OUP user menu

Characterization of two operons that encode components of fructose-specific enzyme II of the sugar:phosphotransferase system of Streptococcus mutans

Zezhang T. Wen , Chris Browngardt , Robert A. Burne
DOI: http://dx.doi.org/10.1111/j.1574-6968.2001.tb10969.x 337-342 First published online: 1 December 2001


Three genes, designated as fruC, fruD and fruI, were predicted to encode polypeptides homologous to fructose-specific enzyme II (IIFru) of the phosphoenolpyruvate-dependent sugar:phosphotransferase system, and were cloned from Streptococcus mutans, the primary etiological agent of human dental caries. The fruC and fruD genes encoded domains BC and domain A of IIFru, respectively. The fruI gene encoded IICBAFru. Northern hybridization and slot blot analysis showed that expression of fruI was inducible by sucrose and fructose, while fruCD were expressed constitutively and at much lower levels. Inactivation of either fruI or fruCD alone, or of both fruCD and fruI, had no major impact on growth on fructose at a concentration of 0.5% (w/v). However, when the strains were grown with 0.2% fructose as the sole carbohydrate source, a significant decrease in the growth rate was seen with the fruCD/fruI double mutants. Assays of sugar:phosphotransferase activity showed that the fruCD/fruI double mutants had roughly 30% of the capacity of the wild-type strain to transport fructose via the phosphoenolpyruvate-dependent sugar:phosphotransferase system. Xylitol toxicity assays indicated that the inducible fructose permease was responsible for xylitol transport.

  • Streptococci
  • Fructose–phosphotransferase system
  • Xylitol toxicity
  • Dental caries

1 Introduction

Streptococcus mutans, the principal etiological agent of dental caries, is capable of transporting and fermenting a wide variety of sugars. The high-affinity phosphoenolpyruvate (PEP)-dependent sugar:phosphotransferase system (PTS) is the principal route for transport of most sugars in oral streptococci and is primarily responsible for sugar transport at low sugar concentrations [13]. The capacity to scavenge carbohydrates at low concentrations enhances survival of oral streptococci during periods between meals, and the ability to produce acids from many different sugars contributes directly to human tooth decay [4].

Because of its importance in the physiology and virulence of oral streptococci, molecular aspects of a number of PTSs have been studied in detail, including the lactose–, mannose–, sucrose–, sorbitol–, mannitol–, β-glucoside–, and glucose–PTS of Streptococcus salivarius or S. mutans. Using spontaneous mutants, Bourassa and Vadeboncoeur [5] biochemically characterized a 57.5-kDa fructose permease in S. salivarius. Gauthier et al. [6] isolated a 12-kDa domain A of an inducible fructose–PTS and also provided indirect evidence for the presence of two fructose–PTSs in S. mutans, one inducible and one constitutive [7]. However, little information is available about molecular aspects of fructose utilization by oral streptococci. As a major constituent of the human diet, fructose has been shown to have the cariogenic potential of glucose in a variety of experimental caries models [8,9]. Fructose can also be liberated from sucrose, a highly cariogenic sugar, by the glucosyltransferases and from fructans by fructanases of oral streptococci [4]. Given the importance of fructose in the human diet and its contribution to human dental caries, a more detailed understanding of the molecular mechanisms of fructose uptake by oral streptococci is warranted.

Notably, the fructose–PTS of oral streptococci, as well as a number of other bacteria, is thought to be responsible for the phosphorylation and uptake of xylitol [6,7,10]. Xylitol is a naturally occurring five-carbon sugar alcohol that has proven to be a non-cariogenic sweetening agent. Total or partial substitution of dietary sucrose with xylitol results in considerable decreases in the number of mutans streptococci in both saliva and dental plaque [11]. Previous studies by Trahan et al. [7] suggested that xylitol was transported via the constitutive fructose–PTS of S. mutans. However, the xylitol-5-phosphate thus created was not metabolizable. It was believed that the xylitol-5-phosphate may have undergone eventual dephosphorylation and perhaps was exported at the expense of ribitol-5-phosphate [12]. This so-called xylitol futile cycle is thought to consume energy and become inhibitory to cell growth, although the actual mechanism(s) for the inhibition of growth of S. mutans by xylitol has not been fully disclosed. We present here the characterization of fructose permease genes of S. mutans and an evaluation of the role of the gene products in fructose and xylitol uptake by this organism.

