Streptococcus dysgalactiae is an important bovine mastitis causing pathogen worldwide (Cameron et al., 2016; Gao et al., 2017; de Campos et al., 2021), which could result in economic loss and deteriorated animal welfare (Heikkilä et al., 2018). Streptococcus dysgalactiae consists of 2 subspecies: Streptococcus dysgalactiae subsp. equisimilis (SDSE) and Streptococcus dysgalactiae subsp. dysgalactiae (SDSD). They can be distinguished based on their hemolytic properties [SDSD (a-hemolytic) and SDSE (b-hemolytic)] on blood agar (Alves-Barroco et al., 2021). SDSD is classically described as an animal pathogen, causing animal diseases, such as bovine mastitis (Zhang et al., 2018), and mostly results in persistent (sub)clinical mastitis (Botrel et al., 2010). Meanwhile, studies indicate that SDSD isolates can also be isolated from human with breast cancer and infect human primary keratinocyte cells in vitro (Alves-Barroco et al., 2019b; Koh et al., 2020), indicating the potential to infect human. Therefore, SDSD in cow milk should be considered a threat to public health.

The pathogenicity of Streptococcus dysgalactiae attributes to various virulence factors (VFs), which involved in the following categories: adherence, enzyme (including hyaluronidase, mitogenic factor, streptococcal enolase, and streptodornase et al.), immune evasion, immunoreactive antigen, iron uptake, manganese uptake, protease, superantigen, and toxin formation process, based on Streptococcus virulence factors database (VFDB). These virulence genes facilitate the process of adherence to epithelial cells and internalization as well as the subsequent dissemination into host cells (O’Halloran et al., 2016; Alves-Barroco et al., 2019a). Abuse of antimicrobials in mastitis treatment contributes the most to the spread of antimicrobial resistance (AMR) in mastitis associated SDSD isolates (McDougall et al., 2014; Cameron et al., 2016; Zhang et al., 2018). Specifically, the emergence of multiple-drug resistant strains can be a major challenge in the treatment of bovine mastitis and is a growing concern for public health (McDougall et al., 2014; El Garch et al., 2020).

Population structure of SDSD remains unknown, although various genotypes of S. dysgalactiae have been described (Rato et al., 2013; Santos-Sanches et al., 2015). Genetic characteristics of SDSD associated with bovine mastitis, including virulence factors and AMR genes, have been profiled using PCR-based methods (Tian et al., 2019). However, these methods are insufficient to capture the subtle genomic differences among strains. Whole genome sequencing (WGS) with high resolution could be the preferred method to determine strain diversity and to infer phylogenetic relationships as well as to identify virulence and resistance genes in SDSD isolates with subtle differences. Therefore, the objectives of this study were to (1) estimate the apparent prevalence of SDSD in samples from (sub)clinical mastitis; (2) determine genetic diversity and evolution of SDSD isolates by 16s RNA sequencing and whole genome sequencing analysis; (3) identify VFs and AMR genes in those SDSD isolates.

Several studies have described the epidemiology of Streptococcus dysgalactiae in dairy herds (Cameron et al., 2016; Horpiencharoen et al., 2019). In addition, studies using PCR-based techniques to identify virulence and AMR genes in bovine S. dysgalactiae isolates have been published (Fernandez et al., 2013; Horpiencharoen et al., 2019). However, no in-depth studies have investigated the population structure, profiles of virulence and AMR genes as well as the phylogenetic relationships of bovine mastitis derived SDSD isolates. Therefore, we conducted phylogenetic analysis based on 16S rRNA and MLST of 51 SDSD isolates obtained from 17 dairy herds in 7 provinces of China. In addition, we also conducted WGS of 12 representative SDSD isolates to determine the distribution of virulence and antimicrobial resistance genes. The core genomes were more related to metabolism, and the unique genomes were more relevant to genetic replication and repair according to functional pan-genomes analysis. Consequently, our study suggests that core genes of metabolism can be used as a tool to identify Streptococcus species in the microbiome. The phylogenetic tree based on the core genome of 24 Streptococcus revealed that human S. dysgalactiae isolates have closer genetic relationship with fish and swine S. dysgalactiae isolates than bovine SDSD isolates, which is consistent with Alves-Barroco et al. (2021).

The apparent prevalence of SDSD in CM was 9.6%, roughly consistent with the figures in previous studies in China, America, and Europe (Cameron et al., 2016; Soltau et al., 2017; Zhang et al., 2018), but lower than the 28.0% in Australia (Shum et al., 2009). The apparent prevalence of SDSD in SCM samples was 1.2%, lower than other studies (Botrel et al., 2010; Andrea Dieser et al., 2014; Leelahapongsathon et al., 2014).

