In bacteria, the RNA polymerase holoenzyme comprises a five-subunit core enzyme and a dissociable subunit, sigma factor, which is responsible for transcriptional initiation. The filamentous bacterium Streptomyces griseus has 52 sigma factors, including one essential ‘principal’ sigma factor (σHrdB) that is responsible for the transcription of housekeeping genes. Here we characterized an alternative sigma factor (σShbA), which is highly conserved within the genus Streptomyces. A σShbA-deficient mutant showed a severe growth defect and transcriptome analysis indicated that many housekeeping genes were downregulated in response to insufficient σShbA production. Biochemical and genetic analyses proved that σShbA is a major determinant of transcription of the σHrdB gene. This observation of a principal sigma factor being governed by another sigma factor throughout growth is unprecedented. We found that increasing σShbA production with mycelial growth maintained a high σHrdB level late in growth. Furthermore, a hrdB-autoregulatable σShbA-deficient mutant, in which the principal sigma factor gene can be transcribed by RNA polymerase containing σHrdB itself, showed several defects: rapid mycelial lysis in stationary phase in liquid culture and delayed morphological development and impaired streptomycin production in solid culture. From these observations, we discuss the biological significance of control of σHrdB by σShbA in S. griseus.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
In bacteria, an RNA polymerase (RNAP) holoenzyme, comprising a five-subunit core RNAP (subunit composition, α2ββ′ω) and a dissociable subunit, sigma (σ) factor, is responsible for transcriptional initiation. The sigma factor contains many, if not all, of the promoter recognition determinants and confers promoter specificity to the RNAP holoenzyme. All bacteria possess at least one essential ‘principal’ sigma factor that transcribes the housekeeping genes required for cell viability, and most bacteria have alternative sigma factors that can redirect RNAP to initiate transcription at alternative promoters after substituting the principal (primary) sigma factor (Gruber and Gross, 2003). In Escherichia coli (Burton et al., 1983), Bacillus subtilis (Wang and Doi, 1987), Pseudomonas aeruginosa (Aramaki and Fujita, 1999), Pseudomonas putida (Fujita et al., 1995), Mycobacterium tuberculosis (Rodrigues et al., 2007) and Microcystis aeruginosa (Asayama et al., 1999), it was experimentally confirmed that each principal sigma factor gene is transcribed by RNAP containing the principal sigma factor itself. Such autoregulation of principal sigma factor was also predicted in the following bacterial species: Synechococcus sp. strain PCC7942 (Tanaka et al., 1992), Listeria monocytogenes (Metzger et al., 1994), Rhodobacter capsulatus (Pasternak et al., 1996), Brevibacterium flavum (Halgasova et al., 2001) and Corynebacterium glutamicum (Pátek and Nešvera, 2011). The vast majority of alternative sigma factors belong to the ECF (extracytoplasmic function) subfamily (Helmann, 2002; Staroń et al., 2009). ECF-subfamily sigma factors consist of only two regions (regions 2 and 4) of four conserved regions of the σ70 family, which are required for both RNA polymerase interaction and recognition of the promoter sequence. ECF-subfamily sigma factors usually activate transcription of specific genes in response to environmental changes.
A soil-dwelling Gram-positive bacterium, Streptomyces undergoes a complex life cycle resembling the morphological development of filamentous fungi. Recent genome sequence analysis of several Streptomyces species revealed that Streptomyces as a whole have a large number of sigma factors (Ômura et al., 2001; Bentley et al., 2002; Ikeda et al., 2003; Ohnishi et al., 2008). Fifty-two sigma factors, for example, are encoded on the chromosome of Streptomyces griseus; 39 of them belonging to the ECF subfamily (Ohnishi et al., 2008). However, their functions remain largely unknown. Highly conserved sigma factors among Streptomyces species seem to play crucial roles in the biology of Streptomyces. An ECF-subfamily sigma factor σAdsA in S. griseus (Yamazaki et al., 2000) and its orthologue σBldN in Streptomyces coelicolor A3(2) (Bibb et al., 2000), for example, are required for aerial mycelium formation. An ECF-subfamily sigma factor, σR, induces genes for thiol-oxidative stress response in S. coelicolor A3(2) (Paget et al., 1998; 2001; Park and Roe, 2008; Kim et al., 2012). An alternative sigma factor σWhiG, which does not belong to the ECF-subfamily, is necessary for the onset of sporulation in S. coelicolor A3(2) (Chater et al., 1989). Orthologues of σR and σWhiG are also conserved in S. griseus and Streptomyces avermitilis with more than 90% amino acid identity (Ômura et al., 2001; Ikeda et al., 2003; Ohnishi et al., 2008).
Here we conducted functional analysis of another highly conserved ECF-subfamily sigma factor (SGR2758) with unknown function in S. griseus. We determined that SGR2758 is a major determinant for the transcription of hrdB, the gene encoding the principal sigma factor σHrdB. We termed SGR2758 σShbA (sigma factor for hrdB). This is the first report of a principal sigma factor being governed by another sigma factor throughout growth. The biological significance of control of σHrdB by σShbA is discussed. Our findings provide new insight into the regulation of a principal sigma factor gene in bacteria and the adaptation to stationary phase in Streptomyces.
