Antimicrobial resistance and molecular typing of Staphylococcus aureus isolates from raw milk in Hunan Province

Introduction

Staphylococcus aureus has long been a threat to human health and the development of animal husbandry. In many countries, 30% to 40% cases of bovine mastitis are caused by S. aureus (Basanisi et al., 2017), mainly through adhesion, which leads to colonization in the mammary epithelial cells of livestock and results in udder abscesses, hardening, fistulae, necrosis, and exfoliation. Diseased bovines present with obvious udder fever, udder hardness and pain, and reddish to black milk with foul-smelling flocs, which not only reduces milk production and quality, but also endangers the animal’s life (Hiramatsu et al., 2002; Sinha & Fraunholz, 2010). Non-standard processing and disinfection procedures can easily result in the production of dairy products contaminated with S. aureus (Schmidt, Kock & Ehlers, 2017; Johler et al., 2018), and many types of S. aureus can be detected in raw dairy products worldwide (Alghizzi & Shami, 2021). S. aureus causes infection in the host body by releasing a variety of toxins, the most important being Panton–Valentine leukocidin (PVL). Encoded by lukF PV and lukS PV, PVL is a poor-forming toxin that causes host immune suppression and infection by attacking leukocytes, such as neutrophils and monocytes (Spaan, Van Strijp & Torres, 2017); S. aureus exfoliative toxin (Ets) causes severe inflammation of the gastrointestinal tract by stimulating intestinal epithelial cells (Banwell & Sherr, 1973). Toxic shock syndrome toxin can promote activation of T lymphocytes and enhance the secretion of inflammatory factors, resulting in a severe immune response of the animal body with subsequent multiple organ failure (Hoppe et al., 2018). Ets reduces the adhesion of mammalian stratum corneum cells by acting on desmosomal core glycoproteins between surface granular cells, resulting in the loosening of keratinocytes, which leads to the staphylococcal scalded-skin syndrome. Bacteria enter the bloodstream, leading to septicemia and endocarditis as the immunity of the infected host decreases (Rodríguez-Lázaro et al., 2015). Recently, foodborne S. aureus infections have been frequently reported in China (Wu et al., 2019). In 2015, microbial food poisoning accounted for 53.7% of all food poisoning emergencies in China, and S. aureus was the main pathogen (Wu et al., 2018). S. aureus not only has a great impact on the dairy industry but also poses a threat to public health and food safety.

The invasiveness of S. aureus depends on its resistance to host phagocytes and its cell adhesion. The capsular polysaccharide (CP) of S. aureus is an important factor related to its anti-phagocytosis and cell adhesion properties, and it can be classified into 11 serotypes based on its antigenic structure (Sompolinsky et al., 1985). CP1 and CP2 are rarely encountered in clinical practice, and approximately 70% to 80% of S. aureus can produce CPs of serotypes 5 and 8, referred to as CP5 and CP8, respectively (Goerke et al., 2005). CP5 and CP8 are expressed by the cp5 and cp8 genes, respectively. Cp5 and cp8 are nucleic acid sequences of approximately 17.5 kb present on bacterial chromosomes. The structures of CP5 and CP8 are 14-3-O-Ac- β-D-Man-NAcA-(1-4)-lFucNAc-(1-3)- β-D-FucNAc-(1-)n and(3)-4-O-Ac- β-D-ManNAcA-(1-3)-L-FucNAc-(1-3)- β-D-FucNAc-(1-), respectively. The structures of CP5 and CP8 mainly differ in terms of the position of sugar linkage and acetylation of amino mannituronic acid residues. It has been shown that CP5 and CP8 are the main factors associated with S. aureus-induced mastitis (Poutrel et al., 1988), and serotyping of S. aureus CPs is of significance to the local dairy industry to better understand the epidemiology of S. aureus.

