Characterisation of Reduced Susceptibility to Metronidazole in Epidemic Clostridium difficile Clinical Isolates

Abstract

Reduced susceptibility in some C. difficile isolates has been observed to metronidazole, a frontline antibiotic for C. difficile infection. Therefore, this study was performed to characterize reduced metronidazole susceptibility in UK C. difficile isolates, using phenotypic and genotypic approaches. This research focused on strains of three UK C. difficile ribotypes 027, 001, 106 including metronidazole reduced susceptible (CDRM) and metronidazole susceptible (CDSM) C. difficile strains, and 2 control strains from ribotype 010 (CDRM) and 038 (CDSM). An agar incorporation method (AIM) was used to determine the susceptibility and heterogeneity of minimum inhibitory concentrations (MICs). Ribotype 001 CDRM strains was observed for metronidazole to have the highest MIC90 (8mg/L) and the highest MIC hetereogeneity distributed across 2 MICs was observed in ribotype 106 at 22% and 11% for metronidazole and vancomycin respectively. All strains were susceptible to vancomycin at MIC ≤ 2mg/L which reflected in the high area under the curve (AUC) results. Population analysis profiling (PAP) incorporating total viable counting on antibiotic-containing agar plates was also employed and ribotype 001 had the lowest AUC (2.32) for metronidazole. Serial passage experiment (5 passages) with and without sub-inhibitory concentrations of metronidazole was performed using AIM and decline in MIC50 from 4mg/L – 0.5mg/L after 5 passages without subinhibitory concentrations of metronidazole was observed. Nitroreductase activity was also evaluated using a spectrophotometric assay and only 3 CDRM strains showed detectable nitroreductase activity at very low levels. Subsequently, metronidazole uptake capacity was determined by employing a microbiological bioassay and AIM was used to evaluate the effect of hemin on the CDRM phenotype using both Wilkins Chalgren and Brucella Agar +/- 5mg/L hemin. Hemin in Brucella agar significantly increased metronidazole MIC (P>0.05) for ribotype 001 and 027 only. The metronidazole uptake assay indicated that the CDRM phenotype was not related to deficiencies in drug reduction and uptake. It does not however eliminate possibility of metronidazole being reduced to its non-toxic amino radical by nim genes. Then analysis was done on the whole genome of two C. difficile reduced susceptible strain, E4 and a clinical strain of ribotype 001 ribotype, compared with C. difficile 630 (CDSM). Whole genome analysis showed single nucleotide polymorphisms (SNPs) in nimB, glyC, nifJ thiH and hemN genes, linked to electron transfer, activation, oxidative stress and DNA repair which could predispose to CDRM phenotype as they are linked with metronidazole mechanism of action. The SNP detected in nimB gene could activate the conversion of metronidazole to its non-toxic amino radical thus contributing to the CDRM phenotype. This study has provided the focal areas that has the potential for predisposing to metronidazole reduced susceptibility. Which are factors affecting DNA repair, oxidative stress and metronidazole reduction. Which is of significance due to the CDRM phenotype compounded by the heterogeneity and stability profile that can give false susceptibility results in clinical settings. Which complicates therapy with metronidazole due to its poor colon penetration