BldC Delays Entry into Development To Produce a Sustained Period of Vegetative Growth in Streptomyces venezuelae

Understanding the mechanisms that drive bacterial morphogenesis depends on the dissection of the regulatory networks that underpin the cell biological processes involved. Recently, Streptomyces venezuelae has emerged as an attractive model system for the study of morphological differentiation in Streptomyces. This has led to significant progress in identifying the genes controlled by the transcription factors that regulate aerial mycelium formation (Bld regulators) and sporulation (Whi regulators). Taking advantage of S. venezuelae, we used ChIP-seq coupled with RNA-seq to identify the genes directly under the control of BldC. Because S. venezuelae sporulates in liquid culture, the complete spore-to-spore life cycle can be examined using time-lapse microscopy, and we applied this technique to the bldC mutant. These combined approaches reveal BldC to be a member of an emerging class of Bld regulators that function principally to repress key sporulation genes, thereby extending vegetative growth and blocking the onset of morphological differentiation.

ABSTRACT

Streptomycetes are filamentous bacteria that differentiate by producing spore-bearing reproductive structures called aerial hyphae. The transition from vegetative to reproductive growth is controlled by the bld (bald) loci, and mutations in bld genes prevent the formation of aerial hyphae, either by blocking entry into development (typically mutations in activators) or by inducing precocious sporulation in the vegetative mycelium (typically mutations in repressors). One of the bld genes, bldC, encodes a 68-residue DNA-binding protein related to the DNA-binding domain of MerR-family transcription factors. Recent work has shown that BldC binds DNA by a novel mechanism, but there is less insight into its impact on Streptomyces development. Here we used ChIP-seq coupled with RNA-seq to define the BldC regulon in the model species Streptomyces venezuelae, showing that BldC can function both as a repressor and as an activator of transcription. Using electron microscopy and time-lapse imaging, we show that bldC mutants are bald because they initiate development prematurely, bypassing the formation of aerial hyphae. This is consistent with the premature expression of BldC target genes encoding proteins with key roles in development (e.g., whiD, whiI, sigF), chromosome condensation and segregation (e.g., smeA-sffA, hupS), and sporulation-specific cell division (e.g., dynAB), suggesting that BldC-mediated repression is critical to maintain a sustained period of vegetative growth prior to sporulation. We discuss the possible significance of BldC as an evolutionary link between MerR family transcription factors and DNA architectural proteins.

INTRODUCTION

The complex Streptomyces life cycle involves two distinct filamentous cell forms: the growing or vegetative hyphae and the reproductive or aerial hyphae, which differentiate into long chains of spores (1–6). Genetic studies identified the regulatory loci that control entry into development, which are called bld (bald) genes because null mutations in these loci prevent the formation of aerial hyphae. However, colony baldness can arise for two different reasons. The larger class of bld mutants, which define positive regulators, fail to initiate development, forming colonies of undifferentiated vegetative mycelium. In contrast, a smaller but growing class of bld mutants, which define negative regulators, enter development prematurely, inducing sporulation in the vegetative mycelium and bypassing the formation of aerial hyphae. Thus, macroscopically these two classes of mutants look similar, forming smooth colonies that lack the “hairy” appearance of the wild type, but microscopically it is apparent that they arise for diametrically opposed reasons (5, 7–9).

BldC is a small, 68-residue protein with a winged Helix-Turn-Helix (wHTH) motif, related to those found in MerR-family proteins (10). The basic structure of classical MerR proteins is a dimer consisting of two identical subunits, each composed of an N-terminal wHTH DNA-binding domain, a C-terminal effector-recognition domain, and an interconnecting linker region that consists of a long α-helix that interacts with the same helix in the other subunit, forming an antiparallel coiled-coil responsible for homodimerization. MerR-like proteins share significant sequence similarity only within their DNA-binding domains; as different family members bind different effectors, their C-terminal domains are variable and show little, if any, similarity to one another.

MerR transcription factors bind to palindromic DNA sequences as homodimers. However, unlike classical members of the MerR family, BldC has neither an effector domain nor the dimerization helix, and BldC behaves as a monomer in free solution (11). As a consequence, how BldC might bind DNA remained unclear. To address this question, Schumacher et al. (11) carried out biochemical and structural studies to characterize the binding of Streptomyces coelicolor BldC to the promoters of two known target genes, whiI and smeA. These studies showed that BldC binds DNA in a completely different way from classical MerR regulators, instead involving asymmetric, cooperative, head-to-tail oligomerization on DNA direct repeats with concomitant pronounced DNA distortion (11). The number of direct repeats present in BldC-binding sites is variable, thus allowing cooperative, head-to-tail binding of additional BldC monomers. Since BldC-like proteins radiate throughout the bacteria, this study identified BldC as the founding member of a new structural family of transcription factors.

Although the work by Schumacher et al. (11) has provided a clear mechanistic understanding of how BldC binds DNA, there has been less insight into its biological role and impact on Streptomyces development. In part, this is because previous studies have focused on the classical model species S. coelicolor, which sporulates only on solid medium. Here we exploit the benefits of the model species Streptomyces venezuelae, which sporulates in liquid culture (12), to study the biological role of BldC. Using ChIP-seq coupled with RNA-seq, we identify the genes under BldC control and show that BldC can function both as a repressor and as an activator of transcription. We show that bldC mutants are bald because they enter development prematurely, bypassing the formation of aerial hyphae. This correlates with the premature expression of BldC target genes with key roles in development, chromosome condensation and segregation, and sporulation-specific cell division, suggesting that BldC-mediated repression is critical to maintain a sustained period of vegetative growth prior to sporulation.

RESULTS

Deletion of bldC causes premature initiation of development.

