This study was approved by the Ethics Committee of the Faculty of Veterinary Medicine, Universitas Gadjah Mada, Indonesia (Approval No. EC 143/EC-FKH/Int./2024). Written informed consent was obtained from all animal owners prior to sample collection using standard clinic consent forms. Clinical samples were collected as part of routine veterinary diagnosis, and no additional experimental procedures were performed.

Sample collection involved non-invasive swabbing (nasal, skin/wound, ear, ocular, and perineal) conducted by trained veterinarians using aseptic techniques, causing minimal discomfort and requiring no sedation or euthanasia. Animal welfare was prioritized at all stages. All laboratory work was carried out in accordance with biosafety guidelines for handling potentially zoonotic pathogens. Animal and owner data were anonymized and used solely for the purposes outlined in the approved study protocol.

This cross-sectional study was conducted between September 2024 and July 2025 at the Clinical Pathology Laboratory, Faculty of Veterinary Medicine, and the Animal Science Learning Center (ASLC), Universitas Gadjah Mada, Yogyakarta, Indonesia. A total of 100 clinical swab samples were collected from 26 dogs, 67 cats, and 7 rabbits presenting to veterinary clinics with clinical signs consistent with bacterial infection. Samples were obtained from 11 veterinary clinics and hospitals located in Central Java and the Special Region of Yogyakarta, Indonesia. All animals enrolled in the study exhibited clinical manifestations suggestive of bacterial infection.

Animals presenting with skin or soft-tissue lesions, otitis externa, nasal discharge, conjunctivitis, perineal infection, or wound-related inflammation were eligible for inclusion. Animals that had received systemic antibiotic treatment within the preceding 7 days or had incomplete clinical records were excluded. Duplicate samples collected from the same anatomical site of an individual animal were not included in the analysis.

For each animal, information on species, age, sex, and, when available, breed was recorded along with the clinical diagnostic category. A targeted symptomatic sampling strategy was applied to ensure the inclusion of clinically relevant Staphylococcus isolates. Clinical specimens, including skin, nasal, ear, ocular, and perineal swabs, were aseptically collected using sterile rayon swabs and placed in Amies transport medium. Samples were transported under refrigerated conditions (4°C–8°C) and processed within 4–6 h of collection. When samples were collected from multiple anatomical sites of the same animal, only the site most representative of the lesion was selected to ensure that each isolate originated from a unique animal. The overall study workflow is illustrated in Figure 1.

All clinical swab samples were inoculated onto Mannitol Salt Agar (MSA) (MerckMillipore™, Germany) and incubated at 37°C for 24 h. Colony morphology was examined, followed by Gram staining and catalase testing. Pure colonies were subsequently subjected to coagulase, ortho-nitrophenyl-β-galactoside (ONPG) (Himedia™, India), and Voges–Proskauer (VP) (Himedia™) tests to differentiate S. aureus from S. pseudintermedius after incubation at 37°C for 24–48 h. VP-positive isolates produced a red coloration at the top of the medium, whereas VP-negative isolates showed no color change. S. aureus is VP-positive, while S. pseudintermedius is VP-negative. In the ONPG test, positive isolates produced a yellow coloration, whereas negative isolates remained clear; S. pseudintermedius is ONPG-positive, whereas S. aureus is ONPG-negative [1, 16]. Hemolytic activity was evaluated using 5% sheep blood agar plates (4 mm thickness) (MerckMillipore™) following incubation at 37°C for 24 h.

Genomic DNA was extracted from confirmed S. aureus and S. pseudintermedius isolates using the DNA Isolation Kit (Geneaid Biotech Ltd., Taiwan) according to the manufacturer’s instructions. Extracted DNA was stored at −20°C until further molecular analysis.

