Antibiotic resistance, biofilm forming ability, and clonal profiling of clinical isolates of Staphylococcus aureus from southern and northeastern India

Abstract Background Staphylococcus aureus is a pathogen endemic in India and sometimes deadly for patients in intensive care units. Objectives To determine the antibiotic-resistance pattern, biofilm forming ability, and clonal type of S. aureus from isolates collected in Tamil Nadu (south) and the Mizoram (northeast) regions of India. Methods We collected S. aureus isolates from diagnostic laboratories in Tamil Nadu and Mizoram. An antibiotic susceptibility test was performed according to Clinical Laboratory and Standards Institute methods. Antibiotic-resistant determinants such as mecA, mecC, blaZ, vanA, vanB, and vanC were confirmed by polymerase chain reaction (PCR). All isolates were further studied for biofilm forming ability. Enterobacterial repetitive intergenic consensus (ERIC)-PCR was used for clonal analysis. Results A study of 206 clinical isolates showed 52.9% prevalence of methicillin-resistant S. aureus in Tamil Nadu and 49.4% in Mizoram. Minimum inhibitory concentration tests showed a high prevalence of 67% oxacillin resistance in isolates from Tamil Nadu and 49% in isolates from Mizoram. PCR showed 53% mecA in Tamil Nadu and 49% mecA in Mizoram. Vancomycin-intermediate resistance S. aureus (VISA) prevalence was lower in isolates from Tamil Nadu (4%) and Mizoram (5%). All methicillin-resistant S. aureus (MRSA) isolates formed biofilms. Clonal analysis revealed a genetic relatedness between the isolates. Conclusions The prevalence of MRSA is high in the regions studied, with most of the clinical isolates being multidrug resistant. Adopting appropriate community-based preventive measures and establishing antimicrobial stewardship is highly recommended to minimize the dissemination in antibiotic resistance.

Globally, antibiotic-resistance patterns and the spread of epidemic clones are continuously monitored by the 2 bodies, the U.S. Centers for Disease Control (CDC) and the World Health Organization (WHO). In India, antimicrobial surveillance is monitored by the Indian Network for Surveillance of Antimicrobial Resistance (INSAR) and Antimicrobial Resistance Research and Surveillance Network (ARRSN), which are actively involved in collecting and analyzing the results from various regional centers located across India. The 2020 annual report by the ARRSN has shown that resistance trends in S. aureus have increased to most antibiotics over the past years. The report also claims that isolation rates or incidence of S. aureus have steadily declined from 2016 to 2020 [5].
India is a geographically wide and densely populated country divided into various regions and states. The distribution of antibiotic-resistant S. aureus in India is known to vary by location and time. The overall prevalence of S. aureus varies by year, with the highest being 40%-50% observed in 2008 and the lowest prevalence of 33% reported in 2016. Recently, from 2018 to 2020, the incidence of MRSA is known to have reduced despite high resistance [6,7]. The current resistance trend of S. aureus in India is due primarily to community-associated (CA)-MRSA [8]. The INSAR found that CA-MRSA is highly distributed among specimens obtained from throat, genital, nasal, and ear swabs of outpatients, and healthcare-associated (HA)-MRSA was more prominent in skin and soft tissue infections, the bloodstream, catheter tips, fluids from sterile body sites, and lower respiratory tract infections of inpatients [6].
The prevalence of the gene mecA in clinical samples among Indian populations is frequently documented; whereas, to our knowledge, there are no reports to date on the prevalence of the mecC in clinical samples from India. Similarly, reports of VRSA are scant. From India, only 2 isolates containing vanA have been reported up to 2015 [9]. In addition to the presence of resistance factors, biofilm forming infections may be critical and challenging for antibiotic treatment. The ability of S. aureus to form biofilms is well-documented [10]. Biofilm forming infections are considered life-threatening because of their complicated architecture that prevents drug penetration. According to a report from the U.S. National Institutes of Health, antibiotic failure may be the cause of death in more than 60% of biofilm forming infections. Therefore, studying the nature of biofilms and their association with antibiotic resistance is critical for addressing biofilm forming infections [11].
The present study aimed to determine the prevalence of MRSA and vancomycin-intermediate S. aureus (VISA) in Mizoram and Tamil Nadu. The objectives of the present investigation were to (1) to assess antibiotic susceptibility patterns of S. aureus to determine the incidence of MRSA in the regions studied; (2) to investigate the nature of biofilms formed by resistant and sensitive isolates of S. aureus; and (3) to assess the genetic relatedness of the bacteria to understand the clonal dissemination of S. aureus in Tamil Nadu and Mizoram.