We constructed an S. venezuelae bldC mutant by replacing the monocistronic bldC coding region with an apramycin resistance (apr) cassette. The resulting mutant was bald, unable to produce the reproductive aerial hyphae that give mature wild-type Streptomyces colonies their characteristic fuzzy appearance (Fig. 1). However, scanning electron microscopy (SEM) of mature colonies of the bldC mutant showed that most of the biomass consisted of spores, rather than undifferentiated vegetative hyphae (Fig. 2). Comparison of the growth of the wild type and the bldC mutant on plates over time showed that after 1 day they looked similar (vegetative growth only) but after 2 days the wild type had produced aerial hyphae while the bldC mutant was still restricted to vegetative growth. After 3 days, the aerial hyphae of the wild type had differentiated into spores, and essentially the entire biomass of the bldC mutant had also differentiated into spores, bypassing aerial mycelium formation. The bldC mutant also produced higher levels of extracellular matrix than the wild type (see the images from the 1-day time point in Fig. 2), a property that might contribute to the failure of bldC mutants to erect aerial hyphae. The bldC mutant phenotype was fully complemented by introducing a single copy of the bldC gene under the control of its native promoter, expressed in trans from the ΦBT1 integration site (Fig. 1 and 2).

FIG 1.

BldC is required for the formation of aerial mycelium. Wild-type S. venezuelae, the bldC mutant, the bldC mutant carrying the empty vector, and the complemented bldC mutant, photographed after 4 days of growth on MYM solid medium.

FIG 2.

Deletion of bldC causes premature initiation of development on solid medium. Scanning electron micrographs showing the phenotypes of the bldC mutant and the wild type after 1, 2, and 3 days of growth on MYM solid medium. The phenotype of the complemented bldC mutant is also shown after 3 days of growth on MYM solid medium. Note the overproduction of extracellular matrix in the bldC mutant relative to the wild type at 1 day and the formation of spores in the bldC mutant at 3 days.

Using an established microfluidic system and methodology (12), we conducted time-lapse fluorescence microscopy to further study the developmental defects associated with deletion of bldC. As in previous studies (7, 12, 13), we introduced an FtsZ-YPet translational fusion into the wild-type, mutant, and complemented mutant strains, allowing us to monitor each of the two distinct modes of cell division that occur in Streptomyces. In Fig. 3, the scattered single Z-rings mark the position of vegetative cross-walls, which do not constrict or give rise to cell-cell separation, but simply divide the vegetative hyphae into long, cylindrical compartments (e.g., Fig. 3A and C, panel 7 h). In contrast, during reproductive growth, long ladders of regularly spaced Z-rings are synchronously deposited along sporogenic hyphae. These Z-rings mark the sites of sporulation septa, which do constrict, ultimately leading to the formation of chains of spores (e.g., Fig. 3A and C, panels 12 h and 21 h). Time-lapse imaging of strains harboring the FtsZ-YPet fusion showed that the duration of vegetative growth was shorter in the bldC mutant than in the wild type (Fig. 3 and Movies S1A and B and S2A and B in the supplemental material) and the complemented mutant (Fig. 3 and Movies S2A and B and S3A and B). Noticeably, following germination, hyphal outgrowth in the bldC mutant was associated with an immediate increase in FtsZ-YPet signal, leading to the precocious formation of ladders of Z-rings (Fig. 3B, 4 h, and Movie S2A and B). However, although ladders of Z-rings were observed as early as 4 h in the bldC mutant, mature spores were not observed in the corresponding DIC images until 21 h, the same time that mature spores were seen in the wild type (Fig. 3A and B). As on plates, the bldC mutant hypersporulates in liquid culture, such that the entire biomass differentiates into spores. Wild-type patterns of FtsZ expression and sporulation were restored in the complemented mutant (Fig. 3C and Movie S3A and B). From these data, we concluded that the overall role of BldC is to sustain filamentous growth and delay entry into development.

FIG 3.

Deletion of bldC causes premature initiation of development in liquid medium. Time-lapse images (4, 7, 12, and 21 h postinoculation) of (A) wild-type S. venezuelae, (B) the bldC mutant, and (C) the complemented bldC mutant, grown in liquid MYM medium in the microfluidic system. All three strains carry the same FtsZ-YPet translational fusion expressed from the native ftsZ promoter, and both the DIC (upper) and fluorescence (lower) images are shown. Evenly spaced FtsZ-rings are a marker for sporulation-specific cell division (see, for example, the 12-h panel for the wild type). Scale bar = 10 µm. For the corresponding movies, please see Movie S1A and B, Movie S2A and B, and Movie S3A and B in the supplemental material.

MOVIE S1A

Time-lapse microscopy of the wild-type strain expressing the FtsZ-YPet fusion. DIC movies are at 5 frames per second. The time following the first image is indicated at the bottom left. Images were taken every 30 minutes (DIC, 150 ms). Movies were assembled in the Fiji software package (http://fiji.sc/Fiji). Download Movie S1a, AVI file, 11.8 MB ^{(12.1MB, avi)} .

Copyright © 2019 Bush et al.

This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.

MOVIE S1B

Time-lapse microscopy of the wild-type strain expressing the FtsZ-YPet fusion. YFP-channel movies are at 5 frames per second. The time following the first image is indicated at the bottom left. Images were taken every 30 minutes (YFP, 100 ms). Movies were assembled in the Fiji software package (http://fiji.sc/Fiji). Download Movie S1b, AVI file, 11.5 MB ^{(11.8MB, avi)} .

Copyright © 2019 Bush et al.

This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.

MOVIE S2A

Time-lapse microscopy of the bldC mutant expressing the FtsZ-YPet fusion. DIC movies are at 5 frames per second. The time following the first image is indicated at the bottom left. Images were taken every 30 minutes (DIC, 150 ms). Movies were assembled in the Fiji software package (http://fiji.sc/Fiji). Download Movie S2a, AVI file, 18.9 MB ^{(19.4MB, avi)} .

