Molecular epidemiology and antibiotic resistance of methicillin- resistant Staphylococcus aureus circulating in the Russian Federation
Abstract
The aim of this study was to investigate the patterns of antimicrobial resistance and molecular features of methicillin-resistant Staphylococcus aureus (MRSA) isolates in Russia. Isolates recovered from hospital patients (n=480), healthy medical personnel (n=25), and healthy carriers (n=13) were included in the study. Hospital-acquired MRSA (HA-MRSA) demonstrated high resistance to ciprofloxacin, gentamicin, and chloramphenicol (76%–92%), moderate – to tetracycline, erythromycin, clindamycin, and rifampicin (38%–54%), and low – to fusidic acid, co-trimoxazole, mupirocin, and daptomycin (2%–7%). Elevated MIC (2.0 µg/ml) of vancomycin was detected in 26% of isolates. All isolates were susceptible to linezolid and tigecycline. Multilocus sequence typing (MLST) revealed that CC8 isolates (ST8 + ST239) constituted 83.1% of HA-MRSA and that this genetic lineage dominated in all regions from Krasnoyarsk to Saint Petersburg. A local ST239 variant harboring the tst gene (ST239Kras) was detected in Krasnoyarsk. The other HA-MRSA isolates belonged to clonal complex 5 (CC5) (21 isolates, 12.2 %) and CC22 (2, 1.2%). The majority of CC5 isolates were affiliated with sequence type 228 (ST228) and were characterized with decreased susceptibility to ceftaroline (MIC=2 µg/ml). We also detected, for the first time in Russia, livestock-associated MRSA (LA-MRSA) from clusters CC398 and CC97 in humans. Among the 2,053 healthy persons screened for nasal carriage of S. aureus, the bacteria were isolated from 426 (21%); among them, 13 carried isolates identified as community-associated MRSA (CA-MRSA). Eleven of 13 CA-MRSA isolates belonged to ST22 (spa types t223, t3243, and t3689; SCCmec types IVa and IVc, agr type I, tst- positive) and were similar to the EMRSA-15/Middle Eastern variant (Gaza strain).
Key words: HA-MRSA; CA-MRSA, Multidrug resistance; Multi-locus sequence typing
1. Introduction
Methicillin-resistant Staphylococcus aureus (MRSA) is one of the leading causes of hospital- and community-acquired infection worldwide. MRSA has evolved from methicillin- susceptible S. aureus (MSSA) lineages through acquisition of the staphylococcal cassette chromosome (SCCmec) element, which carries the mecA gene encoding an alternative penicillin- binding protein, PBP2A, with a low affinity for β – lactam antibiotics (Robinson and Enright, 2003).
Wide application of molecular techniques (Multilocus sequence typing – MLST and spa typing) for genotyping revealed that S. aureus is a clonal bacterium and that a limited number of genetic lineages or clonal complexes (CCs) are predominant in the staphylococcal population worldwide (Feil et al., 2003). However, since SCCmec evolves almost independently from the core genome, genotyping based on this mobile element is also necessary for adequate MRSA isolate characterization.
For decades after the identification of the first MRSA isolate in 1960th, MRSA was regarded as a typical healthcare-associated (HA) pathogen. The epidemiology of HA-MRSA is commonly described by geographical location and temporal dynamics in the distribution of particular clones; however, it is established that CC1, CC5, CC8, CC22, CC30, and CC45 are predominant around the world (Lindsay, 2010). The main changes in the global epidemiology of MRSA during the last decades are associated with the emergence of distinct genetic lineages of community-associated MRSA (CA-MRSA) in 1990th (Mediavilla et al., 2012) and of livestock- associated MRSA (LA-MRSA) in 2000th (Graveland et al., 2011). Recently, a novel variant of the mec gene (mecC) was identified in S. aureus isolated from cattle first in the UK and Denmark (Garcia-Alvarez et al., 2011), and then in many other countries (Diaz et al., 2016).
Due to the geographical position of the country both in Europe and Asia, the characterization of MRSA epidemiology in Russia should present not only local, but also international interest. The country can serve as a donor or as a recipient of successful clones as well as an exchange pathway between remote regions. However, there are no data to what extent the global changes in MRSA epidemiology affect MRSA distribution in Russia. The aim of the present study was to evaluate the diversity and circulation of the global MRSA lineages in the Russian Federation.
