A comparison of methods to determine whether clinical isolates of Staphylococcus epidermidis from the same patient are related

Ecological Overlap and Horizontal Gene Transfer in Staphylococcus aureus and Staphylococcus epidermidis

The opportunistic pathogens Staphylococcus aureus and Staphylococcus epidermidis represent major causes of severe nosocomial infection, and are associated with high levels of mortality and morbidity worldwide. These species are both common commensals on the human skin and in the nasal pharynx, but are genetically distinct, differing at 24% average nucleotide divergence in 1,478 core genes. To better understand the genome dynamics of these ecologically similar staphylococcal species, we carried out a comparative analysis of 324 S. aureus and S. epidermidis genomes, including 83 novel S. epidermidis sequences. A reference pan-genome approach and whole genome multilocus-sequence typing revealed that around half of the genome was shared between the species. Based on a BratNextGen analysis, homologous recombination was found to have impacted on 40% of the core genes in S. epidermidis, but on only 24% of the core genes in S. aureus. Homologous recombination between the species is rare, with a maximum of nine gene alleles shared between any two S. epidermidis and S. aureus isolates. In contrast, there was considerable interspecies admixture of mobile elements, in particular genes associated with the SaPIn1 pathogenicity island, metal detoxification, and the methicillin-resistance island SCCmec. Our data and analysis provide a context for considering the nature of recombinational boundaries between S. aureus and S. epidermidis and, the selective forces that influence realized recombination between these species.

Bacterial substitution of coagulase-negative staphylococci for streptococci on the oral mucosa after hematopoietic cell transplantation

PURPOSE

Coagulase-negative staphylococci (CoNS) are frequently isolated from blood cultures of hematopoietic cell transplantation (HCT) patients. Generally, the use of central venous catheters is recognized as a significant risk factor for CoNS infection, while the impact of CoNS infection from oral ulcerative mucositis, which occurs frequently in HCT, may be underestimated. Here, we examined the bacteria on the buccal mucosa after HCT.

METHODS

Sixty-one patients were examined for bacteria on the buccal mucosa routinely once a week from 1 week before to 3 weeks after allogeneic HCT. Subjects were divided into groups with short and long periods of antibiotic use, and differences in bacterial substitution were evaluated. The relationships between type of HCT (conventional HCT or RIST) and bacterial substitution were also evaluated.

RESULTS

The changes in detection frequencies of CoNS and α-streptococci from before to 3 weeks after HCT were significant (P < 0.05, χ (2) test): 14.5-53.3% and 92.7-53.1%, respectively. Significant bacterial substitution of CoNS for streptococci was observed in the long-term antibiotic use group (P < 0.05, χ (2) test), but also occurred in cases with short-term or no antibiotic use. No relationships between type of HCT (conventional HCT or RIST) were observed.

CONCLUSION

Bacterial substitution of CoNS for streptococci occurred frequently on the buccal mucosa after HCT. In addition to antibiotic use, environmental factors may be involved in bacterial substitution. It is important to consider the presence of oral mucositis in CoNS infection after HCT.

Molecular analysis of coagulase-negative Staphylococcus isolates from blood cultures: prevalence of genotypic variation and polyclonal bacteremia

Fifty-seven coagulase-negative Staphylococcus isolates from 22 inpatients who had > or =2 blood cultures that were positive for Staphylococcus within 24 hours were analyzed to determine the frequency of polyclonal bacteremia. Patients were considered to have bacteremia (14 patients) or contamination of sample (8 patients) on the basis of clinical criteria. Nine colonies were randomly selected from each blood culture and genotyped by means of SmaI digestion/pulsed-field gel electrophoresis. Relatedness was determined by calculation of the Dice coefficient of banding-pattern similarity (S(AB)). Analysis of bacteremic isolates demonstrated the presence of a single species in 35 of 41 blood cultures, 1 related variant in 5 blood cultures (87%-92% S(AB)), and an unrelated strain in 1 blood culture (79% S(AB)). Analysis of contaminated samples demonstrated the presence of a single strain in 10 of 16 blood cultures and 1-3 variants (28%-97% S(AB)) in the remainder. Genotype diversity was significantly more common in the contaminated samples (P=.036). Almost all coagulase-negative Staphylococcus bacteremias were monoclonal.

Polyclonal Staphylococcal endocarditis caused by genetic variability

Cultures of blood obtained from a patient with Staphylococcus epidermidis prosthetic valve endocarditis yielded 15 strains of S. epidermidis. Genome macrorestriction and amplified fragment-length polymorphism analyses of these strains showed that they belonged to 4 different, very closely related clones, suggesting that they were the result of genetic variability of an infecting strain during the infectious episode. In vivo experiments in a rat model for foreign body infections using 1 of the S. epidermidis strains from the patient showed genetic variability similar to that of the infecting strain. In the rat model, we also detected the simultaneous presence of different clones that were identical to those isolated from our patient, thus confirming the possibility of genetic variability. It is important to note that the 4 clones isolated from our patient presented with 2 different antibiograms. Therefore, in cases of foreign device-related infections due to coagulase-negative staphylococci, the possibility of polyclonal infection has to be taken into account, particularly as regards differences in antibiotic susceptibility.

AFLP typing of Staphylococcus epidermidis in multiple sequential blood cultures

AFLP is a novel high-resolution PCR-based DNA fingerprinting method generating complex banding patterns that can be used for comparative analysis. In the present study, the applicability of AFLP in fingerprinting Staphylococcus epidermidis isolates was investigated. The criteria considered were stability of patterns, reproducibility, discriminatory capacity and consistency with epidemiological context. Repeated testing of strains and investigation of subcultures showed that AFLP patterns were reproducible and stable with an intrastrain similarity of S > or = 94% as determined by analysis of digitized patterns. Fifteen unrelated strains were heterogeneous, with a level ranging from 78-93%. The applicability of AFLP in epidemiological studies of S. epidermidis was tested on 11 sets of four blood isolates each, from 11 patients with suspected septicaemia. Nine sets had indistinguishable or highly similar AFLP patterns for each isolate per set, while two sets had heterogeneous patterns. These results show that AFLP has high discriminatory power for strain identification in S. epidermidis.