Copyright © 2019 Bush et al.

This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.

MOVIE S2B

Time-lapse microscopy of the bldC mutant expressing the FtsZ-YPet fusion. YFP-channel movies are at 5 frames per second. The time following the first image is indicated at the bottom left. Images were taken every 30 minutes (YFP, 100 ms). Movies were assembled in the Fiji software package (http://fiji.sc/Fiji). Download Movie S2b, AVI file, 14.8 MB ^{(15.2MB, avi)} .

Copyright © 2019 Bush et al.

This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.

MOVIE S3A

Time-lapse microscopy of the complemented strain expressing the FtsZ-YPet fusion. DIC movies are at 5 frames per second. The time following the first image is indicated at the bottom left. Images were taken every 30 minutes (DIC, 150 ms). Movies were assembled in the Fiji software package (http://fiji.sc/Fiji). Download Movie S3a, AVI file, 19 MB ^{(19.5MB, avi)} .

Copyright © 2019 Bush et al.

This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.

MOVIE S3B

Time-lapse microscopy of the complemented strain expressing the FtsZ-YPet fusion. YFP-channel movies are at 5 frames per second. The time following the first image is indicated at the bottom left. Images were taken every 30 minutes (YFP, 100 ms). Movies were assembled in the Fiji software package (http://fiji.sc/Fiji). Download Movie S3b, AVI file, 10.9 MB ^{(11.1MB, avi)} .

Copyright © 2019 Bush et al.

This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.

BldC levels are highest early in development.

Using an anti-BldC polyclonal antibody, we monitored BldC levels in S. venezuelae during sporulation in liquid culture. Western blotting showed that BldC is abundant throughout the life cycle, but that BldC levels are highest early on, during vegetative growth (Fig. 4). In contrast, levels of the transcription factor WhiA are maintained throughout development (Fig. 4B), as has been reported previously (14).

FIG 4.

Automated Western blot analysis of BldC levels during submerged sporulation in MYM liquid medium. Equal amounts of total protein were loaded for each sample. BldC and WhiA (internal control) were detected with polyclonal antibodies using the quantitative “Wes” capillary electrophoresis and blotting system (ProteinSimple, San Jose, CA). The S. venezuelae bldC mutant (14 h postinoculation) was used as a negative control. (A) Quantitation of BldC levels (area under each peak; arbitrary units). All experimental samples were analyzed in triplicate, and the mean value and its standard error are shown for each sample. (B) Virtual Western blot for BldC (top) and the internal control, WhiA (bottom). Each time point is indicated in hours, along with its relation to the developmental stage (V, vegetative growth; F, fragmentation; S, sporulation), as determined by microscopy. Cultures used to analyze BldC levels were the same as those used to prepare RNA for qRT-PCR analysis (Fig. 6).

Defining the BldC regulon in S. venezuelae.

Previously, ChIP-seq (or ChIP-chip) coupled with transcriptional profiling has enabled us to define the regulons of several key developmental regulators in S. venezuelae (14–17). Here, we employed the same approach, using an anti-BldC polyclonal antibody to identify the promoters directly bound by BldC. We performed ChIP-seq at two distinct stages of vegetative growth when BldC was abundant: early vegetative growth (10 h) and presporulation (14 h). This work revealed 367 potential gene targets, 90% of which were bound by BldC at both time points (Table S1A). These targets include many genes encoding key transcriptional regulators of the Streptomyces developmental cascade (e.g., bldM, whiB, wblA, whiD, whiH, whiI, sigF, and bldC itself), others encoding proteins involved in chromosome condensation and segregation during sporulation (e.g., hupS, smeA-sffA), and those directly involved in cell division during sporulation (e.g., dynAB, ssgB) (Fig. 5 and Table 1). Schumacher et al. (11) characterized the interaction of S. coelicolor BldC with the promoters of two of its previously known targets, whiI and the smeA-ssfA operon. whiI encodes an orphan response regulator that is essential for the later stages of sporulation, when it forms a functional heterodimer with a second orphan response regulator, BldM, enabling WhiI to bind to DNA and regulate the expression of ∼40 sporulation genes (16). The smeA-sffA operon encodes a small membrane protein (SmeA) that recruits a DNA translocase (SffA) to sporulation septa (18). Deletion of smeA-sffA results in a modest defect in spore chromosome segregation and has pleiotropic effects on spore maturation (18).

FIG 5.

BldC ChIP-seq analysis in S. venezuelae. ChIP was conducted in the wild type and the bldC mutant using an anti-BldC polyclonal antibody. Color-coding of the plots is as follows: S. venezuelae wild type 10 h (WT 10 h), red; S. venezuelae wild type 14 h (WT 14 h), green; and ΔbldC mutant 14 h (ΔbldC), black. Plots span approximately 5 kb of DNA sequence. Genes running left to right are shown in green, and genes running right to left are shown in red. The black arrow indicates the gene subject to BldC regulation, as determined by RNA-seq transcriptional profiling. Note that the scale of the y axis varies between panels.

TABLE 1.

