Comprehensive whole genome analysis of Staphylococcus aureus isolates from dairy cows with subclinical mastitis

Staphylococcus species are the primary cause of mastitis in dairy cows across the world. Staphylococcus aureus has recently become a pathogen that is zoonotic and multidrug resistant. This study aimed to sequence whole genomes of 38 S. aureus isolates from 55 subclinical mastitis dairy cows of 7 small-scale farmers in the Free State Province, South Africa and document and their antimicrobial and virulence genes. The 38 isolates were grouped by the in silico multi-locus sequencing types (MLST) into seven sequence types (STs), that is (ST 97, 352, 152, 243) and three new STs (ST8495, ST8500, and ST8501). Thirty-three S. aureus isolates were divided into 7 core single-nucleotide polymorphism (SNP) clusters. Among the 9 distinct spa-types that were detected, Spa-types t2883 accounted for the majority of isolates at 12 (31.57%), followed by t416 with 11 (28.94%) and t2844 with 5 (13.15%). The data also revealed the identification of four (4) plasmids, with Rep_N (rep20) accounting for the majority of isolates with 17 (44.73%), followed by Inc18 (repUS5) with 2 (5.26%). These isolates included 11 distinct antimicrobial resistance genes and 23 genes linked to bacterial virulence. Surprisingly, no methicillin resistance associated genes were detected in these isolates. Genome data of the current study will contribute to understanding epidemiology S. aureus genotypes and ultimately aid in developing treatment and control plans to stop the spread of mastitis in the Free State province and South Africa as a whole.


Introduction
Staphylococcus species are known to cause acute to chronic infections/diseases that are related to increased morbidity to infected hosts such as humans and animals (Pattabhiramaiah and Mallikarjunaiah, 2023).There are about 53 species with 28 sub-species within the Staphylococcus genus with S. aureus being the main cause of persistent clinical and subclinical intramammary infections (IMI; Vanderhaeghen et al., 2015).The pathogenesis of S. aureus starts with teat colonization, through the intramammary space by either progressive colonization or changes in intramammary pressure caused by the milking machines (Maity et al., 2020;Vargová et al., 2023).In the mammary alveolus, S. aureus adheres to and enters mammary epithelial cells, which serve as the site for multiplication, eventually resulting in a chronic IMI (Maity et al., 2020).The molecular mechanisms underlying S. aureus IMI still need to be fully deciphered.Generally, bacteria sense host signals and adapt gene expression to match environmental conditions to cause infection (Elhawy et al., 2021).Numerous virulence factors (VFs) involved in adhesion, invasion, and host defense evasion are known and wellstudied in S. aureus.These VFs are either found in bacterial genomes or are within the transmissible genetic elements in a bacterium (Naushad et al., 2019).The emergence of drug resistance is a serious challenge for mastitis control due to their extensive use in the dairy industry, for example, through the dry cow therapy, contributes to the emergence of antimicrobial determinants in S. aureus, including development of multiple drug resistance (Klibi et al., 2018;Liu et al., 2020;Omwenga et al., 2021;Gelalcha et al., 2022).Molecular epidemiology of staphylococcal species involved in IMI of dairy cattle have focused more on the use of multi-locus enzyme electrophoresis (MLEE), pulsed-field gel electrophoresis (PFGE), sequence-based typing schemes, such as multiple-locus sequence typing (MLST), multiple-locus VNTR (variable number of tandem repeats) analysis (MLVA), random amplification of polymorphic DNA (RAPD) analysis and staphylococcal protein A (spa) typing (Li, 2008).Although these methods are helpful in genetic analysis, their resolution is often not strong enough to reveal genetic differences between strains.However, whole-genome sequencing (WGS) of bacterial genomes has become the preferred method to understand microevolution, phylogenies, and inter and intraspecies differences (Sivakumar et al., 2023).Thus, in the current study we utilized WGS to characterize and understand the virulence and antimicrobial resistance mechanisms in S. aureus isolates of subclinical mastitis (SCM) dairy cows from small-scale farmers of the Free State Province.

