Heat Shock Repressor HspR Directly Controls Avermectin Production, Morphological Development, and H₂O₂ Stress Response in Streptomyces avermitilis

ABSTRACT

The heat shock response (HSR) is a universal cellular response that promotes survival following temperature increase. In filamentous Streptomyces, which accounts for ∼70% of commercial antibiotic production, HSR is regulated by transcriptional repressors; in particular, the widespread MerR-family regulator HspR has been identified as a key repressor. However, functions of HspR in other biological processes are unknown. The present study demonstrates that HspR pleiotropically controls avermectin production, morphological development, and heat shock and H₂O₂ stress responses in the industrially important species Streptomyces avermitilis. HspR directly activated ave structural genes (aveA1 and aveA2) and H₂O₂ stress-related genes (katA1, catR, katA3, oxyR, ahpC, and ahpD), whereas it directly repressed heat shock genes (HSGs) (the dnaK1-grpE1-dnaJ1-hspR operon, clpB1p, clpB2p, and lonAp) and developmental genes (wblB, ssgY, and ftsH). HspR interacted with PhoP (response regulator of the widespread PhoPR two-component system) at dnaK1p to corepress the important dnaK1-grpE1-dnaJ1-hspR operon. PhoP exclusively repressed target HSGs (htpG, hsp18_1, and hsp18_2) different from those of HspR (clpB1p, clpB2p, and lonAp). A consensus HspR-binding site, 5′-TTGANBBNNHNNNDSTSHN-3′, was identified within HspR target promoter regions, allowing prediction of the HspR regulon involved in broad cellular functions. Taken together, our findings demonstrate a key role of HspR in the coordination of a variety of important biological processes in Streptomyces species.

IMPORTANCE Our findings are significant to clarify the molecular mechanisms underlying HspR function in Streptomyces antibiotic production, development, and H₂O₂ stress responses through direct control of its target genes associated with these biological processes. HspR homologs described to date function as transcriptional repressors but not as activators. The results of the present study demonstrate that HspR acts as a dual repressor/activator. PhoP cross talks with HspR at dnaK1p to coregulate the heat shock response (HSR), but it also has its own specific target heat shock genes (HSGs). The novel role of PhoP in the HSR further demonstrates the importance of this regulator in Streptomyces. Overexpression of hspR strongly enhanced avermectin production in Streptomyces avermitilis wild-type and industrial strains. These findings provide new insights into the regulatory roles and mechanisms of HspR and PhoP and facilitate methods for antibiotic overproduction in Streptomyces species.

INTRODUCTION

Streptomyces species are Gram-positive, filamentous soil bacteria that undergo complex morphological development involving the formation of substrate hyphae, aerial hyphae, and spore-bearing hyphae (1). They are an economically important group that produce a wide variety of secondary metabolites, notably, antibiotics having antibacterial, antifungal, antiviral, anthelmintic, anticancer, or immunosuppressive activity (2, 3). Antibiotic biosynthesis is generally associated with development and controlled by multiple levels of transcriptional regulators (TRs), including cluster-situated regulators (CSRs) and higher-level pleiotropic/global regulators, in response to environmental and physiological cues such as low nutrient availability, temperature changes, and pH changes (4–6).

All cellular organisms respond to sudden temperature increases by substantially raising levels of heat shock proteins (HSPs). HSPs are divided into two major classes: (i) molecular chaperones that promote correct folding of denatured and newly synthesized proteins and prevent formation of insoluble protein aggregates (e.g., DnaK, DnaJ, GrpE, GroES, and GroEL), and (ii) ATP-dependent proteases involved in the degradation of denatured proteins (e.g., ClpAP, ClpXP, ClpB, and Lon) (7–9). These HSPs are essential for normal cell growth and are synthesized at a basal level under physiological conditions. Induction of HSPs under stress conditions (particularly heat shock) is a universal response, but synthesis of HSPs is controlled by a variety of mechanisms depending on the bacterial species. In Escherichia coli, two alternative sigma factors, Sig32 and SigE, activate heat shock genes (HSGs) in response to denatured proteins generated in cytoplasm (Sig32) or periplasm (SigE) following stimulation (10). In Streptomyces, HSGs are mainly negatively regulated by three repressors: HrcA, RheA, and HspR (10). HrcA in Streptomyces albus represses hrcA-dnaJ2, groES-groEL1, and groEL2 operons (11). The RheA regulon of S. albus consists of its own gene rheA and nearby divergently transcribed gene hsp18 (12). The HspR repressor is a member of the MerR-family TRs that contain conserved helix-turn-helix (HTH) DNA-binding motif at the N terminus and function as homodimers. HspR represses the dnaK-grpE-dnaJ-hspR operon and clpB and lon genes by interacting with HspR-binding site HAIR (HspR-associated inverted repeat [5′-TTGAGY-N₇-ACTCAA-3′]) at target promoter regions in Streptomyces coelicolor, S. albus, and Streptomyces lividans (13–17). The S. coelicolor DnaK chaperone functions as a corepressor of the HspR regulon by binding to HspR at HAIR sites (14, 18). HspR itself can bind DNA targets, but formation of a stable complex with the DNA targets requires HspR-DnaK interaction. During heat shock, DnaK is titrated by denatured proteins and unable to bind to HspR, leading to the induction of target HSGs (14, 18). The reported studies have only addressed the role of Streptomyces HspR in the heat shock response (HSR), and nothing is known regarding its role in Streptomyces development and antibiotic production.

HspR is present in other actinomycetes besides Streptomyces, including Mycobacterium tuberculosis (19). HSPs contribute to M. tuberculosis virulence and are induced following entry into host cells to protect M. tuberculosis against stress imposed by host macrophages. Among M. tuberculosis HSPs, Acr2 is a member of the α-crystallin family of molecular chaperones and plays a key role during the M. tuberculosis stress response. HspR functions as a repressor of acr2 expression (19). Another acr2 repressor, PhoP, the response regulator of the widespread PhoPR two-component system, interacts with HspR at the acr2 promoter region to coregulate acr2 expression (20). PhoP is not necessary for the DNA-binding activity of HspR; however, HspR-PhoP interaction stabilizes the higher-order DNA-protein complex to prevent access of RNA polymerase to the acr2 promoter, thus repressing the initiation of transcription. M. tuberculosis dnaK is repressed by HspR (19) but is not regulated by HspR-PhoP interactions (20).

The PhoPR two-component system, which senses and responds to phosphate limitation stress, is well conserved in Streptomyces. PhoP acts as a master regulator for the coordination of a variety of physiological processes, including phosphate metabolism, nitrogen metabolism, respiration, development, and antibiotic biosynthesis (21–24). PhoP cross talks with GlnR (main nitrogen metabolism regulator) and AfsR (response regulator of AfsRK two-component system that promotes antibiotic biosynthesis in S. coelicolor) by competing for binding to the same target promoter regions in control of nitrogen metabolism and antibiotic biosynthesis (25, 26). A role of PhoP in HSR and its cross talk with other regulators in Streptomyces has not been reported.

The industrial species Streptomyces avermitilis produces avermectins, a series of 16-membered macrocyclic anthelmintic antibiotics that are widely used in agricultural and medical fields and have great commercial importance (27, 28). The heat shock repressor HspR has not yet been studied in S. avermitilis. The role of HspR in antibiotic production and development in this species is of interest in view of previous findings that it is involved in stress response and that antibiotic biosynthesis in the genus is associated with development and occurs under stress conditions.