Selected BldC ChIP-seq targets and corresponding RNA-seq data^a

min apv	Gene	Product	LogFC
			10 h	14 h
2.42E−10	vnz_18945	BldC
3.03E−18	vnz_35035	EshA	−8.15	−5.34
	vnz_35040	2-Methylisoborneol synthase	−4.89	−2.80
	vnz_35045	2-Methylgeranyl diphosphate synthase	−6.42	−5.19
3.19E−17	vnz_16495	WblA	−4.36	−5.53
1.40E−06	vnz_27205	WhiH	−4.22	−3.77
1.33E−07	vnz_22005	BldM	−2.67	−3.02
5.39E−22	vnz_18205	SsgA	−1.81	−0.56
	vnz_18200	SsgR	−2.30	−2.32
4.48E−32	vnz_04805	CvnA4	−0.18	−2.14
	vnz_04800	CvnB4	0.00	−2.62
	vnz_04795	CvnC4	0.19	−2.54
	vnz_04790	CvnD4	−0.10	−2.79
3.37E−15	vnz_25880	CvnA1	0.01	−1.15
	vnz_25875	CvnB1	−0.08	−1.52
	vnz_25870	CvnC1	−0.16	−2.26
	vnz_25865	CvnD1	−0.10	−2.27
1.33E−07	vnz_22000	WhiD	6.89	2.44
1.33E−15	vnz_18620	SigF	4.42	3.85
2.52E−07	vnz_12110	DynA	3.10	2.39
	vnz_12105	DynB	3.18	2.10
1.23E−12	vnz_04885	SmeA	2.73	0.14
	vnz_04890	SffA	2.70	0.34
2.33E−12	vnz_05545	SsgB	2.55	0.74
4.06E−09	vnz_13800	SigI	2.54	2.76
1.48E−09	vnz_24690	Cell division protein FtsW-like	2.49	1.86
	vnz_24685	Penicillin-binding protein	2.14	2.04
2.50E−15	vnz_22740	Diguanylate cyclase	2.27	2.01
1.87E−27	vnz_28820	WhiI	1.98	3.03
1.22E−35	vnz_12970	Penicillin-binding protein	1.87	1.66
7.74E−05	vnz_25950	HupS	1.86	1.48
7.17E−25	vnz_23255	Penicillin-binding protein	1.45	1.48
1.04E−17	vnz_15130	d-Alanyl-d-alanine carboxypeptidase	0.82	0.71
4.46E−06	vnz_13645	WhiB	0.62	0.51

^aListed are the minimum average P value (min apv) for the ChIP-seq peak, the gene, and the gene product. For each target, the RNA-seq data, showing the relative expression values (logFC) for the ΔbldC mutant compared to the wild type at the 10-h and 14-h time points, are also listed. Significant decreases in relative expression (<−1) are indicated in bold. Significant increases in relative expression (>1) are indicated in italic. Where BldC binding is likely to exert control over multiple genes in a single operon, the data for those genes are also listed.

TABLE S1

(A) ChIP-seq data set for S. venezuelae BldC. Only those peaks with significance P < E−04 for at least one of the time points are included in the analysis. “Pos” = position in the S. venezuelae genome in bases. “lndiff” = the difference between the local normalized (ln) values of the immunoprecipitated wild-type samples and ΔbldC mutant for each of the time points (lndiff 10hr and lndiff 14hr). “min apv” = minimum adjusted P value for the 10-h and 14-h time points. “Closestgene” = nearest annotated gene relative to the position of significance. “lgene” = the identity of a gene where present on the left of the significant position and in the 5′ to 3′ direction. “lproduct” = the predicted gene product of the lgene. “ldist” = the distance between the significant position and the predicted start codon of the lgene. “igene” = the identity of a gene where the significant position is found within (“in”) a coding region. “iproduct” = the predicted gene product of igene. “idist” = the distance between the significant position and the predicted start codon of the igene. “rgene” = the identity of a gene where present on the right of the significant position and in the 5′ to 3′ direction. “rproduct” = the predicted gene product of the rgene. “rdist” = the distance between the significant position and the predicted start codon of the rgene. Where present, for each of lgene, igene, and rgene, the relative expression values generated by RNA-seq are listed for each time point (lgene/igene/rgene logFC 10hr and lgene/igene/rgene logFC 14hr). Where the logFC > 1, cell values are highlighted in red; where the logFC < −1, cell values are highlighted in yellow. (B) Analysis of BldC ChIP-seq “peaks.” For each maximum significant position (Max Sig Pos) or ChIP-seq “peak” (as listed in Table S1A), the genomic positions of the leftmost (Pos L) and rightmost (Pos R) significant positions are recorded as well as the distance between these positions (Width) at both the 10-h and 14-h time points. “FASTA” = nucleotide sequences between pos L and pos R. “Nearest Gene” = closest gene to the maximum significant position. “Class” = the class of BldC enrichment observed upon manual inspection of the ChIP-seq peak. BldC binding generates either a narrow or broad region of enrichment. BldC targets with broad regions of enrichment generally correlate to “peak” widths > 300 bp, and examination of the nucleotide sequences reveals multiple AT-rich sequences that would support BldC multimerization similar to that observed in the smeA-BldC structure (Schumacher et al., 2018). BldC targets with narrow regions of enrichment generally correlate to “peak” widths < 300 bp, and examination of the nucleotide sequences reveals fewer AT-rich sequences that would support BldC multimeriszation similar to that observed in the whiI-BldC structure (Schumacher et al., 2018). Peaks recorded as “Broad*” are broad and >300 bp upon manual inspection, but their significance based upon the threshold applied in our analysis means that a much narrower region is considered bioinformatically. (C) Complete RNA-seq data for BldC. Shown is the relative gene expression for the ΔbldC mutant compared to the wild type at the 10-h and 14-h time points. For each gene, the log fold change (logFC) and average P value (apv) is listed at each time point. The canonical gene name (if known) and expected gene product are also listed. Where the logFC > 1, cell values are highlighted in red; where the logFC < −1, cell values are highlighted in yellow. (D) BldC represses the transcription of genes during vegetative growth. Listed are BldC ChIP-seq targets that are significantly upregulated (logFC > 1) in the absence of bldC during vegetative growth at either the 10-h or 14-h time points, as determined by RNA-seq, ordered by logFC at 10 h and then by logFC at 14 h. For each ChIP-seq target, the position, minimum adjusted P value (min apv), the gene, expected product, and distance to the predicted start codon (dist) are also listed. Where the logFC > 1, cell values are highlighted in red; where the logFC < −1, cell values are highlighted in yellow. (E) Selected RNA-seq data for the dcw cluster, the conservons, and genes involved in aerial mycelium formation. For each gene, the gene product and log fold change (logFC 10 hr and logFC 14 hr) are listed. Where the logFC > 1, cell values are highlighted in red; where the logFC < −1, cell values are highlighted in yellow. (F) BldC activates the transcription of genes during vegetative growth. Listed are BldC ChIP-seq targets that are significantly downregulated (logFC < −1) in the absence of bldC during vegetative growth, as determined by RNA-seq, ordered by logFC at 10 h and then by logFC at 14 h. For each ChIP-seq target, the position, minimum adjusted P value (min apv), the gene, expected product, and distance to the predicted start codon (dist) are also listed. Download Table S1, XLSX file, 4 MB ^{(3.9MB, xlsx)} .

