Genomic analysis revealed a novel genotype of methicillin-susceptible Staphylococcus aureus isolated from a fatal sepsis case in dengue patient

Staphylococcus aureus (S. aureus) is an opportunistic pathogen capable of causing serious health implications in susceptible individuals once it invades the host’s protective barriers. Methicillin-susceptible S. aureus (MSSA) often receives lesser attention although it has been frequently associated with serious infections in human. We aim to investigate the genomic features of a highly virulent yet pan susceptible MSSA strain (coded as HS-MSSA) which caused concurrent bacteraemia in a dengue patient, ultimately resulted in sepsis death of the patient. Whole genome sequence analysis was performed. The draft genome of HS-MSSA is approximately 2.78 Mb (GC content = 32.7%) comprising of 2637 predicted coding sequences. In silico genotyping of the HS-MSSA strain revealed a novel combined genotype (t091/ST2990). The HS-MSSA carries a SaPIn1-like pathogenicity island that harbours the staphylococcal enterotoxin and enterotoxin-like genes (sec3 and selL). The strain-specific β-lactamase (blaZ)-bearing plasmid region was identified in HS-MSSA. Core genome phylogeny showed that the HS-MSSA strain shared a common ancestry with the European MRSA clone. We report herein the genomic features of an MSSA lineage with novel genotype previously not reported elsewhere.

Whole genome sequencing, assembly, and annotation. The genomic DNA of the HS-MSSA strain was extracted using DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany) and the quality of the extracted genomic DNA was evaluated based on spectrophotometric measurements 20 . The HS-MSSA genomic DNA was subsequently subjected to next-generation sequencing via Illumina Genome Analyzer IIx (GA2x, pipeline version 1.80) by a commercial sequencing vendor. Paired-end 2 × 150 bp reads were generated using Illumina chemistry. The quality of the sequence reads was assessed and assembled de novo using the CLC Genomics Workbench version 5.1 (CLC Bio, Aarhus, Denmark) with default parameters. Contigs with an average coverage of 30 and above were selected to represent the bacterial draft genome for further analyses. Reordering of the HS-MSSA contigs was achieved via genomic mapping with the reference genome S. aureus subspecies aureus NCTC 8325 (National Center for Biotechnology Information (NCBI) GenBank accession number: NC007795) by using Mauve software version 2.4.0 21 . The subset of the contigs that were mapped to the complete chromosome of the S. aureus NCTC 8325 was identified as the chromosomal region of the HS-MSSA strain. Genomic annotation was performed using the webserver Rapid Annotations using Subsystems Technology (RAST; http://rast.nmpdr .org/) 22 . The viewing and preliminary sequence-based comparison between the annotated HS-MSSA and reference genomes were done using SEED Viewer version 2.0 (http://pubse ed.these ed.org/) 23 , and further validated by using Mauve.
Whole genome sequence analyses. The chromosomal region of the HS-MSSA was subjected to a microbial genome search using the NCBI Basic Local Alignment Search Tool (BLAST; https ://blast .ncbi.nlm. nih.gov/Blast .cgi) to identify genetically similar S. aureus strains (taxonomy ID: 46170) from the complete and draft genomes database curated in the NCBI GenBank. The BLAST search was performed using the discontinuous megablast algorithm with default parameters. Multiple genomes alignment and visualization were performed using Mauve and BLAST Ring Image Generator (BRIG) software 24 . The genotype of the HS-MSSA strain was inferred in-silico by using web-based servers. The conventional seven-gene multi-locus sequence typing (MLST), whole genome MLST (wgMLST), and ribosomal MLST (rMLST) predictions made use of the S. aureus database in PubMLST (https ://pubml st.org/saure us/) sited at the University of Oxford 25 . The online analysis tools available from the Center for Genomic Epidemiology (CGE) server (http://www.genom icepi demio logy. org/) were used to predict the staphylococcal protein A (spa) type, presence of SCCmec elements, plasmid elements, and pathogenicity genes in the HS-MSSA genome [26][27][28][29] . Phage-associated genes and genomic regions in the HS-MSSA genome was identified using PHAge Search Tool Enhanced Release server (PHASTER; http:// phast er.ca/) 30,31 . The genomic islands were predicted using IslandViewer 4 (http://www.patho genom ics.sfu.ca/ islan dview er/) 32 . All genomic predictions were completed using default parameters, validated by cross-examination with RAST-annotation output, and manually interrogated using NCBI BLAST nucleotide search tool.
