Search tips
Search criteria 


Logo of jbacterPermissionsJournals.ASM.orgJournalJB ArticleJournal InfoAuthorsReviewers
J Bacteriol. 2012 March; 194(5): 1127–1135.
PMCID: PMC3294772

Oligomerization of the Response Regulator ComE from Streptococcus mutans Is Affected by Phosphorylation


We have previously characterized the interactions of the response regulator ComE from Streptococcus mutans and DNA binding sites through DNase I footprinting and electrophoretic mobility shift assay analysis. Since response regulator functions are often affected by their phosphorylation state, we investigated how phosphorylation affects the biochemical function of ComE. Unlike many response regulators, we found that the phosphorylation state of ComE does not likely play a role in DNA binding affinity but rather seems to induce the formation of an oligomeric form of the protein. The role of this oligomerization state for ComE function is discussed.


One of the ways for bacteria to monitor external conditions and adjust their structure, physiology, and behavior is through the expression or repression of genes by a network of environmental sensors and response regulators collectively known as two-component signal transduction systems (TCSTS) (28). The prototypical TCSTS has two protein components: a sensor histidine kinase (HK), often located in the membrane, which monitors an environmental parameter(s), and a cytoplasmic response regulator (RR), which mediates changes in gene expression in response to specific signals (27, 31). These TCSTS are associated with a variety of domains that function as “input” and “output” elements (28). A typical HK contains an N-terminal, membrane-associated sensor domain and a C-terminal phosphotransferase domain, made up of a cytosolic H-box and an ATPase domain. Upon detecting a specific environmental stimulus (pH, limiting or excess nutrient, peptide, temperature, osmolarity, etc.), the ATPase domain mediates autophosphorylation of the HK at a conserved histidine residue in the H-box (36). This phosphoryl group is subsequently transferred to an aspartic acid residue of the RR receiver domain, leading to activation of the RR. A typical RR consists of an N-terminal receiver domain and a C-terminal effector domain. Once activated, the RR then binds to specific regions on the DNA, which leads to the activation/repression of genes involved in the adaptive response (19). In addition, the ability of these HKs to be faithful to their cognate RR is still an open question since there is evidence of cross talk in vitro between noncognate HKs and RRs (34). For example a Bacillus subtilis RR has been shown to be sufficiently phosphorylated when expressed and purified from Escherichia coli by either acetyl-P or another HK (17). Finally, there are even some HKs that respond to multiple signals, like Salmonella enterica serovar Typhimurium PhoQ, which can sense and respond to both magnesium and pH (1), thus adding further complexity to understanding TCSTS.

Currently, there are 14 known TCSTS in Streptococcus mutans (2). We are particularly interested in the ComED system involved in the development of competence for natural transformation, bacteriocin production, biofilm formation, and acid tolerance in S. mutans (20, 22). Competence is a physiological state in which bacteria are able to take up and integrate exogenous free DNA from their environment, which enables the recipient organism to acquire novel genes such as those encoding antibiotic resistance and other virulence factors (6, 13, 14, 16, 20, 21, 38). Therefore, natural genetic transformation can be an important mechanism to allow bacteria to adapt to a changing environment (6, 16).

In our previous work, we presented quantitative DNA binding studies to show that ComE binds to two imperfect direct repeats within the intergenic region of nlmC-comC (10). We further characterized the binding of ComE to other related sites and identified a ComE consensus binding site. We extend this work here with mutational and cross-linking analyses to characterize the phosphorylation state of ComE in S. mutans for the first time. We show that phosphorylation of ComE has little effect on DNA binding but rather strongly promotes oligomer formation. A model for ComE regulation is proposed and discussed.


Construction of D60A and D60E mutations in ComE.

To generate two separate point mutations in ComE, comE was amplified from S. mutans UA159 chromosomal DNA using ComE-F and ComE-R (Table 1) and cloned into pCR2.1-Topo (Invitrogen, Carlsbad, CA). To modify the putative conserved aspartic acid residue (Asp60) required for phosphorylation, primer ComE-DA (Table 1) changes the GAT (Asp) triplet to GCT (Ala) whereas primer ComE-DE introduces GAG (Glu). Inverse PCR was performed with the Elongase enzyme (Invitrogen) to amplify the whole plasmid with the specific mutagenesis primers (ComE-DA or ComE-DE) and the inverse primer ComE-inv. The primers were then phosphorylated for subsequent ligation of the linear plasmid carrying the mutagenized comE using standard methods (32). PCR products were purified, ligated, transformed into E. coli DH5α, and selected with ampicillin, and the introduction of the mutation was confirmed by sequencing. Subsequently, comE D60A and D60E fragments were cloned into the BamHI and HindIII sites of expression vector pQE30 (Qiagen) and transformed into E. coli DH5α to generate SG482 and SG500, ComE:DE and ComE:DA, respectively, as done previously with wild-type ComE, SG481 (10). Overexpression and purification of these proteins were performed as described previously for wild-type ComE (13).

