Resistance and virulence in Staphylococcus aureus by whole-genome sequencing: a comparative approach in blaZ-positive isolates

Abstract

Mastitis caused by Staphylococcus aureus is a worldwide problem in dairy farms, in part because of the pathogenicity of the bacteria, biofilm formation, and mechanisms of antimicrobial resistance that make the disease difficult to diagnose and treat, which is typically done with the use of beta-lactam antibiotics. The aim of the present study was to determine the virulence and resistance factors of S. aureus isolates from subclinical mastitis, blaZ + /mecA − /mecC − , resistant and sensitive to oxacillin. All isolates were classified as CC97 by MLST analysis, a clonal complex well adapted to the mammary gland and although STAU23 and STAU73 were resistant to oxacillin while STAU32 and STAU78 were sensitive, the genomic analysis identified only the blaZ operon corresponding to resistance to beta-lactams. However, the presence of the sdrC gene was revealed exclusively in resistant isolates, an important adhesin in the colonization process that potentiates pathogenicity in S. aureus. In addition, resistance islands (REIs) were identified in these isolates, suggesting more conserved REIs. In the analysis of SNPs throughout the genome, mutations were found in the trmB and smpB genes of the resistant isolates and in the murD and rimM genes of the sensitive isolates. This study highlights the potential benefit of genome-wide characterization tools to identify molecular mechanisms of S. aureus in bovine mastitis.

Supplementary Information

The online version contains supplementary material available at 10.1007/s42770-023-01243-4.

Introduction

Mastitis in dairy cows is one of the most common and highly debilitating diseases in the dairy industry. In addition to the significant negative impact on milk production, it increases susceptibility to other diseases and has a significant impact on the reproductive system [1].

The subclinical form has a higher incidence in mastitis cases and often progresses to chronic infections due to its asymptomatic nature [2]. In association with this chronicity, Staphylococcus aureus is an opportunistic pathogen often associated with subclinical bovine mastitis, exhibiting low cure rates and widespread distribution in dairy herds worldwide [3].

S. aureus possesses various mechanisms that contribute to its pathogenicity, facilitating invasion, colonization, dissemination, modulation of the host immune system, and tissue survival [4]. Numerous studies have aimed to enhance our understanding of the molecular mechanisms employed by S. aureus, and such knowledge could be instrumental in developing therapeutic strategies for mastitis and other infections caused by this pathogen. However, issues such as the phenotype-genotype relationship of antimicrobial resistance, persistence, and factors associated with the clinical and subclinical manifestations of the infection remain unclear.

Beta-lactams are widely used as antimicrobials in the treatment of mastitis. However, reports of beta-lactam-susceptible blaZ-positive S. aureus represent a diagnostic challenge for routine clinical microbiology laboratories and for clinicians treating infections caused by S. aureus that have the resistance marker gene; however, phenotypic tests show susceptibility [5–8]. Rocha et al. [9] evaluated beta-lactam-resistant and sensitive isolates that contained the blaZ gene and detected polymorphisms in the blaZ operon associated with beta-lactam resistance and sensitivity.

This study aimed to analyze the complete genome of S. aureus isolates obtained from bovine mastitis cases, both beta-lactam-resistant and -sensitive, carrying the blaZ gene. The objective was to understand the mechanisms associated with virulence and antibiotic resistance, with a particular focus on beta-lactams.

Methods

Bacterial isolates

Four Staphylococcus aureus isolates from cattle known to have mastitis in the state of Pernambuco, Brazil, were used. Bacteria were isolated from the milk collected from the streak canal, and their identification was performed by assessing morphological, biochemical, and staining characteristics [10]. The isolates were cultivated on brain heart infusion (BHI) agar and stored under refrigeration (± 4°C). These isolates were previously evaluated and classified as containing the blaZ gene [11] and negative for the mecA [11] /mecC gene [12]; two isolates were resistant (STAU23 and STAU73), and two were sensitive (STAU32 and STAU78) to oxacillin by minimum inhibitory concentration (MIC) [9]; additionally, beta-lactamase production was confirmed using nitrocefin disks [9].

Antibiotic sensitivity test

Sensitivity testing was performed using the disk diffusion method on Mueller–Hinton Agar medium for penicillin G (10 IU), gentamicin (10 μg), ciprofloxacin (5 μg), tetracycline (30 µg), tylosin (30 µg), and ceftiofur (30 µg) representative of the most commonly used classes of antimicrobials in animal production, to assess sensitivity to a variety of antimicrobials. Plates with inoculated cultures were incubated at 37 °C for 24 h. Sensitivity was assessed based on the diameters of the inhibition zones, according to the standards for antimicrobial disk susceptibility testing [13, 14].