AMR genes prediction. The online tool, Resistance Genes Identifier version 5.1.0 (RGI; https ://card.mcmas ter.ca/analy ze/rgi) was used to identify AMR genes in the assembled HS-MSSA genome 33 . The AMR genes selection criteria were set to perfect (100% identity) and strict (> 95% identity) hits only to the curated reference sequences in the Comprehensive Antibiotic Resistance Database (CARD; https ://card.mcmas ter.ca/). Prediction of partial genes was accepted as the HS-MSSA genome examined in this study resembles an incomplete genome. The RGI predictions were cross-checked with ResFinder version 3.2 (https ://cge.cbs.dtu.dk/servi ces/ ResFi nder/). AMR genes prediction results obtained using default parameters 34 . All predicted AMR genes were validated by cross-examination with RAST-annotation output and manually interrogated using NCBI BLAST.  37 . 100 bootstrap replicates were performed to support the phylogenetic inferences made in this study for practical reasons. A bootstrap analysis with 100-500 replicates is sufficient to produce qualitatively comparable support values with more than 99.5% correlation with the reference values in ML trees with higher numbers of bootstrap replicates 38 .

7-loci MLST-based phylogenetic analysis.
The concatenated allelic sequences of the seven housekeeping genes (arcC, aroE, glpF, gmk, pta, tpi, and yqiL) for the central STs of the major clonal complexes (CC1, CC5, CC8, CC15, and CC97) and other frequently encountered clonal complexes (CC22, CC30, CC45, and CC93) were used to construct a phylogenetic tree. Multiple alignments of the concatenated sequences were performed by Molecular Evolutionary Genetics Analysis (MEGA) X 39 . Next, the aligned sequences were subjected to a statistical selection of best-fit phylogenetic model using jModelTest version 2.1.6 40 . An unrooted tree was subsequently inferred via the ML method using MEGA X software. www.nature.com/scientificreports/ Malaysia (Fig. 1). The HS-MSSA strain was seen sharing more common regions of genetic variation with the MRSA clones than with the local MSSA strain at different sites on the genome. The great genomic resemblance between the HS-MSSA strain and the MRSA clones from the American continent indicates the possible common ancestry between these lineages, despite the wide geographical distance where the strains were originated. The MRSA clones originated from the USA that showed high genetic homology with HS-MSSA were all community acquired, capable of causing SSTIs and sepsis, and were largely resistant to chlorhexidine [41][42][43][44] . The South American MRSA strain (USA300_SUR1) that caused an outbreak of SSTIs shared common MLST type (ST8) with the community acquired USA300 clone from Texas, USA 43 . The complete genome analysis revealed that the South American outbreak strains were genetic variants of the North American USA300 clone 45 . Previous genomic studies have proven that USA300 clones are clonally related to the European nosocomial MRSA strain COL (NC_002951) emerged from a methicillin-susceptible ancestral strain via the acquisition of methicillin resistance determinant following the introduction of methicillin in 1959 43,46 . Therefore, the MRSA clones USA300 and COL are most likely sharing a common MSSA ancestry but evolved separately through different MRSA lineages 47 . Accordingly, the close genomic resemblance of the HS-MSSA strain to these MRSA clones suggests that this MSSA lineage in Malaysia could have been the less evolved (with reference to AMR) descendant of the common MSSA ancestry with the MRSA clones in the USA and Europe. This inference was made on the basis that chromosomal recombination occurs in S. aureus at low frequencies, except for regions associated with virulence and pathogenicity 48 . The descendants of this MSSA ancestral strain have spread widely over different continents and persisted in the human populations over the years, causing community acquired SSTIs and invasive infections in the susceptible hosts. www.nature.com/scientificreports/ Presence of strain-specific blaZ-bearing plasmid in HS-MSSA. Detailed examination of the 15 contigs that were not included in the HS-MSSA chromosomal region revealed a 19,763 bp region (contig-33) identified as a plasmid-associated genomic region. BLAST analysis of contig-33 revealed the closest plasmid homolog as pl1_M2024 (CP047022), at 100% query coverage with 99% nucleotide identity (E-value = 0). The pl1_M2024 is a complete plasmid retrieved from the nosocomial MRSA clone with spa type t304 and MLST type ST6 (clone t304/ST6) which caused recurrent outbreaks in a hospital in Denmark since 2011 49 . This MRSA clone