Here, we characterize HspR (SAV_4487) in S. avermitilis as a dual repressor/activator of avermectin production, development, and heat shock and H₂O₂ stress responses and identify HspR target genes associated with these biological processes. HspR interacted with PhoP at the dnaK1 regulatory region to corepress the dnaK1-grpE1-dnaJ1-hspR operon. PhoP had exclusive target HSGs (htpG, hsp18_1, and hsp18_2) different from those of HspR (clpB1p, clpB2p, and lonAp). We propose a novel strategy for enhancing antibiotic production through overexpression of the hspR gene.

RESULTS

HspR represses S. avermitilis development but activates avermectin production.

The gene hspR (sav_4487) in S. avermitilis contains 450 nucleotides (nt) and encodes a 16.8-kDa protein, HspR. The S. avermitilis chromosome has two copies of HSGs dnaK, grpE, and dnaJ. hspR is clustered with dnaK1 (sav_4484), grpE1 (sav_4485), and dnaJ1 (sav_4486) to form the operon dnaK1-grpE1-dnaJ1-hspR (Fig. 1A). Protein alignment study showed that HspR is a highly conserved protein in Streptomyces; S. avermitilis HspR has 90.8%, 90.3%, 90.3%, 92%, and 90.8% amino acid identity with its homologs in S. coelicolor, S. venezuelae, S. griseus, S. scabies, and S. lividans, respectively, reflecting the biological importance of this protein in the genus.

FIG 1.

Effects of HspR on morphological development, avermectin production, and cell growth in S. avermitilis. (A) Genetic organization of the dnaK1-grpE1-dnaJ1-hspR operon. (B) Phenotypes of WT strain, the hspR deletion mutant (ΔhspR), complemented strain (ChspR), and overexpression strain (OhspR) grown on YMS agar at 28°C. (C) SEM images of WT, ΔhspR, ChspR, and OhspR strains grown on YMS agar for 2 or 4 days. (D) Avermectin yield in WT, ΔhspR, ChspR, OhspR, and control strains (WT/pKC1139 and WT/pSET152) after 10-day culture in FM-I. (E) Growth curves of WT, ΔhspR, and OhspR strains cultured in FM-II. (F) Avermectin yield in industrial strain A229 and its derivatives OhspR/A229-1, -2, and -3 (hspR overexpression strains) after 10-day culture in FM-I. Error bars (panels D, E, and F) indicate standard deviations (SDs) from three replicates. NS, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001 for comparison with WT (D) or A229 (F) (Student’s t test).

To elucidate the functions of HspR in S. avermitilis, we constructed hspR deletion mutant ΔhspR (see Fig. S1 in the supplemental material), complemented strain ChspR, and overexpression strain OhspR. hspR transcription was undetectable in the ΔhspR strain as shown by reverse transcription and real-time quantitative PCR (RT-qPCR) analysis, and its levels were ∼1.8-fold higher on day 2 (exponential phase) and ∼3.1-fold higher on day 6 (stationary phase) in OhspR than in the wild-type (WT) strain (see Fig. S2), confirming the successful deletion or overexpression of hspR in the above-described strains.

The effect of HspR on morphological development was investigated by growing WT, ΔhspR, OhspR, and ChspR strains on solid sporulation yeast extract-malt-starch (YMS) plates. OhspR and ChspR strains were phenotypically similar to the WT, whereas the ΔhspR strain showed earlier and enhanced formation of spores (Fig. 1B). Spore numbers on days 2 and 4 were higher for the ΔhspR strain than for WT, as shown by scanning electron microscopy (SEM) (Fig. 1C). These findings indicate that HspR functions as a repressor during S. avermitilis development.

The effect of HspR on avermectin biosynthesis was investigated by high-performance liquid chromatography (HPLC) analysis of 10-day insoluble FM-I fermentation broth. In comparison to the WT value, the avermectin yield of the ΔhspR strain was ∼43% lower, that of the OhspR strain was ∼154% higher, and those of the ChspR and control strains (WT/pKC1139, WT/pSET152) were not significantly different (Fig. 1D and see Fig. S3). Time course measurements of growth in soluble FM-II showed that biomass (dry cell weight) values of the ΔhspR and OhspR strains were similar to that of the WT (Fig. 1E), ruling out the possibility that altered avermectin yields in the ΔhspR strain and the OhspR strain were due to changes in growth. These findings demonstrate that HspR functions as an activator in avermectin production.

In view of the finding that hspR overexpression increased avermectin yield in the WT, we transformed hspR overexpression vector pKC-erm-hspR into industrial strain A229 to construct OhspR/A229. Yield of avermectin B1a, the most effective component, was ∼14% to 17% higher for OhspR/A229 than in A229 (Fig. 1F). This finding indicates that avermectin yield could be substantially enhanced in high-producing industrial strains by hspR overexpression.

HspR directly activates ave structural genes.

The transcription profile of hspR in FM-I culture of the WT was monitored by RT-qPCR to further clarify the regulatory role of HspR in avermectin production. The hspR transcription level increased to maximum on day 5, followed by a gradual decrease, and the level on day 7 was similar to that on day 8 (Fig. 2A). The HspR protein expression level was examined concurrently by Western blotting. Avermectin production in strain ΔhspR/hspR-3FLAG (fusion protein HspR-3FLAG expressed in the ΔhspR strain) was similar to that in the WT (see Fig. S4), indicating that HspR-3FLAG complemented HspR function and that the HspR expression profile could be examined using anti-FLAG antibody against HspR-3FLAG in the ΔhspR/hspR-3FLAG strain. Consistent with the transcription profile, the HspR protein level was maximal on day 5 and then declined gradually (Fig. 2B). These findings suggest that HspR plays its regulatory role particularly during the middle fermentation stage or that there is something about this stage in growth that is particularly relevant to one of the stresses modulated by this protein.

FIG 2.

Direct activation of aveA1-aveA2 by HspR. (A) RT-qPCR analysis of hspR transcriptional pattern in the WT grown in FM-I. The hspR transcription level on day 1 was set to 1. (B) Western blotting analysis of the HspR protein level during the fermentation process. The HspR expression level in strain ΔhspR/hspR-3FLAG grown in FM-I was determined using anti-FLAG MAb. (C) Schematic diagram of promoter probes used in EMSAs. (D) In vitro EMSAs of interactions of His₆-HspR with indicated probes (hrdBp as negative-control probe). A 0.15 nM concentration of labeled probe and various amounts of His₆-HspR were used for each binding reaction. In competition experiments, ∼300-fold unlabeled nonspecific probe hrdBp (lane N) or specific probe aveA1p (lane S) was used. Arrows indicate free probes. Bracket indicates the HspR-DNA complex. (E) In vivo ChIP-qPCR assays of HspR binding to aveA1p and aveRp using ΔhspR/hspR-3FLAG and WT (negative control) strains grown in FM-II for the indicated times. The y axis indicates the relative binding level of HspR-FLAG on each site, determined by recovery of target sequence with anti-FLAG MAb. (F) RT-qPCR analysis of aveA1 and aveA2 in WT and ΔhspR strains grown in FM-I. WT value on day 2 for each gene was set to 1. Error bars (panels A, E, and F) indicate SDs from three replicates. NS, not significant; ***, P < 0.001 (t test).

