Characterization of Streptococcus criceti insertion sequence IS Scr 1

A novel insertion sequence element of the IS982 family, ISScr1, was previously identified in Streptococcus criceti strain E49 as a disrupted paaB gene encoding an antigen I/II homologous protein. In this study, we identified two divergent inserted regions of ISScr1 in S. criceti E49 by inverse polymerase chain reaction, the gene-walking method, and screening of the partial plasmid library. Nucleotide sequence analysis indicated the possible explanation that transposition generated 8and 9-bp direct repeat sequences. DNA hybridization analysis revealed that an identical hybridization pattern to ISScr1 was observed in the four S. criceti strains studied and that at least three copies of ISScr1 were preserved in S. criceti strains. In addition, we found different susceptibility to erythromycin and diverse agglutination properties induced by dextran in S. criceti. Furthermore, DNA hybridization analysis showed that no ISScr1-like copy was detected in the other 14 strains of oral streptococci tested.

S. criceti is a bacitracin-sensitive cariogenic agent isolated from the oral cavities of hamsters, wild rats, and humans in North Africa (Loesche, 1986) and inheres a trait of aggregation in a dextran-dependent manner (Drake et al., 1988).As S. criceti is sensitive to bacitracin, this agent is excluded by using MSB media.The initial question of this study was why S. criceti is sensitive to bacitracin in contrast to S. mutans and S. sobrinus.
Furthermore, there are no antibiograms on S. criceti strains but S. criceti strain E49 is a streptomycin resistance strain (Smith et al., 1978).The susceptibility to antibiotics of S. criceti might be associated with the inactivation of antibiotic resistance transporter systems by acquisition of IS.Regarding IS, ISScr1, a member of the IS982 family, has only been identified from S. criceti E49.The 962-bp ISScr1 was observed to inactivate the paaB gene encoding a cell-surface protein of antigen I/II family protein (Tamura et al., 2004) (Fig. 1).IS982 was first identified in lactococci (Yu et al., 1995) and IS982 family elements are distributed in Gram-positive bacteria (Mahillon and Chandler, 1998).To our knowledge, besides ISScr1 of S. criceti, IS982 members identified from streptococci are ISSa4 of S. agalactiae (Spellerberg et al., 2000), which is a causative organism of invasive neonatal human diseases (Gottschalk et al., 2006) and ISSmu5 of S. mutans (GenBank accession no.AF104381).
To date, little is known about the characteristics of ISScr1 and its involvement in the properties of S. criceti.We decided to identify insert sites and copy numbers of ISScr1 in its genome and performed inverse PCR (IPCR), gene walking, screening of plasmid clones from S. criceti E49, and DNA hybridization analysis for ISScr1.Herein we report two discrete insertion sites of ISScr1 in S. criceti E49 and the distribution of ISScr1 in S. criceti strains.

IPCR and gene walking
To amplify flanking regions of ISScr1 in S. criceti E49, IPCR was first achieved with two outward-pointing primers P1 (5'-AAGAAGTTGGC-CGCAAGGA-3') and P2 (5'-GCGTCAAAACATGGCAG-GAG-3') designed on the basis of the nucleotide sequence of the putative transposase gene, tnp, of the 962-bp ISScr1 (Tamura et al., 2004) (Fig. 1), and Taq DNA polymerase (Invitrogen, Carlsbad, CA, USA).The 962-bp ISScr1 has no restriction site of HindIII and DraI restriction enzymes.Chromosomal DNA of S. criceti was completely digested with HindIII or DraI and then selfligated with T4 DNA ligase (Toyobo Biochemicals, Osaka, Japan).The circularized DNA libraries were utilized for templates.The conditions of IPCR were 94°C for 1 min; 25 cycles of 94°C for 30 sec, 50°C for 30 sec, and 72°C for 4 min; and finally 72°C for 10 min.The amplified IPCR fragments were gel-purified using a GeneClean II kit (MP Biomedicals, Solon, OH, USA).The resultant products were directly sequenced.The flanking regions were identified using the gene-walking method as described previously (Tamura et al., 2004).
Construction of a partial plasmid library and screening E49 genomic DNA was completely digested with DraI restriction enzyme and separated on 0.7% agarose gel.The DNA fragments fractionated on agarose gels (around 2.0 kb in size, which were thought to be candidates by DNA hybridization analysis of S. criceti E49 for ISScr1) were ligated into the EcoRV site of pBluescript II SK (+) (Agilent Technologies, Santa Clara, CA, USA) treated with bacterial alkaline phosphatase (Takara Bio, Japan).The resultant plasmid library was screened in E. coli by sequencing.
Sequencing and DNA analysis DNA amplicons were sequenced on both strands using a 3130 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).DNA analysis was performed with computational packages MacVector 7.2 (Accelrys, Inc., San Diego, CA, USA) and Genetyx 11 (Genetyx, Tokyo, Japan).A phylogenetic tree was obtained with the neighbor-joining method using Genetyx 11 (Genetyx) with 1000 trials for bootstrap analysis and the Kimura-2-parameter model.