Study design
A total of 206 nonduplicate S. aureus isolates was collected during the study period from 2018 to 2019. We conducted this cross-sectional study with the isolates collected from Hi-Tech diagnostic centers in Chennai and Tiruchirappalli, in Tamil Nadu (southern region), and a clinical diagnostic laboratory in Aizawl, in Mizoram (northeastern region). We choose the largest, most densely populated cities in the 2 states and collected isolates at various times throughout 2018 to 2019. The diagnostic centers were selected based on the number and diversity of samples received and ensured that a pan regional representation of isolates was attained. The Institutional Ethical Committee (IEC) of Vellore Institute of Technology, Vellore approved the study protocol (VIT/IECH/004/ Jan28.2017). The present study involved laboratory-based analysis of anonymized isolates only. This reporting of the obtained microbiological data is based on the MICRO (Microbiology Investigation Criteria for Reporting Objectively) and STROBE-AMS (Strengthening the Reporting of Observational Studies in Epidemiology-antimicrobial stewardship) statement reporting guideline checklists [12,13].

Isolate collection and processing
All S. aureus isolates were identified using a VITEK 2 system (bioMérieux) and were transported from the diagnostic centers to the Antibiotic Resistance and Phage Therapy Lab, Vellore Institute of Technology in leak proof, screw capped, cryovial tubes and processed using standard microbiological techniques. Upon receiving the isolates at the Vellore Institute of Technology laboratory facility, we streaked all the isolates onto brain-heart infusion agar. Each isolate was assigned a unique identifier number (SA-1 to SA-206, where SA indicates Staphylococcus aureus) and stored in 20% glycerol at low temperature for further analysis. All reported laboratory studies of antibiotic profiling by phenotypic and genotypic analysis were conducted by the Antibiotic Resistance and Phage Therapy Lab, Vellore Institute of Technology. According to the proposed study, MRSA is defined as the presence of mecA validated by polymerase chain reaction (PCR).

Sample source of the isolates
Of 206 S. aureus isolates, 119 isolates were collected from Tamil Nadu, and 87 isolates were collected from Mizoram. These isolates were obtained from clinical sources such as pus, sputum, wound sepsis, urine, catheter tip, blood, synovial fluid, throat swab, wound swab, bronchial fluid, umbilical swab, dysuria, and stool. All the isolates were identified using the VITEK 2 GP ID card system (bioMérieux).
The methicillin sensitive S. aureus subsp. aureus Rosenbach (Seattle 1945) from the American Type Culture Collection (ATCC) No. 25923 was used as a quality control strain. Experiments were conducted and interpreted following Clinical and Laboratory Standards Institute (CLSI) guidelines [14]. The antibiotic resistance of each isolate from the disk-diffusion test was recorded independently and multiple-antibiotic resistance (MAR) indices were calculated. The MAR index is defined as "a/b, where a represents the number of antibiotics to which the isolate was resistant, and b represents the number of antibiotics to which the isolate was exposed" [15].