At present, the main method for preventing and treating dairy cow mastitis caused by S. aureus is udder infusion with antibiotics. However, with the extensive use of antibiotics, S. aureus has acquired resistance to a variety of antibiotics by changing its phenotype, regulating metabolic pathways, and horizontally transferring resistant plasmids. Methicillin-resistant S. aureus (MRSA) has spread across the globe since its discovery. Despite its resistance to new semisynthetic β-lactams, MRSA is resistant to a variety of antibiotics, such as macrolides, aminoglycosides, and tetracyclines (Hassoun, Linden & Friedman, 2017). In the 1990s, MRSA infection, acquired immune deficiency syndrome, and hepatitis B were considered the three most difficult diseases to treat worldwide (Nada et al., 1996). In the animal husbandry industry, MRSA was first isolated from dairy cows with mastitis in 1972 (Devriese, Van Damme & Fameree, 1972), and cows have become its natural hosts (Spohr et al., 2011). The penicillin-binding protein (PBP) 2a encoded by mecA is the main cause for MRSA, and related studies have indicated that the detection of mecA is the main criterion for identifying MRSA (Kobayashi et al., 1994). Oxacillin-sensitive mecA-positive S. aureus (OS-MRSA) has been detected in human clinical and animal products (Conceição et al., 2015). Although OS-MRSA carries mecA, it is often misidentified as methicillin-sensitive S. aureus (MSSA) (Quijada et al., 2019) and is likely to become a subgroup with high resistance under the pressure of β-lactam selection (Martineau et al., 2000). Vancomycin is considered the last line of defense against Gram-positive bacteria and is used to treat infections that do not respond to other antibiotics. However, vancomycin-resistant S. aureus (VRSA) has become increasingly common in China and abroad (Haseeb et al., 2019). The resistance of S. aureus to vancomycin is usually mediated by the vanA operon on transposable Tn1546, which results from the transfer of resistant plasmids from vancomycin-resistant enterococci (Arthur et al., 1993; Lazaris et al., 2017). Under the combined action of vanA and vanH expression products, d-Lac replaces d-Ala in the cell wall peptidoglycan precursor peptide to form d-Ala-d-Lac, which alters the vancomycin binding site and leads to drug failure (Bugg et al., 1991). As the incidence of multidrug resistance has increased all over the world and is considered a significant public health threat, the proper use of antibiotics has become even more necessary. Moreover, the emergence of multidrug-resistant (MDR) and extensively drug-resistant(XDR) strains greatly increases the risk of the resistance spreading to humans and having an impact on the effectiveness of current antibiotic treatments; in fact, this is often a major factor in the failure of treatment and infection control (Algammal et al., 2020; Helmy et al., 2018). Additionally, studies have shown that MDR MRSA isolated from milk carries a variety of major virulence factors, such as sea and coa, that can lead to food poisoning in consumers who ingest the contaminated milk (Algammal et al., 2020). The misuse of antibiotics is an important factor leading to the emergence of MDR and XDR pathogens that threaten human public health as well as livestock-derived and aquatic products (Algammal et al., 2022). Therefore, effective and rigorous monitoring methods need to be established to control their growth in humans and animals (Algammal et al., 2023) and for routine application of antimicrobial susceptibility testing to detect the antibiotic of choice and screen emerging resistant strains. In this context, monitoring the antimicrobial resistance of S. aureus and prevalence of MRSA and VRSA in raw milk is of great significance for the research and development of antibiotics and for identifying the most appropriate drug treatment for animal diseases from the perspective of food safety.