Polymerase chain reaction (PCR) assays targeting the 23S rRNA and nuc genes were performed to confirm S. aureus isolates. The 23S rRNA gene is a conserved component of the bacterial ribosome and is commonly used for genus- and species-level identification [16]. The nuc gene encodes a thermonuclease virulence factor and is widely used as a molecular marker to distinguish S. aureus from other Staphylococcus species. Identification of S. pseudintermedius was performed by targeting the pta gene, which contains a restriction site for the MboI enzyme that is absent in other members of the S. intermedius group (SIG). Therefore, pta gene detection was followed by PCR–RFLP analysis to differentiate S. pseudintermedius from other SIG members.

Each 25 μL PCR reaction mixture consisted of 1 μL of forward primer, 1 μL of reverse primer, 12.5 μL of MyTaq Red Mix, 2 μL of DNA template, and 8.5 μL of nuclease-free water. Amplification was carried out using a thermal cycler (Benchmark Scientific, Inc.). PCR products were resolved by electrophoresis on 1.5% agarose gels stained with RedSafe (Intron Biotechnology, Korea) at 100 V for 30 min using Tris–borate–ethylenediamine-tetraacetic acid buffer and visualized under a UV transilluminator [16–18]. Primer sequences and amplification conditions are summarized in Table 1 [16, 18].

Amplified pta gene products were subjected to RFLP analysis using the MboI restriction enzyme (10 U/µL; Thermo Fisher Scientific, Waltham, MA, USA). For digestion, 25 µL of PCR product was mixed with 0.5 µL (5 U) of MboI and 5 µL of 5× digestion buffer and incubated at 37°C for 2 h, corresponding to the optimal activity temperature of the enzyme. Digested products were separated on 1.5% agarose gels (GeneDirex, USA) using a 50 bp DNA ladder (Invitrogen, USA) and visualized under UV illumination.

The 320 bp pta gene fragment of S. pseudintermedius contains an MboI restriction site, resulting in two fragments of 213 bp and 107 bp following digestion, whereas other SIG members lack this site and remain undigested [18, 19].

Antimicrobial susceptibility testing was performed using the Kirby–Bauer disk diffusion method in accordance with CLSI M100 guidelines [19]. The antimicrobial agents tested included oxacillin (5 µg), tetracycline (30 µg), erythromycin (15 µg), cefoxitin (30 µg), ciprofloxacin (5 µg), gentamicin (10 µg), ampicillin (10 µg), clindamycin (10 µg), penicillin G (10 µg), vancomycin (30 µg), and amoxicillin (25 µg). Mueller–Hinton agar plates were maintained at a uniform thickness of 4 mm. Bacterial suspensions were vortexed for 15–20 s, adjusted to 0.5 McFarland turbidity, and evenly inoculated onto agar surfaces using sterile cotton swabs. Plates were incubated at 37°C for 24 h, after which inhibition zone diameters were measured in millimeters and interpreted as susceptible, intermediate, or resistant following CLSI M100 criteria [19].

Isolates identified as S. aureus or S. pseudintermedius were initially screened for methicillin resistance using Oxacillin Resistance Screening Agar Base (ORSAB; Oxoid, UK), a chromogenic medium containing oxacillin and mannitol indicators. The formation of distinct blue colonies was interpreted as presumptive MRSA or MRSP. ORSAB contains oxacillin (2 mg/L) and polymyxin B (50,000 IU/L), allowing selective growth of methicillin-resistant isolates. Blue colony coloration indicates mannitol fermentation and uptake of the aniline dye present in the medium.

Phenotypic confirmation of methicillin resistance was performed by disk diffusion testing using cefoxitin (30 µg) for S. aureus and oxacillin (5 µg) for S. pseudintermedius in accordance with CLSI M100 guidelines [19]. Genotypic confirmation was achieved by PCR amplification of the mecA gene using specific primers (Table 1). Isolates positive for mecA were classified as MRSA or MRSP [20]. MDR was defined as resistance to at least one antimicrobial agent in three or more antimicrobial classes, following the criteria of Magiorakos et al. [21].