t304/ST6 was detected and predominated among the MRSA strains isolated in a teaching hospital in Oman within the same period 50 . However, this clone remains rarely reported outside these two geographical regions. The finding of similar plasmids in the HS-MSSA strain in Malaysia and MRSA clone in European region suggests possible horizontal gene transfer event, which could have occurred due to chance events after the divergence from the common MSSA ancestral strain. One other hypothesis is that the HS-MSSA strain could have evolved from an MSSA lineage carrying this plasmid, of which some of its descendants might have diverged to another evolutionary path via the acquisition of SCCmec gene cassette. The latter hypothesis is more plausible as this plasmid carries the penicillin-resistant blaZ gene. Based on an earlier study using Staphylococcus species mainly from the bovine origin, the authors demonstrated that plasmid-borne blaZ is often species-specific and lack conjugative transfer mechanisms, thus making horizontal gene transfer an extremely rare event 51 . Therefore, the presence of this blaZ-bearing plasmid in the HS-MSSA strain supports our inference on the common ancestry of this local MSSA with the MRSA clones in European regions.

Results and discussion
In silico genotyping of HS-MSSA revealed novel S. aureus genotype t091/ST2990. In silico MLST analyses showed that HS-MSSA belongs to the sequence type ST2990, with its identity confirmed as S. aureus by rMLST. The 7-loci allelic profile for HS-MSSA is 1-1-1-1-330-1-10, arranged in the order arcC-aroE-glpF-gmk-pta-tpi-yqiL. All loci MLST (wgMLST) results showed 1358 exact matches to sequence type ST2990 (Supplementary Table S2). This sequence type (ST) belongs to MSLT clonal complex CC1. The S. aureus strain 128A deposited in the database (PubMLST ID: 5521) shares the same sequence type as HS-MSSA and was also isolated in Malaysia, earlier in 2014. S. aureus 128A was isolated from the infected wound of an elderly diabetic patient (66 years old) examined in a tertiary hospital in the capital Kuala Lumpur. The strain was found susceptible to oxacillin, methicillin, and vancomycin upon investigation. The S. aureus 128A was confirmed as community acquired MSSA which causes wound infection. The spa type of HS-MSSA is t091 (spa repeat succession 07-23-21-17-34-12-23-02-12-23). Therefore, the combined genotype of the HS-MSSA strain is noted as t091/ ST2990 (spa type/MLST type). Similar to the S. aureus strain 128A which shares the identical sequence type (ST2990), the HS-MSSA strain examined in this study was isolated from a diabetic patient who was admitted to the hospital due to dengue fever 19 . However, instead of causing SSTI (infected diabetic wound) as strain 128A did, the HS-MSSA was more invasive and caused a bloodstream infection that resulted in severe sepsis. The common STs of S. aureus in Malaysia, especially the most studied MRSA strains, mainly comprised of the pandemic clone ST239, with ST22 clone being increasingly documented in the past decade [52][53][54] . Similar to the global epidemiology of MSSA, the few studies that examined the MSSA genotypes in Malaysia documented a higher genetic diversity compared to the MRSA populations in this region [55][56][57] . MRSA with spa type t091 has been reported in Malaysia, but S. aureus strain with genotype t091/ST2990 has not been previously documented 58 . Nonetheless, the isolation of ST2990 strains from different geographical regions suggests that this MSSA genotype may already be widely disseminated in Malaysia. Although S. aureus of CC1 complex (including ST1, ST188, ST2990, etc.) does not predominate in Malaysia, the persistence of these genotypes throughout the years may indicate their evolutionary success 55,57,58 . Absence of staphylococcal chromosome cassette mec elements throughout the evolutionary process of HS-MSSA. The staphylococcal chromosome cassette mec (SCCmec) elements were not detected in the HS-MSSA genome based on SCCmecFinder analysis at 90% identity threshold and a minimum length of 60%. The absence of SCCmec elements in the HS-MSSA genome is reflected in the methicillin-susceptible phenotype of this strain. The SCCmec genomic cassette is a mobile genomic region with two important gene complexes, mec-gene complex encoding methicillin resistance and cassette chromosome recombinase (ccr)-gene complex encoding site-specific recombinase(s) for the mobility of this DNA region 59 . The presence of SCCmec region in the bacterial chromosome is a signature of MRSA strains and the acquisition of which results in the emergence of new MRSA lineage from an ancestral MSSA strain.
