Introduction

In 1986, Devriese et al. [1] designated a species of Lancefield carbohydrate antigenicity group G streptococci from animals as Streptococcus canis. On sheep blood agar plates, this microorganism forms large, smooth, gray/white-colored colonies with β-hemolysis. In healthy dogs, S. canis constitutes part of the resident microflora of the oropharynx, skin, urogenital tract, and anus [2]. This bacterium is an emerging pathogen causing self-limiting dermatitis among companion animals (i.e., dogs or cats) [3]. However, S. canis infection can occasionally result in severe diseases in dogs and cats, including arthritis, streptococcal toxic shock syndrome, necrotizing fasciitis, septicemia, and pneumonia [4, 5]. We have also reported a case of severe soft tissue infection with septic shock caused by S. canis in a miniature dachshund [6]. Additionally, S. canis can infect humans who have been in close contact with companion animals and cause either local or systemic diseases [7]. This species was recovered from two Japanese patients with bacteremia who were in deep contact with or bitten by pet dogs [8, 9]. Periprosthetic joint infection with S. canis has been described in a man undergoing elective primary total hip arthroplasty [10]; a pet dog had frequently licked his legs. Many Japanese individuals keep dogs or cats in their homes. Moreover, medical institutes and nursing homes are introducing animal-assisted therapy as a mental health service for hospital patients and elderly individuals. Companion animals and humans are living closely. Thus, it is important for veterinary and medical doctors to be aware of the possibility of S. canis-related zoonotic infections being underdiagnosed.

The S. canis M-like protein (SCM), which is encoded by the scm gene, can bind to plasminogen and immunoglobulin G and confers antiphagocytic properties [11]. We performed polymerase chain reaction (PCR)-based amplification of scm (with amplicon sizes of 1,700–2,100 bp) and conducted direct sequencing [12, 13]. We constructed an unrooted phylogenetic tree of the deduced amino acid (AA) sequences using the neighbor-joining method. SCM allele typing was performed based on different/similar positions using variable/conserved AA sequences in the phylogenetic tree. Allele types were classified into two groups: group I, with relatively similar sequences (consisting of allele types 1–9) [14] and group II, with diverse sequences (consisting of allele types 10–15) [12]. The typing in group I was performed based on variable AA sequences with signal peptide types at the amino terminus [14].

Streptococcus pyogenes, a virulent human pathogen exhibiting carbohydrate group A, also possesses an antiphagocytic M protein (encoded by the emm gene). The M protein’s trans-acting positive regulator, also known as the multiple gene activator (Mga), is a DNA-binding transcriptional activator protein. Mga can enhance the expression of multiple genes, such as emm, scpA, which encodes C5a peptidase, and mga itself, implying that it is an autoregulator [15]. The mga-emm-scpA genes are closely linked and arranged in tandem, and these genes are referred to as the ‘mga regulon’. Helix-turn-helix (HTH) AA secondary structures have DNA-binding activities. The amino-terminus of Mga contains four potential HTH DNA-binding domains (HTH1–HTH4); two of these domains, HTH3 and HTH4, are needed for direct activation of the ‘mga regulon’ in vivo [16]. Furthermore, it has been established that the conserved Mga domain 1 (CMD-1) likely contributes to the protein stability with (auto)activation [17]. Thus, HTH3/HTH4 and CMD-1 are targeted for Mga functional analysis.

Streptococcus dysgalactiae subsp. equisimilis (SDSE) and subsp. dysgalactiae, S. equi, S. gordonii, S. mitis, and S. pneumoniae also contain Mga ortholgues [17]. However, there are very few descriptions regarding the genetic organization of the scm gene region, which is similar with the mga–emm–scpA linkage. Thus, the purpose of this study was to examine the presence of an Mga orthologue and its related SCM alleles in S. canis.

Methods

Comparison of genomic structures from S. pyogenes and S. canis strains

We performed the comparison of genomic structures from S. pyogenes strain JRS4 and S. canis National Collection of Type Cultures (NCTC) 12,191(T). Additionally, the comparison of genomic structures from other S. pyogenes strains and other S. canis strains was carried out. Genomic structures were constructed based on the whole-genome sequence (WGS) graphics specified in the GenBank descriptions of the National Center for Biotechnology Information (NCBI) database.