To determine whether HspR regulates avermectin production through CSR gene aveR (29, 30) or structural genes, we performed a series of in vitro electrophoretic mobility shift assays (EMSAs) using soluble His₆-HspR expressed in and purified from E. coli. The four genes in the ave gene cluster (aveA1, aveA2, aveA3, and aveA4) encode polyketide synthases (PKSs) responsible for the synthesis of the avermectin polyketide backbone. aveA1 is cotranscribed with aveA2, and aveA4 is cotranscribed with aveA3 (31). We therefore designed promoter probes aveRp, aveA1p (for aveA1-aveA2), and aveA4p (for aveA4-aveA3) for EMSAs (Fig. 2C) and used nonspecific probe hrdBp as the control. His₆-HspR clearly retarded aveA1p but did not bind to aveA4p, aveRp, or hrdBp (Fig. 2D). Binding specificity was confirmed by adding ∼300-fold unlabeled specific probe aveA1p (lane S), which abolished the retarded band, or unlabeled nonspecific probe hrdBp (lane N), which had no effect on the delayed signal. Because of the cotranscription of aveA1 and aveA2, aveA2 is also targeted by HspR. These findings indicate that HspR directly regulates ave structural genes (aveA1 and aveA2) but not CSR gene aveR.

In vivo binding of HspR to aveA1p was confirmed by chromatin immunoprecipitation-quantitative PCR (ChIP-qPCR) assays. Samples were taken from WT and ΔhspR/hspR-3FLAG strains grown in FM-II for various durations. HspR bound to aveA1p at various time points. Binding was strongest on day 5 (consistent with HspR expression profile), whereas HspR was not enriched on aveRp (Fig. 2E), indicating dynamic binding of HspR to the target promoter aveA1p in vivo.

The effect of HspR on expression of targeted aveA1 and aveA2 genes was assessed by RT-qPCR, using RNAs isolated from 2-day and 6-day FM-I cultures of WT and ΔhspR srains. aveA1 and aveA2 transcription levels were lower in the ΔhspR strain than in the WT on both days (Fig. 2F), consistent with avermectin yields for the strains, indicating that HspR activates transcription of these two genes.

Determination of precise HspR-binding site on the aveA1 promoter region.

To identify the precise HspR-binding site on aveA1p and clarify the mechanism whereby HspR regulates the aveA1 gene, we performed DNase I footprinting assays. HspR protected a 31-nt region containing a 19-nt sequence (TTGAACGTCTTCAACTCTT) (Fig. 3A) similar to the conserved HspR-binding site HAIR in S. coelicolor (14), suggesting that the DNA-binding property of HspR is conserved.

FIG 3.

HspR-binding site on aveA1 promoter region. (A) DNase I footprinting assay of HspR on aveA1p. Protection patterns were acquired with increasing His₆-HspR concentrations (reaction without His₆-HspR used as a control). (B) EMSAs using 50-nt WT probe aveA1p-1 and its mutated probe aveA1p-1m. Imperfect inverted repeats in probe aveA1p-1 were replaced with HindIII and BamHI sites to produce mutated probe aveA1p-1m. Each lane contained 0.15 nM labeled probe. (C) Nucleotide sequences of aveA1 promoter region and HspR-binding site. Numbers indicate the distance (nucleotides) from aveA1 TSS. Box, aveA1 translational start codon (TSC); bent arrow, aveA1 TSS; shading, probable −10 and −35 regions; solid line, HspR-binding site; straight arrows, inverted repeats.

The importance of the 19-nt HAIR-like sequence in HspR binding was evaluated by performing EMSAs using 50-nt probes that contained either the intact sequence (termed probe aveA1p-1) or the mutated sequence (termed probe aveA1p-1m; lacking inverted repeats) (Fig. 3B). His₆-HspR bound to WT probe aveA1p-1 but not to mutated probe aveA1p-1m (Fig. 3B), indicating that the 6-nt inverted repeats within the HAIR-like sequence are essential for HspR binding.

The transcriptional start site (TSS) of the aveA1 gene was determined by our group previously (32), and −35 and −10 promoter sequences were predicted on this basis (Fig. 3C). The 19-nt HspR-binding site on aveA1p is very close to the −10 region (3 nt downstream) and overlaps the aveA1 TSS (Fig. 3C). The HspR-binding site on aveA1p is unusual in regard to transcriptional activation; however, it is analogous to the binding sites of Streptomyces antibiotic regulatory protein (SARP)-family regulators AfsR (33) and OtcR (34), which are close to the −10 regions. These regulators presumably activate target transcription by recruiting RNA polymerase to the promoters. The mechanism of such activation of aveA1 by HspR remains to be elucidated.

Identification of S. avermitilis HspR target HSGs.

To examine response of HspR to heat shock stress in S. avermitilis, we measured the sensitivity of WT and ΔhspR strains to heat treatment. Growth of the ΔhspR strain was more resistant to heat shock stress than that of the WT (Fig. 4A), indicating that HspR represses the HSR, consistent with reported functions of its homologs in S. coelicolor and S. albus (13–16).

FIG 4.

Identification of HspR target HSGs. (A) Sensitivity of WT and ΔhspR strains to heat stress. Spore suspensions were treated at 42°C, 50°C, or 60°C for 10 min, spotted on YMS agar, and photographed after 3-day growth at 28°C. (B) Promoter probes of HSGs used in EMSAs. (C) EMSAs of His₆-HspR with indicated probes. Notations as in Fig. 2D. (D) ChIP-qPCR assays of HspR binding to dnaK1p, clpB1p, clpB2p, lonAp, and dnaK2p. Notations as in Fig. 2E. (E) RT-qPCR analysis of five HspR target HSGs in WT and ΔhspR strains grown in FM-I. Error bars (panels D and E) indicate SDs from three replicates. ***, P < 0.001 (t test).

To identify HspR target HSGs, we performed EMSAs using His₆-HspR and promoter probes of potential HSGs based on the S. avermitilis genome database (http://avermitilis.ls.kitasato-u.ac.jp): dnaK1p (for the dnaK1-grpE1-dnaJ1-hspR operon), dnaK2p (for the dnaK2-grpE2-dnaJ2 operon), clpB1p, clpB2p, lonAp, and groES1p (for the groES1-groEL1 operon), groEL2p, htpGp, hsp18_1p, and hsp18_2p (Fig. 4B). His₆-HspR bound specifically to dnaK1p, clpB1p, clpB2p, and lonAp but not to other probes (Fig. 4C). hspR belongs to the dnaK1-grpE1-dnaJ1-hspR operon; therefore, HspR is autoregulated. Direct binding of HspR to dnaK1p, clpB1p, clpB2p, and lonAp in vivo was confirmed by ChIP-qPCR assays (Fig. 4D). Transcription levels of dnaK1, hspR, clpB1, clpB2, and lonA were shown by RT-qPCR analysis to be strongly upregulated in the ΔhspR strain on days 2 and 6 (Fig. 4E), indicating that HspR functions as a repressor of these target HSGs, consistent with the ΔhspR phenotype in heat stress assays.

To determine whether dnaK1, hspR, clpB1, clpB2, and lonA were induced in an HspR-dependent manner under heat stress, WT and ΔhspR strains were cultured in FM-II for 2 days, followed by heat (50°C) treatment for various durations. For the WT, heat treatment resulted in maximal induction of dnaK1 (∼49-fold), hspR (∼43-fold), clpB1 (∼84-fold), clpB2 (∼1.4-fold), and lonA (∼52-fold) within 10 min (see Fig. S5), indicating that HSPs DnaK1, ClpB1, and LonA play key roles in the HSR. For the ΔhspR strain, transcription levels of these five genes were all higher than the levels for the WT and also reached maximal values within 10 min of heat treatment. These findings confirm the negative role of HspR in controlling these target HSGs and suggest that these genes are also controlled by other regulator(s).

