摘要
Data on the usefulness of whole-genome sequencing (WGS) for invasive Campylobacter fetus infections are scarce. However, such methods are of great value for clinical and epidemiological evaluation considering the lack of standardized microbiological testing protocols for this pathogen. We present two cases of disseminated C. fetus infections, highlight the pitfalls of microbiological diagnostics and compare resistance genes detected by WGS with phenotypic susceptibility testing in five clinical isolates. We further demonstrate the application of WGS for epidemiological evaluation of transmission. A 71-year-old male presented with fever and headaches. He was diagnosed with meningitis, multiple mycotic aneurysms and tricuspid valve endocarditis with C. fetus and treated with imipenem, leading to clinical improvement. The outcome was fatal because antibiotic therapy was prematurely stopped by the patient. An 89-year-old woman presented with fever and pain of the right hip and lower back. She was diagnosed with septic arthritis of the lumbar facet joints, hip arthroplasty infection with bursitis and mitral valve endocarditis with C. fetus. Treatment included surgical revision arthroplasty of the hip and bursa and imipenem for eight weeks. Outcome was favourable. Both patients show a broad range of clinical manifestations, as previously reported in invasive C. fetus infections [[1]Pacanowski J. Lalande V. Lacombe K. Boudraa C. Lesprit P. Legrand P. et al.Campylobacter bacteremia: clinical features and factors associated with fatal outcome.Clin Infect Dis. 2008; 47: 790-796Crossref PubMed Scopus (142) Google Scholar,[2]Gazaigne L. Legrand P. Renaud B. Bourra B. Taillandier E. Brun-Buisson C. et al.Campylobacter fetus bloodstream infection: risk factors and clinical features.Eur J Clin Microbiol Infect Dis. 2008; 27: 185-189Crossref PubMed Scopus (69) Google Scholar]. However, multifocal infections with C. fetus might be underrepresented in the current literature, potentially leading to inadequate therapy and higher risk of relapse. Growth of C. fetus in blood cultures was essential for correct diagnosis of disseminated infection in both patients. While conventional cultures and 16S rRNA gene PCR were unable to detect C. fetus, reprocessing of the samples with microaerophilic incubation led to the correct diagnoses. As a result, we suggest that in patients with C. fetus bacteraemia, disseminated infection must be sought. Of the five isolates analysed, four strains were initially grown from standard aerobic and anaerobic blood culture bottles. One strain was cultured from a stool sample in microaerophilic environment. Subcultures were incubated in CO2-enriched atmosphere and in microaerophilic atmosphere. All five strains showed growth at 36°C and 41°C, but in a microaerophilic environment only. The two case isolates and three control strains of C. fetus from the University Hospital Basel were subject to WGS on an Illumina MiSeq 2 × 300bp after NexteraXT library preparation (Table 1). All isolates were sequenced to a coverage in excess of mean 80×. Assembly of the case strain data was performed as previously reported [[4]Wick R.R. Judd L.M. Gorrie C.L. Holt K.E. Unicycler: Resolving bacterial genome assemblies from short and long sequencing reads.PLoS Comput Biol. 2017; 13 (Published 2017 Jun 8)e1005595https://doi.org/10.1371/journal.pcbi.1005595Crossref PubMed Scopus (2888) Google Scholar]. The assemblies were annotated with Prokka v1.13 [[5]Seemann T. Prokka: rapid prokaryotic genome annotation.Bioinformatics. 