2 Materials and methods

2.1 Bacterial strains and cultivation

S. mutans UA159 and its derivatives were grown and maintained on brain heart infusion (BHI), supplemented with erythromycin (Em; 10 μg ml−1), tetracycline (Tc; 10 μg ml−1), or Tc and Em (10 μg ml−1, each), if necessary. Preparation of competent cells and transformations of S. mutans was carried out as previously described [13]. For growth studies and enzyme assays, S. mutans strains were grown in a tryptone–vitamin base medium (TV) [13] with various sugars added as the sole carbohydrate source(s) and, when required, antibiotics were included at the concentrations mentioned above. Escherichia coli strains were grown and maintained in Luria Bertani medium and, if needed, Em (500 μg ml−1) or Tc (10 μg ml−1) was included.

2.2 DNA manipulation, Northern hybridization and slot blot analysis

Unless otherwise stated, standard recombinant DNA procedures were used [13,14]. For quantitative slot blotting and Northern hybridizations, S. mutans UA159 and its derivatives were grown to mid-exponential phase (OD600nm≅0.3–0.4) in TV medium supplemented with the carbohydrate of interest at a concentration of 0.5% (w/v). Total RNAs were prepared [13], denatured and transferred to nylon membranes (GeneScreen Plus, NEN Life Science Products, Inc.) by using standard procedures [14]. Probes were labeled with [α-32P]dATP using the Random Primers DNA Labeling System (Life Technologies). All washes were at 65°C in 0.1× SSC with 0.1% SDS [14]. Hybridization signals were quantified using an IS1000 digital imaging system (Alpha Innotech).

2.3 Measurements of PTS activity

To measure PEP-dependent PTS activity, cultures were grown in TV medium with glucose or fructose (0.5%, w/v) as the sole carbohydrate source [13]. Cells were harvested at mid-exponential phase (OD600nm≅0.3–0.4), washed and permeabilized with toluene, and the glucose– and fructose–PTS activities were assayed using the method of Leblanc et al. [15]. PTS activity was expressed as nmol NADH oxidized in a PEP-dependent manner min−1 mg−1 of cell dry weight.

2.4 Xylitol toxicity assay

Overnight cultures that were grown in TV broth plus 0.5% (w/v) glucose were diluted 1:10 into pre-warmed TV broth with 0.2% (w/v) glucose and incubated at 37°C in a 5% CO2 atmosphere. Growth was monitored by measuring optical density of the cultures at 600 nm. When the cultures reached early exponential phase (OD600nm≅0.2–0.3), xylitol was added to a final concentration of 1% (w/v) and growth was monitored for an additional 6 h. Controls received the equivalent amount of sterile distilled water instead of the xylitol solution.

3 Results and discussion

3.1 Structure of the IIFru genes and flanking DNA

Two open reading frames (ORFs) (Fig. 1), designated as fruC (nt 113761–115161, relative position in the chromosome) and fruI (nt 818833–820800), were found to encode polypeptides homologous to IIFru using the IIFru of Bacillus subtilis in a search of the S. mutans UA159 genome database (http://www.genome.ou.edu/smutans.html). The fruC gene was 1401 bp and encoded a 467-amino acid polypeptide with 49 and 40% identity to IIBCFru of Bacillus halodurans and IIB′BCFru of E. coli, respectively. The fruI gene was 1965 bp and the deduced amino acid sequence showed 62% identity to Lactococcus lactis IIBCFru, 42% identity to B. subtilis IIABCFru and 32% identity to E. coli IIB′BCFru. A comparison of the BC domains encoded by fruI with that from fruC indicated that the deduced amino acid sequences shared 43% identity. Fourteen nucleotides downstream of fruC was a gene we designated as fruD, which was predicted to encode a polypeptide of 150 amino acid residues with 33, 31, and 25% identity to IIAFru of B. subtilis, S. aureus and E. coli, respectively. There was 18% identity between FruD and the A domain of FruI.

