Modulating the body's networks could become mainstream therapy for many health issues.
Image credit: The Feinstein Institutes for Medicine Research.
New Research In
Physical Sciences
Featured Portals
Articles by Topic
Biological Sciences
Featured Portals
Articles by Topic
-
Edited by Carol A. Gross, University of California, San Francisco, CA, and approved January 31, 2003 (received for review December 16, 2002)
Abstract
Transcriptional activator proteins in bacteria often operate by interaction with the C-terminal domain of the α-subunit of RNA polymerase (RNAP). Here we report the discovery of an “anti-α” factor Spx in Bacillus subtilis that blocks transcriptional activation by binding to the α-C-terminal domain, thereby interfering with the capacity of RNAP to respond to certain activator proteins. Spx disrupts complex formation between the activator proteins ResD and ComA and promoter-bound RNAP, and it does so by direct interaction with the α-subunit. ResD- and ComA-stimulated transcription requires the proteolytic elimination of Spx by the ATP-dependent protease ClpXP. Spx represents a class of transcriptional regulators that inhibit activator-stimulated transcription by interaction with α.
Transcriptional activation in bacteria involves contacts between DNA-bound activators and promoter-bound RNA polymerase (RNAP). Most such interactions require the α-subunit of RNAP, which possesses several activator–interaction surfaces within its C-terminal domain (CTD) (1, 2). One such activator is ComA of the bacterium Bacillus subtilis (3–5), a response regulator, required for transcription of genes involved in the development of genetic competence (6). In response to high cell density, ComA becomes phosphorylated by interaction with its cognate histidine kinase, ComP, that is activated when it binds the pheromone ComX (7–9). ComA then activates the transcription initiation of the srf operon, which encodes the competence regulatory peptide ComS (10, 11). ComS serves to release the transcriptional activator ComK from its inhibitory complex composed of the proteins MecA and ClpCP (10–13), so that ComK can stimulate transcription of genes required for DNA uptake in competent cells (14). Interestingly, ComA-dependent transcription of srf requires the ATP-dependent protease ClpXP (15, 16), which functions to eliminate the 15.4-kDa Spx protein (refs. 17 and 18; see Results). A mutation in clpX blocks ComA-activated transcription and has severe effects on growth and development (17). These pleiotropic effects of ClpXP absence can be suppressed either by the elimination of Spx or by missense mutations in the rpoA gene that encodes the RNAP α-subunit (refs. 16 and 17; see Results). This latter finding suggested that Spx exerts its negative effect on ComA-mediated transcription, and other transcriptional activation systems, by interaction with RNAP. A similar relationship between Spx and the α-subunit of RNAP was observed for ResD-activated transcription (see Results). ResD, like ComA, is a response regulator and transcriptional activator. It is part of the ResDE two-component signal transduction system that is required for the transcription of genes that are induced in response to oxygen limitation (19).
In this article we show that Spx interferes with activator-stimulated transcription by interaction with the RNAP α-CTD, a mechanism of transcriptional repression not observed before in studies of prokaryotic transcriptional regulation. A model is presented in which Spx may function to globally reduce transcription of genes involved in growth- and development-promoting processes during periods of extreme stress.
Methods
Strains.
B. subtilis strains were derived from JH642 (trpC2 pheA1) and include MAB188 (comK-lacZ), ORB3087 (comK-lacZ clpX), ORB3247 (srf-lacZ), and ORB3249 (srf-lacZ clpX) (16). ORB3841 (srf-lacZ clpX spx) was constructed by transforming ORB3834 (spx∷neo) (17) with DNA from ORB3247 and LAB2876 (clpX∷spc) (20). ORB4135 bears the cxs-16 allele of spx. To obtain a srf-lacZ spxcxs-16clpX strain, ORB4045 (spx∷neo) was constructed by transforming ZB307A (prototroph) (21) with ORB3834 DNA. ORB4135 was then transformed with ORB4045 DNA with selection for Trp+. The transformants were screened for sensitivity to spectinomycin (Spc) (indicative of clpX+) and sensitivity to Neo (indicative of cxs-16), which yielded ORB4055 [comK-lacZ (chloramphenicol resistance cassette, or Cmr) pheA1 spxcxs-16]. ORB4055 was transformed with ORB4045 DNA, and Phe+ transformants were screened for Neo and Cm sensitivity to obtain ORB4149 (prototroph spxcxs-16). ORB4264 (srf-lacZ clpX spxcxs-16) was obtained by transforming ORB4149 with DNA from ORB3247 and LAB2876.
