RESEARCH ARTICLE Earls et al., Microbial Genomics 2021;7:000601 DOI 10.1099/mgen.0.000601 OPEN OPEN DATA ACCESS Exploring the evolution and epidemiology of European CC1-­ MRSA-­IV: tracking a multidrug-­resistant community-­associated meticillin-­resistant Staphylococcus aureus clone Megan R. Earls1†, Eike J. Steinig2†, Stefan Monecke3,4,5†, José A. Samaniego Castruita6, Alexandra Simbeck7, Wulf Schneider-­Brachert7, Teodora Vremerǎ8, Olivia S. Dorneanu8, Igor Loncaric9, Michèle Bes10, Alicia Lacoma11, Cristina Prat Aymerich11,12, Ulrich Wernery13, Marc Armengol-­Porta14, Anita Blomfeldt15, Sebastian Duchene16, Mette D. Bartels6, Ralf Ehricht3,5,17 and David C. Coleman1,* Abstract This study investigated the evolution and epidemiology of the community-­associated and multidrug-­resistant Staphylococcus aureus clone European CC1-­MRSA-­IV. Whole-­genome sequences were obtained for 194 European CC1-­MRSA-­IV isolates (189 of human and 5 of animal origin) from 12 countries, and 10 meticillin-­susceptible precursors (from North-­Eastern Romania; all of human origin) of the clone. Phylogenetic analysis was performed using a maximum-­likelihood approach, a time-­measured phy- logeny was reconstructed using Bayesian analysis, and in silico microarray genotyping was performed to identify resistance, virulence-­associated and SCCmec (staphylococcal cassette chromosome mec) genes. Isolates were typically sequence type 1 (190/204) and spa type t127 (183/204). Bayesian analysis indicated that European CC1-­MRSA-­IV emerged in approximately 1995 before undergoing rapid expansion in the late 1990s and 2000s, while spreading throughout Europe and into the Middle East. Phylogenetic analysis revealed an unstructured meticillin-­resistant S. aureus (MRSA) population, lacking significant geo- graphical or temporal clusters. The MRSA were genotypically multidrug-­resistant, consistently encoded seh, and intermittently (34/194) encoded an undisrupted hlb gene with concomitant absence of the lysogenic phage-­encoded genes sak and scn. All MRSA also harboured a characteristic ~5350 nt insertion in SCCmec adjacent to orfX. Detailed demographic data from Denmark showed that there, the clone is typically (25/35) found in the community, and often (10/35) among individuals with links to South-­Eastern Europe. This study elucidated the evolution and epidemiology of European CC1-­MRSA-­IV, which emerged from a meticillin-­susceptible lineage prevalent in North-­Eastern Romania before disseminating rapidly throughout Europe. Received 29 October 2020; Accepted 05 May 2021; Published 05 July 2021 Author affiliations: 1Microbiology Research Unit, Division of Oral Biosciences, Dublin Dental University Hospital, Trinity College, University of Dublin, Dublin, Ireland; 2Australian Institute of Tropical Health and Medicine, Townsville, Queensland, Australia; 3Leibniz Institute of Photonic Technology (IPHT), Jena, Germany; 4Institute for Medical Microbiology and Hygiene, Medical Faculty "Carl Gustav Carus", Technische Universität Dresden, Dresden, Germany; 5InfectoGnostics Research Campus Jena, Jena, Germany; 6Department of Clinical Microbiology, Copenhagen University Hospital, Hvidovre, Denmark; 7Institute for Clinical Microbiology and Hygiene, University Hospital Regensburg, Regensburg, Germany; 8Microbiology Unit, Department of Preventive and Interdisciplinary Medicine, University of Medicine and Pharmacy "Grigore T Popa", Iaşi, Romania; 9Institute of Microbiology, Department for Pathobiology, University of Veterinary Medicine Vienna, Vienna, Austria; 10Centre National de Référence des Staphylocoques, Institut des Agents Infectieux, Hospices Civils de Lyon, Lyon, France; 11Microbiology Department, Hospital Universitari Germans Trias i Pujol, Institut d’Investigació Germans Trias i Pujol, CIBER Enfermedades Respiratorias, Universitat Autònoma de Barcelona, Badalona, Spain; 12Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands; 13Central Veterinary Research Laboratory, Dubai, UAE; 14 Microbiology Department, Dresden Municipal Hospital, Dresden, Germany; 15Department of Microbiology and Infection Control, Akershus University Hospital, Lørenskog, Norway; 16Peter Doherty Institute for Infection and Immunity, University of Melbourne, Melbourne, Australia; 17Institute of Physical Chemistry, Friedrich-­Schiller University, Jena, Germany. *Correspondence: David C. Coleman, ​david.​coleman@​dental.​tcd.​ie Keywords: CA-­MRSA; European CC1-­MRSA-­IV clone; evolution; epidemiology; phylogenomics; transmission. Abbreviations: CA, community-­associated; CC, clonal complex; HPD, highest posterior density; MRCA, most recent common ancestor; MRSA, meticillin-­resistant Staphylococcus aureus; MSSA, meticillin-­susceptible Staphylococcus aureus; NCBI, National Center for Biotechnology Information; PVL, Panton–Valentine leukocidin; SCCmec, staphylococcal cassette chromosome mec; ST, sequence type. Sequence read sets are available in the NCBI Sequence Read Archive under BioProject accession number PRJNA494507. †These authors contributed equally to this work Data statement: All supporting data, code and protocols have been provided within the article or through supplementary data files. Four supplementary figures and one dataset are available with the online version of this article. 