Substrate induction of carbohydrate catabolism in B. subtilis
is mediated primarily by transcriptional activators and antitermination (35
). Neither the putative transcriptional terminators (SL1 and SL2) nor the RAT-like sequence appears to play a significant role in regulation of fruA
in response to carbohydrate source or availability. Further supporting this idea, in a search of the S. mutans
UA159 genome database (www.genome.ou.edu/smutans.html
) we have identified a gene that we have designated smaT.
gene is predicted to encode a polypeptide of 281 amino acid residues with 44, 33, 30, and 30% identity to LicT of B. subtilis
, BglG of E. coli
, and SacY and SacT of B. subtilis
, respectively. Inactivation of this gene by allelic exchange had no significant effect on growth on fructans or expression of PfruA
fusion under all conditions tested (Wen and Burne, unpublished data). Inactivation of other transcriptional antiterminators in the genome also had no influence on fruA
expression, further confirming that the role of the 5′UTR in the fruA
mRNA is unrelated to attenuating transcription in response to carbohydrate source or availability.
Considering the fact that the 5′UTR does not have any role in regulation of fruA
, it was logical to explore whether induction of fruA
could be mediated through a target site located 5′ to the promoter. The data presented in this communication clearly show that a dyadic sequence positioned at −72 to −59 relative to the TIS is required for fruA
expression and is probably the target for a transcriptional activator. In some ways, then, the fruA
operon of S. mutans
is regulated similarly to the levanase (sacC
) operon of B. subtilis
), albeit with some important differences. First, sacC
induction requires LevR, which is a sigma 54-like regulator that is genetically linked to sacC.
No regulatory genes are linked to the fruA
operon, nor is fruA
transcribed from a −12/−24-like promoter. Also, there are no genes encoding PTS-like components in the fruA
operon, as there are in the B. subtilis
levanase operon, and there is no gene in S. mutans
that appears to be homologous to levR
by use of computer algorithms or Southern hybridization. Further, the target sequence for binding of LevR and the dyadic sequence required for fruA
show no similarity. On the other hand, the transcriptional activator LicR of the B. subtilis licBCAH
operon binds to an inverted repeat just upstream of the promoter (38
), similar to fruA.
However, there is no apparent homology between the binding site of LicR and the region upstream of the fruA
promoter (data not shown).
Inactivation of the fruI
genes resulted in constitutive elevation of expression of the PfruA
fusion. In some aspects, control of fruA
expression via the fructose PTS makes it again similar to levanase expression in B. subtilis
). Levanase induction in B. subtilis
is controlled by a phosphorelay circuit involving four gene products encoded in the levanase operon that phosphorylate LevR when fructose is absent from the growth medium. When low levels of fructose are added, a preferential transfer of phosphate to the incoming fructose leaves the PRD domain of LevR in an unphosphorylated state, which renders the protein competent for activation of sacC
transcription. Unlike for the B. subtilis
levanase operon, there are no genes in the fruA
operon for EII-like gene products that regulate transcription of the operon, yet an apparently similar phosphorelay circuit involving IIFru
proteins exists for induction and repression of fructanase in S. mutans.
Notably, growth with fructose alone does not induce fruA
expression, but we believe that this is a technical anomaly and that the genetic evidence for fructose induction through fructose EIIs provides a more credible picture of the fruA
regulatory pathway. Specifically, the reason that there is no apparent induction of fruA
in cells grown on fructose is that for the cells to grow well, it is necessary to provide relatively high concentrations of the hexose, which results in catabolite repression of fruA.
In contrast, cells growing on inulin or levan, which optimally induce fruA
expression, are exposed to much lower steady-state levels of fructose because the hexose is liberated from the fructans at a rate lower than the optimal rate for fructose transport.
Utilization of secondary carbon sources in B. subtilis
and other gram-positive bacteria is governed primarily by CcpA, which binds to CREs (1
), although CcpA-independent mechanisms have also been reported (14
). As observed with other catabolite-repressible systems, deletion or mutation of a promoter-proximal CRE, in this case CRE-S, resulted in a dramatic decrease in CCR of fruA
, indicative of the central role of this element in CCR of fruA.
Interestingly, inactivation of ccpA
in S. mutans
had little impact on CCR of fruA
(Table ), consistent with other studies that explored the role of CcpA in CCR in streptococci. For example, in S. mutans
GS-5, disruption of the ccpA
had no effect on diauxic growth when the strain was grown on a variety of nonpreferred carbohydrate sources and glucose. In fact, ccpA
inactivation resulted in increased glucose repression of α-galactosidase, mannitol-1-phosphate dehydrogenase, and phospho-β-galactosidase (34
). Similarly, sucrose-mediated repression of α-galactosidase (aga
) expression by Streptococcus pneumoniae
was not affected by mutation of a gene that encodes an apparent CcpA homologue (26
). Nevertheless, to our knowledge, this communication provides the first evidence that a putative CRE actually functions in CCR of an operon and yet a CcpA deficiency does not have an impact on expression of the operon containing that CRE. Thus, there is the distinct possibility that CcpA, which we have shown is in fact expressed in UA159, fulfills some other function in the cells and that some other repressor(s) acts at the CRE to exert CCR.
In attempting to reconcile the lack of involvement of CcpA in catabolite repression of an operon with a functional CRE, we identified an open reading frame corresponding to a product with a significant degree of similarity to the CcpB protein of B. subtilis
). The B. subtilis
CcpB protein has 30% identity to CcpA of the same organism and is responsible for CCR when cells are grown on solid medium or in liquid medium with little agitation (8
). A homologue of CcpB was identified in S. mutans
UA159, and the gene was inactivated using a strategy similar to that for ccpA
, but mutation of ccpB
alone, or of both ccpA
, had no effect on expression of the PfruA-
fusion (Wen and Burne, unpublished data). Yet another protein in B. subtilis
, CcpC, is a member of the LysR family of transcriptional regulators that mediates repression of citB
expression by glucose and sources of 2-ketoglutarate (18
). However, we could not identify a CcpC paralogue in the genome database. Therefore, if there is another repressor controlling CCR of fruA
, and possibly other catabolite-sensitive operons in S. mutans
, it is probably not a CcpA, -B, or -C homologue.
We also believe that, like for lev
, and some other catabolic operons, EI and HPr may modulate the transcription of the fruA
operon by influencing the activity of the transcriptional activator, which could explain the residual CCR in strains carrying gene fusions to the fruA
promoter with mutations or deletion of CRE-S (22
) (Fig. ) or in the ccpA
mutant of S. mutans.
Such a phenomenon, called CcpA-independent CCR, occurs in regulation of licS
, and lev
operons of B. subtilis
and the lac
operon of L. casei
). Isolation of the fruA
regulator of transcription will allow us to investigate this hypothesis in more detail.
In summary, expression of fruA in S. mutans UA159 is inducible by inulin and subject to catabolite repression. We present evidence here that expression of fruA requires a transcriptional activator that is probably negatively regulated by components of a fructose-specific PTS and that CCR of fruA occurs through a CRE, as well as through yet-undisclosed mechanisms. Thus, there are fundamental differences between substrate induction and catabolite repression of fruA of S. mutans and other genes encoding polysaccharide-degrading enzymes of eubacteria.