Table 1
Primers used for DNase I footprinting and mutagenesis of ComE

Construction of ComE in an E. coli strain lacking pta-ackA background.

AJW2013, a gift from Alan Wolfe (University of Illinois at Chicago), is an E. coli strain lacking pta and ackA, resulting in no internal acetyl-P production (11, 42). To aid in the expression of ComE, AJW2013 was transformed with a repressor plasmid, pREP4 (Qiagen), which confers a high level of lac repression in trans (SG611, Table 2). Next, pQE30+comE was transformed into SG611 and ComE612 was subsequently purified from this mutant strain (SG612) as described previously (13).

Table 2
Bacterial strains used in this study

In vitro phosphorylation of ComE by phosphoramidate (PA) and cross-linking by dimethyl suberimidate (DMS).

To phosphorylate ComE, 11.6 μM ComE was incubated with 62.5 mM PA (gift of Linda Kenney, University of Illinois at Chicago) in phosphorylation buffer (50 mM Tris [pH 7.5], 20 mM MgCl2, 50 mM KCl) in a total volume of 20 μl for 2 h at room temperature. For the cross-linking reaction, 8.7 μM phosphorylated ComE was incubated with 25 mM DMS in 50 mM boric acid (pH 10) (Pierce Biotechnologies Inc.) in a total volume of 10 μl for an additional 2 h at room temperature. Cross-linking reactions were mixed with an equal volume of 2× sample buffer (0.005% bromophenol brilliant blue, 4% SDS, 125 mM Tris, 20% glycerol) and separated by electrophoresis on a 4 to 20% Tris-glycine gel (Invitrogen). Gels were stained with Coomassie blue to observe ComE in its unreacted and oligomerized state.

To verify the identity of ComE oligomers, Ni-NTA-Atto conjugates (Sigma), which specifically detect polyhistidine-tagged proteins, were used according to the manufacturer's protocol. Briefly, gels were fixed in 40% ethanol-10% acetic acid for 1 h, washed twice with water and incubated with 1:1,000 Ni-NTA-Atto in PBST (136 mM NaCl, 2.6 mM KCl, 10 mM Na2HPO4, 1.5 mM KH2PO4, 0.2% Tween 20) in the dark overnight. Finally, gels were washed in water for 2 h in the dark and scanned with the PharosFX Molecular Imager system (Bio-Rad).

DNase I footprinting assay.

Footprinting assays were done as described previously (10). Briefly, labeled substrate of comC (oSG316-oSG317) was incubated with ComE in footprinting buffer at room temperature for 30 min. After incubation, DNase I was added for an additional minute before the reaction was quenched with DNase I stop buffer. The DNA was extracted with phenol-chloroform, ethanol precipitated, and resuspended in sequencing stop buffer. Reactions were separated on a 6% sequencing gel and run at 40 V/cm for approximately 3 h. The gel was dried and scanned with the PharosFX Molecular Imager system.


Electrophoretic mobility shift assays (EMSAs) were done as previously described (13). Briefly, 15 nM ComE was incubated at room temperature for 30 min in reaction buffer (52.5 mM HEPES [pH 6.5], 50 μM EDTA, 9.5% glycerol, and 50 μg/ml bovine serum albumin [BSA]), 100 ng salmon sperm DNA, and 1 nM isotopically labeled DNA substrate in a final volume of 20 μl. Following incubation, EMSA reactions were analyzed on a 6% nondenaturing polyacrylamide gel, run at 10 V/cm for 3 h, subsequently dried, and visualized with the PharosFX Molecular Imager system.


Quantitative binding analysis and footprinting of mutant proteins ComE:DE and ComE:DA.