2014; 30: 2068-2069https://doi.org/10.1093/bioinformatics/btu153Crossref PubMed Scopus (7502) Google Scholar] and ABRicate v 0.8.10 (https://github.com/tseemann/abricate) was used to search for antimicrobial resistance genes, virulence genes and plasmids using the NCBI or CARD AMR gene databases, VFDB and PlasmidFinder respectively. The antimicrobial resistance genes identified within the genome assemblies (all at 100% identity) were aminoglycoside nucleotidyltransferase ANT(6)-Ib in both case strains conferring streptomycin resistance (https://aac.asm.org/content/54/7/3052.long), and an adjacent gene encoding tetracycline resistance ribosomal protection protein Tet(44) in the isolate from case 2. These genes were also found in control 1 and 3 isolates; control 2 isolate carries a chromosomally located tet(O). Case 2 and control 3 isolates also carry a D91E mutation in the quinolone determining region (QRDR) of gyrA, matching the raised ciprofloxacin MIC. All isolates carry genes encoding a potential type IV secretion system (T4SS) and flagella. No plasmids were identified in any of the five sequenced isolates.Table 1MICs of clinical isolatesYear of isolationCase 1Case 2Control 1Control 2Control 320192019201620162018Source of strainBlood cultureBlood cultureStoolBlood cultureBlood cultureAntibioticsMIC (mg/L)InterpretationMIC (mg/L)InterpretationMIC (mg/L)InterpretationMIC (mg/L)InterpretationMIC (mg/L)InterpretationAmoxicillin/Clavulanic acid0.125SaPK-PD (non-species related) breakpoints and0.19SaPK-PD (non-species related) breakpoints and0.016SaPK-PD (non-species related) breakpoints and<0.016SaPK-PD (non-species related) breakpoints and<0.016SaPK-PD (non-species related) breakpoints andErtapenem0.19SaPK-PD (non-species related) breakpoints and0.19SaPK-PD (non-species related) breakpoints and0.125SaPK-PD (non-species related) breakpoints and0.25SaPK-PD (non-species related) breakpoints and0.19SaPK-PD (non-species related) breakpoints andImipenem0.064SaPK-PD (non-species related) breakpoints and0.047SaPK-PD (non-species related) breakpoints and0.064SaPK-PD (non-species related) breakpoints and0.064SaPK-PD (non-species related) breakpoints and0.047SaPK-PD (non-species related) breakpoints andMeropenem0.064SaPK-PD (non-species related) breakpoints and0.047SaPK-PD (non-species related) breakpoints and0.064SaPK-PD (non-species related) breakpoints and0.064SaPK-PD (non-species related) breakpoints and0.032SaPK-PD (non-species related) breakpoints andErythromycin1.5SaPK-PD (non-species related) breakpoints and3SaPK-PD (non-species related) breakpoints and1.5SaPK-PD (non-species related) breakpoints and1SaPK-PD (non-species related) breakpoints and2SaPK-PD (non-species related) breakpoints andClarithromycin34223Azithromycin11.510.752Tetracycline6RbCampylobacter jejuni and coli breakpoints in The European Committee on Antimicrobial Susceptibility Testing. Breakpoint tables for interpretation of MICs and zone diameters. Version 9.0, 2019. http://www.eucast.org.6RbCampylobacter jejuni and coli breakpoints in The European Committee on Antimicrobial Susceptibility Testing. Breakpoint tables for interpretation of MICs and zone diameters. Version 9.0, 2019. http://www.eucast.org.6RbCampylobacter jejuni and coli breakpoints in The European Committee on Antimicrobial Susceptibility Testing. Breakpoint tables for interpretation of MICs and zone diameters. Version 9.0, 2019. http://www.eucast.org.96RbCampylobacter jejuni and coli breakpoints in The European Committee on Antimicrobial Susceptibility Testing. Breakpoint tables for interpretation of MICs and zone diameters. Version 9.0, 2019. http://www.eucast.org.6RbCampylobacter jejuni and coli breakpoints in The European Committee on Antimicrobial Susceptibility Testing. Breakpoint tables for interpretation of MICs and zone diameters. Version 9.0, 2019. http://www.eucast.org.Doxycycline232244Amikacin0.750.750.50.751Gentamicin0.250.250.190.190.25Ciprofloxacin0.38IaPK-PD (non-species related) breakpoints and1RaPK-PD (non-species related) breakpoints and0.38IaPK-PD (non-species related) breakpoints and0.38IaPK-PD (non-species related) breakpoints and0.75RaPK-PD (non-species