Introduction

Staphylococcus aureus is one of the leading causes of severe bacterial infections which may lead to life-threatening conditions, including sepsis, pneumonia, endocarditis, osteomyelitis, and implant-associated diseases [1,2,3]. The emergence of antibiotic resistance in S. aureus has posed a significant impact on the treatment and infection control practices in hospitals worldwide [4, 5]. Methicillin-resistant S. aureus (MRSA), vancomycin-resistant S. aureus (VRSA) and vancomycin-intermediate S. aureus (VISA) are among the pathogens listed as “High” priority in the World Health Organization (WHO)’s priority pathogens list for research and development of new antibiotics [6].

The ability of S. aureus to resist antimicrobials is further enhanced by the strategy of biofilm formation. Currently available antibiotics cannot eradicate biofilms, especially of ESKAPE pathogens, which includes MRSA [7]. Several studies found no statistically significant difference in the biofilm formation among MRSA and MSSA strains [8, 9]. In contrary, a study reported that MRSA strains showed enhanced biofilm formation as compared to MSSA strains [10]. Staphylococcal polysaccharide intercellular adhesin (encoded by icaABCD) [11], collagen-binding protein (cna), fibrinogen binding protein (fib), elastin binding protein (ebpS), laminin binding protein (eno), fibronectin binding proteins A and B (fnbA and fnbB), and clumping factors A and B (clfA and clfB) [12] have been reported to play important roles in S. aureus adherence, which is the first step in biofilm production. The icaABCD operon is also known for its function in slime production [13].

Quorum sensing is a mechanism, whereby bacterial cells communicate and coordinate their behaviours based on population density [14]. The accessory gene regulator (agr) quorum-sensing system plays a key role in S. aureus pathogenesis, while the staphylococcal accessory regulator (sarA) gene is essential in controlling staphylococcal virulence factors [15]. Both agr and sarA quorum sensing genes have been reported to regulate S. aureus biofilm formation [15,16,17,18]. To date, four polymorphic agr types (agrI, agrII, agrIII, and agrIV) have been reported [19].

Previously, a high prevalence of icaADBC genes and varied occurrence of biofilm associated genes, i.e., cna (42.7–93%), fib (24.7–90%), ebps (11.1–100%), fnbA (0–100%) and fnbB (1.1–53.33%) have been reported in Malaysian S. aureus clinical isolates [20,21,22]. The agr1 was the most prevalent type reported in Malaysian isolates of S.aureus, followed by agrII and agrIII; however, no agrIV was detected [21, 23]. Understanding differences in biofilm and slime production between MRSA and MSSA and the associated genetic elements contributes to a better understanding of the epidemiology and spread of S. aureus infections, further aiding in developing more targeted surveillance and treatment strategies. Hence, this study was performed to analyze biofilm and slime production of a collection of MSSA and MRSA clinical isolates and to investigate possible correlations between biofilm-associated genes and the agr/sar quorum sensing systems in relation to antibiotic resistance.

Methods

Collection of clinical isolates

A total of 68 MSSA and 54 MRSA isolates collected from patients attending Universiti Malaya Medical Centre (UMMC) from August 2020 to June 2022 were investigated in this study. The isolates were primarily collected from the blood (n = 38, 31.1%), and tissue (n = 36, 29.5%), followed by pus (n = 15, 12.3%), wound swab (n = 11, 9%), and lower respiratory tract (n = 18, 14.8%) (Additional file 1: Table S1). The identity of the isolates was confirmed using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (VITEK MS system, bioMérieux Clinical Diagnostics, France).

Antibiotic susceptibility testing

The minimum inhibitory concentration values (MIC) of S. aureus against 14 antibiotics, i.e., clindamycin, penicillin, erythromycin, gentamicin, linezolid, oxacillin, rifampicin, cotrimoxazole, tetracycline, vancomycin, ciprofloxacin, levofloxacin, moxifloxacin, and cefoxitin were determined using VITEK® AST-GP67 card (bioMérieux Clinical Diagnostics, France), and interpreted according to the Clinical Laboratory Standards Institute (CLSI-M100) guidelines [24]. Methicillin susceptibility of the clinical isolates was determined using the CLSI disk–diffusion method with cefoxitin 30-µg disk and VITEK® AST-GP67 card. For 10 isolates with missing MIC data, vancomycin susceptibility testing was carried out using microbroth dilution method, as recommended by CLSI-M100 guidelines.