PCR-based amplification and direct sequencing of mga gene

We enrolled S. canis isolates collected during the three previous study periods in 2015 (n = 17), 2017 (n = 6), and 2021 (n = 16) (Table 1) [18,19,20]. The isolates were identified based on the 16 S sequencing results. The corresponding animal information regarding sex and year-age is shown in Table 1. The thirty-nine isolates contained the determined SCM alleles (including the truncated variants) (Table 1). Streptococci genomic DNA was extracted by suspension in 10 mM Tris-1 mM EDTA (pH 8.0), followed by boiling at 97 °C for 10 min and a brief microfuge step after the boiling lysis [21]. Two amplifying primers and one sequencing primer (mga-F1, mga-R2, and mga-F2 shown in Fig. 1) were designed based on the WGS of S. canis NCTC 12,191(T) using the web-based application Primer3Plus [22]. NCTC 12,191(T)-origin DNA was used as a positive control, and DNase/RNase/protease-free water was used as a negative control in each PCR assay. PCR was performed with an initial denaturation step at 94 °C for 1 min, followed by 30 cycles (consisting of denaturation at 94 °C for 1 min, annealing at 50 °C for 1 min, and extension at 72 °C for 2 min), and a final extension step at 72 °C for 10 min. PCR products with the expected amplicon size (1801 bp) were separated using 1.5% agarose gel electrophoresis in Tris-acetate-EDTA buffer. Direct sequencing after amplicon purification by QIAquick PCR Purification Kit (Qiagen, Tokyo, Japan) was conducted on the Applied Biosystems 3730xl DNA Analyzer with the BigDye Terminator V3.1 (Thermo Fisher Scientific, Waltham, MA, USA). We obtained the coding DNA sequences and deduced the AA sequences.

Table 1 M protein trans-acting positive regulator (Mga) sequence and its adjacent M-like protein (SCM) allele of Streptococcus canis
Full size table
Fig. 1
figure 1

Two genomic structures from Streptococcus pyogenes strain JRS4 and Streptococcus canis strain National Collection of Type Cultures (NCTC) 12,191(T). These structures were constructed based on the whole-genome sequence (WGS) graphics specified in the GenBank descriptions (accession numbers CP011414.1 and LR134293.1) of the National Center for Biotechnology Information database

Full size image

WGS-based detection of mga and scm genes

We retrieved WGSs from S. canis strains (n = 22), along with the WGS of NCTC 12,191(T) (gray shading in Table 1), which were deposited in the NCBI database (updated August 1, 2023) for this retrospective study. The Japanese animal information regarding sex and year-age is shown in Table 1. Putative Mga-related nucleotide/AA sequences were obtained by inserting NCTC 12,191(T) Mga sequence into the NCBI Nucleotide/Protein Basic Local Alignment Search Tool [23]. We also retrieved SCM nucleotide/AA sequences adjacent to Mga. Allele typing was performed based on two previous typing methods, and the alleles were classified into the groups I–II [12, 14].

Determination of Mga-specific AA motifs and percent identity to Mga AA sequence in the type-strain

We examined the presence of Mga-specific AA motifs in S. canis as compared to those of S. pyogenes [16, 17]. The percentage identity to Mga AA sequence in NCTC 12,191(T) was examined, along with its AA size in each strain.

All analyses were conducted at the Graduate School of Infection Control Sciences and Ōmura Satoshi Memorial Institute, Kitasato University.

Results

Comparison of genomic structures from S. pyogenes and S. canis strains

Figure 1 shows two genomic structures from S. pyogenes JRS4 and S. canis type-strain. These structures were made based on the WGS graphics specified in the GenBank descriptions (accession numbers CP011414.1 and LR134293.1). The genetic organization between the mga-emm locus is consistent with that between the mga-scm (spaZ) locus. In contrast, the downstream gene arrangements were different. The scpA/B is located at 367 nucleotides downstream of emm in S. pyogenes, whereas the relA (encoding bifunctional (p)ppGpp synthetase/guanosine-3’,5’-bis(diphosphate) 3’-pyrophosphohydrolase) is located at 302 nucleotides downstream of scm (spaZ) in S. canis. In other S. pyogenes strains (NCTC 8198/Culture Collection University of Gothenburg 4207/1085), there was the organization between the mga-emm locus and the different downstream arrangement (including scpA/B) of emm. For example, these three strains had the mga-emm-gene (encoding YSIRK-type signal peptide-containg protein)-sic./gene (encoding lysis inhibitor protein)-gene (encoding IS1182 family transposase)-scpA/B arrangement. In other S. canis strains (NCTC 6198/OT1/TA4), there was the organization between the mga-scm (spaZ)-relA locus. Furthermore, we found the relA possession in S. pyogenes strains and the scp possession in S. canis strains. The relA was shown to be located at distant position from the mga-emm locus and the scp was shown to be located at distant position from the mga-scm locus.

