Recent Trends of Antibiotic Resistance in Staphylococcus aureus Causing Clinical Mastitis in Dairy Herds in Abruzzo and Molise Regions, Italy

This study aimed to investigate the recent trends of antibiotic resistance (AR) prevalence in Staphylococcus aureus isolated from the milk of animals with clinical mastitis in areas of the Abruzzo and Molise regions in Central Italy. Fifty-four S. aureus isolates were obtained from routine testing for clinical mastitis agents carried out in the author institution in the years 2021 and 2022 and were analyzed for phenotypic resistance to eight antibiotics recommended for testing by European norms and belonging to the antibiotic classes used for mastitis treatment in milk-producing animals. Moreover, the presence of 14 transferable genetic determinants encoding resistance to the same antibiotics was analyzed using qPCR tests developed in this study. Phenotypic resistance to non-β-lactams was infrequent, with only one 2022 isolate resistant to clindamycin. However, resistance to the β-lactam cefoxitin at concentrations just above the threshold of 4 µg/mL was observed in 59.2% of isolates in both years, making these isolates classifiable as methicillin-resistant. The AR genotypes detected were the blaZ gene (50% of 2021 isolates and 44.4% of 2022 isolates), aphA3-blaZ- ermC/T (one 2021 isolate), aphA3-ant6-blaZ-ermC/T (one 2021 isolate), blaZ-ermB (one 2022 isolate) and mecA-mph (one 2022 isolate). An inquiry into the veterinarians who provided the samples, regarding the antimicrobials prescribed for mastitis treatment and criteria of usage, indicated a possible causal relation with the AR test results. The occurrence of AR genotypes did not increase in time, most probably reflecting how mastitis was treated and prevented in farms. However, the frequently observed cefoxitin resistance needs to be explained genotypically, further monitored and limited by modifying antibiotic usage practices. The identification of a mecA-positive isolate in 2022 suggests further investigation if this genotype is emerging locally.


Introduction
Staphylococcus aureus is one of the main causative agents of mastitis in milk-producing animals and the first for clinical bovine mastitis with the ability to give causation to persistent intramammary infections [1]. This bacterial species is also a major human pathogen capable of causing food poisoning for the production of multiple heat-stable enterotoxins (SE), localized soft tissue or skin infections and systemic infections triggered by virulence factors comprising staphylococcal superantigens (SAgs) [2][3][4], cytotoxic proteins [5] and factors that favor colonization and immune evasion [6]. The emergence of antibiotic-resistant (AR) S. aureus strains, among which the methicillin-resistant Staphylococcus aureus (MRSA) are listed by the World Health Organization among the pathogens of "high priority" against which new antibiotics are urgently needed, worsens the threat to human health posed by this bacterial species, since MRSA cause infections with high mortality rates [7]. The transferable genetic element providing the MRSA phenotype is the chromosomal cassette mec (SCCmec) that most often carries the mecA or the mecC gene, and sometimes other rare homologues, encoding for additional penicillin-binding proteins (PBP2a) with reduced affinity for β-lactams plus genes for site-specific recombinases [8,9].
Use of antibiotics in the animal farming sector to treat conditions such as mastitis can cause MRSA that is transmissible to humans through raw milk and derived products [10]. Some risk factors have been identified for MRSA transmission in dairy farms, such as poor milking hygiene. However, the role of antimicrobial usage has been little investigated and only one study reported an increase in antibiotic minimum inhibitory concentration (MIC) values in the occurrence of AR genes tetK, tetM, and blaZ after enrofloxacin treatment of persistent mastitis in goats [11]. This study underlined the role of antimicrobial usage on the emergence of AR S. aureus strains.
Phylogenetic analysis based on multi-locus sequence typing (MLST) gives evidence that some S. aureus lineages are found both in human and animal hosts, in particular strains from bovine mastitis, as a consequence of transference from humans to animals and vice versa. Moreover, it was demonstrated that S. aureus has the capacity to switch hosts [12], therefore S. aureus with a resistance to antibiotics circulating among farm animals must be considered a threat to public health.
Analysis of the prevalence of antibiotic-resistant strains of S. aureus can indicate if risk factors that favor their increase in farms are active and allow for the adoption of measures to reduce the dissemination of the genetic determinants encoding resistance. Therefore, this study was undertaken to analyze the prevalence of AR S. aureus in farms by taking into account isolates from the milk of animals affected by clinical mastitis and requiring antibiotic treatment. The study was carried out in areas of the Abruzzo and Molise regions in Central Italy, and aspects of mastitis management in the sampled farms were taken into account to understand if these can be linked by a causal relation to phenotypic and genotypic AR prevalence.

