Molecular characteristics and risk factor analysis of Staphylococcus aureus colonization put insight into CC1 colonization in three nursing homes in Shanghai

Research indicates that Staphylococcus aureus colonization in the elderly with predisposing risks is associated with subsequent infection. However, the molecular epidemiology and risk factors for S. aureus colonization among residents and staff in nursing homes (NHs) in China remain unclear. A multicenter study was conducted in three NHs in Shanghai between September 2019 and October 2019. We explored the prevalence, molecular epidemiology, and risk factors for S. aureus colonization. All S. aureus isolates were characterized based on antimicrobial resistance, virulence genes, multilocus sequence typing (MLST), staphylococcus protein A (spa) typing, and staphylococcal cassette chromosome mec (SCCmec) typing. NH records were examined for potential risk factors for S. aureus colonization. S. aureus and methicillin-resistant S. aureus (MRSA) isolates were detected in 109 (100 residents and 9 staff, 19.8%, 109/551) and 28 (24 residents and 4 staff, 5.1%, 28/551) subjects among 496 residents and 55 staff screened, respectively. Compared to methicillin-susceptible S. aureus isolates, all 30 MRSA isolates had higher resistance rates to most antibiotics except minocycline, rifampicin, linezolid, vancomycin, and teicoplanin. Sequence type (ST) 1 (21.3%) was the most common sequence type, and t127 (20.5%) was the most common spa type among 122 S. aureus isolates. SCCmec type I (70%) was the dominant clone among all MRSA isolates. CC1 (26/122, 21.3%) was the predominant complex clone (CC), followed by CC398 (25/122, 20.5%), CC5 (20/122, 16.4%) and CC188 (18/122, 14.8%). Female sex (OR, 1.70; 95% CI, 1.04–2.79; P = 0.036) and invasive devices (OR, 2.19; 95% CI, 1.26–3.81; P = 0.006) were independently associated with S. aureus colonization.

Introduction Staphylococcus aureus is a gram-positive bacterium known to secrete infection-related toxins and invasive enzymes such as hemolysin, Panton-Valentine leukocidin, plasma-coagulase, and enterotoxins [1]. Infections caused by S. aureus range from skin and soft tissue infections (SSTIs) to invasive diseases, such as endocarditis, lung abscesses, and osteomyelitis, especially in hospitals and community settings [2][3][4]. In fact, S. aureus often colonizes asymptomatically on different body sites in healthy individuals, which significantly increases the chances of infection by providing a pathogen reservoir [5]. Methicillin-resistant S. aureus (MRSA) was first reported at a hospital in England in 1961, and it quickly became an important global pathogen [6]. MRSA is resistant to all currently available β-lactam antimicrobial agents, including β-lactamase-stable penicillins, and cephalosporins, which presents a challenge in terms of treatment and infection control [7]. To date, virulent MRSA strains are prevalent in the general community [8]. Furthermore, multidrug resistance has increased globally and is now considered a public health threat. Several previous studies revealed the emergence of multidrugresistant bacterial pathogens from different origins especially birds, animals, and food chains which may be transmitted to human consumers resulting in severe illness [9][10][11]. Nursing homes (NHs) are implicated as important regional reservoirs of S. aureus [12], which poses a threat to elderly individuals [13]. Old and frail adults receiving NH care are at a higher risk for S. aureus infection due to age-related immune senescence, accumulation of comorbid conditions, impaired mobility, and frequent hospital admissions [14]. Living in small NH facilities was independently associated with MRSA colonization among residents, as reported in a Brazilian study [13]. Transmission dynamics within NHs may be explained by frequent close contact and potentially via fomite contamination in crowded living environments, shared physical therapy devices, shared bath equipment, and group dining facilities [15]. There are also frequent daily contacts between the staff and the residents. Gowns and gloves should be used during specific care activities, especially for residents with chronic skin issues such as pressure ulcers, which have a higher transmission risk [16]. Thus, it is necessary to evaluate the S. aureus colonization rate of NH staff as well.
Various S. aureus clones circulate in different countries or regions associated with specific geographical locations. CA-MRSA (community-associated MRSA) is characterized by strong pathogenicity, a broad drug resistance spectrum, a high drug resistance rate, and complex drug resistance mechanisms [17], which have been studied extensively. For example, the ST1 and ST8 clones are mainly found in the United States and Canada, while ST80 clones are more prevalent in Europe; moreover, ST59 clones are the most common MRSA clones in China and several other Asian countries [18]. Some STs are significantly associated with the occurrence of complications, disease severity, and mortality [19]. Therefore, adding improved surveillance for the molecular characteristics of S. aureus in NHs perhaps will better predict the prognosis of subsequent infections. Most epidemiological studies of S. aureus performed focus on MRSA emergence and spread in healthcare settings. However, few microbiological screening studies have been performed within NHs in China or regarding the molecular epidemiology and risk factors for S. aureus colonization. CC1 was found to be the predominant clone among residents in seven NHs in Shanghai in 2015 [20]. A further confirmatory investigation was conducted in 2019 in three NHs in Shanghai with staff being screened concurrently. The present study of S. aureus in NHs is important for resident health and for providing a better understanding of the successful epidemic clones in NHs, which play an important role in medical care [21].
The objective of this study was to explore the risk factors for S. aureus colonization among residents and to investigate the molecular epidemiology of S. aureus isolates derived from residents and staff in three NHs in Shanghai. The predictors identified will help form infection control strategies for minimizing S. aureus prevalence and transmission in NHs, thus reducing the social and economic impact of NH-associated S. aureus in nearby hospitals and the local community.

