Introduction

Methicillin-resistant Staphylococcus aureus (MRSA) poses a significant public health concern due to its resistance to multiple antibiotics and its ability to cause infections in both humans and animals. The mecA gene, which encodes an altered penicillin-binding protein (PBP2a), is the primary determinant of methicillin resistance. This altered PBP2a reduces the effectiveness of β-lactam antibiotics1,2. In addition to antimicrobial resistance, MRSA possesses multiple virulence factors, including adhesins, immune evasion mechanisms, toxins, and biofilm-forming capabilities. These factors collectively contribute to its pathogenicity and host adaptation3,4,5,6. The widespread prevalence of multidrug-resistant (MDR) MRSA complicates treatment in both human and veterinary medicine, underscoring the urgent need for a comprehensive understanding of its epidemiology1.

Initially confined to healthcare settings, MRSA has now evolved into a prevalent community-acquired pathogen. Although MRSA colonization is more common among healthcare workers and hospitalized individuals, recent reports suggest its increasing presence in community settings, including households with pets7,8,9,10. Companion animals, particularly dogs and cats, may serve as reservoirs for MRSA and facilitate the transmission of the infection within households, posing a One Health concern that interconnects human, animal, and environmental health11. MRSA colonization in pets is less common than in humans, but its prevalence is rising, influenced by study populations and geography7,10.

In Bangladesh, pet ownership-especially of cats and dogs-has traditionally been uncommon but is growing rapidly in urban areas12. While several studies have reported high MRSA prevalence in human clinical samples in Bangladesh, ranging from 43.7 to 53.1%13,14,15,16,17,18, research on MRSA in companion animals remains limited12,17,19,20. The concurrent detection of MRSA in both pets and humans raise concerns about bidirectional transmission within households and potential occupational exposure among veterinary staff10,11,21,22. Despite these concerns, no systematic investigation has been conducted to assess MRSA transmission dynamics between pets, their owners, and veterinary professionals in Bangladesh. This study aims to fill this gap by investigating MRSA prevalence and potential transmission pathways among pets, their owners, and veterinary personnel. Additionally, it examines antimicrobial resistance profiles and virulence determinants of MRSA isolates, offering valuable insights into zoonotic transmission within the One Health framework.

Materials and methods

Ethical statement

Verbal and written informed consents were obtained from pet owners and veterinary personnel during sample collection. The methodologies and related protocols used in this study were approved by the Institutional Ethical Committee (Animal Welfare and Experimental Ethics Committee guidelines at Bangladesh Agricultural University) [Approval No.: AWEEC/BAU/2021(55)]. All methods were performed in accordance with the relevant guidelines and regulations.

Sample size calculation and sample collection

The sample size for estimating MRSA prevalence among humans and animals was calculated using the following formula

$$n=\frac{{Z}^{2}\cdot p\cdot q}{{d}^{2}}$$

where Z = 1.96 for a 95% confidence level, p = 0.05 as the expected prevalence for both pets19 and contact persons23, q = 1 − p, and d = 0.03 as precision. These assumptions resulted in a sample size of 202 for each group.

We used a purposive sampling strategy to collect data from pet owners, their pets (dogs and/ or cats), or veterinary professionals. However, a sample frame for such pair was not available, which prevented us from adhering to a random sampling protocol. Some pet owners declined to provide samples from their animals. As a result, the number of pet samples was slightly lower than initially expected. A total of 213 humans (185 pet owners and 28 veterinary personnel including the veterinary surgeon and attendants), and 191 animals (154 cats and 37 dogs) samples were collected from veterinary teaching hospital (VTH), BAU, Mymensingh, and Central Veterinary Hospital (CVH), Dhaka, during the period between July 2022 and June 2024. Pets visiting the hospitals for illness or regular vaccination, or surgical intervention were included in this study. Pets with no visible illness or complaints were classified as apparently healthy. Considering the primary colonization site and transmission risk nasal swabs were collected from both apparently healthy humans and pets. In clinical cases, samples such as pus, wound swabs and exudates were collected from lesions. All samples were collected using sterilized cotton swabs, immediately inoculated in nutrient broth (Himedia, India), and transported to the laboratory in an ice box at 4 °C. For comparative purposes, the following letter abbreviations were used for origin of samples/isolates: P; people, C; cats, D; dogs, along with respective sample number.