σShbA has a pivotal role in cellular growth of S. griseus
Of 52 sigma factors in S. griseus, six sigma factors (σHrdB, σAdsA, σWhiG, σR, σShbA and SGR3370) are highly conserved in S. coelicolor A3(2) and S. avermitilis (having more than 90% amino acid identity) (Figs S1A and S2A). Among them, only σShbA and SGR3370 had not been previously analysed. σShbA is widely conserved within other genera of actinobacteria, while SGR3370 is found in very few actinobacterial species (Figs S1B and S2B). Gene organization of the shbA locus is highly conserved among Streptomyces species (Fig. S1C). Two genes involved in morphological development are located just upstream of shbA: bldM, encoding a putative transcriptional regulator that is required for aerial mycelium formation (Molle and Buttner, 2000), and whiD, encoding a protein that is involved in spore maturation (Molle et al., 2000). Some neighbouring genes around SGR3370 are also conserved among Streptomyces species (Fig. S2C).
To examine the functions of σShbA and SGR3370, shbA and SGR3370 null mutant strains (ΔshbA and ΔSGR3370) were generated (Figs S1D and E and S2D and E). The ΔshbA mutant formed very small, upheaving and non-differentiating colonies on nutrient-rich YMPD agar (Fig. 1A and C), while ΔSGR3370 showed similar growth to the wild-type (wt) strain when cultivated on YMPD agar (Fig. 1B). Therefore, we examined only the ΔshbA mutant further. Characteristically, the vegetative hyphae of ΔshbA were unable to penetrate sufficiently into the agar medium. Therefore, colonies of the ΔshbA mutant upheaved into the air and the colony surface was rough. When a single copy of shbA was integrated into the chromosome of the ΔshbA mutant (generating the ΔshbA::pTYM19-shbA strain), colony morphology and growth were restored to normal (Fig. 1A). In contrast, the ΔshbA mutant formed small, thin, but differentiating colonies on a supplemented minimal medium (SMM) (Fig. 1D and E). These results indicate that σShbA plays an important role in mycelial growth but is not necessarily required for morphological development or sporulation. The pivotal role of σShbA in cellular growth was confirmed by the observation that the growth rate of the ΔshbA mutant in submerged culture was much slower than that of the wt strain (Fig. 1F).
Expression of shbA is enhanced in the late growth stage
We used quantitative RT-PCR analysis to examine the time-course of shbA transcription, by using 16S rRNA as an internal control. Transcription of shbA was increased along with the growth in both solid and liquid culture, although the amounts of the shbA transcript in liquid culture were much less than those in solid culture for some reason (Fig. 2A, upper panels). shbA was revealed to have two possible transcriptional start points, as determined by S1 nuclease mapping (Fig. S3A), although the number of shbA transcripts was very small, and relatively large quantities of total RNA (10 times greater than usual) were required for detection. We determined two possible transcriptional start points by high-resolution S1 nuclease mapping: p1 (corresponding to the first nucleotide of the translational start codon) and p2 (−143, taking p1 as +1) (Fig. S3B). There were no typical promoter sequences for housekeeping genes transcribed by σHrdB-containing RNAP (TTGACR-N16–18-TAGRRT, where R is A or G) in the appropriate regions upstream from these possible transcriptional start points (Fig. S3C), which might cause the very low level of shbA transcript. It is possible that the short transcript could be a processing product of the long transcript.
σShbA regulates many housekeeping genes
We cloned shbA with its upstream 250 bp and downstream 200 bp regions in a high-copy-number plasmid pIJ486, resulting in pIJ486-shbA. Although we intended to overproduce σShbA by introducing pIJ486-shbA into the ΔshbA mutant, the shbA transcripts in the recombinant strain were significantly less throughout growth than those in the wt strain, as determined by quantitative RT-PCR analysis (Fig. 2B). Probably because of the insufficient shbA transcripts, the recombinant strain (ΔshbA [pIJ486-shbA], square brackets denoting a plasmid-carrying state) formed smaller (but differentiating) colonies on YMPD agar (Fig. 1A). Growth of the ΔshbA [pIJ486-shbA] strain was delayed for approximately 6 h in submerged culture when compared with the wt strain (Fig. 1F). Although this result was unexpected (transcription of plasmid-born shbA appeared to be much lower than that of intact shbA on the chromosome), the ΔshbA [pIJ486-shbA] strain was used for the transcriptome analysis described below. Judging from the severe growth defect of the ΔshbA strain, we assumed that a huge number of genes should be differentially expressed between the wt and ΔshbA strains and therefore it should be very difficult to find σShbA-target genes by a comparative transcriptome analysis between these two strains. In contrast, we assumed that the ΔshbA [pIJ486-shbA] strain, which expressed shbA in a lower level than the wt strain, might be appropriate for such a transcriptome analysis, because the wt and ΔshbA [pIJ486-shbA] strains showed a similar growth rate during their exponential growth phase in submerged culture (Fig. 1F).
Transcriptomes of the wt and ΔshbA [pIJ486-shbA] strains were compared using a DNA microarray. For this purpose, RNAs were obtained from growing mycelia in submerged culture (without thiostrepton) at the early exponential phase; 24 h growth for the wt, and 30 h growth for the ΔshbA [pIJ486-shbA] strain (see Fig. 1F). We employed the following criteria to discriminate genes influenced by insufficient shbA transcription. Those that showed an increase or decrease in expression greater than 2.5-fold and a P-value of < 0.05 when analysed with a t-test (n = 3) were considered influenced. When these criteria were used, 209 genes were positively regulated and 212 were negatively regulated by σShbA (Tables S2 and S3). The accuracy of the DNA microarray analysis was confirmed by quantitative RT-PCR (Fig. S4). Using the COG database (http://www.ncbi.nlm.nih.gov/COG/), we classified the 209 genes that appeared to be positively regulated by σShbA, on the basis of their putative functions (Table 1). Many housekeeping genes, including genes involved in transcription and translation, were downregulated in response to the insufficient shbA transcription. This suggested that the severe growth defect in the ΔshbA mutant could be ascribed to a drastic decrease in the transcription of these housekeeping genes.