In addition to standard reference strains, previously characterized S. aureus isolates reported by Fitranda et al. [16], including validated MRSA and MSSA strains, were used as quality control (QC) isolates. These strains had been confirmed by biochemical identification, cefoxitin or oxacillin susceptibility testing, and mecA PCR and were routinely employed to verify culture performance, PCR accuracy, and antimicrobial susceptibility testing reproducibility. The QC S. aureus isolates had undergone whole-genome sequencing (WGS) as part of laboratory validation, providing high-resolution confirmation of species identity and resistance gene content, thereby ensuring methodological robustness and minimizing analytical variation.

All data were compiled and cleaned using Microsoft Excel 2021 and analyzed using IBM SPSS Statistics version 26 (IBM Corp., Armonk, NY, USA) and GraphPad Prism version 10 (GraphPad Software, San Diego, CA, USA).. Continuous variables are presented as mean ± standard deviation, while categorical variables, including species distribution, prevalence of S. aureus and S. pseudintermedius, and proportions of MRSA, MRSP, and MDR isolates, are expressed as frequencies and percentages. Percentages were calculated using both overall sample denominators (n = 100) and species-specific denominators for AMR analyses. Ninety-five percent confidence intervals (95% CI) were calculated for key prevalence estimates. Comparative analyses between species and sample types were conducted using the chi-square test or Fisher’s exact test when expected cell counts were <5. Statistical significance was set at p < 0.05.

A total of 100 clinical swab samples were collected from 67 cats (67%), 26 dogs (26%), and 7 rabbits (7%). Most animals presented with mild to moderate clinical infections, including skin or soft-tissue lesions, otitis externa, respiratory or nasal discharge, conjunctivitis, and perineal dermatitis. The median age of the sampled animals was 1.5 years (range: 1 month–14 years). Overall, 58% of the animals were male and 42% were female. Sample types consisted of nasal swabs (42%), skin swabs (19%), ear swabs (18%), ocular swabs (16%), and perineal swabs (5%). A summary of sample categories and demographic characteristics is presented in Table 2.

Based on biochemical characteristics, 41 isolates (41%) were identified as S. aureus and 14 isolates (14%) as S. pseudintermedius. Molecular confirmation was subsequently performed. S. aureus isolates were verified by PCR amplification of the 23S rRNA (1250 bp) and nuc (279 bp) genes (Figure 2). In contrast, S. pseudintermedius isolates were confirmed using PCR–RFLP analysis of the pta gene with the MboI restriction enzyme, which yielded two characteristic fragments of 213 bp and 107 bp (Figure 3).

Figure 4 illustrates the distribution of confirmed S. aureus and S. pseudintermedius isolates according to sample origin. S. aureus was most frequently isolated from feline nasal swabs (17/41; 41.46%), followed by rabbit skin swabs (6/41; 14.63%), feline skin swabs (5/41; 12.20%), canine skin swabs (4/41; 9.76%), canine ocular swabs (4/41; 9.76%), feline ear swabs (2/41; 4.88%), canine ear swabs (1/41; 2.44%), and feline ocular swabs (1/41; 2.44%).

Assessment of hemolytic activity on sheep blood agar revealed heterogeneous patterns among isolates. S. aureus isolates exhibited predominantly β-hemolysis in 56.10% (23/41), followed by γ-hemolysis in 39.02% (16/41) and α-hemolysis in 4.88% (2/41). In contrast, S. pseudintermedius isolates most frequently demonstrated α-hemolysis (57.14%, 8/14), followed by β-hemolysis (28.57%, 4/14) and γ-hemolysis (14.29%, 2/14).

These findings indicate that nasal swabs from cats and skin samples from rabbits were the most common sources of Staphylococcus isolates. The observed differences in hemolytic patterns between S. aureus and S. pseudintermedius suggest species-specific expression of hemolysins that may be influenced by host adaptation. Further genomic characterization of hemolysin-associated genes was not conducted and is acknowledged as a limitation of this study.

The antimicrobial susceptibility profiles of S. aureus and S. pseudintermedius isolates are summarized in Table 3 [19] and illustrated in Figure 5. Antimicrobial susceptibility testing was performed using the Kirby–Bauer disk diffusion method against 11 antibiotics representing multiple antimicrobial classes.