Background information about enrolled strains

Table 1 lists the strain background information recorded in our previous investigations (2015–2017–2021) or in the NCBI database. The enrolled strains were from animals (n = 58) and humans (n = 4); the collection years were from 1999 to 2021/2022; the geographic location included forty-nine isolates from Japan and 12 overseas strains; and the isolation sources constituted seven invasive strains (from blood and uterus) and fifty-four noninvasive strains (mainly from ear, pus, urogenital tract, eye, and nose).

Characterization of an Mga orthologue and SCM alleles

The detailed results regarding Mga nucleotide/AA sizes, percentage identity to the type-strain Mga sequence, along with AA sequences at CMD-1 and HTH3/HTH4 domains in each strain are shown in Table 1. We observed mga nucleotide size of 1,530 bp (except for 1,590 bp and 1,529 and 1,531 bp resulting in two truncated variants) and Mga AA size of 509 (except for 529 AA and 10–126–400 AAs of three truncated variants). The percentage identity ranged from a minimum of 96.66% to a maximum of 100%. We found the presence of CMD-1 (including two flanking AAs: Q–L) and two HTH3/HTH4 domains (containing YK and LS motifs) at the amino-terminus to assess the potential Mga orthologous structure associated with its function, because the AA sequences at positions 10–15, 53–72, and 107–126 of S. pyogenes strain JRS4 Mga (530 AAs) were QQWREL, MQFMKEVGGITYKNGYITIW, and LEELAEELFVSLSTLKRLIK, respectively (Fig. 2). Almost all the strains (except for a truncated variant strain KU109 shown in Table 1) had the CMD-1 (including two flanking AAs: Q–L) at AA positions 11–16 or 31–36. Additionally, almost all the strains (except for the truncated variant KU109) possessed the HTH3 domain (containing YK/HK motifs) at positions 54–73, 64–83, or 74–93. Furthermore, almost all the strains (except for the truncated variant KU109) contained the HTH4 domain (containing LS motif) at positions 108–127, 118–137, or 128–147. Thus, we confirmed the potential Mga orthologous structure associated with its function among the registered strains. In contrast, we observed the diverse SCM allele populations consisting of groups I (n = 33) and II (n = 26), along with three truncated variants. Group I included alleles 1–9, whereas group II included alleles 10–15.

Fig. 2
figure 2

Multiple gene activator (Mga) amino acid (AA) structure (530 AAs) of Streptococcus pyogenes strain JRS4 is shown on the upper side. Potential three functional domains are conserved Mga domain 1 (CMD-1) and helix-turn-helix (HTH) DNA-binding domain 3–4 (HTH3–HTH4) that are located at the amino terminus [15, 16]. AA residues composing the three functional domains are shown on the lower side. AA positions are indicated in parentheses. HTH3/HTH4 with YK and LS motifs and CMD-1 with two flanking (Q–L) are targeted for Mga functional analysis

Full size image

Discussion

Group C SDSE, which is closely related to S. canis, has an orthologous gene (mgc), a multigene regulator. Mgc (513 AA) in SDSE strain H46A was 51.3% identical to Mga in S. pyogenes strain D471 [24]. The phylogenetic analysis indicated that Mgc in SDSE constituted a distinct cluster separated from Mga in S. pyogenes [24]. It seems likely that the SDSE/S. canis mgc/mga and S. pyogenes mga have undergone a considerable period of independent evolutionary development.

We searched for related articles by entering the keywords “streptococcus canis, transcriptional regulator,” “streptococcus canis, multiple gene activator,” and “streptococcus canis, DNA-binding” in the PubMed [25]. However, there appear to be no adequate hits in related manuscripts (as of January 11, 2024). To the best of our knowledge, this is the first report of a homologous sequence of Mga and its adjacent diverse SCM alleles in S. canis, suggesting its operon, which is similar with the S. pyogenes ‘mga regulon’. Based on the diversity, we further should establish the SCM allele typing for molecular epidemiological approaches. Two mga alleles (mga-1 and mga-2) are found within S. pyogenes based on their ability to bind to an oligonucleotide probe [26] and are associated with different genetic patterns at mga locus and different tissue tropisms [27]. Therefore, it is important to carry out sequential analysis among additional S. canis strains to monitor the development of MGA alleles in our future observations.

Limitations

We need to further determine whether this molecule has the functional ability to bind to scm, mga, and other genetic regions including their promoter sequences and to activate their transcription by in vitro/in vivo experiments.