Correspondence: Terutoyo Yoshida, Faculty of Agriculture, University of Miyazaki, Miyazaki 889-2192, Japan. Tel.: +81 985 58 7231; fax: +81 985 58 2884; e-mail: firstname.lastname@example.org
The Lancefield group C α-hemolytic Streptococcus dysgalactiae ssp. dysgalactiae (GCSD) causes systemic granulomatous inflammatory disease and high mortality rates in infected fish. Superantigen and streptolysin S genes are the most important virulence factors contributing to an invasive streptococcal infection. PCR amplification revealed that all strains isolated from moribund fish harbored the streptolysin S structural gene (sagA). GCSD fish isolates were PCR negative for emm, speA, speB, speC, speM, smeZ, and ssa. However, the size of the streptococcal pyrogenic exotoxin G (spegg) locus, a superantigen, in positive S. dysgalactiae fish and pig strains was variable. The ORF of the spegg locus of 26 GCSD fish strains and one GCSD pig strain was inserted with IS981SC. Interestingly, the ORF of the spegg locus of two fish strains of GCSD collected in Malaysia was inserted with an IS981SC–IS1161 hybrid IS element. The hybrid IS element was found in all of the GCSD fish isolates and one GCSD pig through PCR screening. Although no insertion sequence (IS) was detected in the spegg locus of S. dysgalactiae ssp. equisimilis (GCSE) strains, a five-nucleotide deletion mutation was detected in the ORF of the spegg locus of one GCSE strain at the supposed site of IS981SC insertion, resulting in a frameshift mutation.
Streptococcus dysgalactiae ssp. dysgalactiae is a Gram-positive bacterium belonging to α-hemolytic Lancefield group C streptococci (GCSD) (Vieira et al., 1998). Animals such as cows and sheep are natural reservoirs of GCSD (Woo et al., 2003). GCSD is mainly associated with mastitis, subcutaneous cellulitis, and toxic shock-like syndrome in bovines (Chénier et al., 2008); suppurative polyarthritis in lambs; and other animal infections (Scott, 2000; Lacasta et al., 2008). GCSD occasionally causes cutaneous lesions, lower limb cellulitis, meningitis, and bacteremia in humans (Bert & Lambert-Zechovsky, 1997; Woo et al., 2003; Fernández-Aceñero & Fernández-López, 2006). The first epizootic outbreak caused by α-hemolytic GCSD among cultured fish populations took place in southern Japan in 2002. The infected yellowtail (Seriola quinqueradiata) and amberjack (Seriola dumerili) exhibited a typical form of necrosis in their caudal peduncles and high mortality rates (Nomoto et al., 2004, 2006, 2008; Abdelsalam et al., 2009b). Mortality is considered to be caused by systemic granulomatous inflammatory disease and severe septicemia (Hagiwara et al., 2009). This pathogen has been isolated from kingfish Seriola lalandi in Japan; gray mullet Mugil cephalus, basket mullet Liza alata, and cobia Rachycentron canadum in Taiwan; hybrid red tilapia Oreochromis sp. in Indonesia; pompano Trachinotus blochii and white-spotted snapper Lutjanus stellatus in Malaysia; pompano T. blochii in China (Abdelsalam et al., 2009a, b, 2010); and Amur sturgeon Acipenser schrenckii in China (Yang & Li, 2009), indicating the increasing importance of this pathogen. In addition, Koh et al. (2009) reported that GCSD caused ascending upper limb cellulitis in humans engaged in cleaning fish and hence may be considered an emerging zoonotic agent. Despite its clinical significance, the fish GCSD genome and the genetic basis of its virulence remain unknown. Therefore, the development of a vaccine against this pathogen is hindered in aquaculture due to the lack of knowledge regarding its pathogenesis and virulence determinants. M protein (emm), superantigen, and streptolysin S genes are important virulence factors in group A Streptococcus pyogenes (GAS) and group C and G S. dysgalactiae ssp. equisimilis (GCSE and GGSE, respectively) due to the contribution of these factors to invasive infections in humans and mammals (Proft et al., 1999; Igwe et al., 2003; Woo et al., 2003; Zhao et al., 2007). Phylogenetic analysis of 16S rRNA gene in different streptococcal species revealed that GCSD clustered together with GGSE, while GAS was found in a closely related neighboring cluster (Facklam, 2002). Because the genetic material of closely related pathogens are important pools from which novel genetic traits can be acquired, in this study, we investigated the occurrences of M protein (emm), superantigen genes, and streptolysin S structural gene (sagA), none of which had been demonstrated to exist in piscine isolates of GCSD. We also analyzed the prevalence of the streptococcal pyrogenic exotoxin G gene (spegg) in piscine GCSD.