DNA extraction

DNA was extracted from the isolates using the protocol described by Regitano et al. [15]. The integrity of the extracted DNA was verified by electrophoresis in a 1% agarose gel stained with ethidium bromide. The concentration of DNA present in the sample, as well as its purity, was determined with Qubit Fluorometric Quantification (Invitrogen®).

Genome sequencing of isolates

Sequencing libraries (850 bp fragments) were prepared using the Nextera® DNA Library Prep kit (Illumina®). For the genomic DNA fragmentation step and addition of adapters, DNA at 0.2 ng/µL was used according to the Illumina® protocol.

After preparation, the libraries were analyzed for their concentration by quantification in Qubit® and by electrophoresis in a 2% agarose gel stained with ethidium bromide for 2 h (80 V, 120 mA, and 20 W) to assess their integrity and size. The samples were sequenced on an Illumina MiSeq platform with Reagent Kits v2 (Illumina®).

Assembly and annotation of genomes

The quality of the reads was verified with FastQC software [16], and low-quality regions with Phred < 20 and adapters were removed with Trimmomatic 0.39 software [17]. The filtered reads were used in the de novo assembly process using SPAdes version 3.6.0 software [18]. The quality of the assembled genomes and assembly metrics were determined using Quast [19].

Identification of coding sequences (CDS) and genome annotation were performed using Prokka 1.12 [20], using the Staphylococcus database provided by the software, in which the protein-coding genes, tRNAs, and rRNAs were predicted. Prior to the annotation step, the assembled genomes were sorted using CONTIGuator software (https://contiguator.sourceforge.net/).

Determining sequence types and spa types

Multilocus sequence typing (MLST) was performed to determine sequence types (STs). The MLST technique for S. aureus is based on seven genes: conserved (housekeeping), arcC (carbamate kinase), aroE (shikimate dehydrogenase), glpF (glycerol kinase), gmk (guanylate kinase), pta (phosphate acetyltransferase), tpi (triosephosphate isomerase), and yqi (acetyl coenzyme A acetyltransferase). The STs of the isolates were determined based on WGS data available on the Central for Genomic Epidemiology-CGE server [21]. This method uses short sequence reads from sequencing or pre-assembled genomes that are compared at each locus with those of known alleles in the S. aureus MLST database (https://pubmlst.org/saureus) to obtain allelic and determine STs using a BLAST-based classification method. Based on ST clustering, clonal complexes (CCs) were predicted using eBURSTv3 [22].

Spa types were predicted using spaTyper v1.0 webserver from the Center of Genomic Epidemiology (https://cge.cbs.dtu.dk/services/spatyper). The spa typing technique compares the 21- to 27-bp polymorphic variable number of tandem repeats (VNTR) of the Staphylococcus protein A (spa) gene to assign a unique repeat code corresponding to their spa type.

Identification of virulence and antimicrobial resistance genes

The contig-based search method ABRicate v.0.8.13 (https://github.com/tseemann/abricate), with identity and minimum coverage of 80%, was used to identify virulence factors in the Virulence Factor Database (VFDB) and antibiotic resistance genes by combining four databases: ARG-ANNOT v3 (Antibiotic Resistance Gene ANNOTation), MegaRES v1.0.1, Comprehensive Antibiotic Resistance Database v1.1.6 (CARD), and ResFinder of the Center for Genomic Epidemiology.

Genomic islands prediction

The prediction of genomic islands (GEI), which include islands of pathogenicity (PAI) and islands of resistance (REI), was performed using GIPSy-Genomic Island prediction Software v.1.1.2 [23]. For this analysis, the file in gbk format of each isolate and the reference genome of S. aureus NCTC 8325 (https://www.ncbi.nlm.nih.gov/nuccore/NC_007795.1) with the parameters recommended by the authors were used as described in the manual. The software BRIG-BLAST Ring Image Generator v.0.95 [24] was used to generate the circular image of all genomes with the visualization of GEIs, PAIs, and REIs found.

SNP detection

Reads filtered by Trimmomatic v.0.39 were aligned by bowtie2 against the S. aureus NCTC 8325 reference genome (NC_007795.1). For the detection of SNPs, each file in the Binary Alignment Map (BAM) format was ordered by coordinates. In addition, PCR duplicates were identified and marked using Picard v.2.18.2 (broadinstitute.github.io/picard/). Descriptive statistical tabulation of the BAM dataset resulting from sequence mapping was performed using Samtools v.1.9. Calling SNP-type variants was performed using GATK v3.8 with a minimum mapping quality of 30 (Phred scale) and using HaplotypeCaller, which builds all haplotype combinations, identifies the most likely ones, and calculates mutations. Mutations were filtered based on the quality of the bases with the parameter “MQ > 40” (MQ = root mean square of mapping quality). The resulting variant call file (VCF) format files were annotated using the SnpEff v 4.3.1 T tool (http://snpeff.sourceforge.net/), which predicts the effects of genome-wide SNPs (SNPome).

