Abstract
Salmonella enterica is one of the most common causes of foodborne illness in the United States. Although salmonellosis is usually self-limiting, severe infections typically require antimicrobial treatment, and ceftriaxone, an extended-spectrum cephalosporin (ESC), is commonly used in both adults and children. Surveillance conducted by the National Antimicrobial Resistance Monitoring System (NARMS) has shown a recent increase in ESC resistance among Salmonella Heidelberg isolated from food animals at slaughter, retail meat, and humans. ESC resistance among Salmonella in the United States is usually mediated by a plasmid-encoded bla CMY β-lactamase. In 2009, we identified 47 ESC-resistant bla CMY-positive Heidelberg isolates from humans (n=18), food animals at slaughter (n=16), and retail meats (n=13) associated with a spike in the prevalence of this serovar. Almost 90% (26/29) of the animal and meat isolates were isolated from chicken carcasses or retail chicken meat. We screened NARMS isolates for the presence of bla CMY, determined whether the gene was plasmid-encoded, examined pulsed-field gel electrophoresis patterns to assess the genetic diversities of the isolates, and categorized the bla CMY plasmids by plasmid incompatibility groups and plasmid multi-locus sequence typing (pMLST). All 47 bla CMY genes were found to be plasmid encoded. Incompatibility/replicon typing demonstrated that 41 were IncI1 plasmids, 40 of which only conferred bla CMY-associated resistance. Six were IncA/C plasmids that carried additional resistance genes. pMLST of the IncI1-bla CMY plasmids showed that 27 (65.8%) were sequence type (ST) 12, the most common ST among bla CMY-IncI1 plasmids from Heidelberg isolated from humans. Ten plasmids had a new ST profile, ST66, a type very similar to ST12. This work showed that the 2009 increase in ESC resistance among Salmonella Heidelberg was caused mainly by the dissemination of bla CMY on IncI1 and IncA/C plasmids in a variety of genetic backgrounds, and is likely not the result of clonal expansion.
Introduction
S
The most common non-typhoidal Salmonella serotypes causing human disease in the United States are Enteritidis, Typhimurium, Newport, Javiana, and Heidelberg. However, serotypes Typhimurium, Enteritidis, and Heidelberg tend to be more invasive and are the most common serotypes isolated from blood (Crump et al., 2011). Invasive Salmonella are more likely to require antimicrobial treatment, and bloodstream isolates are more likely to be antimicrobial resistant to one or more drugs, further complicating treatment (Crump et al., 2011). Salmonella serotype Heidelberg is one of the most common serotypes isolated from human cases of salmonellosis, Heidelberg is more common among bloodstream infections, and it is more likely to be resistant to antimicrobials. In 2009, Salmonella Heidelberg increased significantly to become the third most common serotype among retail meat and food animal isolates, and fifth most common among isolates from humans (CDC, 2009; USDA, 2009; FDA, 2009a).
ESC resistance among Salmonella in the United States is associated with the production of an AmpC-like (CMY) β-lactamase, conferred mostly by bla CMY genes (Philippon et al., 2002). CMY β-lactamases confer resistance to ESCs and β-lactam/β-lactamase inhibitor combinations (Dunne et al., 2000). The bla CMY gene is usually encoded on various plasmids, which can be distinguished by their incompatibility/replication features (Carattoli et al., 2002, 2005; Frye et al., 2008; Zhao et al., 2008; Folster et al., 2010; Sjolund-Karlsson et al., 2010). Plasmids with the same replication controls are incompatible and can therefore be grouped into several replicon (Inc) groups. Replicon type IncA/C and IncI1 are the predominant plasmid types encoding CMY β-lactamases (Hopkins et al., 2006; Welch et al., 2007; Baudry et al., 2009; Folster et al., 2010, 2011). IncI1 plasmids are commonly found among poultry-associated Salmonella serotypes, including Heidelberg (Folster et al., 2010, 2011).
In the United States, antimicrobial resistance among non-typhoidal Salmonella is monitored by the National Antimicrobial Resistance Monitoring Systems (NARMS), a collaboration among the Food and Drug Administration Center for Veterinary Medicine (FDA-CVM), United States Department of Agriculture (USDA), and Centers for Disease Control and Prevention (CDC), which performs surveillance on Salmonella isolates from retail meats, food animals, and humans, respectively. Animal isolates originate from federally inspected slaughter and processing plants throughout the United States; retail meat isolates are collected from 11 states, including 10 Foodborne Diseases Active Surveillance Network (FoodNet) sites and one state public health laboratory; and human isolates are collected from 54 NARMS-participating public health laboratories from all 50 states (FDA, 2009b). The purpose of this surveillance is to monitor trends in antimicrobial resistance among different sources and geographic locations over time.