Among S. aureus isolates, high resistance rates were observed against penicillin (78%), ampicillin (76%), and amoxicillin (68%). Moderate resistance was detected to clindamycin (49%), tetracycline (49%), cefoxitin (41%), erythromycin (32%), oxacillin (29%), gentamicin (27%), vancomycin (22%), and ciprofloxacin (15%).

Similarly, S. pseudintermedius isolates demonstrated pronounced resistance to penicillin (93%), amoxicillin (79%), and ampicillin (71%). Moderate resistance was observed to tetracycline (57%), clindamycin (50%), oxacillin (43%), cefoxitin (43%), erythromycin (43%), vancomycin (29%), gentamicin (21%), and ciprofloxacin (14%).

Overall, 93% of S. aureus (38/41) and S. pseudintermedius (13/14) isolates were classified as multidrug-resistant (MDR), exhibiting resistance to three or more antimicrobial classes. MDR was more prevalent among MRSA and MRSP isolates than among MSSA and MSSP isolates when stratified (Table 2). These findings highlight widespread β-lactam resistance and underscore the importance of prudent antimicrobial use in companion animals to limit the dissemination of resistant Staphylococcus strains. Minimum inhibitory concentration (MIC) testing was not performed and is acknowledged as a limitation.

Initial screening using Oxacillin Resistance Screening Agar Base (ORSAB) indicated a high proportion of presumptive methicillin-resistant isolates, with 95% (39/41) of S. aureus and 79% (11/14) of S. pseudintermedius producing characteristic blue colonies. However, phenotypic confirmation by disk diffusion yielded lower resistance rates, with only 41% (17/41) of S. aureus classified as MRSA based on cefoxitin resistance and 43% (6/14) of S. pseudintermedius classified as MRSP based on oxacillin resistance (Figure 5).

Genotypic analysis provided greater sensitivity, with PCR detection of the mecA gene confirming methicillin resistance in 66% (27/41) of S. aureus and 93% (13/14) of S. pseudintermedius isolates through amplification of the 532 bp mecA fragment (Figure 6).

Comparison of screening, phenotypic, and genotypic methods revealed several phenotype–genotype discrepancies, including mecA-positive isolates that remained susceptible to cefoxitin or oxacillin and isolates exhibiting phenotypic resistance without detectable mecA (Table 4). These discordant profiles likely reflect heterogeneous expression of penicillin-binding protein 2a (PBP2a), borderline resistance, or alternative β-lactam resistance mechanisms.

Notably, eleven S. aureus and seven S. pseudintermedius isolates were mecA-positive but phenotypically susceptible to cefoxitin or oxacillin. Such discrepancies have been previously reported in staphylococci isolated from companion animals and may result in underestimation of methicillin resistance when reliance is placed solely on phenotypic assays. In contrast, no isolates in this study were mecA-negative while displaying phenotypic resistance, suggesting the absence of borderline oxacillin-resistant strains or alternative resistance mechanisms such as β-lactamase hyperproduction. Although mecC was not included in the molecular panel, the presence of undetected mecC-harboring isolates cannot be excluded and is acknowledged as a methodological limitation.

Overall, the combined application of culture-based screening, disk diffusion testing, and PCR-based detection demonstrates that molecular methods remain the most reliable approach for identifying methicillin resistance in Staphylococcus isolates obtained from companion animals.

In Indonesia, research on MRSA has largely focused on livestock-associated strains, food-chain contamination, and human hospital-acquired isolates [16, 19, 22], whereas molecular investigations involving companion animals remain scarce. Waruwu et al. [23] reported MRSA in cats from a veterinary clinic in Sidoarjo, East Java, where S. aureus was isolated from 82% of sampled cats and 8.5% were identified as MRSA. In our previous study, 4.5% (6/11) of isolates from companion animals with clinical signs of bacterial infection in Yogyakarta were classified as MRSA, while 27.2% (3/11) were identified as MRSP [24]. Overall, data regarding the occurrence of MRSA and MRSP in companion animals in Indonesia remain limited.

