing in Salmonella from human clinical isolates, retail meats, and food animals. In 2014, 57% of ser. Kentucky isolates from chickens in the United States were resistant to two to three classes of antimicrobials and 11% were resistant to four to five classes (NARMS, 2014). CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats)-based typing has been used in many different types of pathogenic bacteria, including Salmonella, 1 Department of Biology, Gettysburg College, Gettysburg, Pennsylvania. 2 Pennsylvania Veterinary Laboratory, Pennsylvania Department of Agriculture, Harrisburg, Pennsylvania. 3 Bureau of Laboratories, Pennsylvania Department of Health, Exton, Pennsylvania. 4 Division of Infectious Disease Epidemiology, Pennsylvania Department of Health, Harrisburg, Pennsylvania. 5 Center for Clinical Epidemiology and Biostatistics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania. 1

2 2 VOSIK ET AL. Escherichia coli, and Campylobacter jejuni (reviewed in Shariat and Dudley, 2014). There are two CRISPR arrays in Salmonella, CRISPR1 and CRISPR2, and each comprises invariant direct repeat sequences of 29 nucleotides, separated by highly variable spacer sequences of 32 nucleotides (Touchon and Rocha, 2010). CRISPR spacer content is typically serovar specific as unrelated serovars rarely contain the same spacers (Fabre et al., 2012). Within a serovar, spacer duplication or deletion, or the presence of single-nucleotide polymorphisms (SNPs) allow for precise subtyping of isolates and has been demonstrated in multiple Salmonella serovars to date (Fabre et al., 2012; Shariat et al., 2015). CRISPR-based approaches have been successfully used to subtype Salmonella, including outbreak isolates (Liu et al., 2011; Fabre et al., 2012; Shariat et al., 2013b; Li et al., 2014). CRISPR multi-virulence-locus sequence typing (MVLST) combines a multilocus sequence typing (MLST) approach using sequence analysis of two virulence (MVLST) genes, fimh and ssel. Comprehensive analysis of four major serovars shown to cause illness (serovars Enteritidis, Heidelberg, Newport, and Typhimurium) demonstrated that CRISPR MVLST and pulsed-field gel electrophoresis (PFGE) are comparable methods for strain discrimination (Liu et al., 2011; Shariat et al., 2013a, 2013b, 2013c). Like some Salmonella serovars, such as ser. Newport, ser. Kentucky is polyphyletic (Sukhnanand et al., 2005; Sangal et al., 2010; Achtman et al., 2012; Timme et al., 2013; Haley et al., 2016), although the CRISPR patterns and antibiotic resistance profiles across both groups from isolates in the United States have not been investigated. Our goal, in this study, was to use CRISPR MVLST to subtype human clinical isolates of ser. Kentucky collected from Pennsylvania. We wanted to use this approach to examine strain diversity and determine whether antibiotic Table 1. List of the 40 Salmonella Serovar Kentucky Isolates Isolated by the Pennsylvania Department of Health Between 2007 and 2015, Analyzed by CRISPR MVLST and Pulsed-Field Gel Electrophoresis Isolate CRISPR MVLST PFGE Isolation date PA region Source M KST1 JGPX March 2015 NW Urine 08E02438 KST1 JGPX December 2008 SE Stool M A KST1 JGPX September 2009 SE Stool M A KST1 JGPX December 2009 SE Stool M A KST1 JGPX February 2010 SE Stool 08E00076 KST1 JGPX January 2008 SE Stool M A KST1 JGPX August 2010 NC Stool M A KST2 JGPX June 2011 SC Stool M A KST2 JGPX August 2009 NE Stool M KST2 JGPX December 2014 SW Stool M A KST2 JGPX October 2013 SE Stool M KST3 JGPX March 2015 NW Stool M A KST4 JGPX December 2009 NE Stool 07E00390 KST5 JGPX March 2007 SE Stool M A KST6 JGPX November 2010 NC Stool M KST7 JGPX March 2014 SE Stool M KST8 JGPX December 2014 SC Urine 08E00093 KST8 JGPX January 2008 SE Stool M A KST9 JGPX January 2011 SE Stool M A KST9 JGPX December 2012 SE Stool M A KST9 JGPX April 2013 SE Stool M A KST9 JGPX April 2013 SE Stool M A KST9 JGPX May 2013 SE