Materials and methods
Table 1 lists the 44 strains used in this study. The fish isolates of GCSD (n=30) investigated in this study were obtained from clinical specimens of infected fish in Japan (yellowtail, amberjack and kingfish), Taiwan (mullet), and Malaysia (pompano and white spotted snapper), from the summer of 2002 to the end of 2007. Mammalian isolates of both the α-hemolytic GCSD (n=9) and the β-hemolytic Lancefield group C S. dysgalactiae ssp. equisimilis GCSE (n=5) collected from pigs with endocarditis were kindly provided by the Kumamoto Prefecture Meat Inspection Office in Japan. These isolates were also used for comparison.
Table 1. Streptococcus dysgalactiae strains used in this study
Stock cultures of GCSD and GCSE isolates were maintained in Todd–Hewitt broth (Difco, Sparks, MD) at −80 °C. All isolates were routinely aerobically grown on Todd–Hewitt agar (THA; Difco) or blood agar (Columbia agar base; Becton Dickinson, Cockeysville, MD) containing 5% sheep blood (Nippon Bio-Test Laboratories, Japan), and incubated at 37 °C for 24 h. Lancefield serotyping C (Lancefield, 1933) was confirmed for GCSD and GCSE using Pastorex Strep (Bio-Rad, Marnes-la-Coquette, France). Genomic DNA was extracted from bacterial colonies using DNAzol® reagent (Invitrogen, Carlsbad) according to the manufacturer's protocol. To discriminate fish isolates of GCSD from mammalian isolates of GCSD and GCSE, the specific PCR detection of fish isolates of GCSD using fish sodA gene primers has been performed according to the previously described method (Nomoto et al., 2008).
Detection of virulence genes by PCR
Genomic DNA of GCSD fish isolates was screened by PCR for the presence of emm genes (Zhao et al., 2007), streptococcal pyrogenic exotoxin genes including speA, speB, speC (Hashikawa et al., 2004), speM, smeZ, ssa (Igwe et al., 2003), spegg, and sagA (Ikebe et al., 2004). This PCR assay was performed as described in the references. The primers used by Ikebe et al. (2004) for the amplification of spegg and sagA were designed to yield bands of 205 and 113 bp, respectively. Therefore, the sagA and spegg primers are redesigned for amplifying larger-size bands to examine these gene sequences. The primer pairs of sagaF (5′-TACTTCAAATATTTTAGCTACT-3′) and sagaR (5′-GATGATACCCCGATAAGGATAA-3′) for amplifying a 487-bp segment of sagA were designed based on the streptolysin S genes of S. dysgalactiae ssp. equisimilis (AY033399). The thermal scheme of PCR cycle was as follows: denaturing for 4 min at 95 °C, 35 cycles of denaturing for 30 s at 95 °C, annealing for 30 s at 50 °C, elongation for 50 s at 72 °C, and a final cycle for 10 min at 72 °C. Specific primer pairs Seg1 and Seg12 were used to amplify the full lengths of the spegg genes. These primer pairs were designed based on the spegg gene sequence of S. dysgalactiae ssp. equisimilis (AB105080) (Table 2). PCR was performed under the same conditions as those used for the sagA gene, except that the annealing temperature was set at 48 °C and the elongation was set for 2 min.