Swab samples were collected from companion animals exhibiting clinical signs consistent with S. aureus or S. pseudintermedius infection. Colonies of both species were capable of fermenting mannitol, producing a yellow color change on mannitol salt agar. Gram staining revealed clusters of Gram-positive cocci. Differentiation between S. aureus and S. pseudintermedius based solely on phenotypic characteristics is challenging due to their close resemblance. However, biochemical assays such as ortho-nitrophenyl-β-galactoside (ONPG) and VP tests provide useful discriminatory value [18, 25, 26]. Both species are catalase-positive; however, S. aureus produces acetoin and is VP-positive, whereas S. pseudintermedius is VP-negative. Conversely, S. pseudintermedius is ONPG-positive, while S. aureus is ONPG-negative.

Phenotypically identified isolates were subsequently subjected to molecular confirmation. In this study, S. aureus was confirmed by PCR amplification of the 23S rRNA and nuc genes. The 23S rRNA gene encodes a conserved structural component of the bacterial ribosome, making it a reliable target for species-level identification [17]. The nuc gene encodes thermonuclease, a virulence-associated enzyme that serves as a hallmark marker of S. aureus and enables differentiation from coagulase-negative staphylococci [22]. Amplification of these targets yielded the expected amplicon sizes of 1250 bp (23S rRNA) and 279 bp (nuc), confirming that 41 isolates were S. aureus (Figure 2).

Identification of S. pseudintermedius was performed using PCR–RFLP analysis of the pta gene with the MboI restriction enzyme. This method allows differentiation of S. pseudintermedius from other members of the Staphylococcus intermedius group (SIG). The pta gene of S. pseudintermedius contains an MboI restriction site that is absent in other SIG species; therefore, digestion produces two characteristic fragments of 213 bp and 107 bp, whereas other SIG members remain undigested [18, 20]. All 14 isolates displayed the expected two-band pattern, confirming their identity as S. pseudintermedius (Figure 3).

In total, 55 isolates were successfully characterized, comprising 41 S. aureus and 14 S. pseudintermedius isolates derived from dogs, cats, and rabbits (Figure 4). The inclusion of rabbits as clinical cases adds novelty, as MRSA and MRSP epidemiology in small mammals remains underexplored both regionally and globally. The relatively high rate of S. aureus isolation from rabbits may be linked to their susceptibility to skin lesions and husbandry-related stress, which can facilitate colonization. The predominance of nasal isolates, particularly in cats, supports previous findings that the nasal cavity serves as a major reservoir and transmission source in both veterinary and household environments. These observations are consistent with earlier reports indicating that S. aureus commonly colonizes the skin, nasal passages, and urogenital tract of various animal species [4]. Similarly, S. pseudintermedius is frequently isolated from companion animals, especially dogs, and is associated with pyoderma, otitis externa, keratitis, metritis, and urinary tract infections [6–8]. The anatomical distribution of isolates provides insight into colonization dynamics and aids in distinguishing colonization from active infection. The combined phenotypic and molecular approach used in this study ensured accurate species identification, which is essential for appropriate antimicrobial therapy and epidemiological interpretation.

This study represents one of the first comparative analyses of hemolytic profiles of S. aureus and S. pseudintermedius isolates from companion animals in Indonesia. Methicillin resistance was assessed using a tiered approach consisting of initial screening on oxacillin resistance screening agar base (ORSAB), phenotypic confirmation by disk diffusion testing with cefoxitin (FOX) for S. aureus and oxacillin (OX) for S. pseudintermedius, and genotypic detection of the mecA gene. ORSAB screening classified 95% of S. aureus isolates as MRSA and 5% as MSSA, while 79% of S. pseudintermedius isolates were classified as MRSP and 21% as MSSP. Disk diffusion testing identified 41% of isolates as MRSA and 43% as MRSP, whereas PCR detection of mecA confirmed 66%, 34%, 93%, and 7% of MRSA, MSSA, MRSP, and MSSP isolates, respectively.