Multilocus sequence typing (MLST) is based on the technology of multilocus enzyme electrophoresis. It involves seven pairs of relatively conservative housekeeping genes in the bacterial genome that are sequenced directly and uploaded to an international public database for analysis, which greatly improves the accuracy of multilocus enzyme electrophoresis (Maiden et al., 1998). In addition to the advantages of general genotyping methods, MLST has high resolution and repeatability, making it more convenient and accurate than general genotyping methods (Tivendale, Allen & Browning, 2009). More importantly, MLST overcomes the space restrictions of laboratories, regions, and countries, helps share typing data through large international databases, and enables analysis of the phylogeny, population structure, biological evolution, epidemiology, and molecular genetic evolution of pathogenic microorganisms (Van Wamel et al., 2010). spa typing is used to sequence the X variable region gene of S. aureus protein A; it also reveals changes in base pair numbers, sequence of X region repeat succession, and actions, such as base insertion and deletion. The X region is polymorphic; therefore, the isolates are typed according to the polymorphism of this region (Moosavian, Baratian Dehkordi & Hashemzadeh, 2020). This typing method based on single-base sequence analysis has a standardized international nomenclature and can be integrated into the international spa database, which is convenient for exploring molecular correlations between S. aureus from different regions, environments, and sources (Rodriguez et al., 2015). spa typing can usually be mapped to MLST in the form of spa-MLST-mapping (Strommenger et al., 2008), which has become the main method for typing S. aureus. It makes the research results highly reproducible and the results of regional and international S. aureus pedigree monitoring and source tracing interchangeable (Gurjar et al., 2012).

Consumption of contaminated milk is an important way by which S. aureus can infect humans, and several cases of bacterial food poisoning related to S. aureus have been reported (Fetsch et al., 2014). The fact that raw milk is particularly vulnerable to S. aureus contamination, coupled with the increasingly serious antimicrobial resistance of S. aureus, makes it a potential threat to the development of the dairy industry and public health safety. Milk forms a significant component of the Chinese balanced diet standards, and the consumption of milk is increasing annually (Qian et al., 2011). As the dairy industry in Hunan Province has advanced, large dairy farms have been built in many cities and counties. However, few studies have examined S. aureus isolates from raw milk in Hunan. The purpose of this study was to determine the antibiotic resistance, serotype distribution, and molecular epidemic characteristics of S. aureus isolated from raw milk of dairy cows in Hunan Province and provide a theoretical basis for rational use of antibiotics in dairy cows in Hunan. These findings could help promote animal welfare and food safety management through rational drug use in the dairy industry.

Materials & Methods

Results

Determining the main prevalent molecular types of S. aureus and their antibiotic resistance is necessary to formulate effective antibiotic treatment strategies and reduce the impact of S. aureus infections on the dairy industry and public health. We therefore examined the CPs, molecular types, and antibiotic resistance characteristics of S. aureus isolated from raw milk of dairy farms in Hunan Province.

CP types of S. aureus

All isolates were classified as having CP5 or CP8. The detection rates of CP5 for samples from Changsha, Changde, Shaoyang, Yongzhou, and Chenzhou were 75% (9/12), 35.29% (12/34), 40% (10/25), 0.00% (0/2), and 33.33% (1/3), respectively, while those of CP8 were 25% (3/12), 64.70% (22/34), 60% (15/25), 100% (2/2), and 66.66% (2/3), respectively (Fig. 1).

Antibiotic resistance genes of the isolates

All 76 strains carried two or more antibiotic resistance genes, with a maximum of eight genes. All isolates carried femB with a detection rate of 100%, followed by ermB (80.26%), aacA-aphD (76.32%), femA (75.00%), ermC (68.42%), mecA (47.37%), flor (42.11%), and tetM (38.16%); however, no isolates carried the glrA, ermA, or vanA genes. In this study, only the detection of mecA, ermC, and aacA-aphD in the corresponding antimicrobial resistance phenotypes was statistically significant (Table 5). The findings related to 11 resistance genes for seven types of antibiotics are shown in Fig. 2.

Table 5:

Comparison of genotypic and phenotypic characteristics for antimicrobial resistance in Staphylococcus aureus isolates.

The phenotypes of β-lactam resistance were based on penicillin, oxacillin, cefoxitin, ceftiofur, and augmentin, while those of macrolide-lincomide resistance were based on erythromycin, tilmicosin, and clindamycin; tetracycline resistance phenotype was based on doxycycline; aminoglycoside resistance phenotype was based on gentamicin; and chloramphenicol resistance phenotype was based on florfenicol. The association between resistance phenotype and resistance genes was significant (p < 0.05).