Rate of Mastitis Caused by S. aureus in 2021 and 2022
The number of farms with clinical mastitis caused by S. aureus were 16 among 56 analyzed in 2021 and 13 among 52 analyzed in 2022, accounting for 28.5% and 25% of mastitis outbreaks, respectively. More than one isolate was obtained from the same sample if colonies were of a different dimension and appearance, and hemolysis halo aspects were observed on the blood agar, and this led to obtaining 27 strains per year.

Phenotypic AR of S. aureus Isolates
In this study, isolates with cefoxitin MIC 6-8 µg/mL accounted for 59.2% of isolates in both years. According to the European Committee on Antimicrobial Susceptibility Testing (EUCAST) indications, these isolates are resistant to this antibiotic, though at low levels, and must be considered methicillin-resistant [13]. However, all of them were sensitive to oxacillin. The percentages of isolates assigned to groups with different cefoxitin MIC values in the years 2021 and 2022 are shown in Figure 1.
The distributions of the S. aureus isolates in different cefoxitin MIC groups in the two years were highly correlated (r = 0.99), thus indicating that there was very little variation in the percentage of isolates with a given cefoxitin resistance level in the investigation period.
The distribution of MIC values for all antibiotics tested in the two years is shown in Figure 2.
The MIC distributions for S. aureus isolates between the years 2021 and 2022 were not significantly different for any of the antibiotics considered. However, it can be noted that MICs for norfloxacin showed a shift to higher values in 2022, though resistance was not detected using the disc diffusion assay. According to the MIC values, all the isolates were susceptible to oxacillin, norfloxacin, erythomycin, gentamicin, kanamycin, tobramycin and clindamycin, except for one isolate that was resistant to the latter antibiotic.

Occurrence of AR Genes in the S. aureus Isolates
The AR genes sought in this study were those encoding resistance to the antibiotics of interest and found most frequently in S. aureus, as deduced from the consultation of the sequence databases. In addition, the cfr gene was sought in this study since it codes for a 23S rRNA methyltransferase that provides resistance to different antibiotic classes, among which are lincosamides and phenicols [14]. The primer/probe systems used are reported in Table 1.
For the ermC/ermT genes and for the mecA/mecC genes, a unique primer/probe system was designed in order to merge the diagnostic tests. A mecA-specific test was carried out on the sole mecA/C positive isolate to identify the mec gene homolog, which was then identified as mecA. Table 1. Primers and probes designed in this study for the detection of AR genes, respective targets and dimensions of the amplification products.

Primer and Probe Labels and Sequences (5 -3 ) Target Gene Amplicon Size (bp)
AadA12f a This primer/probe system targets both ermC and ermT genes; b these genes have alternative names in the sequence databases; c this primer/probe system targets both mecA and mecC genes and, coupled with the mecA-specific test, can allow for the detection of the mecC gene.
The AR gene detected in this study and the phenotypic AR for each isolate are reported in Table 2.
The blaZ gene occurrence was frequent and found in 59.2% of 2021 isolates and in 48.1% of 2022 isolates. Differences in the occurrence of this gene between the two years were not statistically significant according to the Student's t test. None of the isolates overexpressed blaZ to levels determining a borderline oxacillin resistant S. aureus (BORSA) phenotype [13]. Only a few genetic determinants beyond blaZ were identified in the isolates studied here. In particular, the mecA gene was found only in isolate six from 2022 (Table 2), which was resistant to cefoxitin (MIC 6 µg/mL) but sensitive to oxacillin, though the MIC value (2 µg/mL) was equal to the cut-off value above which S. aureus strains must be considered resistant [13].
Other AR genes occurring in the S. aureus isolates examined, namely aph3, ant6, ermB, ermC/T and mph, with the exception of ermB which was detected in a 2022 isolate harboring only this gene, were found mostly in association with other AR determinants. In particular, MDR genotypes ant6-aph3-blaZ-ermC/T and aph3-blaZ-ermC/T were each found in one 2021 isolate ( Table 2).
The gene mph for resistance to macrolides was found in the sole mecA-positive strain. This strain was also resistant to clindamycin, possibly for the presence of a genetic determinant different from the genes lnuB and cfr, tested in this study.