Table 1. Functional categories of genes positively regulated by σShbA
Intracellular trafficking, secretion and vesicular transport
Post-translational modification, protein turnover, chaperones
hrdB is transcribed by RNAP containing σShbA
In the DNA microarray analysis, we noticed that hrdB encoding the principal sigma factor σHrdB was downregulated in response to insufficient shbA transcription. Because σHrdB must be required for the transcription of housekeeping genes (Buttner et al., 1990; Shinkawa et al., 1995a), insufficient transcription of hrdB could cause decreased transcription of many housekeeping genes. We proposed, therefore, that σShbA would be directly required for the transcription of hrdB. To test this, we performed the following in vivo and in vitro experiments.
First, we performed qRT-PCR and Western blot analyses to examine the amounts of hrdB transcript and σHrdB protein, respectively, in the wt and ΔshbA strains. When 16S rRNA was used as an internal control, the amount of hrdB transcript was not constant in the wt strain; it was enhanced in the late growth stage in both liquid and solid culture (Fig. 2A, lower panels). In contrast, transcription of hrdB was greatly reduced throughout growth in the ΔshbA mutant compared with the wt strain (Fig. 3A). hrdB transcription was shown to occur from the known single promoter (Shinkawa et al., 1995b), which corresponds to the primary promoter p1 in S. coelicolor A3(2) (Buttner et al., 1990; Kim et al., 2012), in both wt and ΔshbA strains, by 5′-rapid amplification of cDNA ends (RACE). In contrast to the enhanced transcription of hrdB, the amount of σHrdB in the wt strain was almost constant throughout growth in liquid culture, as determined by Western blotting analysis using a newly generated anti-σHrdB antibody (Fig. 3B). We postulated that the enhanced transcription of hrdB could be required for maintaining the level of σHrdB in the late growth stage. As expected, the amount of σHrdB was greatly reduced in the ΔshbA mutant compared with the wt strain (Fig. 3B). Two bands were detected at the expected position for σHrdB in the Western blotting analysis, indicating that σHrdB could be partially processed; degradation of its N- or C-terminal portion is most likely.
Second, transcription of hrdB by an RNAP holoenzyme containing σShbA was shown by an in vitro transcription run-off assay (Fig. 3C). In this assay, we used recombinant σShbA and σHrdB, both of which were produced in E. coli and purified (Fig. S5), and a commercially available E. coli core RNAP. We used the rpoB promoter as negative and positive controls of σShbA and σHrdB-dependent transcription respectively. Note that transcription of rpoB, which encodes the RNAP β subunit, was downregulated in response to insufficient shbA transcription in the transcriptome analysis, but rpoB seemed to be transcribed by RNAP holoenzyme containing σHrdB but not by RNAP holoenzyme containing σShbA. As shown in Fig. 3C, the RNAP holoenzyme containing σShbA was able to initiate in vitro transcription at the hrdB promoter but not at the rpoB promoter. In contrast, RNAP holoenzyme containing σHrdB was able to initiate in vitro transcription at the rpoB promoter but not at the hrdB promoter.
Finally, in vivo binding of σShbA to the hrdB promoter was examined by chromatin affinity precipitation (ChAP) analysis. For ChAP analysis, we constructed C-terminally 12 × His-tagged σShbA-producing strain (ΔshbA::pTYM19-shbA-12his). Colony morphology and growth of this recombinant strain were the same as those of the wt strain (Fig. 1A), indicating that the His-tagged σShbA was fully functional in vivo. Using ChAP, we collected DNA fragments bound by σShbA in the ΔshbA::pTYM19-shbA-12his strain, and examined the constituents of specific regions using quantitative PCR analysis to estimate affinity precipitation efficiency. The hrdB promoter region (−85 to +229, taking the transcriptional start point as +1) could be amplified by PCR from the ChAP-ed DNA sample of the ΔshbA::pTYM19-shbA-12his strain, but not from that of the ΔshbA::pTYM19-shbA strain, in which σShbA was produced without the 12 × His tag (Fig. 3D). This confirmed that ChAP was efficiently performed. As shown in Fig. 3E, the affinity precipitation efficiency of the hrdB promoter region (−71 to −15) was significantly higher than that of upstream (−1188 to −1100) and downstream (+1059 to +1246) regions from the hrdB promoter; and also higher than that of the other promoters (for rpoB, rpsJ, rpmH/dnaA and rpsF) that are apparently downregulated in response to insufficient shbA transcription. These results indicate that σShbA-containing RNAP formed the transcription initiation complex specifically at the hrdB promoter.
From the results of the in vivo and in vitro experiments described above, we concluded that the σShbA-containing RNAP was mainly required for hrdB transcription.
Growth of the ΔshbA mutant was recovered to normal by the introduction of rpoB promoter-driven hrdB
As described above, we hypothesized that insufficient amounts of σHrdB would reduce the transcription of many housekeeping genes and cause the severe growth defect in the ΔshbA mutant. To increase the amount of σHrdB in the ΔshbA mutant, an additional copy of the hrdB gene was placed under the control of the rpoB promoter and integrated into the chromosome of the ΔshbA mutant, generating the ΔshbA::pTYM19-rpoBp-hrdB strain. As expected, this strain grew normally and formed fully developed colonies similar to those of the wt strain on YMPD agar (Fig. 1A). This result supported our hypothesis above and indicated that hrdB is the sole target of σShbA at least to account for the growth defect of the ΔshbA mutant.