Table 2. Primers used for sequencing the spegg gene with associated IS-like elements
Primer name (direction)
The primers pairs (Seg1 and Seg12) and (Seg3 and Seg4) designed from the spegg gene sequence of Streptococcus dysgalactiae ssp. equisimilis (AB105080).
To elucidate the mechanism of the size variation of the spegg loci in GCSD and GCSE isolates, we cloned and sequenced the spegg gene extracted from three fish isolates (94414, KNH07901, and PP1398) and two pig isolates (PAGU656 and PAGU657). Table 2 lists the primers used for sequencing the complete spegg gene in the fish isolates. The purified PCR fragments were directly cloned into the pGEM-T Easy vector® plasmid using T4 ligase (Promega, Madison, WI), and the cloned plasmid was then transformed into Escherichia coli DH5α using the heat shock method. The transformed clones were screened by colony PCR with oligonucleotide primers SP6 (5′-ATTTAGGTGACACTATAGAA-3′) and T7 (5′-TAATACGACTCACTATAGGG-3′). Plasmid DNAs of the clones containing the correct insert segments were then purified and sequenced using the QIAprep Spin Miniprep kit (Qiagen, Germantown, MD). Sequencing reactions were then performed using the GenomeLab DTCS Quick Start Kit (Beckman Coulter, Fullerton, CA) with oligonucleotide primers SP6 and T7. The samples were then loaded into the CEQ 8000 Genetic Analysis System (Beckman Coulter) for the determination of nucleotide sequences. The determined nucleotide sequences were analyzed using bioedit version 7.0 (Hall, 1999). The sequenced spegg was phylogenetically analyzed by the neighbor-joining method using mega version 3 (Kumar et al., 2004).
The presence of the IS981SC–IS1161 hybrid insertion sequence (IS)-like element in S. dysgalactiae
Searching for the presence of the IS981SC–IS1161 hybrid IS-like element in GCSD and GCSE isolates, PCR amplification was carried out using the primer pairs Seg8 and Seg9, which were derived from the nucleotide sequence of the IS981SC–IS1161 hybrid IS element of fish isolate PP1398.
Nucleotide sequence accession numbers
The complete nucleotide sequence of the spegg locus with IS of fish strains 94414, KNH07901, and PP1398 and from GCSE pig strains PAGU656 and PAGU657 were submitted to the DNA Data Bank of Japan under the accession numbers AB452994, AB470100, AB476406, AB518059, and AB448732, respectively.
GCSD fish isolates were PCR negative for emm, speA, speB, speC, speM, smeZ, and ssa. However, all the GCSD fish isolates were PCR positive for the sagA gene (Table 1). On the other hand, 28 fish isolates of GCSD, one pig isolate of GCSD, and three pig isolates of GCSE were PCR positive for the spegg gene (Table 1). Interestingly, size variation was observed in the amplified fragments obtained from organisms having the spegg gene when the primer pairs Seg1 and Seg12 were used (Fig. 1). The positive fish and pig isolates could be divided into three groups. The first group consisted of three strains of GCSE with the expected 833-bp PCR products. The second group consisted of 26 fish isolates of GCSD, and one strain of pig GCSD that had PCR products above 1 kb, which was markedly larger than expected. The third group included two fish isolates (PF880 and PP1398) of GCSD that had PCR products above 2 kb, which was also larger than expected.
