Sequence and phylogenetic analyses of the water-soluble glucan synthesizing-glucosyltransferase genes of Streptococcus dentirousetti



Noriko Shinozaki-Kuwahara, Department of Oral Microbiology, Nihon University School of Dentistry at Matsudo, Matsudo, Chiba 271-8587, Japan. Tel: +81 47 360 9342; fax: +81 47 360 9343; email:


Two tandemly aligned glucosyltransferase (GTF) genes whose gene products are responsible for water-soluble glucan synthesis were isolated from Streptococcus dentirousetti NUM1303 and sequenced. One of the GTF genes of S. dentirousetti consisted of a 4110 bp open reading frame (ORF) that encoded for a 1369 amino acid protein and was revealed to be a S. sobrinus gtfS homolog. The percent similarity of amino acid sequences of the GTF-S from S. dentirousetti compared to those from S. sobrinus was 99%. In addition, a putative gtfT was found in tandem in the downstream region of the S. dentirousetti gtfS. The gtfT of S. dentirousetti consisted of a 4527 bp ORF encoding for 1508 amino acids. The similarity of amino acid sequences of the GTF-T from S. dentirousetti and S. sobrinus was 94%. Phylogenetic analysis based on amino acid sequences from other related streptococcal GTFs suggested that both GTF-S and GTF-T of S. dentirousetti are closely related to S. sobrinus.

List of Abbreviations

DNA Data Bank of Japan




monocolonal antibody


open reading frame


polycolonal antibody


periodic acid–Schiff base

S. criceti

Streptococcus criceti

S. dentirousetti

Streptococcus dentirousetti

S. downei

Streptococcus downei

S. mutans

Streptococcus mutans

S. orisuis

Streptococcus orisuis

S. sobrinus

Streptococcus sobrinus


water-insoluble glucan


water-soluble glucan

Glucosyltransferases (EC of mutans streptococci are responsible for synthesis of WIGs and WSGs from sucrose and are considered significant virulence factors for the initiation of dental caries [1]. Among GTF enzymes, the WIG-synthesizing GTFs, such as GTF-B and GTF-I secreted by S. mutans and S. sobrinus, respectively, have been considered some of the most important cariogenic factors in humans [2]. Recently, the nucleotide sequences of the WIG-synthesizing GTF genes of S. orisuis NUM1001, S. criceti HS6, and S. dentirousetti NUM1303, isolated from pig, hamster and fruit bat, respectively, have been analyzed [3, 4] and these GTF genes found to belong to the S. sobrinus gtfI homolog. WSGs, produced by WSG-synthesizing GTF enzymes whose products differ in both molecular weight and the degree of branching of their glucosyl residues, are also important because they are essential for the primer-dependent WIG-synthesizing GTFs that start WIG synthesis [5]. It is known that S. mutans possess a WSG-synthesizing GTF, GTF-D, and that S. sobrinus possess three types of WSG-synthesizing GTFs, GTF-S, -T, and -U, which are encoded by gtfD, gtfS, gtfT, and gtfU, respectively [6]. WIG- and WSG-synthesizing GTF enzymes interact with each other to facilitate production of adhesive WIGs, which facilitate colonization of tooth surfaces by oral bacteria [7, 8]. The presence in S. criceti HS6 of tandemly aligned WSG-synthesizing GTF genes, designated gtfS and gtfT, has been confirmed [5] and PCR sequencing suggested there could be a portion of gtfT in S. dentirousetti [4]. The purpose of this study was to determine the WSG-synthesizing GTF gene sequences in S. dentirousetti and to elucidate the antigenic properties of this organism's WSG-synthesizing GTF gene products. Further, we attempted construction of a phylogenetic tree based on deduced amino acid sequences of other oral streptococcal GTFs to investigate the origin of cariogenic factor in human.