Isolation and identification of S. aureus, and MRSA

Samples collected in nutrient broth were incubated overnight at 37 °C for enrichment. S. aureus was then isolated on Mannitol Salt Agar (Himedia) and Blood Agar (Himedia) to assess colony morphology and hemolytic properties. S. aureus produces golden yellow color colonies on MSA, golden yellow to off-white (with some exceptions of white appearance) and β-hemolysis on blood agar. Colonies with the above characteristics were subjected to Gram staining and biochemical tests, including catalase and coagulase tests for identification24,25. A total of 404 samples were tested using the slide coagulase test, and 135 were positive (33.42%). S. aureus isolates were further confirmed by PCR targeting nuc (thermostable nuclease) (Supplementary Fig. 1, Supplementary Table 1)17. For the identification of MRSA cefoxitin resistance test as well as methicillin resistance gene mecA screening by PCR was performed (Supplementary Fig. 2, Supplementary Table 1)17,26. Cefoxitin resistance test was performed with cefoxitin (30 µg) using disk diffusion method on Mueller Hinto agar. Isolates showing a zone diameter of ≤ 21 mm were classified as cefoxitin resistant and designated as MRSA26.

Antimicrobial susceptibility testing

Antimicrobial susceptibility testing was performed using the disc diffusion method recommended by the Clinical & Laboratory Standards Institute (CLSI)26. The antibiogram of the isolated MRSA was determined using fifteen antimicrobial agents in accordance with CLSI’s recommended Gram-positive panel. The antimicrobial panel included amikacin (30 μg), amoxycillin + clavulanic acid (30 μg), azithromycin (15 μg), cefoxitin (30 μg), ceftriaxone (30 μg), cephradine (30 μg), ciprofloxacin (5 μg), clindamycin (10 μg), doxycycline (10 μg), gentamicin (10 μg), linezolid (10 μg), penicillin (10 unit), sulfamethoxazole + trimethoprim (25 μg), tetracycline (30 μg), and vancomycin (30 μg). S. aureus ATCC 25,923 was used as a control and isolates showing resistance to three or more antimicrobial classes were classified as multi-drug resistant27.

Whole genome sequencing (NGS) and analysis

A total of 12 strains (6 pairs) of MRSA were subjected to whole genome sequencing (WGS) to investigate potential transmission between pets and their owners. The isolates were selected from cases where both the pet and owner tested positive for MRSA, ensuring relevance to the study’s objectives. WGS was outsourced to the NGS facility at the Child Health Research Foundation (CHRF), and the International Centre for Diarrhoeal Disease Research, Bangladesh (icddr,b). Adapter trimming and assembly were performed using the Galaxy Europe platform (https://usegalaxy.eu/), using Trimmomatic (Galaxy version 0.39 + galaxy2) for trimming and Unicycler (Galaxy version 0.5.0 + galaxy1) for assembly. For long read data generated via Nanopore sequencing, assembly was performed using Flye (Version 2.9.3)28. The genomes have been deposited into GenBank under the following Accession Numbers: JBGVSC000000000–JBGVSK000000000, JBJHFC000000000, JBJHFD000000000, CP169233-CP169235.

Pangenome analysis was carried out using the Roary pangenome pipeline29. The resulting pangenome matrix and gene breakdown figures were generated using the roary_plots.py script. Functional annotation was conducted through Prokka30, and the Rapid Annotation using Subsystem Technology (RAST) server31. To ascertain clonality or possible horizontal transmission, whole genome average nucleotide identity (ANI), SNP analysis, multi-locus sequence typing (MLST), and Staphylococcal protein A (spa) typing were performed. Whole genome ANI was calculated on CLC Genomics Workbench 24.0.2 (Qiagen, Aarhus, Denmark). Pairwise SNP differences were determined from the core-genome alignment using the snp-dists version 0.7.0 (https://github.com/tseemann/snp-dists). For spa typing, spaTyper32,33 was used. Source tracking, identification of antimicrobial resistance, and virulence genes were accomplished via PubMLST (https://pubmlst.org/bigsdb?db=pubmlst_saureus_seqdef), BacWGSTdb2.0 server (http://bacdb.cn/BacWGSTdb/), the Comprehensive Antibiotic Resistance Database (CARD) (https://card.mcmaster.ca/), and Virulence Finder 2.0 http://cge.food.dtu.dk/services/VirulenceFinder/). A phylogenetic tree based on cgMLST was constructed using Phyloviz (https://online.phyloviz.net/index). The pathogenic potential of the isolates was predicted by PathogenFinder 1.134.