2. Material and methods
2.1. Bacterial isolates and antibiotic susceptibility testing
MRSA isolates were collected in participating centers and sent to the Scientific Research Institute of Children’s Infection, Saint Petersburg (SRICI) together with record forms. Personal data of patients were not included in record forms, ethical approval was not required. Participating centers were represented by large multidisciplinary hospitals with capacity more than 1000 beds and with average 60,000 hospitalized in-patients per year. Three groups of MRSA isolates were included in the study. The first comprised non-duplicate HA-MRSA isolates (n=480) from patients with different healthcare-associated infections (HAI) obtained during 2011–2014 in 15 medical centers from 12 cities (Segeja, Petrozavodsk, Saint Petersburg, Murmansk, Moscow, Yaroslavl, Samara, Perm, Chelyabinsk, Kurgan, Khanty-Mansiysk, and Krasnoyarsk – see map, supplement S1). The mean (±95% confidential interval) age of the patients was 42±20 years (from 18 to 80 years) with sexes ratio of male and female respectively 2:1. HAI were defined by the local physicians as infections occurring ≥48 h after hospital admission. MRSA isolates were recovered from different clinical samples (surgical wounds accounted 27%; bone infections – 23%; sputum – 13.5%; blood – 13%; burns – 10.5%, skin and soft tissue infection – 7% and other – 6%) from patients with HAI and were identified in local laboratories by standard methods (supplement S2). The second group consisted of HA-MRSA isolates (n=25) recovered from healthy medical personnel during routine screening for nasal carriage of S. aureus in three centers (Moscow, St. Petersburg, and Perm) during 2013–2014. Screening for nasal carriage is obligatory for surgical staff according to national regulations. Finally, the third group included MRSA isolates (n=13) recovered from healthy employees of a grocery supermarket in Moscow during routine screening for nasal carriage of S. aureus in the frame of the annual state preventive healthcare check-up program (community-associated MRSA, CA-MRSA). Two main criteria of the participants were used for inclusion in the study: absence of any clinical signs of staphylococcal infections and/or other infections and absence of a history of hospitalization within the last year. Nasal swab samples were collected from the anterior nares using the eSwab transport system (Copan, Italy). The swabs were introduced into nostrils, gently rotated about four times, and sent to the laboratories of participating centers within 24 h. The swabs were inoculated into mannitol-salt agar and incubated at 37°C for 24–48 h; the resultant characteristic yellowish colonies were further analyzed by conventional methods. MRSA detection was performed using oxacillin or cefoxitin disks. Isolates were sent to the SRICI to confirm the identification. To identify the isolates, matrix-assisted laser desorption/ionization mass spectrometry was performed on a Microflex LT (Bruker Daltonics, Germany) in accordance with the manufacturer’s instructions (Sogawa et al., 2011). The presence of the mecA gene was confirmed by PCR. Only mecA-positive S. aureus isolates were subjected to further analyses. Minimum inhibitory concentrations (MICs) of antibiotics were determined by the broth microdilution method in cation-adjusted Mueller-Hinton broth (Bio- Rad, Marnes-la-Coquette, France), and interpreted according to the EUCAST guidelines (EUCAST, 2016).
2.2. MLST, typing of spa, SCCmec, and agr, and detection of resistance and virulence genes
MLST and spa typing were performed as described (www.pubmlst.org/saureus and www.spaserver.ridom.de, respectively). Sequencing of amplicons in both directions was performed by the Sanger method (ABI DNA Analyzer 3730, Applied Biosystems, Foster City, CA). Multiplex PCR was used for genotyping SCCmec (according to the recommendations of the International Working Group on the Classification of Staphylococcal Cassette Chromosome Elements (IWG-SCC, 2009) and agr. The genes associated with antibiotic resistance (mecA, ermA, ermB, ermC, lnuA, fusB, fusC, fusD, mupA, cat, aac(6’)-aph(2’’), aadD, aphA3, tetK, tetM) and virulence (sea, seb, lukSF, tst, arcA-opp3) were detected by PCR. Genes involved in fluoroquinolones resistance (gyrAB, parCE) were sequenced in randomly selected isolates with high level MIC. Primer sequences are presented in supplement S3.
2.3. Statistical analysis and software.
MIC data were statistically analyzed using the Whonet 5.6. software based on EUCAST clinical breakpoints (version 6.0) and epidemiological cut-off criteria for mupirocin. Two-tailed Fisher’s exact test was used in comparison analysis.
3. Results
3.1 Resistance profile of MRSA
The results of antibiotic susceptibility testing of different groups of MRSA isolates included in the study are presented in Table. 1.
