Abstract
Background/Objectives:Staphylococcus aureus (S. aureus) intramammary infection remains a major global dairy problem due to its contagious nature, its ability to persist and colonize teat/skin and mucosal niches, and the often-limited bacteriological cure achieved with antimicrobial therapy. Beyond udder health, it is relevant to public health because it can enter raw milk chains and serve as a reservoir for antimicrobial resistance determinants that may circulate between dairy animals and humans. Methods: We assessed S. aureus' ecology in raw ewe milk from 75 sheep farms in Epirus (Greece) by sampling clinically healthy controls (group A) and clinical mastitis cases pre-treatment (group B), followed by resampling at the first post-withdrawal milking after penicillin/streptomycin treatment (group C1-therapeutic protocol 1), oxytetracycline treatment (group C2-therapeutic protocol 2), or enrofloxacin treatment (group C3-therapeutic protocol 3). Results:S. aureus detection was high and comparable across groups (A 23.0%, B 22.0-30.0%, C 20.0-22.0%), and paired analyses showed no significant pre-post shifts in detection/burden within therapeutic protocols (all p > 0.05). Nevertheless, persistence remained evident. The chromosomal gene mecA was detected in S. aureus strains in all groups, ranging from 13.6% in controls to 54.5% post-withdrawal in group C1, and was also present in the pre-treatment group. In paired sampling animals, mecA was mostly stable, with rare emergence or loss. Across antibiotic classes, within-animal resistance transitions were generally uncommon and non-significant (p > 0.05); β-lactam resistance was fully stable (p = 1.00). Descriptively, resistance to protein synthesis inhibitors tended to decline after therapy in protocol 1 and protocol 3, while protocol 3 showed post-treatment gains in fluoroquinolone resistance. By contrast, virulence-associated phenotype traits shifted after therapy: enterotoxigenicity increased post-withdrawal (especially in the C3 group), Staphylococcal Enterotoxin A (SEA) and Staphylococcal Enterotoxin B (SEB) appeared only post-therapy, Staphylococcal Enterotoxin D (SED) increased significantly in paired isolates (p = 0.002), and strong biofilm adherence increased (in C3, p = 1.5 × 10(-5)). Conclusions: The detection of S. aureus after therapy suggests that one possibility is that antimicrobial exposure may select for, or otherwise reshape, the residual intramammary population, rather than reliably eliminating it-an outcome that remains clinically relevant for udder health. Moreover, the persistence of mecA/methicillin-resistant Staphylococcus aureus (MRSA)-compatible profiles indicates that milk released to the food chain after withdrawal compliance may still harbor S. aureus with enhanced preservation capacity and significant food safety relevance.