Abstract
Non-Typhi Salmonella cause over 1.7 million cases of gastroenteritis in North America each year, and food-animal products are commonly implicated in human infections. For invasive infections, antimicrobial therapy is indicated. In North America, the antimicrobial susceptibility of Salmonella is monitored by the U.S. National Antimicrobial Resistance Monitoring System (NARMS) and The Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS). In this study, we determined the susceptibility to cephalosporins by broth microdilution among 5,041 non-Typhi Salmonella enterica isolated from food animals, retail meats, and humans. In the United States, 109 (4.6%) of isolates collected from humans, 77 (15.7%) from retail meat, and 140 (10.6%) from food animals displayed decreased susceptibility to cephalosporins (DSC). Among the Canadian retail meat and food animal isolates, 52 (13.0%) and 42 (9.4%) displayed DSC. All isolates displaying DSC were screened for β-lactamase genes (blaTEM, blaSHV, blaCMY, blaCTX-M, and blaOXA-1) by polymerase chain reaction. At least one β-lactamase gene was detected in 74/109 (67.9%) isolates collected from humans, and the blaCMY genes were most prevalent (69/109; 63.3%). Similarly, the blaCMY genes predominated among the β-lactamase-producing isolates collected from retail meats and food animals. Three isolates from humans harbored a blaCTX-M-15 gene. No animal or retail meat isolates harbored a blaCTX-M or blaOXA-1 gene. A blaTEM gene was found in 5 human, 9 retail meat, and 17 animal isolates. Although serotype distributions varied among human, retail meat, and animal sources, overlap in blaCMY-positive serotypes across sample sources supports meat and food-animal sources as reservoirs for human infection.
Introduction
The predominant mechanism of antimicrobial resistance to extended-spectrum cephalosporins is the production of β-lactamases, enzymes that inactivate the antimicrobial agent. Three main groups of β-lactamases have been identified among cephalosporin-resistant Salmonella: the Ambler class C cephamycinases (CMY), the carbapenemases, and the Ambler class A extended-spectrum β-lactamases (ESBLs).33,34 The latter comprises many different families of β-lactamases, including TEM, SHV, OXA, and CTX-M enzymes. 33
Reports on β-lactamase-producing isolates of Salmonella in North America started to appear in the mid 90s. In the United States, a national survey of 4,003 Salmonella isolates collected in 1995 found three ceftriaxone-resistant isolates, each of which had been acquired outside the United States. 24 The first domestically reported ceftriaxone-resistant isolate was a Salmonella serotype Typhimurium isolated from a 12-year-old boy in 1998. 16 This isolate produced a cephamycinase of the CMY-2 type. 16
The first ESBL-producing Salmonella isolate detected in Canada was a serotype Typhimurium identified in 2000. 32 This isolate harbored a blaSHV-2a and blaTEM-1 gene and carried two plasmids of 70 and 120 kb in size, respectively. 32 Two years later, the first CMY-2-producing Salmonella was reported during a small outbreak of Salmonella serotype Newport in Alberta, Canada. 35 Since then, multiple studies have reported on the occurrence of CMY and ESBL enzymes among Salmonella isolates collected from human, animal, and retail meat sources in Canada and the United States.1,13,17,18,30,41–43,50–52
In North America, the antimicrobial susceptibility of Salmonella isolates is being monitored by two main surveillance programs. In the United States, the National Antimicrobial Resistance Monitoring System (NARMS) performs surveillance of Salmonella collected from humans, food animals, and retail meat. 19 This surveillance program was initiated in 1996 and is a collaboration among three federal agencies: the U.S. Food and Drug Administration (FDA), the U.S. Department of Agriculture (USDA), and the Centers for Disease Control and Prevention (CDC). 19 In Canada, the Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS) monitors the antimicrobial susceptibility of select bacteria, including Salmonella collected from humans, food animals, and retail meat. 38 This surveillance program was initiated in 2002, and is led by the Public Health Agency of Canada (PHAC). 38 The PHAC administers this program with the assistance of the Veterinary Drugs Directorate (Health Canada), Canadian Food Inspection Agency, provincial health and agriculture ministries, academic institutions, and private industry.