Response regulators are often phosphorylated at a conserved aspartate residue, e.g., CheB and OmpR are phosphorylated at D56 and D55, respectively (18, 35). Using the SIM alignment tool for protein sequences (, we were able to align D56 of CheB and D55 of OmpR with D60 of ComE (data not shown). Based on this alignment, we constructed two point mutations where D60 was changed to either glutamic acid or alanine. The change from aspartic acid to glutamic acid extends the negative charge of the side chain and has previously been shown to mimic the phosphorylated state of aspartate (12). The change from aspartic acid to alanine removes the negative charge of the side chain preventing ComE phosphorylation. To find out if ComE:DE and ComE:DA bind with different affinities than ComE, the equilibrium dissociation constant of the two ComE mutants was determined, as described previously for ComE (10). Briefly, a binding isotherm was created by keeping the concentrations of purified ComE:DE and ComE:DA proteins constant and varying the concentration of isotopically labeled comCΔ substrate (a defined ComE single binding site derived from the upstream region of comC) (13). The concentration of DNA, which produced the half-maximal amount of shifted complex (ComE bound to comCΔ substrate), was used to estimate the Kd (equilibrium dissociation constant). Although this approach does not identify the oligomeric state of the protein protomer (monomer, dimer, etc.), it does indicate the active ratio of protomer in our preparations; this quantity is identical to the maximal amount of molar equivalents of shifted DNA (Bmax). We found that ComE had a Kd of 3.4 × 10−9 M and a Bmax value of 5.9 × 10−9 M (10), indicating that only 39% of the added 15 nM ComE is active in binding DNA, given that at this point we have no evidence to suggest the oligomeric state of ComE under physiological conditions, we have assumed here that ComE is present as a monomer. Therefore, in order to normalize the amount of active protein in each binding assay, all our protein preparations were assumed to be in a monomeric state. The Bmax for each binding isotherm was utilized to determine the percentage of active protein, and then the measured Kd was divided by this number to determine the adjusted Kd value for each protein (Table 3). As shown in Table 3, ComE:DE has a Kd value of 3.7 × 10−9 M (adjusted Kd value is 11.6), which is very similar to the ComE Kd value (3.4 × 10−9 M, adjusted Kd value is 8.7) (10). In contrast, ComE:DA has an adjusted Kd value approximately 144% and 87% higher than ComE and ComE:DE, respectively. Should the oligomeric state of any of these proteins vary from the monomer form, it would of course alter the true value of the Kd. Given our simple assumptions, it seems that the changes at this putative phosphorylation site had a decidedly modest effect on binding affinity.

Table 3
Kd, Bmax, and adjusted Kd values of ComE, ComE:DE, ComE:DA, and ComE612

To further characterize this binding, we performed DNase I footprinting. As shown in Fig. 1, DNase I footprint analysis of ComE:DE with the wild-type upstream sequence of comC and comCΔ sites showed that ComE:DE protected the same regions compared to ComE (10). Interestingly, when ComE:DA was used in the DNase I footprinting analysis with comC and comCΔ substrates (Fig. 2), the same region of protections were observed as those of ComE and ComE:DE; however, the protections were significantly weaker and no hypersensitive sites were observed on comC. ComE and ComE:DE required 97.5 nM (10) and 80 nM, respectively, for strong protection, whereas 240 nM ComE:DA was required to produce only weak protection on either comC or comCΔ substrate. These weaker footprints correspond to the relative Kd values (Table 3), i.e., the Kd of ComE:DA is higher than that of ComE and ComE:DE, which means ComE:DA has a lower binding affinity and explains why a greater concentration of protein was required to generate a weak footprint compared to ComE and ComE:DE (Fig. 2). More importantly, the hypersensitive sites are present in both ComE and ComE:DE, indicating a similarly formed protein-DNA complex.

Fig 1
DNase I footprint assay with ComE:DE. comC (A), comCΔ (B). The protected regions are shown on the right side of the figure. The solid lines on the sequence represent the direct repeats of the ComE binding sites (dotted lines on the gel), bold ...
Fig 2
DNase I footprint assay with ComE:DA. comC (A), comCΔ (B). The protected regions are shown on the right side of the figure. The solid lines on the sequence represent the direct repeats of the ComE binding sites (dotted lines on the gel), bold ...

In vitro phosphorylation by small molecule phosphate donors.

RRs typically receive a phosphate group from HKs and often this phosphorylation activates the RRs for function. Studies have shown, however, that RRs can also be phosphorylated by small molecule phosphodonors such as acetyl-P (41). Typically, a phosphorylated RR binds to a DNA binding site with a higher affinity than that of an unphosphorylated RR (7, 26). Although we have shown that ComE binds to its binding site with high affinity (Kd, ~10−9 M), the phosphorylation state of ComE is unknown. To investigate whether the addition of acetyl-P can increase the binding affinity of ComE, we performed EMSA analysis of ComE binding in the presence of acetyl-P. We observed that ComE binding to the comC DNA substrate was not significantly improved with increasing concentrations of acetyl-P (up to 100 mM), and in fact, there appears to be diminution of the extent of the shifted complex (data not shown). It is possible that ComE might have already been phosphorylated endogenously when it was purified from E. coli DH5α. Ladds et al. have shown that purified Spo0A, a RR involved in sporulation of B. subtilis, is sufficiently phosphorylated by either acetyl-P or another E. coli HK (17). Although to date, in vivo cases of cross talk with native HKs and RRs have only been reported in mutant backgrounds under specific conditions (34). To exclude that ComE may have been phosphorylated by endogenous acetyl-P, ComE was purified from an E. coli strain, SG612 (henceforth referred to as ComE612), which lacks the two genes, ackA and pta, that are responsible for the production of acetyl-P (11, 15, 41).