Antibiotics	Genes	Characteristics of S. aureus isolates		Association p
		G+	P+
		No(%)	No(%)
β-lactams	mecA	36(47.37%)	63(82.89%)	0.0002
	femA	57(75.00%)		0.861
Macrolides-Lincoamines	ermB	61(80.26%)	25(32.89%)	0.229
	ermC	52(68.42%)		0.011
Tetracyclines	tetM	29(38.16%)	0(0.00%)	1.000
Aminoglycosides	aacA-aphD	58(76.32%)	15(19.73%)	0.017
Phenicols	Flor	32(42.11%)	6(7.89%)	0.685

DOI: 10.7717/peerj.15847/table-5

Notes:

G +resistance gene positive

P +phenotypic resistance

MLST

MLST was successfully performed for all (n = 76) isolates, and the results are summarized in Table 6 and Figs. 3 and 4. Twenty-two STs were detected, of which ST7 accounted for 23.68% (18/76), ST97 for 14.47%(11/76), ST188 for 9.21% (7/76), and ST398 for 6.57% (5/76) of the isolates. The remaining STs were diverse; six novel alleles and eight new STs were detected in this study, with ST7737, ST7738, ST7739, ST7740, ST7756, and ST7757 carrying new alleles (Fig. 5). Although no new alleles were found for ST7735 and ST7736, new arrangements of alleles existed and were also considered to be new STs. All novel alleles and STs were uploaded to the PubMLST database (https://pubMLST.org/). The Simpson’s diversity index of the 76 isolates based on MLST was 0.907, which conformed to a 95% confidence interval (0.869–0.946). Clustering these STs using the eBURST algorithm revealed that ST97, ST7735, and ST7756 belonged to Clonal Complex (CC)97; ST1, ST81, ST188, and ST7737 belonged to CC1; ST398 belonged to CC398; ST5 and ST6 belonged to CC5; and ST2454 belonged to CC8.

Table 6:

MLST distribution of Staphylococcus aureus isolates.

STsa	clonal complex	No (%) of isolates	Allelic profilesb							Simpson’s diversity index (95% CIc)
			arcC	aroE	glpF	gmk	pta	tpi	yqiL
ST7		18(23.68%)	5	4	1	4	4	6	3	0.907(0.869-0.946)
ST97	97	11(14.47%)	3	1	1	1	1	5	3
ST188	1	7(9.21%)	3	1	1	8	1	1	1
ST398	398	5(6.57%)	3	35	19	2	20	26	39
ST705		4(5.26%)	6	72	50	43	49	67	59
ST7735a	97	3(3.94%)	3	1	605	1	1	5	3
ST7738a		3(3.94%)	5	1084b	1	4	4	6	3
ST7756a	97	3(3.94%)	3	1	1	1	1	843b	3
ST7757a		3(3.94%)	6	125	213	27	13	13	1015b
ST1	1	3(3.94%)	1	1	1	1	1	1	1
ST81	1	2(2.63%)	1	1	1	9	1	1	1
ST479		2(2.63%)	52	79	54	18	56	32	65
ST1921		2(2.63%)	4	9	1	8	1	10	218
ST7740a		2(2.63%)	5	4	1	599b	4	6	3
ST5	5	1(1.31%)	1	4	1	4	12	1	10
ST6	5	1(1.31%)	12	4	1	4	12	1	3
ST629		1(1.31%)	12	89	1	1	4	5	90
ST1458		1(1.31%)	5	4	1	4	4	6	39
ST2454	8	1(1.31%)	3	3	1	1	264	1	10
ST7736a		1(1.31%)	3	35	19	1	1	26	39
ST7737a	1	1(1.31%)	899b	1	1	9	1	1	1
ST7739a		1(1.31%)	18	95	941b	2	7	15	5

DOI: 10.7717/peerj.15847/table-6

Notes:

aNovel ST types.

bNovel alleles.