Copyright © 2019 Bush et al.

This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.

Both BldC and BldD inhibit entry into development, and previously we identified 170 genes directly controlled by BldD (5, 9). Comparison of the BldC and BldD regulons shows only a small overlap of 15 genes/operons bound by both proteins. Strikingly, however, targets regulated by both BldC and BldD include whiB, whiD, bldM, smeA-sffA, bldC, and cvnA1.

Schumacher et al. (11) showed that BldC binds to DNA in a head-to-tail fashion at a variable number of direct repeats. So, for example, in the whiI-BldC structure, there are two direct repeats resulting in the head-to-tail oligomerization of two BldC monomers, whereas in the smeA-BldC structure there are four direct repeats, resulting in the head-to-tail oligomerization of four BldC monomers. In line with this, our data show a broader BldC ChIP-seq peak at the smeA promoter than the whiI promoter (Fig. 5). Indeed, regions of BldC enrichment across the S. venezuelae genome were often noticeably broad. Approximately 60% of BldC targets were defined by narrow, whiI-like ChIP-seq peaks, but the remaining ∼40% showed ChIP-seq peaks at least as broad as the peak observed at the smeA promoter (Fig. S1 and Table S1B). The cooperative binding of BldC to DNA revealed by structural analysis (11) suggested that dimerization on DNA would be the minimum requirement for DNA binding and that extended multimerization would occur at target promoters carrying additional direct repeats. The BldC-DNA structures identified two major elements that define the specificity of BldC binding, a 4-bp AT-rich sequence followed by a C or G four or five nucleotides downstream. The consensus direct repeat is 5′-AATT(N_3-4)(C/G)-3′, but the BldC-smeA structure showed that conservation of even this degenerate consensus is not critical for BldC binding. In particular, for the AT-rich sequence, it is the narrowing of the minor groove caused by the AT-rich nature of that sequence that is important, rather than direct base reading by BldC (11). Because of this plasticity, it is not possible to predict BldC binding sites bioinformatically. Nevertheless, using this loose consensus as a guide, it seems likely that the BldC targets with narrow ChIP-seq peaks have two appropriately spaced direct repeat sequences, whereas BldC targets with broad ChIP-seq peaks, such as smeA-sffA, cdgE, and dynAB, have more (Fig. S1).

FIG S1

BldC ChIP-seq peaks fall into two classes. BldC binding upstream of some targets generates a broad region of enrichment (right column, labeled Broad). For smeA, this likely corresponds with the binding of four direct repeats by BldC, observed in vitro (M. A. Schumacher, C. D. den Hengst, M. J. Bush, T. B. K. Le, N. T. Tran, et al., Nat Commun 9:1139, 2018, https://doi.org/10.1038/s41467-018-03576-3). Other BldC ChIP-seq targets, e.g., cdgE and dynAB, display similarly broad regions of enrichment, and examination of the nucleotide sequence in these regions likewise reveals four similar and appropriately spaced direct repeats. At other BldC ChIP-seq targets, much narrower regions of enrichment are observed (left column, labeled Narrow). For whiI, this likely corresponds with the binding of just two direct repeats by BldC, observed in vitro (Schumacher et al., 2018). Other BldC ChIP-seq targets, e.g., vnz23255 and cvnA4-D4, display similarly narrow regions of enrichment, and examination of the nucleotide sequence in these regions likewise reveals a pair of similar and appropriately spaced direct repeats. The ChIP-seq panels are identical to those shown in Fig. 5. The AT-rich sequences of the verified (in the case of whiI and smeA) and candidate direct repeats to which BldC binds are highlighted below in yellow in the 5′ to 3′ direction. Download FIG S1, DOCX file, 0.2 MB ^{(260.7KB, docx)} .

Copyright © 2019 Bush et al.

This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.

BldC represses the transcription of a subset of target genes.

whiI and the smeA-sffA operon were originally identified as BldC targets in S. coelicolor (11), and the ChIP-seq analysis presented here shows that they are also BldC targets in S. venezuelae (Fig. 5 and Table 1). To assess the regulatory influence of BldC on the whiI and smeA promoters, we performed qRT-PCR using RNA isolated from wild-type S. venezuelae and the bldC mutant during vegetative growth (10 h), when BldC is abundant in wild-type cells (Fig. 6). Under these conditions, expression of both whiI and smeA is significantly higher in the bldC mutant than the wild type (20-fold and 60-fold increases, respectively). This suggests that BldC functions to repress the transcription of these developmental target genes during vegetative growth, consistent with the premature initiation of development seen in a bldC mutant. To gain a global view of the regulatory impact of BldC, we conducted RNA-seq to compare the transcriptomes of wild-type S. venezuelae and the bldC mutant at 10-h and 14-h time points, when both strains were still growing vegetatively (Table S1C). The RNA was prepared from the same cultures used to make protein extracts for the BldC Western blotting shown in Fig. 4. In line with the qRT-PCR data, whiI and smeA showed significant increases in expression in the bldC mutant compared with the wild type. The smeA and sffA genes form an operon, and consistent with this, both genes showed similar upregulation of expression at the 10-h and 14-h time points (Table 1). In total, 156 of the genes we identified as BldC targets in ChIP-seq showed a greater than 2-fold increase in expression in the bldC mutant (Table S1D and Fig. 7). These included the key developmental genes sigF, whiD, and hupS, which showed a greater than 2-fold increase in expression (logFC, >1) in the bldC mutant, compared to the wild type at both the 10-h and 14-h time points (Table 1). qRT-PCR confirmed the upregulation of sigF, whiD, and hupS expression in the bldC mutant relative to the wild type, as was observed for whiI and smeA (Fig. 6).