000601 © 2021 The Authors This is an open-­access article distributed under the terms of the Creative Commons Attribution NonCommercial License. This article was made open access via a Publish and Read agreement between the Microbiology Society and the corresponding author’s institution. 1 Earls et al., Microbial Genomics 2021;7:000601 DATA SUMMARY This study generated sequencing data for 85 Staphylococcus Impact Statement aureus isolates and used published data for 122 isolates. All Community-­associated (CA) meticillin-­resistant Staphy- new sequence read sets are available in the National Center for lococcus aureus (MRSA) strains tend to spread in both Biotechnology Information (NCBI) Sequence Read Archive healthcare and community settings, where they continue under the BioProject accession number PRJNA494507. Refer- to pose a serious challenge to public health worldwide. ences and BioSample accession numbers for all 207 isolates Maintaining surveillance of these strains across borders are detailed in supplementary Dataset S1. enhances our understanding of CA-­MRSA, enabling the development of effective strategies to limit its spread and, thus, reducing our dependence on antibiotics. Euro- INTRODUCTION pean CC1-­MRSA-­IV is a multidrug-­resistant and CA-­MRSA Meticillin-­resistant Staphylococcus aureus (MRSA) are a major strain. Previous work has shown that this strain may cause of hospital-­associated and community-­associated (CA) have originated in South-­Eastern Europe, and is now infections worldwide. These infections are caused by clones endemic in certain regions of Germany, Ireland, Italy and that initially emerged in healthcare facilities, the community Romania. Through the whole-­genome analysis of 194 or livestock, but which tend to spread beyond these environ- isolates, mostly from humans but also from animals, ments. The CA-­MRSA population in Europe is remarkably this study showed that European CC1-­MRSA-­IV emerged diverse [1]. Continuous surveillance of this population is from a multidrug-­ resistant but meticillin-­ susceptible required to increase our overall understanding of CA-­MRSA, Staphylococcus aureus strain prevalent in North-­Eastern and to develop effective strategies to limit its spread. Romania, in approximately 1995. In the late 1990s and 2000s, the strain then spread throughout Europe and into We recently described a Panton–Valentine leukocidin (PVL)-­ the Middle East, expanding particularly rapidly in 2004 negative MRSA clone belonging to clonal complex (CC) 1, and 2007/2008, and occasionally acquiring additional which is associated with the emergence of MRSA in the resistance or virulence determinants. Overall, this study community, and harbouring a type IVa staphylococcal cassette chromosome mec (SCCmec) element [2, 3]. We showed that has examined the evolutionary and epidemiological this clone is endemic in Ireland, Romania and the German history of a multidrug-­resistant and CA-­MRSA strain that state of Bavaria, and revealed that it likely emerged from a has received relatively little attention to date despite its meticillin-­susceptible CC1 lineage [2, 3] known to be highly importance in Europe, and significance to both humans prevalent in North-­Eastern Romania [3]. Considering these and animals. findings, we refer to this clone as European CC1-­MRSA-­IV. European CC1-­MRSA-­IV is a multidrug-­resistant clone that typically encodes resistance to penicillin (blaZ), tetracycline characteristic SCCmec insertion has recently been shown to [tet(K)], aminoglycoside-­streptothricin antibiotics (aadE-­ give rise to false-­negative results with a widely used, PCR-­ sat-­aphA3) and MLSB (macrolide, lincosamide and strep- based commercial MRSA identification assay [5]. togramin B) compounds [erm(C)]. Additionally, the clone consistently harbours the enterotoxin gene seh (but usually Considering the above points, we sought to further examine lacks other enterotoxins), and often carries sak and scn, which European CC1-­MRSA-­IV. Specifically, we aimed to investi- are co-­located on lysogenic hlb-­converting bacteriophages gate the evolution and epidemiology of the clone in order to [2–4]. This clone is further characterized by a SCCmec type better understand its expansion and spread. This study has IVa element that harbours an insertion of approximately 5350 built upon our previous work by including a larger number nt in its downstream constant segment (dcs), adjacent to orfX of isolates from more countries, several isolates from non-­ [5]. To date, this clone has been associated predominantly human hosts, and a more advanced and comprehensive with sequence type (ST) 1 and spa type t127 [2]. phylogenetic analysis. Together, these factors have enabled us to estimate the year this clone emerged, and identify the In recent years, European CC1-­ MRSA IV has been dissemination pattern it has followed to date. recognized as a somewhat problematic clone. A variant harbouring plasmid-­encoded iles2/mupA and qacA (confer- ring mupirocin and chlorhexidine resistance, respectively) was responsible for a protracted outbreak spanning multiple METHODS hospitals and involving the community, in Ireland [6]. Furthermore, the clone is common in a children’s hospital Isolates in Italy [7], and its prevalence increased from 1 % in 2010 to Two hundred and four study isolates, and three outgroup 11 % in 2019 in Bavaria, Germany (our unpublished obser- isolates, were investigated. All isolates are described in detail vations). In addition, we have speculated that this clone is in Dataset S1. Species identification and meticillin-­resistance common in the German state of North Rhine-­Westphalia detection was undertaken as described previously [10]. and among livestock in Italy [2, 8, 9]. Finally, the clone’s Isolates were stored at −80 °C on cryogenic beads. 