CI

confidence interval

MLST

multi-locus sequence typing

ST

sequence typing

Discussion

The isolation rate of S. aureus from raw milk in Hunan Province was 11.16%, which is lower than those observed from milk sources in northern Xinjiang, China (43.1%), by Kou et al. (2021) and bulk canned milk samples from Shandong Province, China (28.9%), by Zhao et al. (2021). However, our results are similar to those of Bissong et al. (2020), who reported a prevalence of 11% for S. aureus in beef and milk in Cameroon. S. aureus, an important multi-host pathogen, can be isolated from dairy cows and contaminated milk (Gurjar et al., 2012; Fetsch et al., 2014). It can spread between people and livestock during grazing, activities involving human–animal interactions, and movement of herds between farms (Zadoks et al., 2011). Although the isolation rate of S. aureus from raw milk in Hunan Province is not high, these bacteria were successfully isolated from the samples collected, suggesting that farmers should strictly control the cleaning and disinfection of the feeding environment of dairy cows. In addition, individuals involved in the dairy industry should implement standardized procedures for raw milk collection, processing, and packaging and take precautionary measures to reduce environmental pollution and avoid storing dairy products in unfavorable conditions to safeguard the health of consumers.

Serotyping of S. aureus CPs is usually performed to predict and determine the epidemiologic trends of strains isolated in an area and the incidence of mastitis in dairy cows before molecular typing (Poutrel et al., 1988). In this study, all S. aureus isolates exhibited either the CP5 or CP8 type; the overall detection rate of CP8 was higher than that of CP5, indicating that CP8 was the main epidemic serotype of S. aureus isolated in Hunan Province. Poutrel et al. (1988) found that CP5 was the main epidemic serotype in France. In a study by Han, Pak & Guidry (2000) in South Korea, the detection rate of CP5 was similar to that of CP8, and 26% of the strains were of the CP5 type and 24.2% of the CP8 type. The serotype distribution of CPs differs between countries and regions, indicating that the distribution of CPs of S. aureus is likely related to geographical location. In addition to serotyping S. aureus, CPs can help identify the antibody induced by its protein-coupled vaccine, which has immunogenicity and has been proven to protect mice from fatal S. aureus infection (Han, Pak & Guidry, 2000). A targeted vaccine developed according to the prevalence of CP serotypes in particular areas can yield an improved immune effect.