DNA hybridization analysis
Chromosomal DNA (5 μg) of streptococci was digested with the indicated restriction enzymes, BamHI, EcoRI, HindIII, or DraI (Toyobo Biochemicals), electrophoresed through 0.7% agarose gels, transferred to Hybond-N + nylon membranes (GE Healthcare UK Ltd., Little Chalfont, Buckinghamshire, UK), and probed with a 0.3-kb digoxigenin-labeled NdeI-PstI fragment of plasmid pPCRE4041-1 comprising a portion of the 962-bp ISScr1.Plasmid pPCRE4041-1 was constructed as follows.PCR was performed with primers 40 and 41 using KOD-Plus-DNA polymerase (Toyobo Biochemicals).The PCR products were gel-purified and cloned at the SrfI site of a pPCR-Script Amp SK(+) cloning vector (Agilent Technologies).DNA hybridization and detection were performed using the DIG High Prime DNA Labeling and Detection Starter Kit I (Roche Diagnostics, Mannheim, Germany) as described previously (Tamura et al., 2004).Briefly, hybridization was performed at 42°C for 16 h, followed by washing according to the manufacturer's instructions (Roche Diagnostics), and hybridized fragments were visualized with a chromogenic alkaline phosphatase substrate NBT/BCIP after 16-h incubation.

Antibiotic sensitivity tests
For determination of the minimum inhibitory concentration (MIC) of streptococci, bacitracin, erythromycin, and streptomycin were used.The MICs of strains were determined by broth dilution after 20-h incubation at 37°C, as described previously (Kitagawa et al., 2011).
Dextran-induced agglutination assay Aliquots (500 μl) of overnight cultures adjusted to OD 600 of 1.0 were transferred to a glass tube (10 × 75 mm) with dextran (Mr ∼2000000) from Leuconostoc ssp. at a concentration of 200 μg/ml, vortexed for 20 sec, and left to stand for 30 min at ambient temperature.The agglutination was examined visually.