Detailed phenotypic analysis of the ΔshbA::pTYM19-rpoBp-hrdB strain
In many bacteria, a principal sigma factor-encoding gene is usually transcribed by RNAP containing the principal sigma factor itself, as described in the Introduction. However in S. griseus, hrdB, encoding the principal sigma factor σHrdB, is transcribed by RNAP containing the alternative sigma factor σShbA throughout growth. Therefore, we questioned the biological significance of such a unique system in S. griseus. To help answer this, we further characterized the ΔshbA::pTYM19-rpoBp-hrdB strain, in which an additional copy of hrdB could be transcribed by σHrdB-containing RNAP. Defects found in the ΔshbA::pTYM19-rpoBp-hrdB strain could shed light on the importance of the original system, whereby the principal sigma factor σHrdB is controlled by the alternative sigma factor σShbA throughout growth.
First, we compared the growth of the wt and the ΔshbA::pTYM19-rpoBp-hrdB strains in liquid culture (Fig. 1F). Growth kinetics for the two strains were very similar up to 72 h post inoculation; both strains had a similar lag phase, showed a similar growth rate, and entered into stationary phase with a similar timing. However, the wet weight of cells rapidly decreased in the ΔshbA::pTYM19-rpoBp-hrdB strain after 72 h, suggesting more pronounced cell lysis compared with the wt strain. To examine the cell lysis, we observed mycelial pellets following SYTO 9 and propidium iodide staining (viability staining) by confocal laser scanning fluorescence microscopy. In the presence of both dyes, cells with intact membranes appear fluorescent green, whereas cells with damaged membranes appear red. Both strains formed mycelial pellets in YMPD liquid from the early exponential growth phase, although the size of pellets of the ΔshbA::pTYM19-rpoBp-hrdB strain was generally smaller than that of the wt strain. Representative pellets are shown in Fig. 4. Many round pellets composed of dead hyphae surrounded by a thick surface layer of living hyphae were observed in the 48 and 72 h culture samples of the wt strain. In the 96 h culture wt sample, relatively small pellets composed of dead hyphae surrounded by a thick surface layer of a mixture of non-viable and viable hyphae, which appeared yellow, were observed most frequently, with a few larger pellets. Compared with the wt strain, hyphae in the pellets of the ΔshbA::pTYM19-rpoBp-hrdB strain had more damaged membranes, especially in the later growth stages (after 48 h); many hyphae in the surface layer of the pellets appeared non-viable at both 72 and 96 h. Furthermore, in the 72 and 96 h samples of the ΔshbA::pTYM19-rpoBp-hrdB strain, the numbers of pellets were greatly decreased. These observations confirmed that cell death was accelerated in the ΔshbA::pTYM19-rpoBp-hrdB strain. We noted that the ΔshbA::pTYM19-shbA strain, which possessed an extra copy of shbA in the ΔshbA strain, showed a growth curve very similar to that of the wt strain; rapid decrease of the wet weight of cells in the stationary phase was not observed in this complemented strain (Fig. S6A). This result eliminates the possibility that integration of the pTYM19-derived plasmid into chromosome should cause a negative effect on the viability of the ΔshbA::pTYM19-rpoBp-hrdB strain in stationary phase. We also noted that the wet weight of cells rapidly decreased after approximately 70 h in the ΔshbA [pIJ486-shbA] strain (Fig. 1F), in which transcription of shbA was downregulated throughout growth.
Then we performed qRT-PCR and Western blot analyses to examine the amounts of hrdB transcript and σHrdB protein, respectively, in the ΔshbA::pTYM19-rpoBp-hrdB strain in liquid culture. The amounts of hrdB transcript and σHrdB protein were similar to those in the wt strain (Fig. 3A and B) in the 24 h culture. However, enhancement of hrdB transcription in the later growth stages (48 and 72 h) was not observed in the ΔshbA::pTYM19-rpoBp-hrdB strain (Fig. 3A). In accordance with this, the amount of σHrdB in the ΔshbA::pTYM19-rpoBp-hrdB strain was considerably reduced in the later growth stages (Fig. 3B). We noted that the amount of σHrdB in the complement strain (ΔshbA::pTYM19-shbA) was almost constant throughout growth as in the wt strain (Fig. S6B). These results strongly supported the hypothesis that the enhanced transcription of hrdB is required for maintaining the level of σHrdB in the late growth stage (i.e. stationary phase in liquid culture). Therefore, we postulate that the shortage of σHrdB results in the accelerated cell lysis in stationary phase.
In addition, we compared the morphological development and streptomycin production of the wt and ΔshbA::pTYM19-rpoBp-hrdB strains in solid culture. As described above, the ΔshbA::pTYM19-rpoBp-hrdB strain formed fully developed colonies on YMPD agar. However, morphological development was delayed for approximately one day in the ΔshbA::pTYM19-rpoBp-hrdB strain compared with the wt strain when both strains were cultivated from a single spore, although the growth rate of vegetative mycelium appeared to be almost identical (Fig. 5A). Furthermore, the ΔshbA::pTYM19-rpoBp-hrdB strain produced only small amounts of streptomycin, as determined by bioassay using B. subtilis as an indicator (Fig. 5B). Streptomycin production by the ΔshbA::pTYM19-rpoBp-hrdB strain was not increased even when cultivation was prolonged to 7 days. These results indicate that the lack of σShbA negatively affects morphological development and streptomycin production in solid culture.
