Entry of Panton–Valentine leukocidin-positive methicillin-resistant Staphylococcus aureus into the hospital: prevalence and population structure in Heidelberg, Germany 2015–2018

Staphylococcus aureus is one of the major pathogens causing community—and healthcare-acquired infections. The presence of the virulence factor Panton–Valentine leukocidin (PVL) is associated with recurrent infection and clinical severity and generally regarded as a feature of community associated-methicillin-resistant Staphylococcus aureus (MRSA). To date, the focus of PVL-positive MRSA in hospitalized patients has been on outbreaks. We aimed to investigate whether PVL-positive MRSA has penetrated the community-hospital barrier by determining the prevalence of PVL in MRSA of hospitalized patients. MRSA strains isolated from patients hospitalized > 48 h in Heidelberg University Hospital between 2015 and 2018 Isolates were analysed for the presence of PVL and subjected to spa-typing. PVL-positive MRSA were then characterized by whole genome sequencing. We analysed 740 MRSA isolates in the study period and identified 6.2% (n = 46) PVL-positivity. 32.6% of PVL-positive MRSA met the criteria for nosocomial acquisition. The most frequent clones among the PVL-positive strains were ST80-t044 (21.7%, n = 10/46) and ST8-t008 (19.5%, n = 9/46). WGS identified three possible transmission clusters involving seven patients. In conclusion, we found successful epidemic PVL-positive MRSA clones entering the hospital and causing nosocomial infections. Preventive measures and constant surveillance should be maintained to prevent transmissions and clonal outbreaks.

Nevertheless, systematic data on the prevalence and molecular characteristics of PVL-bearing MRSA in hospitalized patients in Germany is scarce. Surveillance data and other published report suggest that a high percentage of imported epidemic MRSA clones are circulating and causing community acquired infections [2][3][4] . More importantly, patients with CO-MRSA SSTI often seek medical care in out-patient units and emergency units of tertiary hospitals 10,11 , providing a port of entry for these MRSA clones to the hospital setting.
In this study, we analyzed and characterized MRSA isolates from hospitalized patients with a special focus on PVL-positive MRSA by WGS retrospectively to understand the population structure and dynamics of PVL positive MRSA entering and circulating in our hospital. Clonal shift of MRSA population in the healthcare setting should be monitored closely to prevent clonal spread of highly virulent and epidemic MRSA clones.
For 740 of 756 MRSA, data regarding the presence of PVL was available. Of 740 isolates, 46 (6.2%) were PVL-positive with 2.7% (n = 5/188), 10.8% (n = 17/158), 6.2% (n = 13/210) and 6.0% (n = 11/184) in the respective years (Table 1). clinical and molecular characteristics of pVL-positive MRSA. We identified 46 (6.2%) patients with PVL-positive MRSA in the study period (Table 1). Most patients were males (58.7%, n = 27/46) and mean age was 36 years with a range of 1 month to 96 years. Patients with PVL-positive MRSA were significantly younger than those with PVL-negative MRSA (mean age 36 years and 53 years, respectively, p < 0.0001). The proportion of males did not differ significantly between the two groups (p = 0.93). 58.7% (n = 27/46) of the patients had an infection with PVL-positive MRSA, of which 69.6% (n = 32/46) were simultaneously colonized while 26% (n = 12/46) were free of nasal and rectal PVL-positive MRSA on screening swabs. Two patients with PVLpositive MRSA infection were not screened for rectal or nasal MRSA colonization. 54.3% (n = 25/46) of patients had a documented history of migration. 15 patients (32.6%) met the criteria for nosocomial MRSA acquisition of being detected > 48 h after hospital admission (Fig. 1). Thus, the majority of patients (67.4%, n = 31/46) was already infected or colonized with PVL-positive MRSA on admission to hospital. Skin and soft tissue infections were the most common clinical presentation (68.8% of cases, multiple clinical presentations possible). Two cases presented with blood stream infections and two with bone and joint infections ( Table 2).
possible transmission events. SNP analysis identified three possible transmission clusters involving 7 patients (Fig. 1, red rectangles). These included three patients with ST80 (t044) in cluster 1, two patients with ST8 (t008) in cluster 2 and another two patients with ST5 (t002) in cluster 3. All three patients with t044 had overlapping hospitals stays on the same wards. The two patients with t008 and both patients with t002 had no temporal or spatial (same ward) overlap during their hospital stay.