Analysis of SCCmec (Staphylococcal cassette chromosome) elements

The Staphylococcal cassette chromosome mec (SCCmec) was identified from the WGS data using SCCmecFinder 1.2, provided by the Center for Genomic Epidemiology35,36. Phylogenetic analysis of the SCCmec elements was conducted on CLC Genomics Workbench 24.0.2 (Qiagen Aarhus), and linear comparison were generated with Easyfig 2.2.5 (http://mjsull.github.io/Easyfig/), respectively.

Statistical analysis

Data on pets, pet owners, and veterinary professionals were entered into Microsoft Excel 2010 and then transferred to R (version 4.3.1) for analysis. The prevalence of MRSA among these groups was estimated using the ‘tabpct’ function from the ‘epiDisplay’ package in R. The 95% confidence interval for prevalence was calculated using the ‘prop.test’ function. We used univariable logistic regression to compare the prevalence of MRSA among pets, pet owners, and veterinary professionals, and to assess whether the differences observed between these groups were statistically significant.

Results

Prevalence of S. aureus and MRSA

The overall prevalence of S. aureus and MRSA among humans (pet owners and veterinary personnel) was recorded at 31.9% and 15%, respectively. In cats, S. aureus was found in 18.2% of cases, while MRSA was present in 5.8%. In dogs, S. aureus was detected in 32.4% of cases, and MRSA was found in 13.5% (Table 1). Dogs showed a higher MRSA prevalence (13.5%, 95% CI 5.1–29.6) compared to cats (5.8%, 95% CI 2.9–11.1); however, this difference was not statistically significant (p = 0.20). The overall MRSA prevalence among pets was 7.3% (95% CI 4.2–12.2). Notably, pet owners demonstrated a significantly higher MRSA prevalence (14%, 95% CI 9.5–20.1) compared to pets. This difference resulted in pet owners being 2.1 times more likely to test positive for MRSA than their pets (OR 2.1; 95% CI 1.1–4.2; p = 0.04). The prevalence of MRSA among veterinary personnel was 21.4% (95% CI 9.0–41.4), which was significantly higher than that of pets (OR 3.4; 95% CI 1.1–9.6; p = 0.02).

Table 1 Prevalence of S. aureus and methicillin-resistant S. aureus (MRSA).
Full size table

Antibiotic susceptibility profiles of MRSA

All the S. aureus (n = 108) were subjected to cefoxitin screening where forty-six (46) showed cefoxitin resistance indicating MRSA. During antibiotic susceptibility testing, unfortunately we couldn’t recover seven (7) of the MRSA isolates. Thus, 39 of the 46 MRSA isolates were subjected to susceptibility testing against 15 antimicrobials from different classes. As expected, all the MRSA isolates were resistant to cefoxitin (100%), and most were resistant to penicillin (94.9%), azithromycin (82%), and ciprofloxacin (53.9%). In contrast, all isolates were sensitive to amikacin, clindamycin, linezolid, and vancomycin. Additionally, a high level of sensitivity was noted for trimethoprim-sulfamethoxazole (92%) and gentamicin (87%). (Supplementary Fig. 3). The study also showed that 94.9% (37/39) of the MRSA isolates were resistant to 3 or more classes of antimicrobial agents indicating multidrug-resistant (MDR).

General characteristics of the MRSA genomes

Sequencing and assembly of 12 MRSA genomes revealed sizes ranging from 2,758,146 to 2,843,370 bp, with GC content between 32.6 and 32.9%. Pangenome analysis of these genomes identified 4368 gene clusters, including 1989 core genes, 1418 shell genes, and 961 cloud genes (Supplementary Fig. 4A). This suggests moderate or lower clonality among the isolates. On the other hand, gene breaks within pet-owner pairs indicated high level of homology especially in C134-P134 (99.6% core genes) and C185-P185 (99.7% core genes) pairs (Supplementary Fig. 5) indicating clonality within the pairs. The gene clustering matrix also indicated a notable genetic variation among isolates from different sources (Supplementary Fig. 4B) with similar genetic profiles of C134-P134, and C185-P185 pairs (Supplementary Fig. 4B).

