Identification of sigma factors for growth phase-related promoter selectivity of RNA polymerases from Streptomyces coelicolor A3(2)
Identification of sigma factors for growth phase-related promoter selectivity of RNA polymerases from Streptomyces coelicolor A3(2)Ju-Gyeong Kang, Mi-Young Hahn, Akira Ishihama1 and Jung-Hye Roe*
Department of Microbiology, College of Natural Sciences, and Research Center for Molecular Microbiology, Seoul National University, Seoul 151-742, Korea and 1Department of Molecular Genetics, National Institute of Genetics, Mishima, Shizuoka 411, Japan
Received April 28, 1997;Accepted May 19, 1997
ABSTRACT
We examined the promoter selectivity of RNA polymerase (RNAP) from Streptomycescoelicolor at two growth phases by in vitro transcription. Distinct sets of promoters were preferentially recognized by either exponential or stationary phase RNAP. No change in molecular weight or net charge of the core subunits was observed, suggesting that the associated specificity factors determined phase-specific promoter selectivity of the holoenzyme. Five different specificity factors and their cognate promoters were identified by in vitro holoenzyme reconstitution and transcription assays. [sigma]66 ([sigma]hrdB) and [sigma]46 ([sigma]hrdD) recognized promoters (rrnDp2and dagAp4for [sigma]66, actII-orf4p and whiBp2for [sigma]46)preferentially transcribed by the exponential phase RNAP. [sigma]52 recognized promoters (dagAp3and actIIIpx1)preferentially transcribed by the stationary phase RNAP. [sigma]28([sigma]sigE)recognized promoters (hrdDp1, whiBp1and dagAp2)transcribed equally by both RNAPs. A novel 31 kDa specificity factor recognized actIIIpx2, glnRp2and hrdDp2promoters preferentially transcribed by the stationary phase RNAP. This factor was isolated from the stationary phase RNAP and reconstituted holoenzyme in vitro as a sigma factor. The N-terminal sequence suggests that it is a novel factor. By examining phase-specific promoter recognition pattern we can predict that holoenzyme E[sigma]52 and E[sigma]31 activities are higher in the stationary phase, whereas E[sigma]66 and E[sigma]
46 activities are higher in the exponential phase. Possible promoter sequences recognized by some of these sigma factors were suggested.
INTRODUCTION
Various [sigma]factors in eubacterial RNA polymerase (RNAP) confer promoter selectivity on the core RNAP, thereby enabling the transcription pattern to change in response to environmental cues. At least eight different [sigma]factors have been found in Streptomyces coelicolor,which undergo complex mycelial growth and sporulation (1 ,2 ). Four homologues of Escherichia coli rpoD gene called hrdA-D were cloned and characterized (3 -5 ). The hrdB gene has been proved to be essential for the survival of S.coelicolor,whereas the functions of hrdA, -C and -D genes are not clear (6 ). Recently, it was demonstrated that HrdD actually functions as a sigma factor to transcribe redD and actII-orf4 promoters in vitro (7 ).Two sigma factors have been implicated in the control of the morphological differentiation of this organism. The whiG gene encodes an alternative [sigma]factor required for triggering the onset of sporulation whereas the sigF gene is required for normal spore maturation (8 ,9 ). In vitro transcription studies revealed that at least three different forms of RNAP holoenzyme exist, which are able to recognize the promoters of dagA gene encoding agarase (3 ,10 ).[sigma]hrdB ([sigma]66),the major vegetative sigma factor, recognizes dagAp4as well as the veg gene promoter of Bacillus subtilis. [sigma]52was identified as a [sigma]factor which recognizes dagAp3. [sigma]sigE ([sigma]28),the Streptomyces homologue of a [sigma]factor family thought to regulate extracytoplasmic functions (11 ), recognizes dagAp2. However, the promoter selectivity of various sigma factors as well as the relative abundance and/or activity under different growth conditions and during differentiation have not been systematically investigated.
