Isolation and nucleotide and amino acid sequence analysis of the putative L. reuteri strain ATCC 55730 glucosyltransferase gene/enzyme.
Degenerate primers were designed and used for PCR with chromosomal DNA of L. reuteri
strain ATCC 55730 as a template (Fig. ) (12
). The method used resulted in identification of a single gtf
gene in LB BIO (gtfO
Using iPCR, a total of 9,419 bp of LB BIO DNA was cloned and sequenced, revealing the presence of three complete open reading frames (ORFs) and two partial ORFs (including a putative transposase) on the compiled sequence (Fig. ; Table ). The presence of transposases near (putative) gtf
genes has been previously observed in different lactobacilli (10
). The gtfO
gene encodes a putative protein of 1,781 amino acids, with a deduced molecular mass of 197,170 Da and a pI of 5.49.
Overview of the highest identity and similarity scores of GTFO and surrounding ORFs of Lactobacillus reuteri ATCC 55730
Amino acid sequence alignments of L. reuteri
ATCC 55730 GTFO with other glucosyltransferases from Streptococcus
, and Lactobacillus
subsp. using BLAST (1
) revealed clear similarities. Highest similarity at the amino acid level was found with reuteransucrase of L. reuteri
121 (GTFA), which synthesizes a reuteran with α-(1→4) and α-(1→6) linkages (11
) (Table ). The putative protein structure of GTFO was very similar to that of GTFA, GTF180, and GTFML1, containing (i) an N-terminal signal sequence of 38 amino acids, (ii) a variable N-terminal domain of 705 amino acids, (iii) a catalytic domain of 771 amino acids, and (iv) a C-terminal domain of 267 amino acids (10
The deduced N-terminal amino acid sequence of GTFO contains a putative secretion peptide with a predicted signal peptidase cleavage site (SPase) between amino acid 38 and 39 (http://www.cbs.dtu.dk/services/SignalP/
). Within the deduced N-terminal variable region of GTFO, a series of five RDV repeats, R
-R(Y/F)S (x, nonconserved amino acid residue) were found, as previously observed only in GTFA, GTF180, and GTFML1 from other L. reuteri
The C-terminal domain (glucan binding domain) of GTFO contains four YG-repeating units, NDGYYFxxxGxxH°x(G/N)H°H°H° (x, nonconserved amino acid residue; H°, hydrophobic amino acid residue) according to the definition in reference 9
, and seven YG-repeating units which are less conserved. In the deduced amino acid sequence of the catalytic domain of GTFO, three (putative) catalytic residues were identified: Asp1024
, and Asp1133
In the region downstream of the (putative) catalytic nucleophile Asp1024
DAPDNI, GTFO differs in two out of five amino acids conserved in virtually all studied GTF enzymes from Streptococcus
, and Lactobacillus
). In GTFO, Pro1026
is found in a position where a conserved Val is present in all other glucansucrases (except GTFA of L. reuteri
121, which also possesses a Pro residue at this position). Another conserved amino acid substitution in this region of GTFO, Ile1029
(instead of Val), was also found in amylosucrase, a glucosyltransferase from Neisseria polysaccharea
synthesizing an α-(1→4) glucan (7
), CD2 of DSRE of L. mesenteroides
NRRLB-1299, responsible for the synthesis of α-(1→2) linkages (4
), and GTFA of L. reuteri
). This region in GTFO is identical to that in GTFA of L. reuteri
121, which also synthesizes a reuteran; this region thus may be responsible for or contribute to the α-(1→4) bond specificity in both enzymes.
The region following the putative acid base catalyst E1061 (putative acceptor substrate-binding region) is not highly conserved in glucansucrases (19
). Immediately following the motif 1061
E(A/D)W(S/N), two serine residues are found in GTFO, whereas GTFA of L. reuteri
121 possesses a His and an Ala residue at the corresponding positions, respectively. We speculate that differences in this (putative) acceptor substrate binding region between GTFO and GTFA may at least partly explain the larger amount of α-(1→4) bonds synthesized by GTFO in its products (see below) and/or its higher hydrolytic activity (see below).