FIG 6.

qRT-PCR data showing normalized mRNA accumulation for the BldC target genes whiI, smeA, sigF, whiD, hupS, whiH, and bldM in the wild type (white bars) and the bldC mutant (black bars). Strains were grown in MYM liquid medium, and RNA samples were taken at 10 h. Expression values were calculated relative to the accumulation of the constitutively expressed hrdB reference mRNA and normalized to the wild-type value at 10 h. Note that the scale of the y axis varies between panels.

FIG 7.

The bldC mutation is highly pleotropic. Volcano plots of the RNA-seq data at the 10-h (left panel) and 14-h (right panel) time points with significance (−log₁₀ P value) plotted against differential expression (log fold change). The thresholds for significant differential expression (>1 or <−1 log fold change) are indicated via vertical dashed lines. Genes with log fold change >1/<−1 show at least a 2-fold increase/decrease in expression in the ΔbldC mutant relative to the wild type. Genes that are BldC ChIP-seq targets in S. venezuelae are indicated by red dots.

Among the other BldC targets identified by ChIP-seq were a number of genes encoding members of the penicillin-binding protein (PBP) family, required for the synthesis of peptidoglycan (19, 20). Our data indicate that the vnz12970 and vnz23255 genes, encoding class A high-molecular-mass (HMM) PBPs, and the vnz15130 gene, encoding a low-molecular-mass (LMM) PBP, are all targets of BldC (Table 1). Expression of these PBP-encoding genes is significantly upregulated in a bldC mutant during vegetative growth compared to the wild type (1.75-to 3.6-fold at 10 h), showing that BldC functions to repress their transcription (Table 1).

Peptidoglycan synthesis is required to produce new wall material during cell elongation and to produce septa during division (21). Most rod-shaped bacteria possess distinct gene pairs to control these two processes, a protein of the SEDS (shape, elongation, division, and sporulation) family and its cognate class B PBP. In Escherichia coli, the RodA-PBP2 and FtsW-FtsI pairs control elongation and division, respectively (22). In S. venezuelae, there are four equivalent SEDS-PBP pairs and our data indicate that one of these pairs, vnz24690 and vnz24685, is under BldC control (Fig. 5 and Table 1). BldC binds upstream of this gene pair, and both genes show a >4-fold increase (logFC, >2) in expression in the bldC mutant compared to the wild type.

In Streptomyces, the ftsW-ftsI SEDS-PBP gene pair is specifically required for cell division at sporulation septa (23, 24). Both genes are found in the division and cell wall (dcw) gene cluster. This cluster encodes many proteins that play critical roles in hyphal polar growth, peptidoglycan biosynthesis, and cell division, including DivIVA, SepF, SepG, FtsW, and FtsZ. Closer inspection of the ChIP-seq data showed that BldC binds at multiple positions across the dcw cluster (Fig. 8), although these peaks all fall just below the significance threshold we applied (P < E−04) (Table S1A). The majority of genes within this cluster show a modest increase in expression (logFC, 0.5 to 1.5) during vegetative growth in the bldC mutant relative to the wild type (Table S1E), suggesting that BldC functions to repress the transcription of genes within the dcw cluster during vegetative growth, in line with the premature initiation of cell division during sporulation that we observe in a bldC mutant.

FIG 8.

BldC binds at multiple positions across the dcw operon. The genes found in the dcw cluster are annotated. Color-coding of the ChIP samples is as follows: S. venezuelae wild type 10 h (WT 10 h), red; S. venezuelae wild type 14 h (WT 14 h), green; and ΔbldC mutant 14 h (ΔbldC), black. Genes running left to right are shown in green, and genes running right to left are shown in red. The black arrowheads indicate BldC binding sites identified in this analysis.

Our data also indicate BldC-mediated repression of other genes with critical roles in cell division and sporulation, such as the dynAB operon and ssgB (Fig. 5 and Table 1). dynA and dynB encode two dynamin-like membrane-remodeling proteins that stabilize FtsZ rings during sporulation septation via protein-protein interactions with other divisome components, including FtsZ, SepF, SepF2, and SsgB (13). In S. coelicolor, the actinomycete-specific proteins SsgA and SsgB positively control the spatial distribution of FtsZ rings during sporulation-specific cell division. SsgA binds and recruits SsgB, which in turn recruits FtsZ, determining the future sites of sporulation septation (25). Our data indicate that ssgB is a target of BldC-mediated repression (Table 1).

BldC activates the transcription of a subset of target genes.

Strikingly, our RNA-seq data also reveal that large numbers of genes are significantly (greater than 2-fold) downregulated in the bldC mutant during vegetative growth (Table S1C). Many of these genes are not direct BldC targets but nevertheless encode proteins important for the formation of an aerial mycelium, consistent with the bypassing of aerial hypha formation in the bldC mutant (Table S1E). For example, for aerial hyphae to break surface tension and grow into the air, they must be covered in an extremely hydrophobic sheath that is composed of two families of developmentally regulated proteins, the chaplins and the rodlins. In wild-type S. venezuelae, expression of the chp and rdl genes is activated at the onset of development, both on plates and during submerged sporulation (15). In contrast, the chp and rdl genes were not activated during submerged sporulation in the bldC mutant (Table S1E). It seems likely that the lack of expression of the chp and rdl genes will be an important contributing factor to the failure of bldC mutants to erect aerial hyphae.