Biofilm quantitation assay

Quantitation of S. aureus biofilm production was performed as described by Atshan et al. [22] and Stepanović et al. [25], with slight modifications. Briefly, 100 µl of bacterial suspension (adjusted to 1 × 106 CFU/ml in Mueller Hinton broth containing 1% glucose) were seeded into each well of a sterile 96-well flat bottom microtitre plate (BIOFIL®, Guangzhou, China) and incubated at 37ºC for 24 h. After incubation, the wells were washed thrice, fixed with methanol, and stained using 0.1% (v/v) crystal violet (Cat. No: C6158, Sigma, USA). S. aureus ATCC® 29213™ (MSSA) and ATCC® 33591™ (MRSA) were used as biofilm-producing controls, while microtiter wells with no inoculum served as negative controls. The amount of biofilm was quantitated by measuring the absorbance of each well at 570 nm using a microplate reader (Tecan, Sunrise™, Swiss). Biofilm was graded into four categories as described by Moghadam et al. [26]: no biofilm (ODs ≤ ODc), weak (ODc ≤ ODs ≤ 2 × ODc), moderate (2 × ODc ≤ ODs ≤ 4 × ODc), and strong (4 × ODc < ODs). ODc and ODs represent the OD of the negative and the test isolates, respectively.

Congo red agar assay for determination of slime production

Bacterial slime production was determined qualitatively as described by Freeman et al. [27] and Thilakavathy et al. [28]. Congo red agar was prepared using brain heart infusion (BHI) broth (37 g/L), sucrose (50 g/L), agar no.1 (10 g/L), and Congo red stain (0.8 g/L). Slime producers are expected to form black colonies with a dry, crystalline consistency, while non-slime producers form pink coloured colonies. Intermediate slime production is indicated by the growth of smooth blackish-red colonies. The positive and negative control strains included in the test were Staphylococcus epidermidis ATCC® 35984™ and Staphylococcus hominis ATCC® 35982™, respectively.

Bacterial genomic DNA extraction

Genomic DNA was extracted from overnight cultures of S. aureus in Luria–Bertani broth, using either MasterPure™ Complete DNA and RNA Purification Kit (Lucigen, Middleton, WI, USA) or QIAamp DNA Mini Kit (Qiagen, Germany) following manufacturers’ instructions. Amplification of the 16S rRNA gene from the bacterial DNA extract was performed to rule out the possibility of having PCR inhibitors, using universal oligonucleotide primers (27F and 1492R) as described by Gumaa et al. [29].

PCR detection of biofilm-associated genes

PCR profiling of bap, cna, icaA, and icaD genes was performed using singleplex PCR assays, while ebpS, eno, fnbA, clfA, clfB, fib, and fnbB genes were amplified using multiplex PCR assays as described by Tristan et al. and Vancraeynest et al. [30, 31]. The primers and PCR thermal cycling conditions are shown in Additional file 1: Table S2. sarA gene was amplified using sarAF and sarAR primers as described by Gowrishankar et al. [32]. Meanwhile, agr typing (types I–IV) was performed using primers and amplification conditions as described by Shopsin et al. [19]. The amplified products were then subjected to electrophoresis using 1% (w/v) agarose gel, pre-stained with nucleic acid staining dye (Bioteke Corporation, China). Sequence analyses were performed to confirm that correct genes were amplified.

Statistical analysis

Paired sample t tests were used to compare biofilm and slime production between MSSA and MRSA isolates. Pearson’s Chi-square test was used to determine the correlation of antibiotic resistance with other parameters. Statistical analysis was performed using SPSS software version 20.0 (IBM, Armonk, USA). A p value of less than 0.05 was considered statistically significant.