Identification of insertion sites of ISScr1 in S. criceti E49
To localize insertion sites for ISScr1 copies in the chromosomal DNA of S. criceti E49, IPCR was performed.From HindIII and DraI libraries, two IPCR products (0.9 and 1.4 kb, respectively) were sequenced and it was found that they were identical regions.No other region was detected under the IPCR conditions.DNA amplification using Taq DNA polymerase was limited up to approximately 3 kb during the elongation time (4 min) and the IPCR condition was not enough to amplify the flanking region of the 962-bp ISScr1 in the paaB gene.The flanking region of the IPCR products was sequenced by gene walking.In the determined region (4859 bp; GenBank accession no.AB182586) comprising the second copy of ISScr1, there were four open reading frames (ORFs) other than ISScr1: one was incomplete (ORF1) and three were complete (ORF2-4).ORF1 lacked an initiation codon but contained a stop codon (TAG).ORF1, ORF2, ORF3, and ORF4 coded for proteins of 113, 275, 366, and 270 amino acids, respectively.The 961-bp ISScr1 was situated 548 bp downstream of ORF3 and no gene disrupted by ISScr1 was presented as there was a 15-bp space between ISScr1 and ORF4 (Fig. 2A top).Based on the nucleotide sequence of the insertion-carrying allele, a possible explanation for the boundary was that the ISScr1 duplicated 8-bp target DNA (sequence 5'-TAACCCTT-3') resulting from its insertion and possessed inverted repeats (IRs): 25-bp left inverted repeat (IRL) and 24-bp right inverted repeat (IRR).The IRL sequence had one substituted nucleotide sequence (from C to A) compared to IRL of the 962-bp ISScr1 in paaB (Tamura et al., 2004) (Table 1).
One of 130 clones from the partial plasmid library was identified as comprising the third copy of ISScr1 in the 2112-bp region (GenBank accession no.AB257318).The 960-bp ISScr1 was situated in the region downstream of ORF5 and ORF6 encoding hypothetical proteins composed of 162 and 198 amino acids, respectively (Fig. 2A bottom).A possible explanation for the boundary was that the ISScr1 contained 24-bp imperfect IRs and duplicated 9-bp target DNA (sequence 5'-TAAACACCA-3') resulting from its insertion (Table 1), and ORF5 was found to be disrupted by insertion of ISScr1 because of a 13-bp overlap between ISScr1 and ORF5.Analysis of the deduced amino acid sequence showed that a nonsense mutation at codon 81 (CAA → TAA) occurred in the tnp gene, resulting in a truncated transposase composed of 80 amino acids.Furthermore, the mutation was positioned in the expected region binding primer P1.
Protein sequence search against the data of S. mutans UA159, a genome sequenced strain (Ajdic ´ et al., 2002), revealed that ORF1 and ORF3 in S. criceti exhibited 68.1 and 59.2% identity with the hypothetical protein coded by SMU.1463c and the putative oxidoreductase coded by SMU.1462c, respectively, whereas ORF2, ORF4, ORF5, and ORF6 in S. criceti showed no homology to the proteins of S. mutans UA159.In S. mutans UA159, SMU.1463c and SMU.1462c are arranged in tandem in a region upstream of the rmlABC gene cluster involved in the anabolism of dTDP-L-rhamnose from D-glucose-1phosphate (Tsukioka et al., 1997).In S. criceti, genes homologous to rmlA, rmlB, and rmlC of S. mutans were not found in the corresponding region.

ISScr1 in IS982 family
The 961-bp ISScr1 possessed the tnp gene encoding a putative transposase composed of 287 amino acids (Fig. 2B).Comparison of the nucleotide sequence of 962-bp ISScr1 to that of the 961-bp ISScr1 identified in this study showed that the 201-codon glycine of the tnp gene of 961-bp ISScr1 was altered from GGA to GGC, resulting in no substitution of the amino acid sequence.The stop codon (TAG) of the tnp gene was located within IRR.The calculated molecular mass of the protein was 33.8 kDa and the theoretical pI was 9.99.The DDE catalytic triad motif conserved in transposases (Mahillon and Chandler, 1998) was observed in the deduced amino acid sequence (positions 119, 199, and 245).The deduced amino acid sequence analysis of transposases (287-302 amino acids) in the IS982 family revealed that the transposase of the 961-bp ISScr1 has 88.5, 60.6, 50.0, 38.5, and 34.2% identity to transposases of ISSa4 of S. agalactiae (Spellerberg et al., 2000), ISSmu5 of S. mutans (GenBank accession no.AF104381), ISLpl4 of Lactobacillus plantarum (de las Rivas et al., 2005), IS982 of Lactococcus lactis subsp.cremoris (Yu et al., 1995), and ISEfm1 of Enterococcus faecium (Boyd et al., 2000), respectively.As shown in Fig. 2C, phylogenetic analysis of tnp gene sequences of IS982 members showed that ISScr1 (961-bp long) and ISSa4 formed a