In addition to these indirect effects, 91 direct BldC target genes showed a greater than 2-fold reduction in expression (logFC, <−1) in the bldC mutant compared to the wild type during vegetative growth, implying that BldC functions as an activator of these genes (Table S1F and Fig. 7). Two of these BldC target genes encode the developmental regulators BldM and WhiH, both of which showed significant downregulation in the absence of bldC during vegetative growth in the RNA-seq data (Table 1), a result confirmed by qRT-PCR (Fig. 6). Therefore, BldC functions to activate the transcription of bldM and whiH, which contrasts with its repression of other key developmental genes (e.g., whiI, smeA, sigF, whiD, and hupS) and the observed premature initiation of development in a bldC mutant. Our ChIP-seq data coupled with our RNA-seq data also suggest that BldC binds and activates the transcription of ssgA and ssgR, the latter encoding the sporulation-specific activator of ssgA (26) (Fig. 5 and Table 1).

Other noticeable targets of BldC-mediated activation include members of a family of highly conserved operons, known as the “conservons.” Each conservon (cvn) consists of four or five genes encoding proteins that, based on biochemical studies of Cvn9, may collectively form complexes of proteins at the membrane with roles in signal transduction (27). Our RNA-seq data indicate that each of the seven conservons present on the S. venezuelae chromosome (cvn1, cvn2, cvn3, cvn4, cvn5, cvn7, and cvn9) shows a significant reduction in expression during vegetative growth in the bldC mutant compared to the wild type (Table S1E). Two (cvn1 and cvn4) are direct targets of BldC, as determined by ChIP-seq (Fig. 5 and Table 1). The promoter upstream of cvn1 is also bound by WhiAB (17), and a cvn1 mutant of S. coelicolor is impaired in aerial mycelium formation (28), collectively suggesting that Cvn1 (and perhaps other members of the conservon family) may play a significant but as-yet-undefined role in Streptomyces differentiation.

The importance of cyclic-di-GMP (c-di-GMP) in the control of Streptomyces differentiation became clear with the discovery that engineering high levels of this nucleotide second messenger blocks entry into development, resulting in a classic bald phenotype, whereas engineering low levels of c-di-GMP causes precocious hypersporulation (5, 9). These phenotypes arise, at least in part, because the ability of the master repressor, BldD, to dimerize and repress a suite of sporulation genes during vegetative growth depends on binding to c-di-GMP (5, 9, 29, 30). c-di-GMP metabolism therefore plays a critical role in coordinating entry into reproductive growth. c-di-GMP is synthesized from two molecules of GTP by diguanylate cyclases (DGCs), and our ChIP-seq and RNA-seq data show that expression of cdgE, encoding a predicted DGC, is directly activated by BldC (Fig. 5 and Table 1). cdgE is present in at least 90% of Streptomyces strains, suggesting the role of this DGC will be widely conserved in the genus (31).

Streptomyces spp. are noted producers of the terpene 2-methylisoborneol (2-MIB), one of the volatiles that give soil its characteristic earthy odor. Our RNA-seq data show that expression of the genes required for 2-MIB biosynthesis (mibA-mibB) was significantly reduced (30-fold and 85-fold decreases, respectively, at the 10-h time point) in the bldC mutant compared to the wild type (Table S1C). Unexpectedly, the mibA-mibB genes were found to form an operon with eshA. eshA encodes a putative cyclic nucleotide-binding protein of unclear function that is not required for the biosynthesis of 2-MIB (32–34). The effect of BldC on mibAB expression is direct. ChIP-seq analysis showed that BldC binds to the promoter of the eshA-mibA-mibB operon (Fig. 5 and Table 1) and that all three genes show a similar reduction in expression in the bldC mutant, indicating that BldC serves to activate transcription of the operon (Table 1).

DISCUSSION

One of the most striking aspects of the bldC phenotype is that FtsZ ladders appear at an early growth stage but that sporulation septation does not happen until much later, approximately at the same time it is observed in the wild type. For cell division to occur, FtsZ must recruit other components of the divisome. It therefore seems likely that the delay that occurs in the bldC mutant between FtsZ ladder formation and sporulation septation arises because one or more of those other divisome components is absent.

Canonical bld mutations block entry into development, and so the resulting colonies do not form aerial hyphae and spores. These mutations typically define positive regulators such as the response regulator BldM (16) or the sigma factor BldN (15). Although our data indicate that BldC can function as both an activator and a repressor, we have shown that S. venezuelae bldC mutants are bald because they enter development prematurely, bypassing the formation of aerial hyphae, and that this correlates with premature expression of a subset of BldC target genes with roles in Streptomyces differentiation. Thus, phenotypically, BldC functions as a repressor to sustain vegetative growth and delay entry into development. As such, BldC joins a growing class of Bld regulators known to function as a developmental “brake” (8).

BldD was the first Bld regulator of this alternative class to be clearly recognized. BldD sits at the top of the developmental cascade and represses a large regulon of ∼170 sporulation genes during vegetative growth. BldD activity is controlled by the second messenger c-di-GMP, which mediates dimerization of two BldD protomers to generate a functional repressor. In this way, c-di-GMP signals through BldD to repress expression of the BldD regulon, extending vegetative growth and inhibiting entry into development (5, 9, 29, 30). Because a BldD-(c-di-GMP) complex represses the BldD regulon and not BldD alone, engineering the degradation of c-di-GMP in vivo also causes a precocious hypersporulation phenotype like that of a bldD null mutant (9).