2 Earls et al., Microbial Genomics 2021;7:000601 Study isolates well-­known CC1 clones/lineages. Isolate 124 889 belongs to The study isolates comprised (i) 109 isolates definitively the PVL-­negative CC1-­MRSA-­IV clone known as Western identified previously as European CC1-­MRSA-­IV, (ii) 10 Australian MRSA-1 [2], MW2 belongs to the PVL-­positive meticillin-­susceptible S. aureus (MSSA) isolates previously CC1-­MRSA-­IV clone known as USA400 [13], and MSSA476 identified as precursors of European CC1-­MRSA-­IV, and is a PVL-­negative CC1-­MSSA that was recovered in the UK (iii) 85 MRSA isolates tentatively identified as European [14]. CC1-­MRSA-­IV. The 109 isolates definitively identified previously as European CC1-­MRSA-­IV included 14 isolates Whole-genome sequencing and sequence read investigated by Manara et al. [7], 38 isolates examined by processing Monecke et al. [5] and 57 isolates from our previous study Genomic DNA was extracted from isolates using the S. [2]. While the isolates investigated by Manara et al. and aureus Genotyping Kit 2.0 (Abbott, Alere Technologies), Monecke et al. had been acquired directly from hospital and the Qiagen DNeasy blood and tissue kit (Qiagen). collections, the 57 isolates from our previous study had DNA quality was assessed as previously described [10]. been selected from an international S. aureus microarray The Nextera DNA Flex library preparation kit (Illumina) database. This database comprises genotypic profiles for was used according to the manufacturer’s instructions, over 28 000 MRSA and MSSA isolates recovered from and libraries underwent paired-­end sequencing using the humans and animals worldwide [11]. By analysing these 500-­c ycle MiSeq reagent kit v2 (Illumina). Libraries were profiles, it is possible to determine the clone or lineage to scaled to exhibit at least 50× coverage and the quality of which an isolate belongs [11]. The 57 isolates investigated each sequencing run was assured following cluster density had constituted a representative selection of all European and Q30 assessment. Sequence read sets were submitted to CC1-­MRSA-­IV in the database; they had been recovered in the National Center for Biotechnology Information (NCBI) Germany, Ireland and Romania [2]. Sequence Read Archive (BioProject accession number: The 10 MSSA isolates previously identified as precursors PRJNA494507). of European CC1-­MRSA-­IV had also been selected from For the 119 previously sequenced study isolates (109 the microarray database [2, 3]. Romania had been the only MRSA, 10 MSSA), sequence read sets were acquired from country from which MSSA precursor isolates could be our own collection, or contigs were downloaded from the selected; this lineage had not been detected among the isolates NCBI GenBank (Dataset S1). For outgroup isolates MW2 of any other country represented in the database. and MSSA476, complete genomes were downloaded from The present study included these previously examined the NCBI GenBank (accession numbers: NC_00392 and isolates, but aimed to expand the collection by acquiring NC_002953, respectively). For outgroup isolate 124 889, the isolates from additional countries. Accordingly, 85 addi- sequence read set was acquired from our own collection as, tional MRSA isolates were obtained by requesting that to our knowledge, there are currently no complete Western clinical laboratories or research groups throughout Europe, Australia MRSA-1 genomes publicly available. All sequence and in Saudi Arabia and the UAE, provide isolates (or read sets were trimmed using fastp version 0.20.0 [15]. sequence reads of isolates) matching the description of Trimmed reads were assembled using SPAdes version 3.13.1 the clone. Isolates were required to fit the description of [16], and contigs under 1000 bp were removed. PVL-­negative CC1-­MRSA-­IV, harbour the resistance gene cluster aadE-­sat-­aphA3 and/or belong to spa type t127. A spa typing, MLST and genotypic analysis representative selection (based on year of recovery, recovery Isolates underwent spa typing and multilocus sequence site and/or genotypic variations) of isolates fitting this typing (MLST) using Ridom SeqSphere+ version 7.0 description underwent whole-­genome sequencing at the (Ridom). Genotypic analysis was performed in silico by Dublin Dental University Hospital microbiology labora- mapping sequences to (i) virtual versions of the probes of tory (Ireland). This selection comprised 80 human isolates the S. aureus Genotyping Kit 2.0 (Abbott, Alere Technolo- and 5 animal isolates, from 10 different countries. Four gies) DNA microarray [17], and (ii) the virtual probes of a of the animal isolates were previously described, but not previously described SCCmec subtyping in