While the acquisition of SCCmec elements provides a selective advantage to S. aureus, it is not uncommon for it to spontaneously excised from the MRSA in human hosts 60 . Partial excision is sufficient to result in methicillinsusceptible phenotype, and most often such MSSA strains showed great genetic homology to its MRSA counterpart, although less prevalent 60,61 . Given the genomic resemblance of the HS-MSSA strain to the global MRSA clones, we decided to investigate whether such an excision event has occurred in this local MSSA strain. In order to identify possible remnants of SCCmec cassette or its homologs in the genome, the database search was repeated at 60% identity threshold with 30% minimum query cover length. The search result remains the same, i.e. intact or partial mec genes and the whole cassette was not detected, but a 1.5 kb DNA region associated with the insertion sequence-like IS1272 in SCCmec type I was detected in contig-4, albeit with considerable genetic variation (94% query coverage and 82% nucleotide identity). BLAST analysis of this 1.5 kb region revealed the presence of a pseudogene of IS1182 family transposase in S. aureus (CP039848; 98% query coverage, 99% nucleotide identity, E value = 0). In short, none of the SCCmec elements was detected except for the 1.5 kb DNA region homologous to IS1272 commonly found in the mec-gene complex in MRSA that may also present infrequently in MSSA 62 .  19 . Genomic sequence analysis revealed a low number of AMR-conferring genes compared to virulence genes, and a majority of which are involved in bacterial efflux systems (Table 2). Partial genes were not identified in the HS-MSSA genome.
The pan-susceptibility of the HS-MSSA strain to antimicrobial agents is reflected in its genomic composition. Although multiple efflux pumps and its regulatory systems involve in antimicrobial agents and biocides resistance are present in the HS-MSSA genome, it is not surprising to observe a susceptible phenotype in the strain. Similar to the Gram-negative bacteria, efflux pumps in Gram-positive organisms often confer only low-level resistance and are only effective when overexpressed due to accumulation of mutations in response to environmental stresses 63 . A recent study has reported an interesting phenomenon, whereby the prevalence of certain efflux pump genes was found to be associated with the genotypes of S. aureus 64 . The presence of norA and mepA genes in the HS-MSSA strain is typical of S. aureus lineages in the Asian region, while norB or mdeA is common among European S. aureus 64 . Unlike plasmid-borne genes, these efflux pump genes are chromosomally located hence is fairly conserved within the S. aureus lineages. The close genetic proximity of the HS-MSSA strain carrying norA and mepA genes to the European MRSA lineages suggests that the divergence of these two S. aureus lineages might have occurred early in the evolutionary timeline, followed by local adaptations in the different geographical regions.
