Prevalence and molecular characterization of clinical isolates of Escherichia coli expressing an AmpC phenotype

J Antimicrob Chemother 2010; 65: 460 464 doi:10.1093/jac/dkp484 Advance publication 22 January 2010 Prevalence and molecular characterization of clinical isolates of Escherichia coli expressing an AmpC phenotype Rikke Lind Jørgensen, Jesper Boye Nielsen, Alice Friis-Møller, Hans Fjeldsøe-Nielsen and Kristian Schønning* Department of Clinical Microbiology 445, Hvidovre Hospital, 2650 Hvidovre, Denmark *Corresponding author. Tel: þ45-3632-6161; Fax: þ45-3632-3357; E-mail: kristian.schoenning@hvh.regionh.dk Received 14 August 2009; returned 9 November 2009; revised 10 December 2009; accepted 10 December 2009 Objectives: To establish the prevalence of the AmpC b-lactamase phenotype in clinical isolates of Escherichia coli and characterize the genetic resistance mechanisms causing the observed phenotype. Methods: Clinical E. coli (n¼74) with reduced susceptibility to third-generation cephalosporins and resistance to cefoxitin were collected from the Department of Clinical Microbiology at Hvidovre Hospital, Denmark, in 2006. The AmpC disc test was used to confirm expression of AmpC, and test-positive strains were selected for further antimicrobial susceptibility testing and molecular characterization. Hyperproduction of AmpC b-lactamase was confirmed by isoelectric focusing (IEF). The presence of a plasmid-mediated ampc gene (pampc) was detected by multiplex PCR. The promoter and the entire reading frame of the chromosomal ampc gene were sequenced to identify promoter mutations associated with hyperproduction and gene mutations associated with extended-spectrum AmpC (ESAC) b-lactamase activity. Results: Twenty-four isolates exhibited a positive AmpC disc test. IEF confirmed AmpC expression in all isolates except one. Four isolates contained a bla CMY-2 gene. These were not clonally related by multilocus sequence typing (MLST). The remaining isolates all had mutations or insertions in the promoter region, which could explain increased expression of the chromosomal AmpC enzyme. Mutations in the ampc gene associated with extended activity were rare and did not cause resistance to cefepime. Sequencing of ampc showed that most isolates were not clonally related. Conclusions: E. coli expressing an AmpC phenotype occur sporadically and cause significant resistance to cephalosporins. The majority of these are hyperproducing chromosomal ampc although some isolates have acquired pampc. Keywords: CMY-2, b-lactamases, promoter mutations, ESAC Introduction AmpC b-lactamases belong to Ambler class C and are causing concern because they mediate resistance to a broad spectrum of antibiotics comparable to the extended-spectrum b-lactamases (ESBLs). In Escherichia coli the chromosomal ampc gene is normally expressed in low amounts not causing clinical resistance, because the promoter is weak and the gene contains a transcriptional attenuator. However, mutations in the promoter have been reported to play an important role in E. coli resistance to b-lactams by increasing ampc transcription. 1 Such genes conferring AmpC hyperproduction are here designated campc. Additionally, extended-spectrum AmpC (ESAC) b-lactamases with increased hydrolytic activity against ceftazidime and cefepime have been described in E. coli. The broadened substrate activity of these enzymes is due to structural modifications in the active site. 2 Finally, the presence of a plasmid-mediated ampc gene (pampc) is increasingly frequently identified in E. coli. 3 Originally such genes are thought to have been mobilized to plasmids from the chromosomal ampc genes of organisms such as Citrobacter freundii, Enterobacter cloacae, Hafnia alvei, Morganella