Frequency and distribution of ISScr1 in S. criceti E49 and other strains
To assess the frequency of ISScr1 in S. criceti E49, DNA hybridization analysis was performed.Three hybridized bands were detected in each lane (Fig. 3).Since ISScr1 possesses no restriction sites of BamHI, EcoRI, HindIII, and DraI restriction enzymes, each band should represent at least one element.Based on the nucleotide sequence of 962-bp ISScr1 and flanking regions (Tamura et al., 2001(Tamura et al., , 2004)), the 5.6-kbBamHI, 6.4-kb EcoRI, 6.7-kb HindIII, and 9.7kb DraI fragments correspond to the segment of the 962bp ISScr1.Meanwhile, the 1.3-kb HindIII and 1.9-kb DraI fragments were estimated by the region determined by IPCR and the 2.1-kb DraI fragment was expected to be the region identified by screening; therefore, at least three copies of ISScr1 were preserved in the genome of S. criceti E49.
Furthermore, the copy number of ISScr1 was estimated in the four S. criceti strains by DNA hybridization analysis of EcoRI-or HindIII-digested genomic DNA hybridized with an ISScr1 probe (Fig. 4).The pattern of fragments hybridized to the probe was the same in the four strains.Thus, at least three copies of ISScr1 were preserved in chromosomal DNA of the S. criceti strains examined.

Disruption of paaB gene by insertion of ISScr1 in S. criceti strains
As the paaB gene was disrupted by insertion of the 962-bp ISScr1 in S. criceti E49 (Tamura et al., 2004), we preferentially performed PCR and sequenced a portion of the paaB gene in three S. criceti strains HS-1 (GenBank accession no.AB694740), HS-6 T (AB694741), and OMZ61 (AB694742).The 1279-bp nucleotide sequence determined in each strain was identical to that of strain E49.Based on the nucleotide sequence of the ISScr1-carrying paaB gene in each strain, one possible explanation was that the 9-bp direct repeat (5'-TAGCTAAAT-3') was generated in the flanking region of 962-bp ISScr1 in each strain (Table 1).These findings  indicated that the paaB gene was disrupted in the three strains studied, as in strain E49.

Characterization of S. criceti strains
As 1279-bp nucleotide sequences of the inactivated paaB genes were identical among all strains, we further characterized the strains from two aspects: MICs of antibiotics and agglutination induced by dextran.MICs of bacitracin for S. criceti strains ranged from 0.39 to 1.56 μg/ml, indicating the high sensitivity of S. criceti strains to bacitracin compared with S. mutans MT8148 and S. sobrinus OMZ65 (MIC, 200 μg/ml) (Table 2).There was a marked difference in the MICs of erythromycin for S. criceti strains: OMZ61 exhibited high resistance (MIC >20 μg/ml) but the others were sensitive (MIC 0.156 μg/ml).MICs of streptomycin for S. criceti strains ranged from 25 to 50 μg/ml.Of note, S. sobrinus OMZ65 exhibited high resistance to streptomycin.
As agglutination upon the addition of dextran is characteristic of S. criceti strains and S. sobrinus strains (Drake et al., 1988), it was evaluated (Table 2).Agglutination was observed in S. criceti strains HS-6 T and OMZ61 and S. mutans MT8148, whereas S. criceti strains HS-1 and E49 and S. sobrinus OMZ65 failed to be aggregated.These findings showed the variability of S. criceti strains in dextran-induced agglutination.

Distribution of ISScr1-like copy in oral streptococci
To clarify the distribution of ISScr1-like copy in oral streptococci, DNA hybridization analysis was carried out.Under high stringency conditions, no fragment hybridized to the probe was detected in oral streptococci, except for S. criceti E49 (Fig. 5); thus, no ISScr1-like copy was preserved in the oral streptococci studied.