In E.coli,the RNAP holoenzyme changes in the molecular characteristics of both core enzyme (E) and the associated [sigma]factors as cells enter into the stationary phase. More than 30genes which are induced upon entry into stationary phase are controlled by the stationary phase-specific [sigma]factor, [sigma]rpoS ([sigma]38). The intracellular level of [sigma]38 increases to 30%of the level of major sigma factor [sigma]70 upon entry into stationary phase whereas the level of [sigma]70 stays constant at the stationary phase (13 ). Modified forms of core RNAP were isolated from E.coli,each being separable by phosphocellulose column chromatography and exhibited different promoter selectivities from the exponential phase RNAP (14 ).
Streptomycete genes thus far identified exhibit a wide diversity in promoter sequences and transcription patterns. Among 139 Streptomycete promoters previously compiled, ~20% have sequence elements similar to those recognized by E.coli E[sigma]70 (15 ). The rest of the promoters contain enormously diverse sequences and the cognate [sigma] factors recognizing them have been hardly identified. In order to probe and characterize structural and functional changes in RNA polymerases in S.coelicolor cells during growth transition from the exponential to stationary phases of liquid culture, we prepared in this study RNAPs from the two growth phases and examined their promoter selectivity using a variety of promoters. As a result, the promoters could be classified on the basis of recognition patterns by the two RNAPs. Furthermore, by using invitro transcription assays by holoenzymes reconstituted from core enzyme and proteins eluted from the gel, we identified several [sigma] factors including a novel [sigma] factor.
MATERIALS AND METHODS
Bacterial strains and culture conditions
Streptomyces coelicolor A3(2) M145 cells were grown in YEME medium (16 ) containing 5 mM MgCl2 and 10% sucrose in a fermenter aerated at 0.5 vol air/vol media/min and agitated at 250 r.p.m. at 30oC. Freshly grown seed culture (200 ml) was inoculated to 4 l culture broth in 5 l fermenter. The mycelium was harvested from the fermenter at 12 h after inoculation for the exponential phase and 32 h for the stationary phase culture, and was stored at -70oC until use.
Preparation of RNA polymerase
RNA polymerase was purified from cell pellets by slight modifications of the procedures developed for the purification of E.coli RNAP (17 -19 ). Approximately 20 g of wet mycelial cells were disrupted in a French pressure cell at 1000 p.s.i. The crude extract was subjected to polyethyleneimine (PEI) precipitation, salt fractionation and ammonium sulfate precipitation. The sample was applied to a heparin Sepharose CL-6B column (Pharmacia) which was washed with TGED buffer (10 mM Tris-HCl, pH 7.9 at 4oC, 0.1 mM EDTA, 0.1 mM DTT and 10% glycerol) containing 0.2 M NaCl. Proteins were eluted using a gradient of 0.2-0.7 M NaCl in TGED buffer. The fractions containing RNAP activity were pooled and concentrated by ammonium sulfate precipitation. The precipitate was dissolved in TED buffer (TGED without glycerol) and dialyzed against storage buffer (10 mM Tris, pH 7.9 at 4oC, 10 mM MgCl2, 0.1 M KC1, 0.1 mM EDTA, 1 mM DTT and 50% glycerol). For further purification, the proteins were diluted in TED buffer and chromatographed through Superdex HR 200 (Pharmacia) with TGED containing 0.3 M NaCl and MonoQ (Pharmacia) anion exchange columns using a gradient of 0.3-0.6 M NaCl in TGED buffer. The trailing fractions (higher salt eluates) of the RNAP peak eluted from Mono-Q column were used as core RNAP.