Compared with 2008, 2009 showed a substantial increase in resistance to ceftriaxone (minimum inhibitory concentrations [MIC] ≥4 μg/mL) among Heidelberg isolated from retail meats (from 9.3% to 27.3%) food animals (9.4% to 17.3%), and humans (8% to 20.9%) (CDC, 2009; USDA, 2009; FDA, 2009b). The increase in ESC resistance among isolates from humans appeared mainly in western states (CDC, 2009). To determine whether this increase was driven by the emergence of a new variant of ESC-resistant Heidelberg, we screened NARMS isolates for the presence of bla CMY, determined whether the gene was plasmid-encoded, examined pulsed-field gel electrophoresis (PFGE) patterns to assess the genetic diversities of the isolates, and categorized the bla CMY plasmids by plasmid incompatibility groups and plasmid multi-locus sequence typing (pMLST).
Methods
Isolate collection and testing
Salmonella isolates from ill persons were obtained from specimens submitted to clinical laboratories in the United States and subsequently forwarded to state public health laboratories. Participating state public health laboratories serotyped and submitted every 20th non-typhoidal Salmonella (NTS) to the CDC NARMS laboratory for susceptibility testing. NARMS retail meat monitoring was conducted by the United States FDA-CVM in collaboration with FoodNet (
PCR amplification of bla CMY
For each isolate, DNA template for PCR was prepared by lysing the bacteria at 95°C and collecting the supernatant following centrifugation for 10 min at 20,000×g (Sorvall RC5B Plus, SS-34 rotor; Thermo Fischer Scientific Inc., Waltham, MA). Polymerase chain reactions (PCRs) contained 2× Hot-Start PCR Master Mix (Qiagen Inc., Valencia, CA), 0.4 μM of each primer, 5 μL of template DNA and sterile PCR water to a final volume of 50 μL. Thermal cycling was performed using the following conditions: 15 min at 95°C, followed by 30 cycles of 95°C for 30 s, 56°C for 30 s, and 72°C for 90 s. To determine the presence of bla CMY genes, primers ampC1 (5′-ATGATGAAAAAATCGTTATGC-3′) and ampC2 (5′-TTG CAGCTTTTCAAGAATGCGC-3′) were used (Winokur et al., 2001).
Plasmid purification and characterization
Purified plasmid DNA was used to transform laboratory E. coli, to separate the bla
CMY plasmids from other plasmid types prior to replicon typing, and for pMLST. Plasmids were purified using the QiaFilter Midi kit (Qiagen Inc.), following a modified manufacturer's protocol (Folster et al., 2010). Electroporation of each plasmid into E. coli DH10B Electromax competent cells (Invitrogen, Carlsbad, CA) was performed as previously described (Folster et al., 2010). Cells were plated on LB agar plates containing 100 mg/L of ampicillin or 4 mg/L ceftriaxone (Sigma-Aldrich, St. Louis, MO). Plasmids were re-purified from a single bla
CMY PCR-positive transformant to isolate a single plasmid from each isolate. Purification was performed as described above with the additional modification of growing the cells overnight in 25 mL of LB broth with 100 μg/mL of ampicillin or 4 μg/mL ceftriaxone. Plasmid PCR-based replicon typing (PBRT) was performed as previously described (Carattoli et al., 2005). pMLST was performed on IncI1 plasmids as previously described (Garcia-Fernandez et al., 2008). Sequencing was performed using Big Dye version 3.1 (Applied Biosystems, Foster City, CA), and sequence reactions were cleaned with Centri-sep plates (Princeton Separations, Adelphia, NJ). The reactions were electrophoresed through POP-7 polymer (Applied Biosystems) on a 3730 DNA Analyzer (Applied Biosystems) equipped with a 48-capillary, 50-cm array. Sequence analysis was performed using Lasergene 8 software (DNASTAR Inc., Madison, WI). Sequences were submitted to the pMLST web page (
PFGE
Two-enzyme (XbaI and BlnI) PFGE was performed according to the CDC PulseNet protocol, and all PFGE profiles generated were submitted to the PulseNet national database administered by CDC (NARMS-FDA and NARMS-CDC) or USDA VetNet (NARMS-USDA) (Ribot et al., 2006; Jackson et al., 2007). Gel images were captured using the GelDoc XR system (Bio-Rad Laboratories) and Quantity 1-D analysis software (Bio-Rad Laboratories). Pattern analysis and unweighted pair group method with arithmetic mean (UPGMA) dendrogram generation were performed using BioNumerics software (Applied Maths, Saint-Martens-Latem, Belgium) with the Dice coefficient and tolerance of 1.5%.