DISCUSSION
With the aim of characterizing ISScr1, we first performed IPCR and determined a discrete integration site other than the paaB gene locus in S. criceti E49 (Fig. 2A).Furthermore, a third insertion site was identified by screening clones.Nucleotide sequence analysis of the third site revealed a nonsense mutation of the sequence that was expected to bind to primer P1.This might account for the fact that only one site was identified by IPCR.This implied the possibility of the existence of undetected fragments of ISScr1 in the genome.Hybridization analysis in S. criceti E49 indicated that at least three copies of ISScr1 were preserved (Fig. 3).Nucleotide sequence analysis of ISScr1 provided one possible explanation for the boundary: ISScr1 elements comprised 23-25 bp imperfect IRs at the ends and were flanked with 8-9 bp DRs (Table 1).To our knowledge, IS982 family members in streptococci have been identified in ISSmu5 of S. mutans (GenBank accession no.AF104381) and ISSa4 of S. agalactiae (Spellerberg et al., 2000), except for ISScr1.If IS was intact functionally, the distribution and location of IS element would be diverse in strains.ISSmu5, formerly IS1216, comprised 15-bp terminal IRs flanked with 10-bp DR sequences (GenBank accession no.AF104381) and was a rarely found IS in 2 of 33 isolates (6.1%) of S. mutans; the copy number was determined as 10 and 2 in each isolate by hybridization analysis (Argimón and Caufield, 2011).It has been demonstrated that ISSa4 is found in 21 of 113 strains (18.6%) of human origin and the presence of ISSa4 is limited in strains of serotypes II and II/c (Dmitriev et al., 2004).Furthermore, the copy number of ISSa4 varies from more than 15 copies to only 2-4 copies (Spellerberg et al., 2000).In our study, hybridization analysis showed that the hybridized fragment pattern was the same among S. criceti strains (Fig. 4), suggesting that at least three copies of ISScr1 might be positioned on the genome in each strain.Further studies are necessary to determine whether ISScr1 is preserved in clinical isolates.
Transposases of IS elements preserve a catalytic domain with a DDE motif (Mahillon and Chandler, 1998).In the present study, the tnp gene encoding a 287 amino-acid protein of 961-bp ISScr1 was thought to be intact and the tnp gene of 960-bp ISScr1 was predicted to be inactivated because it encoded a truncated nonfunctional protein (Fig. 2, A and B; Table 1).The predicted transposase of 961-bp ISScr1 showed 88.5% amino acid identity to that of 963-bp ISSa4 of S. agalactiae (Spellerberg et al., 2000).Phylogenetic analysis of tnp gene sequences in the IS982 family showed the high resemblance of ISScr1 to ISSa4 (Fig. 2C), suggesting that ISScr1 and ISSa4 might be derived from the same origin.Interestingly, the recent acquisition of ISSa4 by S. agalactiae was presumed by the results of hybridization analyses of clinical isolates (Spellerberg et al., 2000).It remains to be resolved how the acquisition of IS elements by S. criceti and S. agalactiae occurred.
As the homogeneity of S. criceti strains was assumed, the MIC for antibiotics was examined.In accordance with the previous finding that S. criceti is sensitive, whereas S. mutans and S. sobrinus are resistant to bacitracin (Loesche, 1986), the high susceptibility of S. criceti strains to bacitracin was observed.Our results indicated no discernible properties in strain E49 to other strains of S. criceti in regard to resistance to streptomycin (Table 2).This disagreed with the previous report that S. criceti E49 is resistant at a concentration of 2 mg streptomycin per ml (Smith et al., 1978).It is of note that S. criceti OMZ61 was resistant to erythromycin in contrast to other strains.In S. mutans, bacitracin transporters MbrAB and response regulator MbrC are related with resistance (Tsuda et al., 2002;Kitagawa et al., 2011).Our results indicated that there were no homologous transporters to MbrAB or response regulator to MbrC in the sequenced region of S. criceti E49.It is speculated that lack of a mechanism such as bacitracin transporters and a regulator accounts for the bacitracin susceptibility of S. criceti.Further sequence analysis of the genome would resolve these discrete properties among S. criceti strains.
We further investigated dextran-induced agglutination, which is characteristic of S. criceti strains and S. sobrinus strains (Drake et al., 1988).Indeed, S. criceti HS-6 T and OMZ61 were agglutinated, whereas S. criceti E49 and HS-1 and S. sobrinus OMZ65 were not agglutinated under the examined conditions, suggesting that the binding of lectins for dextran varied among strains.This might be supported by the finding that the same pattern was observed in S. criceti E49 and HS-1, but not in HS-6 T , regarding the electrophoretical migration pattern of dextranases (Tamura et al., 2007).
The 962-bp ISScr1 was first found in the paaB gene in S. criceti E49 (Tamura et al., 2004).PCR and sequence analysis revealed that 962-bp ISScr1 was inserted in the same position as paaB in the four studied S. criceti strains.It is noteworthy that the nucleotide sequence of the 1279-bp region was identical in the studied strains.Second, sequence analysis of S. criceti E49 demonstrated that 961-bp ISScr1 was inserted in the region downstream of the genes homologous to SMU.1463c and SMU.1462c in S. mutans UA159 and showed that rmlA, rmlB, and rmlC genes, which are involved in the biosynthesis pathway of dTDP-L-rhamnose in S. mutans (Tsukioka et al., 1997), were not found in the determined region of S. criceti.Finally, the third insertion site of ISScr1 (960-bp ISScr1) was genetically defined but we could not identify the region functionally.Further sequence analysis would reveal information on the gene locus.
In hybridization analyses, no hybridized band was detected in other oral streptococci studied, except for S. criceti E49 (Fig. 5).This finding suggested that the ISScr1-like element was not preserved in oral streptococci; therefore, the use of PCR specific for ISScr1 might be a tool for the detection of S. criceti from dental plaque samples.
It is of interest whether ISScr1 has a transposition activity.We tried to detect the activity of the 962-bp ISScr1 by PCR excision assay using primers 40 and 41 in the four S. criceti strains tested; however, no band corresponding to the fragment containing the restored integra-tion site of the paaB gene was found (data not shown), suggesting that the transposition activity of 962-bp ISScr1 was undetectable under the examined PCR conditions.
In conclusion, we identified two discrete insertion regions of ISScr1 in S. criceti E49 and characterized ISScr1 in four strains of S. criceti.The copy number was estimated to be at least three in chromosomal DNA, and the distribution pattern was identical in the S. criceti strains studied.Although three copies were detected in S. criceti, it remains possible that discrete IS strongly homologous to ISScr1 was preserved in S. criceti.

