Salmonella enterica subsp. enterica producing VIM-1 carbapenemase isolated from livestock farms.

Sir, Thirdand fourth-generation cephalosporins and carbapenems are ‘critically important’ antimicrobials as classified by the WHO (www.who.int). In fact, carbapenems are last-line clinical antibiotics against infections caused by multidrug-resistant Gram-negative bacteria. In contrast to cephalosporins, carbapenems are not hydrolysed by most b-lactamases, including AmpC b-lactamases and extended-spectrum b-lactamases (ESBLs). However, during the last decade the prevalence of carbapenem resistance in Enterobacteriaceae has increased worldwide. Whereas the increase in the prevalence of ESBL-producing Enterobacteriaceae isolated from livestock is becoming an important public health problem, the increasing prevalence of carbapenemases has only affected hospitals and the community. Recently, however, the occurrence of carbapenemase-carrying commensal Escherichia coli isolated from livestock and their environment has been reported, and this could be the beginning of a new era in the antibiotic resistance field. Within the national RESET project (www.reset-verbund.de) several longitudinal and cross-sectional studies, collecting potential ESBL-carrier organisms from German farms, have been performed (using MacConkey agar with 1 mg/L cefotaxime as the selective medium). From the 221 isolates collected during 2011, 3 of them were ascribed to Salmonella enterica subsp. enterica (Table 1). The three Salmonella isolates (R3, R25 and R27) were obtained from two pig-fattening farms (R25 was collected outside the farm) and one broiler farm (Table 1). The three farms were distributed in different locations in the same German federal region, and although there was no apparent link between them, a common source cannot be excluded. The three isolates were tested for their susceptibility to 35 antimicrobials, including b-lactams/b-lactamase inhibitors (Table 1), phenicols, aminoglycosides, quinolones/fluoroquinolones, tetracycline, folate pathway antagonists, lipopeptides and fosfomycin, as previously described. For the present study, tigecycline (15 mg) and nitrofurantoin (300 mg) were included as well. The presence of ESBLs, AmpC b-lactamases and/or carbapenemase-encoding genes, class 1 and 2 integrons and other resistance genes was screened by PCR/sequencing, as previously described (Table S1, available as Supplementary data at JAC Online). The MIC values for some carbapenemase producers can be lower than the currently recommended breakpoints, and the results of the carbapenem susceptibility tests can be influenced by the genetic background. The Salmonella isolates R3, R25 and R27 showed decreased susceptibility to these antimicrobials [non-wild-type by the EUCAST epidemiological cut-off (ECOFF), but susceptible or intermediate according to the CLSI clinical breakpoint; Table 1]. This ‘decreased susceptibility’ could be transformed to a competent E. coli recipient, but conjugation or mobilization under the conditions used was unsuccessful. The three isolates carried both the AmpC-encoding gene blaACC-1 and the carbapenemase gene blaVIM-1, like in the previously reported E. coli isolates R178 and R29. When Salmonella R3 and the control strain E. coli R178 were grown in liquid medium with carbapenems (Luria-Bertani broth with 16 mg/L imipenem or 8 mg/L ertapenem inoculated with 1:1000 overnight culture), both isolates grew well, showing full carbapenem resistance (clinical breakpoints, CLSI versus EUCAST: imipenem ≥4 versus .8 mg/L; and ertapenem ≥1 versus .1 mg/L). Several class 1 integrons (In31, In70, In71, In110 and In450), transposons (Tn3, Tn402 and Tn21) and plasmids (incompatibility groups IncHI2, IncN, IncI1 and IncW) carrying blaVIM genes have been described. Like in E. coli R178 and R29, in the three Salmonella isolates the blaVIM-1 gene was located on a class 1 integron (variable region with blaVIM-1-aacA4-aadA1 gene cassettes) harboured by an 300 kb IncHI2 plasmid (determined by S1-nuclease PFGE analysis, PCR-based replicon typing and Southern blot hybridization, as previously described; HI2 double plasmid sequence typing failed). The plasmid also carried blaACC-1, strA/B, catA1 and a trimethoprim resistance gene (not identified with the primers used; see Table S1, available as Supplementary data at JAC Online). The sequence of the complete integron of Salmonella R3 (5436 bp, including complete sul1 and orf5, obtained using as template the pRHR3 plasmid of this isolate; see Table S1, available as Supplementary data at JAC Online) was identical to the one from E. coli R178 (accession number HE663536) and was related to Tn402 (like in GQ422826). Salmonella R27 was isolated from the same farm as both E. coli R178 and R29 (Table 1). However, the IncHI2 plasmids harboured by these Salmonella and E. coli isolates were different in size and gene content ( 300 versus 220 kb and presence versus absence of chloramphenicoland trimethoprimresistance genes), suggesting different plasmid evolutions. The three isolates were classified as S. enterica group C, antigenic formula ‘6,7:–:– ’ (www.pasteur.fr) at the National Salmonella Reference Laboratory (NRL-Salm, BfR). The sequence type ST32 (http://mlst.ucc.ie/mlst/dbs/Senterica) and the PFGE patterns found (Figure S1, available as Supplementary data at JAC Online) are typical of Salmonella Infantis (6,7:r:1,5) and have also been detected in German isolates from humans, poultry/ poultry meat and pig/pork meat. Salmonella Infantis is among the top 10 Salmonella serovars implicated in human salmonellosis worldwide (ranking in third place in 2011 in Europe; www.ecdc. europa.eu). Isolates from this serovar also caused disease with Research letters