The alternative sigma factor σShbA governs the principal sigma factor σHrdB
It has long been believed that principal sigma factors are constitutively produced and are responsible for the transcription of housekeeping genes, including the principal sigma factor-encoding gene itself. However, here we have shown that this is not the case for S. griseus. In this species, the alternative sigma factor σShbA governs the principal sigma factor σHrdB throughout growth. Although some alternative sigma factors in B. subtilis (Carter et al., 1988; Qi and Doi, 1990), E. coli (Dartigalongue et al., 2001; Wade et al., 2006) and S. coelicolor A3(2) (Kim et al., 2012) have been reported to regulate the principal sigma factors under specific conditions, this is the first report of an alternative sigma factor controlling the principal sigma factor throughout growth under normal conditions.
Shinkawa et al. (1995a) reported that RNAP holoenzyme containing σHrdB purified from the vegetative-phase mycelia of S. griseus could initiate in vitro transcription at the hrdB promoter. Contrary to this, RNAP holoenzyme containing σHrdB could not initiate in vitro transcription at the hrdB promoter in our study. We used E. coli core RNAP in our assays, but the RNAP holoenzyme containing σHrdB was active enough to initiate in vitro transcription at the rpoB promoter. Thus, it appears that the σHrdB-containing RNAP holoenzyme has much greater affinity for the rpoB promoter than for the hrdB promoter. hrdB promoter sequences are highly conserved among Streptomyces species, and have little similarity to the consensus sequence (TTGACR-N16–18-TAGRRT, where R is A or G) of probable σHrdB-dependent promoters (Fig. 3F) (Buttner et al., 1990; Strohl, 1992; Shinkawa et al., 1995b; Shinkawa, 1996). This also indicates that σHrdB-containing RNAP holoenzyme fails to initiate efficient transcription at the hrdB promoter. Therefore, we speculate that RNAP holoenzyme purified from the vegetative-phase mycelia of S. griseus, reported by Shinkawa et al. (1995a), might contain not only σHrdB but also small but significant amounts of σShbA. Low-level transcription from the hrdB promoter was detected in the ΔshbA mutant, possibly initiated by another sigma factor. However, the efficiency (or amount) of the substitute sigma factor seemed insufficient for high-level transcription from the hrdB promoter. We assume that the ΔshbA strain is not lethal (but severely deficient in growth) because of this leaky expression of hrdB, although we have not directly proven that hrdB is an essential gene in S. griseus.
σShbA proteins and hrdB promoter sequences are highly conserved among Streptomyces species. Therefore, control of σHrdB by σShbA is likely to be a common feature in Streptomyces. M. tuberculosis has a close homologue of σShbA, which has been named σD (Raman et al., 2004; Calamita et al., 2005). However, M. tuberculosis σD, which shows 47% amino acid identity to σShbA, appears to have alternative functions; σD is required for full virulence of this bacterium, and disruption of the σD gene does not affect the in vitro growth rate (Raman et al., 2004; Calamita et al., 2005). Thus, σD seems not to regulate the principal sigma factor in M. tuberculosis. Furthermore, putative σD-recognition consensus sequences (Raman et al., 2004; Calamita et al., 2005) are not similar to the hrdB promoter sequences in Streptomyces. Therefore, we assume that Streptomyces has evolved σShbA along with the hrdB promoter sequence to generate a unique regulatory system, in which the principal sigma factor σHrdB is governed by the alternative sigma factor σShbA.
Biological significance of control of σHrdB by σShbA
Streptomyces griseus uses the alternative sigma factor σShbA for the transcription of the principal sigma factor gene hrdB, while a principal sigma factor-encoding gene is usually transcribed by RNAP containing the principal sigma factor itself in many other bacteria. The biological significance of such a unique system was examined by the characterization of the ΔshbA::pTYM19-rpoBp-hrdB strain, in which the principal sigma factor gene was transcribed by RNAP containing σHrdB itself, as in many other bacteria. In the wt strain, the amount of hrdB transcript was increased late in growth and accordingly, the level of σHrdB was maintained throughout growth. In contrast, in the ΔshbA::pTYM19-rpoBp-hrdB strain, hrdB transcription was almost constant and σHrdB levels were reduced late in growth. From these results, we assume that the enhanced transcription of hrdB by σShbA in the late growth stage is significant as a strategy to maintain a high σHrdB level throughout growth. In this context, elucidation of the molecular mechanism for the enhancement of shbA transcription in the late growth stage is an important challenge for our future research.
The ΔshbA::pTYM19-rpoBp-hrdB strain showed rapid mycelial lysis in stationary phase in liquid culture. Although we cannot completely eliminate the possibility that some genes required for mycelium survival in stationary phase are directly transcribed by σShbA-containing RNAP, we assume that a shortage of σHrdB is the main cause of the accelerated cell lysis in stationary phase. Very recently, Kim et al. (2012) found the σR-dependent p2 promoter of hrdB in S. coelicolor A3(2). Upon exposure to oxidative stress conditions, transcription from the known σR-independent promoter (p1) rapidly decreases, and transcription from the p2 promoter is induced to compensate for loss of transcription from the primary promoter p1 and to maintain σHrdB levels. This result suggests that maintaining a σHrdB level is important for adaptation to oxidative stress. By analogy with this, we think that maintaining a σHrdB level in stationary phase is important for mycelial survival.