In Fig. 3, ComE612 generated similar footprints compared to those by ComE (10), including the extent of protection and the hypersensitive sites. Furthermore, we show that this protein binds with a Kd of 4.9 × 10−9 M (adjusted Kd value is 16.3), indicating a modest decrease similar to ComE:DA (ComE612's adjusted Kd is 83% higher than ComE's adjusted Kd) but otherwise generates similar hypersensitive cleavages in DNase I footprinting compared to ComE. We did try to phosphorylate ComE612 with acetyl-P but observed no improvement in the DNA binding activity in EMSA experiments (data not shown) and therefore did not attempt footprints with acetyl-P-treated ComE612.

Fig 3
DNase I footprint assay of ComE612 with comC substrate. The protected regions are shown on the right side of the figure. The solid lines on the sequence represent the direct repeats of ComE binding sites (dotted lines on the gel), bold bases indicate ...

Even though acetyl-P serves as a phosphate donor for many RRs, there are examples of RRs that cannot be phosphorylated with acetyl-P, such as PhoP of B. subtilis (23). Therefore, we utilized another high-energy phosphate donor, PA, which has been shown to successfully phosphorylate other RRs in vitro (25). In Fig. 4, we demonstrate that when ComE is incubated with PA for 2 h at room temperature, the binding declines at the higher concentrations of treated ComE. To further test whether PA-treated ComE binds to DNA like untreated ComE, we performed DNase I footprinting analysis. As shown in Fig. 5, PA-treated ComE generated the same footprint and hypersensitive sites as seen for untreated ComE. However, a higher concentration of protein was required to observe the same protection for PA-treated ComE than was required for ComE. Overall, we have shown that under our experimental conditions, acetyl-P and PA treatment appears to modestly decrease the binding affinity of ComE and ComE612 but does not diminish its ability to protect comC. In addition, treatment of ComE with PA and untreated ComE612 showed a decrease in binding and thus more ComE was required for protection in a DNase I experiment. Although we found no evidence that phosphorylation of ComE affected DNA binding affinity, we cannot rule out that our in vitro binding conditions might vary sufficiently compared to in vivo such that we could not resolve a difference between the phosphorylated and unphosphorylated state of ComE in target binding.

Fig 4
EMSA analysis of PA treated ComE with comC substrate. Increasing concentration of PA treated ComE at room temperature for 2 h. EMSA of ComE without PA (A) and EMSA of ComE-PA (B). Lanes 1 to 7: 0, 1, 2, 4, 8, 16, and 32 nM ComE.
Fig 5
DNase I footprint assay of ComE with preincubation of PA. The protected regions of the comC substrate are shown on the right side of the figure. The solid lines on the sequence represent the direct repeats of ComE binding sites (dotted lines on the gel), ...

Phosphorylation stimulates ComE dimerization.

We wanted to investigate how the phosphorylated form of ComE differs from that of the unphosphorylated state. Since E. coli-expressed ComE (presumably unphosphorylated) already has a high affinity for DNA, we tested whether phosphorylation affects the oligomeric state of ComE in an in vitro cross-linking experiment. To do this, we treated ComE in the presence and absence of PA with a cross-linking reagent, DMS, a homobifunctional reagent with imidoester reactive groups that react with primary amines of amino acid residues (5, 25). Reactions were then analyzed by SDS-polyacrylamide gel electrophoresis, followed by Coomassie brilliant blue or Ni-NTA-Atto conjugate staining, the latter being specific for His-tagged proteins. As shown in Fig. 6A, two additional bands with slower electrophoretic mobility appeared in the presence of PA and DMS. These two bands are unique and are absent when ComE was treated with either PA or DMS alone. This suggests that these bands represent the dimeric form of ComE since their apparent molecular weight is ~64 kDa, which is twice the size of a His-tagged ComE monomer (33 kDa). It is plausible that one band is a dimer formed between two phosphorylated ComE monomers, while the other band is a heterodimer of one phosphorylated and one unphosphorylated ComE monomers; the doubly phosphorylated species being the faster of the two owing to its higher charge-to-mass ratio. To clarify that this doublet is indeed an oligomeric form of ComE, we used an Ni-NTA-Atto conjugate specific for the histidine tag. As shown in Fig. 6B, we demonstrated that in addition to the ComE monomer, both shifted bands were detected with the histidine tag-specific probe, indicating the bands were ComE dimers. Given that the cross-linking occurs only under conditions of phosphorylation, this suggests that some of ComE is likely to be in an at least a dimer state upon phosphorylation. In addition, ComE:DE and ComE:DA were analyzed under the same conditions to test for their oligomeric state and neither formed a dimer species in the presence of phosphoramidate and DMS (Fig. 6C and D). The band visible in Fig. 6C and D (Fig. 6) below the 64 kDa marker is an unknown contaminating protein from the purification of ComE:DE and ComE:DA and not an oligomer of ComE, as it is present in the absence of both PA and DMS. These results are consistent with D60 acting as the site of phosphorylation and phosphorylation affecting the oligomeric state of ComE. It also demonstrates that the D60E mutation is not sufficient to elicit dimerization, suggesting it is not a good substitute for the phosphorylated form of ComE.