DNase I footprinting assays revealed that HspR protects two sites (dnaK1-I, dnaK1-II) on dnaK1p and that both sites contain a 19-nt HAIR-like sequence (Fig. 5A). The role of HAIR-like sequences in HspR binding was investigated by site-directed mutagenesis of 50-bp WT probes dnaK1p-1 and dnaK1p-2 on the inverted repeats to generate mutated probes dnaK1p-1m and dnaK1p-2m, respectively (Fig. 5B). EMSAs revealed that His₆-HspR affinity for the mutated probes was abolished in comparison to that for the corresponding WT probes (Fig. 5B), reflecting an essential role of HAIR-like sequences in HspR binding. Site dnaK1-I extends from positions −63 to −39, and site dnaK1-II extends from positions −12 to +15 relative to the dnaK1 TSS (35) (Fig. 5C). Site dnaK1-I is close to the putative −35 region, and site dnaK1-II overlaps the putative −10 region of the dnaK1 promoter and dnaK1 TSS, indicating that HspR represses dnaK1 by blocking RNA polymerase access.

FIG 5.

HspR-binding site on dnaK1 promoter region. (A) DNase I footprinting assay of HspR on dnaK1p. (B) EMSAs using 50-nt WT probes dnaK1p-1 and dnaK1p-2 and their mutated probes dnaK1p-1m and dnaK1p-2m. Each lane contained 0.15 nM labeled probe. (C) Nucleotide sequences of dnaK1 promoter region and HspR-binding sites. Numbers indicate distance (nucleotides) from dnaK1 TSS. Black box, dnaK1 TSC; bent arrow, dnaK1 TSS; red boxes, PHO box. Other notations as in Fig. 3C.

The clpB1, clpB2, and lonA promoter regions also contain a 19-nt sequence similar to the HAIR motif, indicating that S. avermitilis HspR binds to these target DNAs at HAIR sites.

HspR responds to H₂O₂ stress.

Possible regulation of other types of stress response by S. avermitilis HspR was evaluated on YMS plates. Relative to that of the WT, the ΔhspR strain showed greater sensitivity to H₂O₂ (which causes peroxidative stress) but similar sensitivity to tert-butyl hydroperoxide (TBHP; causes organic peroxidative stress), diamide (causes thiol-oxidative stress), and NaCl and KCl (cause osmotic stress) (Fig. 6A). S. avermitilis HspR evidently plays a role in resistance to H₂O₂ stress.

FIG 6.

Identification of HspR target genes involved in H₂O₂ stress response. (A) Sensitivity of WT and ΔhspR strains to various stress conditions. Spore suspensions were diluted serially and spotted on YMS agar containing NaCl, KCl, TBHP, H₂O₂, or diamide at the indicated concentrations. (B) Promoter probes of H₂O₂ stress-related genes used in EMSAs. (C) EMSAs of His₆-HspR with indicated probes. Notations as in Fig. 2D. (D) ChIP-qPCR assays of HspR binding to katA2p, katA3p, catR_katA1, and oxyR_ahpCD. Notations as in Fig. 2E. (E) RT-qPCR analysis of six HspR target H₂O₂ stress-related genes in WT and ΔhspR strains grown in FM-I. Error bars (panels D and E) indicate SDs from three replicates. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (t test).

Bacteria typically respond to H₂O₂ stress by producing peroxidases and catalases that degrade H₂O₂. In S. avermitilis, one ahpCD operon (for alkyl hydroperoxide reductase and alkylhydroperoxidase), three catalase genes (katA1, katA2, and katA3), and two TR genes (oxyR and catR) are involved in H₂O₂ stress response (36). Possible interactions of His₆-HspR with promoter probes of these H₂O₂ stress-related genes (Fig. 6B) were investigated by EMSAs. His₆-HspR bound specifically to probe katA3p and bidirectional promoter probes catR_katA1 and oxyR_ahpCD but not to katA2p (Fig. 6C). HspR-binding promoter regions katA3p, catR_katA1, and oxyR_ahpCD all contained HAIR-like sequences, as expected. Direct binding of HspR to katA3p, catR_katA1, and oxyR_ahpCD was confirmed in vivo by ChIP-qPCR assays (Fig. 6D). Consistent with the H₂O₂ stress phenotype for the ΔhspR strain, transcription levels of katA1, catR, katA3, oxyR, ahpC, and ahpD were all reduced in the ΔhspR strain relative to that in the WT grown in FM-I (Fig. 6E), indicating that HspR acts as an activator of these genes.

For analysis of H₂O₂ stress responses, WT and ΔhspR strains were treated with 0.6 mM H₂O₂ for various durations. In the WT, H₂O₂ treatment caused notable induction of katA1 (∼47-fold) and catR (∼8.7-fold) within 10 min and slight induction of katA3 (∼2.7-fold), ahpC (∼1.39-fold), and ahpD (∼1.36-fold) within 30 min and of oxyR (∼2.2-fold) within 40 min, whereas H₂O₂ treatment of the ΔhspR strain had a much lower inducing effect on katA1 (∼5-fold), catR (∼4.6-fold), and oxyR (∼1.7-fold) and no effect on katA3, ahpC, or ahpD expression (see Fig. S6). These findings indicate that HspR promotes H₂O₂ stress resistance in S. avermitilis by activating transcription of target genes (katA1, catR, katA3, oxyR, ahpC, and ahpD).

The effect of H₂O₂ on avermectin production was also investigated by HPLC analysis of 10-day cultures of the WT in FM-I containing various concentrations of H₂O₂, and the results showed that H₂O₂ addition did not promote avermectin production (see Fig. S7).

HspR interacts with PhoP at the dnaK1 promoter region.

Cross talk between different regulatory systems commonly occurs in bacteria for the coordination of cell growth and metabolism. In M. tuberculosis, HspR interacts with PhoP to coregulate expression of acr2, which encodes an essential pathogenic determinant (20). We examined the possibility that HspR also interacts with PhoP in Streptomyces. The Streptomyces PhoP-binding sequence (termed PHO box) consists of 11-nt consecutive direct repeat units (DRus; G^G/TTCAYYYR^G/CG) (37). We found that dnaK1p contains a PHO box sequence formed by four DRus and that the PHO box overlaps site dnaK1-II (Fig. 5C), suggesting that HspR-PhoP interaction may occur at dnaK1p. To test this possibility, we coexpressed His₆-HspR with glutathione transferase (GST) or GST-PhoP in E. coli and performed GST pulldown assays. His₆-HspR was pulled down by GST-PhoP but not by GST tag (negative control) (Fig. 7A), indicating that HspR interacts with PhoP. Specific binding of purified GST-PhoP (38) to dnaK1p was confirmed by EMSAs (Fig. 7B).

FIG 7.

Interaction of HspR with PhoP at dnaK1p. (A) GST pulldown assays of HspR and PhoP from E. coli whole-cell lysate containing both His₆- and GST-tagged proteins. Lane 1, flowthrough of cell lysate after incubation with glutathione-Sepharose beads; lane 2, washing buffer following fourth wash of beads; lanes 3, GST pulldown. (B) EMSAs of GST-PhoP with probe dnaK1p. A 0.15 nM concentration of labeled probe and various amounts of GST-PhoP were used for each binding reaction. For nonspecific (lane N) or specific (lane S) competition assays, ∼500-fold unlabeled competitor DNA was used. (C) Competitive EMSAs of probe dnaK1p with His₆-HspR and GST-PhoP. A 0.15 nM concentration of labeled probe dnaK1p was incubated with indicated concentrations of His₆-HspR and GST-PhoP. (D) ChIP-qPCR assays of HspR-3FLAG binding to dnaK1p in WT (negative control), ΔhspR/hspR-3FLAG, WT/hspR-3FLAG, and ΔphoP/hspR-3FLAG strains grown in FM-II for the indicated times. Error bars indicate SDs from three replicates. ***, P < 0.001 (t test).

