|Home | About | Journals | Submit | Contact Us | Français|
The incidence of methicillin-resistant Staphylococcus aureus (MRSA) has been increasing yearly at Peking Union Medical College Hospital (PUMCH). In order to understand the molecular evolution of MRSA at PUMCH, a total of 466 nonduplicate S. aureus isolates, including 302 MRSA and 164 methicillin-susceptible (MSSA) isolates recovered from 1994 to 2008 were characterized by staphylococcal cassette chromosome mec (SCCmec) typing, spa typing, pulsed-field gel electrophoresis (PFGE), and multilocus sequence typing (MLST). The 302 MRSA isolates were classified into 12 spa types and 9 sequence types (STs). During the years from 1994 to 2000, the most predominant MRSA clone was ST239-MRSA-III-spa t037. Since 2000, ST239-MRSA-III-spa t030 has rapidly replaced t037 and become the major clone. Another clone, ST5-MRSA-II-spa t002 emerged in 2002 and constantly existed at a low prevalence rate. The 164 MSSA isolates were classified into 62 spa types and 40 STs. ST398 was the most common MLST type for MSSA, followed by ST59, ST7, ST15, and ST1. Several MSSA genotypes, including ST398, ST1, ST121, and ST59, were identical to well-known epidemic community-acquired MRSA (CA-MRSA) isolates. MLST eBURST analysis revealed that the ST5, ST59, and ST965 clones coexisted in both MRSA and MSSA, which suggested that these MRSA clones might have evolved from MSSA by the acquisition of SCCmec. Two pvl-positive ST59-MRSA-IV isolates were identified as CA-MRSA according to the clinical data. Overall, our data showed that the ST239-MRSA-III-spa t037 clone was replaced by the emerging ST239-MRSA-III-spa t030 clone, indicating a rapid change of MRSA at a tertiary care hospital in China over a 15-year period.
Methicillin-resistant Staphylococcus aureus (MRSA), which can cause various diseases, from skin and soft-tissue infections to bacteremia, necrotizing pneumonia, endocarditis, and toxic shock syndrome (TSS), has become a serious clinical problem worldwide since the 1980s. Since the 1990s, the worldwide emergence of community-acquired MRSA (CA-MRSA) has been a threat to individuals in both the community and the hospital environment, since these strains are known to be more virulent than hospital-acquired MRSA (HA-MRSA). Our previous study found that the mean prevalence of MRSA across China was over 50% in 2005, and in Shanghai, the rate was over 80% (32). Therefore, it is very important to control the spread of MRSA in both hospital and community settings. Knowledge about the nature and the number of MRSA clones that are disseminating around the world is required to implement strategies to control the transmission of MRSA, either in hospitals or communities.
The major MRSA clones that cause infectious diseases worldwide are reported to belong to only a few pandemic clones. Using multilocus sequence typing (MLST) and eBURST, the five pandemic MRSA lineages can be grouped as clonal complexes: CC5, CC8, CC22, CC30, and CC45 (9). Along with the limited number of MRSA clones circulating in the world, the phenomenon of clonal replacement has been observed in several studies. For example, in Portugal, the Brazilian MRSA clone, which was the leading clone in the late 1990s, was declining and being progressively replaced by the EMRSA-15 and New York/Japan clones (5), and in Argentina, this clone was displaced by the Cordobes clone (sequence type 5 [ST5], spaA type TIMEMDMGMGMK, staphylococcal cassette chromosome mec [SCCmec] type I) (28). Similarly, the Hungarian clone (ST239-III) was replaced by the Southern German clone (ST228-I) and the New York/Japan epidemic clone (ST5-II) in Hungary (7). All of these observations indicate that clonal replacement has occurred or is occurring in the world. Therefore, understanding these changes at the local and international levels is of great significance. However, there are few reports on the dynamics of MRSA over long periods of time in China, mainly because in the past, few hospitals stored the MRSA isolates and it is also difficult to perform a nationwide collection of MRSA isolates from multiple centers over a long period in China. Therefore, the purpose of this study was, first, to investigate the molecular evolution of MRSA in a tertiary care hospital in China over a long period of time (1994 to 2008) and, second, to study whether the genetic background of methicillin-susceptible Staphylococcus aureus (MSSA) is related to current MRSA epidemic clones.