Understanding the phylogenetic relationships among strains is important to determine the transmission of pathogens. The phylogenetic tree based on 16S rRNA sequences revealed 4 phylogenetic clusters. Cluster A-D contained 15, 15, 10, and 11 isolates derived from 3, 5, 5, and 3 provinces, respectively. Some identical STs (ST521, ST453, ST523, ST454, and ST527) were identified in multiple herds in this study, whereas the other STs were only detected in single herds. That identified STs were in different herds, which is consistent with a previous study (Lundberg et al., 2014). Furthermore, movement of infected animals likely promotes spreading of pathogens between herds (Widgren and Frössling, 2010).

Using the whole genome sequences of 12 representative SDSD genomes, we observed differences in the number of predicted prophage regions in 12 SDSD genomes among different strains and STs. A total of 35 prophage regions were detected in 12 SDSD genomes. The average number of prophages per genome was 2.9-the highest number of prophages was presented in ST453 (5.3) and the lowest was 1.3 in ST521. These values suggest that the number of prophage regions in SDSD could possibly be associated with ST/CC, which is in line with Lichvariková et al. (2020).

Adherence-related genes could facilitate adhesion and biofilm formation, which are important factors in Streptoccoci pathogenesis (Widgren and Frössling, 2010). Adhesion is the first step in biofilm formation or invasion into host cells, promoting survival of microorganisms in infected tissues and facilitating development of mastitis (Dahesh et al., 2012; Alves-Barroco et al., 2019a). Eleven adhesion-related genes [abpB (Huang et al., 2015), fbpA (Bao et al., 2014), fnBA (Yeswanth et al., 2017), shr (Dahesh et al., 2012), slrA (Bober et al., 2011), cbpD (Roig-Molina et al., 2020), fnz2 (Yi et al., 2013), fneB (Lannergård et al., 2005), lmb (Zhang et al., 2014), seM (Timoney et al., 2010), srr-1 (Sheen et al., 2011)] encode a number of surface proteins. These surface proteins are identified as important virulence factors that involve bacterial adhesion to the epithelium of the host cell mediated by microbial surface components recognizing adhesive matrix molecules, consequently contributing to host cell attachment and tissue colonization. Lmb was the most abundant adherence related gene in the 12 SDSD isolates. In previous studies, the lmb gene had a relatively high prevalence in Streptococcus uberis isolated from cattle (Fessia et al., 2019), but was less prevalent in bovine-associated Streptococcus dysgalactiae and Streptococcus agalactiae (Ding et al., 2016; Tian et al., 2019). Three biofilm-related genes (srtA, aspA, and emm) were identified in 12 SDSD genomes. Biofilms are communities of microorganisms attached to a surface and are involved in chronic and recurrent infections in animals and humans (Veerachamy et al., 2014). Compared to planktonic bacteria, those with biofilms are more resistant to antibiotics (Rabin et al., 2015). Therefore, these genes may contribute to the persistent infection induced by SDSD (Kaczorek et al., 2017).

Enzyme (dltA, eno, sda, hylB, and mf/spd), immune evasion (hasC, rgpA to G, except rgpC, oppF etc.), immune reactive antigen (sip), iron and manganese uptake (shp, htsA to C), protease (htrA, clp), and peptidase (lspA) genes were widely distributed among all SDSD. These genes enable S. dysgalactiae to colonize (Roy et al., 2019; Diaz-Dinamarca et al., 2020), spread (Florindo et al., 2018) and survive in host tissues (De et al., 2017; Roy et al., 2019) and cause infection (Calonzi et al., 2020). HylB was the most abundant enzyme-related gene, and this high prevalence was consistent with previous studies (Ding et al., 2016; Calonzi et al., 2020; Lin et al., 2021). Hyaluronidase encoded by hylB promoted intracellular survival of Streptococci and expression of pro-inflammatory cytokines, indicating it may have a key role in pathogenesis of Streptococci mastitis, including SDSD (Wang et al., 2014).