The present study was undertaken to investigate the presence of decreased susceptibility to extended-spectrum cephalosporins among non-Typhi Salmonella enterica (NTS) isolated from humans in the United States and to explore the molecular mechanisms underlying resistance. We further aimed to investigate isolates obtained from food animals and retail meats in the United States and Canada to determine the role these sources play as reservoirs of cephalosporin-resistant Salmonella isolates.
Methods
In 2008, clinical diagnostic laboratories submitted Salmonella isolates to state and territorial public health laboratories, where they were confirmed and serotyped according to the Kauffmann–White scheme. The 54 NARMS-participating public health laboratories from all 50 states forwarded every 20th human isolate of NTS to CDC. Similarly, NTS isolated from raw retail meat (chicken breasts, ground turkey, ground beef, and pork chops) were submitted by 11 sites (10 FoodNet-participating states and Pennsylvania) for analysis at the FDA Center for Veterinary Medicine (FDA-CVM). In Canada, isolates were obtained through the CIPARS. Retail meat isolates (chicken [legs and wings, skin-on] and pork chops) were obtained from five regions across Canada (British Columbia, Saskatchewan, Ontario, Québec, and the Maritimes) and forwarded to the Department of Pathobiology at the Ontario Veterinary College in Guelph, Ontario, Canada, testing for further analysis. NTS from food animals in the United States were obtained from carcass rinsates (chicken), carcass swabs (turkey, cattle, and swine), and ground products (chicken, turkey, and beef). These animal samples were collected by USDA's Food Safety Inspection Service (FSIS) from federally inspected slaughter and processing plants, and NTS recovered from the samples were forwarded to the USDA Agricultural Research Service (ARS) in Athens, GA, for further analysis. In Canada, Salmonella from food animals were isolated from cecal samples collected from federally inspected abattoirs through the CIPARS Abattoir component (chicken and swine) and pooled fecal samples collected from a sentinel farm network through the CIPARS Farm component (swine). After susceptibility testing, isolates were forwarded to the Department of Pathobiology at the Ontario Veterinary College in Guelph, Ontario, Canada, for further analysis. Information on the specific methods used to isolate Salmonella from various animal and retail meat sources in the United States and Canada has been published elsewhere.15,37,45 For each positive sample, one representative colony of Salmonella was picked for further analysis.
Minimum inhibitory concentrations (MIC) to 15 antimicrobial agents were determined by broth microdilution (Sensititre®; Trek Diagnostics). Testing was performed according to the manufacturer's instructions adhering to the guidelines of the Clinical and Laboratory Standards Institute (CLSI). The following quality control strains were used: Escherichia coli ATCC 25922, Staphylococcus aureus ATCC 29213, E. coli ATCC 35218, Pseudomonas aeruginosa ATCC 27853, and Klebsiella pneumoniae ATCC 700603. When available, CLSI recommendations were used to define resistance. 11 For drugs lacking such criteria, consensus breakpoints used by NARMS were applied. 44 Human, animal, and retail meat isolates of NTS showing decreased susceptibility (MIC≥2 mg/L) to ceftriaxone (DSC) or ceftiofur (an extended-spectrum cephalosporin used in veterinary medicine) were included in the present study and tested for susceptibility to additional β-lactam drugs, including cefepime and imipenem.
To investigate molecular mechanisms mediating β-lactam resistance, genomic DNA was prepared for each isolate by using the UltraClean® Microbial DNA Isolation Kit (MO BIO Laboratories, Inc.) or by lysing the bacteria at 95°C and collecting the supernatant after centrifugation. Polymerase chain reaction assays using previously described primers and conditions were used to screen isolates for the following genes: blaTEM, blaSHV, blaCMY, blaCTX-M, and blaOXA-1.4,10,27 Amplicons were visualized on 1.5% agarose gels stained with ethidium bromide or GelRed nucleic acid gel stain (Biotium, Inc.). β-lactamase genes, other blaCMY, and a subset of the blaCMY genes, were confirmed by nucleic acid sequencing. Sequencing was performed by using a 3730 DNA Analyzer (Applied Biosystems) or by submitting DNA samples to Retrogen (Retrogen, Inc.) or the Laboratory Services of the University of Guelph, Canada, for sequence analysis.