Figure 1

Schematic diagrams of the fruCD (panel A; accession number BK000030) and fruI (panel B; accession number BK000031) gene clusters and their flanking regions. Panel A: fruC (IIBCFru) and fruD (IIAFru) are apparently co-transcribed with fruP (FruK) and tdpA (tagatose 1,6-diphosphate aldolase). Upstream of fruK there are two other ORFs: rgl encodes a RggA-like transcriptional regulator and rpl encodes an RpiR-like transcriptional regulator. Panel B: The fruI (IIABCFru) gene is apparently co-transcribed with fruR (a DeoR-like transcriptional regulator) and fruK. Upstream of the fruI gene cluster are two ORFs: mtfA is predicted to encode a tRNA methyltransferase and trdA, a thioredoxin reductase. Arrows represent the orientation of each individual ORF and the numbers underneath indicate their respective positions in the chromosome as of June, 2001. The lollipop-like structure indicates potential transcriptional terminators.

Analysis of the flanking regions revealed several ORFs that appeared to be linked to fruCD and fruI (Fig. 1). An ORF of 909 bp immediately upstream of fruI was found to be highly similar to fructose-1-phosphate kinase (FruK) of B. subtilis and S. aureus (46 and 44% identity, respectively). The first gene in the apparent operon for fruI encoded a polypeptide with 39% identity to B. subtilis FruR, a transcriptional regulator of the DeoR family [16]. The predicted size of the transcript of this fruRKI cluster was consistent with the 3.7-kb transcript observed in Northern blots (Fig. 2). The smaller mRNA (∼2.3 kb) hybridizing with the fruI probe could be a processed transcript consisting of fruKI, but its actual identity remains to be determined. Just four nucleotides 5′ of fruC, an ORF of 930 bp that would encode another protein highly similar to FruK of B. halodurans (38% identity) and S. aureus (37% identity), was designated as fruP. FruP is 34% identical to FruK. The fruC, fruD and fruP genes appeared to be co-transcribed with another ORF, designated as fdpA, consistent with the detection of a transcript of 3.8 kb seen in Northern blots with RNA from S. mutans UA159 (Fig. 2). FdpA was most similar to LacD, a tagatose 1,6-diphosphate aldolase of L. lactis (73% identity),

Figure 2

Northern hybridization (panels A1, A2, and A3) and slot blot (panels B1 and B2) of total RNA isolated from cultures grown on 0.5% (w/v) glucose (G), fructose (F), inulin (I), or sucrose (S) as the sole carbohydrate source. Panel A1 was probed with a 16S rDNA probe; panels A2 and B1 were probed with a fruI-specific probe; and A3 and B2 were probed with a fruC-specific probe. See text for additional details.

It is interesting that S. mutans possesses two distinct gene clusters for fructose uptake, perhaps reflecting the importance of fructose utilization in the oral cavity, as well as the necessity for the organism to have a high capacity of fructose uptake and to be able to differentially regulate assimilation of this abundant dietary sugar. In S. mutans and some other bacteria, fructose can be phosphorylated at different positions, which may be accomplished by FruP and FruK. The finding that fruCD and fruI genes are immediately preceded by genes that encode proteins homologous to FruK further supports the idea that the EII enzymes encoded by fruCD and fruI could be the primary conduits for fructose transport in this organism. Similarly, the occurrence of a gene for a DNA-binding protein within the fruI operon (Fig. 1) with homology to E. coli FruR, a known regulator of fructose transport, emphasizes that FruI probably plays a role in fructose transport, and is consistent with the fact that fruI expression is up-regulated in response to growth on fructose or sucrose (Fig. 2), whereas the apparently constitutively expressed fruCD operon lacks a regulatory gene (Fig. 2).