Stool M A KST9 JGPX July 2012 SE Stool M KST9 JGPX July 2015 SW Stool 09E00042 KST9 JGPX January 2009 SE Stool M A KST9 JGPX November 2010 SE Urine M A KST9 JGPX January 2012 SE Stool M KST9 JGPX October 2014 NE Stool M A KST10 JGPX May 2011 SE Stool M A KST10 JGPX September 2011 SC Stool M A KST10 JGPX August 2012 SE Stool M A KST10 JGPX January 2013 NW Urine M KST10 JGPX May 2014 SE Blood 07E01747 KST11 JGPX August 2007 SE Gallbladder fluid 07E02462 KST11 JGPX October 2007 Unknown Stool 07E02258 KST11 JGPX November 2007 SE Urine M KST12 JGPX January 2015 NC Stool M A KST13 JIXX01 April 2013 SE Stool M A KST14 JGPX July 2013 NE Stool CRISPR, Clustered Regularly Interspaced Short Palindromic Repeats; KST, Kentucky sequence type; MVLST, multi-virulence-locus sequence typing; PFGE, pulsed-field gel electrophoresis.

3 SALMONELLA KENTUCKY POLYPHYLY AND ANTIBIOTIC RESISTANCE 3 resistance correlates with subtyping analysis. Our results show that CRISPR-typing identifies two distinct groups, Groups I and II, reflecting ser. Kentucky polyphyly, and which exhibit differential patterns of antibiotic resistance. Materials and Methods Bacterial isolates Forty clinical isolates of ser. Kentucky were obtained from the Pennsylvania Department of Health (Table 1), and comprise all ser. Kentucky isolates submitted to the Pennsylvania Department of Health between March 2007 and July The majority of isolates (24/40) was isolated from patients in southeastern Pennsylvania. After streaking on LB plates, single colonies were propagated overnight in 2 ml LB at 37 C. One milliliter of the culture was cryopreserved in 20% glycerol. Six hundred microliters of the culture was used to isolate genomic DNA using the Promega (Madison, WI) Genomic DNA Isolation Kit, following the manufacturer s directions. DNA pellets were resuspended in 200 ll of sterile water and stored at-20 C. Polymerase chain reaction amplification and DNA sequencing Polymerase chain reaction (PCR) amplification was performed on four genomic loci: CRISPR1, CRISPR2, fimh, and ssel, as previously described (Shariat et al., 2013b). Five microliters of each PCR was run on a 1% gel to confirm presence of a product. The remaining PCR was treated with exonuclease and Antarctic alkaline phosphatase (New England BioLabs) as previously described (Shariat et al., 2013b). Purified PCR products were sequenced at Eton Bioscience (Union, NJ). Some CRISPR1 and CRISPR2 arrays were longer (up to 2 kb) and necessitated additional sequencing primers designed to match internal spacers for sequencing. All primers used are indicated in Table 2. Sequence type analysis and pulsed-field gel electrophoresis Sequences were assembled and aligned using SeqMan and MegAlign (Lasergene 13; Madison, WI). Both CRISPRs were further analyzed using CRISPR-finder ( psud.fr/server/) (Grissa et al., 2007) to determine the spacer content. Spacers were visualized using a CRISPR macro. For each of the four loci, allelic identifiers were given to unique sequences and these are available on GenBank (MF MF104698). For each individual ser. Kentucky isolate, the combination of allelic types at each of the four loci defined the Kentucky sequence type (KST). Sequence types (STs) were analyzed using eburst ( using default parameters for four genes. PFGE using XbaI was performed using standard protocols (Ribot et al., 2006; Sandt et al., 2006). Antibiotic resistance testing Salmonella isolates were plated onto 5% sheep blood agar, and incubated overnight at 35 C. Suspensions equivalent to a 0.5 McFarland standard were prepared and adjusted to colony-forming unit [CFU]/mL in Cation-Adjusted Mueller-Hinton Broth. The suspension was used to inoculate the CMV3AGNF panel (SENSITITRE Ô ; Trek Diagnostics, Table 2. List of Polymerase Chain Reaction and Sequencing Primers Used for CRISPR MVLST Primer use ser. Kentucky PCR Sequencing group References Annealing temperature Marker