Fig 6
SDS-PAGE of cross-linking of ComE, ComE:DE, and ComE:DA with PA and DMS. The molecular weights are indicated in kilodaltons on the left of the figures, and the ComE dimers are indicated on the right. Coomassie-stained gels of ComE (A), ComE stained with ...


The physiologic state of a protein can dictate its biological function. For RRs, phosphorylation plays an essential role in the signal relay of the TCSTS. In this work, we examined the S. mutans RR ComE, focusing on its phosphorylation state and its capacity to form oligomers. We provide evidence that phosphorylation fails to strongly affect both the specificity and affinity of binding of ComE to its cognate binding site. In contrast, phosphorylation appears to affect the oligomeric state of the protein, perhaps being critical for downstream functions after binding.

A variety of mechanisms have been identified by which phosphorylation activates RRs. Here we examined if the positional negative charge was necessary and sufficient to replace phosphorylation and the active site aspartate. A change in the conserved aspartic acid to glutamic acid can mimic the phosphorylated form of some RRs, and the consequences of this change differs from one species to another (3, 8, 12, 29, 33, 35). To examine the effect of this mutation on ComE, the putative active site aspartic acid was altered to glutamate or alanine. From binding affinity experiments, we found that ComE:DE has a similar equilibrium dissociation constant as ComE (Table 3). However, ComE:DA has a higher Kd than either ComE or ComE:DE, which suggests that the replacement of aspartic acid with alanine, in either negative charge or configuration, has a marginal effect on ComE binding, demonstrating that phosphorylation is not an intrinsic quality of protein-DNA interactions.

We further tested these two mutant ComE proteins in DNase I footprinting assays with two comC substrates. We found the identical protection pattern for both wild-type ComE and ComE:DE (Fig. 1). Interestingly, the footprint of ComE:DA showed a difference compared to ComE and ComE:DE (Fig. 1 and and2).2). Similar protection of the comC substrate was observed; however, the protection was modestly weaker and no hypersensitive sites were found on the ComE:DA footprint (although there was a hypersensitive site on the comCΔ substrate) (Fig. 2). Hypersensitive spots from footprinting analysis usually indicate the bending of the DNA by the protein of interest. As described previously by Hoover et al., integration host factor (IHF) facilitated the activation of nitrogen fixation nif operon by bending the regulatory region of the DNA to promote the interaction of transcription factor NifA and RNA polymerase (9). Although the role of DNA bending by ComE has not been elucidated, the lack of hypersensitive spots by ComE:DA could mean that the aspartic acid plays a significant role in the biological function of ComE. It is plausible that the change from aspartic acid to alanine results in a less active form of ComE, which is consistent with the finding that ComE:DA has a slightly higher equilibrium dissociation constant and a lack of hypersensitive sites as judged by DNase I footprinting. The similarity in binding pattern and affinity for ComE:DE and ComE is not unusual and has been seen with Caulobacter crescentus CtrA D51E, which binds to DNA with similar affinity as the wild-type unphosphorylated CtrA (33).

As previously indicated, phosphorylation of an RR plays an essential part in activation and often increases its affinity for target DNA. In order to test the effects on binding activity of phosphorylated ComE, we directly phosphorylated ComE with small molecule phosphate donors and purified ComE from an E. coli pta-ackA double mutant strain, which eliminated the production of acetyl-P and reduced the likelihood of an endogenously phosphorylated ComE. ComE612 showed identical protection and hypersensitive site cleavages of the comC substrate as the wild-type ComE (Fig. 3). This indicates that the gross binding and structural imposition of each protein on the DNA are indistinguishable. However, we did note a 2-fold difference in binding affinity of the ComE612 protein to DNA but in the absence of any additional data cannot determine if this difference is intrinsic to the protein itself or to the preparations of protein purified; therefore, the binding affinity difference remains an open question. We do, however, distinguish the DNA binding abilities of ComE:DA and ComE612 despite similar DNA binding affinities; the DNase I footprint of ComE612 appears identical to ComE while the ComE:DA footprint does not. Finally, we tried to phosphorylate ComE612 with acetyl-P but observed no improvement in the DNA binding activity in EMSA experiments (data not shown).