silico array [18]. sequenced [12]. Other previously described animal isolates The in silico microarray method and the probe sequences that fulfilled the selection criteria were also identified [8]; used for both arrays have been previously described however, these were unavailable for inclusion in this study. [17, 18]. The five animal isolate genomes were also searched As South-­Eastern Europe was underrepresented in our final for additional genes, namely, the animal-­ associated isolate collection, the Staphopia database was searched for enterotoxin A gene variant, sea-­320E (GenBank accession ST1 genomes from this region; however, no such genomes number: AY196686.1), the bovine-­associated leukocidin were identified. genes, lukM/F-­P83 (D83951.1), and the equid-­associated leukocidin genes, lukP/Q (LT671578.1). Any unusual geno- Outgroup isolates typic characteristics identified were quickly investigated The outgroup isolates were 124889, MW2 and MSSA476; using the in silico microarray approach. Alternatively, the these were included in the analysis to represent the other sequence data were examined manually. 3 Earls et al., Microbial Genomics 2021;7:000601 Phylogenetic analysis t127. None of the study isolates harboured PVL-­encoding Sequence alignment and SNP identification were performed genes and all MRSA study isolates harboured SCCmec type using Snippy version 4.4.5 (https://​github.​com/​tseemann/​ IVa. All conventional molecular typing data are detailed in snippy). Default settings were used and MW2 was designated Dataset S1. the reference genome. A maximum-­likelihood tree based on core-­genome SNPs was generated using iq-­tree version Phylogenetic insights into the European CC1-MRSA- 1.6.12 [19]. The general time reversible (GTR) model was used IV clone with an ascertainment bias correction and 1000 bootstraps. An in-­depth phylogenetic analysis revealed significant The phylogenetic tree was visualized using Interactive Tree of details regarding the origin and dissemination of this Life version 5.5.1 [20]. A SNP distance matrix was generated clone. Using a maximum-­likelihood approach, phylogenetic using SNP-­dists (https://​github.​com/​tseemann/​snp-​dists). analysis based on 7066 SNPs grouped all 204 study isolates into a single clade (blue and red branches in Fig. 1). The 10 Time-scaled phylogenetic analysis MSSA precursors of the European CC1-­MRSA-­IV clone Molecular dating analysis of major divergences and the most formed the base of this large clade (blue branches in Fig. 1), recent common ancestor (MRCA) of the lineage was based while the remaining study isolates derived from the base on a recombination-­pruned core-­genome SNP alignment (red branches in Fig. 1). This tree topology confirmed the of a reduced set of 190 study isolates, including all MSSA, identity of all 194 MRSA study isolates as European CC1-­ for which sequence read sets were available. Root-­to-­tip MRSA-­IV. Typically, isolates did not group according to regression analysis and date randomization testing were their country of origin or year of recovery; however, several used to confirm the data had a sufficient temporal signal small country-­specific clusters were evident (Fig. 1). Simi- (Fig. S1, available with the online version of this article) larly, the only two isolates from the Middle East investigated [21]. A time-­scaled phylogenetic tree and effective popula- grouped into the same three-­isolate sub-­clade. Molecular tion size over time (including the age of the MRCA) were clock analysis based on 7330 SNPs indicated that the study estimated in beast-2 under a coalescent Bayesian skyline isolates shared their MRCA around 1987 [95 % highest tree prior and a strict molecular clock model [22]. The GTR posterior density (HPD): 1984–1990) (Fig. 2a, b). The (general time reversible) substitution model with a gamma European CC1-­MRSA-­IV clone then diverged from MSSA distribution for among site rate heterogeneity was used and in approximately 1995 (95 % HPD: 1992–1996) before a Markov chain Monte Carlo analysis of length 500 million spreading throughout Europe, largely in the late 1990s and steps was run; this was performed for various numbers early 2000s (Fig. 2a). Correspondingly, reconstruction of of coalescent intervals of the Bayesian skyline (known as the effective population size over time estimated a sharp dimensions in the model) (see the Supplementary Informa- rise in genetic diversity during this time (particularly in tion and Figs S2 and S3, available withe the online version 2004 and 2007/2008) (Fig. 2c). The genomic clock rate of of this article for details). Effective sample size values for all the clone was estimated at approximately 1.498×10−6 (95 % parameters were over 200 and sufficient chain mixing was HPD: 1.368×10−6–1.626×10−6) substitutions per site per confirmed using Tracer v1.7 [23]. year, when assuming a core genome of 2.1 MB (Fig. 2d). This rate estimate is in line with previous estimates for other S. aureus lineages [24–27]. RESULTS Isolates Epidemiological characteristics of the European The genomes of 204 study isolates and three outgroup CC1-MRSA-IV clone isolates were investigated. The study isolates included 109 The epidemiological data available for each MRSA study known European CC1-­MRSA-­IV and 