A total of 302 nonrepetitive MRSA (including the first MRSA isolate in 1994) and 164 nonrepetitive MSSA isolates were selected from the Peking Union Medical College Hospital (PUMCH) between January 1994 and December 2008. PUMCH consists of two campuses comprising an area over 170,000 m2 with 1,800 hospital beds. There are 40 clinical departments and 15 adjunct departments, including 10 national key disciplines (in 12 departments) and 2 municipal key disciplines. Over the past decade, patient visits steadily increased, last year alone tallying up to 1,850,000 outpatient visits, 48,000 inpatient admissions, and 28,000 surgical operations, with inpatient admission occupancy above 90%. From 1994 to 2000, the incidence of MRSA was relatively low, and we selected all the stored viable MRSA isolates (n = 94) during this period. After 2000, the rate of MRSA was increasing rapidly. In order to exclude isolates affected by colonization, after 2000, we chose MRSA and MSSA isolates from sterile body sites. Between 2001 and 2008, about 100 MRSA and 80 MSSA strains were isolated from sterile body sites each year in PUMCH. We selected about 50 MRSA and 40 MSSA strains isolated from a variety of sources and different wards of PUMCH every even year, and these strains covered all kinds of antibiograms. The isolates were recovered from various clinical sources, including the respiratory tract (n = 92), blood (n = 87), secretions (n = 55), drainage (n = 48), pus (n = 57), wounds (n = 62), abdominal fluid (n = 33), and other sources (n = 32). Isolates were confirmed to be S. aureus by the positive tube coagulase test and by use of VITEK2 automated systems (bioMérieux). Resistance to methicillin was determined by measuring the oxacillin MIC against the organism. The selected MRSA isolates were confirmed by multiplex PCR for the detection of the mecA and femB genes. All these selected isolates were stored at −80°C and grown overnight on blood agar plates at 37°C.
Antimicrobial susceptibility profiles were determined by the agar dilution method on Mueller-Hinton agar. The antimicrobial agents tested included penicillin (Sigma Chemical Co., St. Louis, MO), oxacillin (Sigma), clindamycin (Sigma), ciprofloxacin (Bayer AG, Leverkusen, Germany), chloramphenicol (Sigma), erythromycin (Sigma), gentamicin (Sigma), rifampin (Sigma), tetracycline (Sigma), trimethoprim-sulfamethoxazole (Sigma), and vancomycin (Sigma). Clinical and Laboratory Standards Institute breakpoints were used for MIC interpretation. S. aureus ATCC 29213 and Enterococcus faecalis ATCC 29212 were used as quality controls in each set of tests.
One loopful of bacteria was harvested from blood agar plates. Cell suspensions were successively incubated with lysostaphin and proteinase K, boiled, and centrifuged. The cell lysates were used directly in PCRs.
All of the 302 MRSA and 164 MSSA isolates were analyzed by SCCmec typing, spa typing, and Panton-Valentine leukocidin (pvl) gene detection. One isolate representative of each major spa type was further characterized by MLST. In order to verify the homology of spa t030 and t037 isolates, pulsed-field gel electrophoresis (PFGE) was performed on 30 randomly selected spa t030 and t037 isolates.
The SCCmec types were determined by a multiplex PCR strategy developed by Oliveira et al. (24). Nontypeable (NT) types were defined as isolates showing unexpected fragments or lacking some fragments as determined by multiplex PCR. International clones of SCCmec types I to V were used as quality controls.
spa typing was performed as described previously (14). Purified spa PCR products were sequenced, and short sequence repeats were assigned by using the spa database website (http://www.ridom.de/spaserver). The spa complex was defined by visual analysis, whereby spa types with similar short sequence repeats were clustered into the complexes previously described by Ruppitsch et al. (27).
MLST was carried out as described previously (10). The sequences of the PCR products were compared with the existing sequences available on the MLST website (http://saureus.mlst.net) for S. aureus, and the allelic number was determined for each sequence. The clustering of related STs, which were defined as clonal complexes (CCs), was determined by using the program eBURST (based upon related sequence types) (11).
DNA extraction and SmaI restriction were performed as described previously (17). A bacteriophage lambda DNA PFGE molecular size standard was included in each gel. The PFGE patterns were examined visually and were interpreted according to the criteria of Tenover et al. (30).
The detection of pvl genes was performed by PCR, as described previously (20). The identity of the PCR products was confirmed by sequencing with an ABI 3700 sequencer.
In order to track the original MRSA clone and the substitute clone, the medical records of the patients from whom spa t037 isolates were obtained in 1994 and the patients from whom spa t030 isolates that emerged in 2000 were obtained were reviewed. In addition, the medical records of two patients from whom pvl-positive and SCCmec type IV isolates were obtained were also reviewed. The following variables of data were collected: patient demographics (gender and age); ward transfer; underlying diseases; the length of time after admission that a sample for culture was obtained; the presence of an invasive device (e.g., a vascular catheter or a gastric feeding tube) at the time of admission; and a history of MRSA infection or colonization, surgery, hospitalization, dialysis, or residence in a long-term care facility in the 12 months preceding the culture.