Results

Antibiogram

The isolate STAU23 exhibited resistance to three antibiotics (penicillin G, ciprofloxacin, and gentamicin) belonging to three distinct classes. Isolate STAU32 demonstrated resistance to penicillin G and gentamicin as well as intermediate sensitivity to ciprofloxacin. The isolate STAU78 was resistant exclusively to penicillin G and displayed intermediate sensitivity to gentamicin. All the isolates were sensitive to tetracycline, tylosin, and ceftiofur (Table 1).

Table 1.

Antibiogram results for penicillin G, ciprofloxacin, gentamicin, tetracycline, tylosin, and ceftiofur of isolates STAU23, STAU32, STAU73, and STAU78

Antimicrobial class	Antibiotic	STAU23	STAU32	STAU73	STAU78
Beta-lactam	Penicillin G 10 μg	14 mm (R)	12 mm (R)	10 mm (R)	12 mm (R)
Fluoroquinolone	Ciprofloxacin 5 μg	12 mm (R)	20 mm (I)	20 mm (I)	22 mm (S)
Aminoglycoside	Gentamicin 10 μg	8 mm (R)	12 mm (R)	12 mm (R)	14 mm (I)
Tetracycline	Tetracycline 30 μg	20 mm (S)	26 mm (S)	20 mm (S)	20 mm (S)
Macrolide	Tylosin 30 μg	22 mm (S)	25 mm (S)	24 mm (S)	24 mm (S)
Cephalosporin	Ceftiofur 30 μg	23 mm (S)	32 mm (S)	29 mm (S)	28 mm (S)

R resistent, I intermediate, S susceptible

Genome assembly of S. aureus

The sequencing data resulted in sequences with coverage ranging from 33.96 to 179.07 times the size of the Staphylococcus aureus genome, with an average Phred quality score of 38 and an average of 581,272 reads (paired end) (S1).

The generated sequences were sufficient for the de novo genome assembly. Genomic assembly quality descriptions, including genome size, guanine-cytosine (GC) content, and N50 and L50 values, are presented in S2. In the ordering of the contigs performed using the software CONTIGuator (S3), the similarity of the alignments of the sequenced isolates with the reference genome was observed. The descriptions of the annotations are presented in Table S4.

Identification of STs and spa types

MLST analysis (Table 2) grouped the four S. aureus isolates into two distinct STs and one CC. The isolates STAU23, STAU32, and STAU78 were assigned to ST126, while isolate STAU73 was assigned to ST97. eBURST analysis of pooled STs classified all isolates as CC97. Spa typing identified two distinct types, t605 (isolated from STAU23, STAU32, and STAU78) and t521 (isolated from STAU73), which corroborated the MLST results, with spa type t605 being related to ST126 and t521 to ST97.

Table 2.

Allele prediction of the seven housekeeping genes considered in the MLST analysis

Isolates	arcC	aroE	glpF	gmk	pta	tpi	yqiL	ST*	CC**
STAU23	3	68	1	4	1	5	40	126	97
STAU32	3	68	1	4	1	4	40	126	97
STAU73	3	1	850	1	1	5	3	97	97
STAU78	3	68	1	4	1	5	40	126	97

*Sequences type

**Clonal complex

Virulence factors

An overview of the identified virulence genes in the evaluated isolates is presented in Fig. 1 using a Venn Diagram. Among the 84 virulence factors, 52 genes were detected in all isolates.

Fig. 1.

Venn diagram with virulence factors identified in STAU23, STAU32, STAU73, and STAU78

Among the 28 adhesion-related genes, 12 (atl, ebp, efb, fnbA, fnbB, eap-map, icaA, icaB, icaC, icaD, icaR, and sdrE) were present in all the isolates sequenced in this study (Table 3).

Table 3.