Sample selection and bacteriological analysis
The S. aureus isolates used in the present study were all obtained from milk samples collected from seven (7) small scale farms in three local Municipalities (Maluti-A-Phofung, Mantsopa and Setsotso) in the Free State Province of South Africa between year 2021-2022.The sample collection was conducted according to the guidelines of National Mastitis Council (2004).A total of 166 composite milk samples from individual cows were randomly screened for intramammary infection by means of somatic cell count (SCC) assay using flow cytometry (Mérieux NutriSciences, South Africa).Thereafter, based on the SCC results, only 220 individual quarters from 55 of 166 cows were subjected to California mastitis test (CMT) according to manufacturer's instructions (DeLaval, South Africa) on farm and subsequently only 160 quarter milk samples were collected for another round of SCC and microbiological analysis.The CMT results were scored and interpreted as recommended by Karzis et al. (2017).Isolates were defined as S. aureus on the basis of being gram-positive cocci and catalase positive.Thereafter, matrix-assisted laser desorption ionization-time-of-flight mass spectrometry (MALDI-TOF MS) and gene sequencing were employed for further identification of the isolates as reported in our previous study (Ozbey et al., 2022;Khasapane et al., 2024).Furthermore, we performed phenotypic and genomic antimicrobial resistance based on diskdiffusion and PCR techniques (Khasapane et al., 2024).To evaluate the susceptibility of Staphylococcus isolates to widely used antimicrobial drugs, the single disk diffusion technique was utilized.Antibiotic discs (ThermoFischer, South Africa) comprising of gentamicin (10 μg), ampicillin (10 μg), tetracycline (30 μg), penicillin (10 μg), erythromycin (15 μg), ciprofloxacin (5 μg), and cefoxitin (15 μg) were utilized according to the Clinical Laboratory Standards Institute (CLSI, 2023) guidelines which are interpreted as intermediate (I), sensitive (S), and resistant (R).

Whole-genome sequencing
Whole genome sequencing of S. aureus isolates was conducted at the National Institute of Communicable Diseases (NICD) Sequencing Core Facility, South Africa.Briefly, multiplexed, paired-end libraries (2 × 150 bp) were prepared using the Illumina DNA Prep kit (Illumina, San Diego, United States), followed by sequencing on the Illumina NextSeq 2000 platform (Illumina, San Diego, United States) at 100× coverage.

Antimicrobial resistance prediction
Detection of antimicrobial resistance determinants was performed using a combination of three popular tools, namely, AMRFinderPlus (Feldgarden et al., 2021), ABRicate 3 and staramr, 4 by scanning the assembled contigs against ResFinder (Bortolaia et al., 2020), PointFinder 5 and AMRFinderPlus databases.The outputs from these tools were summarized using HAMRonization. 6

Phylogenetic analysis
Core genome SNPs (single nucleotide polymorphisms) were used to investigate the phylogeny and genetic relatedness of isolates.Briefly, whole genome alignments were performed using scapper7 and Staphylococcus aureus strain NCTC 8325 was used as a reference.Recombinant regions were removed using Gubbins v3.2.1 (Croucher et al., 2015) and variable sites were obtained using snp-sites v2.5.1 (Page et al., 2016).Pairwise SNP distances were calculated using snp-dist v0.8.28 and a normalized pairwise SNP distance matrix was used as input for the cluster analysis using the R software environment v4.2.1.Assignment of SNP clusters was achieved by a combination of K-means clustering implemented in the eclust function (factoextra package; v1.0.7;Kassambara, 2016) and custom functions written in R using a silhouette score and SNP cut-off of 0.5 and 20, respectively.Visualization of cluster heat maps was performed using the ComplexHeatmap package v.14.0 (Gu, 2022).IQ-TREE v2.0.3 (Minh et al., 2020) was used to generate a maximum-likelihood phylogenetic tree using the GTR + F + ASC + R4 with 1,000 bootstrap approximations using UFBoot2 (Hoang et al., 2018).The phylogenetic tree was visualized and annotated using Microreact.9