ORB3560 (hmp-lacZ clpX) was obtained by transforming ORB3555 (hmp-lacZ) with LAB2876 DNA. ORB3562 (hmp-lacZ clpX cxs-1) was constructed by transforming ORB3317 (clpX∷spc rpoAcxs1) (16) with ORB3555 DNA. ORB4128 (hmp-lacZ clpX spx∷neo) was obtained by transforming ORB3555 with LAB2876 and ORB3834 DNA. ORB4265 (hmp-lacZ clpX spxcxs-16) was obtained from ORB4149 (spxcxs-16) in a similar way as ORB4264 by transforming with ORB3555 and LAB2876 DNA.
Assays of β-Galactosidase Activity.
srf-lacZ strains were grown aerobically in competence medium (22). Strains carrying the hmp-lacZ fusion were grown anaerobically in 2× yeast extract/tryptone supplemented with 1% glucose and 0.2% KNO3. β-Galactosidase activity was measured as described (23).
Western Blot Analysis.
Cell extracts prepared from B. subtilis cells grown in competence medium to T2 (2 h after the end of log phase) were applied to 15% SDS/PAGE. Proteins were transferred to a nitrocellulose membrane and probed with anti-Spx antibody as described (17).
Yeast Two-Hybrid Procedure.
Plasmid pSN11 was constructed by subcloning spx from pMMN470 (16) into pGBKT7 (CLONTECH). The spxcxs-16 gene, amplified by PCR using DNA prepared from ORB4055 (cxs-16), was inserted into pGBKT7 to generate pSN18. A part of rpoA was amplified by PCR using JH642 DNA as template. The PCR fragment was inserted into pGADT7 (CLONTECH) generating pSN25. pSN26 was constructed in the same way as pSN25 except DNA from ORB3317 (clpX cxs-1) was used as template to obtain the cxs-1 allele of rpoA. The RpoA products encoded by the plasmids include amino acids between 213 and 291.
Plasmid pSN11, a pGBKT7 derivative carrying spx-GAL4 DNA-binding domain fusion, was used to transform Saccharomyces cerevisiae PJ69-4A (His−, Ade−, Trp−, and Leu−) (24), which carries a GAL2-ADE2 fusion controlled by GAL4. A genome library of B. subtilis 168 DNA (25) in a Gal4 activation domain fusion vector pGADT7 (carrying LEU2) was introduced by transformation into PJ69-4A carrying pSN11 with selection for growth on SD (synthetic defined) medium (26). Plasmids from the positive clones were used to transform PJ69-4A, together with pSN11 or pSN18.
Protein Purification.
The IMPACT self-cleavable, affinity tag system (New England Biolabs) was used to purify ResD, ResE (27), ComA, ClpX, ClpP, and RpoA. Escherichia coli ER2566 (New England Biolabs) or BL21(DE3)pLysS (28) was used for overproduction of the proteins. The ComA and RpoA proteins obtained have a Pro-Gly extension at the C termini. ClpX and ClpP proteins, containing no extra residues, were further purified by elution with a 100–600 mM KCl gradient from a High Q column (Bio-Rad). Fractions containing ClpX were applied to a hydroxyapatite column (Bio-Rad) and eluted with 0–250 mM sodium phosphate, pH 6.8 gradient.
PCR products of spx WT and cxs-16 alleles were inserted into pPROEX-1 (Life Technologies, Rockville, MD) to generate plasmids that were used for the production of His-6-tagged Spx and Spxcxs-16. Proteins were purified from Ni-NTA (Qiagen, Valencia, CA) columns, followed by High Q columns with a 50–500 mM KCl gradient. The His-6 tag was cleaved with recombinant tobacco-etch virus protease (Life Technologies), which left a Gly-Ala-His extension at the N termini of Spx and Spxcxs-16.