Evaluation of Antibiotic Management via Veterinarian Interview
In order to understand if the results of AR screenings might be linked with a causal relation to the antibiotic usage practices adopted locally, the 18 veterinarians providing medical care to the sampled farms were interviewed via a questionnaire regarding the antibiotics used, farm hygiene and the criteria adopted for decisions surrounding antibiotic use in clinical mastitis. The results of the interviews are presented in Table 3. It is possible to observe that all the antibiotic classes allowed for mastitis treatment were used; however, penicillins, cephalosporins and enrofloxacin prevailed.
Hygiene in the farms was considered good or acceptable in most cases, and the milking hygiene was found to be adequate in all instances. However, four veterinarians observed insufficient mastitis prevention measures in some farms. In addition, most farms were found to adopt adequate mastitis prevention measures and protocols for antibiotic usage. However, four veterinarians reported inadequate mastitis prevention measures in the farms, and these were the same veterinarians who also reported antibiotic treatment failure. Fourteen veterinarians defined farm and milking hygiene and mastitis prevention management good or adequate. Nevertheless, they reported cases of recidivating mastitis. Management of the total bacterial and somatic cell counts in bulk tank milk was referred to be good or adequate, thus showing little concern for the occurrence of subclinical mastitis.
Half of the interviewed veterinarians declared to be committed to the reduction of antibiotic usage and all of them declared to use antibiotics based on the antibiogram outcomes for the specific pathogens. Notably, all veterinarians observed cases of AR, though most of them defined those cases as rare.

Discussion
The AR genes sought in this study were those encoding resistance to the antibiotic classes used for mastitis therapy in dairy herds in the area of interest and found most frequently in S. aureus, as deduced from consultation of the sequence databases and from a recent survey on identity and the frequency of AR genes in 29,679 genomes of S. aureus isolated worldwide [16]. The qPCR technique was introduced in this study for some of the AR genes for which only end point PCR tests were available because they are more rapid, specific and sensitive. The design of new tests for genes for which qPCR tests were already described [17][18][19] was decided to ensure specificity for S. aureus after the BLASTn alignment of sequences available for the species in the public domain database. Oligonucleotides were designed to match all gene variants retrieved. In particular, for the blaZ and the mecA genes, qPCR tests were previously reported as primer/probe systems were redesigned in this study after the alignment via BLASTn of one thousand sequences for each gene. Therefore, the results can be considered comparable to those of former investigations carried out with different primer/probe systems. In addition, unique primer/probe systems were designed in this study to detect couples of genes with sequence similarity, i.e., ermC/ermT and mecA/mecC, to reduce the number of tests necessary to identify isolates with transferable AR genotypes.
The classes of β-lactams and fluoroquinolones prevailed among the antibiotics prescribed by veterinarians, and this could explain the high prevalence of strains harboring blaZ genes and the increase in norfloxacin MIC values in 2022. In investigations carried out in different countries, the blaZ gene was found at high frequencies as well, with a maximum of 95.7% [20]. The increase in norfloxacin MIC values could be indicative of an increase in resistance to quinolones that in S. aureus is mediated by the core gene norA, encoding different variants of an efflux pump, as well as other efflux systems [14]. These are increasingly expressed under antimicrobial pressure and can lead to the emergence of resistant phenotypes [21,22].
The isolation of only one strain harboring the mecA gene indicated a frequency lower than that reported in other studies [23][24][25][26] but similar to that reported in Southern Italy for the bulk tank milk of small ruminants [27], thus indicating that its prevalence can vary on a local basis. The mecA-positive strain was not resistant to oxacillin, showing an MIC equal to the S. aureus epidemiological cut-off (ECOFF) for this antibiotic. The occurrence of mecA-positive and oxacillin-sensitive strains was reported recently [28].
The percentage of isolates resistant to cefoxitin observed in this study was among the highest reported for European countries [29]. According to the EUCAST AST guidelines, S. aureus strains resistant to cefoxitin have an MIC > 4 µg/mL, a value that coincides with the ECOFF of the species for cefoxitin, and in most cases they harbor a mecA or mecC gene [13]. However, in this study cefoxitin resistance was not associated with the presence of the mecA or the mecC gene, resulting in an example of a mec-independent β-lactam resistant phenotype. The occurrence of cefoxitin resistant isolates without the mec genetic determinants was described previously [30][31][32][33][34], and different genetic features were found to determine resistance only to cefoxitin with an MIC of 6 µg/mL, namely mutations in native pbp1, pbp2, pbp3 and pbp4 penicillin-binding protein genes, mutations in the pbp4 promoter and in gene gdpP for a phosphodiesterase c-di-AMP regulator [30]. Further investigations should be devoted to defining the genetic basis of cefoxitin resistance in the isolates examined in this study.
An increase in AR to clindamycin, erythromycin, gentamycin and oxacillin was not observed, though it was reported to occur globally for S. aureus strains, causing bovine mastitis or isolated from milk and dairy products [35,36]. In addition, a recent meta-analysis carried out in China suggested how the AR rate can be particularly high in some geographical contexts, suggesting local misuse of antibiotics [37]. Therefore, it can be hypothesized that the low frequency of phenotypic AR for some antibiotic classes observed in this study is a consequence of the low usage of those drugs, as corroborated by veterinarian's statements. Nevertheless, isolates harboring AR genes encoding resistance to macrolides and aminoglycosides were identified in this study, though these were susceptible to the antibiotics for which resistance was encoded. Further experiments aimed at elucidating if those genes can be induced upon gradual exposure to antimicrobials should be carried out. Moreover, the occurrence of multiresistance-encoding mobile genetic elements should be investigated to assess the risk of MDR genotype dissemination.
According to the veterinarian statements, strains causing mastitis were isolated and tested via antibiogram only in the case of recidivating mastitis or in the case of treatment failure. This could imply that initial treatments were carried out without antibiogram execution, thus possibly leading to the selection of antibiotic resistant strains that become difficult to eradicate. Changes in this practice, together with improvements in mastitis management, could reduce the prevalence of AR S. aureus in farms. The answers of veterinarians to questions regarding hygiene in the farms, milking hygiene and mastitis prevention measures indicated good or acceptable levels in most cases. This might imply a low usage of antibiotics, since good farming practices reduce the occurrence of infections and need for antibiotic treatment [10]. On the other hand, the veterinarian's statements indicated that clinical mastitis was recidivating and sometimes impossible to treat. This could be a consequence of antibiotic misusage, so the protocols of antibiotic usage adopted in farms need to be examined and, if necessary, revised to reduce the risk of AR bacteria selection.