10 MSSA previously isolate were examined. While the clone had been previ- identified as precursors of this clone, recovered in Germany, ously identified in Ireland, Germany, Romania and Italy, Ireland and Romania [2, 5, 7]. The present study included an expanded search led to its identification in six other these 119 previously investigated isolates, but expanded European countries (Austria, Cyprus, Denmark, France, the collection in order to help elucidate the evolutionary Norway and Spain) and two Middle Eastern countries dynamics of European CC1-­MRSA-­IV. Hence, the study (Saudi Arabia and the UAE). Additional European CC1-­ isolates also included 85 MRSA tentatively identified MRSA-­IV isolates from Germany and Romania were also as European CC1-­MRSA-­IV, recovered in 10 different identified. The oldest study isolate was recovered in 2006, countries. The three outgroup isolates were included in Denmark. Fifty-­five study isolates (26.9 %) were from for comparative purposes, and represented the Western colonization sites, 105 (51.5 %) from infection sites and Australian MRSA-1 clone, USA400 clone and MSSA476 44 (21.6 %) from unknown sites. Fifty-­five (52.4 %) of the lineage. The majority (190/204; 93.1 %) of study isolates infection isolates were from skin and soft tissue infections, were assigned to ST1, with the remaining 14 assigned and 50/105 (47.6 %) from other infections. The majority to 10 different single locus variants of ST1. A total of 14 (199/204; 97.5 %) of study isolates were from humans, while different spa types were identified among the study isolates, 5 were from wild or domesticated animals. Two of these although the majority (182/204; 89.2 %) exhibited spa type animal isolates were from wild rooks (Corvus frugilegus; 4 Earls et al., Microbial Genomics 2021;7:000601 Fig. 1. A maximum-­likelihood phylogeny based on 7066 SNPs. The red branches represent the European CC1-­MRSA-­IV clone, the blue branches represent CC1-­MSSA precursors of European CC1-­MRSA-­IV and the black branches represent CC1 outgroup isolates. The countries and sites of isolate recovery are indicated in the key. The presence or absence of erm(C), tet(K) and sak/scn is indicated by the black or white circles, respectively. Isolate pairs or clusters containing isolates that differed by 10 or less SNPs are shaded in grey. The pink shapes on branch tips indicate that the isolate was recovered from an animal: circle, horse; rectangle, cat; triangle, wild rook. The majority of isolates were ST1 (MLST profile: 1-1-1-1-1-1-1) and spa type t127 (repeats succession: 07-23-21-16-34-33-13); non-­ST1 and/or t127 variants are highlighted. One variant cluster (highlighted in turquoise) comprised one t7096 (14-13), one t398 (26-33-13) and four t174 (14-21-16-34-33-13) isolates. Four of these six isolates were also ST4110 (1-1-1-1-1-1-558). Another variant cluster (highlighted in green) comprised two German isolates, a Saudi Arabian isolate and a UAE isolate, all of which were spa type t386 (07-23- 13). One of these isolates was also ST6001 (1-1-794-1-1-1-1). The remaining variants (highlighted in yellow) did not cluster into groups. Branches with 100 % bootstrap support are indicated with black dots. The tree scale denotes substitutions per site. both in Austria), two from horses (one in Austria; one in the data. The majority (25/35; 71.4 %) of Danish isolates UAE) and one from a cat (in Austria). The extent of meta- were from samples taken by community general medical data available for the study isolates varied; Danish isolates practitioners, while the remaining samples were taken in were associated with particularly detailed demographic eight different hospitals. Thirteen (37.1 %) of the Danish 5 Earls et al., Microbial Genomics 2021;7:000601 Fig. 2. Bayesian analysis of 180 European CC1-­MRSA-­IV isolates and 10 MSSA precursors of the clone. (a) A Bayesian consensus tree based on 7330 non-­recombinant core-­genome SNPs and including 95 % HPD intervals across different dimensional partitioning of population sizes. (b) Posterior density distribution showing tree height (MRCA). (c) Estimated effective population size of the clone over time (dark line) including 95 % HPD intervals (grey shading) denoting uncertainty. (d) Posterior density distribution showing clock-­rate estimates for the core-­genome SNP alignment. isolates were recovered from patients with international Figs 1 and S4). Notably, one European CC1-­MRSA-­IV links, e.g. they were known to have migrated from a specific isolate lacked the aadE-­sat-­aphA3 gene cluster (according country, or had a name associated with a specific country. to both microarray analysis and in silico microarray anal- Nine (25.7 %) of these links were to Romania (including ysis). This unique isolate (2546-­va69252) was recovered in the oldest study isolate), while the remaining four were to Regensburg, Germany, in 2019. Excluding mecA, the MSSA Afghanistan, Macedonia, Nigeria and Ukraine. precursor isolates harboured similar resistance genes to the European CC1-­MRSA-­IV isolates (Table 1). Genotypic analysis of European CC1-MRSA-IV The genotypic profiles of the study isolates were examined to Virulence-associated genes help determine the clone’s resistance and virulence potential, This clone consistently