Results
Among 2,380 NTS collected from humans in the United States in 2008, 109 (4.6%) displayed DSC. Among 491 isolates from retail meats and 1,326 isolates from animals in the United States, 77 (15.7%) and 140 (10.6%) displayed DSC, respectively. Among 399 isolates collected from retail meat in Canada, 52 (13.0%) showed DSC, and among 446 isolates collected from Canadian abattoirs and farms, 42 (9.4%) isolates displayed DSC (Table 1).
Decreased susceptibility to cephalosporins (ceftriaxone or ceftiofur MIC≥2 mg/L).
Each isolate contained one or more β-lactamase genes. A subset (n=25) of the blaCMY genes were sequenced and confirmed as blaCMY-2. All blaTEM genes were blaTEM-1; the two blaSHV genes from humans were blaSHV-12, whereas the blaSHV-positive isolate from retail meat contained a blaSHV-2a. All blaCTX-M genes were blaCTX-M-15.
Number of isolates with negative results in β-lactamase gene screen.
MIC, minimum inhibitory concentration.
Among the 77 U.S retail meat isolates displaying DSC, 48 (62.3%) were obtained from chicken breast, 24 (31.2%) from ground turkey, and 5 (6.5%) from ground beef. Among the 140 U.S food-animal isolates with DSC, 73 (52.1%) were obtained from cattle, 54 (38.6%) from chicken, 8 (5.7%) from turkey, and 5 (3.6%) from swine. The 52 retail meat isolates with DSC collected in Canada were all obtained from chicken, whereas the 42 food animal isolates with DSC originated from ceca of chickens (71.4%) and swine (28.6%).
A β-lactamase gene was detected in 74/109 (67.9%) NTS collected from humans in the United States, and the blaCMY genes were most prevalent (69/109; 63.3%) (Table 1). The predominant serotype detected among the blaCMY-positive isolates collected from humans was Newport (Table 2). Three isolates from humans harbored a blaCTX-M-15 gene. These were of serotype Concord and were obtained from three patients: two female and one male patient aged 5–8 months. All three patients were Ethiopian adoptees. Two of the blaCTX-M-15-positive isolates also harbored a blaSHV-12. Among the U.S. retail meat and animal NTS displaying DSC, 63 (82.9%) and 138 (98.6%), respectively, harbored one or more β-lactamase genes; 57 retail meat and 138 animal isolates were positive for blaCMY. Serotype Typhimurium var O:5– was the most common serotype detected among the blaCMY-positive isolates from U.S. retail meats, whereas Dublin and Kentucky were most common among U.S. food animals (Table 2). None of the animal or retail meat isolates harbored a blaCTX-M, blaSHV, or blaOXA-1 gene. Among all isolates collected in the United States, a blaTEM-1 gene was detected in 5 human, 8 retail, and 15 animal isolates.
Among the retail meat and animal NTS collected in Canada, 48 (92.3%) and 28 (66.7%), respectively, harbored one or more β-lactamase gene. All of these isolates were positive for blaCMY. In addition, two animal isolates and one retail meat isolate harbored a blaTEM-1 gene. A single retail meat isolate (chicken) harbored a blaSHV-2a gene. No blaCTX-M orblaOXA-1 genes were detected among the animal and retail meat isolates collected in Canada. The most prevalent serotypes among the blaCMY-positive retail meat and food animal isolates from the United States and Canada are presented in Table 2. A subset of all blaCMY-positive isolates detected in this study (n=25) was sequenced and confirmed as blaCMY-2.