One possibility is that ComE cannot be phosphorylated by acetyl-P. In fact, not all RRs are phosphorylated when pretreated with acetyl-P (23, 24). However, other phosphodonors, such as carbamyl phosphate and PA, have been used to phosphorylate RRs in vitro (24, 26, 44). We used PA as an alternative donor to phosphorylate ComE. Treatment with PA reduced ComE binding activity to comC substrate in EMSA. There are at least three possibilities for why we may not have seen an enhancement of binding affinity when ComE was phosphorylated by a phosphate donor. First, unphosphorylated ComE may be the active form of ComE as has been observed with B. subtilis DegU (4). Another possibility is that ComE is efficiently phosphorylated but has a fast and spontaneous autodephosphorylation rate as reported recently by Thomas et al. (37). However, since ComE lacks some of the important amino acids identified in the Thomas study, which were indicative of high autodephosphorylation rates, we were unable to predict with confidence at what relative rate ComE is predisposed to autodephosphorylate. An empirical experiment would have to be performed to determine the autodephosphorylation rate of ComE. A third possibility is that while phosphorylation might be important to activate RRs for gene expression, it may not play a role in ComE binding, as has been observed for S. Typhimurium NtrC (43). Currently, we do not know whether phosphorylation is required for ComE to regulate gene expression, but our analysis of ComE binding strongly favors that phosphorylation does not play a role in DNA binding affinity.

In addition to an increase in binding affinity and activation of gene regulation by phosphorylation, some RRs oligomerize upon phosphorylation. NtrC of enteric bacteria oligomerizes when phosphorylated, and this oligomer catalyzes the isomerization of closed complexes between σ54 holoenzyme and the promoter to open complexes, which activate transcription (40). On the other hand, simple DNA recognition by the RR can promote oligomerization (25). In E. coli, the RR involved in osmotic regulation, OmpR, was shown to dimerize upon phosphorylation or by DNA binding (25). To address if ComE forms an oligomer when phosphorylated, we used a homobifunctional cross-linker, DMS, to examine the effect of phosphorylation on ComE oligomerization. We showed that when ComE is phosphorylated by PA, a shift to a dimer state becomes favored. However, we observed dimerization only in a small fraction of ComE. There may be a kinetic barrier preventing dimer formation, for example, this phosphorylation-induced dimer could be very unstable or the conformation necessary for dimerization has too short a half-life. Alternatively, it is possible that the DMS reacts nonproductively so that dimers are not prone to form. A similar phenomenon was observed for OmpR in which only a portion of OmpR became dimerized in the presence of PA and the cross-linking reagent (25). However, in contrast to OmpR (25), addition of DNA did not increase the dimer formation in the cross-linking reaction (data not shown). Interestingly, when ComE:DE and ComE:DA were incubated with PA and DMS, we did not observe oligomerization of either protein. This result is consistent with D60 of ComE being the active-site aspartic acid for phosphorylation and phosphorylation facilitating oligomerization.

The significance of the phosphorylation-induced ComE dimer remains unknown, but we considered two possibilities. One model is that unphosphorylated ComE binds to one repeat with high affinity as a monomer, and this binding induces a cooperative interaction to the second direct repeat. Once two ComE protomers occupy each direct repeat, phosphorylation of the ComE monomers would regulate gene expression. This model seems unlikely since we were able to show that phosphorylation of ComE by PA also induces cooperative binding to approximately the same degree as unphosphorylated ComE (data not shown). A second possibility is that phosphorylation induces the dimer formation of ComE before binding to the direct repeats and then this dimer binds to the target site to regulate gene expression. In addition, the orientation of the ComE dimers with respect to the RNA polymerase might serve as a mechanism to distinguish between activation and repression of genes. Studies have shown that NtrC activates transcription by contacting RNA polymerase by means of a DNA loop, which allows the polymerase to gain access to the template DNA strand in a productive way (30, 39, 43). It is possible that ComE binds in a head-to-tail fashion, and the formation of the oligomer (greater than dimers) extends ComE monomers to accomplish a similar contact with RNA polymerase, e.g., when the head of ComE contacts the RNA polymerase, the gene is turned off, and when the tail contacts the RNA polymerase, gene expression is activated or vice versa (Fig. 7). Previously, we have published that ComE acts bifunctionally where it both activates mutacin (nlmC) production and represses CSP biosynthesis through the same intergenic space (13). This model accounts for how ComE could regulate these two genes differently while utilizing the same intergenic sites.

Fig 7
Proposed mechanism on how ComE activates nlmC and represses comC. ComE (open arrow bar) binds to a consensus binding site in a head-to-tail fashion and forms oligomers to extend and contact with RNA polymerase (RNAP; diamond shape). When the head contacts ...

In summary, we have characterized conditions that lead to the oligomeric state of ComE. We determined that phosphorylation can facilitate dimer formation of ComE and is likely responsible for downstream activities, possibly through interactions with RNA polymerase.


We thank Linda Kenney (University of Illinois at Chicago) for kindly providing phosphoramidate, Alan Wolfe (University of Illinois at Chicago) for providing bacterial strains, and Eduardo A. Ayala for the critical reading of our manuscript.

This study was supported by NIH grants 5R01DE013230 (D.G.C. and S.D.G.), 4R00DE018400 (J.K.), RO1-DE014757 (F.Q.), and 1R01DE020102-01 (W.S.).