RNAP containing a His-10-tagged RpoC subunit was purified from B. subtilis MH5636 (WT) or ORB4123 (rpoAcxs-1) strains by using a procedure described in refs. 29 and 30.
In Vitro Proteolysis.
The reaction was carried out in 50 mM Hepes/KOH (pH 7.6), 50 mM KCl, 10 mM Mg acetate, 5 mM DTT, 5% glycerol, 5 mM ATP, 10 mM creatine phosphate, and 0.1 μg/μl creatine kinase (Sigma). Spx or Spxcxs-16 (5 μM) was incubated at 30°C in the presence of ClpP (3 μM) and ClpX (3 μM) in a 50-μl reaction mixture. At time intervals, 10 μl of the samples was collected and analyzed on a 15% SDS/PAGE followed by staining with Coomassie blue.
Electrophoretic Mobility-Shift Assays (EMSAs).
srf and hmp probes were obtained by PCR with oligonucleotide primers that were radioactively labeled by using T4 polynucleotide kinase and γ-32P-ATP, followed by purification using 6% nondenaturing PAGE. The srf probe was incubated with or without RNAP (0.025 μM for WT RNAP and 0.2 μM for RNAP with RpoAcxs-1), ComA [2.5 μM, phosphorylated by using acetyl phosphate (4)], and different amounts of Spx in 20 μl of buffer A [25 mM Tris⋅HCl, pH 7.5/100 mM KCl/0.1 mM EDTA/0.5 mM DTT/5 mM MgCl2/50 μg/ml poly(dI-dC)/50 μg/ml BSA/10% glycerol] at room temperature for 20 min. In the hmp reaction 0.25 μM ResD, 0.25 μM ResE, and 0.25 mM ATP were preincubated at room temperature for 15 min before the addition of other proteins and the labeled probe. DNA and DNA–protein complexes were resolved by 4% nondenaturing PAGE in TAE buffer (40 mM Tris-acetate, pH 8.0/1 mM EDTA). The gel was dried and analyzed with a PhosphorImager (Molecular Dynamics).
In Vitro Transcription.
Linear DNA for in vitro transcription was generated by PCR. The rpsD promoter fragment used was the same as reported (20). A 476-bp fragment containing the srf promoter encodes a 104-bp run-off transcript. A 270-bp fragment containing the hmp promoter encodes a 79-bp run-off transcript. The transcription reactions (20 μl) contained 40 mM Tris⋅HCl (pH 7.9), 10 mM NaCl, 6 mM MgCl2, 2 mM spermidine, 10 mM DTT, 10 units RNasin (Promega), 5 nM srf template, 0.025 μM RNAP, and 2.5 μM phosphorylated ComA. The mixtures were incubated at 37°C for 10 min with or without Spx before the addition of 37.5 μM ATP, CTP, and GTP, 18.75 μM UTP, and 3.75 μCi [α-32P]UTP. After 20 min, the mixtures were precipitated with ethanol. Electrophoresis was performed as described (29). The transcription of hmp was carried out in a 20-μl reaction in a similar way by using 5 nM template, 0.025 μM WT RNAP, or 0.1 μM of the enzyme with RpoAcxs-1 and 0.25 μM ResD phosphorylated with 0.25 μM ResE as described above. An RNAP concentration of 0.05 μM was used in the reactions of Fig. 3I.
Mixed template in vitro transcription reactions containing either srf or hmp promoter DNA were assembled as described above except that both reactions contained rpsD promoter DNA that was synthesized by PCR. The rpsD/srf reaction contained 0.025 μM rpsD and srf, 0.2 μM RNAP, and 1.6 μM ComA∼P. The rpsD/hmp reaction contained 5 nM rpsD and hmp, 0.05 μM RNAP, and 0.5 μM ResD and ResE.
Protein Interaction Experiment.