Primer sequence Orientation CRISPR1-1 GATGTAGTGCGGATAATGCT Reverse 55.5 Yes Yes Group I/II Fabre et al. (2012) CRISPR1-5 GGTTTCTTTTCTTCCTGTTG Forward Yes Yes Group I Fabre et al. (2012) CRISPR1-7 CGTATTCCGGTAGATBTDGATGG Forward Yes Yes Group II Liu et al. (2011) CRISPR1-8 ACGGTAGGAATATTCACGTTTTTTA Forward Yes (i) Group II This paper CRISPR1-9 TCGATATGAACGACGCGCAAG Reverse Yes (i) Group II This paper CRISPR2-3 ATTGTTGCGATTATGTTGGT Forward 55 Yes Yes Group I/II Shariat et al. (2013a) CRISPR2-4 TCCAGCTCCCTTATGATTTT Reverse Yes Group I/II Shariat et al. (2013b) CRISPR2-1 GCAATACCCTGATCCTTAACGCCA Reverse Yes Group I/II Shariat et al. (2013a) CRISPR2-11 GGATCGTCAGCACTCTGAACC Forward Yes (i) Group II This paper fimh-1 AGGTGAACTGTTCATCCAGTGG Forward 56.7 Yes Yes Group I/II Shariat et al. (2013a) fimh-2 GCGGGCTGAACAAAACACAA Reverse Yes Yes Group I/II Shariat et al. (2013a) ssel-1 AAAATCAGGTCTATGCCTGATTTAATATATC Forward 60 Yes Group I/II Shariat et al. (2013a) ssel-5 GGCTCTAAGTACTCRCCATTACT Reverse Yes Group I/II Shariat et al. (2013a) ssel-3 ACCAGGAAACAGAGCAAAATGAATATATGT Forward Yes Group I/II Shariat et al. (2013a) ssel-4 TTCTCTCGGTAAACTATCCTATTGGGC Reverse Yes Group I/II Shariat et al. (2013a) (i), internal seq primer; PCR, polymerase chain reaction.

4 4 VOSIK ET AL. Cleveland, OH) and incubated at 35 C for 18 h. All isolates were tested using a NARMS protocol for 14 drugs, amoxicillin, ampicillin, augmentin (amoxicillin/clavulanic acid) azithromycin, cefoxitin, ceftriaxone, ciprofloxacin, chloramphenicol, gentamicin, nalidixic acid, streptomycin, sulfisoxazole, tetracycline, and trimethoprim/sulfamethoxazole. Collectively, these 14 antibiotics represent 9 distinct classes of antimicrobials. Minimum inhibitory concentrations for drugs were tested and deduced based on defined breakpoints (NARMS, 2014; CLSI, 2015a, 2015b). A two-tailed t-test was used to determine significance between antibiotic susceptibility of the two ser. Kentucky groups. For multidrug resistance, intermediate resistance was not considered. Results CRISPR MVLST and ST distribution We used CRISPR MVLST to subtype 40 ser. Kentucky isolates, which include all ser. Kentucky samples submitted to the Pennsylvania Department of Health over a 9-year period. Among these, we identified 13 CRISPR1 alleles, 8 CRISPR2 alleles, 2 fimh alleles, and 2 ssel alleles. CRISPR alleles were characterized by deletion or duplication of individual spacers, whereas allelic differences in fimh and ssel were defined by at least one SNP. By combining the allelic variation at each of the 4 loci, we defined 14 unique ser. KSTs (Fig. 1). The three most frequent KSTs we identified were KST9 (11/40 isolates; 27.5%), KST1 (7/40; 17.5%), and KST10 (5/40; 12.5%). Eight KSTs were represented by single isolates. The calculated discriminatory power using CRISPR MVLST was The CRISPR profiles (i.e., the spacer composition) of different Salmonella serovars are distinct. Within this collection of ser. Kentucky isolates, the CRISPR profiles separate the collection into two groups, Group I and Group II, with no single spacer found in both groups (Fig. 1). This analysis reflects the polyphyletic nature of this serovar, which is also observed in PFGE analysis of the same isolates (Fig. 2). The distinct composition of spacers within the CRISPR arrays demonstrates the separate evolutionary paths of the two groups. The two groups are evenly represented in our isolate collection: there are 18 Group I isolates (KST1-8) and 22 Group II isolates (KST9-14). There was no allelic overlap in any of the four loci across the two groups. Isolates within each group possess a single allele for fimh and also for ssel, showingthatforser.kentucky,thesetwoalleles do not contribute to the subtyping scheme. Using Basic Local Alignment Search Tool (BLAST), we determined that the nucleotide identity between the two alleles of fimh was 1017/1035 (98%) and 968/981 (99%) between the two ssel alleles. BURST (Based Upon Related Sequence Types) analysis demonstrates similarities between MLST groups, so we applied this to our CRISPR MVLST analysis. We showed that KST1 6, and -8 group fall into a single BURST group (BURST Group1), and that KST10 and -12 fall into BURST Group2 (Supplementary Fig. S1; Supplementary Data are available online at KST7, -9, -11, -13, and -14 were singletons. Unsurprisingly, there was no overlap between Group I and II KSTs with respect to BURST group allocation. Additionally, there was no correlation between location or date of isolation among the two groups (data not shown). FIG. 1. CRISPR MVLST separates ser. Kentucky into two distinct groups. Graphic representation of the CRISPR arrays for CRISPR1 and CRISPR2 for the 14 ser. KSTs identified in this study. The number of isolates, out of the 40 analyzed, with a given KST is shown in italics. The number that identifies the fimh or ssel allele is shown and the number of the CRISPR allele is listed to the left of the respective CRISPR array. Each unique spacer is represented by a uniquely colored box, and the direct repeats are not shown. The spacers are aligned and the gaps represent the absence of a particular spacer. The direction of the spacers is from 5 to 3 with respect to transcription of the CRISPR loci. CRISPR, Clustered Regularly Interspaced Short Palindromic Repeats; KST, Kentucky sequence type; MVLST, multi-virulence-locus sequence typing. Color images available online at

5 SALMONELLA KENTUCKY POLYPHYLY AND ANTIBIOTIC RESISTANCE 5 isolates are sensitive to these antibiotics. Conversely, Group II isolates show resistance to nalidixic acid, sulfisoxazole, gentamicin, and ciprofloxacin, whereas Group I isolates are largely susceptible to these antibiotics (Fig. 3b). The antibiotic resistance profiles for individual isolates are shown in Supplementary Table S1. In total, 50% of the ser. Kentucky isolates (20/40) submitted to the Pennsylvania Department of Health during exhibited resistance to more than three classes of antibiotics. Of these, 60% were resistant to five or more classes of antibiotics (Supplementary Table S1). Importantly, 10/11 KST9 isolates (representing 25% of all isolates) are resistant to ciprofloxacin. Only a quarter of the strains (10/40) were susceptible to all antibiotics tested. FIG. 2. Polyphyly is reflected in PFGE analysis. PFGE using XbaI of all 40 isolates is able to separate the 2 groups of ser. Kentucky as indicated. PFGE, pulsed-field gel electrophoresis. Antimicrobial susceptibility testing We sought to determine the antibiotic resistance profiles for all 40 isolates in our collection. Among the 14 antibiotics tested, we observed broad resistance against tetracycline (56% of isolates), ampicillin (48%), and streptomycin (43%) (Fig. 3a). A single isolate was resistant to chloramphenicol and trimethoprim sulfamethoxazole (M A; KST9) and all isolates were susceptible to azithromycin. Interestingly, resistance profiles against additional antibiotics correlate with the polyphyletic group. Isolates from Group I are resistant to amoxicillin, cefoxitin, and ceftriaxone and show intermediate resistance to augmentin, whereas Group II Correlation with MLST studies Aseven-geneMLSTschemeisroutinelyusedtotypeSalmonella and two common ser. Kentucky STs include ST152 and ST198 (EnteroBase; Available at: We performed in silico analysis to determine the consensus of CRISPR1 and CRISPR2 arrays from published genomes of 110 ser. Kentucky (Supplementary Fig. S2). All 13 consensus array pairs (i.e., both CRISPR1 and CRISPR2) from ST152 isolates (representing 101 genomes) fall into Group I and both ST198 consensus array pairs (nine genomes) fall into Group II. ST198 and ST152 do not have any allelic overlap (0/7 alleles), similar to the CRISPR MVLST data, where 0/4 alleles overlap across the two Groups I and II. Discussion Data