spa typing alone is not sufficient to demonstrate endemic establishment of meticillin-resistant Staphylococcus aureus in a low-prevalence country

Rapid nanopore-based DNA sequencing protocol of antibiotic-resistant bacteria for use in surveillance and outbreak investigation

Outbreak investigations are essential to control and prevent the dissemination of pathogens. This study developed and validated a complete analysis protocol for faster and more accurate surveillance and outbreak investigations of antibiotic-resistant microbes based on Oxford Nanopore Technologies (ONT) DNA whole-genome sequencing. The protocol was developed using 42 methicillin-resistant Staphylococcus aureus (MRSA) isolates identified from former well-characterized outbreaks. The validation of the protocol was performed using Illumina technology (MiSeq, Illumina). Additionally, a real-time outbreak investigation of six clinical S. aureus isolates was conducted to test the ONT-based protocol. The suggested protocol includes: (1) a 20 h sequencing run; (2) identification of the sequence type (ST); (3) de novo genome assembly; (4) polishing of the draft genomes; and (5) phylogenetic analysis based on SNPs. After the sequencing run, it was possible to identify the ST in 2 h (20 min per isolate). Assemblies were achieved after 4 h (40 min per isolate) while the polishing was carried out in 7 min per isolate (42 min in total). The phylogenetic analysis took 0.6 h to confirm an outbreak. Overall, the developed protocol was able to at least discard an outbreak in 27 h (mean) after the bacterial identification and less than 33 h to confirm it. All these estimated times were calculated considering the average time for six MRSA isolates per sequencing run. During the real-time S. aureus outbreak investigation, the protocol was able to identify two outbreaks in less than 31 h. The suggested protocol enables identification of outbreaks in early stages using a portable and low-cost device along with a streamlined downstream analysis, therefore having the potential to be incorporated in routine surveillance analysis workflows. In addition, further analysis may include identification of virulence and antibiotic resistance genes for improved pathogen characterization.

Repeated introduction and spread of the MRSA clone t304/ST6 in northern Europe

OBJECTIVES

During the last decades several methicillin-resistant Staphylococcus aureus (MRSA) clones with the capability of global spread have emerged in the community. Here, we have investigated a large collection of clinical isolates belonging to MRSA clone t304/ST6, which has emerged in many European countries over the last years, in order to retrace its phylogeny and its spread.

METHODS

We characterized 466 ST6 isolates from Denmark (n = 354), France (n = 10), Norway (n = 24), Sweden (n = 27) and the UK (n = 51). All had spa-type t304 (n = 454) or t304-related spa-types (n = 12) and whole genome sequencing (WGS) was carried out on Illumina Miseq or Hiseq with 100-300 bp reads. cgMLST was performed using Ridom SeqSphere.

RESULTS

A minimum spanning tree (MST) of all 466 isolates showed one large cluster including 182 isolates collected only from Denmark and related to a long-term neonatal outbreak in Copenhagen. This cluster contrasted with numerous small clusters, including the remaining Danish isolates and isolates from the other countries that interspersed throughout the tree. Most isolates were Panton-Valentine leukocidin (PVL) negative (95%) and harboured SCCmec IVa. One genome was closed using Oxford Nanopore technology and Illumina MiSeq. It contained a plasmid of 19.769 bp including the blaZ gene. A similar plasmid was found in 78% of all isolates.

DISCUSSION

t304/ST6 is a successful emerging clone and the fact that isolates from five countries are interspersed throughout the MST indicates a common origin. This clone is commonly described in the Middle East and its emergence in Europe coincides with influx of refugees from the Syrian Civil War.

Early-Stage Staphylococcus aureus Bloodstream Infection Causes Changes in the Concentrations of Lipoproteins and Acute-Phase Proteins and Is Associated with Low Antibody Titers against Bacterial Virulence Factors

S. aureus sepsis has a high complication and mortality rate. Given the limited therapeutic possibilities, effective prevention strategies, e.g., a vaccine, or the early identification of high-risk patients would be important but are not available. Our study showed an acute-phase response in patients with S. aureus bloodstream infection and evidence that lipoproteins are downregulated in plasma. Using immunoproteomics, stratification of patients appears to be achievable, since at the early stages of systemic S. aureus infection patients had low preexisting anti-S. aureus antibody levels. This strengthens the notion that a robust immune memory for S. aureus protects against infections with the pathogen.