The ΔshbA::pTYM19-rpoBp-hrdB strain also showed delayed morphological development and impaired streptomycin production in solid culture. Both morphological development and streptomycin production are induced by the microbial hormone A-factor (2-isocapryloyl-3-R-hydroxymethyl-γ-butyrolactone) in S. griseus (Horinouchi and Beppu, 1994; Horinouchi, 2007). The ΔshbA::pTYM19-rpoBp-hrdB strain formed aerial mycelium (though somewhat delayed), which indicated that A-factor could be produced sufficiently to trigger morphological development and streptomycin production. Because all of the seven promoters in the streptomycin biosynthetic gene cluster, which were determined in our recent mRNA-seq analysis (T. Tezuka and Y. Ohnishi, unpubl. data), are very different from the hrdB promoter, σShbA should not be directly required for the transcription of the streptomycin biosynthesis genes. Nevertheless, we cannot eliminate the possibility that some genes required for streptomycin production and normal morphological differentiation are directly transcribed by σShbA-containing RNAP. It is an important challenge for our future research to identify other σShbA target genes than hrdB, if they exist. Indeed, it is likely that disturbance of the physiology caused by the shortage of σHrdB in the late growth stage affects morphological development and streptomycin production in the ΔshbA::pTYM19-rpoBp-hrdB strain.
We also speculate that the unique system described here has further benefits for S. griseus during the exponential growth phase. Because σShbA functions as the primary sigma factor of hrdB transcription throughout growth, S. griseus can modulate the amount of σHrdB in mycelium by altering the amount (or activity) of σShbA. For example, if transcription of shbA was repressed in response to some environmental condition, the amount of σHrdB would soon decrease and growth would slow because of poor transcription of many housekeeping genes. Thus, control of the principal sigma factor σHrdB by the alternative sigma factor σShbA may enable S. griseus to control mycelial growth more sensitively in response to environmental changes. Further studies of the involvement of σShbA in the global gene expression of S. griseus will shed light on the novel aspects of the biology of Streptomyces, a model prokaryote for the study of morphogenesis and an important industrial microorganism able to produce a wide variety of biologically active compounds.
Bacterial strains, plasmids and growth conditions
S. griseusIFO13350 was obtained from the Institute of Fermentation, Osaka (IFO), Japan. YMPD is a nutrient-rich medium (pH 7.0) and SMM is a minimal medium (pH 7.2) (Akanuma et al., 2009; Hara et al., 2009). Thiostrepton (5 mg l–1) and apramycin (50 mg l–1) were added when necessary. Spores of S. griseus strains were prepared from cells grown on SMM agar. For submerged culturing, approximately 106 spores were inoculated into YMPD (100 ml) and cultivated in a shaking flask at 30°C. Media and growth conditions used for E. coli were described by Maniatis et al. (1982). Ampicillin (50 mg l–1) and apramycin (50 mg l–1) were used when necessary. For DNA fragments amplified by PCR, the absence of PCR errors was confirmed by nucleotide sequencing on a CEQ 8000 DNA analysis system (Beckman, Brea, CA, USA). All primers used in this study are shown in Supplemental Table S1.
Gene disruption and complementation
pIJ486 is a high-copy-number vector (Ward et al., 1986) and pTYM19 is an integration vector (Onaka et al., 2003). E. coli JM109 and JM110, and pUC18, were purchased from Takara Biochemicals. E. coli IR539 was obtained from RIKEN (Suzuki et al., 2011). This was used to prepare plasmids for gene disruption and transformation in S. griseus. DNA was manipulated in Streptomyces (Hopwood et al., 1985; Kieser et al., 2000) and in E. coli (Maniatis et al., 1982; Ausubel et al., 1987) as described previously.
For construction of the ΔshbA mutant, upstream and downstream regions of shbA were amplified by PCR using primers shbA-FF plus shbA-FR, and shbA-RF plus shbA-RR, and digested by EcoRI plus XhoI, and XhoI plus HindIII. These fragments were cloned together between the EcoRI and HindIII sites of pUC18. A 0.9 kb fragment harbouring aac(3)IV [PCR-amplified from pIJ773 (Gust et al., 2003) using primers apra-F plus apra-R and cloned into pGEM-t easy (Promega) through TA cloning, generating pGEM-aac(3)IV] was digested with EcoRI and cloned into the resulting plasmid at the EcoRI site, generating pUC18-ΔshbA. Plasmid pUC18-ΔshbA, which had been isolated from E. coli IR539, was introduced into S. griseus by protoplast transformation. Apramycin-resistance colonies resulting from a single cross-over were isolated. One of the apramycin-resistant transformants was cultured several times in the absence of apramycin. Apramycin-sensitive colonies (due to a second cross-over event) were isolated as candidates for shbA disruptants. Disruption of shbA was confirmed by Southern hybridization with probe 1 (PCR-amplified using primers Southern-F-1 plus Southern-R-1) and probe 2 (Southern-F-2 plus Southern-R-2) against SalI-digested chromosome DNA (Supplementary Fig. S1E).
For construction of the ΔSGR3370 mutant, upstream and downstream regions of SGR3370 were amplified by PCR using primers SGR3370-FF plus SGR3370-FR, and SGR3370-RF plus SGR3370-RR, and digested by XbaI plus XhoI, and XhoI plus HindIII. These fragments were cloned together between the XbaI and HindIII sites of pUC18. A 0.9 kb fragment harbouring aac(3)IV was digested with EcoRI and cloned into the resulting plasmid at the EcoRI site, generating pUC18-ΔSGR3370. Plasmid ΔSGR3370 was introduced into S. griseus and the ΔSGR3370 mutant was selected as described above. Disruption of SGR3370 was confirmed by Southern hybridization with probe 3 (PCR-amplified using primers Southern-F-3 plus Southern-R-3) and probe 4 (Southern-F-4 plus Southern-R-4) against SphI-digested chromosomal DNA (Supplementary Fig. S2E).