WT His-6-Spx (6 μM) or His-6-Spxcxs-16 was incubated with WT RpoA or RpoAcxs-1 at 37°C for 15 min in a 150-μl reaction (25 mM Tris⋅HCl, pH 8.0/100 mM KCl/5 mM MgCl2/0.1 mM DTT). The reactions were applied to 30-μl Ni-NTA columns, which were prewashed with the incubation buffer. After washing with 150 μl of buffer five times, His-6-Spx was eluted in buffer containing 200 mM imidazole. Aliquots of 15 μl from final wash and the elution sample were analyzed by 15% SDS/PAGE gel followed by Coomassie staining.
Results
spx Mutations Eliminate the Defect in srf and hmp Expression of a clpX Null Mutant.
The reduced expression of a transcriptional srf-lacZ fusion, observed in clpX mutants, is relieved by rpoAcxs mutations (16) that confer amino acid substitutions in the α-CTD (31, 32). The repression of srf was also relieved by either the spx∷neo null or spxcxs-16 missense mutation (Fig. 1A). In experiments to determine whether other activator-dependent promoters were affected by Spx, the expression of the hmp gene (33), a ResDE-controlled gene encoding flavohemoglobin and induced under oxygen limitation (27), was also examined. It is repressed in clpX mutant cells under anaerobic conditions, but spx mutations eliminate the clpX-conferred defect (Fig. 1B). As in the case of srf transcription (16), introduction of the rpoAcxs-1 mutation into the clpX background also resulted in elevated hmp-lacZ expression.
Figure 1
Effect of clpX and spx mutations on srf-lacZ and hmp-lacZ. (A) Expression of srf-lacZ was measured in cells grown aerobically in competence medium. ○, ORB3247 (WT); ●, ORB3249 (clpX); □, ORB3841 (clpX spx); ■, ORB4264 (clpX spxcxs-16). (B) hmp-lacZ was measured in cells grown anaerobically in 2× yeast extract/tryptone plus 1% glucose and 0.2% KNO3. ○, ORB3555 (WT); ●, ORB3560 (clpX); ▵, ORB3562 (clpX rpoAcxs-1); □, ORB4128 (clpX spx); ■, ORB4265 (clpX spxcxs-16). Time 0 indicates the end of log-phase growth.
The Spx protein is barely detectable in a Western blot of WT cell extract, but is abundant in clpX mutants (Fig. 2A; refs. 17 and 18). An spx mutant allele, spxcxs-16, conferring an Arg substitution at the highly conserved Gly-52, encodes an inactive form of Spx (Fig. 1) that is also abundant in a clpX mutant (Fig. 2A). The clpX spxcxs-16 strain shows no growth defect like that observed in clpX mutants (data not shown) and was complemented by the WT allele (data not shown), indicating that Spxcxs-16 is unable to engage in the target interaction by which WT Spx causes impairment of developmental processes.
Figure 2
(A) Spx levels as determined by Western blot analysis using anti-Spx antibody. Lane 1, MAB188 (WT); lane 2, ORB4135 (spxcxs-16clpX); lane 3, ORB3087 (clpX). (B) In vitro proteolysis of Spx. Reactions were incubated at 30°C for indicated times as described in Methods. The faint band visible between CrK (creatine kinase) and ClpP is from the creatine kinase used for regeneration of ATP.
In Vitro Proteolysis of Spx and Spxcxs-16 by ClpXP.
The above and previous results (16) strongly suggested that Spx is degraded by ClpXP in vivo. During the course of the present study, it was discovered that the in vitro degradation of Spx by ClpXP had been previously unsuccessful because the Spx substrate had an extra two amino acids appended to its C terminus (16). This Spx protein was degraded by ClpCP protease in vitro (18), probably because the degradation requires the adapter protein MecA. Spx protein, having no extra amino acids at the C terminus (see Methods), was efficiently degraded by ClpXP in vitro (Fig. 2B). Spxcxs-16 was degraded by ClpXP as efficiently as WT Spx (Fig. 2B).
Spx Negatively Affects Activator-Dependent Transcription.