Preparation of template promoter fragments
The promoters examined in this study are listed in Table 1 . DNA fragments containing the dagA (20 ), rrnD (21 ), glnR (22 ) and actII-orf4 (23 ) promoters were PCR-amplified from the genomic DNA of S.coelicolor A3(2) strain M145 using the following pairs of oligonucleotide primers: ATCAGCCGGAGTGAACCGTT and GAGTGCGACGGCACTCCAC for dagA;CTGGCCTACGTCTACGTTGT and CGATCAGGTCGGGGTATCAA for rrnD;CGACGAACACCAGGTCAG and GCAGCAGAGAACTCATCG for glnR; GAGGACCCAGCCGTATCAG and GTACACGTACGTCTGCAG for actII-orf4. The resulting PCR products were cloned into pUC18 or pTZ-18R. The glkA, hrdD and whiB genes were provided by Dr M.Bibb at John Innes Center. The actI/III gene was provided by Dr S.K.Hong at Myeong Ji University. Each promoter fragment was subcloned from the larger parental plasmid into pUC18; glkA from pIJ2420 (24 ), hrdD from pIJ2036 (25 ), whiB from pIJ558 (26 ) and actI/III from pIJ2303 (27 ). Promoter DNA fragments ranging from 346 to 578 nucleotides in length were cut out from the resulting recombinant pUC18 plasmids using the restriction enzymes listed in Table 1 .
In vitro transcription assay
In vitro run-off transcription assay was performed using the combined conditions of Fujita et al. (18 ) and Buttner et al. (10 ). RNAP (1.5 pmol) was incubated at 30oC for 5 min in 15 [mu]l transcription buffer [40 mM Tris pH 7.9, 10 mM MgCl2, 0.6 mM EDTA, 0.4 mM potassium phosphate, 1.5 mM DTT, 0.25 mg/ml BSA and 20% (v/v) glycerol] with 0.15 pmol of template DNA. RNA synthesis was initiated by the addition of 3 [mu]l of substrate mixture containing 2 [mu]Ci[[alpha]-32P]CTP (400 Ci/mmol) and 0.4 mM each of UTP, ATP and GTP. An aliquot of 3 [mu]l of heparin (0.1 mg/ml final concentration) was added after 2 min to prevent further reinitiation and the incubation was continued for 5 min before adding 2 [mu]l of cold CTP (1.0 mM final concentration). After 10 min incubation, the reaction was terminated by adding 50 [mu]l of stop solution (375 mM sodium acetate pH 5.2, 15 mM EDTA, 0.15% SDS and 0.1 mg/ml calf-thymus DNA). Transcripts were precipitated with ethanol, resuspended in formamide sample buffer [80% (v/v) formamide, 8% glycerol, 0.1% SDS, 8 mM EDTA, 0.01% bromophenol blue and 0.01% xylene cyanol] and electrophoresed on 5% polyacrylamide gel containing 7 M urea.
Holoenzyme reconstitution assay
About 1 mg of RNAP partially purified through heparin Sepharose CL-6B chromatography was subjected to preparative SDS-PAGE. The gel was stained with 0.25 M KCl and 1 mM DTT and proteins were eluted by the procedure of Hager and Burgess (28 ). Gel slices (2-8 mm) were cut out and placed in dialysis bags containing 300 [mu]l of 2:5 diluted SDS-running buffer and 3 [mu]l of 10 mg/ml BSA. Electroelution from the gel was carried out at 30 mA for 3 h at 4oC. The eluates were recovered and precipitated by adding 1.2 ml of cold acetone. Renaturation of the eluted proteins was performed in the presence of GroEL as described (3 ). GroEL was purified from E.coli containing pGroELS (a gift from Dr Lorimer, E.I.Dupont de Nemours and Co.) through DEAE-Sepharose, heparin Sepharose CL-6B, Superose 6B (Pharmacia) column chromatographies (29 ,30 ). The acetone precipitate was dissolved thoroughly in 20 [mu]l of 6 M guanidine-HCl buffer (20 mM Tris-HCl pH 7.8, 150 mM NaCl, 5 mM DTT, 0.1 mM EDTA, 6 M guanidine-HCl) and incubated at room temperature for 20 min. The resuspended pellet was then diluted into 1 ml of GroEL incubation buffer (50 mM Tris pH 7.8, 12 mM MgCl2, 9 [mu]g/ml GroEL) and incubated at 22-25oC for 2 h. It was then dialyzed against renaturation buffer [20 mM Tris-HCl, pH 7.8 at 4oC, 10 mM MgCl2, 10 mM KCl, 0.1 mM EDTA and 50% (v/v) glycerol] for 12-16 h with one change of the buffer. An aliquot of 5 [mu]l of the renatured proteins was added to ~1 pmol of core RNA polymerase and the mixture was incubated on ice for 10 min. After addition of 0.15 pmol of DNA template, the mixtures were incubated at 30oC for 30 min and subjected to in vitro transcription assay.