A total of 480 from 716 suspected HA-MRSA isolates submitted from participating centers were confirmed as MRSA in the central laboratory. Despite the presence of the mecA gene, MICs of oxacillin for 1.5% of the isolates were below the resistance breakpoint (0.25–2 μg/ml); at the same time, 5% isolates demonstrated decreased susceptibility to ceftaroline (MIC=2 μg/ml). MRSA demonstrated high level of resistance to fluoroquinolones, gentamicin, chloramphenicol, tetracycline, erythromycin, clindamycin and rifampin (92% – 38%), no more than 7% isolates were resistant to mupirocin, co-trimoxazole, and fusidic acid. The MIC of vancomycin was 3–4 μg/ml for only few isolates; however, for 26% isolates it was 2 μg/ml, among them 3% of isolates demonstrated decreased susceptibility to daptomycin (MIC=2 μg/ml).
Twenty-five HA-MRSA isolates were recovered from medical personnel in three centers (Moscow, St. Petersburg, and Perm). Their patterns were the similar with HA-MRSA from hospitalized patients, and majority of the isolates were resistant to erythromycin, clindamycin, tetracycline, gentamicin and fluoroquinolones.
Among healthy employees (n = 2,053) of a grocery supermarket in Moscow screened for nasal carriage of S. aureus, the bacteria were recovered from 426 (21%) individuals. Thirteen (0.6%) isolates were mecA-positive, demonstrated borderline oxacillin-resistant phenotype with MIC range from 0.25 to 8 μg/ml and were recognized as CA-MRSA. The majority CA-MRSA isolates were susceptible to non-β-lactam antibiotics, but 46.2% of them were resistant to co- trimoxazole, and 15.4% – to erythromycin, and tetracycline.All MRSA isolates were fully susceptible to linezolid and tigecycline.
3.2 Genotyping of MRSA
All 480 HA-MRSA isolates were subjected to SCCmec, agr typing and detection of virulence and resistance genes. A representative set of isolates (172) from different geographical regions of different SCCmec types, as well as of different resistance and virulence patterns were selected for spa typing, and 87 were additionally analyzed by MLST (Table 2). The phylogenetic analysis based on MLST profiles dataset with epidemiology data can be found in supplement S4. Of the 172 isolates, 88 (51.1%) belonging to two major spa types (t008 and t024) and several minor types corresponded to ST8. All of them belonged to agr group I, and harbored SCCmec IVce and a similar set of resistance-associated genes. In addition, ermB and lnuA were detected in some t024 isolates. The isolates belonging to ST8-t008\t024 – SCCmec IVce were dominant in all regions. Another HA-MRSA group (n=55, 32.0%) comprised three major (t632, t037, and t030) and two minor (t4238, t138) spa types corresponding to ST239, all of them harbored SCCmec III. Peculiar features of SCCmec III from ST239 isolates were the presence of an additional recombinase ccrC and absence of the mer complex. Finally, all ST239 isolates (n=13) from Krasnoyarsk (Siberia) contained toxic shock syndrome (tst) genes.
CC5 was represented by ST228, ST5, and ST764. The majority of isolates from this group (14 of 21) belonged to European clone ST228-t041-SCCmec IA, and demonstrated decreased susceptibility to ceftaroline (MIC=2 μg/ml). One ST5-t02460-SCCmec III (ccrC+) isolate was identified in Ural region (Kurgan), and all other CC5 isolates were detected in the European part of Russia. Two isolates belonged to СС22 (ST22); they were represented by different spa types (t223 and t688), harbored SCCmec IVce and agr I type, and were susceptible to non-β – lactam antibiotics.
Two genetic lineages of LA-MRSA (ST398 and ST97) were detected. ST398 isolates were recovered from hospitalized patients in Saint Petersburg, isolates belonged to different SCCmec and spa types, and agr groups but carried the same set of drug resistance genes. ST97 isolates were from patients with osteomyelitis residing in geographically distant regions (St. Petersburg and Kurgan). ST97 isolates belonged to the same agr I type, but demonstrated different spa and SCCmec types; one isolate was susceptible to non-β – lactam antibiotics, and another harbored the ermC, tetK, and aac(6’)-aph(2’’) genes.
The group of HA-MRSA from medical personnel was represented mostly by CC8 isolates (n=15), with main features similar to those of the CC8 HA-MRSA from hospitalized patients: spa types t008 and t024, SCCmec IVce type, agr I type, and the presence of the ermC, cat, tetK, and aac(6’)-aph(2’’) resistance genes. Two t008 isolates belonged to single locus variants (SLV) of ST8 (ST569 and ST1497) and lacked genes encoding resistance to non-β – lactam antibiotics. Sporadic isolates were represented by ST5, ST22, ST97, ST398, and ST121 and were similar to HA-MRSA of corresponding STs from hospitalized patients.