Multi-locus sequence (MLS) and Staphylococcal protein A (spa) typing

Multi-locus sequencing typing (MLST) classified the genomes into four (4) distinct sequence types (ST), with 66.67% (8/12) belonging to sequence type 6 (ST6) (Fig. 1). Notably, isolates from the same pet-owner pairs shared the same ST type, with the exceptions of C52-P52 and D88-P88. Phylogenetic analysis using concatenated MLST gene sequences grouped the isolates into four distinct clusters, alongside S. aureus from different regions worldwide (Fig. 2). Similarly, spa typing categorized the isolates into four types, with 58.33% identified as spa type t304 (Fig. 1). Consistent with the MLST findings, isolates from the same pet-owner cohort, except for C52-P52 and D88-P88, exhibited identical spa types. Additionally, this study identified two previously unknown spa types (Fig. 1).

Fig. 1
figure 1

Phylogenetic relatedness, phenotypic and genotypic characteristics of MRSA isolated from pet and their owners in Bangladesh. The tree was constructed from the core-genome alignment of the twelve (12) MRSA sequences from this study. Core genome alignment was performed using Rorary Pangenome pipeline. The maximum likelihood tree was constructed from the core genome alignment file using RAxML Next Generation V1.2.2. The tree is showing the isolated ID, sampling information, genotypic and phenotypic characteristics. ST, spa, and SCCmec types were determined using PubMLST server, spaTyper, and SccmecFinder 1.2, respectively. Antimicrobial resistance (AMR) phenotypes: CIP, ciprofloxacin; CE, cephradine; Fox, cefoxitin; P, penicillin; TE, tetracycline. Antimicrobial resistance genes were identified using Comprehensive Antibiotic Resistance Database (CARD). Abbreviations: AH, apparently healthy; NS, nasal swab; PS, pus swab; UT, untypable; WS, wound swab.

Full size image
Fig. 2
figure 2

MLST-phylogeny of S. aureus sequences. The minimum spanning tree was constructed using the concatenated sequences of 7 MLST genes on Phyloviz online platform (https://online.phyloviz.net/). Concatenated sequences of related S. aureus were downloaded from PubMLST website (https://pubmlst.org/organisms/staphylococcus-aureus). Isolates from different countries were indicated by different color codes with isolates from this study as red. ID’s with ST inside each circle indicates sequence type (ST) of the entire circle.

Full size image

Antimicrobial resistance genes (ARGs)

All MRSA isolates (n = 46) carried the mecA gene, confirming their methicillin resistance. On the other hand, tetracycline resistance genes tetA and tet (38) were universally present in all the 12 sequenced genomes. However, only one isolates (D88) showed phenotypic resistance to tetracycline (Fig. 1). Notably, the tetracycline resistance isolate D88 carried an additional tetK gene (Fig. 1), suggesting that tetK rather than tetA or tet(38), was associated with tetracycline resistance. However, we don’t have experimental evidence to establish the association between tetK and tetracycline resistance. The beta-lactam resistance gene blaZ was found in 75% (9/12) of the genomes (Fig. 1). Among the isolates, D88 of CC88 showed higher AMR resistance, however, no particular correlation was observed between resistance profiles and genetic lineages.

Virulence factors (VFs)

The WGS (n = 12) were analyzed for the presence of virulence factors associated with blood cell lysis, enterotoxicity, exoenzyme production, and immune modulation. Several virulence factors were detected in the genomes (Fig. 3). All 12 genomes (100%) harbored genes associated with blood cell lysis, including hlgA, hlgB, hlgC, lukD, and lukE. Additionally, two strains (D88 and P88) contained the genes for Panton–Valentine Leukocidin (lukF-PV and lukS-PV), which are important virulence factors in community-associated MRSA (CA-MRSA) (Fig. 3). Furthermore, all the genomes, except for C52, contained the serine protease-encoding genes splA and splB, which are believed to contribute to various aspects of S. aureus pathogenesis. The presence of aureolysin, an important virulence factor involved in immune evasion, tissue invasion, biofilm formation, and antimicrobial resistance, was also noted in all genomes. In addition, multiple enterotoxin genes were identified, with sea being the most prevalent (66.67%), followed by sek (25%) and seq (25%). These findings suggest that these isolates may pose a risk as potential foodborne pathogens. Interestingly, one genome (C52) contained seven enterotoxin encoding genes along with the tst gene, which encodes toxic shock syndrome toxin-1, a factor associated with severe complications such as toxic shock and multiple organ failure.