A total of 23 AMR genes were identified with various copy numbers in 12 SDSD isolates according to the annotation of CARD. According to previous surveys, 5 categories of antimicrobials, including β-lactams (penicillin G, cefalexin, and ceftriaxone), macrolide (erythromycin), aminoglycoside (kanamycin and streptomycin), and tetracyclines, are frequently used in treatment of bovine mastitis on Chinese dairy farms (Shi et al., 2010; Zhang et al., 2018). The emergence of AMR genes are mainly through spontaneous mutations and acquisition of AMR genes from other bacteria (MacLean and Millan, 2019). The high prevalence and breadth diversity of AMR genes, which encode proteins conferring resistance to the 5 categories of antibiotics above, might be attributed to abuse of antimicrobials in Chinese dairy herds. In this study, only one integrative conjugative element ICETn6009 was detected in SDSD20 and SDSD37, both of which belonged to ST453 and harbored the most abundant AMR genes. ICETn6009, a Tn916 based element with Gram-positive mer operon and directly linked to the tet(M) gene, was identified from two Gram-positive and three Gram-negative genera, including Streptococcus (Soge et al., 2008). Meanwhile, 4 AMR genes were located in insertion sequences from 4 isolates in ST453 and ST521 and one in transposon from SDSD 34 in ST521. This may indicate isolates of these two STs were more capable in acquisition of antimicrobial resistance genes from MGEs (Oppegaard et al., 2020).

In general, the average identity values were relatively high for the AMR genes relevant to specific antibiotics: Aminoglycoside [ANT6-IA, APH(3′)-IIIa, APH(3′)-Ia; 100.00%], Peptide (MCR-4.1; 100.00%), Lincomycin (lnuB; 100.00%), and Nucleoside (SAT-4; 99.43%). These high identity values indicate that these genes were highly likely to be AMR genes (Velez et al., 2017). Four antimicrobial resistance mechanisms are involve in the AMR genes identified in 12 SDSD genomes, including antibiotic inactivation [aad(6), APH(3′)-IIIa, APH(3′)-Ia, ant6-IA, aad9-IB, lnuB, and SAT4], antibiotic target alteration (vanA, vanRA, vanRB, vanRG, bacA, MCR-4.1, and pbp2X), antibiotic target protection [tet(W/N/W), tetM, mefA and IsaE], and antibiotic efflux (bcrA, patA, patB, and pmrA) (Christaki et al., 2020). Aminoglycoside phosphotransferase encoded by APH(3′)-IIIa in gram-positive bacteria could phosphorylate the aminoglycoside agents, such as amikacin, neomycin, and kanamycin (Woegerbauer et al., 2014). And lnuB, encoding a 3-lincosamide-O-nucleotidyltransferase which could inactivate lincosamides by adenylation in position 3, was reported in S. agalactiae (Haenni et al., 2018). Mutation in pbp2x could reduce the affinity of the bacteria to β-lactam antibiotics, which has been considered as the major mechanism of Streptococcus drug resistance (McDougall et al., 2020; Vannice et al., 2020). TetM directly interacts and alters the conformation of nucleotide within 16S rRNA that comprises part of the tetracycline binding site, leading to tetracycline dissociation from the ribosome and prevents rebinding (Dönhöfer et al., 2012). PatA and patB work as heterodimers to interact together to make a functional drug efflux transporter, conferring efflux-mediated fluoroquinolone resistance in Streptococcus (Boncoeur et al., 2012; Baylay et al., 2015). However, the expression of AMR gene is regulated by other components, such as membrane sensor protein (Depardieu et al., 2007). For example, component regulatory systems bceRS take part in the regulation of bacitracin resistance in Streptococcus (Ouyang et al., 2010). Therefore, AMR phenotypes should be determined to further disclose the relationship between the phenotype and the genotype of SDSD with respect to AMR.

To the best of our knowledge, this is the first report on characterization of Streptococcus dysgalactiae subsp. dysgalactiae isolated from mastitis cows in China using whole-genome sequencing. The apparent prevalence of SDSD was estimated at 7.6%. Eleven sequence types (ST298, ST305, ST453, ST454, ST460, ST461, ST521, ST523, ST526, ST527, and ST529) were determined according to MLST. A total of 108 VFs in 11 categories (adherence, enzyme, immune evasion, immune reactive antigen, iron and manganese uptake, protease, peptidase, superantigen, toxin, and others) and 23 AMR genes in 11 categories (aminoglycoside, tetracycline, vancomycin, bacitracin, fluoroquinolone, lincomycin, nucleoside, peptide, macrolide, β-lactams antimicrobials, and multidrug) were identified according to whole genome sequence analysis. These results can be used to study bovine-associated SDSD virulence and resistance genotypes, as well as to better understand the phylogenetic relationships among mastitis-derived SDSD in Northern China.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

This study was conducted in accordance with ethical guidelines and standard biosecurity and institutional safety procedures of China Agricultural University (CAU; Beijing, China). Prior to the start of the study, ethical approval was granted by the Departmental Committee of the College of Veterinary Medicine, CAU.

BH conceived and designed the experiment. SX, YL, MZ, JY, FH, JG, and ZD performed the experiments, analyzed the data, and contributed to drafting the manuscript. SX, JG, JK, ZD, and BH wrote, reviewed, and edited the manuscript. All authors approved the final version of the 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.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.