harboured seh, and often harboured identify any significant patterns among the isolates, and aid sak and scn (Table 1). The pattern of variants in the phylogeny identification of the clone. All genotypic data are shown in suggests that the clone has lost sak and scn on many occasions, Fig. S4, specific points are highlighted in Fig. 1 and selected and has acquired sek and seq on three separate occasions frequencies are detailed in Table 1. (Figs 1 and S4). Isolates lacking sak and scn harboured an undisrupted β-haemolysin gene (hlb), indicating the absence Antimicrobial-resistance genes of lysogenic phages that encode combinations of human-­ Genotypic data indicated that this clone is typically resistant specific immune modulator genes (sea, sak, chp, scn). The to β-lactams, aminoglycosides, streptothricin, tetracycline animal-­associated enterotoxin A gene variant, sea-320E, the and MLSB (macrolide, lincosamide and streptogramin B) bovine-­associated leukocidin genes, lukM/F, and the equid-­ compounds (Table 1). As indicated by the sporadic occur- associated leukocidin genes, lukP/Q, were not detected in rence of variants in the phylogeny, the clone has a tendency the animal isolate genomes investigated. The sek and seq to lose erm(C) and tet(K), and gain other resistance genes genes were universally present in the three outgroup isolate such as aacA-­aphD, dfrA, mupA/ileS2 and qacA (Table 1, genomes. The MSSA precursor isolates harboured similar 6 Earls et al., Microbial Genomics 2021;7:000601 Table 1. Prevalence of major antimicrobial-­resistance and virulence-­ DISCUSSION associated genes identified in European CC1-­MRSA-­IV and ten MSSA precursors Using a combination of maximum-­likelihood, time-­scaled phylogenetic, genotypic and epidemiological analyses, this Gene(s) Frequency study elucidated the evolutionary history of the epidemic MRSA (n=194) MSSA (n=10) CA-­MRSA clone known as European CC1-­MRSA-­IV. Phylog- enomic analyses indicated that European CC1-­MRSA-­IV No. % No. % emerged from an MSSA lineage prevalent in Romania, in Resistance approximately 1995. Upon acquiring SCCmec, the clonal population underwent rapid expansion, while spreading aacA-­aphD 7 3.6 1 10 to at least 10 European and 2 Middle Eastern countries aad6 1 0.5 0 0 (Fig. 2a). During this time, it appears that no major temporal aadD 1 0.5 0 0 or geographical variants of the clone evolved within Europe. In fact, the CC1-­MRSA-­IV population investigated lacked aadE-­sat-­ 193 99.5 8 80 any significant structure, resembling a star-­like phylogeny aphA3 (Fig. 1). Although the clone was typically identified as ST1-­ blaZ 194 100 10 100 MRSA-­IV-­t127, variants such as ST4110-­MRSA-­IV-­t174 were cat 2 1 1 10 also identified (Fig. 1), and additional variants may have been overlooked if they did not fit the inclusion criteria during dfrA 3 1.5 1 10 isolate collection. erm(C) 179 92.3 8 80 The phylogenomic data were in line with the available lnuA 2 1 0 0 epidemiological information, previous studies and known human migration patterns. Firstly, the oldest study isolate mph(C) 1 0.5 0 0 was recovered in 2006, shortly after a sharp clonal popula- msrA 1 0.5 0 0 tion increase was first observed according to our estimations mupA 3 1.5 0 0 (Fig. 2c). Secondly, the grouping of the Romanian MSSA isolates at the base of European CC1-­MRSA-­IV clade in both qacA 4 2 0 0 phylogenies (rather than among the outgroup isolates) was in qacC 1 0.5 1 10 agreement with previous work determining that European CC1-­MRSA-­IV may have emerged in South-­Eastern Europe tet(K) 175 90.2 8 80 [2, 3]. Among other evidence, this previous work took into Virulence consideration that (i) the CC1-­MSSA precursor lineage is common in North-­Eastern Romania, and (ii) European CC1-­ sak-­scn 170 87.7 8 80 MRSA-­IV was the predominant clone in a hospital in Iaşi, Undisrupted 24 12.3 2 20 Romania, by 2008, at which point it had been recognized in hlb few other countries [3]. Finally, sharp population increases seh 194 100 10 100 were observed in approximately 2004 and particularly in 2007/2008 (Fig. 2c), corresponding with periods of increased sek-­seq 7 3.6 0 0 migration from Eastern Europe and South-­Eastern Europe, respectively, into the rest of Europe [28]. Certain characteristics of European CC1-­MRSA-­IV resemble virulence genes to the European CC1-­MRSA-­IV isolates those associated with the PVL-­positive ST80-­MRSA-­IV clone (Table 1). known as European CA-­MRSA. For example, the European CA-­MRSA population also appears to lack any significant SCCmec element structure, exhibiting a star-­like phylogeny reminiscent of An SCCmec type IVa element harbouring the mvaS-­SCC gene that described here [24]. However, more structure may be and a characteristic dcs insertion was typically identified in revealed in the future as additional European CC1-­MRSA-­IV isolates of the European CC1-­MRSA-­IV clone. This inser- isolates are identified for analysis. Furthermore, evidence tion encodes six hypothetical proteins and is approximately suggests that European CA-­MRSA derived from a single 5350 nt in size [5]. According to in silico microarray mapping multidrug-­ resistant progenitor [24]. Likewise, genotypic analysis, however, three isolates lacked C5QAP8 (part of the data from this study suggest