Fig. 3
figure 3

Distribution of virulence genes in S. aureus. Virulence genes were identified by VirulenceFinder 2.0 and the heat map was generated online on DISPLAYR (https://southeastasia.displayr.com/). Dark shed indicates presence and light shed indicates absence of the genes. Hostimm*, host immune modulators.

Full size image

Characteristic of the SCCmec elements

Analysis using SCCmecFinder identified SCCmec type IV in all isolates, which was further confirmed by alignment and phylogenetic analysis of the SCCmec elements (Fig. 4). The genetic organization of the mec and ccr gene complexes (ccrA-ccrB-IS1272-mecR1-mecA-IS431) (Fig. 5) was identical to that of SCCmec type IV (2B) described earlier35. Interestingly, isolates from the same pet-owner cohort carried similar genetic arrangements at both the up-stream and downstream ends of the SCCmec elements.

Fig. 4
figure 4

Phylogenetic relationship of SCCmec clusters. The SCCmec references sequences (denoted by accession numbers) were retrieved from the NCBI nucleotide database followed by alignment with that of the study genomes (indicated by strain IDs) and phylogeny on CLC Genomics Workbench 23.0.5. Phylogenetic tree was constructed using UPGMA method and the nucleotide distances were calculated using Jukes–Cantor model with bootstrapping on 1000 replicates.

Full size image
Fig. 5
figure 5

Linear comparison of SCCmec clusters with adjacent upstream and downstream sequences. The image was created using Easyfig 2.2.5 (http://mjsull.github.io/Easyfig/). A type IV SCCmec cluster (Accession No. AB063172) was used as the reference sequence. SCCmec-related genes are indicated by different colored arrows, while other genes are represented by light ash-colored arrows. The color intensity band between the sequences indicates nucleotide homology per site.

Full size image

Genetic relatedness of the S. aureus

Whole genome based average nucleotide identity (WGS-ANI), pairwise SNP analysis was performed to determine genetic homology of the isolates. The isolates showed considerable diversity. Interestingly, the pet-owner pairs C134-P134, C151-P151, C185-P185 and C186-P186 showed ≥ 99.9% identity within the pairs based on WGS-ANI (Supplementary Fig. 6). On the other hand, SNP analysis revealed 18, 84, 0, and 154 nucleotide differences within the pairs C134-P134, C151-P151, C185-P185, and C186-P186, respectively (Fig. 6). The MLST, spa and SCCmec patterns within the pairs C134-P134; C151-P151; C185-P185; C186-P186 were also identical. Considering the SNP cutoff threshold37,38 these results suggest that the cohorts C134-P134, and C185-P185 were clonal, indicating the potential transmission of S. aureus between the owners and their pets.

Fig. 6
figure 6

Heatmap of SNP Distances among S. aureus genomes in this study. Core genome alignment was performed using Roary pangenome pipeline. Pairwise SNP distances from the core genome alignment was calculated using snp-dists v0.7.0 and visualized with plot_matrix_heatmap.

Full size image

Discussion

We observed a high prevalence of multidrug-resistant MRSA strains in both pets and their owners, demonstrating the zoonotic potential of MRSA within the One Health framework. The close genetic relatedness observed between paired human and pet isolates (C134-P134 and C185-P185) supports the likelihood of interspecies transmission, reinforcing the role of human-animal interactions in the transmission dynamics of these pathogens10,39,40,41,42,43,44,45. While humans are recognized as both transient and permanent carriers of S. aureus and MRSA, pets are generally considered transient hosts of coagulase-positive staphylococci, including MRSA46,47,48. However, the detection of clonal MRSA strains in both pets and their owners suggest that close contact facilitates transmission, likely from humans to their companion animals. The genetic diversity observed in the isolates (C52, P52, P88, D88, C151, P151, C186, P186) indicates independent adaptation or horizontal gene transfer events, reflecting the complexity of MRSA’s evolutionary dynamics. Whether pets serve as reservoirs for prolonged MRSA carriage or act primarily as vectors for transient colonization remains an open question that warrants further longitudinal studies.