Table Table11 shows the antibiotic resistance profile, SCCmec type, spa type, ST, CC, and pvl status for 302 MRSA isolates recovered from 1994 to 2008. Table Table22 displays the clonal distribution of 164 MSSA isolates recovered in this study year by year. The results concerning significant S. aureus clones are as follows.
Of the 302 MRSA isolates, 106 (35.1%) belonged to the clone of ST239-MRSA-III-spa t037. All isolates of this clone were pvl negative. Most of the t037 strains were resistant to tetracycline, erythromycin, clindamycin, gentamicin, chloramphenicol, trimethoprim-sulfamethoxazole, and ciprofloxacin and susceptible to rifampin and vancomycin. In PUMCH, the first MRSA isolate was found in 1994 and belonged to this clone. According to the medical record, this patient, who had no history of hospitalization, received surgical repair of the sigmoid colon. The MRSA strain was isolated from the peritoneal drainage fluid the fourth day after the operation. The patient was treated with gentamicin, imipenem, and metronidazole, and, finally, the patient was cured. spa t037 was the predominant clone in PUMCH from 1994 to 2000, accounting for 92.6% (87/94) of all the MRSA isolates during the period. However, after 2000, this clone decreased rapidly, accounting for 9.1% (19/208) between 2002 and 2008, maintaining at a low proportion.
ST239-MRSA-III-spa t030 was the most frequent clone, accounting for 52.0% (157/302) of the 302 MRSA isolates. However, to our surprise, before 2000, none of the MRSA isolates was found to belong to this clone. In 2000, this clone emerged but accounted only for 8.2%, and the prevalent clone was still ST239-MRSA-III-spa t037, accounting for 91.8%. The first patient with t030 MRSA was transferred from another hospital near PUMCH and admitted to the intensive care unit (ICU). Subsequently, the patient got better and was transferred to the surgical ward. Soon, t030 began to spread in the ICU and surgical wards. Then, in 2002, the drastic clonal replacement occurred. The prevalence of the t030 clone, accounting for 89.6% (43/48), increased significantly and established its dominant status. From 2002, ST239-MRSA-III-spa t030 has been the leading clone in MRSA infections in PUMCH and accounted for 64% to 89.6% every year. Antibiogram analysis revealed that t030 was resistant to rifampin. In order to get high-resolution analysis of the relationship among these t030 isolates, 15 t030 isolates were randomly selected for PFGE. The result demonstrated that these isolates shared identical or closely related PFGE types which were obviously different from that of t037.
Overall, ST5-MRSA-II-spa t002, which belonged to the CC5 clonal complex, accounted for 7.0% (21/302). Compared to ST239-MRSA-III-spa t037/t030, this clone was resistant to tetracycline, erythromycin, clindamycin, gentamicin, and ciprofloxacin but susceptible to chloramphenicol, rifampin, trimethoprim-sulfamethoxazole, and vancomycin. This clone was not detected in PUMCH before 2002. Then, in 2002, a very small number of isolates of ST5-MRSA-II-spa t002 emerged. Since then, t002 was found every year but took up only a very small proportion, from 2.1% to 28%.
Along with the three prevalent MRSA clones, several minor clones were also found in PUMCH. Two isolates with spa t2164 and one isolate with spa t539 belonged to ST5, which had SCCmec II. Four strains with spa t062 belonged to ST965, which was a single-locus variant (SLV) of ST5. The four t062 isolates carried different SCCmecs, 2 isolates had SCCmec IV, and the other two isolates had SCCmec II and SCCmec III. Another isolate with an SLV of ST5 was identified as ST641 (SLV5)-MRSA-II-spa t214. All of these isolates belonged to the same clonal complex (CC5). Of the sporadic isolates, 7 isolates belonged to CC8, including ST158 (SLV239)-MRSA-III-spa t632 (n = 4), ST585 (SLV239)-MRSA-III-spa t3518 (n = 2), and ST1097 (SLV239)-MRSA-III-spa t2270 (n = 1). Two isolates with the pvl gene were ST59-MRSA-IV-spa t437, and one isolate belonged to ST723 (singleton)-MRSA-III-spa t254.