morganii and Aeromonas spp. 3 The aim of the present study was to examine the prevalence of E. coli isolates with an AmpC phenotype among clinical isolates in our department. Additionally, we analysed the causes for the observed AmpC phenotype molecularly, determined the relative contribution of pampc and campc and compared the susceptibility patterns associated with the different genotypes. Materials and methods Between 1 January and 1 August 2006, non-replicate clinical isolates of E. coli that exhibited reduced susceptibility to at least one cephalosporin were collected at the Department of Clinical Microbiology, Hvidovre # The Author 2010. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org 460

AmpC-hyperproducing E. coli JAC Hospital, Denmark. All isolates displaying reduced susceptibility to cefpodoxime, cefuroxime, ceftriaxone or ceftazidime were screened for AmpC hyperproduction by cefoxitin susceptibility testing. Isolates with a cefoxitin inhibition zone,20 mm were designated as possible AmpC hyperproducers and were further tested by the AmpC disc test method as described previously. 4 Isolates with a positive AmpC disc test were selected for further analysis. MICs were determined using Etest (AB Biodisk) according to the manufacturer s instructions. For isoelectric focusing (IEF) analysis bacterial suspensions were disrupted by vortexing with Garnet Bead Tubes 0.15 mm Sample (Mo Bio Laboratories, USA) and cleared by centrifugation. IEF was done using standard methods. 5 PCR amplification for bla SHV, bla TEM and bla CTX-M was performed as previously described. 5,6 PCR-positive samples for bla TEM were sequenced using the same primers as for PCR. Positive samples in the multiplex PCR for bla CTX-M were re-amplified using primers specific for either group 1 or group 9 where appropriate [Table S1, available as Supplementary data at JAC Online (http://jac.oxfordjournals.org/)], encompassing the entire reading frame and sequenced using the same primers as for amplification. Multiplex amplification of the six families of plasmid-mediated ampc was performed as described previously. 7 PCR-positive strains were re-amplified by use of specific primers for amplification of the entire bla AmpC gene as previously described. 7 Sequencing of the resulting PCR products was done using the same primers as for PCR together with a set of internal primers designed in this study (Table S1). Chromosomal ampc promoter mutations were examined by PCR and sequencing, as previously described. 8 Briefly, a set of primers were used to amplify and sequence a 271 bp fragment (position 2150 to þ120) from the E. coli ampc promoter containing the 235 box, the 210 box and the attenuator. 1 PCR amplifications of ampc genes were performed with primers described by Mammeri et al., 2 yielding a 1315 bp amplification product that contained the entire ampc gene of E. coli. Strains positive for bla CMY-2 were analysed by multilocus sequence typing (MLST) following the procedure described by Wirth et al. 9 Primers specified at the E. coli MLST web site (http://mlst.ucc.ie/mlst/ dbs/ecoli/documents/primerscoli_html) were used for gene amplification and sequencing except for mdh and icd for which primers described by Nicolas-Chanoine et al. 10 were used. Results and discussion In the study period, specimens obtained from 6356 individual patients yielded positive cultures for E. coli. Seventy-four isolates displaying decreased susceptibility to third-generation cephalosporins and cefoxitin were further tested using the AmpC disc test. In all, 24/6356 (0.4%) isolates displayed a positive AmpC disc test and were selected for further analysis. IEF was done to confirm the expression of AmpC-type b-lactamase in the selected strains. AmpC-type