RNAP fractions eluted from Mono-Q column with peak levels of the hrdD transcribing activity were pooled, electrophoresed, and electroblotted onto polyvinylidene difluoride membrane (pore size, 0.1 [mu]m, Millipore) in CAPS buffer [10 mM 3-(cyclohexylamino)-1-propanesulfonic acid, 10% methanol, pH 11.0]. The first 22 residues of [sigma]31and 10 residues of [sigma]28were determined by Edman degradation using Procise Protein Sequencing System (Applied Biosystems).
RESULTS
Collection of template promoters
To examine promoter selectivities of RNAPs in vitro,we prepared a number of DNA fragments, each carrying specific promoter(s) from various Streptomycete genes which are involved in such diverse functions as transcription or translation (hrdD, rrnD),carbon and nitrogen metabolism and their regulation (dagA, glkA, glnR)and morphological and physiological differentiation (whiB, actI/III, actII-orf4). Most of these genes have multiple promoters as listed in Table 1 . Of 21promoters examined, only dagAp4, rrnDp2and whiBp2have sequences similar to E.coli E[sigma]70-recognized promoters (15 ). The cognate [sigma] factor recognizing each promoter in our collection has not been identified except for the dagA and actII-orf4 promoters (3 ,7 ,10 ). The collection of promoter templates provides a representative example of the promoter sequence heterogeneity in S.coelicolor and thus was expected to serve as diverse templates to probe the RNAPheterogeneity.
Differential promoter selectivities of RNAPs from two growth phases of S.coelicolor
When S.coelicolor A3(2) M145cells were grown in YEME medium in a fermenter, they entered into the stationary phase ~20h after inoculation, as judged by the optical density of the culture. RNApolymerases were prepared from either exponentially growing cells (12h culture) or the stationary phase cells (32h culture) and compared for their promoter selectivity. Partially purified RNAPsat the step of heparin Sepharose CL-6B column chromatography were used for in vitro transcription assay. Both enzyme preparations contained [beta]', [beta] and [alpha] subunits as the major components but a number of additional proteins were identified by SDS-PAGE, including at least five sigma subunits (see below for details).
Figure 1 demonstrates representative transcription pattern of promoters from eight genes. Transcripts in vitro of the sizes expected from transcription initiation sites in vivo (Table 1 ) were observed for all the test promoters except for the actI/III and rrnDp1promoters. Promoters such as actII-orf4p, dagAp1, p4, glkAp23, glnRp3, rrnDp3, p4and whiBp2 promoters were preferentially recognized by the exponential phase RNAP whereas promoters actIIIpx1, px2, dagAp3and hrdDp2were preferentially recognized by the stationary phase RNAP.
Examination of growth-coupled changes in core RNAP
In order to find the structural basis for the observed differences in promoter selectivities, we examined the chromatographic behaviors and subunit compositions of the RNAPs prepared from the exponential and the stationary phase. In E.coli,upon transition from exponential to stationary growth phase, RNAP has been found to convert into different forms which could be separated by phosphocellulose column chromatography (14 ,32 ). They suggested that the change is due to association of poly- or oligo-phosphates with core subunits (Kusano,S. and Ishihama,A., in preparation). The altered stationary phase forms of RNAP manifested different promoter recognition properties from the exponential phase RNAP. In contrast to E.coli, S.coelicolor RNAPs were eluted as a single, broad peak from the phosphocellulose column. The conductivity measurement of the peak fractions revealed that the RNAPs from both phases were eluted at the same salt concentration (data not shown). Thus both S.coelicolor RNAPs seem to have similar net charge. Two dimensional SDS-PAGE analysis of both enzymes also showed no difference in the migration pattern of core subunits, suggesting no change in the molecular weight and net charge of each subunit. Therefore, it seems most likely that the two RNAP preparations differ not in core enzyme subunits but in the associated specificity factor.