Results
Identification of bla CMY-positive Salmonella ser. Heidelberg isolates
NARMS received and performed antimicrobial susceptibility testing on 223 isolates of Salmonella Heidelberg from food animals, retail meat, and humans in 2009. Of these, 47 isolates (21.1%) displayed resistance to ceftriaxone, ceftiofur, and amoxicillin-clavulanic acid, suggesting the presence of a bla CMY allele. Isolates from humans (n=18) made up the largest proportion of these resistant isolates (38.3%). Two-thirds of these isolates (n=12) were obtained from patients in western states (California, Oregon, and Washington) over a 5-month period. Thirteen of the human isolates were from male patients, and five were from females; the median age was 26 years. Among animal isolates, 96.1% (74/77) of the Heidelberg isolates and 92.3% (12/13) of the resistant isolates were obtained from chickens. Among retail meat isolates, 80.4% (45/56) of Heidelberg isolates were obtained from chicken breast samples, while 17.9% came from ground turkey. Most of the cephalosporin-resistant isolates (14/16; 87.5%) were obtained from chicken breast samples. PCR analysis confirmed that all 47 ESC-resistant isolates were positive for bla CMY.
Characterization of the bla CMY plasmids
Plasmids were purified from the transformants and typed by PBRT. Forty-one of 47 bla CMY plasmids were replicon type IncI1, and the remaining six plasmids were replicon type IncA/C (Table 1). The six IncA/C plasmids were identified among two human isolates, one pork chop isolate, and three chicken isolates. Antimicrobial susceptibility testing (AST) of the bla CMY plasmid transformants, along with a comparison to the resistance phenotypes of the original isolates, identified resistance phenotypes conferred by the plasmids (Table 1). Forty of the 41 IncI1-bla CMY plasmids only conferred resistance to drugs associated with presence of a bla CMY resistance determinant (ampicillin, amoxicillin/clavulanic acid, cefoxitin, ceftriaxone, and ceftiofur). The IncI1-bla CMY plasmid identified in isolate B095270 also conferred resistance to kanamycin. In contrast, all of the IncA/C-bla CMY plasmids conferred multi-drug resistance (MDR), defined as resistance to at least one antimicrobial in three or more antimicrobial classes (CDC, 2009). The most common additional resistance conferred by the IncA/C plasmids was resistance to sulfisoxazole and tetracycline (Table 1). Resistance to chloramphenicol, trimethoprim-sulfamethoxazole, gentamicin, and kanamycin was less common.
State of isolation is given for human and retail meat sampling, whereas region where isolate was obtained is given for sampling of food animals at slaughter. Region 1 includes ME, VT, NH, NY, MA, CT, RI, PA, MD, DE, NJ, OH, IN, MI, DC; region 2 includes VA, KY, TN, NC, SC, GA, AL, WV, FL, Puerto Rico; region 3 includes ND, SD, NE, KS, MN, IA, MO, WI, IL; region 4 includes OK, AR, LA, TX, MS; region 5 includes WA, MT, OR, ID, WY, CO, UT, NM, AZ, NV, CA, AK; region 6 includes Hawaii, Guam, U.S. Virgin Islands, Mariana Islands, and American Samoa.
All isolates were resistant to ampicillin, amoxicillin/clavulanic acid, cefoxitin, and ceftiofur.
All transformants were resistant to ampicillin, amoxicillin/clavulanic acid, cefoxitin, and ceftiofur. Additionally, all transformants were resistant to streptomycin due to the natural resistance of DH10B cells.
CHL, chloramphenicol; FIS, sulfisoxazole; GEN, gentamicin; KAN, kanamycin; SXT, trimethoprim-sulfamethoxazole; TET, tetracycline; I, intermediate.