Many antibiotics are used in the animal breeding industry as therapeutics or growth enhancers every year in China. However, with the evolution of bacteria and abuse of antibiotics, the antibiotic resistance rate of S. aureus has gradually increased. It is generally believed that the antibiotic resistance mechanism of S. aureus mainly includes the following: (1) drug hydrolase and modification (such as the aacA-aphD gene code modifies specific groups of aminoglycosides to make them lose their affinity for bacterial target ribosomes) (Wang et al., 2015); (2) the active efflux system pumps drugs out of bacterial cells (for example, the transporter encoded by msr gene can expel intracellular lincoamines to reduce intracellular drug concentration); and (3) changes in the target sites of antibacterial drugs (such as erm gene encoding methylase, which methylates the target of macrolide drug action and reduces its affinity with bacterial ribosomes) (Di Modugno et al., 2002). In this study, S. aureus isolated from raw milk from Hunan Province was highly resistant to β-lactam antibiotics, with 82.89% of the isolates were resistant to penicillin. This finding is similar to that of studies from Shandong (74.4%) (Zhao et al., 2021) and the Ningxia Hui Autonomous Region (94.3%) (Wang et al., 2016), suggesting that S. aureus isolated from Chinese dairy cows is generally resistant to penicillin. The mechanism of penicillin resistance in S. aureus is mainly through the synthesis of β-lactam hydrolase, which hydrolyzes the β-lactam ring, or mecA-encoded PBP2a, which does not easily bind with β-lactams, to replace the effect of transpeptidase PBPs, thus inhibiting the ability of β-lactams to interfere with bacterial cell wall synthesis and reducing the antimicrobial effect of β-lactam drugs (Hassanzadeh et al., 2020). Of the 76 isolates, 36 were found to carry mecA and were identified as MRSA, of which 27.77% were OR-MRSA and 72.22% were OS-MRSA. As Quijada et al. (2019) showed in their study, deletions or mutations in the bla system (blaI-blaR1-blaZ) of some isolates may downregulate the expression of mecA, thereby affecting the resistance of the MRSA strain to oxacillin and resulting in OS-MRSA. In addition, it was also found that 22.50% (9/40) of MSSA isolates were resistant to oxacillin, and mecA- negative but oxacillin resistance-positive phenotypes have been reported in India (Kumar, Yadav & Singh, 2010) and Turkey (Turutoglu et al., 2009). This is possibly due to the role of PBP2c, which is expressed by mecC in some strains; overexpression of β-lactamase in borderline oxacillin-resistant S. aureus; or mutations at some amino acid sites in the PBP1, PBP2, and PBP3 transpeptidase domains, which enhance the resistance of bacteria to β-lactam antibiotics, resulting in oxacillin resistance despite mec gene negativity (Hryniewicz & Garbacz, 2017). S. aureus isolates were also resistant to macrolides, lincosamides, aminoglycosides, florfenicol, and quinolones, which may be related to the use of these drugs as the main antibiotics for treating bovine mastitis and other diseases at some dairy farms. Nevertheless, there is usually no significant correlation between antibiotic resistance genes and phenotypes (Bissong & Ateba, 2020) because bacterial antibiotic resistance phenotypes may be determined by a variety of factors, such as environmental changes, strain activity, and frequent use of antimicrobial agents, whereas genotypes are decided only by internal gene factors. In this study, a strain of MRSA was also found to be resistant to vancomycin. The strain was identified as VRSA, but vanA was not detected. Bhattacharyya et al. (2016) found that the emergence of VRSA was not completely related to vanA positivity, and Roch et al. (2014) proved that S. aureus could trigger the vancomycin-resistant phenotype through exposure to non-glycopeptide antibiotics, such as β-lactams. The development of vancomycin resistance in S. aureus could also be related to epigenetic processes and not only to gene expression (Haaber et al., 2015), which could explain why the only strain of VRSA in this study was also an MDR OR-MRSA.

A large international database containing the sequence data of thousands of S. aureus strains has been established. It is an important data source for large-scale epidemiological and genetic correlation research on S. aureus (Van Wamel et al., 2010). In this study, we found abundant and diverse STs of S. aureus isolated from raw milk in Hunan Province, with ST7 being the most common. Some studies have shown that ST398 is the main ST related to livestock infections, and ST5, 45, and 239 are related to hospital-acquired infections. ST1, 8, 80, and 59 are mainly associated with community-acquired infection (Skov & Jensen, 2009), while ST97 is commonly associated with S. aureus- induced mastitis worldwide (Smith et al., 2005). In this study, CC398-ST398, CC5-ST5, CC1-ST1, and CC97-ST97 were all detected, which further confirmed the threat posed by raw milk infected with S. aureus to the dairy industry and public health. It is inferred that pathogenic bacteria may spread to the environment and infect humans during milking and transportation and via inadequately sterilized dairy products. Additionally, a previous study reported that alleles having only a single nucleotide mutation should be labeled as a novel ST (Enright et al., 2000). In this study, CC1-ST7737, ST7738, ST7739, ST7740, CC97-ST7756, and ST7757 showed altered nucleotide sequences, but only CC97-ST7735 and ST7736 showed a new allele arrangement. The emergence of these new STs is not accidental, and related mutants might have appeared a long time ago and spread among dairy cows in Hunan Province. However, these new STs were only discovered recently, possibly due to the lack of research on the molecular typing of S. aureus in raw milk from Hunan. The detection rate of MRSA in the 17 strains with newly discovered STs was 52.94%, of which ST7738 and 7740 were MRSA, indicating that these new STs may become an important molecular type of MRSA in Hunan Province. Furthermore, the only VRSA strain exhibited ST7739, indicating that the monitoring of new STs from isolates obtained during epidemics using MLST should be strengthened.