3.2 Isolation and characterization of FruI- and FruCD-defective mutants

Several fruI and fruCD single, and fruI/fruCD double mutants, were generated by allelic exchange using an Em-resistance determinant (for fruI) or a Tc-resistance element (for fruCD) to replace portions of the coding sequences. To inactivate fruC, a 5′-fragment of 313 bp (135–427 relative to the start codon) and a 3′-fragment of 354 bp (366–720 downstream from the stop codon) were amplified by PCR reactions, and the fragments were then directionally cloned into pUT, a pUC-based plasmid with a Tc-resistance element cloned in the unique BamHI site in such a manner that the cloned fruC fragments were in the proper orientation with respect to one another and flanked the Tc marker. After sequencing to confirm that the PCR products were identical to the targets, the resulting plasmid was used to transform S. mutans UA159 and fruC mutants arising as a result of double-crossover recombination identified by Southern hybridizations. Because the 3′-fragment that was amplified contained sequence encoding the N-terminal region of FruD, the mutants that were generated had the majority of the BC domain and a large portion of domain A, replaced with the Tc-resistance element. A similar strategy was used to generate mutants with interruption of the fruI gene. Briefly, a 339-bp 5′-fragment (7–346 relative to the start codon ATG) and a 3′-fragment of 348 bp (1458–1806 relative to the start codon) generated by PCR were directionally cloned into pGEM-Em, a pGEM7-based vector, flanking an Em-resistance element capable of functioning in streptococci. Em-resistant fruI mutants, which had the entire B domain and part of domains A and C deleted and replaced with the Em-resistance element, were isolated following transformation of competent S. mutans UA159. For construction of fruCD/fruI double mutants, a fruCD mutant strain was transformed with genomic DNA isolated from a fruI mutant, and the transformation mixture was plated on BHI-TcEm plates. Mutants with inactivated fruI/fruCD genes were selected for Tc and Em resistance, and Southern blot analysis was used to confirm that gene inactivation had occurred at the expected sites.

The mutants were evaluated for the ability to grow on fructose as the sole carbohydrate source, to transport fructose by the PTS, and to grow in the presence of xylitol. When fructose was supplied in the growth medium at a concentration of 0.5% (w/v), the fruI and fruCD single mutants and fruI/fruCD double mutants grew as well as the wild-type strain. However, growth of the fruI/fruCD double mutants in TV medium with 0.2% fructose was substantially slower than the wild-type strain (96 min doubling time vs. 60 min for the wild-type). The doubling times for the fruI and fruCD single mutants were 90 and 81 min, respectively.

Consistent with the growth characteristics, both the fruCD and fruI mutants had substantial fructose–PTS activities (Table 1). Fructose–PTS activity in the fruCD/fruI double mutants was roughly 30% of that found in the wild-type. Interestingly, the fructose–PTS activity in fructose-grown fruI was increased by 77% and by 69% in the fruCD mutant, compared with those same strains grown on glucose. This observation is readily explained in the case of the mutant retaining an intact copy of the inducible fruI, but it is not immediately apparent why there is increased fructose–PTS activity in the fruI mutant, which retains only the constitutive IIFru. Preliminary Northern blot experiments indicate that there is not a greater amount of fruCD mRNA found in the fruI mutant, when compared to the wild-type organism grown under the same conditions (data not shown). Therefore, either the activity of FruCD can be regulated post-transcriptionally or other PTS enzymes that can transport fructose may be aberrantly regulated in the fruI strains. Since there is not significant up-regulation of fructose-specific PTS activity in the double mutant, it seems that modulation of FruCD activity may be the most logical explanation for our findings. Of note, spontaneously occurring xylitol-resistant strains of Streptococcus sobrinus[17], which presumably had a mutation(s) in one of the IIFru, grew on trypticase–yeast extract medium supplemented with fructose as fast or even faster than the wild-type strain on fructose. Thus, the compensation or adjustment of fructose transport activity in response to mutation of a IIFru may not be unusual in oral streptococci.

View this table:
Table 1

PTS activity of S. mutans IIFru mutants

  • Note: S. mutans UA159 and its mutant strains were grown on TV medium [13] with indicated sugar as the sole carbohydrate source. PEP-dependent fructose– and glucose–PTS activities, F–PTS and G–PTS, respectively, were measured using the method of Leblanc et al. [15] and expressed as nmol min−1 mg−1 cell dry wt. FruI and FruCD are inducible and constitutive IIFru, respectively. Data represent means (±S.D.) of no less than four separate experiments. Values within a column, with different markers (a,b,c), differ (P<0.01).