The PhoP-binding site on dnaK1p (PHO box) overlaps the HspR-binding site dnaK1-II; these two regulators may therefore affect each other’s binding to dnaK1p. We examined this possibility by applying His₆-HspR and GST-PhoP, both separately and together, with probe dnaK1p in EMSAs. When applied separately, both proteins retarded dnaK1p migration (Fig. 7C). In the presence of 2 μM GST-PhoP, increasing amounts of His₆-HspR resulted in a reduction of the PhoP-dnaK1p complex, an increase of the HspR-dnaK1p complex, and formation of a new retarded band located between those for PhoP-dnaK1p and HspR-dnaK1p (Fig. 7C, left), which was presumably formed by interaction of the HspR-PhoP complex with dnaK1p. In the presence of 0.05 μM His₆-HspR, increasing amounts of GST-PhoP resulted in reduction of the HspR-dnaK1p complex, appearance of PhoP-dnaK1p, and appearance of a new band, presumably HspR-PhoP-dnaK1p (Fig. 7C, right). These findings suggest that HspR and PhoP cooperate as a complex for DNA binding besides competing for binding to dnaK1p.

We used ChIP-qPCR assays to examine binding of HspR-3FLAG to dnaK1p in WT, ΔhspR/hspR-3FLAG, WT/hspR-3FLAG, and ΔphoP/hspR-3FLAG strains. HspR-3FLAG on dnaK1p was enriched (relative to that for negative-control WT) in ΔhspR/hspR-3FLAG, WT/hspR-3FLAG, and ΔphoP/hspR-3FLAG strains grown in FM-II for various durations (Fig. 7D), indicating that DNA-binding activity of HspR does not depend on PhoP. Enrichment levels of HspR-3FLAG on dnaK1p were highest in the ΔhspR/hspR-3FLAG strain and higher in the ΔphoP/hspR-3FLAG strain than in the WT/hspR-3FLAG strain at various time points (Fig. 7D), indicating competitive binding of HspR and PhoP to dnaK1p. HspR evidently plays a dominant role relative to PhoP in the regulation of dnaK1p.

PhoP represses target HSGs.

The observed binding of PhoP to dnaK1p suggests that PhoP may also be involved in the HSR. We evaluated this possibility by comparing the growth of WT and ΔphoP strains (38) on YMS plates following heat treatment. Resistance to heat shock stress was greater for the ΔphoP strain than for the WT (Fig. 8A), indicating that PhoP (like HspR) has a negative effect on HSR. In contrast to HspR, PhoP had no effect on the H₂O₂ stress response (Fig. 8A).

FIG 8.

Identification of PhoP target HSGs. (A) Sensitivity of WT and ΔphoP strains to heat and H₂O₂ stresses. (B) EMSAs of GST-PhoP with promoter probes of HSGs. Notations as in Fig. 7B. (C) RT-qPCR analysis of five PhoP target HSGs in WT and ΔphoP strains grown in FM-I. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (t test). (D) Induction of five HSGs by heat treatment (50°C) in WT and ΔphoP strains grown in FM-II. For each gene, the transcription level in the WT before temperature rise (0 min) was set to 1. Error bars (panels C and D) indicate SDs from three replicates.

PhoP bound to probe dnaK1p (Fig. 7B). EMSAs with GST-PhoP and other promoter probes of HSGs (Fig. 4B) revealed specific binding of PhoP to htpGp, hsp18_1p, and hsp18_2p but not to other tested probes (Fig. 8B). HspR and PhoP thus have exclusive target HSGs in addition to their common target HSGs (genes in the dnaK1-grpE1-dnaJ1-hspR operon); i.e., clpB1, clpB2, and lonA are targets of HspR but not of PhoP, whereas htpG, hsp18_1, and hsp18_2 are targets of PhoP but not of HspR.

Consistent with the ΔphoP phenotype observed in heat stress tests, transcription levels of PhoP target HSGs dnaK1, hspR, htpG, hsp18_1, and hsp18_2 were much higher in the ΔphoP strain than in the WT grown in FM-I (Fig. 8C), indicating negative regulation of these targets by PhoP. Transcription levels of these five genes in WT and ΔphoP strains were recorded under heat shock stress. In the WT, levels increased to maximum for dnaK1 (∼49-fold) and hspR (∼43-fold) within 10 min, for hsp18_1 (∼22-fold) and hsp18_2 (∼25-fold) within 20 min, and for htpG (∼25-fold) within 30 min (Fig. 8D), indicating important roles of DnaK1, Hsp18_1, Hsp18_2, and HtpG in HSR. In the ΔhspR strain, all five genes showed increased transcription levels (Fig. 8D), confirming repression of these genes by PhoP.

Prediction of HspR regulon and identification of HspR targets involved in development.

To further clarify the roles of HspR in S. avermitilis, we predicted its regulon. WebLogo (http://weblogo.berkeley.edu) analysis of the 19-nt HAIR sequences in the above-mentioned HspR-binding promoter regions (aveA1p, dnaK1p, clpB1p, clpB2p, lonAp, catR-katA1, katA3p, and oxyR-ahpCD) revealed a consensus sequence, 5′-TTGANBBNNHNNNDSTSHN-3′ (N is A/T/C/G, B is T/C/G, H is A/C/T, D is A/G/T, and S is C/G) (Fig. 9A). Scanning of the S. avermitilis genome with the 19-nt consensus HspR-binding sequence by PREDetector (39) identified 155 putative HspR target genes (cutoff; score ≥ 8) (see Table S1). Of these, 60 were unclassified or unknown, and the remaining 95 were assigned to 15 functional groups on the basis of the KEGG pathway database for S. avermitilis.

FIG 9.

Identification of HspR target genes associated with development. (A) Analysis of consensus HspR-binding sequence by WebLogo. Arrows, conserved 6-nt inverted repeats; asterisks, consensus bases. (B) EMSAs of His₆-HspR with promoter probes of four developmental genes. Notations as in Fig. 2D. (C) RT-qPCR analysis of wblB, ssgY, and ftsH in WT and ΔhspR strains grown on YMS agar. Error bars show SDs from three replicates. **, P < 0.01; ***, P < 0.001 (t test).

Our phenotypic observations revealed the negative role of HspR in S. avermitilis development. We therefore performed EMSAs on several predicted HspR target developmental genes listed in Table S1: whiB (sav_5042) for sporulation regulator WhiB (40), wblB (sav_4997) for putative WhiB-family TR (putative control of cell cycle), ssgY (sav_4267) for putative sporulation-specific cell division protein SsgY, and ftsH (sav_4666) for putative cell division protein FtsH. His₆-HspR bound to promoter probes wblBp, ssgYp, and ftsHp but not to probe whiBp (Fig. 9B). Transcription levels of wblB, ssgY, and ftsH were determined by RT-qPCR using RNAs prepared from WT and ΔhspR strains grown on YMS plates for 2 (aerial hypha growth stage), 4 (middle sporulation stage), or 6 (spore maturation stage) days. Levels of these three genes were higher in the ΔhspR strain than in the WT at three time points (Fig. 9C), consistent with the earlier differentiation phenotype of the ΔhspR strain. These findings indicate that HspR negatively regulates development by directly repressing wblB, ssgY, and ftsH.

DISCUSSION

HspR has been reported to function as an HSR repressor in S. coelicolor, S. albus, and S. lividans (13–17); however, its roles in antibiotic production, development, and other stress responses of Streptomyces had not been studied until now. Results of the present study of molecular mechanisms underlying HspR function in avermectin production, development, and heat shock and H₂O₂ stress responses in S. avermitilis demonstrate that HspR acts as a dual repressor/activator in these important physiological processes through control of corresponding target genes. HspR was shown to interact with another important regulator, PhoP, to coregulate the HSR. These findings bring to light previously unrecognized roles and regulatory mechanisms of HspR in this genus and provide a basis for the construction of antibiotic-overproducing strains in Streptomyces species.