presented in this study shows that CRISPR MVLST is able to separate 40 ser. Kentucky isolates into 14 distinct KSTs, KST1 KST14. The majority of strain diversity is resolved by analysis of the CRISPR1 locus, which shows the greatest polymorphisms across different strains. As seen in other Salmonella serovars, neither CRISPR locus appears to be adapting through acquisition of new spacers (Shariat et al., 2015). Rather, allelic differences arise from deletion or duplication of individual spacer-repeat units. These data are comparable to CRISPR MVLST analysis of other serovars (also human derived), for example, analysis of 141 ser. Enteritidis isolates generated 22 Enteritidis STs, 89 ser. Heidelberg isolates generated 21 Heidelberg STs, and 86 ser. Typhimurium generated 37 Typhimurium STs (Shariat et al., 2013a, 2013c). Serovar Kentucky is polyphyletic (Sukhnanand et al., 2005; Achtman et al., 2012; Timme et al., 2013; Haley et al., 2016) and both phyletic groups are equally represented in our collection of isolates, which consists of all ser. Kentucky isolates that were submitted to the Pennsylvania Department of Health over a 9-year period. We termed these two lineages Group I and Group II. KST1 (17.5% of all isolates) and KST9 (27.5%) were the two most frequent types of each group. Whole genome sequence data of multiple ser. Kentucky isolates have shown that the two groups bear less resemblance to each other than to other Salmonella serovars (Timme et al., 2013; Haley et al., 2016), and this is reflected in our allelic data, where no alleles are shared between the two groups. Antibiotic susceptibility testing highlighted differences between Groups I and II. While many isolates exhibit resistance to the antibiotics tetracycline (58% of isolates), ampicillin (50%), streptomycin (43%), and intermediate or full resistance

6 a b FIG. 3. Antibiotic resistance correlates with phyletic group. (a) Heat map showing the percentage of isolates in each KST that are resistant to the indicated antibiotic. The inset shows how the color-coding presents the percentage ranges of resistance. (b) Graph showing the collective antibiotic resistance of isolates from each group. For both the heat map and bar graph, Group I isolates are shown in blue and Group II in orange. Intermediate resistance is shown in gray. b t-tests showed significance between Group I and II resistance ( p < 0.05); a p < 0.05 when considering full resistance, but not when considering intermediate resistance. The abbreviations for the antibiotics are as follows: Amp, ampicillin; Aug, augmentin; Axo, amoxicillin; Azi, azithromycin; Chl, chloramphenicol; Cip, ciprofloxacin; Fis, sulfisoxazole; Fox, cefoxitin; Gen, gentamicin; Nal, nalidixic acid; Str, streptomycin; Sxt, trimethoprim sulfamethoxazole; Tet, tetracycline; Xnl, ceftriaxone. KST, Kentucky sequence type. Color images available online at 6

7 SALMONELLA KENTUCKY POLYPHYLY AND ANTIBIOTIC RESISTANCE 7 to augmentin (45%), there were surprising differences when the groups were considered separately. Group I isolates exhibited resistance against amoxicillin and two cephems (cefoxitin and ceftriaxone) and are largely resistant to cephems and penicillins. Conversely, Group II isolates are resistant to quinolones (nalidixic acid and the fluoroquinolone, ciprofloxacin), gentamicin, and sulfisoxazole. The fact that half the isolates in our collection showed multidrug resistance to three or more classes of antibiotics complicates treatment options in patients with invasive ser. Kentucky infections. Other studies examining ser. Kentucky have used MLST analysis to identify two common STs, ST198 and ST152. Our in silico data, using genome sequences available on National Center for Biotechnology Information (NCBI) (Haley et al., 2016) (Supplementary Fig. S2), show that all ST152 isolates (using 13 consensus sequences that were derived from 101 individual genomes) were