Bacterial Strains and Culture Conditions
The bacterial strains used in this study were all isolates from mastitic milk samples analyzed upon the request of veterinarians by the Istituto Zooprofilattico Sperimentale dell'Abruzzo e del Molise (IZSAM), Campobasso and Lanciano branches, for identification of the infectious agent and antibiogram execution. Strains phenotypically identified as S. aureus in routine analysis were obtained from 56 farms and 52 farms in 2021 and 2022, respectively. These were propagated by streaking on blood agar (10/L g tryptose, 10 g/L meat extract, 5 g/L NaCl, 15 g/L agar, 100 mL of defibrinated sheep blood added aseptically after autoclaving and cooling of the base medium) incubated in aerobic conditions at 37 • C for 24-48 h. Cell biomass from a colony isolated after two subsequent streaks on blood agar was used for each phenotypic or genotypic test. For long-term storage, the isolates were maintained in Microbank (Biolife Italiana, Milan, Italy) at −80 • C.

Phenotypic AR Testing
The antibiotics tested phenotypically, i.e., cefoxitin (FOX), clindamycin (CD), erythromycin (ERY), gentamicin (CN), kanamycin (KAN), norfloxacin (NOR), oxacillin (OXA) and tobramycin (TOB), were those of human usage belonging to the classes of antibiotics used for mastitis treatment and recommended for testing by the EUCAST to deduce resistance to antibiotic classes [13]. The minimum inhibitory concentration (MIC) values were determined by using the Liofilchem ® MIC Test Strips (Liofilchem, Roseto degli Abruzzi, TE, Italy) according to the instructions. The MIC values were assigned in accordance with EUCAST guidelines on antimicrobial susceptibility testing (AST) [38]. For norfloxacin, resistance was defined by using discs with 10 µg of the antibiotic (Liofilchem) as recommended [13]. The reference to the ECOFF values [13] was used to define the position of the new isolates in the range of observed MIC values for the species S. aureus.