For construction of the ΔshbA::pTYM19-shbA strain, a 1172 bp fragment containing shbA with its upstream and downstream regions, was PCR-amplified using primers shbA-F plus shbA-R. The amplified fragment was digested with EcoRI plus XbaI and cloned between the EcoRI and XbaI sites of pTYM19, resulting in pTYM19-shbA. Non-methylated pTYM19-shbA was prepared from E. coli JM110, and introduced into the ΔshbA mutant by protoplast transformation.
For construction of the ΔshbA::pTYM19-shbA-12his strain, the shbA sequence (with its promoter region) was PCR-amplified using primers shbA-F plus shbA-12his, containing an His12-coding sequence. The amplified fragment was digested with EcoRI and HindIII. In addition, the putative terminator region of shbA was PCR-amplified using primers shbA-F-term plus shbA-R-term and digested with HindIII plus XbaI. These two fragments were cloned together between the EcoRI and XbaI sites of pTYM19, resulting in pTYM19-shbA-12his. Plasmid pTYM19-shbA-12his was introduced into the ΔshbA mutant as described above.
For construction of the ΔshbA [pIJ486-shbA] strain, the 1172 bp shbA fragment containing shbA with its upstream and downstream regions was digested with EcoRI plus XbaI and cloned into pIJ486, resulting in pIJ486-shbA. Plasmid pIJ486-shbA was isolated from Streptomyces lividans TK21 and introduced into the ΔshbA mutant by protoplast transformation.
For construction of the ΔshbA::pTYM19-rpoBp-hrdB strain, the hrdB-coding sequence with its 5′ UTR and terminator sequences (from +1 to +1800, taking the transcriptional start site of hrdB as +1) was placed under control of the rpoB promoter sequence (from −199 to −1, taking the transcriptional start site of rpoB as +1). The former hrdB DNA was PCR-amplified using primers hrdB-F/hrdB-R. The amplified fragment was digested with HindIII and EcoRI and cloned between the HindIII and EcoRI sites of pTYM19, resulting in pTYM19-hrdB. The latter rpoB promoter DNA was PCR-amplified using primers rpoBp-F/rpoBp-R. The amplified fragment was fused with pTYM19-hrdB using an In-Fusion® PCR Cloning Kit w/Cloning Enhancer (Clontech). The resulting plasmid, pTYM19-rpoBp-hrdB, was introduced into the ΔshbA mutant as described above.
Scanning electron microscopy
S. griseus wt and ΔshbA strains were grown on YMPD agar or SMM agar. Scanning electron microscopy was performed as described previously (Yamazaki et al., 2000).
For liquid culture, 106 spores were inoculated into YMPD (100 ml) and cultivated at 30°C. For solid culture, 2.5 × 105 spores were inoculated on cellophane on the surface of 25 ml of solid YMPD medium and grown at 28°C. Total RNA was extracted from collected cells using the RNAqueous kit (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions.
S1 nuclease mapping
S1 nuclease mapping was performed using the method described by Bibb et al. (1986) and Kelemen et al. (1998). Hybridization probes were amplified using primer pairs shbAp-F/shbAp-R* (for low resolution), P1-F/P1-R* and P2-F/P2-R* (for high resolution). Primers indicated by an asterisk were labelled at the 5′ end with [γ-32P]-ATP (220 TBq mmol–1) by T4 polynucleotide kinase, prior to PCR. For hybridization, 500 μg of each total RNA sample was used.
Quantitative RT-PCR was performed as described previously (Ohnishi et al., 2008). We used glkA as an internal standard for confirmation of accuracy of DNA microarray analysis, because glkA was similarly expressed between the wt and ΔshbA [pIJ486-shbA] strains (fold change; 1.3). For other experiments, 16S rRNA was used as an internal standard. All reactions were performed in triplicate, and the data were normalized using the average for the internal standard.
DNA microarray analysis
S. griseus wt and ΔshbA [pIJ486-shbA] strains were cultivated in YMPD liquid medium for 24 h and 30 h, respectively, and harvested by centrifugation. Total RNA was extracted as described above and cDNA was synthesized and labelled with Cy3 and Cy5, as described previously (Ohnishi et al., 2008). Microarray preparation and the method for microarray hybridization have also been previously described (Ohnishi et al., 2008). Among the triplicate cDNA samples of the wt strain, two were labelled with Cy5 and the other was labelled with Cy3. Among the triplicate cDNA samples of the ΔshbA [pIJ486-shbA] strain, two were labelled with Cy3 and the other was labelled with Cy5. We performed duplicate competitive hybridization experiments using the Cy5-labelled cDNA of the wt strain plus the Cy3-labelled cDNA of the ΔshbA [pIJ486-shbA] strain, and a single competitive hybridization experiment using the Cy3-labelled cDNA of the wt strain plus the Cy5-labelled cDNA of the ΔshbA [pIJ486-shbA] strain. Equal amounts (50 pmol) of Cy3- and Cy5-labelled cDNA were used in all hybridizations. Expression ratios (wt/ΔshbA [pIJ486-shbA]) were calculated for each gene as described previously (Ohnishi et al., 2008). Details of the microarray design, transcriptome experimental design and transcriptome data have been deposited in the NCBI Gene Expression Omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/) under Accession No. GSE27426.