Distribution of virulence factors related to adhesion in isolates of S. aureus (STAU23, STAU32, STAU73, and STAU78)

Virulence factors	Genes	STAU23	STAU32	STAU73	STAU78
Accumulation of virulence-associated proteins	aap	−	−	−	−
Biofilm-associated surface protein	bap	−	−	−	−
Autolysine	atl	+	+	+	+
Agglutination factors (A, B)	clfA	−	+	+	+
	clfB	−	+	−	+
Collagen adhesion	cna	−	−	−	−
Elastin-binding protein	ebp	+	+	+	+
Fibronectin-binding proteins	ebh	+	+	−	+
	efb	+	+	+	+
	uafA	−	−	−	−
	fnbA	+	+	+	+
	fnbB	+	+	+	+
Extracellular adhesion/major histocompatibility complex analog protein	eap-map	+	+	+	+
Cell wall anchor surface proteins	sasC	−	−	−	−
	sasG	−	−	−	−
	sasP	−	−	−	−
Intercellular adhesins	icaA	+	+	+	+
	icaB	+	+	+	+
	icaC	+	+	+	+
	icaD	+	+	+	+
	icaR	+	+	+	+
Serine-aspartic acid rich fibrinogen binding proteins	sdrC	+	−	+	−
	sdrD	−	−	+	−
	sdrE	+	+	+	+
	sdrF	−	−	−	−
	sdrG	−	−	−	−
	sdrH	−	−	−	−
	sdrI	−	−	−	−

Gene presence ( +); gene absence ( −)

We highlight here that the clfB gene was exclusive to oxacillin-sensitive isolates (STAU32 and STAU78), while the sdrC gene was exclusive to oxacillin-resistant isolates (STAU23 and STAU73). The remaining 11 adhesion genes (aap, bap, cna, uafA, sasC, sasG, sasP, sdrF, sdrG, sdrH, and sdrI) were absent in all the isolates. All genes that include the ica operon were well characterized in all isolates evaluated.

Genes related to exoenzyme production (adsA, sspA, sspB, sspC, hysA, lip, geh, splA, splB, splC, splD, splE, splF, coa, and nuc) were present in all isolates (Table 4). Five of the 21 genes related to this category of virulence factors were not identified in any of the isolates.

Table 4.

Distribution of virulence factors related to exoenzymes in isolates of S. aureus (STAU23, STAU32, STAU73, and STAU78)

Virulence factors	Genes	STAU23	STAU32	STAU73	STAU78
Adenosine synthase A	adsA	+	+	+	+
Aureolysin	aur	+	+	−	−
Cysteine proteases	sspA	+	+	+	+
	sspB	+	+	+	+
	sspC	+	+	+	+
	sspD	−	−	−	−
	sspE	−	−	−	−
	sspF	−	−	−	−
Hyaluronate lyase	hysA	+	+	+	+
Lipases	lip	+	+	+	+
	geh	+	+	+	+
Serine proteases	splA	+	+	+	+
	splB	+	+	+	+
	splC	+	+	+	+
	splD	+	+	+	+
	splE	+	+	+	+
	splF	+	+	+	+
Staphylocoagulase	coa	+	+	+	+
Staphylokinase	sak	−	−	−	−
Thermonuclease	nuc	+	+	+	+
Von Willebrand factor-binding protein	vWbp	−	−	−	−

Gene presence ( +); gene absence ( −)

None of the isolates showed capH, capl, capJ, chp, or scn genes corresponding to immune evasion, while the other related genes were present in all isolates (Table 5). The isdA, isdB, isdC, isdD, isdE, isdF, isdG, and srtB genes involved in iron acquisition and metabolism were present only in the STAU23 isolate (S5).

Table 5.

Distribution of Virulence Factors related to evasion of the immune system in isolates of S. aureus (STAU23, STAU32, STAU73, and STAU78)

Virulence factors	Genes	STAU23	STAU32	STAU73	STAU78
Capsular genes	capA	+	+	+	+
	capB	+	+	+	+
	capC	+	+	+	+
	capD	+	+	+	+
	capE	+	+	+	+
	capF	+	+	+	+
	capG	+	+	+	+
	capH	−	−	−	−
	capI	−	−	−	−
	capJ	−	−	−	−
	capK	+	+	+	+
	capL	+	+	+	+
	capM	+	+	+	+
	capN	+	+	+	+
	capO	+	+	+	+
	capP	+	+	+	+
Chemotaxis inhibitor protein	chp	−	−	−	−
Staphylococcal complement inhibitor	scn	−	−	−	−
Staphylococcal protein A	spa	+	+	+	+
Staphylococcal immunoglobulin ligand	sbi	+	+	+	+

Gene presence ( +); gene absence ( −)

Among the leukocidins, lukF-PV was identified in all isolates: hemolysin alpha (hly-hla), delta (hdl), and gamma (hlgA, hlgB, and hlgC). Beta hemolysin (hlb) was present only in STAU23. Type VII secretion system genes (esaA, esaB, essA, essB, and essC) were identified in STAU73 and STAU78, esxA only in STAU78, and esxB in STAU73 (Table 6).

Table 6.