The hypothesis that Spx-dependent repression was caused by disruption of activator-stimulated transcription initiation whereas the mutant Spxcxs-16 lacked this activity was supported by in vitro run-off transcription analysis. RNAP required ComA∼P and ResD∼P to transcribe srf (Fig. 3A) and hmp (Fig. 3C), respectively. Transcription from the rpsD [ribosomal S4 protein (35)] promoter is not affected by a clpX mutation in vivo (20) and was used as a negative control of Spx activity (Fig. 3 E and G). Increasing amounts of Spx led to a decrease in the ComA-dependent transcription of srf (Fig. 3 B, E, and F) and ResD-dependent transcription of hmp (Fig. 3 D, G, and H) relative to the rpsD control. Spxcxs-16 had no significant effect on transcription of srf and hmp (Fig. 3 B and D), a result in keeping with its inactivity in vivo. Notably, transcription using RNAP carrying RpoAcxs-1 (Fig. 3 B and D) was not affected by WT Spx. These results are in good agreement with the phenotype of spxcxs-16 and rpoAcxs-1. Spx showed no effect on the level of basal, ResD-independent hmp transcription (Fig. 3I), indicating that Spx affects only activator-dependent transcription.
Figure 3
Effect of a α-CTD mutation and Spx on srf and hmp transcription in vitro. (A) In vitro run-off transcription reactions contained srf template and WT RNAP in the presence or absence of ComA∼P. (B) srf transcription in the presence of ComA∼P by using RNAP containing WT or cxs-1 mutant α-subunit with increasing amounts of WT or cxs-16 mutant Spx. (C) In vitro transcription using hmp DNA and WT RNAP in the absence and presence of ResD∼P. (D) hmp transcription in the presence of ResD∼P was assayed by using WT or cxs-1 mutant RNAP with increasing amounts of WT or cxs-16 mutant Spx. (E) Mixed template in vitro reactions containing srf and rpsD promoter DNA, RNAP, ComA∼P, and increasing amounts of Spx as indicated. (F) Plot of the ratio of srf transcript vs. rpsD transcript as a function of Spx protein amount. Values in the y axis were determined by quantifying the bands with National Institutes of Health image j and represent the area of peaks in pixels. The transcript ratio value determined when Spx is absent is expressed as 100. (G) Mixed template reaction containing hmp and rpsD promoter DNA, RNAP, ResD∼P, and increasing amounts of Spx as indicated. (H) Plot of the ratio of hmp transcript to rpsD transcript as a function of Spx amount (see F). (I) Effect of Spx on hmp transcription in the absence of ResD.
Spx Destabilizes a Complex of RNAP and Transcriptional Activators.
The effect of Spx on the formation of transcriptional activation complexes consisting of RNAP, ComA or ResD, and their cognate promoters was examined. In the EMSA of Fig. 4A, 2.5 μM ComA∼P did not bind to the srf promoter. This finding is consistent with the previous result that showed no binding of ComA∼P to srf promoter DNA when present in a concentration of 5 μM (4). RNAP was able to bind to the srf promoter and a complex with a slower mobility than the RNAP–DNA complex was detected when 2.5 μM ComA∼P was added. Spx disrupted the ComA∼P–RNAP–DNA complex, but Spx apparently did not dissociate RNAP from srf promoter DNA. The same amount of Spx had no significant effect on binding of RNAP to the srf promoter in the absence of ComA∼P (data not shown). Spxcxs-16 did not show any effect on ComA∼P–RNAP–DNA complex formation. The formation of the ComA∼P–RNAPcxs-1–DNA complex was not affected by Spx (Fig. 4A), suggesting that the amino acid substitution conferred by the rpoAcxs-1 caused reduced affinity of RNAP for Spx. Together, these data indicated that Spx, by binding to RNAP, weakened the interaction between ComA∼P and RNAP at the srf promoter.
