Molecular Diversity of Staphylococcus aureus Colonizing the Upper Respiratory Tract of Residents and Staff in A Nursing Home

Staphylococcus aureus is an opportunistic human pathogen known for its increasing prevalence in both hospitals and the general community. An incremental rise in drug-resistance and high disease-causing capabilities has also been observed (Zhang et al. 2015; Zurita et al. 2016). The most common anatomical location for the commensal carriage of S. aureus is the anterior nares (Kluytmans et al. 1997; Pollitt et al. 2018), although recent epidemiological studies highlight the role of pharyngeal carriage in S. aureus transmission (Nilsson and Ripa 2006; Hamdan-Partida et al. 2010). The complexity of host-colonizer interactions has led researchers to investigate alternative contexts in pursuit of the mechanism for colonization. Long term care facilities (e.g. nursing homes) constitute communities in which crowding and inadequate hygienic conditions might contribute to higher MSSA (methicillin-sensitive S. aureus) and MRSA (methicillin-resistant S. aureus) prevalence rates. Chronic diseases, immune system disorders, and frequent hospitalization make elderly people vulnerable to S. aureus infection (Ledell et al. 2003; Chuang et al. 2015; Peters et al. 2017).

S. aureus is also one of the most genetically diverse bacteria and the mechanisms by which the predominant strains are selected within each biological niche depends on the geographic location as well as the studied human host (Zurita et al. 2016). MSSA strains are more prevalent than MRSA strains and have a higher genetic diversity, whereas MRSA outbreaks are usually caused by only a few different MRSA clones (Goering et al. 2008). Therefore, analyzing the genetic relatedness of colonizing MSSA strains in nursing homes might result in a better understanding of S. aureus epidemiology.

We aimed at determining the genetic diversity and relatedness of MSSA and MRSA strains colonizing the upper respiratory tract of nursing home residents and staff using the MLVF (multiple-locus variable-number tandem-repeat fingerprinting) method. We hypothesized that there would be the same genetic MLVF banding patterns of S. aureus retrieved from different individuals, and the presence of genetic homogeneity might indicate that bacterial transmission has occurred.

Prevalence ofS. aureuscolonization. Among all participants, 46.1% (47/102) were colonized with S. aureus (Table I). The pattern of colonization was divided into three categories based on the anatomical location. The category that registered the highest number of individuals was the combination of the anterior nares and the throat (40.4% of colonized individuals), followed by the anterior nares only (36.2%), and the throat only (23.4%).

S. aureus colonizes multiple anatomical locations and the frequency of colonization in various niches might vary depending on the methodology and the context of the study. Our study demonstrated that the anterior nares are colonized more frequently than the throat, which is consistent with the findings of previous studies on the prevalence of both MSSA and MRSA occurring together (Roghmann et al. 2015; Zhang et al. 2015; Peters et al. 2017). Alternatively, some researchers have reported that colonization only in the throat is the most prevalent pattern (Nilsson and Ripa 2006; Hamdan-Partida et al. 2010) or have observed the high prevalence of extra-nasal colonization as well, i.e., in the groin and the perianal regions (Mody et al. 2008).

In the current study, 31.6% (6/19) of the participants colonized with two S. aureus strains (one in the anterior nares and one in the throat) registered dissimilar MLVF banding patterns, implying that the two S. aureus strains were unrelated. A similar observation has been made by (Mongkolrattanothai et al. 2011), who noted that simultaneously carriage of two or more unrelated S. aureus isolates was observed in 30.4% of positive swab samples collected from children. Colonization with more than one strain of bacteria entails a risk of heteroresistance, i.e., a situation in which an antibacterial agent is effective against one strain, but not against another strain of the same bacteria. Truly, in cases of co-colonization with MSSA and MRSA both together, heteroresistance might pose a therapeutic challenge.

Antimicrobial resistance profile ofS. aureus. Table II shows the resistance profiles of S. aureus strains with the information about their location in the dendrogram. Among all S. aureus strains tested, 10.6% (7/66) were resistant to tetracycline; 6.1% (4/66) to macrolides, lincosamide, and streptogramin; 4.5% (3/66) to tobramycin and/or gentamicin, and 1.5% (1/66) to ciprofloxacin, levofloxacin, and daptomycin. We observed the presence of the MLS_B-constitutive resistance only. The share of S. aureus isolates susceptible to all of the antimicrobial agents tested was high (77.3%, 51/66), although the presence of MDR (multi-drug resistant) MRSA strain (Table II) suggests the need of conducting longitudinal studies investigating its transmission and long-term colonization patterns (Sun et al. 2017). The antimicrobial susceptibility testing identified 98.5% (65/66) of S. aureus strains as MSSA, whereas only one isolate was identified as MRSA. Consequently, the MRSA prevalence rate in our studies was low when compared to similar studies: 1% among all participants and 1.7% specifically among the residents. The prevalence of MRSA colonization among residents of long term care facilities across the world varies within a wide range: from 0% in nine Swedish nursing homes (Andersson et al. 2012), 3.7% in nine nursing homes in Brasil (da Silveira et al. 2018), 7.2% in 19 long term care facilities in Luxemburg (Mossong et al. 2013), 20.4% in 36 residential care homes for the elderly in Hong Kong (Chuang et al. 2015), to 28% in 13 nursing homes in Michigan and Maryland (Roghmann et al. 2015).