Results

Antibiotic susceptibility profiling of S. aureus clinical isolates

MRSA isolates exhibited higher rates of resistance to erythromycin (53.7% vs 17.6%), ciprofloxacin (83.3% vs 2.9%), levofloxacin (83.3% vs 1.5%) and moxifloxacin (75.9% vs 0%), compared to MSSA isolates (Table 1). Clindamycin resistance was observed in 16.2% and 7.6% of MSSA and MRSA isolates, respectively, while inducible clindamycin resistance was detected in 23 (42.6%) MRSA isolates and 1 (1.5%) MSSA isolate. The MRSA MIC90s against clindamycin (0.5 vs 8 µg/ml), erythromycin (0.5 vs 8 µg/ml), gentamicin (0.5 vs 8 µg/ml), ciprofloxacin (0.5 vs 8 µg/ml), and levofloxacin (0.25 vs 8 µg/ml) were 16–32 folds higher than those of MSSA isolates (Additional file 1: Table S3). Meanwhile, all isolates exhibited high susceptibility towards linezolid (100%), vancomycin (100%), rifampicin (99.2%), cotrimoxazole (86%), tetracycline (84.4%) and gentamicin (83.6%). In this study, no isolate showed resistance to vancomycin and linezolid. The MRSA vancomycin and linezolid MICs ranged from 0.5 to 2 μg/ml and 1 to 2 μg/ml, respectively.

Table 1 Antibiotic susceptibility profiles of S. aureus isolates investigated in this study
Full size table

Biofilm production of MRSA and MSSA isolates

Of the 122 S. aureus isolates tested, a majority (79.5%) were identified as strong biofilm producers. A total of 57 (83.8%) biofilm-producing isolates were MSSA and 41 (75.9%) isolates were MRSA (Table 2). In addition, 12.3% of S. aureus isolates were identified as moderate biofilm producers, 5.7% were identified as weak biofilm producers and 1.64% of strains did not produce biofilms.

Table 2 Distribution of biofilm and slime producers among MRSA and MSSA isolates
Full size table

Slime production of MRSA and MSSA isolates

Using Congo Red agar assay, most S. aureus isolates (72.1% MSSA and 72.2% MRSA isolates, respectively) were regarded as intermediate slime producers. There was no significant difference between MSSA and MRSA isolates in slime production (p = 0.19). Only 4 (6.0%) MSSA isolates demonstrated strong slime production after 24 h of incubation (Table 2).

Distribution of biofilm-associated genes and agr/sar quorum sensing genes in MSSA and MRS isolates

In this study, the successful amplification of the 16S rRNA gene from all S. aureus isolates indicated the absence of PCR inhibitors in the bacterial DNA extracts. The amplification of biofilm-associated genes from MSSA and MRSA isolates using various singleplex and multiplex PCR assays is shown in Additional file 1: Fig. S1.

The presence of the intracellular adhesion A (icaA) gene was observed in all S. aureus isolates (100%). There was variability in the distribution of other biofilm-associated genes in MSSA and MRSA (Fig. 1). Overall, the intracellular adhesion A and D (icaA and icaD), laminin-binding protein (eno), clumping factors A and B (clfA and clfB), and fibronectin-binding protein A (fnbA) were the most prevalent biofilm-associated genes in S. aureus isolates, regardless of MSSA or MRSA.

Fig. 1
figure 1

Prevalence of biofilm-associated genes among MSSA and MRSA clinical isolates. Paired t tests were performed to determine significant difference between MSSA and MRSA isolates. (*p < 0.05, **p < 0.01, ***p < 0.001)

Full size image

Compared to MSSA, the detection rates of agrI, icaD, cna, clfA, and clfB genes were significantly higher in MRSA, while the fibronectin-binding protein B (fnbB) gene was absent in all MRSA isolates. All MSSA isolates harbored the laminin-binding protein (eno) gene, while all MRSA isolates harbored intracellular adhesion D (icaD), clumping factors A and B (clfA and clfB) genes. Intriguingly, the bap gene (encoding biofilm matrix protein) was not amplified from any of the isolates.

The number of biofilm-associated genes detected in S. aureus varied from three to eleven, with most isolates having 10 genes (including 24 MSSA and 10 MRSA isolates). However, the number of genes detected from an isolate was not significantly associated with biofilm production (p = 0.299, Pearson’s Chi-square, Table 3). Interestingly, the presence of agrI in MSSA (p = 0.018), and ebpS in MRSA isolates was significantly associated with biofilm production (p = 0.006) (Table 4).