Minimum inhibitory concentration
The minimum inhibitory concentration (MIC) of antibiotics was determined following the CLSI M07 guidelines [14]. Briefly, the MIC of the isolates was determined using a microbroth dilution method in MHA No. 2 cation-controlled broth. The antibiotics oxacillin, ciprofloxacin, and vancomycin were used at concentrations ranging from 128 mg/mL to 0.125 mg/mL. The findings were interpreted following the CLSI guidelines [14]. S. aureus ATCC 25923 was used as the standard laboratory testing quality control strain.
The MIC 50 and MIC 90 were calculated based on the method described by Schwarz et al. [16] to determine the lowest concentrations of an antibiotic capable of inhibiting 50% and 90% of the bacterial isolate.

Genomic analysis
Molecular analysis by conventional PCR was used to screen for resistance genes using the following antibiotic-resistant determinants: penicillin-binding protein genes (mecA and mecC) [17,18], vancomycin-resistance genes (vanA, vanB, vanC) [19], and b-lactamase gene (blaZ) [20]. The primer pairs used in this analysis were synthesized according to sequences published in the literature. The list of primers used in the study is tabulated ( Table 1). The bacterial DNA was isolated using a phenol-chloroform method, and the extracted DNA was used in the PCR reaction as template DNA. The amplified PCR product was analyzed by 1% agarose gel electrophoresis.

Biofilm analysis-microtiter plate assay
Biofilm formation in vitro was measured by absorption spectroscopy of crystal violet-stained biomass in a 96-well polystyrene microtiter plate assay. Biofilm formation was studied by mimicking the internal body system, in which the polystyrene surface of a 96-well plate was primed with media containing human plasma using a protocol described by Cardile et al. [22]. Briefly, pooled human plasma was diluted to 10% with 50 mM sodium bicarbonate (as a buffering agent). Each well in a microtiter plate was incubated overnight at 4 °C with 100 mL of the 10% plasma. Following the cold incubation, the diluted plasma was removed, and the wells were refilled with 100 mL freshly prepared tryptic soy broth supplemented with 1% glucose. Overnight culture of S. aureus was diluted (1:10), and 100 mL of the culture was added to the wells. Biomass was allowed to accumulate at 37 °C for 24 h. After incubation, the wells were gently washed 3´ with sterile phosphate-buffered saline (7 mM Na 2 HPO 4 , 3 mM NaH 2 PO 4 , and 130 mM NaCl at pH 7.4) to remove planktonic bacteria and the plates air-dried. Bacteria adherent to the wells were stained for 5 min with 125 mL of a 0.1% solution of crystal violet (Gentian Violet, CAS No. 548-62-9). Finally, 125 mL of 30% acetic acid was added to solubilize the dye, and the optical density (OD) of the well contents determined using a microplate reading spectrophotometer at 570 nm. Acetic acid (30%) served as a blank control, and the broth served as an experimental control. Biofilm formed was categorized as follows: OD test isolate (OD t ) < OD control blank (OD c ) was considered no biofilm formation, OD c < OD t < 2 ´ OD c was considered a weak biofilm producer, 2 ´ OD c < OD t < 4 ´ OD c was considered a moderate biofilm producer, and 4 ´ OD c < OD t was considered a strong biofilm producer.