The region following D1133 (transition state stabilizer) in GTFO differs from the sequence 1133
DSEVQTVI, conserved in many glucansucrases from Streptococcus
, and Lactobacillus
). In GTFO as well as in GTFA, an original tripeptide, NNS is found immediately downstream of this catalytic Asp. Also, GTF180 and GTFML1 both contain an original tripeptide at this position, SNA and NGS, respectively. Finally, also alternansucrase and CD2 of DSRE of L. mesenteroides
NRRLB-1299 also both contain an original tripeptide at this position, YDA and KGV, respectively (4
). Mutational evidence is available that the region following this catalytic Asp1133
is important in glucan structure determination in GTF enzymes (18
). The presence of an original tripeptide, NNS, in both GTFA and GTFO may at least partly explain the unique structure of the glucans synthesized by both enzymes containing high amounts of α-(1→4) linkages.
Purification, pH, and temperature optima. L. reuteri
ATCC 55730 GTFO(-ΔN) enzymes expressed in E. coli
were purified to homogeneity. The predicted Mr
's of the N-terminal deletion mutant (117 kDa, without signal sequence and N-terminal variable region) and full-length protein (194 kDa, without signal sequence) were in agreement with the results obtained by sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis (data not shown). In order to define the best conditions for subsequent kinetic studies, the pH and temperature optima of the full-length GTFO enzyme were examined. The pH optima for both the hydrolyzing and transferase activities were at pH 5.0 (data not shown). The temperature optima for both reactions were 35°C (data not shown). The N-terminal deletion mutant (GTFO-ΔN) showed the same pH and temperature optima (data not shown). The pH optimum of GTFO(-ΔN) was similar to that of GTFA of L. reuteri
121 (pH 4.7). However, the temperature optimum of GTFO(-ΔN) was 15°C lower than that of GTFA (50°C) (13
Kinetic studies with GTFO(-ΔN). (i) Kinetic parameters.
In the presence of sucrose, GTFO displayed a Michaelis-Menten type of kinetics for hydrolysis (VG
) and for total enzyme activity (VF
) (Table ). Transferase activity (initial rates; VF
) was observed only at sucrose concentrations above 10 mM. GTFO thus favors hydrolysis at low sucrose concentrations and polymerization at higher sucrose concentrations. GTFA of L. reuteri
121 clearly displayed transferase activity (initial rates) at sucrose concentrations below 10 mM (13
). GTFO-ΔN showed similar kinetics as full-length GTFO (data not shown). Affinity for the substrate sucrose in the hydrolysis reaction was similar for both GTFO(-ΔN) enzymes. Deletion of the N-terminal variable domain in GTFA resulted in drastic changes in enzyme activity, especially resulting in a strongly increased transferase activity and an approximately two times lower turnover rate for the hydrolysis reaction (13
). Deletion of the N-terminal variable domain in GTFO did not affect its transferase activity. However, the GTFO-ΔN hydrolysis reaction turnover rates (kcat
) were about 1.5 times higher than for full-length GTFO (data not shown).
Comparison of the kinetic properties and product spectra of GTFO of Lactobacillus reuteri ATCC 55730 and GTFA of Lactobacillus reuteri strain 121
(ii) Effect of maltose on initial GTFO activity.
The presence of maltose strongly stimulated GTFO total activity (initial rates) (sevenfold increase to VF
of 95.4 ± 2.2 U · mg−1
). A similar observation was made for GTFA (3.5-fold increase to VF
of 97.7 ± 0 · 9 U · mg−1
Product analysis. (i) Product spectrum from sucrose.