Several major clones have chronologically dominated in PUMCH during the past 15 years: (i) the ST239-MRSA-III-spa t037 clone was the main clone until 2000, (ii) the ST239-MRSA-III-spa t030 clone was introduced into PUMCH in 2000 (this clone has replaced t037 and become the predominant one since 2002), and (iii) the ST5-MRSA-II-spa t002 clone emerged in 2002 and existed at a low prevalence rate until now.
Table Table22 shows the clonal distribution of MSSA isolates in each year. Compared with the relative preponderance of a few MRSA clones, MSSA showed more diverse patterns. The 164 MSSA isolates were classified into 62 spa types and 40 STs. For all the isolates, 12 spa types were identified as the relatively prevalent types, including t571 (18/164, 11.0%), t084 (9/164, 5.5%), t034 (7/164, 4.3%), t127 (7/164, 4.3%), t163 (7/164, 4.3%), t437 (7/164, 4.3%), t377 (7/164, 4.3%), t091 (6/164, 3.7%), t796 (6/164, 3.7%), t002 (5/164, 3.0%), t189 (5/164, 3.0%), and t2091 (5/141, 3.5%). For MLST types, the most frequent MLST type was ST398 (31/164, 18.9%), followed by ST59 (14/164, 8.5%), ST7 (14/164, 8.5%), ST15 (14/164, 8.5%), ST1 (11/164, 6.7%), and ST5 (8/164, 4.9%). ST5, ST59, and ST965 coexisted in both MRSA and MSSA isolates.
The pvl gene was determined for all of the MRSA and MSSA isolates. Overall, 2 MRSA and 9 MSSA isolates were pvl positive. Nine spa types and eight STs were found. Two MRSA isolates that were pvl positive belonged to ST59-MRSA-IV. These 2 isolates were identified as community-associated MRSA (CA-MRSA), according to the medical records.
Among the many MRSA isolates, only a few MRSA clones were widespread as major clones over a long period of time. In order to understand the pandemic MRSA clones and the molecular evolution of MRSA in PUMCH over a long period, we investigated the molecular characteristics of 302 MRSA and 164 MSSA isolates by SCCmec-spa-MLST typing methods from 1994 to 2008.
Between 1994 and 2008, the Brazilian clone ST239-MRSA-III was the most prevalent clone determined by the molecular typing of 302 MRSA isolates from PUMCH. The Brazilian MRSA clone was first discovered in Brazil and widely disseminated in various hospitals (29). Subsequently, this clone spread to neighboring South American countries (Argentina  and Uruguay and Chile ) and to Europe (Portugal , the Czech Republic , and Greece [1)]), where it displaced the local major clones. Recently, ST239 also became the epidemic clone in most Asian countries (12). Our previous study demonstrated that this clone was also the dominant one in most of the cities in China from 2005 to 2006 (21). ST239 represented a distinct branch within CC8 in the evolutionary model of the emergence of MRSA. The previous study proposed that ST239 arose from a single recombination event that involved the exchange of >200 kb of contiguous DNA between ST30 and ST8 (26).
ST239 still presented as the most predominant ST of MRSA over all the years of the study period, but by the higher-resolution spa typing method, the epidemic clone had a rapid change. Our data demonstrated that spa t037, which was prevalent before 2000, was rapidly replaced by the t030 MRSA clone, which emerged in 2000. Since then, t030 has become the most popular MRSA clone. Surprisingly, it was after 2000 that the prevalence of MRSA increased rapidly in PUMCH, which was coincident with the introduction of t030. It seemed that ST239-MRSA-III-t030 had a stronger survival advantage and could easily transmit in our hospital. The most obvious difference between these two clones was that most t030 isolates were resistant to rifampin, while only a few t037 isolates were resistant.
In order to investigate whether t030 evolved from the former epidemic clone or was an alien clone, we reviewed the medical records of the patients from whom the t030 MRSA isolates were first isolated in 2000. Most of these patients were transferred to PUMCH from other hospitals. So we speculated that t030 was introduced in this hospital from outside and replaced the previous epidemic clone t037. In addition, we randomly selected 15 t030 MRSA isolates between 2000 and 2008 for PFGE typing. The results showed that the 15 isolates had identical or similar PFGE types, which suggested that they belonged to the same ST239-MRSA-III-t030 clone that had spread in this hospital since 2000. Besides, there were some minor evolutionary events related to the epidemic clone ST239. Four MRSA isolates evolved from ST239 to ST158 by a single point mutation of yqiL; another two isolates belonged to ST585, which evolved from ST239 by a new single-locus variant (aroE) mutation; and another isolate belonged to ST1097, which evolved from ST239 by a single point mutation of glpF (Fig. (Fig.11).
