Bovine mastitis, an inflammation of the mammary gland, causes physical, chemical, and usually, bacteriological changes in milk and pathological changes in the glandular tissues of the udder that affect the quality and quantity of milk (Sharma and Jeong, 2013). It has a high incidence and prevalence in dairy cows, affecting the net earnings of milk producers worldwide (Frola et al., 2011). Generally, clinical mastitis is easily diagnosed by visible clinical manifestations, such as red, hot, and swollen mammary glands (Sharma et al., 2007). Subclinical mastitis has no visible clinical symptoms in mammary glands and in milk, but milk production decreases, somatic cell count (SCC) increases, pathogens are present in the secretion, and the milk composition is altered (Lafarge et al., 2004). Therefore, development of effective, reliable, and safe mastitis prevention and treatment strategies has become a focus of research.

In China, a teat disinfectant is usually applied to prevent mastitis in pasture. Teat disinfection pre- and post-milking is important not only to reduce the possibility of mastitis, but also to reduce the risk of bacterial contamination of milk (Suriyasathaporn and Chupia, 2011; Zucali et al., 2011). Currently, various teat cleaning disinfectants, including iodophor solution, iodine based gel, sodium hypochlorite, dodecyl benzene sulfonic acid, chlorine, chlorhexidine, phenolic compounds, alcohol, and guava leaf extract, have been used for pre-milking teat dipping (Ingawa et al., 1992; Foret et al., 2005; Gibson et al., 2008). These chemical disinfectants can reduce major pathogen infections; however, the high concentration of chemical substances has raised the concern of potential residues in milk (Galton et al., 1986). As a result, a natural substance that exhibits inhibitory activity against bacteria was developed as a teat disinfectant to minimize bacterial contamination in raw milk. A previous study showed that methanol extracts from guava leaves used as a pre-milking teat disinfectant could significantly reduce teat-end bacterial loads compared with routine udder sanitization without teat dipping (Kummee et al., 2015).

The application of probiotic lactic acid bacteria (LAB) is now considered the best choice for the treatment of many infectious human diseases and for the control of bovine mastitis (Tagg and Dierksen, 2003; Soleimani et al., 2010). It is well known that these bacteria are safe, non-pathogenic, and exhibit many properties that prevent spoilage and the growth of pathogenic bacteria (Mojgani and Ashtiani, 2006; Yezli et al., 2015). The Lactobacillus (L.) plantarum strains IMAU 80065 and IMAU 10155 were isolated from pickles and fermented camel milk from Sichuan and Inner Mongolia, China, respectively (Yu et al., 2012). They were screened from 347 isolates and produce bacteriocins that significantly inhibit the growth of Escherichia (E.) coli, Staphylococcus (S.) aureus, Salmonella, Shigella, and Listeria (Yu et al., 2015). The objective of this study was to compare the effectiveness of a novel teat disinfectant containing two probiotic L. plantarum strains with a commercial disinfectant (CD). We analyzed the SCC, the number of pathogenic bacteria, and the bacterial community of the raw milk. Furthermore, we used Pacific Biosciences (PacBio) single molecule, real-time sequencing technology (SMRT) based on the full-length 16S ribosomal RNA (rRNA) gene to accurately and systematically depict the bacterial profiles of cow milk.

Microorganisms in raw milk directly impact the flavor, quality, and shelf life of milk products. Moreover, the presence of pathogens in raw milk can lead to severe illness (Oliver et al., 2009; Quigley et al., 2013). The application of teat disinfection pre- and post-milking is recommended to reduce the number of bacteria on the teat skin and in milk (Oliver et al., 1993). The aims of this study were to harness the PacBio SMRT sequencing technology to investigate the influence that LAB teat disinfectant and CD have on teat milk microbiota from individual cows. Our findings confirm the efficacy of LAB disinfectant in inhibiting pathogenic bacterial growth and improving the bacterial populations of raw milk during the cleaning process.

Currently, the SCC and the California mastitis test (CMT) are used to diagnose mastitis. Somatic cells are naturally present in milk, which is widely used to distinguish healthy quarters from quarters with an inflammatory response most likely due to an intramammary infection (Schukken et al., 2003; Oikonomou et al., 2014). Usually, an SCC of <100,000 cells/mL is considered healthy, and a threshold of 200,000 cells/mL was shown to have high sensitivity and specificity as an identifier of subclinical mastitis (Hillerton, 1999; Diepers et al., 2016). In this study, cows were selected based on this criterion. During the course of teat disinfectant treatment, the SCC gradually decreased and the SCC of the LAB group was slightly lower than that of the CD group. Although the SCC did not decrease to a healthy level, changes were seen in the microfloral structure of teat milk. Previous research also showed that the microfloral structure of bulk tank milk is associated with the SCC (Rodrigues et al., 2017).