Methicillin-resistant Staphylococcus aureus (MRSA) is recognized as one of the most adaptable and successful pathogens, capable of infecting a wide range of hosts and causing various types of infections2,49. The World Health Organization (WHO) has classified MRSA as a high-priority pathogen for research and the development of new antimicrobials because of its significant threat to public health, rising antimicrobial resistance, and potential to cause life-threatening infections50. Our findings emphasize the zoonotic risk posed by pets, a concern that is amplified by their close and prolonged contact with humans22,51,52. The occupational risks for pet owners and veterinary personnel also warrant attention, as evidenced by the notable prevalence of S. aureus and MRSA (31.9 and 15%, respectively) among these groups40,43.

The key mechanism of methicillin resistance is the mecA gene, which is located on a horizontally transferable element known as Staphylococcal Cassette Chromosome mec (SCCmec)35. Understanding the SCCmec type is essential for epidemiological tracking and infection control, as different SCCmec types are often linked to distinct ecological niches and modes of transmission. For example, SCCmec types I, II and III are predominantly associated with hospital acquired MRSA (HA-MRSA); whereas types IV and V are more common in community acquired MRSA (CA-MRSA)53. Moreover, types I, II and III often carry additional resistance genes such as aadD, aphA3, ermA, tetK, tetM, beyond mecA, resulting in multidrug resistance54. In contrast, type IV typically confers resistance solely to methicillin, making CA-MRSA strains less resistant overall, but often more virulent54. Thus, identifying the SCCmec type can guide treatment decisions by predicting which antibiotics are likely to be effective. In this study, all the 12 sequenced MRSA isolates carried SCCmec type IV, irrespective of their host or clonality. Predominance of SCCmec type IV has been also reported in clinical MRSA isolates from Bangladesh18. The association between SCCmec type IV and CA-MRSA suggests the adaptation of the study strains to community environments, a trend increasingly observed in CA-MRSA55,56. This raises concerns about the potential for widespread dissemination of these strains.

The successful establishment of MRSA in specific geographical locations depends on various factors, including frequency and persistence49. Over time, multiple clones have emerged, each with different virulence determinants that allow them to adapt to particular hosts. Although not all virulence factors have been fully understood, key contributors include adherence mechanisms, immune-evasion, toxin production, biofilm formation and acquired factors such as PVL, ACME, SapInn5449,57,58. Current successful MRSA lineages include CC5, CC8, CC22, CC88, CC30, CC59, CC80, CC93, CC9, and CC398, each with different host specificities and geographical distributions49. The successful lineages ST80 (CC80) have been reported in 38 countries across Europe, the Middle East, North Africa and Asia. ST88 (CC88) has been identified in various regions including Asia, Africa, America, Europe, Oceania indicating their widespread distribution59,60. In our study, we identified four different sequence types (STs), with ST6 being the most prevalent. Our findings align with the recent report of MRSA from clinical cases in Bangladesh18. The presence of ST80 and ST88 in both humans and animals in Bangladesh18, indicates that these globally dominant lineages are circulating within the study population, posing a potential public health threat. However, due to limited genomic information, it is unclear which lineage is dominant in this country. Notably, ST88, typically associated with HA-MRSA18, was found in the nasal swab of an apparently healthy dog. The occurrence of ST88 in dogs is rarely reported, suggesting either a reservoir role for the dog or acquisition through close human contact. Similarly, the identification of ST6-t304 in cats and its clonal relationship with human isolates is a significant finding. Although, MRSA has been reported in clinical cases involving human and cat, the lineage ST6-t304 has not been previously reported in pets other than humans18,61,62,63,64,65. Moreover, some isolates carried previously unreported or unknown spa types. These novel spa types can emerge through genetic recombination, deletion events, or horizontal gene transfer, potentially providing selective advantages like improved colonization, immune evasion, or antimicrobial resistance66,67,68,69. The functional significance of these unknown spa types in MRSA pathogenesis and adaptation is still unclear, highlighting the necessity for further genomic characterization to assess their role in MRSA persistence and spread.