that a single SCCmec element dcs insertion) and three other isolates lacked mvaS-­SCC. was acquired by an ancestral MSSA that had a resistance-­ Upon further analysis, the C5QAP8 gene was identified split gene and virulence-­gene profile closely matching that of between contigs, for all three isolates. In the case of mvaS- European CC1-­MRSA-­IV (Table 1). This is in contrast to ­SCC, no remnants of the gene were identified among the the well-­studied CA-­MRSA clone, USA300, which does three contig sets and its absence was, therefore, considered not inherently harbour antimicrobial-­resistance genes, but representative of an authentic deletion (Fig. S4). has acquired them over time [29]. Worryingly, however, 7 Earls et al., Microbial Genomics 2021;7:000601 European CC1-­MRSA-­IV can also acquire additional resist- emerging in Ireland and Germany likely originated in South-­ ance over time. Indeed, certain study isolates harboured extra Eastern Europe. Infect Genet Evol 2019;69:117–126. antimicrobial-­resistance genes, such as those encoding resist- 3. Monecke S, Müller E, Dorneanu OS, Vremerǎ T, Ehricht R. Molec- ular typing of MRSA and of clinical Staphylococcus aureus isolates ance to trimethoprim, mupirocin and quaternary ammonium from Iaşi, Romania. PLoS One 2014;9:e97833. compounds (Table 1, Figs 1 and S4). Furthermore, a multihos- 4. Carroll D, Kehoe MA, Cavanagh D, Coleman DC. Novel organization pital outbreak of a mupirocin-­resistant variant of European of the site-­specific integration and excision recombination func- CC1-­MRSA-­IV occurred previously in Ireland [6]. tions of the Staphylococcus aureus serotype F virulence-­converting phages φ13 and φ42. Mol Microbiol 1995;16:877–893. Limitations of this study centre mainly around potential gaps 5. Monecke S, König E, Earls MR, Leitner E, Müller E, et al. An epidemic in our isolate collection. Specifically, it is likely that the MSSA CC1-­MRSA-­IV clone yields false-­negative test results in molecular progenitor lineage exists also outside of Romania. Therefore, MRSA identification assays: a note of caution, Austria, Germany, Ireland, 2020. Euro Surveill 2020;25:pii=2000929. while the phylogenomic evidence indicates that European 6. Earls MR, Kinnevey PM, Brennan GI, Lazaris A, Skally M, et al. CC1-­MRSA-­IV emerged from this MSSA lineage, the country The recent emergence in hospitals of multidrug-­ resistant in which this happened cannot be definitively determined. community-­ associated sequence type 1 and spa type t127 Relevant MRSA isolates may also have been missed. Indeed, methicillin-­resistant Staphylococcus aureus investigated by there is a general lack of MSSA data for Europe, and a lack of whole-­genome sequencing: implications for screening. PLoS One 2017;12:e0175542. MRSA data for many parts of South-­Eastern Europe. Further- 7. Manara S, Pasolli E, Dolce D, Ravenni N, Campana S, et al. Whole-­ more, just five animal isolates were included, although we genome epidemiology, characterisation, and phylogenetic recon- have previously speculated that European CC1-­MRSA-­IV is struction of Staphylococcus aureus strains in a paediatric hospital. well-­established among Italian livestock [2, 8]. Genome Med 2018;10:82. 8. Alba P, Feltrin F, Cordaro G, Porrero MC, Kraushaar B, et al. This study demonstrated that European CC1-­MRSA-­IV Livestock-­associated methicillin resistant and methicillin suscep- spread rapidly throughout Europe in the late 1990s and tible Staphylococcus aureus sequence type (CC)1 in European early 2000s. Ongoing surveillance is essential to minimize farmed animals: high genetic relatedness of isolates from Italian cattle herds and humans. PLoS One 2015;10:e0137143. its further spread. 9. Scheithauer S, Trepels-­Kottek S, Häfner H, Keller D, Ittel T, et al. Healthcare worker-­ related MRSA cluster in a German neona- tology level III ICU: a true European story. Int J Hyg Environ Health Funding information 2014;217:307–311. D. C. C. and M. R. E. were supported by the Microbiology Research 10. Earls MR, Coleman DC, Brennan GI, Fleming T, Monecke S, et al. Unit, Division of Oral Biosciences, Dublin Dental University Hospital, Intra-­hospital, inter-­hospital and intercontinental spread of ST78 University of Dublin, Trinity College Dublin, Ireland. S. M. and R. E. were MRSA from two neonatal intensive care unit outbreaks established supported by the German Federal Ministry of Education and Research, using whole-­genome sequencing. Front Microbiol 2018;9:1485. within the framework of a project (ADA; 13GW0456C) on rapid tests for the detection of MRSA. The funders had no role in study design, data 11. Monecke S, Slickers P, Ehricht R. Assignment of Staphylococcus collection and interpretation, nor the decision to submit the work for aureus isolates to clonal complexes based on microarray analysis and publication. pattern recognition. FEMS Immunol Med Microbiol 2008;53:237–251. 