Settings and participants
This cross-sectional study was performed between September 2019 and October 2019, in three NHs in Shanghai, China: one in the Jingan District, and the other two in the Putuo District. This study was a descriptive study that focused on the risk factors for S aureus and MRSA carriage and epidemiological surveillance of S. aureus colonization in specimens sourced from nasal, axillary, and skin samples from residents in three NHs in Shanghai. There are 35 and 43 registered NHs in Jingan District and Putuo District, respectively, that were issued business qualification certificates by the local district government. All three participating NHs (NH1, NH2, and NH3) are privately owned facilities containing 167, 320, and 180 beds, respectively. NHs were randomly selected and asked to participate in the study before it started in 2019. The inclusion criteria that were applied to participant: residents who signed the informed consent, had a normal body temperature during the survey phase, and had no obvious symptoms of infectious disease. Paper-based informed consent was signed by residents enrolled or by a statutory agent for residents with a cognitive deficit. Skin swabs from the wound surface were also collected if the residents were suffering from pressure sores or other skin lesions. Each of the 496 residents was surveyed with a real-name questionnaire to assess age, sex, comorbidities, any invasive devices being used, hospital admissions during the preceding 12 months, previous surgery, and antimicrobial usage in the previous 3 months. Data related to NH information included the number of resident beds, staffing ratio (staff per 10 beds), average monthly expenditure per resident, and sanitary conditions.

Sample collection
This study was approved by the Ruijin Hospital Ethics Committee (Shanghai Jiao Tong University School of Medicine) and the approval code was RJ2019NO140-1. A total of 551 individuals (496 residents and 55 nursing staff), aged 40-102 years, in the three NHs were screened for S. aureus. All swabs were cultured on blood agar for 24 h at 37˚C. Further analysis was based on morphology, Gram staining, catalase and coagulase testing, mannitol fermentation experiment, and S. aureus chromogenic medium at a clinical microbiology laboratory [22,23]. Species confirmation was performed using matrix assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry (VITEK 1 MS, bioMérieux, France) [24].

Detection of toxin genes
DNA was extracted using the simplified alkaline-lysis method [29]. Several clinically significant toxin genes [30,31], including sea-see and seg-sej (which encode staphylococcus enterotoxin SEA-SEE and SEG-SEJ, respectively), lukS/F-PV (which encodes Panton-Valentine leucocidin), tst (which encodes toxic shock syndrome toxin-1), and eta and etb (which encode exfoliative toxin A and B, respectively), were analyzed by PCR assay. The PCR reactions consisted of 12.5 μL 2×Taq Master Mix (Accurate Biology Hunan, China), 1 μL forward primer (10 μM), 1 μL reverse primer (10 μM), 2 μL DNA template (200 ng/μL-600 ng/μL), and 8.5 μL ddH 2 O. The PCR conditions were determined and adjusted according to primers and amplicon lengths as previously described [32]. The reactions were run for 30 cycles (each cycle included 94˚C for 30 s, 55˚C for 30 s, and 72˚C for 1 min), with initial hot start (94˚C for 10 min) and final extension (72˚C for 10 min). All the primers used in this study were synthesized by Sangon Biotech (Shanghai, China). We added 5 μL PCR products into each well, and the agarose gel electrophoresis ran for 40 min in 0.5×TAE solution under 110 v voltage. The PCR products were detected in a 1.5% agarose gel stained with ethidium bromide and visualized under UV light.

