Tigecycline-resistant Enterococcus faecalis strain isolated from a German intensive care unit patient

Sir, Tigecycline is a member of the new group of glycylcyclines and a promising new antibiotic of last resort, active against many bacteria including Enterococcus spp.1,2 Tigecycline acts in a similar way to tetracyclines by binding to the 30S subunit of the bacterial ribosome and thus inhibiting protein biosynthesis;1 however, different binding capacities and kinetics also allow activity against tetracycline-resistant bacteria.1 Enterococcal isolates displaying MICs of tigecycline of ≤0.25 mg/L are considered susceptible.3 The epidemiological cut-off value (breakpoint) for tigecycline is >0.5 mg/L for enterococci.3 Here, we report the first case of a tigecycline-resistant Enterococcus from a German hospital patient isolated in 2007. The 65-year-old patient underwent an intra-abdominal surgery in October 2006 following a femoral neck fracture. Soon after, she suffered from several post-operative complications including colon perforation, peritonitis, nosocomial pneumonia after long-term ventilation and renal failure. The patient showed septic signs for weeks and months. She died of multiorgan failure in June 2007. She was treated with several courses of antibiotics; including, among others, tigecycline for more than 2 weeks to treat multiresistant Stenotrophomonas maltophilia isolated from tracheal secretion. Soon after, enterococci were isolated from catheter urine samples with colony counts of 1 × 105 cfu/mL in January 2007. Species identification revealed Enterococcus faecalis. Initial Etest for tigecycline resulted in an MIC of 6 mg/L. The strain (UW6940) was sent to the Robert Koch Institute for confirmation and further characterizations. Tigecycline susceptibility was determined by broth microdilution and Etest. Non-susceptibility to tigecycline is mediated via efflux porters in Acinetobacter baumannii or via mutations in Tet(A)-mediating tetracycline efflux, e.g. in Escherichia coli. We tested MICs of tigecycline in the presence and absence of several efflux pump inhibitors.4 Concentrations were chosen as given in the literature and were as follows: reserpine, 20 mg/L; verapamil and omeprazole, 60 mg/L; and prochlorperazine, 2 mg/L (16 mg/L prochlorperazine alone already inhibited growth of the test strain). A single tetracycline resistance gene, tetX, encodes an oxygen-dependent monooxygenase conferring tigecycline resistance.5 So, MICs of tigecycline were determined under anaerobic and aerobic growth conditions. In addition, we tested amplification of a PCR product specific for tetX (primers: tetX-F: CAATAATTGGTGGTGGACCC; tetX-R: TTCTTACCTTGGACATCCCG; 468 bp) with DNA from strain UW6940 and a reference E. coli strain Em24 pBSJ possessing tetX cloned into a pBR328 plasmid vector (kindly provided by Professor M. Roberts, WA, USA). Tigecycline interacts with different target nucleotides of the ribosomal 16S rRNA.1 Mutations at those positions could render UW6940 non-susceptible to tigecycline. We sequenced the 16S rDNA of UW6940 and compared it with the sequence of the tigecycline-susceptible, fully sequenced E. faecalis V583. Strain UW6940 was grown for 2 weeks (∼400–500 generations) on brain heart infusion (BHI; Difco labs, Sparks, MI, USA) agar plates and in BHI liquid broth in the absence of any selective pressure to test the stability of tigecycline resistance. Transferability of the resistance trait was identified by in vitro filter- and broth-mating using tigecycline-susceptible E. faecalis recipient strains JH2-2 and OG1X. The MLST sequence type (ST) was indentified and allocated using a web-based internet service (http://efaecalis.mlst.net/). Resistance to tigecycline in strain UW6940 was confirmed by Etest, revealing an MIC of 2 mg/L (Table 1). MIC determination in broth medium revealed an MIC of 1 mg/L. Tigecycline MICs for susceptible reference strains E. faecium ATCC 19434 and E. faecalis ATCC 29212 were between 0.047 and 0.125 mg/L (Table 1). Addition of several efflux pump inhibitor substances did not show any effect (always 1 mg/L), except with omeprazole where addition of it resulted in a several-fold increased MIC (64 mg/L with omeprazole versus 1 mg/L without omeprazole). This intriguing antagonistic effect of omeprazole was obviously concentration-dependent and strain-specific since it did not appear at lower concentrations (20 mg/L) and with enterococcal and staphylococcal reference strains (data not shown). The MIC of tigecycline was not influenced by aerobic or anaerobic growth conditions. In addition, we failed to amplify by PCR an internal fragment specific to the tetX resistance gene with DNA from UW6940, whereas we were able to demonstrate an expected PCR product with DNA from the tetX-positive reference strain. MICs (mg/L) of tigecycline tested by Etest and broth microdilution BHI, brain heart infusion (Difco); MH, Mueller–Hinton (Difco); IST, Iso-Sensitest (Oxoid, Basingstoke, UK). MICs (mg/L) of tigecycline tested by Etest and broth microdilution BHI, brain heart infusion (Difco); MH, Mueller–Hinton (Difco); IST, Iso-Sensitest (Oxoid, Basingstoke, UK). The ∼1.5 kb 16S rDNA fragment was identical in E. faecalis V583 and UW6940 (data not shown). The sequenced fragment covered all relevant sites of interaction between tigecycline and the 16S rDNA of the 30S ribosomal subunit in the model bacterium Thermus thermophilus.1 The strain was grown for 2 weeks (∼400–500 generations) on BHI agar plates and in BHI liquid broth without any selective pressure in order to test the stability of tigecycline resistance. Progenies were sampled after the first and second weeks of passage. MICs of tigecycline for all progenies remained stable at 1 mg/L (data not shown). The resistance trait was not transferable in vitro irrespective of the test method or recipient used. Molecular typing using MLST revealed ST6. ST6 belongs to MLST clonal complex 2 combining mainly isolates from hospital outbreaks.6 Our preliminary results show stable tigecycline resistance in an epidemic, hospital-adapted E. faecalis strain type from an intensive care unit patient after prolonged tigecycline therapy. The basis of tigecycline resistance is not due to mutations in the 16S rDNA, tetX-encoded or efflux-mediated. The exact mechanism of resistance could not be elucidated. None to declare. Excellent technical assistance by Mrs Uta Geringer is highly acknowledged. We thank Professor Dr M. Roberts, University of Washington, USA, for the gift of reference strain E. coli Em24 pBSJ (containing tetX). A special thanks to Dr R. J. Willems, University Medical Center, Utrecht, The Netherlands, for developing and curating the MLST scheme for E. faecalis and the Imperial College, London, UK, for hosting the MLST web-based service as well as the Wellcome Trust, London, UK, for funding these services (http://efaecalis.mlst.net/).