ST5-MRSA-II was the second epidemic MRSA clone in PUMCH which belonged to the New York/Japan international clone. This clone was initially described as the main clone in the United States (25) and Japan (2) and subsequently detected in several European (6) and Asian (18) countries. The previous data also showed that ST5-MRSA-II was the prevalent clone in several cities (such as Shenyang and Dalian) in China (21). In this study, ST965 (a single-locus variant of pta of ST5) and ST5 MSSA were found as early as 1996. Then as time went by, ST5-MRSA-II isolates emerged in 2002, and ST965-MRSA-II was detected in 2008. Another clone, ST641-MRSA-II, evolved from ST5-MRSA-II by aroE mutation (Fig. (Fig.11).
Frequent conversion of MSSA to MRSA by horizontal transfer of SCCmec has been described (13), suggesting that MSSA was the origin of MRSA and that MRSA independently developed by multiple evolutionary events in the accessory genome within a single ST. Thus, to elucidate the origin of MRSA, 164 MSSA isolates recovered between 1994 and 2008 were characterized by spa and MLST. Interestingly, the genotypes of MSSA isolates were much more diverse than those of MRSA. The majority of MRSA isolates typically belonged to one or a few different clonal types. In comparison, MSSA isolates were more often genetically variable. In this study, ST5, ST59, and ST965 coexisted in both MRSA and MSSA isolates, and ST5-MSSA-t002 and ST59-MSSA-t437 had already existed before the emergence of ST5-MRSA-t002 and ST59-MRSA-t437, which might indicate that several MRSA clones had evolved from MSSA in PUMCH.
Another interesting finding was that some high-frequency MSSA genotypes closely resembled those of the well-known epidemic CA-MRSA clones. Of the types determined for the 164 MSSA isolates, ST398 was the most popular type, but this ST was not found with MRSA isolates. Recently, ST398-V MRSA was found to cause community infections in pigs, pig farmers, and the patients who had the pig contact experience (19, 31). Another popular ST was ST1, which bears the same ST as MW2. MW2 was the first CA-MRSA strain reported in the United States and is characterized as PFT USA400, SCCmec IV, and pvl positive (6). Another ST that has been reported to cause skin and soft tissue infections in the community is ST121 (22), and 6 isolates belonged to this ST in this study. Currently, none of the isolates belonging to this ST have been reported as MRSA, but given the opportunity, this type might be converted to CA-MRSA. Therefore, the findings above suggest that these MSSA clones, once they acquire some accessory genes, like, SCCmec, pvl, or the pathogenicity island, would cause serious CA-MRSA infections or even outbreaks; thus, we should pay more attention to these clones.
Through the eBURST analysis of MSSA, ST121 was the predicted founder of CC121, five isolates belonged to ST837, and two isolates belonging to ST95 evolved from this founder by yqiL mutation. Another four isolates belonged to ST965, which is a single-locus variant of ST5, and six isolates belonged to ST188, which was a double-locus variant (DLV) of ST1, which was the predicted founder of CC1. One MSSA strain belonged to ST195, which was a DLV of ST630, and the two STs belonged to CC8. Therefore, MSSA also presented evolutionary events.
In this study, the number of MSSA isolates carrying the pvl gene (n = 9) was higher than that of MRSA isolates (n = 2), which seemed to indicate that toxic gene deletion occurs during the evolution of MSSA to MRSA. We found two pvl-positive MRSA isolates from the outpatients that belonged to ST59-MRSA-IV, and the clinical data confirmed them to be CA-MRSA. ST59-MRSA-IV was reported to be prevalent in skin and soft tissue infections in children in Taiwan (16), Hong Kong (15), and mainland China (33), but in this study, except for these two isolates, the rest that belonged to ST59 were MSSA, which might indicate that ST59-MRSA-IV was not popular in community-acquired infections in adults.
In summary, we have described for the first time the long-term molecular evolution of MRSA in a tertiary care hospital in China. The epidemiology of MRSA has been evolving rapidly. A new clone (ST239-MRSA-III-spa t030) was introduced and replaced the previously predominant clone (ST239-MRSA-III-spa t037). Long-term multicenter MRSA evolution and the genetic evidence for the survival advantage of t030 need further study.
This work was supported by the National Natural Science Foundation of China (no. 30971571) and the Beijing Medical Scientific Development Foundation (no. 2007-2015). It was partially supported by the Beijing Natural Science Foundation (no. 7102130) and Beijing Nova Programme (no. 2006A43).
Published ahead of print on 22 February 2010.