Distribution of virulence factors related to toxins in isolates of S. aureus (STAU23, STAU32, STAU73, and STAU78)

Virulence factors	Genes	STAU23	STAU32	STAU73	STAU78
Alpha hemolysin	hly-hla	+	+	+	+
Beta hemolysin	hlb	+	−	−	−
Delta hemolysin	hld	+	+	+	+
Gamma hemolysin	hlgA	+	+	+	+
	hlgB	+	+	+	+
	hlgC	+	+	+	+
Leukocidins M	lukM	−	−	−	−
	lukF-like	−	−	−	−
Panton-Valentine Leukocidin	lukS-PV	−	−	−	−
	lukF-PV	+	+	+	+
Leukotoxins	lukD	−	−	−	−
	lukE	−	−	−	−
Toxic shock syndrome toxin	tsst	−	−	−	−
Exfoliative toxins	eta	−	−	−	−
	etb	−	−	−	−
	etc	−	−	−	−
	etd	−	−	−	−
Type VII secretion system	esaA	−	−	+	+
	esaB	−	−	+	+
	esaC	−	−	−	−
	essA	−	−	+	+
	essB	−	−	+	+
	essC	−	−	+	+
	esxA	−	−	−	+
	esxB	−	−	+	−
Phenol soluble modulin (PSM)	PSMα1	−	−	−	−
	PSMα2	−	−	−	−
	PSMα3	−	−	−	−
	PSMα4	−	−	−	−
	PSMmec	−	−	−	−
	PSMβ1	−	−	−	−
	PSMβ2	−	−	−	−
	PSMβ3	−	−	−	−
	PSMβ4	−	−	−	−
	PSMβ5	−	−	−	−
	PSMβ6	−	−	−	−

Gene presence ( +); gene absence ( −)

None of the 21 genes encoding enterotoxins (sea, seb, sec, sed, see, seg, seh, sei, sej, selK, selL, selM, selN, selO, selP, selR, selU, selV, yent1, and yent2) were identified. However, some genes related to the production of staphylococcal exotoxins (set) were identified (Table 7).

Table 7.

Distribution of exotoxin-related virulence factors in S. aureus isolates (STAU23, STAU32, STAU73, and STAU78)

Virulence factors	Genes	STAU23	STAU32	STAU73	STAU78
Exotoxins	set1	−	−	−	−
	set2	−	−	−	−
	set3	−	−	−	−
	set4	−	−	−	−
	set5	−	−	−	−
	set6	−	−	−	−
	set7	−	−	+	−
	set8	−	−	−	−
	set9	−	−	−	−
	set10	+	+	+	+
	set11	+	+	−	+
	set12	−	−	+	−
	set13	+	+	−	+
	set14	−	−	−	−
	set15	−	−	−	−
	set16	+	+	−	+
	set17	−	−	−	−
	set18	+	+	+	+
	set19	+	+	+	+
	set20	−	−	−	−
	set21	−	−	−	−
	set22	−	−	+	−
	set23	−	−	−	−
	set24	−	−	−	−
	set25	−	−	+	−
	set26	+	+	+	+
	set27	−	−	−	−
	set28	−	−	−	−
	set29	−	−	−	−
	set30	−	−	−	−
	set31	+	+	−	+
	set32	−	−	−	−
	set33	−	−	−	−
	set34	−	−	−	−
	set35	−	−	−	−
	set36	−	−	−	−
	set37	−	−	−	−
	set38	−	−	+	−
	set39	+	+	−	+
	set40	−	−	−	−

Gene presence ( +); gene absence ( −)

Antibiotic resistance by S. aureus

The resistance profiles of the isolates sequenced in this study are presented in Table 8, with the genes identified by alignment of the databases used.

Table 8.

Resistance genes detected in the genomes of STAU23, STAU32, STAU73, and STAU78

Antimicrobial class	Detected genes	STAU23	STAU32	STAU73	STAU78
Aminoglycosides	aac3	+	+	+	+
	aph3-prime	+	+	+	+
	ant6	−	−	+	−
Βetalactams	blaI	+	+	+	+
	blaR	+	+	+	+
	blaZ	+	+	+	+
Phosphomycin	fosB	+	+	+	+
MDR (multi-drug resistance) efflux pumps	lmrS	+	+	+	+
	arlR	+	+	+	+
	arlS	+	+	+	+
	mepA	+	+	+	+
	mepB	+	+	+	+
	mepR	+	+	+	+
	mgrA	+	+	+	+
	norA	+	+	+	+
	norB	+	+	+	+
	qacG	+	+	−	−
	tet38	+	+	+	+
Macrolides, lincosamides and streptogramins	rlmH	+	+	+	+

Gene presence ( +); gene absence ( −)