Chromosomal DNA of S. dentirousetti NUM1303 [9] was extracted and purified as previously described [10]. The sequences of PCR primers used in this study are listed in Table 1. The target region of the gene was amplified using KOD FX (Toyobo, Tokyo, Japan), the amplification conditions being as described previously [4]. Determination of the nucleotide sequences was also as described previously [4], except that ABI PRISM 3130 was used instead of ABI PRISM 310 (Life Technologies, Carlsbad, CA, USA). Sequence data were analyzed by DNASIS sequence analysis software (Hitachi Software Engineering, Tokyo, Japan) and DDBJ database searches. Amino acid sequences of GTF were aligned with the CLUSTALW software program [11]. To detect the part of the GTF gene(s) responsible for the WSG-synthesizing enzyme, S. mutans gtfD or S. sobrinus gtfT, PCR was carried out previously described [12]. The resulting 326 bp fragment was amplified using a primer set, MKT-F and MKT-R. Sequencing results of this fragment from S. dentirousetti revealed 100% homology to the corresponding region of the S. sobrinus gtfT. When genomic Southern blot analysis was carried out using a portion of gtfD from S. mutans as probe, no positive band was detected (data not shown). To clone the entire coding region of the S. dentirousetti gtfT-homolog, inverse PCR was employed utilizing either HindIII- or PstI-digested chromosome as a template, and a set of oligo-nucleotides, gtfT-316f and gtfT-298r, as primers. As a result, 2.3 and 6.1 kbp fragments were amplified from S. dentirousetti (Fig. 1, lines a and b, respectively) and sequenced. It was found that the 2.3 kb fragment was similar to a part of the gtfT-like ORF (line a), whereas the 6.1 kb fragment contained 1.4 kb of the 3′-region of gtfS-like ORF and the 4.7 kb of gtfT-like ORF (line b). Therefore, two additional primer sets, gtfS-up and gtfS-3395R, and gtfT-4145F and gtfT-R2B, were designed for direct PCR. As a result, 3.4 and 0.7 kb fragments were amplified and sequenced (lines c and d). Utilizing these fragments, two entire WSG-synthesizing GTF gene sequences were determined in tandem. The nucleotide sequences of the S. dentirousetti gtfS-like and gtfT-like genes are composed of 4110 and 4527 nucleotides encoding for 1369 and 1508 amino acids with molecular masses of ≈152 and 166 kDa, respectively. Since the N-terminal region of these GTFs contains a signal peptide, the molecular weights of the mature GTF proteins are ≈148 and 162 kDa, respectively. These GTFs contain the characteristic amino acid sequences, such as the putative active-site (GTF-S; DGVRDAVD or GTF-T; DGIRDAVD), and direct repeating units responsible for glucan binding near the C-terminus [13, 14]. The deduced amino acid sequences show the highest similarity to the S. sobrinus GTF-S (99%) and S. sobrinus GTF-T (94%), respectively. The percent homology of the GTF-S and GTF-T amino acid sequences from S. dentirousetti and S. criceti were 78% and 85%, respectively. These analyses show that two gtfs of S. dentirousetti are homologs of gtfS and gtfT, respectively. The DDBJ accession numbers for gtfS and gtfT of S. dentirousetti in this paper are AB685253 and AB685254, respectively.

Table 1. Primers used in this study
  • , nucleotide sequence from GenBank accession no. M30943 (S. downei gtfS).
  • , nucleotide sequence from GenBank accession no. AB685253 (S. dentirousetti gtfS).
  • §, nucleotide sequence from GenBank accession no. AB685254 (S. dentirousetti gtfT).
  • , nucleotide sequence from GenBank accession no. D13928 (S. sobrinus gtfT).
gtfS-up5′-CATGTTAAAATAAATTGATA-3′164–183This study
gtfS-659R5′-CTGGAAGCCTCATTGCTATAGG-3′659–638This study
gtfS-683F5′-TGACTGCCGATTCTTGGTATCG-3′683–704This study
gtfS-1387F5′-ACGAAAAGAATGCCATTGAC-3′1388–1407This study
gtfS-1779F5′-CGCAACCTACCTCAATGATAG-3′2199–2219This study
gtfS-3084R5′'-AGTAGCACCTGAACGTGTTCTACC-3′2835–2812This study
gtfS-3585F5′'-CAAGGAGGGCATTTTGTACGATG-3′3336–3358This study
gtfS-3395R5′'-GTCCGGCCATTATTGTCATAG-3′3395–3375This study
gtfS-4094R5′-GAAGCTAGACCATCTCGGTTA-3′4094–4074This study
gtfT-289r5′-GAACTTGGCTGCTTTCCGAAG-3′289–269§This study
gtfT-316f5′-CAAGCAAGCGACCTTGCTCCA-3′316–336§This study
gtfT-2239R5′-CATCATCGGTGAAGAGGTCACCATAGT-3′1930–1904§This study
gtfT-2459F5′-CTGACCAAGACATTCGTGTGG-3′2459–2479§This study
gtfT-3477F5′-CTCAGTGATGAAGGTACTGG-3′3169–3188§This study
gtfT-4145F5′-AACCGGTGTCCAAGTTAAAGG-3′3837–3857§This study
gtfT-R2B5′-GAGAGTTAATGTTGCT-3′4743–4728This study
Figure 1.

Chromosomal structure of S. dentirousetti gtfS and gtfT. The shaded bar represents the 326 bp region amplified by MKT-F and MKT-R primers. E, EcoRI; H, HindIII; K, KpnI; P, PstI. Solid lines indicate PCR products for determination of nucleotide sequences of S. dentirousetti gtfS and gtfT. Lines (a and b) HindIII-digested and PstI-digested inverse PCR products, respectively; (c and d) direct PCR products. Arrowheads denote the primers used for PCR in this study. For details, see the text.