Generally, S. aureus, S. uberis, and S. agalactiae are recognized as major virulent mastitis-causing bacteria (Kuehn et al., 2013). Therefore, researchers have focused on developing a safe and effective alternative to the use of chemicals to inhibit these pathogenic bacteria. Previous studies demonstrated the inhibitory effects of secreted bacteriocins, like nisin and lacticin 3147 (Cao et al., 2007; Klostermann et al., 2010), and live probiotic microorganisms, in in vitro and in vivo approaches (Espeche et al., 2009; Diepers et al., 2016). Interestingly, Oikonomou et al. (2014) reported that these bacterial species were present in small quantities in healthy cow milk and hypothesized that they are part of the normal bacterial flora of the mammary gland. In our study, these bacterial species were also detected in lower quantities in the majority of milk samples by SMRT sequencing and qPCR. However, the mean amounts of S. aureus and S. agalactiae gradually decreased during the cleaning protocol, particularly in the LAB group, after 10 days (Figure 2). This result confirms that our LAB disinfectant possesses antibacterial activity primarily against mastitis-causing bacteria in vivo. Moreover, the antibacterial effect of LAB was marginally better than that of the chemical disinfectant.

Probiotic products against udder diseases have been developed over the past few decades and comprise bacteriocins for external application, intramammarily applied bacteriocins, and live probiotic microorganisms. Encouraging results were obtained with the injection of live cultures of selected LAB strains into the bovine udder, which were able to inhibit several mastitis pathogens (Crispie et al., 2008; Frola et al., 2011). Results have shown that preparations of the live culture were as effective as commonly used antibiotics for the treatment of intramammary infections and did not show adverse effects on mammary tissue. Moreover, the bovine teat surface can contain a high diversity of bacteria (Verdiermetz et al., 2012). Monsallier et al. (2012) proposed that the teat skin was a source of microbial populations in raw milk and that farm management and animal grazing practices influenced the diversity and microbiota of raw milk. Another study also highlights that the teat surface and herd habitats are significant drivers of milk microbiota composition (Doyle et al., 2017). Hence, it is beneficial to use a probiotic LAB teat disinfectant as a protective barrier to inhibit pathogens and improve the microbial balance of the teat. Our data provide new insights into probiotic LAB regulation of the bacterial composition of the teat and suggest a role for LAB in the prevention of mastitis. The results revealed differences in the bacterial composition of milk during the cleaning process with two disinfectants. These differences were reflected in the beta-diversity measurements, as illustrated in the PCoA analysis of the UniFrac distance (Figure 6). The UniFrac PCoA showed that the non-cleaned (0 day) and cleaned, LAB cleaned, and chemical cleaned samples mostly fell within separate clusters. That means the LAB teat disinfectant can alter the microbial structure of teat milk.

Analysis of the bacterial composition of the milk samples after 1, 6, 10, and 12 days of teat dipping in the same individuals revealed bacterial species diversity (Figure 4 and Supplementary Figure S1). The relative abundance of the L. plantarum group in the LAB group following cleaning was slightly higher than in the CD group. This demonstrates that the LAB disinfectant could be transferred through the milk ducts into the teat and mammary glands to further regulate the bacterial community of the teat. In addition to its presence in raw milk, O. profundus has been found in deep-sea sediment core samples and wheat rhizospheric soil (Kim et al., 2007; Aislabie et al., 2008; Ranieri et al., 2012). The relatively high abundance of O. profundus in our samples may have been caused by potential contamination from the feeding environment and forage grass. It is worth noting that the relative abundance of O. profundus in the LAB group was significantly higher than in the CD group, suggesting that further research is needed to determine the role O. profundus may play in the microbiota of cow teats.

The relative abundances of Acinetobacter schindleri, Acidovorax radicis, Limnobacter thiooxidans, N. varians, M. timonae, Paenibacillus validus, and Pseudomonas alcaligenes gradually decreased following cleaning (Figure 5). With the exception of Pseudomonas alcaligenes, these species are not commonly found in milk. Acinetobacter schindleri was first found in the urine of a male outpatient with cystitis (Nemec et al., 2001) and the Acinetobacter genus is widely distributed in nature and frequently found in milk, soil, and water (Dortet et al., 2006; Li et al., 2016). Acidovorax radicis is a wheat root-colonizing bacterium that was previously identified from surface-sterilized wheat roots (Li et al., 2011). Limnobacter thiooxidans is a thiosulfate-oxidizing bacterium that was isolated from freshwater lake sediment (Spring et al., 2001). Massilia timonae was isolated from patient blood and soil (Lindquist et al., 2003) and N. varians was identified in clinical specimens and water (Kämpfer et al., 2008). Pseudomonas alcaligenes is a human pathogen, but occurrences are rare. Previous studies have shown that Pseudomonas spp. are responsible for the spoilage of milk and dairy products because of their ability to produce heat-resistant proteolytic and lipolytic enzymes at low temperatures (Ercolini et al., 2009; Zhang et al., 2015). The presence of these species may originate from components of the environment in contact with the teat skin. Our results suggest that teat dipping with LAB and CDs have identical efficacy in the reduction of these contaminating species. Paenibacillus validus was decreased at 10 days with both disinfectants, but increased sharply after cleaning with water for 2 days. This species is abundant in soil but rarely found in milk (Hildebrandt et al., 2006), while other Paenibacillus species have been detected in cow’s milk and various milk products. Paenibacillus spp. are of interest as they have been associated with the biocontrol of pathogenic bacteria and a decreased shelf life of milk products (Ziarno and Zareba, 2010; Debois et al., 2013). In particular, Ranieri et al. (2012) suggested that Paenibacillus spp. could represent an important tool for determining the quality of raw milk, as well as for identifying potential contamination sites at the farm and processing facility. Obviously, both LAB and CDs have the ability to decrease the relative abundance of Paenibacillus validus, although LAB disinfectants are more durable.