Molecular typing
All S. aureus isolates were characterized by staphylococcus protein A (spa) typing [33] multilocus sequence typing (MLST) [34], and staphylococcal cassette chromosome mec (SCCmec) typing, as previously described [20]. The spa types and STs were assigned by uploading Sanger dideoxy DNA sequencing into online databases (http://spa.ridom.de/index.shtml; https:// pubmlst.org/). A minimum spanning tree was generated by the PHYLOViZ web server based on the ST types (http://www.phyloviz.net). PCR assays were performed to identify the SCCmec structural types. The reagents used for PCR were the same as described above, so as the conditions for gel electrophoresis.

Statistical analysis
The data were collected using Epidata 3.1 (Epidata Association, Odense, Denmark) and exported to SAS (version 8.2, SAS Institute Inc. Cary, NC) and SPSS (version 21, IBM-SPSS Inc. Armonk, NY) software for further analysis. Data regarding colonized and uncolonized NH residents were analyzed using univariate analysis with the χ 2 test or Fisher exact test for categorical variables, and Student's ttest or the Mann-Whitney U test for continuous variables. Multilevel regression analysis was performed using SPSS, and multivariate logistic regression analysis was performed with the forward likelihood ratio logistic-regression model included in SAS to identify variables independently related to S. aureus and MRSA colonization. All tests performed were two-sided, and P<0.05 was considered statistically significant.

Prevalence of S. aureus colonization and overview of the NHs
A total of 551 nasal swabs, 551 axillary swabs, and 40 skin swabs were collected from 551 individuals (496 elderly residents and 55 nursing staff) in three nursing homes in Shanghai. S. aureus colonization was detected in 100 residents and 9 nursing staff. The overall prevalence of S. aureus in all subjects was 19.8% (109/551). S. aureus isolates from 96 residents were detected in only one sample area (nasal, n = 77; axillary, n = 18; and skin, n = 1). Additionally, S. aureus isolates from 13 residents were detected in 2 of 3 sample areas (nasal and axillary, n = 12; axillary and skin, n = 1). No S. aureus isolates were detected in all three sample areas. A total of 88 S. aureus isolates were collected from 551 nasal swabs (88/551, 15.97%), including 22 MRSA isolates (22/551, 3.99%). Thirty-two S. aureus isolates were collected from 551 axillary swabs (32/551, 5.8%), including eight MRSA isolates (8/551, 1.45%). Two S. aureus isolates were collected from 40 skin swabs (2/40, 5%), and no MRSA was detected. A total of 109 (100 residents and 9 staff) and 28 (24 residents and 4 staff) subjects were found to be colonized by S. aureus and MRSA, respectively. Hence, the overall prevalence of S. aureus and MRSA colonization was 19.8% and 5.1% in the NHs.
In addition, 113 S. aureus isolates were collected from 100 of the 496 residents screened, including 26 MRSA isolates, and the overall prevalence of S. aureus and MRSA colonization among residents was 20.2% (100/496) and 4.8% (24/496), respectively. Two different genetic phenotypes of S. aureus isolates were obtained from three residents respectively, including one MRSA isolate. Moreover, nine S. aureus isolates were collected from 55 staff screened including four MRSA isolates, and the overall prevalence of S. aureus and MRSA among staff was 16

Antimicrobial resistance
Among the 113 S. aureus isolates collected from the elderly residents, 26 (23%) were MRSA and 87 (77%) were methicillin-susceptible Staphylococcus aureus (MSSA). Only one MRSA isolate was susceptible to penicillin among 26 MRSA. All resident isolates were susceptible to minocycline, sulfamethoxazole-trimethoprim, rifampicin, linezolid, vancomycin, and teicoplanin. Among the nine S. aureus isolates collected from the nursing staff, four strains (44%) were MRSA and five (56%) were MSSA. All S. aureus isolates from staff were susceptible to minocycline, chloramphenicol, rifampicin, quinupristin-dalfopristin, linezolid, fusidic acid, Sanitary level: Level 1: The room has an independent bathroom and can take a shower with a fixed bath time.
Level 2: There is an independent bathroom and shower in the room, but the bath time is not fixed. Level 3: There is a separate bathroom in the room, but no shower and the bath time is not fixed. https://doi.org/10.1371/journal.pone.0253858.t001 vancomycin, and teicoplanin. Compared to MSSA isolates, all 30 MRSA isolates had higher resistance rates to most antibiotics except minocycline, rifampicin, linezolid, vancomycin, and teicoplanin. The antibiotic resistance rates for all 122 S. aureus isolates (113 from residents and 9 from nursing staff) are presented in Table 2.