Discussion
PVL is regarded as one of the main features of CA-MRSA and a clinically relevant virulence marker for S. aureus 3 . In recent years, cases of nosocomial PVL-positive MRSA transmissions and outbreaks in Europe have been accumulating 5,12,13 thus suggesting that the barrier between community-and hospital-associated MRSA has already been penetrated. Although PVL is associated with the severity of infection 3, the characterization of MRSA isolates in hospitalized patients in Germany in terms of PVL is not consequently pursued. To our knowledge, this is the first systematic analysis on the prevalence and molecular structure of PVL-positive MRSA in hospitalized patients in Germany. We found that the clonal distribution as determined by MLST and spa-typing was overall concordant with existing knowledge about the molecular epidemiology of MRSA in Germany: 30.5% of isolates were spa-type t003, consistent with the predominant ST5 "Rhine-Hesse epidemic clone" 14 and 5.8% t002 8 . Likewise ST8 t008 MRSA (4.6%,n = 35/756) is considered one of the most common MRSA clones circulating in Germany 8 . Unexpectedly, our sample contained 2.6% (n = 20/756) of t063 MRSA which has been rarely described in Germany, yet 8 . This was explained by a putative transmission cluster of PVL-negative MRSA in our hospital.
Altogether, we detected 6.2% PVL positivity in MRSA. This was considerably higher that the findings of another study on MRSA prevalence in hospitalized and non-hospitalized patients in Germany 2010-2011, which reported 2.7% being PVL-positive 8 . A possible explanation for this discrepancy could the premise of the analysis. The authors did not differentiate between in-patients and out-patients for their analysis, it is therefore difficult to directly compare the reported rates of PVL-positive MRSA between the two studies. Similar to our findings ,  spatype  t044  t044  t131  t1247  t044  t044  t044  t044  t044  t044  t044  t044  t121  t121  t008  t008  t008  t008  t008  t008  t008  t008  t008  t127  t127  t9736  t690  t690  t325  t18874  t002  t002  t002  t002  t852  t005  t1328  t417  t005  t019  t019  t342  t034  t034  t355  t4690   MLST  80  80  80  80  80  153  80  80  80  80   www.nature.com/scientificreports/ a Swiss study reported 7.7% PVL-positive MRSA in a mixed population of hospitalized and non-hospitalized patients 15 . Similarly, a European study conducted among patients presenting with skin and soft tissue infections in the emergency room reported 8% PVL-positive MRSA for the German study centre 10 . Compared to the 40% prevalence of PVL-positive MRSA in patients presenting with SSTI in outpatient departments of our hospital 2 , these estimates and our findings support, that only a fraction of these patients actually enter the hospital. However, we also found PVL-positive MRSA in blood cultures and tissue samples from bone and joint infections, providing evidence for invasive infections in this population besides the high percentage of skin and soft tissue infections. Over two thirds (67.4%) of patients with PVL-positive MRSA in our study were already colonized or infected on hospital admission. This was somewhat expected, since the presence of PVL has been linked to communityassociated clones 2,16 . In line with our previous findings on the genetic background of MRSA clones in community onset SSTI, the majority of characterized isolates in this study belonged to ST8-t008 and ST80-t044 2 . ST80-t044 MRSA clonal lineage has been described as a frequent CA-MRSA clone circulating in Europe and the Middle East 17,18 . Another dominant clone was the ST8-t008 MRSA belonging to the epidemic USA300 clade (both North-American and Latin American variant). Although the prevalence of this MRSA clone in Europe is generally considered to be low 8,10 , we found a remarkably high proportion of USA300 MRSA clones circulating in the community 2 and USA300 is indeed the most commonly imported MRSA into the European continent through intercontinental travel 19 . Moreover, USA300 MRSA clones can be classified as epidemic, as demonstrated by their expansive properties and dominance on the American continents 20 and further highlighted by accumulating reports on nosocomial transmission events and outbreaks in Europe 5,12,13 .
Another interesting clone is the ST1-t127 MRSA, which is regarded as livestock-associated MRSA (LA-MRSA), made up 5.4% of our study isolates. This particular clone has been linked with bovine mastitis and colonization of farm animals. Although several reports of human infections with ST1-t127 MRSA have been published, this occurrence is still considered rare 21,22 . Moreover, some of the LA-MRSA clones harbor the virulence factor PVL. Although reports on the presence of PVL genes in ST1-t127 have been published, the isolates in those reports harbour the human-adapted SCCmec IV 22 . In contrast our PVL-positive ST1-t127 MRSA were of the SCCmec V, consistent with the porcine ST1-t127 MRSA 22 . The acquisition of PVL was not only restricted to the ST1 clones but several LA-MRSA isolates with ST1232 (t034), a single locus variant of the ST398 LA-MRSA, suggesting human adaptation of LA-MRSA clones 22 . Taken together, these finding demonstrates the relevance of the "One Health" concept for the surveillance of multi-drug resistance bacteria.
Of the three previously undetected potential transmission clusters identified by SNP analysis, one cluster most likely represents nosocomial transmission events, since these three patients had overlapping hospital stays on the same ward (t044, cluster 1, Fig. 1). The two patients with t008 MRSA (cluster 2, Fig. 1) had no temporal or spatial overlap. Comparative analysis of virulence genes revealed slight differences in genes present/absent (Fig. 2), suggesting non-identical isolates despite close clonal relationship. Therefore, patient-to-patient transmission was unlikely. In the third potential transmission cluster (t002, cluster 3, Fig. 1), both patients did not have apparent epidemiological overlap during their hospital stays. However, since both patients were housed in the same residential facility for refugees, a transmission event outside of the hospital setting is a plausible alternative explanation. Our findings underline the importance of reliable epidemiological data for investigations of transmissions and outbreaks of multi-drug resistant bacteria.
Our study has strengths and limitations. It was a single-centre study and may therefore only reflect the local epidemiological situation in a part of Germany. However, this systematic analysis provides evidence on the occurrence of PVL-positive MRSA over a 4-year period in a tertiary care hospital. Molecular typing data have indicated that epidemic PVL-positive MRSA clones have entered the hospital and may cause nosocomial infections and outbreaks. Thus, implementation of pre-emptive measures, such as systematic surveillance of MRSA population structures in hospitalized and non-hospitalized patients is needed, with special emphasis on PVL-positive clones.