Nucleotide sequence analyses of the spegg gene
On the basis of the nucleotide sequences of spegg genes extracted from fish and pig isolates, the size variation was confirmed to be due to the presence of IS in the spegg locus of fish isolates. The spegg locus obtained from GCSD fish strains (94414 and KNH07901) was 2059 bp long due to the presence of a 1224 bp IS. This IS was found to have 99% similarity to IS981SC of Streptococcus iniae (AY904444). The spegg locus sequence was interrupted 604 bp downstream from its start codon by IS981SC, resulting in a 3-bp (5′-ATA-3′) duplication at the insertion site. IS981SC contained two ORFs, designated ORF1 and ORF2, which encode 86 and 279 amino acids, respectively. These ORFs were oriented in the direction opposite to that of the spegg (Fig. 2c). The spegg locus obtained from GCSD fish strain PP1398 was 3236 bp in length due to the presence of two IS: IS981SC and IS1161 (Fig. 3). The IS981SC sequence was interrupted after 50 bp from its noncodon part of the 3′ end by the IS1161-like element, resulting in a 13-bp (5′-ATTTTAATCTATT-3′) duplication at the insertion site. The IS1161-like element sequence has 98% similarity to IS1161 (NC_011375) of the S. pyogenes strain (NZ131). The IS1161-like element was 1164 bp in length. IS1161 has one ORF that encodes 342 amino acids, and this ORF was oriented in the direction opposite to that of spegg, but in the same direction as that of the two ORFs in IS981SC (Fig. 2d). The hybrid IS981SC–IS1161-like element was found to be inserted into the same location as that of IS981SC in the spegg locus of fish strain KNH07901, resulting in an insertion mutation in ORF of spegg (Fig. 3). The spegg locus obtained from the pig GCSE (PAGU657) strain showed a five-nucleotide deletion mutation, from nucleotides 604 to nucleotides 608 (5′-AAGCT-3′), in the ORF of spegg (Fig. 2a). The spegg locus obtained from the pig strain of GCSE (PAGU656) yielded the expected product size and had 100% similarity to spegg variant 4 (AB105080), which has one ORF that encodes 234 amino acids (Fig. 2b). As expected from the sequence similarity results, phylogenetic analysis revealed that spegg of fish isolates (without IS) was related to that of β-hemolytic S. dysgalactiae ssp. equisimilis. Moreover, spegg of fish isolates could be distinguished from that of S. pyogenes (Fig. 4).
Occurrence of the IS981SC–IS1161 hybrid IS element in S. dysgalactiae
The presence of the IS981SC–IS1161 hybrid IS-like element in various isolates of GCSD and GCSE was screened by PCR analysis using specific primers Seg8 and Seg9. All fish isolates of GCSD and one isolate of pig GCSD (dNo. 112-2) contained homologous sequences of the IS981SC–IS1161 hybrid IS-like element (Table 1), and as determined from agarose gel electrophoresis, the amplified DNA fragments of various isolates were of the expected size, 740 bp.
Group C streptococci (GCS) were found in porcine β-hemolytic GCSE strains and in bovine, porcine, and piscine α-hemolytic GCSD strains (Nomoto et al., 2004; Brandt & Spellerberg, 2009). Compared with those of other GCS members, little is known of the virulence factors of α-hemolytic GCSD. Within GCS, superantigenic exotoxins (seeH, seeI, seeL, and seeM) were characterized for the animal pathogenic species Streptococcus equi ssp. equi, while S. equi ssp. zooepidemicus has been shown to possess seeL and seeM (Holden et al., 2009; Paillot et al., 2010). Chénier et al. (2008) and Brandt & Spellerberg (2009) reported that bovine α-hemolytic GCSD screening failed to reveal any superantigen genes. In the present study, GCSD fish isolates were revealed to be PCR negative for emm, speA, speB, speC, speM, smeZ, and ssa when annealing structural gene sequence primers were used. This result indicated that either these genes did not exist within the isolates or that the isolates possessed gene variants that could not be detected by the primers that had been designed based on S. pyogenes sequences. On the other hand, 28 isolates of fish GCSD, one isolate of pig GCSD, and three isolates of pig GCSE were found to contain the spegg gene. Previous studies revealed that only spegg was detected from the clinical isolates of β-hemolytic S. dysgalactiae ssp. equisimilis (Hashikawa et al., 2004; Ikebe et al., 2004; Zhao et al., 2007), but not from α-hemolytic S. dysgalactiae ssp. dysgalactiae (Zhao et al., 2007). The spegg gene of β-hemolytic S. dysgalactiae was found to have properties similar to those of superantigens, and it is likely that the spegg genes play a pathogenic role in animals through having mitogenic activity toward bovine peripheral blood mononuclear cells and selectively activating bovine T cells bearing Vβ1,10 and Vβ4 (Zhao et al., 2007).