For each isolate displaying DSC, additional susceptibility testing was performed using other β-lactam drugs. Three isolates collected from humans displayed resistance (MIC>32 mg/L) to cefepime, a fourth-generation cephalosporin. The highest MIC for the carbapenem imipenem was 1 mg/L and was observed in three isolates collected from humans in the United States, seven isolates from animals in the United States, and one retail meat isolate from Canada.
Discussion
With increasing resistance to traditional agents such as trimethoprim-sulfamethoxazole, chloramphenicol, and ampicillin, extended-spectrum cephalosporins (e.g., ceftriaxone) have become an important treatment option for severe Salmonella infections.12,21 In 2005–2006, 4.1% of all NARMS Salmonella isolates collected from humans in the United States showed decreased susceptibility to the extended-spectrum cephalosporins ceftriaxone or ceftiofur. 42 In the current article, we confirm the continued presence of cephalosporin resistance among clinical isolates of Salmonella collected from humans in 2008 with 4.6% of the isolates exhibiting DSC.
Among all Salmonella isolates collected from humans in the United States, DSC was mainly due to the presence of a blaCMY gene. This is in accordance with previous reports confirming blaCMY genes and blaCMY-2, in particular, to be common among human isolates of Salmonella collected in the United States.13,17,42 The most common serotype among the blaCMY-positive isolates collected from humans in the United States was Newport. This serotype has been associated with multiple outbreaks in this country and usually harbors other antimicrobial resistance genes in addition to blaCMY, making it multidrug resistant.8,9,20
ESBLs are not widely disseminated among human isolates of Salmonella in the United States. Two different types were detected in the present study: SHV and CTX-M. The latter exhibits increased activity against cefotaxime and confers cross-resistance to the fourth-generation cephalosporin cefepime. The CTX-M enzymes have been considered rare in the United States, but recent studies have documented the emergence of CTX-M-producing Enterobacteriaceae.6,28,31,41 In this study, three human isolates of the serotype Concord carried the blaCTX-M-15 genes. These isolates were obtained from two female and one male patient aged 5–8 months. All three patients were adoptees from Ethiopia. For two patients, symptoms, including diarrhea, vomiting, and fever, were reported. The third patient (8-month-old male) was asymptomatic, but tested positive for Salmonella in a routine screening test administered in the United States. Two CTX-M-15-producing isolates of the serotype Concord have been identified in NARMS previously.41,42 Notably, all of the cases have been adoptees from Ethiopia or patients reporting a travel history to Ethiopia in conjunction with illness onset. The emergence of CTX-M-15-producing Concord isolates among children and adults in Ethiopia has been documented in several studies performed in Europe and the United States.3,22,23,47
Nonhuman sources such as food animals and retail meat have been implicated in the transmission of drug-resistant NTS. 5 In the United States, CMY-producing isolates of Salmonella have been recovered from various animal and meat sources on multiple occasions.2,18,39,48,52 In this study, we investigated the presence of β-lactamase-producing Salmonella among food animals and retail meat sources in the United States and Canada. The prevalence of DSC among animal and retail meat samples collected in the United States and Canada was similar. However, the predominant serotype among the blaCMY-positive isolates from animals and retail meats differed between the two countries. This difference can be partly explained by the sample sources that were investigated; the majority of DSC isolates from food animals in the United States were obtained from cattle, whereas chicken was the main sample source of DSC isolates in Canada. However, for the retail meat isolates, different serotypes were observed despite the main sample source being chicken in both countries. The most common serotype among blaCMY-positive isolates from U.S. retail meat was Typhimurium var. O:5–, whereas Kentucky was most common among blaCMY-positive retail meat isolates from Canada. Thus, the differences in ranking of blaCMY-positive serotypes are partly attributable to differences in the serotype distributions themselves.