Core SNP cluster analysis and transmission network reconstruction
Pairwise SNP distances were calculated using snp-dist (v0.8.2; https://github.com/tseemann/snp-dists)using the final alignment file based on variable sites only.A normalized pairwise SNP distance matrix was used as input for the cluster analysis performed using the R software environment (version 4.2.1).Briefly, K-means clustering was done using default parameters in the eclust function from the factoextra package (v1.0.7;Kassambara, 2016), but with bootstrapping set to 500, and only core SNP clusters with a silhouette score ≥ 0.5 and SNP cut-off ≤ 25 were considered.Visualization of cluster heat maps was performed using the ComplexHeatmap package (v.14.0;Gu, 2022).A simple dendrogram showing the SNP clusters was generated using the fviz_dend function from the factoextra package.The minimum spanning tree was generated using the ape package in R (Paradis et al., 2004) based on pairwise SNP distances.Visualization of the minimum spanning tree was done using the visNetwork package (v2.2.2; https://datastorm-open.github.io/visNetwork/).

Staphylococcus aureus isolates
In the current study, whole genomes of 38 S. aureus isolates associated with bovine subclinical mastitis from Free State Province smallholder farms were sequenced.Table 1 provides an overview of genomic sequences.

Discussion
Numerous investigations have examined potential variations in virulence gene profiles between S. aureus isolates from clinical and subclinical mastitis; however, no discernible differences have been found (Rocha et al., 2019;Naushad et al., 2020;Åvall-Jääskeläinen et al., 2021).According to Wellnitz and Bruckmaier (2012), mastitis is a dynamic phenomenon in which the microbe enters the mammary gland, causing leukocytes in the milk and mammary-gland epithelial cells to react and start the immunologic defense process.This causes a large number of neutrophils to migrate to the mammary gland and attempt to kill the microbial cells.In this study, we examined the presence of S. aureus associated with subclinical mastitis in dairy cattle in addition to associated antimicrobial resistance and virulence factors in 38 isolates.
Our identified nine spa types from S. aureus isolates and majority of them were assigned to t2883 (31.57%), followed by t416 (28.94%).Prior research has revealed a high diversity of S. aureus spa types in samples collected along dairy chains, including t2883 (Ben Slama   al., 2011;Boero et al., 2022).Interestingly, a study by Ben Slama et al. (2011) found the t2883 spa type in human hands that had come into touch with animals in Tunisia, suggesting that there may be a chance of transmission from human to animal or vice versa.In the current analysis, the t416 spa type was also discovered, and it appears to be dominant in all 38 isolates of S. aureus.The high prevalence of t416 spa type discovery in our investigation is more than that reported by Mora-Hernández et al. ( 2021), who found t416 in 6.06% of the S. aureus isolates from dairy cows in Mexico that had mastitis.The widespread use of antibiotics to treat S. aureus-caused human infections and cow mastitis is endangering public health since AMR is emerging among dangerous bacteria (Zigo et al., 2019).Finding the ARGs is necessary to evaluate the pathogenic potential of S. aureus during mastitis.Scanning genome sequences facilitates the identification of genetic components linked to virulence and antibiotic resistance (Paramasivam et al., 2023).
The phylogenomic study aligned with the known mechanism of zoonotic transmission of S. aureus (Silva et al., 2023).This may  be explained by data indicating that, although many S. aureus lineages are non-specific, others are suited to colonize and infect particular host species (Schmidt et al., 2017).It has been observed that majority of isolates across all clusters and STs were carrying the multidrug resistance genes lmrS, mepA, and tet, which encode for resistance against macrolide/phenicol, multidrug efflux MATE transporter MepA and tetracycline.Pu et al. (2021) carried out an experiment to identify genes (cadD) that confer resistance to cadmium antibiotics from bacterial plasmids.Subsequent investigations discovered these genes in five strains of MRSA isolates from Spain.In addition, the cadD gene for cadmium resistance was found in ST97, ST152, and ST8500, which together account for 44.74% of the isolates.While this study is among the few that has detected this gene in raw milk, another study conducted in the UK found it in animal feed at a detection rate of only 23% (Smith et al., 2005).Sequence Type 243 of the isolates of S. aureus, carried the gene MurA, which codes for resistance against fosfomycin.At different levels, fosfomycin resistance genes and mutations were found, and they were unmistakably linked to specific clonal complexes.According to Michalopoulos et al. (2011), the antibiotic fosfomycin targets the UDP-N acetylglucosamine enolpyruvyl transferase, which is involved in cell wall construction and encoded by the murA gene.Interestingly, the only strain with all the antibiotic resistance genes in this investigation was ST152, with the exception of the cadmium resistance genes.Quinolone, beta-lactam, ampicillin (blaI, blaR1 and blaZ), and trimethoprim (dfrG and dfrG 1; parE) were also present in this ST152.