Figure 4
Effect of Spx on activator–RNAP–promoter complex formation. (A) An end-labeled DNA fragment containing the srf promoter was incubated with or without ComA∼P and RNAP carrying WT or cxs-1 mutant RpoA. Increasing amounts of WT Spx or Spxcxs-16 were added. (B) An end-labeled DNA fragment containing the hmp promoter was incubated with or without ResD∼P and RNAP as shown in A. (C) Effect of Spx on EMSA of hmp in the presence of RNAP, with and without ResD∼P.
In a similar experiment, 0.25 μM ResD∼P efficiently bound to the hmp promoter (Fig. 4B), unlike ComA∼P, which seems to require RNAP for stable srf promoter interaction. Furthermore, RNAP had less affinity for the hmp promoter than it has for the srf promoter, as the RNAP–hmp complex was detected only after prolonged exposure before phosphorimaging (the position is marked in Fig. 4B). More RNAP bound to the hmp promoter when ResD∼P was present, as evident by the supershifted ResD∼P–RNAP–DNA complex. Therefore, the binding dynamics of RNAP and the transcriptional activators at the srf and hmp promoters showed different characteristics. Nevertheless, Spx acted similarly in disrupting the formation of the ResD∼P–RNAP–DNA complex, resulting in the dissociation of RNAP and leaving intact the ResD∼P–hmp DNA complex (Fig. 4B). Again, Spx has no inhibitory effect on RNAP–promoter DNA interaction (Fig. 4C). The ResD∼P–RNAP–DNA complex remained stable either in the presence of Spxcxs-16 or when RNAP bearing the RpoAcxs-1 subunit was included in the reaction (Fig. 4B).
Spx Interacts with the α-CTD of RNAP.
The above results suggested that Spx disrupted activator-stimulated transcription by direct interaction with the α-CTD of RNAP. To test this, the yeast two-hybrid system was used to search for proteins encoded in a genomic B. subtilis, GAL4 fusion library that interacted with WT Spx but not with Spxcxs-16. Two plasmids (pSN20 and pSN33, Fig. 5A) were among those that activated the reporter gene, GAL2-ADE2, when introduced by transformation into a yeast strain containing pSN11 (carrying spx). Only pSN20 and pSN33, when cotransformed with pSN18, carrying spxcxs-16, failed to result in activation of the reporter gene. Sequence analysis of pSN20 and pSN33 revealed that their inserts encoded overlapping regions of the RNAP α that contained the sites of the cxs-1- and cxs-2-conferred residue substitutions (16). To further examine the effect of cxs-1 on this interaction, parts of rpoA and rpoAcxs-1 encoding polypeptides extending from amino acid positions 213 to 291 were inserted into pGADT7 (see Methods), thus fusing the CTD-coding ends with the Gal4 activation domain. The result showed that RpoA and Spx interact, whereas RpoA was unable to interact with Spxcxs-16 (Fig. 5B). As predicted, RpoAcxs-1 did not interact with WT Spx.
Figure 5
(A) B. subtilis α-subunit of RNAP (RpoA) composed of N-terminal domain (gray box) and CTD (white box). Plasmids pSN20 and pSN33 were recovered from the yeast-two hybrid screen of the B. subtilis genomic library and express activator domain fusions to RpoA that are shown as black boxes with amino acid numbers. The locations and residue changes of the cxs-1 and cxs-2 mutations in the α-CTD are shown. (B) Phenotypes of the yeast strain PJ69-4A on SD medium minus tryptophan, leucine, and adenine. Plasmids used to transform PJ69-4A are shown in each sector of the lower half circle. AD-RpoA (pSN25) carries a fragment encoding a C-terminal region of RpoA (amino acids 213–291) and AD-RpoAcxs-1 (pSN26) carries the cxs-1 mutation. BD-Spx (pSN11) carries the WT spx. BD-Spxcxs-16 (pSN18) is identical to pSN11 except that it carries the cxs-16 allele. (C) Effects of rpoAcxs-1 and spxcxs-16 mutations on Spx–RpoA interaction in vitro. WT and mutant proteins were combined pairwise in solution then applied to a Ni-NTA column (see Methods). The last of five wash fractions (W) and the elution fractions (E) were analyzed by SDS/PAGE.