The pathogenicity genes analysis showed that the HS-MSSA harbours 536 CDSs which matched to known S. aureus pathogenicity families (with 59% uncharacterized or hypothetical proteins). Cross-examination with virulence genes analysis showed that the HS-MSSA strain carries multiple genes encoding for bacterial adherence, enzyme production, host immune evasion, various prominent staphylococcal toxins, a type VII secretion system, and intact regulatory operons for virulence expression (global accessory gene regulator operon, agrABCD; staphylococcal accessory regulator, sar). The virulence analysis of the HS-MSSA strain is reported in detail in our earlier publication 19 (Supplementary Table S3). Majority of the prominent toxin-producing genes are found associated with the highly variable genomic regions in the HS-MSSA genome. The HS-MSSA resembles a community acquired, highly virulent but pan-susceptible S. aureus strain that was capable of causing an invasive bloodstream infection in the human host that resulted in fatal sepsis. It is a common notion that the MSSA lineages that cause human infections often harbour a wider array of virulence genes compared to the MRSA 16 . The combined effect of multiple virulent factors present in the MSSA genomes has resulted in greater severity of infection, especially one that causes invasive infection 17 . This notion is experimentally proven in a study that examined the transcriptional changes in an MRSA strain when exposed to different antibiotics 65 . In this study, Choe and colleagues proved that the expression of the β-lactam resistance (Bla/Mec system) in the MRSA represses the transcription of the agr operon, thus halting the production of the virulence factors that are regulated by this operon. Therefore, the ability of the HS-MSSA to cause severe sepsis in the human host could be related to the lack of SCCmec genes cassette in its genome, thus the expression of its multiple virulence factors was not impeded.  www.nature.com/scientificreports/ genomic region ( Table 3). The PR1 was identified on contig-13 of the HS-MSSA genome which mainly comprised of staphylococcal toxin-production genes that encode for exotoxins and the superantigen enterotoxins (SEC3 and SElL). Genome mapping showed that this genomic region is not present in the staphylococcal reference genome NCTC 8325, thus resembling a genomic insertion (Fig. 2). The intact prophage PR2 was located on contig-4 and found associated with genes encoding for staphylococcal Panton-Valentine leukocidin (lukFS-PV). When mapped to the reference genome S. aureus NCTC 8325, the PR2 showed great genetic similarity with a 72 kb prophage on NCTC 8325. The PR3 was identified on contig-11, closely associated with genes encoding for β-hemolysin (hlb), cytolytic pore-forming protein (lukGH), extracellular adherence protein (eap), and staphylococcal complement inhibitor (scn). A total of six genomic islands (GI1-GI6) were identified in the HS-MSSA chromosome by using S. aureus NCTC 8325 as the reference (Fig. 2). The genomic size, contigs involved, and the types of genes identified in the genomic islands (GIs) are listed in Table 4. Generally, the regions encoding GI1-GI2, GI5, and GI6 correspond with the phage regions PR1, PR2, and PR3, respectively. The genomic region that spanned GI1 and GI2 resembles a staphylococcal pathogenicity island (SaPI) harbouring the enterotoxin-encoding genes sec3 and selL similar to the SaPIn1 in the MRSA strain N315 (BA000018, previously AP003129) 66,67 . An array of staphylococcal virulence genes mainly associated with bacterial adherence and stress response were found in GI4, while GI5 consisted of mainly hypothetical proteins. Majority of the genomic islands identified in the HS-MSSA genome are associated with bacterial virulence and pathogenicity. Genomic islands are specific chromosomal DNA regions that include genes conferring a survival advantage to the bacterial host, such as virulence and resistance 59 . The association of the genomic islands, prophages, and SaPI observed in the HS-MSSA genome is not uncommon as these genetic entities are often Table 3. Phage-associated regions in the chromosome of the HS-MSSA strain. a Identified prophages were over multiple contigs that were concatenated for PHASTER analysis. b Phage identity was calculated based on the percentage of phage-associated genes over the total protein-coding sequences identified in the phage region. c A prediction of whether the phage region contains genes necessary for lysogeny.

Region
Region length (kb) Region position a (bp) Ordered contig numbers GC % PHASTER annotation Phage identity b (%) PHASTER prediction c   Table 4. Genomic islands identified in the HS-MSSA chromosome. a Identified genomic islands were over multiple contigs that were concatenated for IslandViewer4 analysis.  66 . Moreover, prophages and pathogenicity islands are mobile elements in the staphylococcal chromosome that constitutes the genetic diversity among different S. aureus lineages. The prevalence of specific staphylococcal prophages was shown associated with the clonal background of S. aureus, i.e. different lineages may harbour different types of prophage 68 . Meanwhile, the SaPIs actively involved in the pathogenic evolution of S. aureus via the lateral transfer of virulence genes 69 , hence a greater genetic homology in SaPIs may indicate a closer phylogenetic relationship. Therefore, the sharing of genetically similar SaPIs between HS-MSSA and MRSA N315, and common phage regions between HS-MSSA and NCTC 8325 suggest a clonal relationship, or more probably, the sharing of a recent common ancestor that acquired the common phage or SaPI regions prior to evolutionary divergence.