12. Loncaric I, Stalder GL, Mehinagic K, Rosengarten R, Hoelzl F, et al. Acknowledgements Comparison of ESBL and AmpC producing enterobacteriaceae and We thank Peter Slickers at Abbott Rapid Diagnostics, Jena, Germany, methicillin-­resistant Staphylococcus aureus (MRSA) isolated from for performing the in silico microarray analysis. D.C.C. and M.R.E. thank migratory and resident population of rooks (Corvus frugilegus) in Grainne Brennan and the staff of the Irish National MRSA Reference Austria. PLoS One 2013;8:12-e84048. Laboratory (NMRSARL) for technical assistance, and the hospitals in 13. Baba T, Takeuchi F, Kuroda M, Yuzawa H, Aoki K, et al. Genome which Irish isolates were recovered for referring their isolates to the and virulence determinants of high virulence community-­acquired NMRSARL. MRSA. Lancet 2002;359:1819–1827. Author contributions 14. Holden MTG, Feil EJ, Lindsay JA, Peacock SJ, Day NPJ, et al. M. R. E., conceptualization, methodology, formal analysis, writing orig- Complete genomes of two clinical Staphylococcus aureus strains: inal draft preparation. E. J. S., methodology, formal analysis, writing evidence for the rapid evolution of virulence and drug resistance. original draft preparation. S. M., conceptualization, methodology, formal Proc Natl Acad Sci USA 2004;101:9786–9791. analysis, writing original draft preparation. J. A. S. C., formal analysis. 15. Chen S, Zhou Y, Chen Y, Gu J. Fastp: an ultra-­fast all-­in-­one FASTQ A. S., W. S.-B., T. V., O. S. D., I. L., M. B., A. L., C. P. A., U. W., M. A.-P., A. preprocessor. Bioinformatics 2018;34:i884–i890. B. data curation, reviewing and editing of manuscript. S. D., formal analysis, reviewing and editing of manuscript. M. D. B., data curation, 16. Bankevich A, Nurk S, Antipov D, Gurevich AA, Dvorkin M, et al. formal analysis, writing original draft preparation. R. E., conceptualiza- SPAdes: a new genome assembly algorithm and its applications to tion, methodology, formal analysis, writing original draft preparation, single-­cell sequencing. J Comput Biol 2012;19:455–477. supervision. D. C. C., conceptualization, methodology, formal analysis, 17. Monecke S, Slickers P, Gawlik D, Müller E, Reissig A, et al. Molec- writing original draft preparation, supervision, funding. ular typing of ST239-­MRSA-­III from diverse geographic locations and the evolution of the SCCmec III element during its interconti- Conflicts of interest nental spread. Front Microbiol 2018;9:1436. The authors declare that there are no conflicts of interest. 18. Monecke S, Jatzwauk L, Müller E, Nitschke H, Pfohl K, et al. Diver- References sity of SCCmec elements in Staphylococcus aureus as observed in 1. Rolo J, Miragaia M, Turlej-­Rogacka A, Empel J, Bouchami O, et al. South-­Eastern Germany. PLoS One 2016;11:e0162654. High genetic diversity among community-­ associated Staphylo- 19. Nguyen L-­T, Schmidt HA, Von Haeseler A, Minh BQ. IQ-­TREE: a coccus aureus in Europe: results from a multicenter study. PLoS fast and effective stochastic algorithm for estimating maximum-­ One 2012;7:e34768. likelihood phylogenies. Mol Biol Evol 2015;32:268–274. 2. Earls MR, Shore AC, Brennan GI, Simbeck A, Schneider-­Brachert W, 20. Letunic I, Bork P. Interactive Tree of Life (iTOL) v4: recent updates et al. A novel multidrug-­resistant PVL-­negative CC1-­MRSA-­IV clone and new developments. Nucleic Acids Res 2019;47:W256–W259. 8 Earls et al., Microbial Genomics 2021;7:000601 21. Duchêne S, Duchêne D, Holmes EC, Ho SYW. The performance community-­associated methicillin-­resistant Staphylococcus aureus of the date-­randomization test in phylogenetic analyses of time-­ lineage from the Indian subcontinent. mBio 2019;10:e01105-19. structured virus data. Mol Biol Evol 2015;32:1895–1906. 26. Duchêne S, Holt KE, Weill F-­X, Le Hello S, Hawkey J, et al. Genome-­ 22. Bouckaert R, Vaughan TG, Barido-­Sottani J, Duchêne S, Fourment M, scale rates of evolutionary change in bacteria. Microb Genom et al. BEAST 2.5: an advanced software platform for Bayesian evolu- 2016;2:e000094. tionary analysis. PLoS Comput Biol 2019;15:e1006650. 27. van Hal SJ, Steinig EJ, Andersson P, Holden MTG, Harris SR, et al. 23. Rambaut A, Drummond AJ, Xie D, Baele G, Suchard MA. Posterior Global scale dissemination of ST93: a divergent Staphylococcus summarization in Bayesian phylogenetics using Tracer 1.7. Syst aureus epidemic lineage that has recently emerged from remote Biol 2018;67:901–904. northern Australia. Front Microbiol 2018;9:1453. 24. Stegger M, Wirth T, Andersen PS, Skov RL, De Grassi A, et al. 28. Kahanec M, Zimmerman KF. Migration in an enlarged EU: a chal- Origin and evolution of European community-­acquired methicillin-­ lenging solution? Econ Pap 2009;363:16–19. resistant Staphylococcus aureus. mBio 2014;5:e01044-14. 29. Uhlemann AC, Otto M, Lowy FD, DeLeo FR. Evolution of commu- 25. Steinig EJ, Duchene S, Robinson DA, Monecke S, Yokoyama M, nity- and healthcare-­ associated methicillin-­resistant Staphylo- et al. Evolution and global transmission of a multidrug-­resistant, coccus aureus. Infect Genet Evol 2014;21:563–574. Five reasons to publish your next article with a Microbiology Society journal 1. The Microbiology Society is a not-for-profit organization. 2. We offer fast and rigorous peer review – average time to first decision is 4–6 weeks. 3. Our journals have a global readership with subscriptions held in research institutions around the world. 4. 80% of our authors rate our submission process as ‘excellent’ or ‘very good’. 5. Your article will be published on an interactive journal platform with advanced metrics. Find out more and submit your article at microbiologyresearch.org. 9