The proposed interaction of RpoA with Spx was verified by a pull-down experiment using Ni-NTA column chromatography (Fig. 5C). WT His-6-Spx was incubated either with WT RpoA or RpoAcxs-1 and the mixture was applied to a Ni-NTA column. WT RpoA, but not RpoAcxs-1, was present in the eluate, indicating a direct interaction between RpoA and Spx whereas poor interaction between Spx and the mutant cxs-1 α was observed. WT RpoA did not coelute with His-6-Spxcxs-16 from the Ni column, demonstrating that the conserved Gly-52 residue in Spx is important in binding to RpoA. These results fully explain the previous finding that rpoAcxs-1 suppressed the effect of clpX and clpP on the srf and hmp expression by preventing the binding of Spx to the α-CTD of RNAP.
Discussion
Spx is a negative transcriptional regulator in prokaryotes that disrupts activator-dependent transcription initiation by direct interaction with the α-CTD. As an “anti-α” factor, the Spx mode of action is different from that of repressors that interact with promoter DNA. It is unlikely that it is a sequence-specific DNA-binding protein given the different genes that are affected by Spx. Certain regulators of transcription, namely GalR, LacI (36), FNR (37), and the phi29 P4 protein (38), can exert negative control by interacting with the α-CTD, but these are sequence-specific DNA-binding proteins that directly interact with α-CTD to prevent RNAP from initiating transcription. They also exert their repression in a promoter-specific manner. In contrast, Spx interacts with the α-CTD to interfere with activator–RNAP–promoter complex formation and it exerts its effect more globally.
Spx interaction need not result in removal of RNAP from promoter DNA, as is evident in Fig. 4, in which RNAP remains bound to srf promoter DNA in the presence of Spx, whereas ComA∼P is released, rendering RNAP incapable of efficiently initiating transcription. The rpoA mutations that eliminate Spx interaction reside in the part that encodes helix 1 of the first HhH motif, which has been implicated in binding of α to extended promoter DNA (39–41) and is required for productive activator–RNAP interaction (31).
In a previous study (34), Spx was found to act as an antagonist of ComS by interaction with the ComK/MecA/ClpC complex. This interaction resulted in inhibition of ComS-dependent release of ComK, which is required for ComK to activate the transcription of genes that are necessary to establish the competent state (12). The key finding of the study was that rpoAcxs mutations did not suppress clpP with respect to ComK activity, but the spx null mutation did, indicating that Spx can exert a negative effect on the activity of ComK that does not involve its interaction with RNAP. As discussed (34), we do not know whether this reflects a function of Spx as a regulator of MecA activity or is simply a consequence of the fact that Spx is a substrate for MecA/ClpC and its accumulation in clpP mutant cells may result in the interference with the normal ComS-dependent ComK activation process by its interaction with MecA/ClpC.
B. subtilis ComA and ResD, when phosphorylated, are sufficient for in vitro srf and hmp transcriptional activation. EMSA data suggest that ResD and ComA interact with the α-subunit at their cognate promoters (data not shown). ComA binds to two ComA boxes (−118 to −103 and −74 to −59; refs. 3 and 4) and ResD binds to sequences from −80 to −40 near the hmp promoter (27). The locations of the activator-binding sequences relative to the promoter suggests that the α-CTD is the activation target. Contact between activators and the α-CTD is known to increase the affinity of RNAP to promoter DNA (1, 42). According to the EMSA experiment, ResD increases the affinity of RNAP for the hmp promoter, but the activity of ComA in srf transcription initiation does not fit a simple recruiting model. Activator-induced conformation changes through α-CTD interaction are suggested in some cases where RNAP forms a nonproductive complex with promoter DNA. Although RNAP is bound to the malT promoter, CRP might be required to accelerate the formation of the open complex (43). Similarly, Ada does not enhance RNAP binding to the ada and aidB promoters but instead stimulates transcription by stabilizing an intermediate RNAP–promoter complex (44). Involvement of the α-CTD in the transition from closed to intermediate complexes was also proposed as the role of the p4 protein in activation at the phage phi29 A3 promoter (45). ComA could serve a similar function in stimulating RNAP-catalyzed transcription from the srf promoter.