Identification of the RNAP-associated specificity factors
We next set out to find the specificity factors associated with the RNAP from two growth phases of S.coelicolor. RNAP from exponentially growing cells was partially purified through heparin-Sepharose column. In order to identify [sigma] factors, the associated proteins were separated on SDS-PAGE, eluted from gel slices, renatured and mixed with the core enzyme to test the activity of promoter-directed transcription. Such holoenzyme reconstitution experiments have been successfully employed to identify [sigma] factors in several bacterial systems. Successful renaturation of the denatured [sigma]hrdBin the presence of chaperonin, GroEL, has been reported (3 ).
Core enzyme was obtained from highly-purified RNAP preparations following Mono Q column chromatography as previously described (10 ). The trailing fractions of the peak having high non-specific RNA synthesizing activity but little promoter-specific activity were taken as core enzyme. For isolation of [sigma] factors, the partially-purified RNAP sample was run on SDS-8% PAGE and proteins 20-70 kDa in size were eluted from gel slices and renatured as described in Materials and Methods. Figure 2 A demonstrates SDS-PAGE pattern, in which the positions of core subunits and of some known sigma factors predicted from their observed molecular weights are marked within the corresponding gel slices. The eluted proteins were renatured and mixed with the highly purified core enzyme for run-off transcription assay with various promoters. The results are shown in Figure 2 B. The core enzyme alone was inactive in transcription of all the test promoters, but regained the activity upon the addition of gel eluates. Only a subset of the transcripts observed in Figure 1 were detected with the reconstituted enzymes, suggesting that some as yet unidentified factors were not recovered from the gel or not renatured under the condition employed.AB
Growth-related variation in the relative activity level of various sigma factors
After comparison of the selective recognition of each promoter by each of the two partially purified holoenzyme preparations from two growth phases (Fig. 1 ; Table 2 )or each of the reconstituted holoenzymes (Figs 2 -4 ), we discovered that a good correlation exists between the growth phase-related activity of promoters and the selective recognition properties of promoters by different [sigma] factors (Table 3 ). [sigma]52and the putative [sigma]31 only recognized those `stationary' promoters, suggesting that E[sigma]52 and E[sigma]31 are more abundant (or active) in the stationary than in the exponential phase. The promoters recognized by [sigma]sigEwere all transcribed to similar extent by RNAPs from both phases. Therefore the relative activity of E[sigma]sigEis likely to be similar in both phases. E[sigma]hrdBand E[sigma]hrdDrecognized promoters that are preferentially transcribed by the exponential RNAP in vitro. Therefore, it is reasonable to predict that E[sigma]hrdBand E[sigma]hrdD may be relatively more abundant (active) in the exponential phase.