Twenty S. aureus spa types were identified in this study. Diversity analysis showed that the diversity of spa typing was high; t796 was the most common, followed by t189, t359, and t091. Mekonnen et al. (2018) identified 20 spa types from 79 strains of S. aureus isolated from milk samples from dairy farms in northwest Ethiopia, and t042 was the main spa type. Strommenger et al. (2006) studied S. aureus isolated from Dutch cows with mastitis and found that t529 was the dominant spa type. Upon comparing the above results, we speculate that isolates from different countries and regions usually have different molecular types. Many studies have used MLST and spa typing to determine the population structure of highly cloned S. aureus (Bhargava et al., 2011; Basanisi et al., 2017). Malachowa et al. (2005) used pulse field gel electrophoresis, MLST, multiple locus variable-number tandem repeat analysis, and spa typing to compare 59 strains of iatrogenic S. aureus and found that spa typing showed similar results to those of MLST. They suggested that multiple typing methods should be integrated to improve the accuracy of S. aureus typing. In this study, there was a significant correlation between spa typing and MLST. Although spa types and STs with multiple corresponding genotypes were observed, most STs and spa types corresponded to 1 genotype. The existence of different STs associated with the same spa type indicates that 2 STs may have originated from the same clone and evolved in parallel (Strommenger et al., 2008). The spa-MLST-mapping typing of the 76 isolates clearly reflected the molecular characteristics of S. aureus isolates in Hunan Province and that t796-ST7 was the main molecular type of S. aureus. Notably, five isolates exhibited t034-CC398-ST398, which has been reported to be the main molecular type of S. aureus infection in livestock and poultry (Bouiller et al., 2016), and MRSA isolates exhibiting t034-CC398-ST398 have been discovered in domestic pigs (Graveland et al., 2011). It is speculated that t034-CC398-ST398 isolates may spread among different kinds of livestock in different regions, and they may even spread between livestock and humans and cause diseases (Van Cleef et al., 2011); therefore, farmers should be vigilant.

Through the isolation and identification of milk-derived S. aureus in northern Italy, Gazzola et al. (2020) found that CC398, CC1, and CC97 were three common MRSA clones, and Li et al. (2015) found that the main molecular type of MRSA isolated from milk in Shaanxi Province was t524-ST71. These studies suggest that the distribution of MRSA may be related to its molecular typing. In this study, different molecular isolates showed different antibiotic resistance characteristics, and the most common molecular type exhibited by MRSA was t796-ST7. The isolates of the same molecular type may have the same or similar antibiotic resistance spectrum. The isolates obtained from different regions or sources may have different molecular types, and the isolates of different molecular types may have different antibiotic resistance, which is speculated to result from the competition between different molecular types under the screening pressure of antibiotics in different regions, and the prevalence of dominant molecular types can promote further spread of antibiotic-resistant bacteria. In addition, Enright et al. (2000) pointed out that it is usually impossible for unrelated MRSA and MSSA to have the same molecular typing and that this status may be caused by the horizontal transfer of mecA from MRSA to MSSA or the loss or inactivation of the mec gene of some MRSA strains, resulting in some MRSA strains acquiring the same allele map as their MSSA ancestors and thus being classified as having the same STs.

Conclusions

The results showed that S. aureus from raw milk in Hunan area had a typical CP type, rich molecular typing, and different levels of antibiotic resistance to a variety of antibiotics. Our findings could help promote the production of S. aureus vaccine against CP8 in Hunan area and resistance monitoring of t796-ST7 isolates with the prevalent molecular type of MDR strains. The present results suggest reducing the use of β-lactam, macrolide, lincoamine antibacterial drugs and appropriately administering doxycycline, sulfonamides, and glycopeptides as veterinary antibiotics to treat infectious diseases in dairy cows. The antibiotic drug dosage should be strictly controlled, and drug rotation and the withdrawal period of the antibiotics should be strictly implemented. Further, active research, development, and promotion of Chinese herbal medicines, micro-ecological preparations, and other new non-antibiotic drugs are necessary.

Supplemental Information