Toxin genes
The toxin genes eta, etb, see, and lukS/F-PV have not been detected in any S. aureus isolate collected from residents. sec was detected most frequently among the screened toxin genes, occurring in 31 isolates (31/113, 27.4%). Forty-four isolates did not have any of the virulence genes including seven MRSA isolates, which were all SCCmec V. sea and seh were found more frequently among MRSA isolates than among MSSA isolates (P = 0.0011, P = 0.0313, respectively). The toxin genes sed and sej were detected only in MSSA isolates, but there was no statistical difference between MSSA isolates and MRSA isolates (P>0.05). Detailed results about toxin genes are listed in Table 3.

Risk factor analysis
In the univariate analysis, the significant risk factors for MRSA colonization were previous hospitalization (odds ratio [OR], 2.50; 95% confidence interval [CI], 1.07-5.82; P = 0.029), invasive devices (OR, 3.45; 95% CI, 1.45-8.18; P = 0.003), senile pruritus (OR, 2.72; 95% CI, 1.09-6.83; P = 0.027), renal insufficiency (OR, 4.91; 95% CI, 1.69-14.28; P = 0.001), and diabetes mellitus (OR, 3.72; 95%CI, 1.59-8.70; P = 0.001). The only significant risk factor for S. aureus colonization was invasive devices (OR, 1.96; 95% CI, 1.14-3.36; P = 0.014). The results of univariate statistical analysis are shown in Table 7. In terms of NH characteristics, sanitation levels and the number of staff were both significant factors for S. aureus, but not MRSA. The definition of the sanitation level is shown in Table 1. The average monthly expenditure per resident was not a significant factor for S. aureus or MRSA colonization. In multilevel logistic regression analysis, the result of the empty model (model S. aureus and model MRSA) showed that the intraclass correlation coefficient was 0.169 (P = 0.414) and Table 4