According to reports, beta-lactam antibiotics frequently cause staphylococci to exhibit resistance (Olsen et al., 2006;Zigo et al., 2022).The blaZ gene produces a penicillinase, also called a beta-lactamase, that imparts penicillin resistance by hydrolyzing the beta-lactam ring and rendering the drug inert (van den Borne et al., 2010;Zigo et al., 2022).Additionally, the results of this study support the conclusions that there is little to no Staphylococcus that is resistant to beta-lactam or penicillin (van den Borne et al., 2010).However, almost 45% of S. aureus isolates were found to be blaZ positive and phenotypically penicillin-resistant in a follow-up study conducted in New Zealand by Steele and McDougall (2014).This genotype/ phenotype combination was associated with a very low cure following antibiotic treatment.The mecA and mecC genes were not present in any of the S. aureus isolates in the current investigation, which were all methicillin-sensitive S. aureus (MSSA).Furthermore, compared to a study by Haulisah et al. (2021) which revealed 80%-100% resistance to trimethropin, this study did not detect any of these genes from the isolates.Moreover, the production of toxins like hemolysins, leukotoxins, and enterotoxins as well as enzymes like serine proteases, cysteine proteases, and lipases that function as effectors during pathogenicity is what gives S. aureus its pathogenic potential (Sivakumar et al., 2023).
In addition to several enzyme-coding genes, the majority of the 38 S. aureus isolates included in this investigation carried multiple hemolysin genes.The genes encoding enterotoxins, enterotoxin-like proteins, and exfoliative toxins were found in just a small number of carefully selected genomes.Leukocidin D/E was found in all ST152 and ST342, while the Panton-Valentine leucocidin genes were only found in two sequence types (ST152 and ST243).Similar to this, all STs had higher frequencies of genes encoding adhesins and hemolysins, while isolates of S. aureus strains associated with bovine mastitis had lower frequencies of enterotoxins.Another crucial factor in determining virulence in staphylococci, particularly in S. aureus, is toxin synthesis.Inflammation and leukocyte cell death are promoted by these toxins, which include cytotoxins (hemolysins, leukotoxins, and leukocidins) and superantigens (enterotoxins, exfoliative toxins, and toxic shock syndrome toxins; TSST; Haag et al., 2019;Chen et al., 2022).In isolates from pus, skin infections, and abscesses, the lukD and lukE genes exhibited strong self-association.As their products are secreted prior to combining to create the PVL toxin, the genes lukF-PV and lukS-PV were associated in the current groups, which is consistent with the literature (Kaneko and Kamio, 2004;Rodrigues et al., 2022).All S. aureus isolates in all clusters carried the cytotoxins hlgA, hlgB, and hlgC, which encode alpha, beta, and hemolysin, respectively.All S. aureus isolates used in the current investigation included leukocidin genes, such as ST97, ST352, lukS-PV, and lukF-PV (ST152 and ST243).This is consistent with research on bovine S. aureus isolates, which discovered leukocidin and leukotoxin genes in the majority of isolates (Wächter et al., 2021) from cows in India.This may explain similar observations from the current study because many of the lukS-PV and lukF-PV genes in our analysis shared the same percentage identity across different S. aureus isolates, suggesting that the gene sequences may be comparable to those of the genome even if they are absent.Many enterotoxin genes were also found in S. aureus isolates in this current study.The aur gene produces the protein aureolysin, which alters the adhesion factor CflB and triggers additional proteases to increase S. aureus pathogenicity (McAleese et al., 2001).The bi-component leukotoxins that are produced by the hglA, hglB, and hglC genes can create holes in cell membranes, which allows them to lyse cells (Staali and Colin, 2021).Since these gene products are linked to clinical mastitis, they could be valuable targets for the creation of vaccinations and therapeutic drugs (Ahmad-Mansour et al., 2021;Moawad et al., 2023).
The enterotoxins sec, sei, sen, sem, seo, and seu were also shown to be associated with similar frequencies in all analyzed groups; these findings have also been reported by other investigations (Indrawattana et al., 2013;Schwan, 2019;Ren et al., 2020).Additionally, prior research has demonstrated that see and sec, sel, are commonly found in MRSA strains (Hu et al., 2011); however, the current investigation did not find this association.It is crucial to prevent contamination of enterotoxin-producing S. aureus isolates throughout the food production chain since they can cause acute and severe food poisoning.Unpasteurized milk-based cheese and raw milk are wellknown dietary sources of S. aureus food poisoning.It is believed that the chemotaxis inhibitory protein (chp product) and the scn gene product are highly specific for staphylococcal isolates of human origin (Pinchuk et al., 2010).The most typical cause of bovine mastitis is staphylococcal enterotoxin C. The primary enterotoxin gene of S. aureus isolated from cows with mastitis, according to other researchers, is called sea.Three enterotoxin-like genes have shown super-antigenic activity but no emetic qualities, two enterotoxin genes (seg and sei), and two enterotoxin genes make up the cluster (selo, selm, and seln).In the current investigation, 91.7% of the S. aureus isolates tested positive for the SEs coding genes classical expression (sea, seb, sec, or sed).The exfoliative toxins (eta and etb), toxic shock syndrome toxin-1 (tsst-1), staphylococcal enterotoxins (sea, seb, sec, sed, saw, seg, seh, sei, and sej), and pvl are all exoproteins that S. aureus is capable of producing (Ren et al., 2020).On the other hand, it is still unclear how staphylococcal enterotoxins affect mammary epithelial cells.Nineteen (19) serologically unique SEs have recently been found.Moreover, the genes most commonly found in S. aureus isolates from dairy cows with mastitis are sec, sed, seg, and sei and sea (Ren et al., 2020).