Region Region position a (bp) Region length (bp) Ordered contig numbers Gene products
Phylogenetic analysis revealed common ancestry of HS-MSSA with MRSA clones originated from European regions. Representative S. aureus complete genomes were selected for cgSNP-based phylogenetic analysis to elucidate the probable evolutionary origin of HS-MSSA strain. The reference genomes were selected based on greatest nucleotide identity in BLAST analysis (MRSA strain 2148.C01 and USA300 clones), sharing of similar mobile genomic regions (M2024, NCTC 8325 and N315), same geographical origin (KT/314250), same clonal complex (MSSA476), and representative MSSA and MRSA genomes (MSSA476 and COL) from European region (potential common ancestry). The HS-MSSA genome showed ≥ 94% genome coverage and ≥ 99.16% nucleotide identity (E value = 0) with all selected S. aureus genomes. The ST22 reference genome HO 5096 0412 (NC_017763) from a clonal complex (CC22) less closely related to other S. aureus strains was used as an outgroup in constructing the ML tree. The phylogenetic analysis results revealed that the HS-MSSA strain formed a distinctive clade by itself, sharing a more recent common ancestry with the MRSA clones in Europe and Japan (M2024 and N315) (Fig. 3). The MRSA strain N315 isolated from the pharyngeal smear of a Japanese patient in 1982 belongs to the clonotype II-A which is prevalent in Japan and USA 67 . Together with another MRSA strain M2024 (NZ_CP047021) which caused recurrent clonal outbreaks in Denmark 49 , the three S. aureus strains formed a loose cluster with relatively divergent evolutionary paths and geographical origins (Cluster I). Interestingly, the MSSA476 from the United Kingdom that shares a common clonal complex (CC1) with HS-MSSA formed another distinctive cluster (II) although both are community acquired MSSA capable of causing bacteraemia in human hosts and carried the strain-specific plasmid-borne blaZ gene 70 . Similarly, the MSSA strain KT/314250 which was also isolated in Malaysia formed a separate clade distinctive from the HS-MSSA lineage but is tightly clustered with the MSSA476 strain. The representative S. aureus strains from Europe (NCTC 8325 and COL) and MRSA clones from America formed a third cluster (III) more distantly related to the HS-MSSA strain.
The phylogenetic relationships of the HS-MSSA strain with the global S. aureus strains as predicted by the core genome SNPs analysis supported our inference on the HS-MSSA lineage. The sharing of a relatively recent common ancestry with the European-originated MRSA clone M2024 suggests the existence of an ancestral MSSA strain which, in turn, shares a common ancestry with the MSSA476 lineage in Europe. The descendants of this MSSA lineage had since taken separate evolutionary paths, developing antimicrobial resistance or virulence as an   [55][56][57] . Unlike most of the MSSA strains that have circulated in this region, the MSSA strain 128A that shares a common MLST genotype (ST2990) with HS-MSSA only appeared on scientific records in 2014. Based on the limited strain data available in the PubMLST database for the MSSA strain 128A, we hypothesize that the ST2990 strains in Malaysia showed variable clinical manifestations, ranging from non-invasive SSTI to severe bacteraemia. The HS-MSSA lineage has probably accumulated sufficient adaptive features through the acquisition of mobile genetic elements, giving rise to its unusual virulence in causing severe invasive infection in human host 19 . Moreover, the close genetic proximity with the MRSA clones and its genome plasticity also suggests that the HS-MSSA lineage could provide a stable genetic environment to allow the acquisition of genetic elements that confer drug resistance 6 . Therefore, continuous molecular surveillance to detect highly virulent MSSA in clinical settings is essential to allow timely management of invasive staphylococcal infections and the development of potential anti-virulence therapy.