in CRISPR Group I and that ST198 isolates (from 2 consensus sequences/9 genomes) fall into Group II. ST152 isolates are responsible for the majority of ser. KSTs identified in nonhuman samples in the United States (Le Hello et al., 2011; Haley et al., 2016). In a study using whole genome sequencing of 112 poultry and cattle-derived isolates in the United States, 90% were ST152/Group I (Haley et al., 2016). Furthermore, all (31) isolates analyzed in that study from Pennsylvania were Group I (29/31 were ST152, 2/31 were ST318), so it seems that in Pennsylvania at least, the ST152/ Group I reservoir is in domestic food animals. Conversely, in that study, only 5% of nonhuman isolates were typed as ST198/Group II, showing that this ST is infrequently associated with poultry and cattle in the United States. Enterobase also shows low levels of ST198 associated with food animals (EnteroBase). While our sample size, 40, is small (we typed all ser. Kentucky isolates in the Pennsylvania Department of Health collection), the frequency of Group I to Group II infections is comparable (45 55%, respectively), suggesting that there may be an alternative reservoir for Group II/ ST198 that does not exist in domestically produced food. Multiple studies have linked ST198 infections with international travel to North Africa and the Middle East, where ST198 is prevalent in poultry (Le Hello et al., 2011, 2013; Mulvey et al., 2013). More recently, ser. Kentucky cases have been linked to travel to India, as well as domestically within Europe (Le Hello et al., 2013). In the United States, five ST198 infections were related to imported spices from North Africa between 2002 and 2009 (Le Hello et al., 2013). A comprehensive study by Le Hello et al. (2011) subtyping human clinical ser. Kentucky isolates from France, England and Wales, Denmark, and the United States show that 74% of isolates were ST198 and that over a quarter of these were resistant to ciprofloxacin. This study implicates international travel or import of contaminated food as causes for ST198 infection in North America, and while ST198/Group II isolates are present at low levels in food animals in the United States, it is not clear whether there is another, larger, domestic reservoir for Group II (ST198). Unlike serovars Enteritidis and Heidelberg, which are frequently isolated from both poultry and humans, ser. Kentucky is the leading serovar identified in chickens in the United States, and is also found in cattle, yet it is rarely found to cause illness in humans. This suggests some limitation to its pathogenicity in humans. Perhaps the ST152/Group I isolates do not cause severe symptoms in healthy individuals, thus, fewer cases are reported. That ST198/Group II Salmonella are found at lower levels in domestic food animals may be because its colonization is limited by the presence of ST152/Group I isolates. It is possible that although ST198 is infrequently found in nonhuman samples in the United States, that when human infections do occur, they are more severe than ST152, and thus greater number of patients report their illness. Serovar Kentucky ST198 is more commonly found to infect humans elsewhere in the world, suggesting that this strain is indeed more infectious, although this strain is also found internationally in poultry. Continual monitoring of our food supply animals for ST198/Group II isolates, combined with antimicrobial susceptibility testing should be prioritized. The phenotypic and genetic differences between Group I and II ser. Kentucky isolates presented in this study highlight their distinct lineages and suggest they inhabit two distinct ecological reservoirs. Future investigations may address the following outstanding questions: (1) whether Group I (ST152) isolates are better adapted to poultry and cattle than Group II (ST198) isolates (at least in the United States), (2) whether there are other