b-lactamases display a pi of 9. All isolates with an AmpC phenotype except E23 expressed a b-lactamase with a pi of 8.9 (Table 1). No b-lactamases were detected in the fully susceptible control strain E. coli ATCC 25922. Thus, hyperproduction of AmpC was confirmed in 23/24 strains. Several isolates had more than one b-lactamase that could be visualized by IEF. Eleven strains displayed bands with pi 5.4, and sequencing identified these as bla TEM-1. Five strains displayed bands with an apparent pi of 7.6 7.9. Sequencing identified these as two bla CTX-M-14 genes, two bla CTX-M-15 genes and one bla CTX-M-55 gene. E04 and E23 displayed an unidentified b-lactamase with a pi of 6.5 whereas E01, E16, E22, E24 and E33 displayed an unidentified b-lactamase with a pi of 7.4. All strains were negative for bla SHV by PCR. All 24 isolates were screened for pampc by multiplex PCR as described previously. 7 Four isolates (16.7%) were positive for bla CIT. DNA sequencing of the entire gene showed 100% identity with the bla CMY-2 gene in all four isolates. All four bla CMY-2 -positive isolates also contained bla TEM-1. To investigate if these isolates were clonally related MLST was performed on bla CMY-2 -positive strains. All four strains belonged to individual sequence types (STs). E16 belonged to ST46 (clonal complex, ST46 with an insertion sequence IS1 in the fumc gene), E22 belonged to ST48 (clonal complex, ST10), E29 belonged to ST652 (clonal complex, none) and E33 belonged to ST448 (clonal complex, ST448). This indicates that bla CMY-2 -positive strains occurred sporadically in the study period. To investigate the extent to which hyperproduction of campc contributed to the observed phenotype (Table 1) we sequenced the promoter region of the chromosomal ampc gene to identify sequence elements associated with increased strength of the promoter (Table 2). 1 All of the ampc multiplex PCR-negative isolates with an AmpC phenotype and confirmed AmpC expression by IEF had mutations or insertions in their control regions. Group 1 comprising five isolates had a 1 bp insertion at either position 221 or 213, shifting the length of the spacer region to the optimal 17 bp between the 235 and 210 boxes. These displayed moderately decreased susceptibility to cephalosporins [geometric mean MICs: cefuroxime, 16 mg/l; cefotaxime, 2.6 mg/l; and ceftazidime, 1.3 mg/l (Table 1)]. Group 2 comprising three isolates had mutations in the ampc promoter region at position 232, changing the 235 box. Additionally, one of these isolates had mutations in the attenuator region (þ17 to þ37). This group also displayed a moderately decreased susceptibility to cephalosporins [geometric mean MICs: cefuroxime, 32 mg/l; cefotaxime, 5.0 mg/l; and ceftazidime, 2.5 mg/l (Table 1)]. Group 3 comprised two isolates with mutations at both position 232 and 211, changing the 235 box and the 210 box. Additionally, these two isolates had mutations in the attenuator. This group exhibited more pronounced resistance to cephalosporins [geometric mean MICs: cefuroxime, 91 mg/l; cefotaxime, 32 mg/l; and ceftazidime, 4.0 mg/l (Table 1)]. Group 4 comprising five isolates had mutations at positions 288, 282, 242, 218, 21 and þ58, generating alternative 235 and 210 boxes. 1 This group displayed pronounced resistance [geometric mean MICs: cefuroxime, 49 mg/l; cefotaxime, 21 mg/l; and ceftazidime, 6.1 mg/l (Table 1)]. Mutations at position 228 and in the attenuator region were identified in five isolates with an AmpC phenotype. These isolates also contained a gene encoding CTX-M. AmpC-type b-lactamases with broadened substrate activity due to mutations in the structural gene are called ESACs and exhibit extended hydrolysis of b-lactams. 