lsolation of 31 kDa protein and reconstitution of E[sigma]31 holoenzyme
Since the putative [sigma]31 is predicted to be more abundant in the stationary phase, we purified RNAP holoenzyme from the stationary phase mycelial cells. The RNAP activity measured by non-specific assay was eluted as a single broad peak from the heparin-Sepharose column. The RNAP containing [sigma]hrdB as detected by dagAp4-dependent transcription is more enriched in the leading fractions of the peak (lower salt eluate) whereas the RNAP with hrdDp2-transcribing ([sigma]31) activity is more enriched in the trailing fractions. We therefore applied the higher salt eluates to further purification steps through Superdex HR 200 and Mono-Q anion exchange column chromatographies as described above. Figure 5 A demonstrates the SDS-PAGE pattern of each step of enzyme preparation. Upon further purification, we could detect the stained band of putative [sigma]31 as well as [sigma]28 protein. We determined the N-terminal peptide sequences of both 31 and 28 kDa bands. The 28kDa protein contained a sequence of GEVLXXEEYV which matches well with the S.coelicolor SigE peptide (11 ), confirming that the 28kDa protein is indeed [sigma]sigE. The N-terminal peptide sequence of 31 kDa protein was determined to be XGTDAGTEHGQAEQPEGRGTXA, which did not match with any known sequence.AB
Sequences of promoters recognized by each sigma factor
The promoters examined in this study were classified according to their cognate sigmas and compared with respect to the nucleotide sequences. Figure 6 represents the comparison of promoter sequences within each group, including those of known promoters whose recognition by each sigma factor was previously verified. The consensus sequences of promoters recognized by E[sigma]hrdB,E[sigma]hrdD,E[sigma]sigEand E[sigma]31 have been proposed. When the whiBp2promoter was compared with two other promoters known to be recognized by [sigma]hrdD(7 ), the -10 and -35 regions were quite similar. The comparison of three promoters recognized by E[sigma]hrdDdemonstrated that the plausible consensus sequence is tTGAcN-N17-18-tatNaT which is more degenerate than E[sigma]hrdBrecognition sequence, TTGaCA-N17-18-TAgaaT. Conservation of amino acids in the promoter binding regions of [sigma]hrdBand [sigma]hrdDsuggest that they may recognize similar DNA sequences (2 ). The promoter sequences of hrdDp1and whiBp1matches well with dagAp2known to be recognized by [sigma]sigE. All these promoters contain Gg/cAAC at -35 region and TC dimer at -10 region with 19 nucleotides spacing. This is consistent with the consensus promoter sequences for [sigma]sigEpredicted from the computer matches (33 ). Some common sequences found among promoters recognized by [sigma]52or [sigma]31are also presented in Figure 6 . Promoters recognized by E[sigma]31 contained GGgcag sequences in -35 region and Gttgc in -10 region with 17-18 nucleotides spacing. The significance of these sequence matches need further investigation.
DISCUSSION
We monitored the changes in promoter selectivity of RNAP at two growth phases using in vitro transcription assay on various promoters. The difference in promoter selection by partially purified holoenzyme from two growth phases reflects the difference in the distribution of sigma factors within the holoenzyme preparation. The abundance of a particular holoenzyme is primarily determined by the availability of its own sigma factor. However, depending on the concentration and the binding affinity of other competing sigma factors, the relative abundance is bound to change significantly. Our data predicts that E[sigma]hrdBand E[sigma]hrdDbecome less abundant whereas E[sigma]52 and E[sigma]31 become more abundant in the whole holoenzyme population at the stationary phase than at the exponential phase. As for E[sigma]hrdBa preliminary result suggests that its decrease in the stationary phase is not due to the decrease in the intracellular concentration of [sigma]hrdBjudging from the relatively constant level of [sigma]hrdBby immunoblotting (data not shown). In B.subtilis, the level of major sigma factor [sigma]A maintained at a constant level throughout vegetatively growing or sporulating cells (34 ),whereas the level of purifiable E[sigma]A decreases as the cells proceed into sporulation (35 ). Putative inhibitor of [sigma]A at the late sporulation stage has been suggested to regulate the level of functional E[sigma]A holoenzyme (36 ). In E.coli,the level of the major sigma factor [sigma]70(rpoD)stays constant at the same level at all growth phases. However, the level of [sigma]38 increases significantly at the stationary phase and thereby reduces the relative level of E[sigma]70 holoenzyme (13 ). In this respect, it is likely that the decrease in E[sigma]hrdBin the stationary S.coelicolor cells may be due to the reduced relative abundance by competition or the reduced activity of [sigma]hrdB.