Data analysis
All statistical analyses were performed using GraphPad Prism (version 9). MAR indices were determined following the procedure described by Krumperman [15] and Jaja et al. [23]. Significant differences between categorical variables (susceptibility to antibiotic) were determined using a one-way analysis of variance (ANOVA) or a Brown-Forsythe test. The association between the MAR index and antibiotic resistance of isolates from the regions studied was assessed using a c 2 test. P < 0.05 was considered significant. DNA fingerprints were analyzed using BioNumerics (version 8.0, Applied Maths, bioMérieux). Gel images were captured in grayscale TIFF format and imported into the Bio-Numerics database to develop a dendrogram. After alignment and normalization, the similarity index was computed and visualized using cluster analysis. For clonal analysis, the Dice coefficients and unweighted pair group average (UPGMA) were used to measure similarity index, with an index of ³90% being considered the same ERIC type. Banding patterns were defined as present (score = 1) or absent (score = 0).
Disk-diffusion tests showed isolates were resistant or susceptible to various antibiotics ( Table 2). Of the antibiotics tested, isolates showed the highest resistance to penicillin, followed by resistance to erythromycin, and ciprofloxacin, and were highly susceptible to linezolid. MAR indices for the isolates from Tamil Nadu ranged from 0.18 to 0.91, with 118 isolates showing a score >0.25. All 119 isolates were resistant to at least 1 antibiotic under study. The association between the MAR index and antibiotic resistance of isolates from Tamil Nadu was significant (P = 0.001). There were 117 (98%) multidrug resistant (MDR) isolates. As observed from the MIC, 80 (67%) isolates showed resistance to oxacillin, 101 (85%) isolates were resistant to ciprofloxacin antibiotic, and none showed vancomycin resistance. However, 5 (4%) isolates were vancomycin-intermediate resistant S. aureus (VISA).  The MIC 50 and MIC 90 , respectively, were 1 mg/mL and 0.5 mg/ mL for oxacillin, 16 mg/mL and 16 mg/mL for ciprofloxacin, and 2 mg/mL and 2 mg/mL for vancomycin. PCR revealed that 63 (53%) isolates carried mecA, 111 (93%) carried blaZ, and none carried mecC. Vancomycinresistance genes vanA, vanB, and vanC screened for in the VISA isolates were absent.
We analyzed the 119 isolates from Tamil Nadu for genetic relatedness by multiple banding similarity patterns in ERIC-PCR and found 8 distinct clusters (I-VIII). Clusters I, IV, and VI had the least number of methicillin-resistant strains at 1 each, with I and VI having 1 MRSA and IV having 1 methicillin-susceptible S. aureus (MSSA). Clusters II and V each had 3 strains of 1 MRSA and 2 MSSA, and 2 MRSA and 1 MSSA, respectively. Cluster III had 6 strains, 3 of which were MRSA. Cluster VII had the most strains at 99, including 55 MRSA, while Cluster VIII had 5 MSSA strains in total (
Of the antibiotics tested in a disk-diffusion test, isolates showed the highest resistance to penicillin, followed by resistance to cefoxitin, and ciprofloxacin, and were highly susceptible to linezolid ( Table 2). MAR indices for the isolates from Mizoram ranged from 0.18 to 1.00, with 85 isolates showing a score >0. 25. The association between the MAR index and antibiotic resistance of isolates from Mizoram was significant (P < 0.003). We found MDR in 82/87 (94%) of the isolates from Mizoram. As observed from the MIC tests, 43 (49%) isolates showed resistance to oxacillin, 65 (75%) isolates were resistant to ciprofloxacin, and none showed resistance to vancomycin. However, 4 (5%) of the isolates were VISA. The MIC 50 and MIC 90 were, respectively, 0.5 mg/ mL and 0.125 mg/mL for oxacillin, 0.125 mg/mL and 0.125 mg/ mL for ciprofloxacin, and 4 mg/mL and 4 mg/mL for vancomycin.
ERIC-PCR analysis of the isolates from Mizoram showed 8 distinct clusters (I-VIII). Clusters II and IV had the least number of MRSA strains at 1 each, cluster III had 2 strains, of which 1 was MRSA, cluster I had 5 strains, of which 1 was MRSA and the other 4 were MSSA, cluster VI had 15 strains, with 6 MRSA, cluster V had 19 strains with 10 MRSA, cluster VII had 21 strains with 11 MRSA, and cluster VIII had the highest number of strains at 23, of which 12 were MRSA (