Table 3 Cross-tabulation between the number of S. aureus biofilm-associated genes and biofilm production
Full size table
Table 4 Correlations between agr, sarA, and biofilm-associated genes, with biofilm production in 68 MSSA and 54 MRSA isolates investigated in this study
Full size table

In this study, the most prevalent agr type in S. aureus isolates was agrI (56.7%), followed by agrIII (20.5%) and agrII (9.0%). The agrI was detected with a significantly higher rate in MRSA (81.5%) as compared to MSSA (36.8%). In contrast, higher detection rates of agrIII and agrII were found in MSSA (26.5% and 14.7%, respectively) as compared to MRSA (13% and 1.9%, respectively). The agrIV was only detected in only one MSSA isolate (1.5%). Sequence analyses of representative agr alleles in this study demonstrated 100% similarity to agrI (352/352, 100%, GenBank accession no. AJ617710), agrII (472/472, 100%, GenBank accession no. AJ6177170), agrIII (333/333, 100%, GenBank accession no. AJ617723) and agrIV (577/577, 100%, GenBank accession no. AJ617712), as reported by Goerke et al. [33]. In this study, the presence of agrI was significantly correlated with ciprofloxacin (p = 0.000), levofloxacin (p = 0.003), moxifloxacin (p = 0.000), oxacillin (p = 0.000) and cefoxitin (p = 0.000) resistance (Table 5).

Table 5 Association between the presences of agrI with antibiotic susceptibility in S. aureus clinical isolates
Full size table

Discussion

The treatment and management of S. aureus infection pose significant challenges and a big threat in healthcare settings worldwide due to the emergence of antibiotic-resistant strains. In comparison with the Malaysia National Surveillance of Antimicrobial Resistance (NSAR) 2022 report [34], higher resistance rates to clindamycin (12.4% vs 5.9%), erythromycin (33.6% vs 9.9%) and gentamicin (13.1% vs 3.2%) were reported from a collection of clinical S. aureus isolates in this study. No linezolid-resistant strain was identified in this study, consistent with the very low percentage of linezolid resistance (0.4%) documented in the latest national report [34]. So far, the highest linezolid resistance rate was reported in a previous NSAR study (2010) whereby 7.7% in MRSA and 3.3% in MSSA were linezolid resistant [1], while there have been no studies documenting S. aureus resistance to vancomycin in Malaysia [35, 36].

In addition to antibiotic resistance, almost 80% of S. aureus isolates (MSSA and MRSA) in this study exhibited slime and biofilm production. However, no correlation was found between slime and biofilm production among staphylococcal isolates investigated in this study (Table 2). Similar observations have been reported for S. aureus human and animal isolates in earlier investigations [21, 37]. The lack of correlation between slime and biofilm production in S. aureus may be attributed to different measurement methods, i.e., Congo red agar method versus microtiter plate-based crystal violet assay, leading to disparities in the results. In addition, the complex nature of biofilm formation, possibly affected by bacterial genetic diversity, environmental factors, and regulatory mechanisms, may be attributed to the limited correlation between slime and biofilm production in S. aureus.

The most prevalent biofilm-associated genes detected in MRSA isolates in this study were intracellular adhesion A and D (icaA and icaD), laminin-binding protein (eno), clumping factors A and B (clfA and clfB), and fibronectin-binding protein A (fnbA), as shown in Fig. 1. The agrI, icaD, cna, clfA, and clfB genes were detected at significantly higher rates amongst MRSA isolates, while fnbB was detected at a significantly higher rate in MSSA isolates. The variability observed in the frequencies of biofilm-associated genes could be attributed to strain-to-strain difference [22, 38], source of isolation [39], and geographical settings [40]. Amongst the biofilm-associated genes, the elastin-binding protein (ebpS) gene has been significantly associated with biofilm production amongst MRSA isolates in this study (Table 3). Elastin-binding protein facilitates S. aureus-binding to elastin-rich tissues and promotes bacterial colonisation on mammalian tissues [41]. It has been significantly associated with strong biofilm production in S. aureus food isolates in two previous studies [38, 42].