Systemic and quantitative investigations of human plasma proteins (proteomics) and Staphylococcus aureus-specific antibodies (immunoproteomics) provide complementary information and hold promise for the discovery of biomarkers in Staphylococcus aureus bloodstream infection (SABSI). Usually, data-dependent acquisition (DDA) is used for proteome analysis of serum or plasma, but data-independent acquisition (DIA) is more comprehensive and reproducible. In this prospective cohort study, we aimed to identify biomarkers associated with the early stages of SABSI using a serum DIA proteomic and immunoproteomic approach. Sera from 49 SABSI patients and 43 noninfected controls were analyzed. In total, 608 human serum proteins were identified with DIA. A total of 386 proteins could be quantified, of which 9 proteins, mainly belonging to acute-phase proteins, were significantly increased, while 7 high-density lipoproteins were lower in SABSI. In SABSI, total anti-S. aureus serum IgG was reduced compared with controls as shown by immunoproteomic quantification of IgG binding to 143 S. aureus antigens. IgG binding to 48 of these anti-S. aureus proteins was significantly lower in SABSI, while anti-Ecb IgG was the only one increased in SABSI. Serum IgG binding to autoinducing peptide MsrB, FadB, EsxA, Pbp2, FadB, SspB, or SodA was very low in SABSI. This marker panel discriminated early SABSI from controls with 95% sensitivity and 100% specificity according to random forest prediction. This holds promise for patient stratification according to their risk of S. aureus infection, underlines the protective function of the adaptive immune system, and encourages further efforts in the development of a vaccine against S. aureus.

IMPORTANCES. aureus sepsis has a high complication and mortality rate. Given the limited therapeutic possibilities, effective prevention strategies, e.g., a vaccine, or the early identification of high-risk patients would be important but are not available. Our study showed an acute-phase response in patients with S. aureus bloodstream infection and evidence that lipoproteins are downregulated in plasma. Using immunoproteomics, stratification of patients appears to be achievable, since at the early stages of systemic S. aureus infection patients had low preexisting anti-S. aureus antibody levels. This strengthens the notion that a robust immune memory for S. aureus protects against infections with the pathogen.

Development of a rapid MALDI-TOF MS based epidemiological screening method using MRSA as a model organism

In this study we present a method using whole cell MALDI-TOF MS and VITEK MS RUO/SARAMIS as a rapid epidemiological screening tool. MRSA was used as a model organism for setting up the screening strategy. A collection of well-characterised MRSA strains representing the 19 most common Pulsed-Field Gel Electrophoresis (PFGE)-types in the region of South-West Sweden for the past 20 years was analysed with MALDI-TOF MS. A total of 111 MRSA strains were used for creating 19 PFGE-specific Superspectra using VITEK MS RUO/SARAMIS. Prior to performing the final analysis, the 19 Superspectra were combined into ten groups displaying similar peak patterns, hereafter named “MALDI-types”. Two-hundred fifty-five MRSA strains were analysed to test the constructed Superspectra/MALDI-type database. Matches to the Superspectra above a threshold of 65% (corresponding to the number of matched peaks in the Superspectrum) were considered as positive assignment of a strain to a MALDI-type. The median peak matching value for correct assignment of a strain to a MALDI-type was 78% (range 65.3–100%). In total, 172 strains (67.4%) were assigned to the correct MALDI-type and only 5.5% of the strains were incorrectly assigned to another MALDI-type than the expected based on the PFGE-type of the strain. We envision this methodology as a cost-efficient step to be used as a first screening strategy in the typing scheme of MRSA isolates, to exclude epidemiological relatedness of isolates or to identify the need for further typing.

Multiple-locus variable number tandem repeat analysis is superior to spa typing and sufficient to characterize MRSA for surveillance purposes

AIM

Assess the best approach to type methicillin-resistant Staphylococcus aureus (MRSA), Staphylococcal protein A (spa) typing, multiple-locus variable number tandem repeat analysis (MLVA) or both.

MATERIALS & METHODS

Discriminatory power of spa typing and MLVA was determined using 20,771 MRSA isolates.

RESULTS

There were twice as many MLVA types (MTs) as spa types present in the collection. Among the top 70% of the isolates, 37 spa types and 139 MTs were found. MLVA diversity among the top-10 spa types was high (diversity index 0.96), while spa diversity among the top-10 MTs was much lower (diversity index 0.83). The probability that two MRSA isolates with the same spa type also had the same MT was low (Wallace's coefficient 0.27). By contrast, most MRSA isolates yielding the same MT also had the same spa type (Wallace's coefficient 0.90).