Discussion
The present article reports the antibiotic-resistance profiling of S. aureus in 2 Indian states: Tamil Nadu in the south and Mizoram in the northeast. We identified a high prevalence of MRSA and MDR isolates in these regions. These data are essential for public health professionals to determine the prevailing pattern of resistance in S. aureus, which helps these professionals frame policies to control the development of antibiotic resistance.
One of the key findings of the present study are the high MAR indices in the isolates studied. High MAR indices (>0.25) suggest that the isolates come from high-risk sources where antibiotics are used frequently, which indicates that antibiotic usage in the studied regions was high.
The data for antibiotic consumption from the Intercontinental Medical Statistics Health MIDAS (Market Information Data Analytics System) database shows that India is the world's greatest consumer of antibiotics, with the highest consumption observed during 2010, accounting for 10.7 units per person [24,25]. Similarly, a recent spatial model study has shown that antibiotic consumption in India has increased from 48% in 2000 to 67% in 2018 [26]. The high MAR indices in most of the isolates indicate the need to establish antibiotic stewardship to reduce the overuse and misuse of antibiotics. We also found a high prevalence of MDR isolates in both regions, which is attributed to the overuse of antibiotics in India. The proportion of oxacillin resistance from the Tamil Nadu region varied considerably according to method in the phenotypic and genotypic analysis. The MIC method found a larger proportion of oxacillin resistance (67%) than PCR (53%). By contrast, for isolates from Mizoram, we found that every oxacillin-resistant S. aureus (ORSA) identified by MIC had mecA. This variation observed in the isolates of Tamil Nadu is considered to be associated with a borderline oxacillin resistance S. aureus (BORSA) phenotype, which results in increased b-lactamase production. In this context, CLSI guidelines discourage the use of oxacillin disk-diffusion tests as a standard method to identify MRSA due to the difference in the inhibitory zone produced by increased b-lactamase production [27]. Due to such phenotypic variations observed in our study, MRSA was reported based on the presence of mecA only, which was validated by PCR. The BORSA phenotype predicted by the present study was not explored as this was out of the scope of the present investigation.
In the present study, the prevalence of MRSA was 52.9% in Tamil Nadu and 49.4% in Mizoram. The prevalence of MRSA reported is similar to that in reports of prevalence between 2014 and 2019, which showed that the prevalence of MRSA in India was high at 53% [28]. By contrast, a systemic review conducted on the prevalence of MRSA in various zones of India showed a lower prevalence in southern India with 34% MRSA [29]. The variation in the MRSA prevalence reported in the present study could be attributed to the regions included in the analysis. In the present report from southern India, we only assessed prevalence in Tamil Nadu, while the systemic review reported the prevalence of MRSA from 6 states in southern India: Tamil Nadu, Telangana, Karnataka Andhra Pradesh, Kerala, and Puducherry [29].
MIC by the broth microdilution method revealed that oxacillin resistance was high among the isolates collected from the Tamil Nadu (67%) compared with Mizoram isolates (49%). Overall, MIC revealed high ciprofloxacin resistance, as ciprofloxacin is frequently prescribed in MSSA infection [3]. The prevalence of VISA at 4% in Tamil Nadu and 5% in Mizoram indicate that vancomycin resistance is emerging in these regions. Thus, usage of vancomycin should be restricted to reserve the antibiotic for serious S. aureus infections. Tiwari and Sen [30] reported similar prevalence of VISA in the northern part of India. Compared with the prevalence of VISA in Europe and the United States, the prevalence of VISA is higher in Asia [31].
We found that the prevalence of biofilm producers is high in isolates from samples taken in Tamil Nadu. Isolates from Tamil Nadu showed greater prevalence of strong biofilm producers (30%), whereas isolates in Mizoram had comparatively weaker biofilm producers (19%). A study conducted in a tertiary care hospital in the Tripura region in 2015 showed 55% (55/100) of the isolates forming biofilms of various strengths [32].