Published ahead of print 30 December 2011


1. Bearson BL, Wilson L, Foster JW. 1998. A low pH-inducible, PhoPQ-dependent acid tolerance response protects Salmonella Typhimurium against inorganic acid stress. J. Bacteriol. 180:2409–2417. [PMC free article] [PubMed]
2. Biswas I, Drake L, Erkina D, Biswas S. 2008. Involvement of sensor kinases in the stress tolerance response of Streptococcus mutans. J. Bacteriol. 190:68–77. [PMC free article] [PubMed]
3. Bourret RB, Hess JF, Simon MI. 1990. Conserved aspartate residues and phosphorylation in signal transduction by the chemotaxis protein CheY. Proc. Natl. Acad. Sci. U. S. A. 87:41–45. [PubMed]
4. Dahl MK, Msadek T, Kunst F, Rapoport G. 1992. The phosphorylation state of the DegU response regulator acts as a molecular switch allowing either degradative enzyme synthesis or expression of genetic competence in Bacillus subtilis. J. Biol. Chem. 267:14509–14514. [PubMed]
5. Davies GE, Stark GR. 1970. Use of dimethyl suberimidate, a cross-linking reagent, in studying the subunit structure of oligomeric proteins. Proc. Natl. Acad. Sci. U. S. A. 66:651–656. [PubMed]
6. Davison J. 1999. Genetic exchange between bacteria in the environment. Plasmid 42:73–91. [PubMed]
7. Federle MJ, Scott JR. 2002. Identification of binding sites for the group A streptococcal global regulator CovR. Mol. Microbiol. 43:1161–1172. [PubMed]
8. Green BD, Olmedo G, Youngman P. 1991. A genetic analysis of Spo0A structure and function. Res. Microbiol. 142:825–830. [PubMed]
9. Hoover TR, Santero E, Porter S, Kustu S. 1990. The integration host factor stimulates interaction of RNA polymerase with NIFA, the transcriptional activator for nitrogen fixation operons. Cell 63:11–22. [PubMed]
10. Hung DC, et al. 2011. Characterization of DNA binding sites of ComE response regulator from Streptococcus mutans. J. Bacteriol. 193:3642–3652. [PMC free article] [PubMed]
11. Klein AH, Shulla A, Reimann SA, Keating DH, Wolfe AJ. 2007. The intracellular concentration of acetyl phosphate in Escherichia coli is sufficient for direct phosphorylation of two-component response regulators. J. Bacteriol. 189:5574–5581. [PMC free article] [PubMed]
12. Klose KE, Weiss DS, Kustu S. 1993. Glutamate at the site of phosphorylation of nitrogen-regulatory protein NTRC mimics aspartyl-phosphate and activates the protein. J. Mol. Biol. 232:67–78. [PubMed]
13. Kreth J, et al. 2007. The response regulator ComE in Streptococcus mutans functions both as a transcription activator of mutacin production and repressor of CSP biosynthesis. Microbiology 153:1799–1807. [PMC free article] [PubMed]
14. Kreth J, Merritt J, Zhu L, Shi W, Qi F. 2006. Cell density- and ComE-dependent expression of a group of mutacin and mutacin-like genes in Streptococcus mutans. FEMS Microbiol. Lett. 265:11–17. [PubMed]
15. Kumari S, et al. 2000. Regulation of acetyl coenzyme A synthetase in Escherichia coli. J. Bacteriol. 182:4173–4179. [PMC free article] [PubMed]
16. Kuramitsu HK, Trapa V. 1984. Genetic exchange between oral streptococci during mixed growth. J. Gen. Microbiol. 130:2497–2500. [PubMed]
17. Ladds JC, et al. 2003. The response regulator Spo0A from Bacillus subtilis is efficiently phosphorylated in Escherichia coli. FEMS Microbiol. Lett. 223:153–157. [PubMed]
18. Lan CY, Igo MM. 1998. Differential expression of the OmpF and OmpC porin proteins in Escherichia coli K-12 depends upon the level of active OmpR. J. Bacteriol. 180:171–174. [PMC free article] [PubMed]
19. Laub MT, Goulian M. 2007. Specificity in two-component signal transduction pathways. Annu. Rev. Genet. 41:121–145. [PubMed]
20. Li YH, et al. 2002. A quorum-sensing signaling system essential for genetic competence in Streptococcus mutans is involved in biofilm formation. J. Bacteriol. 184:2699–2708. [PMC free article] [PubMed]
21. Li YH, Hanna MN, Svensater G, Ellen RP, Cvitkovitch DG. 2001. Cell density modulates acid adaptation in Streptococcus mutans: implications for survival in biofilms. J. Bacteriol. 183:6875–6884. [PMC free article] [PubMed]
22. Li YH, Lau PC, Lee JH, Ellen RP, Cvitkovitch DG. 2001. Natural genetic transformation of Streptococcus mutans growing in biofilms. J. Bacteriol. 183:897–908. [PMC free article] [PubMed]