The distribution of agr types is variable in S. aureus from different geographical regions [43]. In this study, the most prevalent agr type identified from S. aureus isolates was agrI (56.7%), followed by agrIII (20.5%) and agrII (9.0%), while agrIV (0.8%) has a low occurrence rate. Remarkably, a significantly higher percentage of MRSA isolates in this study was found to harbor agrI, compared to MSSA. The presence of agrI has been significantly associated with biofilm production amongst MSSA isolates in this study (Table 3), corresponding well with another study using nonclinical isolates [42]. Kawamura et al. [44] found that MRSA isolates haboring agrII have a significantly greater ability to produce biofilm, however; Usun Jones et al. [21] and Cha et al. [45] found no variation in MRSA biofilm production among different agr groups. The difference might be attributed to variations between strains, potentially resulting from microbial adaptation and geographical influences.

As the transcription of the agr locus (I–IV) is auto-inducing peptide (AIP)-dependent, the differentiation of staphylococcal strains based on agr typing may provide further insights into the epidemiology and antibiotic resistance. Studies have shown that the mecA gene of MRSA indirectly activates AIPs which significantly affect biofilm production, quorum-sensing and virulence, and antibiotic resistance [17, 18]. As quorum sensing is higly influenced by cell density, high-density colonies can produce numerous small molecule signals, triggering downstream processes, such as virulence and antibiotic resistance mechanisms, which poses a threat to the host and antibiotic efficacy [46]. Biofilm production has been reported to provide a niche for generation of antibiotic resistant subpopulations or persister cells through the exchange of genetic materials [47]. Recent data demonstrated a significant correlation between agrI with cefoxitin and erythromycin resistance [48], as well as tetracycline, erythromycin, clindamycin, and ciprofloxacin resistance in S. aureus [43]. Interestingly, a significant association was found between agrI with fluoroquinolones (ciprofloxacin, levofloxacin, and moxifloxacin) resistance (p < 0.05) for the first time in this study. In addition, the high resistance (75.9%) of MRSA against fluoroquinolones especially moxifloxacin, a fourth-generation fluoroquinolone, is alarming (Table 1).

Fluoroquinolone exposure has been identified as an increased risk factor for MRSA isolation and infection [49,50,51]. The key mechanims to S. aureus fluoroquinolone resistance are through chromosomal point mutations in gyrA/B (DNA gyrase subunits), grlA/B (DNA topoisomerase IV subunits), and the promoter region of norA efflux pump [52]. The accumulation of such mutations may be enhanced in biofilm producing agr1-habouring strains, contributing to a high level of resistance to fluoroquinolones, as observed in the MRSA isolates in this study. However, more extensive studies are required to explore the linkage between agr1, biofilm production and fluoroquinolone resistance.

One of the limitation of this study is its confinement to a single-center setting and convenient sampling of S. aureus isolates, thus the ratio of MSSA to MRSA might not reflect the actual prevalence of multidrug resistant S. aureus in the local setting. For more comprehensive insights, future studies are recommended to include diverse sampling methods and multiple centers, to ensure a more representative analysis of the genetic diversity and prevalence of biofilm-associated genes in the Malaysian isolates. As the antibiotic susceptibility profiling of S. aureus isolates was limited to planktonic cells, future reserach should also include comprehensive assessment of antibiotic susceptibility within biofilm structures to enhance understanding of their impact on biofilm-associated S. aureus infections. In addition, the utilization of mec (SCCmec) typing would be beneficial for identifying distinct MRSA types and establishing correlations with other study variables. As conventional antibiotics do not work effectively against S. aureus biofilm infection, new therapeutic strategies and infection control practices are urgently needed. The genetic profiling of biofilm-associated genes and quorum sensing systems of S. aureus isolates has provided scientific foundation for developing a more targeted approach for surveillance, and treatment against biofilm infection in our clinical setting.

Conclusion

The emergence of multidrug-resistant S. aureus strains has been driven by the use of multiple antibiotic classes over the years. The high rates of resistance against clindamycin, erythromycin, and fluoroquinolones as reported in this study have called for more judicious use of antibiotics for treatment of MRSA infection in this region. More importantly, the identification of prevalent biofilm-associated genes and agr types associated with antibiotic resistance in this study has shed valuable genetic insights into S. aureus biofilm formation, which are important to tailor more focused surveillance and treatment strategies against S. aureus biofilm infection in our setting.