2 Sequencing of the ampc coding region revealed that 10 isolates had mutations in the ampc coding region; however, all remained susceptible to cefepime except for 3 isolates, which contained an additional bla CTX-M to explain resistance (Table 1). Phylogenetic reconstruction based upon the sequence information from the ampc promoter and coding region showed that most isolates yielded unique sequences (results not shown). E04 and E20 yielded identical sequences, as did E06, E12 and E13. In the case of E04 and E20 these were isolated 3 months apart in different wards and contained different bla CTX-M. In the case of E06, E12 and E13 an epidemiological link between E12 and E13 could be suspected, as these were isolated from the same 461

462 Table 1. MIC values and resistance mechanisms detected in the 24 E. coli clinical isolates included in this study MIC Etest (mg/l) Strain FOX CTT CXM CTX CAZ FEP IPM bla SHV type bla TEM PCR and sequencing bla CTX M type plasmidmediated bla AmpC genes Mutations in the ampc promoter at position: Mutations in the AmpC protein pi by IEF Jørgensen et al. campc isolates group 1 E06 16 1 16 2 1 0.0625 0.25 1 bp insertion (213) wt AmpC 8.9 E12 64 8 32 4 2 0.125 0.25 1 bp insertion (213) wt AmpC 8.9 E13 16 2 16 2 1 0.0625 0.25 1 bp insertion (213) wt AmpC 8.9 E34 16 1 8 2 1 0.0625 0.25 1 bp insertion (213) wt AmpC 8.9 E28 32 2 16 4 2 0.125 0.25 bla TEM 1 273/228/1 bp insertion (221) A351T/A359T 5.4/8.9 group 2 E14 16 2 16 2 2 0.0625 0.25 273/232 wt AmpC 8.9 E15 32 2 16 4 2 0.0625 0.25 273/232/þ58 A351T/A359T 8.9 E05 256 16 128 16 4 0.25 0.25 bla TEM 1 232/attenuator (þ22, þ26, þ27, þ32) 3 aa insertion at D123/ Q180H/N185T group 3 E11.256 32 128.32 8 0.25 0.25 232/211/attenuator (þ37) N185T/ T308A 8.9 E35 64 8 64 32 2 0.0625 0.25 273/232/228/211/attenuator (þ17) A351T/A359T 8.9 group 4 E17 16 4 16.32 4 0.031 0.25 288/282/242/218/21/þ58 wt AmpC 8.9 E18 32 8 32 16 8 0.125 0.25 bla TEM 1 288/282/242/218/21/þ58 wt AmpC 5.4/8.9 E21 8 4 16 8 4 0.125 0.25 288/282/242/218/21/þ58 wt AmpC 8.9 E02.256 32.256.32 4 0.5 0.25 288/282/242/218/21/þ58/ wt AmpC 8.9 attenuator (þ20) E07 256 32 128.32 16 0.5 0.25 bla TEM 1 288/282/242/218/21/þ58/ attenuator (þ23) wt AmpC 5.4/8.9 CTX-M and campc E24 16 1.256.32 16 16 0.125 bla CTX M 15 273/228/þ58/attenuator (þ34) wt AmpC 7.4/7.9/8.9 E01 16 1.256.32 16 16 0.25 bla CTX M 15 273/228/þ58/attenuator (þ34) wt AmpC 7.4/7.9/8.9 E20 8 0.5.256.32 1 8 0.25 bla TEM 1 bla CTX M 14 273/228/þ58 A351T/A359T 5.4/7.6/8.9 E04 8 2.256.32 64 64 0.5 bla TEM 1 bla CTX M 55 273/228/þ58 A351T/A359T 5.4/6.5/7.6/8.9 E23 16 2.256.32 4 32 0.25 bla TEM 1 bla CTX M 14 attenuator (þ22, þ26, þ27, þ32) Q180H/N185T 5.4/6.5/7.9 bla CMY isolates E16 64 32 32.32 16 2 0.5 bla TEM 1 bla CMY 2 A351T 5.4/7.4/8.9 E22 64 16 16.32 8 0.25 0.25 bla TEM 1 bla CMY 2 A351T 5.4/7.4/8.9 E29 32 4 16.32 8 0.25 0.25 bla TEM 1 bla CMY 2 288/282/218/21/þ58 wt AmpC 5.4/8.9 E33 32 4 16.32 8 0.25 0.25 bla TEM 1 bla CMY 2 288/282/218/21/þ58 wt AmpC 5.4/7.4/8.9 Cephalosporin-susceptiple control ATCC 25922 2 0.25 2 0.5 0.25 0.0625 0.25 273 wt AmpC 5.4/8.9 FOX, cefoxitin; CTT, cefotetan; CXM, cefuroxime; CTX, cefotaxime; CAZ, ceftazidime; FEP, cefepime; IPM, imipenem; wt, wild-type; aa, amino acid. Downloaded from http://jac.oxfordjournals.org/ by guest on December 31, 2015