Compared to GTFA, GTFO converted less sucrose into glucan but produced larger amounts of isomaltose and leucrose. GTFO action on sucrose resulted in considerably more glucose release from sucrose (>54% of sucrose; Table ) than that for GTFA (22.7% of sucrose; Table ). Two unknown GTFO products eluted after 36 and 54 min, respectively (data not shown). GTFO-ΔN showed a product spectrum similar to that of GTFO (data not shown).
(ii) Enzymatic analysis of the structures of oligosaccharides synthesized from maltose and isomaltose as acceptor substrates.
The major oligosaccharide products (DP3) synthesized by both GTFA and GTFO from sucrose in the presence of maltose as an acceptor reaction substrate eluted at the same position (43.5 min; Dionex analysis) (Fig. ) as panose (α-d
). This oligosaccharide was purified, and analysis of products formed upon its enzymatic degradation in time, by amyloglucosidase and alpha-glucosidase, confirmed its identity as panose (Table ) (13
FIG. 2. Anion exchange chromatography of GTFO and GTFA (data from reference 13) acceptor reaction products. Products formed upon incubation of 30 nM GTFA (A) and 30 nM GTFO (B) enzyme with 100 mM sucrose and 100 mM maltose for 60 h. Products formed upon incubation (more ...)
Identification of major oligosaccharides formed by GTFO from sucrose and different acceptor substrates, deduced from enzymatic degradation productse
The major oligosaccharide products (DP3) synthesized by both GTFA and GTFO from sucrose in the presence of isomaltose as an acceptor reaction substrate eluted after 34 min (Fig. ) (13
). Its identity (α-d
-glucose; isopanose) was deduced from the products formed upon its degradation to isomaltose and glucose by amyloglucosidase and alpha-glucosidase (Table ). Besides isopanose, GTFO synthesized small amounts of a DP4 oligosaccharide, eluting after 46 min. GTFA synthesized larger amounts of this DP4 oligosaccharide (Fig. ) (13
). Therefore, the DP4 oligosaccharide synthesized by GTFA was used for enzymatic analysis of its structure. It was degraded slowly by amyloglucosidase, whereas degradation by alpha-glucosidase was poor (Table ). Isopanose was one of the degradation products from amyloglucosidase; this is taken to suggest that isopanose formed by GTFA and GTFO is used as an acceptor reaction substrate to form α-(1→6)-isopanose (α-d
-glucose). Degradation of this DP4 oligosaccharide by dextranase was not possible, whereas another DP4 oligosaccharide with an α-(1→6) linkage in the middle was cleaved by dextranase (data not shown).
(iii) Comparison of the amounts of oligosaccharides synthesized by GTFA and GTFO.
Maltose is an equally good acceptor reaction substrate for oligosaccharide synthesis by both GTFO and GTFA (Table ). In the presence of sucrose and maltose, GTFA and GTFO formed panose as the most abundant acceptor reaction product (see above, approximately 44 mM and 33 mM, respectively) (Fig. ). Compared to GTFA, GTFO synthesized larger amounts of maltotriose (5 mM and 17 mM, respectively) (Fig. ).
Isomaltose is a two-times-better acceptor reaction substrate for oligosaccharide synthesis by GTFO than for synthesis by GTFA (Table ). In the presence of sucrose and isomaltose, GTFO formed small amounts of isomaltotriose and isomaltotetraose, plus isopanose (Fig. ; see above). GTFO synthesized approximately 2.5 times more isopanose than GTFA (Fig. ). Besides isopanose, GTFO also synthesized minor amounts of α-(1→6)-isopanose. GTFA synthesized significantly larger amounts of α-(1→6)-isopanose (Fig. ) (13
GTFO-ΔN displayed a similar product distribution with maltose and isomaltose as acceptor reaction substrates (data not shown).