The observation of growth-related selectivity in vitro iscorrelated with the observation in vivo for several promoters. For example, the promoters known to be highly expressed in the exponentially growing cells in vivo were transcribed predominantly by the RNAP isolated from the exponential phase. These include rRNA promoters (rrnDp3, p4)and a sporulation gene promoter (whiBp2) (37 ,38 ). rrnDp3and p4are the strong promoters of the rrnD gene contributing >90% of its transcripts in rapidly growing cells. Following nutritional shift-down, transcription initiation decreases dramatically at all rrnD promoters (37 ). In vitro transcription pattern of rrnDp3and p4 isconsistent with the in vivo observation. On the other hand the efficient transcription of rrnDp2 in vitro and its relatively constitutive expression is different from its expression pattern in vivo. Transcription in vivo from whiBp2isknown to increase when aerial mycelial growth begins on surface culture, whereas the level of p1 transcript exhibits no obvious correlation with the developmental stage. In liquid culture, however, the level of p2 transcript is high especially during the exponential growth, whereas the level of p1 transcript is low and varies little with growth phase (38 ). The transcription of whiBp2 in vitro is consistent with the observation in vivo for liquid culture.
Promoter specificities of RNAP defined in vitro, however, may not directly reflect in vivo expression of the promoter. There are several examples where the in vitro promoter recognition pattern is not verified in vivo. [sigma]hrdDrecognizes actII-orf4pand redDppromoters in vitro, whereas hrdD null mutants do not affect biosynthesis of actinorhodin and undecylprodigiosin (red antibiotics) (6 ,7 ). Similarly, [sigma]sigErecognizes phsA promoter in S.antibioticus in vitro, but sigE disruption does not abolish phsA transcription in vivo (39 ). These discrepancies could by due to either recognition of those promoters by multiple sigma factors in vivo,or the lack of any relationship between in vitro transcription conditions currently employed and the intracellular environment. The presence of transcriptional regulators in vivo could also influence the promoter selectivity of various holoenzymes. In order to resolve these apparent contradictions, further systematic experiments need be pursued both in vitro and in vivo.
Several promoters examined in this study were not transcribed by the reconstituted RNAP, indicating that the cognate specificity factors were not recovered from the gel. In particular, the strong promoters rrnDp3and p4 were not recognized by the reconstituted E[sigma]hrdB. This suggests that an as yet unidentified vegetative holoenzyme(s) other than E[sigma]hrdB partakes their transcription or an additional transcriptional activator(s) is required for E[sigma]hrdBto transcribe those promoters. As an attempt to identify the missing factor(s), we pooled all the renatured polypeptides from total gel slices and performed the reconstitution and transcription. Still we were not able to observe any rrnDp3and p4 transcripts in vitro (data not shown).
Enormously diverse sequences of Streptomycete promoters suggest that there are still large numbers of the cognate [sigma] factors and associated transcription factors remaining to be identified. The discovery of [sigma]31as a novel sigma factor recognizing actIIIpx2, glnRp2and hrdDp2promoters demonstrates that the holoenzyme reconstitution assay is a useful tool to identify those factors in Streptomycetes. By employing wider spectrum of promoters and in vitro transcription conditions it is possible to identify more specificity factors. How these specificity factors function in regulating diverse promoters in vivo and thereby perforrn coordinated gene expression remains a formidable task to challenge.
ACKNOWLEDGEMENTS
We thank Dr N.Fujita for his invaluable help in protein purification, in vitro transcription assay and for detailed advice on this work. Much gratitude is directed to Dr M.J.Bibb for providing glkA, hrdD and whiB promoters; Dr S.-K.Hong for providing actI/III promoters; Dr M.J.Buttner for providing HrdB overexpression construct and helpful comments; Dr K.Tanaka for providing antibody against HrdB; J.-B.Bae for technical assistance. This work was supported partly by a grant from both KOSEF and JSPS for Korea-Japan collaborative research between J.-H.R. and A.I. and a grant from the SRC (Research Center for Molecular Microbiology).
REFERENCES
1 Westpheling,J., Ranes,M. and Losick,R. (1985) Nature313, 22-27.MEDLINE Abstract