Discussion
About a third of healthy individuals in the community are colonized by S. aureus in the nostrils [2]. Nasal carriage of S. aureus has been associated with subsequent infection [35], and carriers are an important source of infection spread in communities. Infections caused by MRSA tend to be either widespread or are local outbreaks, which are difficult to treat, have high fatality rates, and increase the socio-economic and medical burden [6]. Compared with healthcareassociated MRSA (HA-MRSA) infections, CA-MRSA infections can occur in healthy individuals with no identified predisposing risk, suggesting that these strains may have greater virulence [36]. Nursing homes increase MRSA transmission and infection risk as a result of admission of elderly residents with premorbid conditions and weakened immune systems [37][38][39]. The MRSA colonization rate among residents in this study (24/496, 4.8%) was much lower than that in other studies showing 9.0% in Belgium [40], 9.04% in Saudi Arabia [41], and 14.5% in Crete; in Greece [12], the colonization rate we observed was slightly higher than that in Brazil (3.7%) [13] which may differ by geographic location. Twenty STs and thirty-two spa types were found in 122 S. aureus isolates collected from the three NHs in Shanghai, indicating high genetic diversity and complexity. In our study, the most prevalent S. aureus lineages were CC1 (21.3%), CC398 (20.5%), CC5 (16.4%), and CC188 (14.8%) which was generally consistent with a 2014 study of seven NHs in Shanghai [20]. We then might be able to hypothesize that CC1 was the dominant clone during the past 5 years in NHs in Shanghai, which may point out the direction for our future research. However, CC1-SCCmec I MRSA is gradually replacing CC1-SCCmec IV-MRSA as the predominant MRSA clone. These results indicated that SCCmec I-MRSA, which was generally associated with HA-MRSA infection, might spread to NHs. Previous studies reported that HA-MRSA infection is generally accompanied with multidrug-resistance [42]. It is necessary to strengthen MRSA admission screening and take appropriate quarantine measures to prevent the spread of HA-MRSA in communities. Moreover, there was a consistent dominant clone in both residents and staff which may indicate the S. aureus transmission dynamics in NHs, was similar to previous study results. In addition, gloves and gown use is highly recommended [16]. CC1-MSSA is among the most frequently observed clonal lineage in animals in some African countries [43]. CC398, which has been identified as a colonizer or infectious agent in livestock, was the second most common CC in our study. The genetic background association between CC1 and CC398 needs further research. There was a strong association between ST 1 and spa type t127 in our present study and our 2014 study [20]. In addition, ST2768 was the second most common ST (14/113, 12.4%), and all ST2768 isolates were MSSA. Importantly, all 12 ST2768 isolates were collected from the residents in NH2, which indicates that NHs should strengthen the awareness of the residents and healthcare workers. ST2768 was first documented in Spain from an ulcer swab of a patient with a diabetic foot infection in 2016, as shown in the S. aureus MLST database (https://pubmlst.org/saureus/). ST188 and ST4863 belong to CC188, of which 66.7% (12/18) of isolates were ST4863. These isolates were all obtained from the residents in NH3. Notably, 11 ST4863 isolates were genotyped as (spa) t189, and the other one was untyped. Moreover, ST4863 was recently added to the PubMLST database by Jie Hong in 2018. This strain was identified in Jiangsu Province, which is west of Shanghai. This result suggests that we should further monitor the molecular characteristics of the S. aureus and work to find the relationship between genetic relatedness and neighborhood of emerging ST clones. Among the 16 residents with MRSA typed SCCmec I, 10 residents had a history of recent hospitalization, which confirmed that recent hospitalization might be an important risk factor for MRSA colonization [13]. Among the four MRSA isolates collected from the staff, three MRSA isolates were identified as SCCmec I. This result was expected as Albrecht et al. [16] reported that MRSA transmission between residents and nursing staff often occurs during activities such as tidying beds, diaper changes, dressing changes, and transfers of position. In addition, nursing staff clothing and hands become contaminated with MRSA after caring for colonized residents. Thus, transmission was likely caused by direct contact with the contaminated hands of the nursing staff or the S. aureus isolates in the nursing staff survived in the environment and were transmitted after contact with the environment.
One hundred and thirteen S. aureus isolates were observed in 100 residents. Thirteen of 100 residents were colonized by S. aureus at two different sites. The molecular characteristics of the S. aureus isolates from different sites among the three resident groups were genetically different, as shown in Table 5. One MSSA isolate and one MRSA isolate were simultaneously obtained from one resident, and the MRSA strain was type SCCmec V, which indicates CA-MRSA [6]. Though it was suggested by Petry et al. [44] that in cases of wound or tissue samples with pending culture results, a negative MRSA nasal swab may influence the decision to withhold or discontinue MRSA-active screening. Samples from different sites should be collected when conducting S. aureus screening and evaluating the elimination of colonization [20,45].
Several studies revealed that lukS/F-PV is more closely associated with skin and soft tissue infections [46,47]. In this study, all S. aureus isolates were negative for genes that encode for Panton-Valentine leucocidin, enterotoxin E, and exfoliative toxins A and B. Interestingly, the frequency of tst in CC5 (9/13, 69.2%) isolates was significantly higher than that in other clones; only one isolate was MRSA, which may indicate that different S. aureus lineages have specific virulence gene patterns.
Similar to previous studies, invasive device presence was independently associated with S. aureus and MRSA colonization [12,48,49]. Although Stensen et al. reported that men have higher S. aureus nasal carrier rates than women [50], female sex was an independent risk factor for S. aureus colonization in the present study. In our previous study, we observed that the incidence of MRSA colonization was high in elderly patients with senile pruritus [51] which was confirmed in this study here. The link between renal insufficiency and MRSA colonization was also reported among hemodialysis patients [52]. Residents needing dialysis periodically go to the hospital, which may partially explain the increased risk. Notably, diabetic patients with specific pathological changes tend to have multiple complications. Previous studies suggest that the levels of MRSA colonization among diabetics with comorbidities in NHs or the wider community (patients and outpatients) reflects the extent of interaction with the healthcare system [53,54].
The main limitation of our study is the inclusion of a relatively small number of subjects. Larger sample sizes from community NHs are required to accurately depict the prevalence of S. aureus colonization in Shanghai. Additionally, a limited number of staff members caring for residents in NHs was included in this study. The genetic relationship between isolates colonized among residents and staff needs further research. Finally, the colonization mechanism of CC1 S. aureus isolates among residents in NHs remains unclear. A future study using wholegenome sequencing may help determine the mechanisms underlying S. aureus colonization.

Conclusions
The results provided by our study indicated that there are still measures we could take to control the high prevalence of S. aureus colonization in NHs. Strengthening the supervision and implementation of personal hygiene for resident individuals who may need assistance from nursing staff and staff individuals could reduce the source of infection. Furthermore, specific nursing management strategies for residents with predisposing factors for S. aureus colonization are indispensable such as rational drug use and regular screening.
Supporting information S1 Table. Information and molecular characteristics of S. aureus isolates among residents and staff of three nursing homes in Shanghai. (XLSX)