The observed variation among diagnostic methods aligns with previous reports demonstrating that ORSAB screening frequently overestimates MRSA prevalence compared with molecular confirmation. Blanc et al. [27] reported that only 42% of ORSAB-positive isolates were confirmed as MRSA by molecular methods. This discrepancy arises because non-S. aureus species can ferment mannitol and exhibit oxacillin resistance. Consequently, ORSAB should be used strictly as a preliminary screening tool and must be followed by molecular confirmation. In this study, only 66% of isolates classified as MRSA by ORSAB were mecA-positive, highlighting the limitations of culture-based detection and reinforcing the importance of molecular diagnostics in Indonesian veterinary laboratories.

Notably, 23 S. aureus and S. pseudintermedius isolates were mecA-positive yet remained phenotypically susceptible to cefoxitin or oxacillin. This phenomenon suggests the presence of silent or low-expression mecA elements in pet-associated staphylococci, an underreported issue in Indonesia. Similar genotype–phenotype discrepancies have been described previously [28]. Silent resistance genes may remain unexpressed under routine testing conditions but pose a latent risk due to their potential for horizontal transfer. Kang et al. [29] described S. pseudintermedius isolates carrying mecA that were oxacillin-susceptible, referring to them as pre-MRSP strains. These isolates possess an intact mecR1–mecI regulatory system that represses mecA transcription. The mecI gene encodes a repressor protein, while mecR1 functions as a membrane-bound signal transducer; when functional, these regulators suppress mecA expression, resulting in phenotypic susceptibility despite gene carriage [30, 31].

Conversely, eight isolates exhibited phenotypic resistance to cefoxitin or oxacillin but were mecA-negative. Phenotypic assays depend on bacterial growth conditions and may be influenced by agar composition, antibiotic diffusion, and inoculum density. Variations in agar depth or antibiotic potency can alter inhibition zones, leading to false resistance profiles. Heteroresistance, in which bacterial subpopulations express variable resistance levels, may also contribute to inconsistent phenotypic results. Furthermore, methicillin resistance is not exclusively mediated by mecA; alternative genetic determinants such as mecC, fem, and aux genes may also confer β-lactam resistance [32].

Antimicrobial susceptibility testing against 11 antibiotics revealed the highest resistance rates to penicillin, detected in 78% of S. aureus and 93% of S. pseudintermedius isolates. Resistance was also observed to ampicillin, amoxicillin, tetracycline, clindamycin, cefoxitin, erythromycin, oxacillin, gentamicin, vancomycin, and ciprofloxacin. These results align with global trends indicating increasing β-lactam resistance among Staphylococcus species isolated from companion animals. As noted by Ventola [33], inappropriate antimicrobial use, including underdosing, overuse, and prolonged treatment duration, creates selective pressure that promotes resistance development.

Overall, this study demonstrates a high prevalence of AMR and MDR among S. aureus and S. pseudintermedius isolated from companion animals. Isolates were classified as MDR when resistance was observed to at least one agent in three or more antimicrobial classes [26]. The comparable MDR burden observed in both species suggests convergent resistance evolution in Indonesian companion animals, a finding not previously documented in regional studies. The emergence of MRSA and MRSP in pets is of substantial concern due to their zoonotic potential and capacity to act as reservoirs of resistance genes at the animal–human–environment interface. By providing molecular evidence of MRSA and MRSP circulation in companion animals, this study strengthens the One Health perspective on AMR transmission in Indonesia.

To the best of our knowledge, this study is the first to integrate multispecies clinical sampling, molecular species confirmation, and comprehensive AMR profiling of MRSA and MRSP in Indonesian veterinary clinics. Nevertheless, several limitations should be acknowledged, including the absence of SCCmec or multilocus sequence typing, lack of sequencing-based confirmation, and omission of mecC screening, which may restrict detailed characterization of resistance mechanisms. Additionally, risk factor analysis and larger sample sizes were not included. Future studies should incorporate broader geographic sampling, longitudinal surveillance, and expanded molecular approaches to better elucidate transmission dynamics and resistance evolution.