S. avermitilis HspR directly activated avermectin production by interacting with the promoter region of structural genes aveA1 and aveA2 and strongly promoted avermectin production in both WT and industrial strains. HspR homologs are widely distributed among Streptomyces. The HspR-mediated activation of antibiotic production may occur in other Streptomyces species, and this possibility requires further investigations.

HspR was shown to play a negative regulatory role in S. avermitilis development, and three developmental genes (wblB, ssgY, and ftsH) were identified as HspR targets. wblB encodes a putative WhiB-family TR homologous to S. coelicolor WhiD, which is required for late-stage sporulation (41). ssgY encodes a putative cell division protein homologous to S. coelicolor SsgA, which is required for synthesis of sporulation septa (42). ftsH encodes a putative cell division protein homologous to E. coli FtsH. Enhanced expression of these three HspR target genes in the ΔhspR strain may account for the phenotype of this mutant. However, we cannot rule out possible contributions to the ΔhspR phenotype by other HspR target developmental genes; further studies are needed to resolve this point.

In regard to control of the HSR, S. avermitilis HspR acts as a direct repressor of several HSGs (the dnaK1-grpE1-dnaJ1-hspR operon, clpB1p, clpB2p, and lonAp), consistent with its reported role in HSRs in other Streptomyces species (13–17). The ability of HspR and chaperone DnaK to form a complex to corepress HspR target HSGs has been well established in S. coelicolor (14, 18). HspR and DnaK are both highly conserved in the genus, and it is therefore likely that HspR-DnaK interaction occurs in other species. The mechanism of such interaction was not investigated in the present study. Our findings reveal an additional layer of complex regulation of HSR, i.e., HspR cross talks with PhoP at dnaK1p to corepress the dnaK1-grpE1-dnaJ1-hspR operon. GST pulldown and competitive EMSAs revealed the existence of the HspR-PhoP complex and the reduction or disappearance of the HspR-PhoP-dnaK1p complex with increasing concentrations of one protein while concentration of the other protein remained constant, indicating that formation of the HspR-PhoP complex is a highly dynamic process. HspR and PhoP do not depend on each other for DNA binding, and they both have exclusive target HSGs (clpB1, clpB2, and lonA for HspR; htpG, hsp18_1, and hsp18_2 for PhoP). The HspR-PhoP complex may therefore play a regulatory role which neither protein by itself is capable; i.e., stabilization of DNA-protein structure to precisely regulate essential genes. Consistent with this possibility, the essential pathogenic determinant gene acr2 in M. tuberculosis is regulated by the HspR-PhoP complex (20). Notwithstanding their cooperative activity, HspR and PhoP compete for binding to dnaK1p, i.e., an increasing concentration of one protein suppresses formation of the complex with dnaK1p by the other protein, reflecting the delicate interplay between these two regulators in the control of the dnaK1-grpE1-dnaJ1-hspR operon. Our findings reveal a novel role of PhoP in the HSR and further demonstrate the importance of PhoP in Streptomyces. Coregulation of the dnaK1-grpE1-dnaJ1-hspR operon by HspR and PhoP suggests that HSGs within this operon play a dominant role over other HSGs in response to heat shock stress.

HspR has a negative regulatory role in the HSR, but we also observed a positive role in the H₂O₂ stress response, i.e., it directly activates the transcription of related genes. HspR target genes (katA1, catR, katA3, oxyR, ahpC, and ahpD) are well conserved in Streptomyces; such an HspR-based mechanism involved in the control of the H₂O₂ stress response is therefore likely to be present in other species. HspR also functions as an activator (it activates transcription of ave structural genes and H₂O₂ stress-related genes), but the mechanism underlying this function remains to be elucidated.

Our findings, taken together, provide a basis for a proposed model of the HspR-mediated regulatory network involved in development, avermectin production, and heat shock and H₂O₂ stress responses in S. avermitilis (Fig. 10). These important physiological processes are coordinated by HspR through its positive or negative effect on target genes. HspR also interacts with PhoP to corepress the dnaK1-grpE1-dnaJ1-hspR operon. Predicted HspR targets are involved in other essential functions such as primary metabolism (see Table S1 in the supplemental material). It is clear that the roles of HspR in Streptomyces are much broader than previously recognized. Identification of additional targets and related molecular processes will require extensive further study.

FIG 10.

Proposed model of HspR-mediated regulatory network in S. avermitilis. Solid arrows, direct activation; bars, direct repression; hollow arrows, avermectin biosynthesis, development, or response to stress.

MATERIALS AND METHODS

Strains, plasmids, primers, and growth conditions.

Strains and plasmids used in this study are listed in Table 1, and primers are listed in Table 2. Culture conditions for S. avermitilis and E. coli were described previously (43). YMS (44) agar was used for observation of the S. avermitilis phenotype. Insoluble FM-I fermentation medium and soluble FM-II fermentation medium (45) were used for routine avermectin production and growth analysis, respectively. FM-II was also used to grow mycelia for ChIP-qPCR assays and for RNA isolation following stress treatment.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Description	Source or reference
Strains
S. avermitilis
ATCC 31267	Wild-type (WT) strain	Laboratory stock
A229	Industrial strain	Qilu Pharmaceutical
ΔhspR	hspR deletion mutant	This study
ChspR	hspR complemented strain	This study
OhspR	hspR overexpression strain	This study
ΔphoP	phoP deletion mutant	38
ΔhspR/hspR-3FLAG	hspR complemented strain with HspR-3FLAG	This study
ΔphoP/hspR-3FLAG	phoP deletion mutant with HspR-3FLAG	This study
WT/hspR-3FLAG	WT strain with HspR-3FLAG	This study
WT/pKC1139	WT strain carrying empty vector pKC1139	This study
WT/pSET152	WT strain carrying empty vector pSET152	This study
OhspR/A229	hspR overexpression strain based on A229	This study
E. coli
JM109	General cloning host for plasmid manipulation	Laboratory stock
BL21(DE3)	Host for protein expression	Novagen
Plasmids
pKC1139	Multiple-copy temperature-sensitive E. coli-Streptomyces shuttle vector	46
pSET152	Integrative E. coli-Streptomyces shuttle vector	46
pET-28a(+)	Vector for His6-tagged protein expression in E. coli	Novagen
pGEX-4T-1	Vector for GST-tagged protein expression in E. coli	GE Healthcare
pCIMt005	Multiple-copy temperature-sensitive E. coli-Streptomyces shuttle vector	47
pJL117	pIJ2925 derivative carrying the Streptomyces strong constitutive promoter ermE*p	49
pKCΔhspR	hspR deletion vector based on pKC1139	This study
pΔhspR	hspR deletion vector based on pCIMt005	This study
pSET152-hspR	hspR complemented vector based on pSET152	This study
pKC-erm-hspR	hspR overexpression vector based on pKC1139	This study
pSET152-hspR-	hspR complemented vector with 3×FLAG-tagged	This study
3FLAG	HspR on pSET152
pIJ10500	Vector carrying 3×FLAG fragment	50
pET28-hspR	His6-HspR expression vector based on pET-28a(+)	This study
pGEX-phoP	GST-PhoP expression vector based on pGEX-4T-1	38

TABLE 2.