Cell extracts for Western blotting were prepared as described by von der Haar (2007) with some modifications. After boiling, cells were disrupted by sonication and subjected to TCA precipitation. Pellets were resolved in a buffer containing 8 M urea, 4% CHAPS and 50 mM DTT. Concentrations of proteins in cell extracts were measured using Bradford reagent (Bio-Rad, Hercules, CA, USA). Cell extracts were separated by SDS-PAGE and then transferred to PVDF membrane (Millipore, Billerica, MA, USA). Anti-σHrdB polyclonal antibody was used to probe σHrdB. Horseradish peroxidase-coupled anti-rabbit IgG (KPL) was used as the secondary antibody and detected by chemiluminescence using SuperSignal West Dura Chemiluminescent Substrate (Pierce).
Production and purification of σShbA and σHrdB
The shbA-coding sequence was amplified using primers shbA-N and shbA-6his, and digested by NdeI and XbaI. The hrdB-coding sequence was amplified using primers hrdB-N and hrdB-6his, and digested by NdeI and EcoRI. Each amplified fragment was cloned into pET26b(+), which had been digested by a pair of appropriate restriction enzymes. Resulting plasmids were introduced into E. coli BL21(DE3). Transformants were cultivated in LB (2 ml) at 37°C overnight. A portion (166 μl) of the culture broth was inoculated into fresh LB (10 ml) and cultivated at 16°C. Four hours later, IPTG was added to the culture with a final concentration of 100 μM and cultivated for another 24 h. Cells were then collected by centrifugation and His-tagged proteins were purified using a Ni-NTA Spin Kit (Qiagen) according to the manufacturer's instructions. Imidazole was removed by dialysis and proteins were concentrated through a Microcon Centrifugal Filter Device (Millipore). Concentrations of purified proteins were measured using Bradford reagent (Bio-Rad).
In vitro transcription run-off assays
In vitro transcription assays were performed using [α-32P]-CTP (30 TBq mmol–1), purchased from Amersham Pharmacia Biotech., as described previously (Kato et al., 2005). Templates for in vitro transcription assay were amplified using primers hrdBp-F and hrdBp-R for the hrdB promoter, and rpoBp-F and rpoBp-R for the rpoB promoter. Each sigma factor (30 pmol) and 1 unit of E. coli RNA polymerase (Epicentre Technologies) were used for the reaction.
Chromatin affinity precipitation (ChAP)
In 100 ml of YMPD, 106 spores of ΔshbA::pTYM19-shbA-12his were cultivated for 27 h. Cells were cross-linked using formaldehyde (1%, final concentration) and cross-linked cells were sonicated to shear the DNA and obtain a size range of approximately 300–1000 bp, with an average fragment size of ≤ 500 bp (Higo et al., 2012). A portion of this extract was used as the input material. This lysate was subjected to affinity precipitation with Dynabeads Talon (Veritas), and DNA was then purified using a Qiaquick PCR purification kit (Qiagen) as described previously (Higo et al., 2012). For each PCR reaction, titrations of a standard DNA solution were used for calculation of the ratio of affinity-precipitated DNA to input DNA. PCR primer pairs used to amplify the respective regions were: hrdBp-upstream-F and hrdBp-upstream-R for a region upstream from the hrdB promoter; hrdBp-q-F and hrdBp-q-R for a region containing the hrdB promoter; hrdB-q-F and hrdB-q-R for a region downstream from the hrdB promoter; rpoBp-q-F and rpoBp-q-R for the rpoB promoter region; rpsJp-q-F and rpsJp-q-R for the rpsJ promoter region; rpmHp-dnaAp-q-F and rpmHp-dnaAp-q-R for the rpmH and dnaA promoter region; and rpsFp-q-F and rpsFp-q-R for the rpsF promoter region. A putative terminator region between SGR3357 and SGR3358 was used as an invariant control using SacI-term-F and SacI-term-R. All reactions were performed in triplicate and the recovery ratio of the hrdB promoter region was compared with that of the other regions through one-tailed one sample t-test.
Cells of the wt and ΔshbA::pTYM19-rpoBp-hrdB strains cultivated in 100 ml of YMPD liquid medium were harvested by centrifugation and washed twice with phosphate-buffered saline (PBS), then resuspended in PBS. Equal volumes of 2× stock solution of the LIVE/DEAD®BacLightTM staining reagent (mixture of 12 μM SYTO 9 and 60 μM propidium iodide, PI; Molecular Probes) were added and incubated at room temperature in the dark for 15 min. Samples were observed under an FV300 laser confocal microscope (Olympus) at excitation wavelengths of 488 and 543 nm and emission of 530 nm (green) or 610 nm (red) (optical sections of approximately 6.8 μm).
The amount of streptomycin produced was measured by a bioassay using B. subtilis as an indicator (Horinouchi et al., 1984).
We thank Dr Hirofumi Hara (Okayama University of Science, Okayama, Japan) for assistance with COG analysis and construction of pGEM-aac(3)IV. This work was supported by the Targeted Proteins Research Program (TPRP) of the Ministry of Education, Culture, Sports, Science, and Technology of Japan (MEXT) and a Funding Program for Next Generation World-Leading Researchers from the Bureau of Science, Technology, and Innovation Policy, Cabinet Office, Government of Japan. H.O. was supported by the Global COE programme ‘Nuclear Education and Research Initiative’ of MEXT.