HS-MSSA (CC1/ST2990) showed a closer phylogenetic relationship with CC5 than CC1 based on cgSNP analysis. Upon inspection of the MLST clonal complexes of the global S. aureus strains selected for cgSNP-based phylogenetic analysis, we observed that the HS-MSSA strain genotyped as CC1/ST2990 was not found on the same clade as other CC1 strains. In order to investigate the accuracy of this phylogenetic clustering, representative genomes of the central STs for the major CCs were retrieved from the PubMLST database for further analysis 71 . Overall, the clustering of the STs corresponded with the CC grouping based on the 7-loci MLST scheme (Fig. 4). However, the HS-MSSA formed a distinctive clade sharing a common node with the CC5/ST5 clade (100% bootstrap support) instead of being grouped in the CC1 cluster. On the contrary, the phylogenetic analysis based on the concatenated sequences of the representative STs revealed that the HS-MSSA indeed showed a closer phylogenetic relationship with CC1/ST1 and CC8/ST8 (92% bootstrap support) (Supplementary Fig. S1). Nonetheless, the multiple genomes alignment of HS-MSSA and the representative CC1/ ST1 genomes MSSA476 and ERR410082 (PubMLST ID: 5965) showed that multiple genomic insertions were observed in both variable (PR1-PR3) and conserved regions of the HS-MSSA genome ( Supplementary Fig. S2). This could have contributed to the greater phylogenetic divergent of HS-MSSA from other CC1 strains based on cgSNP analysis. Our observations support the previous notion that the current grouping of S. aureus strains into clonal complexes based on the 7-loci MLST scheme requires reconsideration to better discriminate the strains 71 . This observation is not unexpected as cgSNP-based analysis infers the phylogenetic relationship between organisms based on thousands of core genes, hence providing a greater resolution than the 7-loci MLST that examines a relatively small portion of the genome. Nonetheless, current advancement in next-generation sequencing The multiple genome sequence alignments of 17 S. aureus strains were generated by the online tool REALPHY using S. aureus NCTC 8325 as the reference genome. An unrooted phylogenomic tree was inferred by approximately Maximum Likelihood method using Generalized Time-Reversible (GTR) model with gamma distribution rates and 100 bootstrap replicates. Bootstrap support value is indicated as a percentage at each node. The representative genomes for each clonal complex are denoted as clonal complex/sequence type (i.e. CC1/ST1 represents MLST sequence type 1 in clonal complex 1). The PubMLST ID and European Nucleotide Archive (ENA) reference numbers of the representative genomes are available in Supplementary Table S1 www.nature.com/scientificreports/ technology has improved the resolution of MLST-based phylogeny. The core genome MLST (cgMLST) which involves the assessment of 1351 loci (for S. aureus) and the whole genome MLST (wgMLST) are the more appropriate approaches for inferring phylogenetic relationship of genetically clonal strains. Previous studies using different bacterial species have observed a high concordance between cgMLST and wgMLST with SNP-based analysis, generating phylogenetic trees with great statistical similarity and almost equivalent clustering of bacterial strains 72,73 . Therefore, the current clustering of ST2990 into CC1 needs to be re-assessed by the scientific community. Nonetheless, more ST2990 genomes should be included for cgSNP-based analysis in the future to verify this notion. The present study is limited by only one ST2990 (HS-MSSA) genome that was available on publicly accessible genomic databases at the time of writing.

Conclusion
We report herein a novel genotype of MSSA (t091/ST2990) of community origin that caused severe sepsis in the human host. Upon examination, the HS-MSSA genome harbours only very few AMR genes, which was reflected in its pan-susceptible phenotype. The HS-MSSA strain carries multiple virulence genes in its genome, including prominent staphylococcal toxin genes that are associated with mobile genetic elements (prophages and SaPI). The HS-MSSA strain shares a common ancestry with the MRSA clone originated from the European region, but instead of developing drug resistance, it has evolved through a different path via the acquisition of multiple virulence and pathogenicity factors. Due to the limited epidemiological data available for this novel genotype (t091/ST2990), we could not conclude that this MSSA lineage has emerged locally or the HS-MSSA strain was just a sporadic strain imported from other regions without molecular typing data. Although the patient reported no history of recent travels upon admission, we cannot rule out the possibility of local transmission. Therefore, increased effort in molecular surveillance is essential to allow the development of appropriate treatment strategies to reduce the mortality risk of severe invasive infections caused by this organism.

Data availability
All data generated or analysed during this study are included in this published article. The genomic data of the HS-MSSA strain generated and analysed during the current study are available in the NCBI GenBank database at https ://www.ncbi.nlm.nih.gov/genom e/, under the accession VCMW00000000.