As expected, the bla operon was identified in all isolates, and the blaZ gene encoding beta-lactamase and its regulators, blaR1 and blaI, were the only mechanisms related to resistance to beta-lactams identified by the databases. Regarding the genes involved in aminoglycoside resistance, all isolates presented aac3 and aph3-prime, and only STAU73 presented a nucleotidyltransferase encoded by ant6. The fosB gene was identified in all isolates, as were the genes involved in MDR efflux pumps (arlR, arlS, mgrA, norA, norB, lmrS, mepA, mepB, mepR, and tet38) and the lmrS gene, which belongs to the superfamily of key facilitators (MFS). All isolates harbored the rlmH gene, which encodes 23S methyltransferases, preventing the action of the MLS group of antibiotics (macrolides, lincosamides, and streptogramins).

Genomic islands

In the GIPsy analysis, two pathogenicity islands (PAIs) were identified in STAU23 and STAU32 and three in STAU73 and STAU78. One resistance island (REIs) was identified in STAU23 and STAU73, in addition to four genomic islands (GEIs) only in STAU73, which, during the PAI prediction step, islands of pathogenicity were not considered because they did not have sufficient virulence factors (Table 9).

Table 9.

Number of PAIs, REIs, and GEIs predicted by GIPSy in STAU23, STAU32, SYAU73, and STAU78

Isolates	PAI	GEI	REI
STAU23	2	–	1
STAU32	2	–	–
STAU73	3	4	1
STAU78	3	–	–

PAI pathogenicity island, GEI genomic island, REI resistance island

The gene content of the predicted islands for all isolates mostly consisted of hypothetical proteins (S9, S10, S11, and S12). REIs were identified only in oxacillin-resistant isolates (Figs. 2 and 3), and the REI predicted in STAU23 contained the bla operon between two Tn552 resolvase transposases, whereas the REI of the STAU73 isolate contained a protein related to the ABC transporter and a protein of permease/ATP binding of macrolide export. The circular genome with islands identified in the oxacillin-sensitive isolates is represented in S13 and S14.

Fig. 2.

Circular visualization of the genomes of isolates STAU23, STAU32, STAU73, and STAU78 with PAIs and REIs predicted by Gipsy in isolate STAU23 is highlighted in red. Reference genome (NCTC8325) represented in the outermost ring. The figure was built using the BLAST Ring Image Generator (BRIG)

Fig. 3.

Circular visualization of the genomes of Isolates STAU23, STAU32, STAU73, and STAU78 with PAIs, GEIs, and REIs predicted by Gipsy in isolate STAU73 highlighted in red. Reference genome (NCTC8325) represented in the outermost ring. The figure was built using the BLAST Ring Image Generator (BRIG)

Variant analysis

The reads of STAU23, STAU32, STAU73, and STAU78 were correctly aligned to the reference genome with mapping of 93.85%, 91.91%, 90.20%, and 94.02%, respectively.

We identified 9077 SNPs in STAU23, 8837 in STAU32, 9127 in STAU73, and 9781 in STAU78. The number of effects as a result of the SNPs found is represented in Figure S6, being the type of modifying effect with the highest percentage (> 90%) in all isolates.

A high-impact effect corresponded to the variant responsible for a probable protein truncation or loss of function, which was observed in a lower percentage, 0.035% (STAU23), 0.033% (STAU32), 0.025% (STAU73), and 0.036% (STAU78). Approximately 5% of all isolates were classified as low-impact variants, in which it was assumed that there was no change in the protein conformation. For moderate-impact variants, approximately 2% in all isolates, the variant is assumed to be non-disruptive; however, it may cause changes in the protein. The variants were also grouped by functional class (S7), with silent variants being the most predominant, followed by missense and nonsense variations for all isolates.

According to the SNPeff annotation, the SNPs that caused the greatest number of effects were downstream (defined as 5-kb downstream of the polyA addition site) and upstream (defined as 5-kb upstream of the transcription site start), followed by synonymous (the variant causes a codon that produces the same amino acid), missense (the variant causes a codon that produces a different amino acid), and SNPs in the intergenic region(S8).

The identified SNPs involved 2262 genes, shown in Figure S15 in a Venn Diagram with each isolate’s unique and shared genes. The SNPs of trmB, smpB, and 10 other genes not yet characterized were exclusive to STAU23 and STAU73 (Tables S16, S17,S18, and S19), and the SNPs of the murD, rimM, and 24 uncharacterized genes were exclusive to STAU32 and STAU78 (S20, S21, S22, and S23).