CONCLUSION

MLVA is superior to spa typing and will suffice to characterize MRSA isolates for surveillance.

Can MLVA Differentiate among Endemic-Like MRSA Isolates with Identical Spa-Type in a Low-Prevalence Region?

The prevalence of methicillin-resistant Staphylococcus aureus (MRSA) in Norway is low, but an endemic-like MRSA clone with Staphylococcal protein A (spa)-type t304 has been established especially in nursing homes in the Oslo region causing several large outbreaks. The challenge was that spa-typing and the gold standard Pulsed-Field Gel Electrophoresis (PFGE) were inadequate in discriminating isolates in outbreak investigations. Additional higher resolution genotyping methods were needed. The aims of this study were a) to evaluate whether Multiple-Locus Variable number of tandem repeat Analysis (MLVA) could differentiate within the PFGE clusters between epidemiologically related and unrelated endemic-like ST8-MRSA-IV-t304-PVL-neg (MRSA-t304) isolates and b) investigate the evolution of the endemic-like MRSA-t304 clone over a 15-year time period. All MRSA-t304 isolates detected in the region from 1998 through April 2013 were included. In total, 194 of 197 isolates were available for PFGE and MLVA analyses. PFGE results on isolates from 1998-2010 have been published previously. Two PFGE clusters subdivided into eight MLVA types were detected. One major outbreak clone (PFGE cluster C2/ MLVA type MT5045) appeared from 2004 to 2011 causing long-lasting and large outbreaks in seven nursing homes and one hospital. Five new MLVA types (N = 9 isolates) differing in only one VNTR compared to the outbreak clone C2/MT5045 were detected, but only one (C2/MT5044) was seen after 2011. We suggest that MLVA can replace PFGE analysis, but MLVA may not be the optimal method in this setting as it did not discriminate between all epidemiologically unrelated isolates. The results may indicate that all eight outbreaks in different locations within the PFGE C2 cluster may be branches of one large regional outbreak. The major outbreak strain C2/MT5045 may now, however, be under control, extinguished or has moved geographically.

Clonality and antimicrobial susceptibility of methicillin-resistant Staphylococcus aureus at the University Hospital Zurich, Switzerland between 2012 and 2014

Background

Methicillin-resistant Staphylococcus aureus (MRSA) is a global epidemic threat. The aim of this study was to determine which globally known MRSA lineages are currently present at our tertiary care hospital in Switzerland, a hospital with low MRSA prevalence. In light of the increasing prevalence of multi drug resistance including vancomycin resistance we also assessed antibiotic susceptibilities.

Methods

The 146 MRSA strains collected over two years (March 2012 until February 2014) at the University Hospital Zurich, Switzerland, were analyzed by PFGE analysis of SmaI digests in combination with spa-typing. In addition, representative isolates were analyzed by multi locus sequence typing (MLST). Susceptibilities to eight antibiotics were assessed using the Kirby-Bauer disc diffusion method.

Results

Isolates showed resistance to erythromycin (48%), ciprofloxacin (43%), clindamycin (31%), tetracycline (22%), and gentamicin (16%). All isolates were susceptible to vancomycin, 95% were susceptible to sulfamethoxazole/trimethoprim and rifampicin, respectively. PFGE analysis revealed 22 different patterns, with four major patterns that accounted for 53.4% of all MRSA isolates, and seven sporadic patterns. Spa typing revealed 50 different spa types with the predominant types being t008 (14%), t002 (10%), and t127 (9%). 82% of the MRSA isolates could be assigned to six clonal complexes (CCs) namely CC1 (10%), CC5 (23%), CC8 (18%), CC22 (17%), CC30 (11%), and CC45 (3%) based on spa-types, PFGE patterns, and MLST. Two isolates could not be typed by either PFGE analysis or spa-typing and three isolates had spa-types that have not yet been described.

Conclusions

The combination of the two typing methods was more discriminatory as compared to the use of a single method. Several of the lineages that are predominant in Europe are present in our hospital. Resistances to antibiotics have decreased in comparison to a study conducted between 2004 and 2006.

Electronic supplementary material

The online version of this article (doi:10.1186/s12941-015-0075-3) contains supplementary material, which is available to authorized users.