Staphylococcus aureus and methicillin resistant staphylococcus aureus (mrsa) in dairy farms of central Italy: a characterization of isolates from milk and dairy products and an in-farm epidemiological investigation on mrsa colonization of sheep and in-contact humans

Staphylococcus aureus (S. aureus) is considered one of the most significant aetiological agent of intramammary infection in dairy ruminants, causing both clinical and subclinical forms accompanied with relevant economic losses due to reduced milk production and quality. S. aureus is also considered a major human foodborne pathogen. Milk and dairy products represent indeed potential sources of contamination by S. aureus preformed enterotoxins, which can be responsible of staphylococcal food-poisoning (SFP) outbreaks. S. aureus pathogenicity is also related to its capacity to quickly develop resistance against new antibiotics used in clinical practices. Among multidrug resistant strains, methicillin resistant Staphylococcus aureus (MRSA) represents a global health threat. In the last few years, the isolation of MRSA strains from both livestock and companion animals has gained growing attention, both in veterinary medicine and from a Public Health perspective. In particular, the presence of MRSA in livestock animals pose a potential risk of infection and/or colonization to humans both through direct contact, and through the consumption of contaminated food. However, there are still many unknowns regarding “MRSA epidemiology”, the risk factors involved in transinfection, and the direction of transmission between animals and humans. This is particularly true for small ruminants due to the fact that they are traditionally not included in national surveillance programs on antibiotic resistance, and related foodstuff are not routinely investigated for this purpose. MRSA population genetics and MRSA epidemiology, have been conventionally studied by using various molecular methods, such as multi-locus sequence typing (MLST) and typing of the S. aureus protein A (spa-typing), often in combination with the staphylococcal cassette chromosome mec (SCCmec) typing. Fingerprinting techniques such as pulsedfield gel electrophoresis (PFGE) have been also frequently implemented to study S. aureus infections in nosocomial settings, or in case of foodborne outbreaks. Currently, the MRSA clonal complex (CC) 398 is the most prevalent lineage among LivestockAssociated MRSA (LA-MRSA), but primarily “human” clones, such as the sequence type (ST) 1, have been also found in livestock. This work describes the genotypic and phenotypic characterization of S. aureus and MRSA isolates from dairy farms of Central Italy, with the main purpose to evaluate the potential zoonotic risks throughout the dairy production chain, in particular focusing on the small ruminants. The work was split in two parts: i) a preliminary study based on the characterization of S. aureus and MRSA strains isolated from a variety of milk and dairy products produced in Central Italy from different animal species; ii) a within-farm epidemiological investigation on the possible MRSA colonization of sheep and in-contact humans. In the first part, a total of 227 S. aureus colonies isolated from 54 samples of raw milk and dairy products of ovine, caprine, bovine and bubaline origin were tested for the presence of genes coding for staphylococcal enterotoxins (SEs/SEls) and for methicillin resistance. Ninety-three colonies from 31 of the 54 samples (57.4%), and from 18 (69.2%) of the 26 farms of origin tested positive for SEs/SEls genes. Most isolates harboured more than one toxin gene and 15 different toxinotypes were recorded. The most frequent were “sec” gene alone (28.6%), “sea, sed, ser, selj” (20%), “seg, sei” and “seh” (8.6%). The 77 colonies harbouring “classical enterotoxins” genes (sea-sed) were further tested for enterotoxin production, which was assessed for 59.2% of the colonies. Three MRSA isolates were detected in three different ovine milk/dairy product samples (1.3%) from two different sheep farms, both located in the province of Rome and apparently not epidemiologically related. All MRSA isolates belonged to spa type t127, ST 1, CC 1, SCCmec type IVa. In the second part, a within-farm epidemiological study was carried out in one of the two MRSA positive dairy sheep farms previously identified. A total of 556 individual milk samples were collected from all the lactating ewes of the herd, together with a bulk tank milk sample. Two weeks later, based on the results obtained from the individual milk samples, specimens (milk, nasal and wound swabs, udder skin samples) were further collected from two MRSApositive animals. Samples (nasal swabs, hand skin samples, and oropharynx swabs) were also collected from three persons in close contact with the animals (the farm-owner and two farm-workers). S. aureus was detected in 12 out of 556 (2.16%) individual samples, with 10 of them (10/556; 1.8%) positive for methicillin- susceptible S. aureus (MSSA) and two (2/556; 0.34%) for MRSA. MRSA was also isolated from the udder skin samples of both MRSA positive animals and from two out of the three persons examined. All the MRSA strains isolated from animal and human sources belonged again to spa type t127, ST 1, CC 1, SCCmec type IVa. However, pulsed field gel electrophoresis (PFGE) analysis performed on MRSA selected isolates of human and animal origin revealed that two closely pulsotypes, a human pulsotype (HP) and an ovine pulsotype (OP), circulated within the farm. The results of microarray analysis on selected strains representative of the two pulsotypes, revealed very similar profiles and only few differences in the gene composition (e.g. genes of the immune evasion cluster, IEC). The presence of IEC genes (sak, scn) was also confirmed by PCR analysis only in MRSA isolates showing the HP. MRSA selected isolates also displayed a very similar resistance phenotype being all resistant to cefoxitin, penicillin, erythromycin, clindamycin, streptomycin, kanamycin, and tetracycline. In conclusion, the occurrence of SEs/SEls positive isolates in a high proportion of milk and dairy product samples of Central Italy, including the detection also from “ready to eat” products of recently discovered SEs/SEls genes, is of concern and underline the need of standardised diagnostic methods to verify and quantify the presence of the “new” enterotoxins directly in food. The identification and genetic characterization of three MRSA isolates from two ovine farms represents the first Italian report on the occurrence of MRSA in ovine milk, and to the Author knowledge, the first genotyping of MRSA strains from ovine dairy products at international level. Although the prevalence was low, the isolation of MRSA from “ready to eat” food is of concern. The results of the within-herd study also demonstrated that the same MRSA strain (i.e., same sequence type, spa type, SCCmec type, PFGE and resistance profile) was able to persist over time at farm level, being detected two years after the first isolation, although the in-herd prevalence was very low. Yet, direction and routes of transmission of the MRSA farmspecific pulsotypes identified in this study remain unclear. Altogether, the results of these studies highlight the importance of applying effective biosecurity and hygiene practices at all times along the dairy production chain. Good hygiene and good manufacturing practices, as well as the use of microbiological testing at every level of the production chain should be implemented in order to control a possible spread of MRSA, and to minimize food poisoning risks.