The presence of virulence factors including hemolysins, leukocidins, enterotoxins, aureolysin, and serine proteases, reflecting their pathogenic versatility of these strains. The ubiquitous presence of hemolysin genes (hlgA, hlgB, hlgC, lukD, lukE) across all the 12 genomes indicates a high potential for these strains to cause severe invasive infections by targeting and lysing host cells70,71. Additionally, the detection of enterotoxin genes, especially in isolates carrying multiple enterotoxins such as sea, sek, and seq, raises concerns about the potential for these strains to cause illnesses, particularly in scenarios where cross-contamination between pets and humans could occur72,73,74. The presence of PVL genes in two isolates (D88 and P88) is particularly alarming due to their association with severe necrotic infections75,76,77. Moreover, isolate D88 is an MDR-MRSA and belongs to the currently dominant HA-MRSA lineage ST88. MRSA with PVL from lineage ST88 has been reported in clinical cases in Bangladesh18, but its occurrence in a pet (dog) is, to the best of our knowledge, being reported for the first time. Similarly, the tst-1 gene in isolate C52 raises concerns about MRSA pathogenesis and toxic shock syndrome, a severe and potentially fatal condition78,79.

We found a high prevalence of MDR MRSA strains (94.8%) indicating the increasing difficulty in managing MRSA infections. These strains showed resistance to critical antibiotics like cefoxitin, penicillin, and ciprofloxacin, which aligns with global trends and emphasizes the need for judicious antimicrobial use52,80,81. However, there is a positive note as these strains remain susceptible to amikacin, clindamycin, and linezolid, providing potential treatment options. Despite this, the prevalence of MDR strains complicates treatment strategies and highlights the importance of routine resistance monitoring2,80,82. A subset of isolates showed resistance to cefoxitin while appearing susceptible to penicillin. These isolates were confirmed to carry the mecA gene, which encodes the penicillin-binding protein PBP2a—known for its low affinity to β-lactam antibiotics, including penicillin, oxacillin, methicillin, and cefoxitin. This unusual phenotype may be attributed to several factors. These include heteroresistance or low-level mecA expression, which could lead to cefoxitin resistance while maintaining partial susceptibility to other β-lactams. Additionally, some strains may harbor a nonfunctional mecA gene or have reduced β-lactamase activity, resulting in apparent susceptibility to penicillin.

Our findings provide critical insights into the genetic and phenotypic characteristics of MRSA in this population. However, the findings are limited by a small sample size and geographical scope, which may affect their generalizability. Future research should include larger-scale, nationwide studies to better understand MRSA epidemiology in both pets and humans. Longitudinal studies are also needed to investigate MRSA transmission dynamics and the role of environmental reservoirs. Additionally, exploring the molecular mechanisms behind resistance and virulence will provide deeper understanding of MRSA’s adaptability and inform targeted interventions.

Conclusions

This study demonstrated the presence of MRSA in both pets and their human companions, with an overall prevalence of 7.3% in pets and significantly higher levels in pet owners (14%) and veterinary personnel (21.4%). Whole-genome sequencing revealed that pet-owner pairs harbored clonal MRSA strains, indicating probable transmission between pets and humans. The isolates displayed multidrug resistance and carried a range of virulence factors, including Panton–Valentine leukocidin (PVL), enterotoxins, and tst, underscoring their pathogenic potential. SCCmec typing identified all isolates as type IV, and WGS-ANI and SNP analyses confirmed high genetic similarity among clonal pairs. These findings provide genomic evidence supporting the zoonotic transmission of MRSA. Given these results, the study highlights the zoonotic potential and public health risks associated with MRSA transmission between pets and their owners. The findings underscore the importance of adopting a One Health approach to tackle MRSA transmission. Policymakers should focus on implementing routine MRSA screening in both veterinary clinics and human healthcare facilities, encourage responsible antibiotic use among pet owners and veterinary professionals, and enhance public awareness of hygiene practices. Collaborative efforts between human and veterinary healthcare sectors are crucial to control the spread of MRSA and mitigate its public health impact.