By contrast, our study of isolates obtained between 2018 and 2019 shows that 100% of the isolates produced biofilms of various strengths. Observing such a high prevalence of biofilm producers could be ascribed to the modified biofilm analysis method used in the present study. However, it remains unclear to what degree the human plasma might affect the strength of the S. aureus biofilms. Nevertheless, as consistent with the models of Cardile et al. [22], it can be concluded that adhesive matrix molecules from the plasma enhance biofilm formation by S. aureus.
We observed that all MRSA form biofilms of varying strengths, indicating that methicillin resistance and biofilm development have a strong association. Despite this association, there was a significant variation in the strength of the biofilms (strong and moderate biofilms) produced by MRSA, implying that additional factors may influence the strength of the biofilms formed. According to Pozzi et al. [33], virulence mechanisms including biofilm forming capacity and antibiotic resistance are interrelated in MRSA, in which increasing antibiotic resistance directly influences and attenuates potential virulence, thereby resulting in altered biofilm formation. The interrelationship between antibiotic resistance and biofilm formation is known to be governed by homo-and heteroresistance to oxacillin. While the relationship between other antibiotics such as vancomycin and biofilm formation is less explored as resistance due to vancomycin is rarely encountered, the molecular mechanism involved in interrelationship shows that a deletion in the accessory gene regulator (arg) system of proteases can enhance biofilm formation by clinical strains of MRSA [34]. However, the precise mechanisms in the interrelationship between antibiotic resistance and biofilm formation are not yet determined [34]. As molecular markers were out of the scope of the present study, their interrelationship with antibiotic resistance was not explored here. Our findings revealed a substantial prevalence of MRSA among strong biofilm producers (58%) in Tamil Nadu. By contrast, isolates from Mizoram revealed a high prevalence of MRSA among weak (60%) and moderate biofilm producers (50%). Overall, the prevalence of biofilm formers and methicillinresistant biofilm isolates was high in Tamil Nadu. Similarly, for burn patients in Chennai, 75% of isolates were MRSA and biofilm producers [35].
The similarity index of SA-18 (MRSA) and SA-119 (MSSA) of the Tamil Nadu isolates from cluster II shared a typical banding pattern with SA-168 (MRSA) and SA-122 (MSSA) of Mizoram in ERIC-PCR analysis. Similarly, in cluster III, SA-12 (MRSA) and SA-42 (MSSA) shared a related banding pattern with the Mizoram isolates SA-125 (MSSA) and SA-134 (MSSA), respectively. These data show that the rise of antibiotic resistance of isolates from the Mizoram population might be due to medical tourism, shared borders, and human migration, which results in the dissemination of clones by carriers who may unknowingly shed the pathogen during their travel [28].
A strength of this work lies in the detection of MRSA by the presence of mecA and the enumeration of the variations observed between the phenotypic and genotypic screenings. Observations of the strengths of the biofilms among the MRSA and clonal dissemination between the 2 regions added background evidence to the study.
We acknowledge several limitations of the present study, notably related to the nonuniformity of isolates from each sample source, many cities not being included in the study as the diagnostic laboratories in those regions refused to share the isolates due to policy restrictions on sharing isolates and secondary data, and the genetic relatedness of isolates being studied by ERIC-PCR (a method not established globally) as it was feasible and cost-effective in our laboratory facility.

Conclusions
Our molecular typing suggests that the clonal dissemination between Tamil Nadu and Mizoram could play a crucial role in the spread of antibiotic resistance between these regions. Prevalence of S. aureus infections in the regions studied are high and emphasize the need for periodic surveillance and monitoring of MRSA outbreaks. In this context, we believe that implementation of a "One Health" approach, in which multiple public health sectors work together to bring about better public health could help control the rise of antibiotic resistance.

Authors contributions.
Both authors contributed to the conception and design of the study. AL collected the data and both authors contributed to their analysis and interpretation. AL wrote the initial manuscript draft. RN obtained funding for the study, and evaluated and revised the manuscript. Both authors approved the final version submitted for publication and take responsibility for statements made in the published article.