Discussion

Mastitis caused by Staphylococcus aureus poses a significant challenge in dairy farms worldwide due to the pathogenicity of the bacteria, biofilm formation, and mechanisms of antimicrobial resistance, which complicate diagnosis and treatment, often relying on beta-lactam antibiotics. In this study, we provided a comprehensive genomic description of S. aureus isolates obtained from cases of subclinical mastitis, specifically focusing on blaZ + /mecA − /mecC − isolates that exhibited resistance or sensitivity to oxacillin.

Through MLST analysis, all isolates were classified as belonging to CC97, a clonal complex well-adapted to the mammary gland. The genotype ST126 has been described as a bovine pathogen that is highly associated with milk, with ST97 being its probable ancestor ST [25]. This genotype is prevalent in several herds in southern Brazil [26], and recently S. aureus ST126 phenotypically resistant to oxacillin was reported for the first time associated with cattle harboring the mecC gene in Pernambuco [27]. Although there have been few studies on the characterization of S. aureus isolates from cases of bovine mastitis in northeastern Brazil, this study has added to our understanding of multilocus sequence typing (MLST) and spa typing based on WGS data, as the availability of these data allows us to compare the sequences of S. aureus from various origins to understand the hosts that serve as pathogen reservoirs.

In the genomic analysis of resistance genes, blaZ was identified in all isolates, the only mechanism of resistance to beta-lactams found in the databases corroborating the observed resistance to penicillin G in all isolates (Table 1). Previous research conducted by Rocha et al. [9] reported that the STAU23 and STAU73 isolates displayed minimum inhibitory concentrations (MICs) of 25 µg ml⁻¹ and 10 µg ml⁻¹ for oxacillin, respectively. According to the guidelines established by the Clinical and Laboratory Standards Institute (CLSI), strains are classified as methicillin-resistant Staphylococcus aureus (MRSA) when their oxacillin MIC is ≥ 8 µg ml⁻¹, indicating the presence of a PBP2a-mediated resistance mechanism [28]. However, it should be noted that these S. aureus isolates, which demonstrate resistance to penicillin-resistant penicillins (PRPs) such as oxacillin, do not possess an altered penicillin-binding protein (PBP2a) encoded by the mecA or mecC genes, as determined by previous PCR analysis and now confirmed by whole-genome sequencing (WGS).

The underlying mechanism responsible for this phenomenon remains undefined, whether it involves the overproduction of beta-lactamase, the synthesis of novel beta-lactamases encoded by plasmids, or the modification of penicillin-binding protein (PBP) genes resulting from spontaneous amino acid substitutions in the transpeptidase domain [29]. In a previous study, we extensively investigated the impact of polymorphisms in the entire bla operon of these isolates, and mutations were identified specifically in the blaZ gene. These mutations resulted in altered binding site positions within the beta-lactamase enzyme encoded by blaZ, distinguishing between oxacillin-sensitive and resistant isolates [9].

Strains of S. aureus that are misclassified as either methicillin-resistant or methicillin-sensitive present significant epidemiological and therapeutic challenges. Moreover, these strains hold a considerable “One Health” significance as they are frequently isolated from both human and animal sources, with prevalence observed in both hospital and community settings [28].

In the genomic analysis of virulence genes, the presence of clfB genes (exclusive to STAU32 and STAU780) and sdrC (exclusive to STAU23 and STAU73) was identified. These genes encode microbial surface components that recognize adhesive matrix molecules (MSCRAMMs), which are surface proteins anchored to the cell wall and bound to peptidoglycan [30]. These surface proteins play a crucial role in facilitating bacterial adhesion to host cells and tissues, serving as the initial step in promoting tissue damage, dissemination, and evading the host immune system [30]. Among the numerous virulence factors expressed by S. aureus, these proteins are particularly important for enhancing the pathogen’s in vivo fitness and success, as they are associated with the release of biofilm-related proteins and enable greater bacterial persistence in the mammary gland [31].

Regarding the sdrC gene, which was exclusively identified in oxacillin-resistant isolates, previous studies have reported its positive expression during the colonization process and invasion of S. aureus, indicating its significant role as an adhesin in the pathogenicity of the bacterium [32]. Furthermore, SdrC has been found to form dimers through intermolecular interactions, suggesting an enhanced efficiency in bacterial-host binding [33]. The findings from this study contribute to the understanding of pathogenic adhesins, which may serve as potential targets for therapeutic interventions.