Multiple-locus variable-number tandem repeat fingerprinting ofS. aureus. Using molecular typing methods to determine the genetic diversity of human pathogens is crucial in identifying the reservoirs and routes of transmission (Grzegorczyk and Malm 2014). Unlike other bacterial genotyping methods, such as PFGE (pulsed-field gel electrophoresis), MLST (multilocus sequence typing) or spa typing, MLVF has a comparable resolution, but is cheaper, less time- and labour-consuming (Luczak-Kadlubowska et al. 2008; Holmes et al. 2010; Kosecka-Strojek et al. 2016). The high discriminatory power of MLVF is its main advantage, enabling precise detection of single events of transmission between patients (Karynski et al. 2008).

The MLVF of the 66 S. aureus strains isolated from 47 individuals resulted in 51 banding patterns: 36 unique banding patterns were represented by only one strain and 15 banding patterns consisted of two identical S. aureus strains each. In accordance with our results, a high genetic diversity of MSSA has been reported in other studies (Emaneini et al. 2011; Karynski et al. 2008; Zurita et al. 2016). (Zurita et al. 2016) observed 69 MLVA (multiple-locus variable-number tandem-repeat analysis) genotypes among 70 tested MSSA strains from various populations.

The MLVF typing allowed us to divide the 66 S. aureus isolates into two main clusters, herein designated as A and B (Fig. 1). One S. aureus strain (R41N) was not compatible with any other banding pattern. In the dendrogram, it represents a singleton with no association to the main clusters. The main cluster A was divided into five subclusters (herein designated I–V); the main cluster B was divided into seven subclusters (herein designated VI–XII). Altogether, the twelve subclusters included 94% (62/66) of all strains (cut-off >35%). Comparing the subclusters concerning the size, the two largest (I and IV) both included nine banding patterns, and 13 and 11 S. aureus strains, respectively. The six smallest subclusters (III, V, VI, IX, X, and XI) each included between one and three banding patterns and two or three S. aureus strains.

Fig. 1.

Exemplary MLVF banding patterns for S. aureus strains are shown in Fig. 2. The number of MLVF bands depends on the variable expression of genes in the sdr region of the bacterial DNA. The expression of between one and three genes (sdrC, sdrD and/or sdrE) translates into a range between five and seven MLVF bands (Sabat et al. 2003). In our study, the six-band and seven-band profile were the most numerous, accounting for 56.0% and 36.4% of all isolates, respectively. The less common five-band profile captured only 7.6% of all isolates.

Fig. 2.

At the level of S. aureus isolates, we were able to group 26 isolates in even pairs based on perfect relatedness (one hundred percent homology). Without exception, each related pair of isolates was retrieved from the anterior nares and the throat of one individual, i.e., 13 individuals contributed one pair each. At the level of individuals, this meant that 27.6% (13/47) of the colonized individuals contributed a related pair of isolates. At the group level, i.e., the serendipitous detection of related isolates across the population, we were able to group four individuals in even couples, each couple sharing one bacterial strain. However, the anatomical location and pattern of co-colonization were inconsistent, rendering the similarities imperfect (for details, see subclusters I, VI and VIII in Fig. 1). These findings suggest that inter-participant transmission of S. aureus has occurred and might result in cross-infection – especially among immunocompromised individuals living in crowded conditions. (Andersen et al. 2002) have also reported that environmental factors, i.e., crowding and insufficient staffing, might increase the risk of S. aureus cross-infection.

Although our study provides valuable information about the genetic diversity, prevalence of colonization, and the occurrence of S. aureus transmission in a nursing home, it has several limitations. The MLVF method returns its most valid results when applied within a limited geographical area. As a result, making comparisons between studies from various countries or regions is difficult. Future research in the field of active microbial surveillance must include longitudinal aspects by subjecting the study populations to repeated sampling over extended periods. Understanding the temporal aspects of colonization might lead to new insights regarding the routes of bacterial transmission.