Quantitative PCR Primer Design
New qPCR tests for the transferable AR genes encoding resistance to the antibiotics tested were designed by searching and aligning sequences in the NCBI databases (https: //www.ncbi.nlm.nih.gov/, accessed on 1 October 2022) and in the National Database of Antibiotic Resistant Organisms (NDARO, https://www.ncbi.nlm.nih.gov/pathogens/ antimicrobial-resistance/, accessed on 2 October 2022). For each gene, a BLASTN analysis restricted to the S. aureus taxon was carried out in order to consider different variants to be aligned so as to design oligonucleotides targeting all of them. The primer/probe systems designed in this study are listed in Table 1 with respective target genes and amplicon dimensions. These were synthetized by Eurofins Genomics (Ebersberg, Germany).
The gene regions comprised between each pair of oligonucleotides, ranging in size between 130 and 246 bp, were synthetized upon request by GenScript Biotech (Rijswijk, The Netherlands) and delivered as pUC57 vector constructs to serve as positive controls in the qPCR runs.

DNA Extraction
DNA was extracted from one loopful biomass resuspended in 200 µL of Macherey Nagel T1 buffer (Carlo Erba, Cornaredo, MI, Italy) containing 100 mg of sterile 200 µm diameter glass beads in safe lock Eppendorf tubes (Eppendorf). The suspension was beadbeaten in a TissueLyser II (Qiagen Italia, Milan, Italy) at 30 hz for 2 min). Then 200 µL of Macherey Nagel B3 buffer (Carlo Erba) were added, and the extraction was continued according to the Macherey Nagel Nucleospin Tissue (Carlo Erba) protocol.

Quantitative PCR Conditions
The qPCR reactions were carried out in a QuantStudio 5 thermal cycler (Thermo Fisher Scientific, Rodano, MI, Italy). Identification of isolates at the species level was carried out as previously described via a nucA-targeted qPCR test [39]. For AR gene detection, a unique program suitable for all the primer/probe systems designed was used. This comprised initial denaturation at 94 • C for 5 min and 40 cycles of denaturation at 94 • C for 15 s and annealing at 51 • C for 30 s. The qPCR reaction of 20 µL volume comprised 10 µL of Takara Premix Ex Taq (Probe qPCR) (Diatech, Jesi, AN, Italy), 0.2 µM primers and probe, TaqMan Exogenous Internal Positive Control Reagents (Thermo Fisher Scientific) in the recommended concentration, 2 µL of DNA sample and Nuclease Free water (Thermo Fisher Scientific) added to the final volume. Four nanograms of a synthetic, positive control construction were used in the positive control reaction.

Veterinarian Questionnaire
The 18 veterinarians who requested the bacteriological examinations and antibiograms for mastitis diagnosis in the years 2021 and 2022 were interviewed to identify the antibiotic classes prescribed, the criteria adopted for antibiotic usage and different aspects of mastitis management in farms by delivering a questionnaire with closed ended questions.

Statistical Analyses
MIC values plots, Student's t test evaluation of the distinctness of the MIC data series obtained in 2021 and 2022 and correlation analyses were carried out by using PAST 4.03 free statistical software downloaded from https://past.en.lo4d.com/windows (accessed on 23 December 2022). The data series were considered distinct for p < 0.05.

Conclusions
This study showed that the prevalence of both genotypic and phenotypic AR is currently low for non-β-lactam antibiotics and with no increasing trend in S. aureus isolates from the areas of Abruzzo and Molise considered. This is probably the consequence of the infrequent usage of those antibiotics, except for enrofloxacin, reported by the veterinarians interviewed. However, strains harboring β-lactam resistance blaZ genes, already known to be widespread in the species S. aureus, occurred frequently, probably for the preferential use of β-lactams in clinical mastitis therapy. Phenotypic resistance to cefoxitin in the mecA/Cnegative isolates was frequent, and its genetic basis needs to be identified in order to understand the molecular bases behind the emergence of this phenotype. Moreover, the occurrence of one MRSA and two genotypically MDR isolates suggest that monitoring the presence of these AR profiles in dairy herds should be continued to understand if these genotypes tend to disseminate. This study was limited to clinical mastitis cases in which the causative agent was isolated for diagnostic and antibiotic treatment purposes. In addition to the characterization of S. aureus strains causing clinical mastitis, investigations should be undertaken for farms with suspected subclinical mastitis to isolate the causative strains and define their AR status. This could help to elucidate if AR genotypes can be spread by strains causing subclinical mastitis. Indeed, these strains are more likely to persist for long time in farms, thus representing a relevant risk of the transfer of AR determinants.