Eleven of 13 CA-MRSA isolates belonged to ST22 (spa types t223, t3243, and t3689; SCCmec types IVa and IVc; agr type I, tst-positive, and no resistance genes); two spa type t223 isolates also carried seb genes. The remaining two CA-MRSA isolates belonged to CC5; one of them was ST5 and another – ST2704 and carried lukSF genes.
The results of fluoroquinolone-resistant isolates gyrA and parC genes sequencing indicated that these isolates had well-characterized non-synonymous mutations: S84L and S80F. All MRSA isolates analyzed here lacked the ACME-complex genes. We were unable to detect mupA gene in isolates resistant to mupirocin (MIC>2 μg/ml).Both phylogenetic analysis and distribution of sources of MRSA isolates are presented in supplements S2 and S5.
3.3 Resistance profiles and SCCmec types
Two major MRSA groups carrying SCCmec III (ST239) and SCCmec IVce (ST8) were included in the analysis. MRSA carrying other types or non-typeable SCCmec were represented by small number of isolates and demonstrated highly variable resistance profiles. Comparison of major groups revealed significant differences of resistance patterns (Table 3). Isolates carrying SCCmec III demonstrated higher mean geometric MIC of oxacillin than isolates with SCCmec IVce: 512 μg /ml and 75 μg /ml respectively. Associated resistance to non-β-lactam antibiotics was also higher among SCCmec III carrying isolates; the differences reached statistical significance for moxifloxacin, rifampicin, tetracycline, mupirocin, and fusidic acid and were less pronounced for co-trimoxazole, ciprofloxacin, gentamicin and chloramphenicol. However, isolates with vancomycin MIC = 2 μg/ml were more predominant among HA-MRSA carrying SCCmec IVce (35% vs. 23%, p=0.02).
4. Discussion
According to a recent review, MRSA prevalence in Russian hospitals varied significantly from 0 to 80% (Gostev and Sidorenko, 2015). Comparison of the present data with previous publications (Dekhnich et al., 2002) revealed increase of the percentage of isolates with vancomycin MIC=2.0 mg/ml from 2% in 2002 to 26% in 2014. Since high probability of treatment failure is associated with elevated vancomycin MIC alternative antibiotics are needed. All MRSA isolates were fully susceptible to linezolid and tigecycline, but isolates with decreased susceptibility to ceftaroline and daptomycin were observed.
However, MRSA population was not homogeneous; CA-MRSA demonstrated lower level of oxacillin MIC90 in comparison with HA-MRSA and were susceptible to non-β – lactam antibiotics. In the present study CA-MRSA were recovered only in Moscow from healthy carriers, isolates belonged mainly to ST22, spa type t223, harbored agr type I and SCCmec IVa, they were tst-positive. This lineage is closely related to EMRSA-15/Middle Eastern variant (Gaza strain) which was originally detected among healthy carriers in the Gaza Strip (Biber et al., 2012). Later, this variant was reported in other countries of the Middle East region (Abou Shady et al., 2015; Aqel et al., 2015; Senok et al., 2016) and occasionally in Europe (Monecke et al., 2016); there is also evidence of its invasion of healthcare settings (Geraci et al., 2014). In the present study, one isolate similar to the Gaza strain was recovered from hospitalized patient also. We can’t reconstruct details of Gaza strain importation into Russia, but at least in Moscow it’s circulation is already established.
Previously there was only one report on CA-MRSA isolation in Russia from Vladivostok (Baranovich et al., 2007), but the isolate belonged to Southwest Pacific clone (ST30, SCCmec IVc, PVL-positive) (Huh and Chung, 2016) and it’s circulation probably didn’t established in Russia. In the present study we detected only two PVL – positive isolates belonging to ST121- t0308-SCCmec V and ST2704-t002-SCCmec IVce isolated respectively from medical personnel and from healthy carrier.
Among HA-MRSA main differences in resistance profiles were associated with the type of SCCmec. Isolates carrying SCCmec IVce were more susceptible to β-lactams and non-β-lactam antibiotics in comparison with SCCmec III carrying isolates. Presented observations are in agreement with known data that SCCmec type III is the largest variant among 12 known cassettes and that it can harbor many different resistance determinants (reference). Whole genome sequencing of ST239 isolate confirmed that majority of resistance genes are localized on SCCmec III.