Primers used in this study

Primer	DNA sequencea (5′→3′)	Use
Gene disruption, complementation,  and overexpression
WQ11	CGGGATCCGATCGCCGAGAACCCCTG (BamHI)	Deletion of hspR gene
WQ12	GCTCTAGAGTACTGACGCAGGGTCTG (XbaI)
WQ13	GCTCTAGAACCAGGAGGTCCAGCAGA (XbaI)
WQ14	CCCAAGCTTAACGAGATGGCAACAGCA (HindIII)
LXR101A	GTCTCGTACGAAGAGCTTTTATAAAAGCTTGATCGCCGAGAACCCCTG
LXR101B	AAAATCCCTTAACGTGAGCCTAGGGCGTGCAACGAGATGGCAACAGCA
WQ15	CCTGCAAGGCCTGTTCGG	Confirmation of hspR deletion in ΔhspR strain
WQ16	AGGGCTATGACGGTACGA
WQ17	ACCGTCTCGGCCTGGTCT
WQ18	CTCGCGCTGCTGCATCAT
LXR133A	CCGGAATTCGATGGCGTCCTCGTGCTC (EcoRI)	Complementation of hspR in ΔhspR strain
LXR133B	CGGGATCCTTCAATTCCTCCAGCAA (BamHI)
LXR134A	CGGGATCCGAGGAGATGGACGGCCGT (BamHI)
LXR134B	CCCAAGCTTCGAAGCCGTCAGTCCGAG (HindIII)
WQ25	CGGGATCCCCGCGAAGGGAGCATGAGG (BamHI)	Overexpression of hspR in Streptomyces
WQ26	CCCAAGCTTCTGGACAGTCTCGGTGCCG (HindIII)
WQ27	CCGGAATTCCATGAGGAGATGGACGGC (EcoRI)	Expression of His6-HspR protein in E. coli
WQ28	CCCAAGCTTACATGCTGGACAGTCTCG (HindIII)
LXR60A	CCCAAGCTTGGAGGTGGCATGGACTAC (HindIII)	Amplification of 3×FLAG fragment
LXR60B	CGGGATCCTCCGGTTGACCCTTATTT (BamHI)
LXR136A	GCTCTAGAGATGGCGTCCTCGTGCTC (XbaI)	Complementation of hspR in ΔhspR strain with 3×FLAG-tagged HspR
LXR136B	TTCAATTCCTCCAGCAA
LXR135A	TCATTGCTGGAGGAATTGAAGAGGAGATGGACGGCCGT
LXR135B	CCCAAGCTTGTCCGAGGACTGGCGCTT (HindIII)
EMSA
aveA1p-Fw	ATGGTCGGGAACCTCCGCAA	Probe aveA1p
aveA1p-Rev	CTGTGTCCTCACCGCTAGGC
LXR59A	ATCTAGTCGTTTTTGAGT	Probe hrdBp
LXR59B	ACGAACAACCTCTCGGAA
aveA4p-Fw	CGACAAGAGAAATCGGAAATT	Probe aveA4p
aveA4p-Rev	GCCTGCACCTGTGACAAG
aveRp-Fw	CCGCACCGCCATACATAC	Probe aveRp
aveRp-Rev	GAAACTCCCTGCATGATGTTC
LXR102A	GGGCATCAATGTCTTGAACGTCTTCAACTCTTCAAAGCTAATGACGGCGC	Probe aveA1p-1
LXR102B	GCGCCGTCATTAGCTTTGAAGAGTTGAAGACGTTCAAGACATTGATGCCC
LXR103A	GGGCATCAATGTCAAGCTTGTCTTCAGGATCCCAAAGCTAATGACGGCGC	Probe aveA1p-1m
LXR103B	GCGCCGTCATTAGCTTTGGGATCCTGAAGACAAGCTTGACATTGATGCCC
WQ39	AGGAGCGGATAAGACTTGAGT	Probe dnaK1p
WQ40	TTAGTCGTGCCCAGGTCG
LXR104A	TAAGGAGCGGATAAGACTTGAGTCGGCTCGACTCAGGTCTGTTGACCCAG	Probe dnaK1p-1
LXR104B	CTGGGTCAACAGACCTGAGTCGAGCCGACTCAAGTCTTATCCGCTCCTTA
LXR105A	TAAGGAGCGGATAAGACAAGCTTCGGCTCGGGATCCGTCTGTTGACCCAG	Probe dnaK1p-1m
LXR105B	CTGGGTCAACAGACGGATCCCGAGCCGAAGCTTGTCTTATCCGCTCCTTA
LXR106A	TTGGGTGATGCACACTTGAGTCCGTTCCACTCAAGTCATTGCTGGAGGAA	Probe dnaK1p-2
LXR106B	TTCCTCCAGCAATGACTTGAGTGGAACGGACTCAAGTGTGCATCACCCAA
LXR107A	TTGGGTGATGCACACAAGCTTCCGTTCCGGATCCGTCATTGCTGGAGGAA	Probe dnaK1p-2m
LXR107B	TTCCTCCAGCAATGACGGATCCGGAACGGAAGCTTGTGTGCATCACCCAA
LXR108A	CATGGGGCCGAACCTAGC	Probe clpB1p
LXR108B	TCGCTGTCTCCTCCTGTG
LXR109A	AAGGAAGAGGCGGGCGTG	Probe clpB2p
LXR109B	CGACCGGATCCGACGCG
LXR110A	GAAGCACCTGGTTCGCCT	Probe lonAp
LXR110B	AAGCCATGATCTCCCCTT
SD335A	CGTTCTCCTCGAGCCTACAGG	Probe dnaK2p
SD335B	CTCCCACACGGCGATCAC
SD339A	ATTGGCACTCCGCTTGAC	Probe groES1p
SD339B	GGCTTGATGGCAACCTTG
SD337A	TTCAGATGGACGGCTACC	Probe groEL2p
SD337B	AGAGTGCTAACGCCAATGA
SD340A	GGAAAGCCGCTGGTCAGAC	Probe htpGp
SD340B	GGGAGTCCATCGTCGCAG
SD341A	ATCGTCCCTGGTCATCGG	Probe hsp18_1p
SD341B	ATCAACATCGTGAAACACCTCC
SD342A	AGTGCTGTCGTCGCTCAT	Probe hsp18_2p
SD342B	CATCAACACGGTAAACACCTC
SD123A	GCGCAGTCGTTCCAGTAG	Probe catR_katA1
SD123B	GGACTCGCTGTTCTGATTGT
SD124A	GGAAAGGTGGTCTGGTTC	Probe katA2p
SD124B	GGTGTACGGAGCCTTCTG
SD122A	AGGTCCGTGGAGCGCCTC	Probe katA3p
SD122B	CGGTGGCTGAGTTCTGGTTGTC
SD236A	GAGAAGCCGAGCACCTGG	Probe oxyR_ahpCD
SD236B	CGCTGGAGAAGGGCAATG
SD210A	CCGATGGCTGTAGGCGTTTAT	Probe whiBp
SD210B	CCCTCGTCTGTCTTCGCG
LXR111A	GAGCGCGTTTTCTCCGAG	Probe ssgYp
LXR111B	ATGCAGCCCCTGAATTTC
LXR112A	ATCTTCGGCACTCTACGT	Probe wblBp
LXR112B	AAATCTGCCATTACGTGT
SD115A	GCCCGGTTCTCGTAAGAC	Probe ftsHp
SD115B	CAGCACGATCCACATGAC
DNase I footprinting
FAM-WQ45	CCCAGCGCACGATTCATGA	aveA1 promoter region
WQ46	CCGTCCATCCTCTGCACCTG
FAM-WQ43	GTCTCGGTCCTCGGGCTG	dnaK1 promoter region
WQ44	TGCTGCGGGTTGAAGTCC
RT-qPCR
16S-QP-Fw	AGCGGAGCATGTGGCTTAAT	16S rRNA ORF
16S-QP-Rev	ACGTATTCACCGCAGCAATG
GJ99	CGGACAGGACTACGCACTTC	aveA1 ORF
GJ100	ACGAGATACGACCGGAGATC
LXR32A	ATCTCGTCAAGTCCCAGA	aveA2 ORF
LXR32B	GTGTAACTGGTTCCTGAG
SD346A	CCAAGAACGGTGAGGTGC	dnaK1 ORF
SD346B	CTGCTGCGGGTTGAAGTC
LXR113A	TCGGAACCCCTATGAACT	hspR ORF
LXR113B	GTACTGACGCAGGGTCTG
LXR114A	CCTCAAGAACAAGCGGCT	clpB1 ORF
LXR114B	CGGCGAGCACGGTCTTCA
LXR115A	ATGAACCGTCTCACCCAG	clpB2 ORF
LXR115B	TCCTCCTGATCGAGAAGT
LXR116A	AAGGGGAGATCATGGCTT	lonA ORF
LXR116B	GGCGTCGTTCAGGTCCAG
katA1-QP-Fw	ACCTCGTCGGCAACAACAC	katA1 ORF
katA1-QP-Rev	TGTACGGGTCGCGCTTCTG
catR-QP-Fw	TGAAGCTGCCCGAGATCTCC	catR ORF
catR-QP-Rev	CTTGTCCGTGGCGACCTCC
katA3-QP-Fw	CATCCACTCCCAGAAGCGC	katA3 ORF
katA3-QP-Rev	TCGCCCATCAGCCAGGTC
oxyR-QP-Fw	GATGGAGGAGGCCGAGGC	oxyR ORF
oxyR-QP-Rev	GCAGGAGGTTGAGCTCACG
ahpC-QP-Fw	CCAGGTGCTCGGCTTCTCC	ahpC ORF
ahpC-QP-Rev	GAAGCACGAGCTGATGCGC
ahpD-QP-Fw	CTCTCGATGAACTGAAGG	ahpD ORF
ahpD-QP-Rev	TGCGGCAGGTCGGAGTTG
LXR117A	TCACAGTCATCCCAGGCAT	htpG ORF
LXR117B	CTCGCGCAGATAGACCTT
LXR118A	CGATGCCAATGGACGCCTA	hsp18_1 ORF
LXR118B	TCAGCATGTTCCGTTCGA
LXR119A	ATCTGATGGGTCCGGGAA	hsp18_2 ORF
LXR119B	AAGGCGATGACGTACTGG
LXR120A	CAGCTCGCTCTTCTTCCA	wblB ORF
LXR120B	TCATGCAGACCTCTTTGG
LXR121A	AATGGATCATCGGACGAGA	ssgY ORF
LXR121B	TGTACGAGTCGGAGCGGTCAG
SD247A	CGGCTACAAGACAGTGGACA	ftsH ORF
SD247B	GACCTTGATGACCTGCTCGT
LXR122A	GCACCGTCCAGCAGTTGT	hspR deletion region
LXR122B	ACTCCAGTTCCACCACGC
ChIP-qPCR
LXR123A	AGGGGTGGATAGGGGTATT	aveA1 promoter DNA
LXR123B	AAGAATGAAAGGAGCGCG
LXR69A	CACTCGTGAATTGTGGAC	aveR promoter DNA
LXR69B	TGACTTTGATGATTAACT
LXR124A	CGTGTAAGGAGCGGATAA	dnaK1 promoter DNA
LXR124B	TTCAATTCCTCCAGCAAT
LXR125A	TTCCCGGCACCCCCGAA	clpB1 promoter DNA
LXR125B	TCTGACATTGGAAGCGTA
LXR126A	ACGAGCGGCGGTGGCGGT	clpB2 promoter DNA
LXR126B	CGACCGGATCCGACGCG
LXR127A	ATTTCCGATGATCGCTGA	lonA promoter DNA
LXR127B	AAGCCATGATCTCCCCTT
LXR128A	TGTGACTTTGGAGCTTTA	catR_katA1 promoter DNA
LXR128B	TCCTGCGTCATTCCTGCAA
LXR129A	GGAACCAACTTCAACAAA	katA3 promoter DNA
LXR129B	TTCGACATCGTGCACCTT
LXR130A	AAGATGACCTTCCACTGA	oxyR_ahpC promoter DNA
LXR130B	TGACAAGTTCCCCGAGTT
LXR131A	TTCGTCACCCTGGGGGTG	dnaK2 promoter DNA
LXR131B	ACTCTCCGGATCACGTTC
LXR132A	CCCAGAACCGAACCGAAG	katA2 promoter DNA
LXR132B	AAAGAAGAGGGTCGTTAG