In the present study, we observed size variation of the spegg locus in positive fish and pig strains. IS981SC was confirmed to be inserted into the spegg locus of positive fish isolates of GCSD by PCR and sequencing of spegg genes. The insertion site of IS981SC was identical in all of the investigated isolates. Another interesting feature is the existence of the IS981SC–IS1161 hybrid IS element inserted into the spegg locus of two fish isolates of GCSD collected from Malaysia. All fish isolates and one isolate of pig GCSD carried the IS981SC–IS1161 hybrid IS-like element, except for other pig GCSD and GCSE. This finding suggested that the IS981SC–IS1161 hybrid IS-like element prevailed in fish GCSD isolates collected in various Asian countries.
IS981 was a widespread insertion element in Lactococcus lactis, Streptococcus thermophilus (Bourgoin et al., 1999; Bongers et al., 2003), S. iniae (Lowe et al., 2007), and fish isolates of GCSD. IS1161 was also a widespread insertion element in S. pyogenes (McShan et al., 2008), the complete genome of GGSE (AP010935), and GCSD fish isolates. Genes that encode virulence traits are often associated with mobile genetic elements such as IS elements that recruit foreign genes. Moreover, IS can contribute to genetic rearrangements such as translocation, duplication, inversion, and deletion (Vasi et al., 2000; Bongers et al., 2003; De Visser et al., 2004). The disseminations of IS981 and IS1161 in various isolates of streptococci collected from different sources suggested that recombination and horizontal gene transfer events might occur in these species. IS can also form compound transposons by flanking other genes to promote the horizontal gene transfer of virulence genes. It may be possible that IS981SC, IS1161, and spegg are the remnants of a compound transposon. Sachse et al. (2002) reported that the origin of spegg in S. pyogenes might be S. dysgalactiae ssp. equisimilis via horizontal gene transfer.
Interestingly, the nucleotide sequence of pig isolate of GCSE PAGU657 revealed a deletion mutation at the supposed site of IS981SC insertion. IS981SC was found to mediate L. lactis mutations, including simple insertions of IS981SC into new sites of bacterial genome and recombinational IS981SC deletion from the bacterial genome (De Visser et al., 2004). This finding might explain the five-nucleotide deletion mutation of GCSE (PAGU657) at the supposed insertion site of IS981SC, suggesting that IS981SC may contribute to virulence. The deletion and insertion mutations may contribute to the evolution of bacterial pathogenesis and could promote recipient pathogen virulence.
The present study also revealed that sagA was also present in all of the GCSD fish isolates using the primer pair sagaF and sagaR, and the sequenced fragments revealed no difference between the predicted amino acids sequences of the sagA gene extracted from fish isolate (AB520742) and that extracted from S. dysgalactiae ssp. equisimilis (AY033399) (data not shown). Woo et al. (2003) reported that the sagA gene was identified in α-hemolytic GGSE. Immunological studies have recently provided convincing evidence that sagA is the structural gene that encodes streptolysin S. This gene was considered to be a factor contributing to the pathogenesis of streptococcal necrotizing soft tissue infection (Humar et al., 2002) and to the virulence potential of S. iniae infection in fish (Locke et al., 2007). Our findings indicate that α-hemolytic fish GCSD isolates carried some virulence genes that may be responsible for S. dysgalactiae ssp. equisimilis virulence and pathogenesis. Therefore, α-hemolytic fish GCSD isolates should not be disregarded as putative infectious disease agents in humans and mammals.
The authors are grateful to Dr Lauke Labrie, head of the aquatic animal health team of Schering-Plough Animal Health, Singapore, for kindly providing S. dysgalactiae isolates. This study was partially supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Culture and Sports, Japan (21580229). The first author of this paper appreciates the financial support from the Ministry of Higher Education, Egypt, during the study period.