Additional drugs tested: AMI, amikacin; CIP, ciprofloxacin; NAL, nalidixic acid.
Inc type, incompatibility/replicon type; ST, sequence type; —, none; NA, not applicable; ND, not determined.
IncI1 plasmids were compared using pMLST (Garcia-Fernandez et al., 2008). Currently, IncA/C plasmids are not included in the pMLST scheme. Of the 41 IncI1 plasmids, 27 (65.8%) were ST12 (Table 2). Ten plasmids had a new ST profile, ST66; however, since this profile differed from ST12 by only one allele (trbA), it is considered to be part of the ST12 clonal complex. We also identified a single ST2, ST23, ST65 (a new profile), and one plasmid that could not be typed due to a large insertion in the trbA allele.
ST, sequence type.
Determining similarity by PFGE of the bla CMY-positive isolates
Two-enzyme PFGE was used to evaluate the genetic relatedness of ESC-resistant strains from different sources (Fig. 1). Twenty-seven patterns were generated for the 47 isolates, indicating the dissemination of multiple distinct isolate types. Four groups containing three or more indistinguishable PulseNet PFGE patterns were identified; VetNet patterns matched those found in PulseNet. The largest group contained seven isolates (JF6X01.122/JF6A26.0058), including a single chicken isolate and isolates from chicken breasts and humans. While most of the animal/retail meat isolates were recovered from chicken or chicken breasts, one isolate each was recovered from pork chop, ground turkey, and a turkey that grouped separately from chickens and chicken breast isolates (Fig. 1).
Pulsed-field gel electrophoresis (PFGE) patterns of bla CMY-positive Salmonella enterica ser. Heidelberg isolated from food animals, retail meat, and humans from the United States in 2009. Dendrogram of percent genetic similarity by PFGE was generated using BioNumerics based on XbaI and BlnI restriction digestion. Pattern analysis and unweighted pair group method with arithmetic mean (UPGMA) dendrogram generation were performed using BioNumerics software (Applied Maths, Saint-Martens-Latem, Belgium) with the Dice coefficient and tolerance of 1.5%. Percent similarity is located above the dendrogram. Antibiogram displays the antimicrobial resistance profile of the isolates; a black box indicates resistance to that antimicrobial. Isolate number, source, XbaI pattern name, BlnI pattern name, plasmid incompatibility type, and sequence type (where applicable) are listed to the right of the antibiogram. The BlnI patterns for isolates B095270 and B094351 have not been identified previously by PulseNet, so they do not have pattern names at this time.
Discussion
In the last decade, the rise in ESC-resistant Salmonella ser. Heidelberg in the United States has been documented, but the recent sharp increase in resistance among retail meat, food animal, and human isolates is especially concerning (FDA, 2009b; Folster et al., 2010). Taken together with studies documenting the invasiveness and antimicrobial resistance observed among isolates of serotype Heidelberg, it is imperative to understand the factors driving this phenomenon and determine what actions may mitigate its impact in the future (Crump et al., 2011).
In this study, we examined 47 CMY-β-lactamase producing Heidelberg isolated in 2009 from retail meat, food animals, and humans, and characterized their bla CMY plasmids. PFGE analysis identified five clusters with indistinguishable patterns for isolates recovered from both human clinical cases and chicken carcasses/chicken breasts. The largest group contained seven isolates with XbaI PulseNet pattern JF6X01.0122 (VetNet pattern JF6X01.0001 ARS), which is one of the more common Heidelberg patterns in PulseNet and VetNet. Most of the isolates could be distinguished by their XbaI pattern (27 different XbaI patterns among 47 isolates). This suggests that the rise in ESC-resistant Heidelberg in human and food animal sources is not due to a single clone expansion, but it is likely due to multiple independent events of ESC resistance acquisition in a serovar associated with poultry, where cephalosporins are used (Silvers, 2002). This observation is consistent with what has been observed in retail meat isolates of Heidelberg during 2004–2009 (FDA, 2009a).