To investigate the enzymatic activity and immunoreactivitiy of the GTF-S and GTF-T in S. dentirousetti, following SDS–PAGE a crude GTFs preparation from M4 [15] was analyzed using PAS and immuno stains, respectively [3, 16]. It has previously been reported that the molecular weights of S. sobrinus GTF-S and GTF-T are 150 and 158 kDa, respectively [17, 18]. As shown in Figure 2a, WSGs synthesized by GTF-S and GTF-T from S. dentirousettei (lane 1) were visualized by PAS stain at approximately the corresponding region of S. sobrinus (lane 2) and were coincident with that of the deduced mature proteins described above. The immunoreactivity of the GTFs was assayed by western blot analysis using anti-GTF-T MAb and anti-rGTF-S PAb that had been prepared previously [17, 18]. Figure 2b, c shows that GTF-S and GTF-T of S. dentirousetti were detected at the corresponding regions of those of S. sobrinus, respectively. These results indicate that the secretion level of S. dentirousetti GTF-S is low, whereas that of GTF-T is almost the same as that of S. sobrinus. The prepared protein sample of S. dentirousetti was recognized with anti-GTF-U MAb [17], suggesting that the gtfU might be maintained in S. dentirousetti (data not shown). These findings indicate that the WSG-synthesizing GTFs, GTF-S, GTF-T and GTF-U in S. dentirousetti, have similar functions to those of S. sobrinus WSG-synthesizing GTFs, namely to supply primers for the GTF-I enzyme in these bacteria.

Figure 2.

WSG-synthesizing GTF activity by (a) PAS staining and (b)western blot analyses reacted with anti GTF-rS PAb and (c) anti-GTF-T MAb of crude GTF preparations following SDS-PAGE. Lane 1, S. sobrinus 6715; lane 2, S. dentirousetti NUM1303. Molecular size markers are shown on the left in panel A.

Tree topology and evolutionary distances were calculated [19], and phylogenetic trees constructed by the neighbor-joining method using MEGA software package, version 4.0 [20]. To study the relationship among GTFs, a phylogenetic tree was constructed based on the amino acid sequences of 32 GTFs, including both the WIG- and WSG-synthesizing GTF enzymes from oral streptococci. The nucleotide sequence of S. sobrinus gtfS was also determined (accession number AB740265), because among the WSG-synthesizing GTFs from S. sobrinus, only the nucleotide sequence of the GTF-S enzyme had previously been reported [16, 18]. The sequences of two dextransucrases, DSRA and DSRB, from Leuconostoc mesenteroides were supplemented into this tree. As expected, GTF-S and GTF-T were closely related to those of S. sobrinus and positioned in the same cluster as those of S. sobrinus, S. downei and S. criceti (Fig. 3). Recently, Hoshino et al. reported that streptococcal GTFs can be classified into three clusters: WIG-synthesizing, WSG-synthesizing and intermediate groups [21]. The GTF-based tree constructed in this study is similar even when GTF sequences of animal origin streptococci are included. As shown in Figure 3, the GTF-S group diverged at a relatively early branch and independent evolutional processes might have occurred in all GTFs. It is likely that oligosaccharides produced by GTF-S function as primer molecules for fully activating primer-dependent GTFs such as GTF-I [22].

Figure 3.

Phylogenetic tree constructed from 32 amino acid sequences of oral streptococcal GTFs. Two dextransucrases of Leuconostoc mesenteroides were used as an outgroup. The lengths of the connecting lines indicate relative phylogenetic distances. Bar, 0.1 substitutions per amino acid position. The numbers in parentheses represent the accession numbers of data bases.

According to the GTF-based phylogenetic tree, it seems that S. mutans GTF-B and GTF-C appeared ahead of S. sobrinus GTF-I. Because our preliminary experiment using a previously described method [23] showed that S. mutans is far more capable of sucrose-dependent firm colonization than is S. sobrinus, it is possible that consumption of large amounts of sucrose facilitates adherence of S. mutans rather than S. sobrinus within human oral cavities. Hoshino et al. also reported that, in humans, consumption of refined sugar completed acquisition of cariogenicity of gtf [21]. These findings may explain why S. mutans is the predominant species of cariogenic streptococci and can be isolated from most human oral cavities, whereas S. sobrinus is detected in only 20% of human oral cavities. Because animals do not have opportunities to consume sucrose, in their oral cavities S. sobrinus-type GTF genes might dominate over mutans streptococci such as S. dentirousetti, S. orisuis and S. criceti. It is possible that gene structures can change during transfer between various oral cavities. Indeed, in the latter strains, the numbers of repeating units in glucan-binding domains of gtfI differ from those of S. sobrinus strains [3, 4]. Phylogenetic analysis based on the 16S rRNA gene has reportedly shown strong sequence similarity with S. downei [9]. The fact that 16S rRNA is indispensable, whereas GTF genes are not essential in bacteria, suggests that their evolutionary processes differed among streptococcal species. To identify the origin of cariogenic factors of streptococcal species in humans, we have isolated streptococci from the oral cavities of various animals. Analyses of their GTF genes may clarify the process by which cariogenic factors develop at a level beyond species or genera.


This study was supported in part by a Grant-in-Aid for MEXT-Supported Program for the Strategic Research Foundation at Private Universities, 2008–2012 (S0801032) and a Grant-in-Aid for Young Researchers of Nihon University School of Dentistry at Matsudo 2012.


The authors declare that they have no conflicts of interest.