Fig. 1 .
Fig. 1.Restriction map of the region of the 962-bp ISScr1 disrupting the paaB gene of S. criceti E49.The region from positions 3000 to 16000 of the nucleotide sequence (GenBank accession no.AB042239) (Tamura et al., 2004) is depicted.Positions of restriction enzyme BamHI, DraI, EcoRI, HindIII, NdeI, and PstI recognition sites are shown.Arrows and open box above the line indicate ORFs and ISScr1 element, respectively.The paaB gene (open arrow) was disrupted by ISScr1.The tnp gene (closed arrow) encoded a putative transposase of ISScr1.Arrows beneath the line indicate the positions and orientations of the primers.Thick bar represents the position of the DNA probe yielded from the NdeI-PstI fragment, 0.3-kb, of pPCRE4041-1.

Fig. 2 .
Fig. 2. A. Genetic maps of the regions comprising ISScr1 determined by IPCR and gene walking (top), and plasmid library screening from S. criceti E49 (bottom).Arrows above and below the line indicate the ORFs and positions of primers, respectively.The orf5 gene disrupted by 960-bp ISScr1 is represented by an open arrow.B. Deduced amino acid sequence of the tnp gene of 961-bp ISScr1.Amino acid residues of the DDE catalytic triad motif conserved in transposases (Mahillon and Chandler, 1998) are shown in bold.The 201-codon glycine (G) is underlined.C. Phylogenetic tree based on tnp sequences of five IS982 family members.GenBank accession numbers of members are as follows: ISScr1 (AB182586), ISSa4 (AF165983), ISSmu5 (AF104381), ISEfm1 (AF138282), ISLpl4 (AJ968657), and IS982 (L34754).Numbers indicate percent support for specific nodes after 1000 replications of bootstrap analysis.Scale bar represents 10% estimated sequence difference.

Table 1 .
Characteristics of ISScr1 copies in strains of S. criceti