Table 2. Mutations detected in the ampc promoter region of the 24 E. coli clinical isolates Mutation at position Strain 288 282 276 273 242 232 228 221 218 213 211 21 þ5 þ17 þ20 þ22 þ23 þ26 þ27 þ32 þ34 þ37 þ49 þ58 þ63 þ70 þ81 K12 a C A G C C T G 2 c G 2 c C C C C C C G T A G G G A C T C G campc isolates group 1 E06 T T T A E12 T T T A E13 T T T A E34 T T A E28 T A T A AmpC-hyperproducing E. coli 463 group 2 E14 T A A E15 T A T C A E05 A A T G T A T group 3 E11 A T A G T E35 T A A T T A group 4 E17 T G T A T T E18 T G T A T T E21 T G T A T T E02 T G T A T A T E07 T G T A T A G T CTX-M/cAmpC E24 T A A T A E01 T A A T A E20 T A T A E04 T A T A E23 A T G T A T bla CMY-2 isolates E16 E22 E29 T G A T T E33 T G A T T Controls ATCC 8739 a ATCC 25922 b T A a The GenBank nucleotide sequence accession numbers for E. coli K12 and E. coli ATCC 8739 were NC000913 and CP000946, respectively. b E. coli ATCC 25922 was sequenced together with the 24 E. coli clinical isolates. c 1 bp insertion. JAC Downloaded from http://jac.oxfordjournals.org/ by guest on December 31, 2015

Jørgensen et al. ward 1 day apart. However, the majority of isolates occurred sporadically during the study period. In conclusion, E. coli strains that hyperproduce AmpC either caused by the acquisition of bla CMY-2 or as a result of promoter mutations associated with hyperproduction of chromosomally encoded bla AmpC occur sporadically in Denmark. In the present study 5/24 isolates with an AmpC phenotype had a 1 bp insertion between the 210 and 235 boxes (group 1 campc) and 3/24 isolates had mutations in the 235 box to a consensus box (group 2 campc). These mutation patterns in the present study were only associated with low-level resistance. In other studies isolates with 242 promoter mutations have predominated. 2,5 It is likely that different selection criteria for study strains contributed to the observed differences in prevalence. As hyperproduction of AmpC is associated with clinically significant resistance, their detection, proper reporting and monitoring of future epidemiology are important. Acknowledgements Arnfinn Sundsfjord and Bjørg Haldorsen are gratefully acknowledged for providing control strains for PCR. Funding This work was supported by Fondation Idella (3.3.6-2007 to K. S.) and Hvidovre Hospitals Forskningsfond (2009020025 to R. L. J.). Transparency declarations None to declare. Supplementary data Table S1 is available as Supplementary data at JAC Online (http://jac. oxfordjournals.org/). References 1 Caroff N, Espaze E, Gautreau D et al. Analysis of the effects of 242 and 232 ampc promoter mutations in clinical isolates of Escherichia coli hyperproducing AmpC. J Antimicrob Chemother 2000; 45: 783 8. 2 Mammeri H, Eb F, Berkani A et al. Molecular characterization of AmpC-producing Escherichia coli clinical isolates recovered in a French hospital. J Antimicrob Chemother 2008; 61: 498 503. 3 Jacoby GA. AmpC b-lactamases. Clin Microbiol Rev 2009; 22: 161 82. 4 Black JA, Moland ES, Thomson KS. AmpC disk test for detection of plasmid-mediated AmpC b-lactamases in Enterobacteriaceae lacking chromosomal AmpC b-lactamases. J Clin Microbiol 2005; 43: 3110 3. 5 Haldorsen B, Aasnaes B, Dahl KH et al. The AmpC phenotype in Norwegian clinical isolates of Escherichia coli is associated with an acquired ISEcp1-like ampc element or hyperproduction of the endogenous AmpC. J Antimicrob Chemother 2008; 62: 694 702. 6 Woodford N, Fagan EJ, Ellington MJ. Multiplex PCR for rapid detection of genes encoding CTX-M extended-spectrum b-lactamases. J Antimicrob Chemother 2006; 57: 154 5. 7 Pérez-Pérez FJ, Hanson ND. Detection of plasmid-mediated AmpC b-lactamase genes in clinical isolates by using multiplex PCR. J Clin Microbiol 2002; 40: 2153 62. 8 Corvec S, Caroff N, Espaze E et al. 211 mutation in the ampc promoter increasing resistance to b-lactams in a clinical Escherichia coli strain. Antimicrob Agents Chemother 2002; 46: 3265 7. 9 Wirth T, Falush D, Lan R et al. Sex and virulence in Escherichia coli: an evolutionary perspective. Mol Microbiol 2006; 60: 1136 51. 10 Nicolas-Chanoine MH, Blanco J, Leflon-Guibout V et al. Intercontinental emergence of Escherichia coli clone O25:H4-ST131 producing CTX-M-15. J Antimicrob Chemother 2008; 61: 273 81. 464