The two most common serotypes detected among blaCMY-positive isolates from humans in the United States were Newport and Typhimurium. These serotypes were also among the top six serotypes of the U.S. animal and retail meat isolates harboring blaCMY, but were not detected among the retail meat and animal isolates from Canada. The fact that the serotype Newport was not detected among Canadian isolates carrying blaCMY is likely due to the different sampling schemes. Cattle and ground beef was not sampled in Canada, since Salmonella rarely have been found among this species. 36 The serotype distributions for all NTS collected by each agency in 2008 are described in NARMS and CIPARS annual reports.7,15,37,46 These reports also entail information on each agency's sampling scheme for Salmonella.
Heidelberg was the third most common serotype detected among blaCMY-positive isolates from humans in the United States. This is of concern, as Heidelberg is more likely to cause invasive disease that requires hospitalization than, for example, serotypes Newport and Typhimurium. 26 Isolates of serotype Heidelberg were also detected among animal and retail meat samples from both Canada and the United States, suggesting that these sources may be reservoirs of cephalosporin resistance. Recent work by Folster et al. corroborated the role of retail meat as a reservoir for cephalosporin-resistant Heidelberg. 17 By performing pulsed-field gel electrophoresis (PFGE) on 54 blaCMY-positive cephalosporin-resistant isolates of Heidelberg collected from humans in the United States, a high degree of genetic relatedness among the isolates was demonstrated. A further comparison to 42 Heidelberg isolates obtained from retail meat in the United States showed that 25 human isolates had PFGE patterns identical to retail meat isolates. 17 The fact that identical PFGE patterns were detected among Heidelberg isolates of human and retail meat origin suggests that poultry is a reservoir for Heidelberg infections in humans. Similar connections between cephalosporin-resistant isolates of human and animal origin have been demonstrated previously in the United States,16,48,49 and in Canada, a strong correlation between ceftiofur-resistant Salmonella ser. Heidelberg in retail chicken and incidence of ceftiofur-resistant Heidelberg infections in humans has been reported. 14
To summarize, 4.6% of all NTS collected from humans in the United States and 11.6% of all NTS collected from animal and retail meat sources in Canada and the United States displayed decreased susceptibility to the extended-spectrum cephalosporins ceftriaxone or ceftiofur. DSC was commonly due to the presence of blaCMY genes. Detection of DSC among Salmonella from food animal and retail meat samples collected in Canada and the United States confirms these sources to be reservoirs of resistant Salmonella. To limit the selection and dissemination of extended-spectrum cephalosporin-resistant Salmonella, judicious use of antimicrobial agents in both human and veterinary medicine is necessary.
Footnotes
Acknowledgments
We thank the NARMS-participating public health laboratories, the Retail Foods Survey Working Group, and the FSIS laboratories for submitting the isolates. We also thank the CO, MO, and NY Divisions of Public Health for providing patient interviews. This work was supported by an interagency agreement between CDC and FDA-CVM. Regarding CIPARS abattoir isolates, we thank the abattoir industry personnel and the Canadian Food Inspection Agency regional directors, inspection managers, and onsite staff for their extensive voluntary participation. The CIPARS farm surveillance component thanks all participating veterinarians and producers, the CIPARS Farm Swine Surveillance Advisory Committee as well as Agriculture and Agri-Food Canada, the provincial ministries of agriculture in BC, SK, AB, and QC, and the USDA, CAHFSE Program. The CIPARS retail component thanks the field technicians as well as the University of Prince Edward Island. Lastly, CIPARS thanks all laboratory technicians and data management staff for their contributions to the overall program. The molecular characterization of Canadian isolates was financially supported by the Laboratory for Foodborne Zoonoses, Public Health Agency of Canada, Guelph, Ontario, Canada.
On behalf of both CIPARS and NARMS, the authors would like to dedicate this article to the memory of Dr. Lucie Dutil (1965–2011). Lucie led the CIPARS analysis group and was a critical member of CIPARS across all program components. She was a widely respected and remarkable colleague and an even better friend. She will be missed greatly.
Author Disclosure Statement
The authors have no conflicts of interest to declare.