It is known that fructose can be taken up by other PTS enzymes, including the mannose– and sorbitol–PTS [1,3], although the affinities of the non-IIFru systems for fructose are lower (approx. 10 mM) when compared to IIFru (10 μM). The presence of EIIs for mannose and sorbitol have been reported in oral streptococci, including S. mutans[18,19]. In fact, ORFs with high degrees of similarity to domains of IIMan of S. salivarius are present in the UA159 genome (nt 1770344–1771333 for domains AB, nt 1771362–1772174 for domain C, and 1772336–1773162 for domain D (with 85, 58, and 78% identity, respectively)). So, a likely candidate for fructose transport in the fruI/fruCD double mutant is IIMan. This may also explain why there was no difference in the fructose–PTS activity of the double mutant in cells grown on glucose or fructose (Table 1). Since the concentration of fructose provided for growth was 0.2% (11 mM) or higher, it would not be surprising to see that the fruI/fruCD double mutants still possessed significant growth capacity, even if the Km for fructose of IIMan or of the EII for sorbitol was comparatively high.

3.3 Xylitol is transported by the inducible fructose–PTS of S. mutans

Addition of xylitol to mid-exponential phase cultures of S. mutans UA159 resulted in an almost immediate and nearly complete cessation of growth (Fig. 3). Inactivation of the fruCD gene did not alleviate sensitivity to xylitol. In contrast, the fruI mutant showed no detectable sensitivity to inhibition by the presence of xylitol as evidenced by the equivalent growth rates and yields of the fruI or fruI/fruCD double mutants in the presence or absence of xylitol in the growth medium. These findings indicate that the inducible enzyme FruI, and not the constitutive permease FruCD, is responsible for xylitol transport in S. mutans UA159. In E. coli[10], inactivation of fruA, which encodes the inducible IIFru, also confers xylitol tolerance. However, in studies with S. mutans strains, Trahan et al. [6,7] observed that xylitol-resistant mutants, which were isolated after repeated transfers of cultures in media containing 0.2% (w/v) glucose and 0.5% (w/v) xylitol, had reduced EII activity for fructose. In contrast to our findings, it was then concluded that xylitol was transported by a constitutive fructose–PTS of S. mutans. Considering that the inducible IIFru (FruI) seems to constitute almost 75% of the measured fructose–PTS activity of the wild-type strain (Table 1), it is possible that low fructose–PTS activity of xylitol-resistant mutants of S. mutans observed by Trahan et al. [7] was actually a result of mutation or aberrant regulation of the gene for the inducible IIFru of this organism.

Figure 3

Xylitol toxicity assay of S. mutans UA159 wild-type (WT) and IIFru mutants. All strains were grown on TV medium with 0.2% glucose as the sole carbohydrate source. FCD, FI, and FCD/FI are FruCD-, FruI- and FruI/FruCD-deficient mutants, respectively. Growth was monitored by measuring optical density at 600 nm. Xylitol (Xy), at 1% (w/v, final concentration), was added at the time indicated by the arrow. Controls received an equal amount of sterile water instead. See text for more detail.

In conclusion, FruCD and FruI appear to be the primary enzymes that function as fructose transporters in S. mutans and inactivation of fruI is sufficient to confer resistance to growth inhibition by xylitol. However, inactivation of either fruCD or fruI alone, or of both fruCD and fruI, did not result in complete loss of the ability of S. mutans to grow on fructose or to transport fructose via the PEP-dependent PTS. Future studies, orientated toward dissecting biochemical aspects of fructose transport, will help to determine the relative contribution of the various PTS permease to fructose transport. Also, the defined xylitol-resistant mutants generated in this study can be used in animal models of dental caries to gain a better understanding of the effects of xylitol on the development of dental caries and on the ecology of the complex populations comprising oral biofilms.


We would like to acknowledge Drs. J. Lemos and S. Bhagwat for their critical evaluation of this manuscript. We also wish to acknowledge the efforts of B.A. Roe, R.Y. Tian, H.G. Jia, Y.D. Qian, S.P. Linn, L. Song, R.E. McLaughlin, M. McShan and J. Ferretti for their immensely valuable contribution to this project through the S. mutans Genome Sequencing Project, which is supported by a grant from the National Institute of Dental and Craniofacial Research. This work was supported by NIDCR Grant number DE12236.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
View Abstract