23. Liu W, Hulett FM. 1997. Bacillus subtilis PhoP binds to the phoB tandem promoter exclusively within the phosphate starvation-inducible promoter. J. Bacteriol. 179:6302–6310. [PMC free article] [PubMed]
24. Lukat GS, McCleary WR, Stock AM, Stock JB. 1992. Phosphorylation of bacterial response regulator proteins by low molecular weight phospho-donors. Proc. Natl. Acad. Sci. U. S. A. 89:718–722. [PubMed]
25. Maris AE, Walthers D, Mattison K, Byers N, Kenney LJ. 2005. The response regulator OmpR oligomerizes via beta-sheets to form head-to-head dimers. J. Mol. Biol. 350:843–856. [PubMed]
26. Mattison K, Oropeza R, Byers N, Kenney LJ. 2002. A phosphorylation site mutant of OmpR reveals different binding conformations at ompF and ompC. J. Mol. Biol. 315:497–511. [PubMed]
27. Nixon BT, Ronson CW, Ausubel FM. 1986. Two-component regulatory systems responsive to environmental stimuli share strongly conserved domains with the nitrogen assimilation regulatory genes ntrB and ntrC. Proc. Natl. Acad. Sci. U. S. A. 83:7850–7854. [PubMed]
28. Parkinson JS, Kofoid EC. 1992. Communication modules in bacterial signaling proteins. Annu. Rev. Genet. 26:71–112. [PubMed]
29. Pazour GJ, Ta CN, Das A. 1992. Constitutive mutations of Agrobacterium tumefaciens transcriptional activator virG. J. Bacteriol. 174:4169–4174. [PMC free article] [PubMed]
30. Popham DL, Szeto D, Keener J, Kustu S. 1989. Function of a bacterial activator protein that binds to transcriptional enhancers. Science 243:629–635. [PubMed]
31. Ronson CW, Nixon BT, Ausubel FM. 1987. Conserved domains in bacterial regulatory proteins that respond to environmental stimuli. Cell 49:579–581. [PubMed]
32. Sambrook J, Russell D. 2001. Molecular cloning: a laboratory manual, 3rd ed, vol 1-3. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
33. Siam R, Marczynski GT. 2003. Glutamate at the phosphorylation site of response regulator CtrA provides essential activities without increasing DNA binding. Nucleic Acids Res. 31:1775–1779. [PMC free article] [PubMed]
34. Siryaporn A, Goulian M. 2010. Characterizing cross-talk in vivo avoiding pitfalls and overinterpretation. Methods Enzymol. 471:1–16. [PubMed]
35. Stewart RC, Roth AF, Dahlquist FW. 1990. Mutations that affect control of the methylesterase activity of CheB, a component of the chemotaxis adaptation system in Escherichia coli. J. Bacteriol. 172:3388–3399. [PMC free article] [PubMed]
36. Stock AM, Robinson VL, Goudreau PN. 2000. Two-component signal transduction. Annu. Rev. Biochem. 69:183–215. [PubMed]
37. Thomas SA, Brewster JA, Bourret RB. 2008. Two variable active site residues modulate response regulator phosphoryl group stability. Mol. Microbiol. 69:453–465. [PMC free article] [PubMed]
38. van der Ploeg JR. 2005. Regulation of bacteriocin production in Streptococcus mutans by the quorum-sensing system required for development of genetic competence. J. Bacteriol. 187:3980–3989. [PMC free article] [PubMed]
39. Wedel A, Weiss DS, Popham D, Droge P, Kustu S. 1990. A bacterial enhancer functions to tether a transcriptional activator near a promoter. Science 248:486–490. [PubMed]
40. Weiss DS, Batut J, Klose KE, Keener J, Kustu S. 1991. The phosphorylated form of the enhancer-binding protein NTRC has an ATPase activity that is essential for activation of transcription. Cell 67:155–167. [PubMed]
41. Wolfe AJ. 2005. The acetate switch. Microbiol. Mol. Biol. Rev. 69:12–50. [PMC free article] [PubMed]
42. Wolfe AJ, et al. 2003. Evidence that acetyl phosphate functions as a global signal during biofilm development. Mol. Microbiol. 48:977–988. [PubMed]
43. Wyman C, Rombel I, North AK, Bustamante C, Kustu S. 1997. Unusual oligomerization required for activity of NtrC, a bacterial enhancer-binding protein. Science 275:1658–1661. [PubMed]
44. Zapf JW, Hoch JA, Whiteley JM. 1996. A phosphotransferase activity of the Bacillus subtilis sporulation protein Spo0F that employs phosphoramidate substrates. Biochemistry 35:2926–2933. [PubMed]

Articles from Journal of Bacteriology are provided here courtesy of American Society for Microbiology (ASM)