Deinococcus grandis and Psychrobacter faecalis increased during the cleaning protocol. Deinococcus grandis is a radioresistant bacterium that was initially isolated from freshwater fish and animal feces (Oyaizu et al., 1987). Some species in the Deinococcus genus are found in milk products. Members of the Deinococcus genus are also used as ingredients in food to ferment soymilk and as a feed additive for hens to enhance yolk coloration to meet customer demand (Wu et al., 2016). Psychrobacter faecalis was first isolated from pigeon feces and has not been detected in milk. While Psychrobacter spp. have been found at a relatively high level in the raw milk of healthy cows compared with cows with mastitis (Kuehn et al., 2013), these bacteria were not previously associated with mastitis (Vacheyrou et al., 2011). Given the change in the amounts of these species during the cleaning protocol, we favor the explanation that in small amounts they are part of the normal bacterial flora of the teat and may encourage teat microflora balance. Future research should aim to investigate the origin of these species and their role in teat microflora.

Three species of Bacillus were dominant in our samples. Previous studies have reported the presence of Bacillus spp. in raw milk (Hou et al., 2015) because of their heat resistance properties and intrinsic antagonism toward other microbes (Griffiths and Phillips, 1990). The season (summer, approximately 30°C) and sanitary conditions of the location of the sample collection may have resulted in the higher relative abundance of Bacillus spp. in the teat, which did not exhibit significant changes during the cleaning process. This suggests that these two different cleaning disinfectants were not able to inhibit Bacillus spp.; hence, it will be important to screen additional LAB strains and develop novel disinfectants aimed specifically at Bacillus spp.

In this study, we used the latest PacBio SMRT sequencing technology to describe the changes in the bacterial community during the cleaning protocol based on the full-length 16S rRNA gene. Compared with previous sequencing approaches, the PacBio SMRT technology has a high capacity for the production of long reads and has been shown to offer higher taxonomic resolution in the profiling of bacterial communities in dairy product samples to the species level (Hou et al., 2015; Li et al., 2016; Zheng et al., 2016). It was therefore not surprising to find more uncommon species and numerous unidentified sequences or sequences identified as uncultured bacteria in our study. Previous data on the bacterial profile of dairy products were generated largely by second-generation sequencing technology, which generates high quality, though short, sequence reads that preclude the accurate assignment of DNA sequences at the species level and even limits classification at the genus level (Amir et al., 2013). Furthermore, PacBio SMRT technology can greatly shorten the time required to detect and analyze the overall bacterial profile of a sample to 24 h (Zheng et al., 2016). By employing SMRT sequencing technology coupled with full-length 16S rRNA sequence determination in this study, clear bacterial population structural differences in the teat were revealed in 69 raw milk samples during the cleaning protocol using LAB or chemical disinfectant.

Rainard (2017) was skeptical about the potential of probiotics as a potential preventive solution for mastitis control. However, the present study demonstrates that teat dips containing LAB possess the ability to reduce both the SSC and pathogenic bacteria, thus regulating the bacterial composition of the teat. Moreover, the ability of LAB to increase beneficial microbes was relatively higher than chemical disinfectants. Overall, our findings indicate that LAB as a teat disinfectant is antimicrobial, non-irritating, and highly efficient and can be used for teat dipping following a routine udder sanitization procedure. Previous studies have been devoted to the development of novel disinfectants, such as plant extracts, to replace chemical disinfectants (Zeedan et al., 2014; Kummee et al., 2015).

This is the first report on the use of PacBio SMRT sequencing technology to evaluate the impacts of LAB disinfectant on the bacterial composition of the teat and comparison with a chemical disinfectant. Our results indicate that the teat bacterial community is highly diverse and complex. Both LAB and CD altered the bacterial composition of the teat and, in particular, reduced the number of mastitis-causing bacteria. Teat dipping with LAB has advantages over chemical disinfectant in that it is harmless to the milker’s health and non-toxic to milk consumers. Therefore, LAB disinfectant could be used as an alternative to chemical pre- and post-milking teat disinfectants to maintain udder health. Some species showed significant differences between cleaning and non-cleaning, suggesting these species could also be used as markers of udder health. In addition, the PacBio SMRT sequencing technology, a powerful and valuable tool, provided accurate microbiota profiling data in our study.

Sequencing data for the 16S rRNA sequences have been deposited in the Metagenomics Rapid Annotation Server with the accession number MGP80281.

JY and HZ designed the experiments. JY, YR, XX, and WH performed the experiments. JY, YR, and HZ drafted the manuscript. All authors read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.