domestic reservoirs of ser. Kentucky, especially since the incidence of Group II is much lower in poultry and cattle, (3) to determine the extent of selective pressure from antibiotics in the different reservoirs, and (4) to determine whether genetic differences in Group II ser. Kentucky are responsible for increased prevalence in causing illness. Succinctly, our findings show that the evolutionarily distinct Groups I and II of ser. Kentucky have both caused illness in humans in Pennsylvania and that the antibiotic resistance profiles are different between the two lineages. These findings may have implications to other polyphyletic Salmonella, including serovars Newport, Muenchen, Seftenberg, Bareilly, and Paratyphi B (Timme et al., 2013). Acknowledgments The authors thank Kelly Kline and members of the Shariat Laboratory for critically reviewing this article, and Drs. Shelly Rankin and Zakiya Whatley for helpful discussions. Funding was provided by Gettysburg College (startup funds to N.S.), the United States Department of Agriculture-National Institute of Food and Agriculture (grant to N.S.), and the Commonwealth of Pennsylvania and the NARMS through a cooperative agreement (CDC ELC-04040). The authors also thank Dr. Phillipe Horvath (Danisco) for providing them with the Macro to visualize CRISPR spacers. Disclosure Statement The authors state that no competing financial interests exist. References Achtman M, Wain J, Weill FX, Nair S, Zhou Z, Sangal V, Krauland MG, Hale JL, Harbottle H, Uesbeck A, Dougan G, Harrison LH, Brisse S, the S. enterica MLST study group. Multilocus sequence typing as a replacement for serotyping in Salmonella enterica. PLoS Pathog 2012;8:e Antunes P, Mourão J, Campos J, Peixe L. Salmonellosis: The role of poultry meat. Clin Microbiol Infect 2016;22: Centers for Disease Control and Prevention. National Enteric Disease Surveillance: Salmonella Annual Report, CDC,

8 8 VOSIK ET AL Available at: Accessed May 15, Centers for Disease Control and Prevention. National Enteric Disease Surveillace: Salmonella Annual Report 2014 (Published in 2017). CLSI. M100-S27: Performance Standards for Antimicrobial Susceptibility Testing; Twenty-Fourth Informational Supplement. Document M100-S27. Wayne, PA: CLSI, 2015a. CLSI. M07-A10: Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria That Grow Aerobically. 10th ed. Wayne, PA: CLSI, 2015b. European Food Safety Authority, European Centre for Disease Prevention and Control. The European Union summary report on trends and sources of zoonoses, zoonotic agents and foodborne outbreaks in EFSA J 2016;14:4634. Fabre L, Zhang J, Guigon G, Le Hello S, Guibert V, Accou- Demartin M, De Romans S, Lim C, Roux C, Passet V, Diancourt L, Guibourdenche M, Issenhuth-Jeanjean S, Achtman M, Brisse S, Sola C, Weill F-X. CRISPR typing and subtyping for improved laboratory surveillance of Salmonella Infections. PLoS One 2012;7:e Grissa I, Vergnaud G, Pourcel C. The CRISPRdb database and tools to display CRISPRs and to generate dictionaries of spacers and repeats. BMC Bioinformatics 2007;8:172. Haley BJ, Kim SW, Pettengill J, Luo Y, Karns JS, Van Kessel JS. Genomic and evolutionary analysis of two Salmonella enterica serovar Kentucky sequence types isolated from bovine and poultry sources in North America. PLoS One 2016; 11:e Le Hello S, Bekhit A, Granier SA, BaruaH, Beutlich J, Zając M, Münch S, Sintchenko V, Bouchrif B, Fashae K, Pinsard JL, Sontag L, Fabre L, Garnier M, Guibert V, Howard P, Hendriksen RS, Christensen JP, Biswas PK, Cloeckaert A, Rabsch W, Wasyl D, Doublet B, Weill FX. The global establishment of a highly-fluoroquinolone resistant Salmonella enterica serotype Kentucky ST198 strain. Front Microbiol 2013;4:395. Le Hello S, Hendriksen RS, Doublet B, Fisher I, Nielsen EM, Whichard JM, Bouchrif B, Fashae K, Granier SA, Jourdan- Da Silva N, Cloeckaert A, Threlfall EJ, Angulo FJ, Aarestrup FM, Wain J, Weill