Conclusion
Understanding the epidemiology of S. aureus genotypes in dairy animals and herds may aid in developing treatment and control plans to stop the illness from spreading.Sequence Type 97, virulence genes like leucocidin, hemolysin, and aureolysin, and AMR genes like lmrS, mepA, and tet (38) were most frequently identified genes in this study.The capacity of S. aureus to colonize and penetrate the host may be influenced by the combination of these genes.Consequently, due to the global distribution of these genes, screening them in S. aureus isolates may be valuable for aiding in clinical outcome prediction and, specifically, for identifying hazardous strains.The current study also showed that all isolates had a nearly identical genotypic pattern and that certain isolates carried virulence factors such as PVL-encoding genes, suggesting that S. aureus isolates in animals should be closely monitored to prevent the spread of these genes.It is noteworthy that all isolates tested negative for mecA and mecC, Furthermore, the current study has also showed or revealed an association between ST97, spa type t2883 and t416 and the plasmid Rep_N.This study has shown that there is a wide range of S. aureus genotypes occurring in dairy cattle in the Free State province and that genetic variations are related to geographic origin of the isolates.This suggests that taking the region of interest and the strain virulence into consideration may help to formulate strategies directed to stop the spread of infection and to set up control measures in accordance with pathogen and host features.Hence, depending on the description of the circulating strain, the farmer would be able to choose whether to slaughter the sick animals or isolate positive cows using sanitary milking practices and an appropriate milking schedule. et
FIGURE 2Depicts core genome SNP clusters.

FIGURE 3
FIGURE 3Distribution of antimicrobial resistance and virulence genes among Staphylococcus aureus isolates.

TABLE 1
Shows the overall sequence types based on MLST, spa-types, and plasmid diversities.