The putative target of Spx interaction is the conserved helix I region that participates in promoter interaction (31). Amino acid positions affected by the cxs mutations are Val-260 and Tyr-263 (Fig. 6, which is published as supporting information on the PNAS web site, www.pnas.org), the latter of which is conserved in Gram-positive organisms such as Bacillus species and other low-GC content bacteria that also carry the spx gene. Modeling of B. subtilis α-CTD according to the E. coli structure (ref. 38; Fig. 6) shows that the Tyr aromatic ring lies near the Val-260 side chain. The yeast two-hybrid and the pull-down experiments (Fig. 5) showed that Tyr-263 is critical for interaction with Spx. Val-260 is another residue that may participate directly in α-CTD–Spx interaction, as the cxs-2 (V260A) mutation, which has the same phenotype as cxs-1, confers a defect in Spx–RNAP interaction as revealed by yeast two-hybrid analysis (unpublished work). Together, the results and the structure model suggest that Val-260 and Tyr-263 might constitute part of the putative binding surface for Spx.
The mechanism of repression proposed here for Spx is somewhat similar to certain eukaryotic factors such as Polycomb (46, 47) and the Groucho family members (48) of corepressors that exert negative control by interaction with the basal transcriptional machinery or mediator (49). However, these factors require a DNA-binding partner to form a repressing complex and none of them target the core subunit's RNAP, features that are in sharp contrast to Spx.
Microarray analysis revealed that spx is induced by heat shock (50) and phosphate limitation (51). Western blots showed high levels of Spx protein after 50°C incubation (unpublished data). Spx protein might accumulate as a result of ClpXP titration by damaged proteins, which diverts the protease from its normal task of removing Spx. This results in Spx-dependent repression of genes that are not needed for coping with harsh conditions, while allowing the stress response to be mobilized. Genes, such as clpP, clpC, dnaK, etc., that are induced under heat shock would not be affected by Spx because their induction is triggered by inactivation of the repressors CtsR and HrcA (52) and not by the action of positive regulatory factors that interact with RNAP.
Spx-dependent negative control likely extends to a wide range of activators considering that most of the pleiotropic phenotypes attributed to clpX and clpP mutations are alleviated by the spx null mutations or the rpoAcxs missense mutations. The conservation of Spx among low-GC content Gram-positive bacteria is evidence for its important role in the bacterium's decision-making process when harsh conditions are encountered.
Acknowledgments
We are grateful to Hirofumi Yoshikawa for his kind gift of the yeast two-hybrid/B. subtilis genomic library, Kürsad Turgay and Alan Grossman for valuable discussion, Richard Losick, Rick Gourse, and Chester Price for helpful comments and critical reading of the manuscript, and Yi Zhu for technical assistance. M.L. is supported by a Royal Thai Government Scholarship. This research was supported by National Institutes of Health Grant GM45898 and National Science Foundation Grant MCB0110513.
Footnotes
-
↵* To whom correspondence should be addressed. E-mail: pzuber{at}bmb.ogi.edu.
-
This paper was submitted directly (Track II) to the PNAS office.
Abbreviations
- RNAP,
- RNA polymerase;
- CTD,
- C-terminal domain;
- EMSA,
- electrophoretic mobility-shift assay
- Received December 16, 2002.
- Copyright © 2003, The National Academy of Sciences
References
A regulatory protein that interferes with activator-stimulated transcription in bacteria
Sign up for Article Alerts
Jump to section
You May Also be Interested in
Adaptations in heart structure and function likely enabled endurance and survival in preindustrial humans.
Image courtesy of Pixabay/Skeeze.
Viscoelastic carrier fluids enhance retention of fire retardants on wildfire-prone vegetation.
Image courtesy of Jesse D. Acosta.
Water requirements may make desert bird declines more likely in a warming climate.
Image courtesy of Sean Peterson (photographer).
QnAs with NAS member and plant biologist Sheng Yang He
Image courtesy of Sheng Yang He.