More recently, bldO was identified as a second member of this emerging class of bld mutants (7, 8). In contrast to BldD and BldC, which both control large regulons, BldO functions to repress a single developmental gene, whiB. The precocious hypersporulation phenotype of the bldO mutant arises from premature expression of whiB, and in line with this, constitutive expression of whiB alone is sufficient to induce precocious hypersporulation in wild-type S. venezuelae (7). WhiA and WhiB act together to cocontrol the same set of promoters to initiate developmental cell division in Streptomyces (14, 17). WhiA is constitutively present throughout the life cycle, but it only binds to its target promoters at the onset of sporulation when WhiB is present (14, 17). This is because WhiA and WhiB function cooperatively and in vivo DNA binding by WhiA depends on WhiB and vice versa (17). As a consequence, the regulation of whiB expression is key in controlling the switch between hyphal growth and sporulation. This critical role for WhiB is reflected in the extensive developmental regulation to which whiB transcription is subject, being directly repressed by BldC, BldD (29), and BldO (7) and directly activated by BldM (16). It should be noted, however, that the repressive effect of BldC on whiB expression is mild (Table 1).

BldC-family members radiate throughout the bacterial domain. Interestingly, some BldC orthologs are annotated as possible DNA resolvase/integrase-associated proteins, consistent with the structural similarity observed between BldC and Xis (11, 35). Xis is a DNA architectural protein that mediates the formation of a nucleoprotein complex required for the phage-encoded Int recombinase/integrase to catalyze the site-specific recombination event that leads to the excision of phage lambda from the E. coli chromosome. Like BldC, Xis binds to DNA in a head-tail fashion to generate a nucleoprotein filament, leading to distortion of the DNA (35). Thus, BldC may represent an evolutionary link between transcription factors of the MerR family and DNA architectural proteins (11). Only one proteomic survey of the Streptomyces nucleoid has been published, in which BldD was detected but BldC was not (36).

There is an interesting analogy between the relationship of BldC to MerR and the relationship of Fis to NtrC. Fis is a 98-residue nucleoid-associated protein found in proteobacteria that is closely related to the DNA-binding domain of the much larger bacterial enhancer-binding protein NtrC (37–39). Like BldC, Fis prefers binding to AT-rich DNA and its interaction with DNA is affected by the width of the minor groove (40). Fis can function in the cell as an architectural protein in the nucleoid, but it can also function as a transcription factor (41, 42). Like BldC, Fis exerts a global influence on the transcription profile of the cell and can have positive or negative effects on the activity of its target promoters (43). Fis does not bind a ligand, and it is not known to be controlled by posttranslational modification. Instead, its influence appears simply to reflect Fis protein concentration, which is high in early log phase but low at other growth stages. In the future, it will be interesting to determine if the activity of BldC is controlled posttranslationally, or whether BldC function is more akin to that of nucleoid-associated proteins like Fis.

MATERIALS AND METHODS

Bacterial strains, plasmids, and oligonucleotides.

Bacterial strains, plasmids, and oligonucleotides are listed in Table S2 in the supplemental material.

TABLE S2

Strains, plasmids, and oligonucleotide primers used in this study. Download Table S2, DOCX file, 0.02 MB ^{(17.5KB, docx)} .

Copyright © 2019 Bush et al.

This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.

The S. venezuelae genome sequence.

The S. venezuelae strain used in this work is the same strain used in previous papers from this lab (7, 9, 12–17). The final and complete genome sequence of this strain has been determined and submitted to the NCBI database under reference NZ_CP018074.1. In parallel, to unambiguously link the two, the sequenced strain has been deposited with the NRRL culture collection under stock code NRRL B-65442.

Construction and complementation of an S. venezuelae bldC null mutant.

Using “Redirect” PCR targeting (44, 45), bldC mutants were generated in which the coding region was replaced with a single apramycin resistance (apr) cassette. A cosmid library that covers >98% of the S. venezuelae genome (M. J. Bibb and M. J. Buttner, unpublished data) is fully documented at http://strepdb.streptomyces.org.uk/. Cosmid 4O24 was introduced into E. coli BW25113 containing pIJ790, and the bldC gene (vnz18945) was replaced with the apr-oriT cassette amplified from pIJ773 using the primer pairs bldCdis_F and bldCdis_R. The resulting disrupted cosmids were confirmed by restriction digestion and by PCR analysis using the flanking primers bldCcon_F and bldCcon_R and introduced into S. venezuelae by conjugation (46). Null mutant derivatives, generated by double crossing over, were identified by their apramycin-resistant, kanamycin-sensitive, and morphological phenotypes, and their chromosomal structures were confirmed by PCR analysis using the flanking primers bldCcon_F and bldCcon_R. A representative bldC null mutant was designated SV25. For complementation, bldC was amplified with the primers bldCcomp_F and bldCcomp_R, generating an 846-bp fragment carrying the coding sequence and the bldC promoter, and cloned into HindIII-KpnI/Asp718-cut pIJ10770 to create pIJ10618. The plasmid was introduced into the bldC mutant by conjugation, integrating in trans at the ΦBT1 attB site, and fully complemented all aspects of the mutant phenotype.

ChIP-seq, RNA-seq, qRT-PCR, Western blotting, time-lapse imaging, and scanning electron microscopy.

For ChIP-seq, RNA-seq, qRT-PCR, Western blotting, time-lapse imaging, and scanning electron microscopy, please see the supplemental materials and methods (Text S1).

TEXT S1

Supplemental materials and methods. Download Text S1, DOCX file, 0.02 MB ^{(24.2KB, docx)} .

Copyright © 2019 Bush et al.

This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.

Data availability.

The BldC ChIP-seq data and RNA-seq transcriptional profiling data have been deposited at the MIAME-compliant ArrayExpress database (https://www.ebi.ac.uk/arrayexpress/) under accession numbers E-MTAB-7450 (ChIP-seq data) and E-MTAB-7457 (transcriptional profiling data). The final and complete genome sequence of the S. venezuelae strain has been determined and submitted to the NCBI database under reference NZ_CP018074.1.