related) breakpoints andAMR genes and mutations identified (100% ID)ant(6)-Ib, tet(44)ant(6)-Ib, gyrA D91E—ctet(44) may be present in the genome of this isolate, given its phylogenetic proximity to control 3 and phenotype, but the gene was not identified in the assembly, with adjacent genes found at contig breaks.ant(6)-Ib, tet(44)tet(O)ant(6)-Ib, tet(44), gyrA D91EMIC, minimal inhibitory concentration. Interpretation according toa PK-PD (non-species related) breakpoints andb Campylobacter jejuni and coli breakpoints in The European Committee on Antimicrobial Susceptibility Testing. Breakpoint tables for interpretation of MICs and zone diameters. Version 9.0, 2019. http://www.eucast.org.c tet(44) may be present in the genome of this isolate, given its phylogenetic proximity to control 3 and phenotype, but the gene was not identified in the assembly, with adjacent genes found at contig breaks. Open table in a new tab MIC, minimal inhibitory concentration. Interpretation according to Further bioinformatic analysis was performed in CLC genomics workbench v10.1.1 using default methods of assembly (slow), mapping, variant calling at minimum coverage of 10×, minimum percentage at 10%, including MNVs. A single nucleotide polymorphism (SNP) comparison of the case isolates (mapping both to the assembly of the case 1 isolate) showed that they differ by over 3500 SNPs, ruling out a transmission or common source. A phylogeny was generated with all five sequenced strains and database isolates selected to represent the eight clusters previously identified in C. fetus [[3]Iraola G. Forster S.C. Kumar N. Lehours P. Bekal S. García-Peña F.J. et al.Distinct Campylobacter fetus lineages adapted as livestock pathogens and human pathobionts in the intestinal microbiota.Nat Commun. 2017; 8: 1367Crossref PubMed Scopus (32) Google Scholar]. All isolates in the tree are subspecies fetus, except for ERR1069014 in cluster 1 and ERR1024002 in cluster 2 (both C. fetus subsp. venerealis (Cfv)) (Fig. 1) [[3]Iraola G. Forster S.C. Kumar N. Lehours P. Bekal S. García-Peña F.J. et al.Distinct Campylobacter fetus lineages adapted as livestock pathogens and human pathobionts in the intestinal microbiota.Nat Commun. 2017; 8: 1367Crossref PubMed Scopus (32) Google Scholar]. This demonstrates that the isolates from case 1, control 1 and control 2 fall within Cluster 2 among European isolates from the last 15 years, with a divergence of approximately 25 SNPs in Cluster 2. The isolates from case 2 and control 3 fall within Cluster 4, within which the diversity is approximately 500 SNPs, with the Swiss isolates being most closely related to each other, separated by 28. However, the different complements of AMR determinants within the isolates also indicate their independent histories. Comparison of the strains by WGS clearly classified the two isolates as unrelated, despite the overlap in geographic and time of presentation. All five Swiss isolates are highly likely to be subsp. fetus, from genomic comparisons. The isolate from case 1 appears to derive from a cluster of European isolates, with the most closely related being from a Swiss case three years previously. The isolate from case 2 is also most closely related to another Swiss isolate, suggesting that this may be a local strain. Phenotypic and genotypic antimicrobial resistance profiles largely agreed. No additional resistance or virulence conferring factors were identified within the genomes of the case isolates compared to control isolates. Our results demonstrate the usefulness of WGS as an additional diagnostic tool for this rare pathogen. Currently, as no breakpoints are defined for C. fetus, detection of resistance genes can guide treatment decisions and possibly improve patient outcomes. Epidemiological evaluation of transmission routes can aid our current knowledge of the disease.