All of the bla CMY genes were located on plasmids, which mediate nearly all ESC resistance in Salmonella in the United States. Most plasmids (41/47) were replicon type IncI1, a common bla CMY-encoding plasmid type along with IncF, IncHI1, and IncA/C (Hopkins et al., 2006; Baudry et al., 2009; Fricke et al., 2009; Folster et al., 2010). All of the IncI1 plasmids, except one, conferred only bla CMY-associated resistance. Plasmid pB095270 also conferred kanamycin resistance, which is the first IncI1-bla CMY plasmid we identified that conferred an additional resistance phenotype (Folster et al., 2010, 2011). IncI1-bla CMY plasmids are common among poultry-associated Salmonella serotypes and Escherichia coli from various agricultural and clinical sources (Baudry et al., 2009). IncI1 plasmids are usually highly mobile and are characterized by the presence of a type IV pilus locus, which may be involved in conjugation and virulence, and have been shown to be more common among pathogenic rather than commensal E. coli (Kim and Komano, 1997; Johnson et al., 2007). The six remaining plasmids were IncA/C, a MDR-plasmid ubiquitous in agricultural settings (Lindsey et al., 2009; Mulvey et al., 2009).
Subtyping using pMLST revealed that most IncI1 plasmids in this study were sequence type 12, consistent with earlier observations (Folster et al., 2010, 2011). The pMLST database (Jolley and Maiden, 2010) also documents additional ST12 IncI1 plasmids, including a bla CMY-2-positive Salmonella Kentucky isolate from poultry (Fricke et al., 2009) and bla CMY-2 plasmids from Salmonella and E. coli isolated from human, animal, and environmental sources in Canada (Fricke et al., 2009; Mataseje et al., 2010).
Interestingly, 10 of the plasmids we characterized in this study were ST66, a novel sequence type differing from ST12 by a single allelic change (trbA3 to trbA11; Table 2). Since five out of six alleles match, ST66 is thought to be related to ST12 and has been placed in the same clonal complex (CC) as ST12, CC12. When we examined the source and state/region of the isolates with the ST66 plasmid, all 10 isolates were from chicken breasts or humans, and nine out of 10 isolates were obtained in western states (California, Colorado, and Washington). These were collected over a 5-month period, suggesting that they were not due to a single outbreak. None of the animal isolates obtained at slaughter, including those from region 5 (western states), contained the ST66 IncI1 plasmid. Additional studies are needed to explain this phenomenon. However, due to our limited number of Heidelberg isolates from animals, it is also possible that we simply missed the ST66 plasmids among this source.
Among the IncI1-bla CMY plasmids, we also identified a ST2, ST23, ST65, and a nontypeable plasmid. Previously identified ST2 IncI1 plasmids include bla CMY-2-positive isolates of S. Heidelberg, S. Typhimurium, and E. coli found in humans, dogs, and environmental sources (Garcia-Fernandez et al., 2008; Mataseje et al., 2010). ST23 IncI1 plasmids have been identified in bla CMY-2-positive isolates of S. Heidelberg and E. coli from humans (Mataseje et al., 2010). ST65 is a new sequence type, and the unidentifiable plasmid had a large insertion into the trbA allele that prevented it from being typed, although the remaining four alleles matched ST12 and ST66, suggesting that it may be related to these sequence types. Further sequencing and categorization are necessary to fully understand the diversity of plasmids that carry ESC and other relevant antimicrobial resistance.
Conclusion
Overall, this work demonstrates that the 2009 increase in ESC resistance among Salmonella Heidelberg in the United States was due to the dissemination of bla CMY on IncI1 and IncA/C plasmids in a variety of genetic backgrounds, and likely not the result of clonal expansion. The IncI1 plasmids showed identical STs in strains from humans, chicken carcasses, and chicken breasts, further supporting chicken products as an important source of human infection with ESC-resistant Salmonella Heidelberg. Both resistant and susceptible strains of Salmonella Heidelberg continue to present a significant public health burden in the United States and elsewhere. Ongoing monitoring of human clinical cases, resistances associated with the food supply, and the characterization of plasmids from different sources will help attribute distinct resistances to different food animal sources, and will help facilitate a more global understanding of the genetics and ecology of antimicrobial resistance in Salmonella Heidelberg.
Footnotes
Acknowledgments
We thank the NARMS participating public health laboratories for submitting the isolates, Anne Whitney for DNA sequencing, Alessandra Carattoli for the plasmid incompatibility typing control strains, and Maria Karlsson for her critical review. This work was partially supported by an interagency agreement between CDC, USDA, and the FDA Center for Veterinary Medicine. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the CDC, FDA or USDA.
Disclosure Statement
No competing financial interests exist.