In our study SCCmec IVce and SCCmec III were associated respectively with ST8 and ST239 lineages belonging to CC8 which are dominant in the extended territory from Krasnoyarsk to Saint Petersburg.According to earlier studies (Afanas’ev et al., 2010; Dmitrenko et al., 2005; Romanov et al., 2013; Romanov et al., 2012; Vorobieva et al., 2008), ST8-spa type t008-SCCmec IVсе is among the most prevalent MRSA lineages in Russia, at least since the middle 1990th. In Europe, such isolates are referred to as the Lyon clone (Ferry et al., 2006), and they are mostly restricted to Belgium and France (Aanensen et al., 2016; Grundmann et al., 2014; Hetem et al., 2016). The data represents strong evidence on the European origin of ST8 circulating in Russia. However there are some signs of local evolution: a variant of ST8 (ST8Kras) characterized by enhanced transcription of genes encoding cytolytic peptides and immune evasion factors was described in the Krasnoyarsk region (Khokhlova et al., 2015).
ST239 was found to be the second most prevalent MRSA lineage. This lineage is widespread in Australian, Asian, and South American hospitals and is sporadically reported in other continents (Harris et al., 2010); however, it has not been established in European hospitals (Knight et al., 2012). In Russia, ST239 appeared to be more successful; spa type t037 was identified in different regions in early 1990th (Dmitrenko et al., 2005) and at present it has spread on a large territory, and is represented by two different variants. Variant ST239Kras, was previously reported in Krasnoyarsk; the isolates of this variant belonged to spa type t037, were tst-positive, but sea-negative, and caused fatal hospital-acquired pneumonia (Khokhlova et al., 2015). In the present study, this variant was also detected only in Krasnoyarsk region. Isolates belonging to spa type t632, tst-negative but sea-positive were detected only in the central region (Saint Petersburg, Moscow, and Yaroslavl). Sporadic cases of spa type t632 were described in Turkey (Kirca Yilmaz et al., 2014), China (Shen et al., 2010) and Middle East (Udo et al., 2014). Ongoing importation of ST239 variants to Russia from Asia can’t be excluded.
СС5 is the third most important HA-MRSA lineage in Russia. In Europe it is represented by three clusters: ST225, ST228, and ST125 (Aanensen et al., 2016). In Russia, CC5 is represented mainly by HA-MRSA ST228 referred to as South German EMRSA (Witte et al., 1997). The majority of isolates belonging to this cluster demonstrated typical features, including spa type t041, SCCmec IA, and multiple drug resistance mediated by the ermA, spc, aphA3, and aac(6’)-aph(2’’) genes. Decreased susceptibility of some isolates to ceftaroline (MIC=2 µg/ml)
was the main feature of this lineage. Possibilities of further selection of high level of ceftaroline resistance and dissemination of this lineage are of major concern.
A number of HA-MRSA isolates in the current study belonged to livestock-associated lineages CC398 and CC97. To the best of our knowledge, this is the first report of СС398 and CC97 from Russia. Four CC398 isolates were recovered in the European part of Russia from patients and medical personnel in different hospitals; they demonstrated features typical to CC398 (Graveland et al., 2011). Three CC97 isolates recovered in remote regions (St. Petersburg and Kurgan) were genetically different and demonstrated different susceptibility patterns. We have no epidemiological data on possible contacts of patients and personnel with animals and on travel history. At present, it is impossible to discriminate sporadic cases of livestock-associated lineages importation and early stages of established circulation.
The major limitation of the study is insufficient geographical coverage, a number of regions in Siberia and Far East with low density of population were not included in the study. The overall number of collected isolates probably not enough for clear representation of the rare clones of MRSA, circulating in our country. Since the majority of isolates under the study originated from a highly selective group of patients, the clinical relevance of the data is limited.
5. Conclusion
MRSA epidemiology around the world is changing rapidly, and antibiotic resistance is a growing problem. Our study showed that the CC8 lineage represented by two clusters of different origin is predominant among HA-MRSA circulating in Russia. The ST8 cluster related to the Lyon clone widely spread in Europe is more prevalent. ST239 is probably originated from Asia and a local tst-positive variant (ST239Kras) has emerged in Russia. To the best of our knowledge, this is the first report of three rare genetic lineages detected in Russia. Among them, ST398 and ST97 represent LA-MRSA and originated from Europe, while ST22 (EMRSA- 15/Middle Eastern variant, Gaza strain) represents CA-MRSA and originated from the Middle East. Further surveillance is necessary to determine whether these lineages have been established in Russia or they represent sporadic imported cases without further transmission.