Methods
Study population. We analysed and characterized MRSA isolates from hospitalized patients in our tertiary care hospital over a 4-year period between 2015 and 2018 retrospectively. The study was performed at Heidelberg University Hospital, located in the South-West of Germany. Clinical and microbiological data were acquired though the routine microbiological diagnostic, surveillance of multidrug resistant bacteria and patient care. Only hospitalized patients (> 48 h) were included in this study.
Microbiological methods. MRSA were cultured and detected from routine screening samples and clinical samples from hospitalized patients seeking medical care, as described previously 2 . Species identification was performed with the MALDI-TOF (Bruker GmbH, Germany), antibiotic susceptibility testing using the VITEK2 (Biomérieux GmbH, Germany) and interpreted according to current EUCAST clinical breakpoints. MRSA was confirmed by the presence of nuc and mecA as described earlier 2 . Isolates were cryopreserved for surveillance purposes.
Spa typing, Sccmec typing and detection of pVL. PVL detection was performed by PCR, as previously described 3 . Spa typing was performed by Sanger sequencing as published elsewhere 23 . SCCmec typing for the cassette types I to V were determined by multiplex PCR as described earlier 24 . All PVL positive strains were subject to WGS for a more detailed analysis.

WGS.
Genomic DNA was extracted from overnight bacterial culture using the DNeasy Blood and Tissue Minikit (Qiagen GmbH, Germany) according to manufacturer's protocol with the addition of a prior lysis step Scientific RepoRtS | (2020) 10:13243 | https://doi.org/10.1038/s41598-020-70112-z www.nature.com/scientificreports/ with lysostaphin (Genaxxon GmbH, Germany). Standard genomic library was prepared from the DNA and sequenced with the Illumina MiSeq platform (2 × 250 bp paired end). Quality control and assembly was performed as described previously 25 . Briefly, raw sequences were trimmed for quality using Sickle 1.33 (parameters, q > 30; 1 > 45) 6 . The cleaned sequences were then assembled using SPAdes 3.13.0 7 . Contigs obtained from the assembly were curated for length (> 500 bp) and coverage (> 10 ×) to ensure no errors and contamination in the draft genome. Annotation was performed using Prokka 1.14.1 (based on Genetic Code Table 11) 8 and the NCBI Prokaryotic Genome Annotation Pipeline. Resistance and virulence genes were found using Abricate 0.8.13 (https ://githu b.com/tseem ann/ abric ate with respectively the antibiotic database from ResFinder 3.0; ARG-ANNOT; CARD; NCBI-BARRGD and the virulence database VFDB. MLST was perform using the MLST 2.0 pipeline from the centre for Genomic Epidemiology 26 . Sequences will be made publically available upon publication of the manuscript. Cluster analysis by SNP and allelic difference. Genomes were compared by calculating the core genomes with Roary and only genes present in all isolates were considered (1965 genes, 74.7 ± 1.7% of the genomes). Phylogenetic distance was calculated with Gubbins 2.3.4 to take in account recombination events and not over-estimate the SNP polymorphism 27 . Clonal groups were defined as genomes distant from less than 10 SNPs.
Data and statistical analysis. Patient data was extracted from the medical record. A history of migration was defined as a patient with home address outside of Germany or living in a community facility for refugees. Descriptive statistics was analyzed using the STATA 13 software (StataCorp, USA). Two-sample t test or chi square test was used calculating the p value for age and sex, p < 0.05 was considered statistically significant. ethical considerations. Ongoing molecular characterization of MRSA isolates is performed as part of obligatory surveillance and infection control measures mandated by the German Infection Protection Act. All methods were carried out in accordance with relevant guidelines and regulations. The Ethical review board of the University of Heidelberg approved the study protocol (S-474/2018) and waived informed consent.