Compared to GTFA, GTFO thus introduced more α-(1→4) linkages in its acceptor substrates, yielding larger amounts of maltotriose from maltose and larger amounts of isopanose from isomaltose and only minor amounts of α-(1→6)-isopanose (Fig. ). The ability of GTFO to synthesize an α-glucan with large amounts of α-(1→4) linkages (see below) thus also is reflected in its oligosaccharide product spectrum.
Analysis of the glucans produced by L. reuteri ATCC 55730 culture supernatants and purified recombinant GTFO(-ΔN) enzymes.
Purified recombinant GTFO(-ΔN) enzymes and supernatants of sucrose grown cultures of L. reuteri
ATCC 55730, incubated with sucrose, produced high-molecular-weight glucans. Using high-performance size exclusion chromatography-multiangle laser light scattering, the average molecular weights of the glucans produced by LB BIO and by the purified GTFO(-ΔN) enzymes were determined (Table ). The identical nature of these glucans was confirmed by methylation analysis (Table ). The 1
H-NMR spectra of the glucans produced by LB BIO culture supernatants and by the purified recombinant GTFO-ΔN enzyme were virtually identical (Table ; Fig. ). Comparison of both 1
H-NMR spectra with that of the reuteran produced by the L. reuteri
121 GTFA enzyme (Table ; Fig. ) (11
) showed that both glucans consist of α-(1→4) (~5.3 ppm) and α-(1→6) (~5.0 ppm)-linked glucopyranosyl units. Due to poor resolution of the NMR spectra, it was not possible to trace the terminal and α-(1→4,6)-linked residues present (as indicated by the methylation analysis) (Table ).
TABLE 5. Methylation analysis, NMR analysis (see also Fig. ), and masses of the glucans produced by culture supernatants of Lactobacillus reuteri strain ATCC 55730 under the conditions tested and the purified recombinant GTFO(-ΔN) and GTFA (more ...)
FIG. 3. 600-MHz 1H-NMR spectra of glucans produced by L. reuteri strain ATCC 55730 culture supernatants (A), by purified recombinant GTFO-ΔN protein (B), and by purified recombinant GTFA protein of L. reuteri strain 121 (C) (13) recorded in D2O at 80°C. (more ...)
NMR (Table ; Fig. ) and methylation analysis (Table ) clearly showed that GTFO and GTFA synthesize different reuteran products. GTFO reuteran contains approximately 20% more α-(1→4) glucosidic linkages [and 20% fewer α-(1→6) glucosidic linkages]. The degrees of branching of the GTFO- and GTFA-synthesized reuterans [amount of both terminal and α-(1→4,6) glucosyl units] are rather similar (see methylation analysis, Table ). The precise structures of both reuterans remain to be elucidated.
LB BIO is able to colonize the human gastrointestinal tract (29
), but a possible role for GTFO and its glucan product remains to be elucidated. The concentrations of sucrose needed to yield significant glucan polymer production may normally not be achieved in the gut. Glucans produced by streptococci play a key role in the development of human dental caries, enhancing the attachment and colonization of cariogenic bacteria to teeth (16
). Recently, different lactobacilli have been identified in carious dentine (5
). LB BIO may also be able to colonize the oral cavity, in which case GTFO may contribute to polymer formation and colonization on oral surfaces.
This paper reports the molecular and biochemical characterization of the novel L. reuteri
ATCC 55730 reuteransucrase gene (gtfO
) and the novel reuteransucrase enzyme (GTFO) encoded. The GTFO, GTFA, GTF180, and GTFML1 structures and amino acid sequences are highly similar (10
) (~60% identity, ~75% similarity), but they synthesize glucans with different glucosidic linkages [mainly α-(1→4), α-(1→4)/α-(1→6), mainly α-(1→6), and mainly α-(1→3) linkages, respectively]. These enzymes thus are very interesting for structure/function studies aiming to identify amino acid residues responsible for glucosidic bond specificity. Furthermore, GTFA and GTFO are interesting candidates for examination of structural differences responsible for hydrolysis/transferase activity ratios.