FX. International spread of an epidemic population of Salmonella enterica serotype Kentucky ST198 resistant to ciprofloxacin. J Infect Dis 2011;204: Li H, Li P, Xie J, Yi S, Yang C, Wang J, Sun J, Liu N, Wang X, Wu Z, Wang L, Hao R, Wang Y, Jia L, Li K, Qiu S, Song H. New Clustered regularly interspaced short palindromic repeat locus spacer pair typing method based on the newly incorporated spacer for Salmonella enterica. J Clin Microbiol 2014;52: Liu F, Barrangou R, Gerner-Smidt P, Ribot EM, Knabel SJ, Dudley EG. Novel virulence gene and clustered regularly interspaced short palindromic repeat (CRISPR) multilocus sequence typing scheme for subtyping of the major serovars of Salmonella enterica subsp. enterica. Appl Environ Microbiol 2011;77: Mulvey MR, Boyd DA, Finley R, Fakharuddin K, Langner S, Allen V, Ang L, Bekal S, El Bailey S, Haldane D, Hoang L, Horsman G, Louis M, Robberts L, Wylie J. Ciprofloxacinresistant Salmonella enterica serovar Kentucky in Canada. Emerg Infect Dis 2013;19: NARMS Report Available at: AnimalVeterinary/SafetyHealth/AntimicrobialResistance/ NationalAntimicrobialResistanceMonitoringSystem/UCM pdf Accessed September 2, Ribot EM, Fair MA, Gautom R, Cameron DN, Hunter SB, Swaminathan B, Barrett TJ. Standardization of pulsed-field gel electrophoresis protocols for the subtyping of Escherichia coli O157:H7, Salmonella, and Shigella for PulseNet. Foodborne Pathog Dis 2006;3: Sandt CH, Krouse DA, Cook CR, Hackman AL, Chmielecki WA, Warren NG. The key role of pulsed-field gel electrophoresis in investigation of a large multiserotype and multistate food-borne outbreak of Salmonella infections centered in Pennsylvania. J Clin Microbiol 2006;44: Sangal V, Harbottle H, Mazzoni CJ, Helmuth R, Guerra B, Didelot X, Paglietti B, Rabsch W, Brisse S, Weill FX, Roumagnac P, Achtman M. Evolution and population structure of Salmonella enterica serovar Newport. J Bacteriol 2010;192: Scallan E, Hoekstra RM, Angulo FJ, Tauxe RV, Widdowson MA, Roy SL, Jones JL, Griffin PM. Foodborne illness acquired in the United States Major pathogens. Emerg Infect Dis 2011;17:7 15. Shariat N, DiMarzio MJ, Yin S, Dettinger L, Sandt CH, Lute JR, Barrangou R, Dudley EG. The combination of CRISPR- MVLST and PFGE provides increased discriminatory power for differentiating human clinical isolates of Salmonella enterica subsp. enterica serovar Enteritidis. Food Microbiol 2013a;34: Shariat N, Dudley EG. CRISPRs: Molecular signatures used for pathogen subtyping. Appl Environ Microbiol 2014;80: Shariat N, Kirchner M, Sandt CH, Trees E, Barrangou R, Dudley EG. Subtyping of Salmonella enterica serovar Newport outbreak isolates by CRISPR-MVLST and determination of the relationship between CRISPR-MVLST and PFGE results. J Clin Microbiol 2013b;51: Shariat N, Sandt CH, DiMarzio MJ, Barrangou R, Dudley EG. CRISPR-MVLST subtyping of Salmonella enterica subsp. enterica serovars Typhimurium and Heidelberg and application in identifying outbreak isolates. BMC Microbiol 2013c;13:254. Shariat N, Timme RE, Pettengill JB, Barrangou R, Dudley EG. Characterization and evolution of Salmonella CRISPR-Cas systems. Microbiology 2015;161(Pt 2): Sukhnanand S, Alcaine S, Warnick LD, Su WL, Hof J, Craver MP, McDonough P, Boor KJ, Wiedmann M. DNA sequencebased subtyping and evolutionary analysis of selected Salmonella enterica serotypes. J Clin Microbiol 2005;43: Timme RE, Pettengill JB, Allard MW, Strain E, Barrangou R, Wehnes C, Van Kessel JS, Karns JS, Musser SM, Brown EW. Phylogenetic diversity of the enteric pathogen Salmonella enterica subsp. enterica inferred from genome-wide referencefree SNP characters. Genome Biol Evol 2013;5: Touchon M, Rocha EPC. The small, slow and specialized CRISPR and anti-crispr of Escherichia and Salmonella. PLoS One 2010;5:e Address correspondence to: Nikki W. Shariat, PhD Department of Biology Gettysburg College 300 N. Washington Street Gettysburg, PA nshariat@gettysburg.edu