Hemolysins are virulence factors of S. aureus that contribute to bacterial invasion and escape from the host immune response, especially in cases of chronic infections of the mammary gland [34]. In STAU23, alpha and beta hemolysins were identified, and the interaction between them can increase adherence to bovine mammary epithelial cells and the proliferation of S. aureus. In addition, these toxins can remain stable at high temperatures, which represents an important public health issue given the considerable consumption of milk and its derivatives by the human population [35]. The lukF-PV gene identified in the isolates encodes Panton-Valentine leukocidin (PVL), a leukotoxin belonging to the family of pore-forming toxins, which leads to cell lysis of mammalian leukocytes, including neutrophils, monocytes, and macrophages [33]. Furthermore, it has been shown that even at low concentrations, PVL toxin can stimulate neutrophils to increase the production of pro-inflammatory factors, which represents an important role of these leukotoxins in the initial phase and in the progression of bovine mastitis because of their influence and ability to modulate the immune system [34].

In the analysis of variants, SNPs were found in smpB and trmB of STAU23 and STAU73 isolates. The smpB gene forms the transfer messenger RNA (tmRNA)–SmpB ribonucleoprotein complex in the trans-translation mechanism, which comprises the key to quality control, ensuring high fidelity in proteins synthesized in bacteria, as well as playing an important role in genetic changes in response to stress, pathogenesis, and differentiation [36]. Because this complex has already been identified in approximately 99% of bacterial species, including S. aureus, transtranslation-targeted inhibitors represent a promising drug target, as tested in community- and hospital-associated MSSA and MRSA strains [37]. The protein encoded by trmB is involved in the biosynthesis of the N(7)-methylguanine-tRNA pathway, which is a part of tRNA modification. RNA modifications aid in the acquisition of antibiotic resistance along with ribosomal protection proteins, and because they are involved in the translation process, they can act as potential antibacterial targets [38]. Although the smpB and trmB genes have been reported in the annotation of SNPs, the mutations were of low, moderate, and modifying impact, making it necessary to verify whether this type of SNP could affect the gene itself and alter the effectiveness of the potential targeted drugs.

The SNPs found in STAU32 and STAU78 were exclusively involved in rimM and murD. The rimM gene encodes a 16S rRNA processing protein (ribosomal maturation factor), and mutations in this gene have already been reported and classified as mutations that affect the survival and fitness of bacteria when subjected to stress and can lead to an inadequate response to antibiotics, contributing to a false-positive effect of the antibiotic in question [39]. The murD gene, on the other hand, encodes an enzyme that catalyzes the formation of a peptide bond between UDP-N-acetyleyl-l-alanine and d-glutamic acid, which is important in peptidoglycan synthesis and cell wall formation by S. aureus and is considered an important target for the development of new antibacterial agents [40]. In this study, we found SNPs that affected this gene, which may be useful to consider when designing MurD inhibitors with high activity against the MurD enzyme of S. aureus.

Conclusion

Considering the high prevalence of S. aureus mastitis in cattle, knowledge of the entire genomic content of the different mechanisms of virulence and resistance is important to understand this pathogen during infection, especially in the subclinical form. The isolates of S. aureus (blaZ + /mecA − /mecC −) belonged to CC97, which represents a bovine pathogen strongly associated with milk and is well adapted to the mammary gland.

The exclusivity of the clfB gene in oxacillin-sensitive isolates (STAU32 and STAU78) and the sdrC gene in oxacillin-resistant isolates (STAU23 and STAU73) was revealed, which is required in the colonization process for invasion in S. aureus, suggesting an important role of SdrC adhesin in the pathogenicity of S. aureus. Furthermore, REIs were identified only in STAU23 and STAU73, which contributes to the success and potential mechanism of these isolates in mastitis in conjunction with adhesin SdrC. Mutations that were present exclusively in the trmB and smpB genes of the resistant isolates (STAU23 and STAU73) and in the murD and rimM genes of the sensitive isolates (STAU32 and STAU78) may be useful for the study and design of potential drugs against S. aureus, as well as for comparative studies focused on the molecular mechanisms of infection, useful in targeting and developing strategies for the prevention or treatment of mastitis.

Nucleotide sequence accession numbers

The draft genome sequences of STAU23, STAU32, STAU73, and STAU78 are available in GenBank under accession numbers JAKJJJ0000000000, JAKKID000000000, JAKRWN000000000, and JAKRWN000000000, respectively.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

Gabriela Dias Rocha: data curation; formal analysis; methodology; validation; writing—review and editing; writing—original draft. João José de Simoni Gouveia: conceptualization, methodology, supervision, writing—review and editing. Mateus Matiuzzi da Costa: conceptualization, methodology, supervision, writing—review and editing. Riani Ananda Nunes Soares: writing—review and editing. Gisele Veneroni Gouveia: conceptualization, methodology, supervision, writing—review and editing.

Funding

This study was funded by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior- CAPES.

Data availability

Data supporting the findings of this study are available upon request.

Declarations

Conflict of interest

The authors declare no competing interests.