^aUnderlining indicates sequences for the restriction enzymes shown in parentheses.

Construction of S. avermitilis mutant strains.

To construct the hspR deletion mutant, a 594-bp 5′ flanking region (positions −527 to +67 relative to the hspR translational start codon [TSC]) and a 454-bp 3′ flanking region (positions +385 to +838) were generated by PCR from S. avermitilis WT genome with primers WQ11/WQ12 and WQ13/WQ14. The two fragments were digested with BamHI/XbaI and XbaI/HindIII, respectively, and then ligated into BamHI/HindIII-digested pKC1139 (46) to generate pKCΔhspR. Temperature-sensitive plasmid pCIMt005, containing the idgS gene that encodes indigoidine synthetase to make colonies blue, was used for hspR deletion by simple blue-white screening (47). A 1,048-bp DNA fragment containing 5′- and 3′-flanking regions of hspR was amplified from pKCΔhspR with primers LXR101A and LXR101B and ligated into NcoI-digested pCIMt005 to generate hspR deletion vector pΔhspR, which was transformed into WT protoplasts. The hspR deletion mutant was screened as described previously (48) and confirmed by colony PCR and DNA sequencing. Use of primers WQ15 and WQ16 (flanking the exchange regions) (see Fig. S1 in the supplemental material) resulted in the appearance of a 1,337-bp band, whereas a 1,655-bp band was observed in the WT. When primers WQ17 and WQ18 (located within the deletion region of hspR) were used, only the WT produced a 385-bp band. We thus generated the hspR deletion mutant (ΔhspR strain), in which a 275-bp fragment within hspR open reading frame (ORF) (positions +112 to +386 relative to the TSC) was deleted (see Fig. S1).

For complementation of ΔhspR, a 465-bp fragment containing the hspR ORF and a 446-bp fragment containing the dnaK1 promoter were amplified with primers LXR134A/LXR134B and LXR133A/LXR133B, respectively. The two PCR fragment were digested with BamHI/HindIII and EcoRI/BamHI, respectively, and ligated simultaneously into pSET152 (46) to generate hspR-complemented vector pSET152-hspR, which was then introduced into the ΔhspR strain to generate complemented strain ChspR.

For overexpression of hspR, a 586-bp fragment carrying the hspR ORF was amplified with primers WQ25 and WQ26 and ligated simultaneously with a 188-bp ermE*p (Streptomyces strong constitutive promoter) fragment from pJL117 (49) into pKC1139 to generate hspR-overexpressing vector pKC-erm-hspR, which was introduced into S. avermitilis WT and industrial strain A229 to generate hspR overexpression strains OhspR and OhspR/A229, respectively.

To express 3×FLAG-tagged HspR in S. avermitilis, the hspR ORF and dnaK1 promoter were amplified with primers LXR135A/LXR135B and LXR136A/LXR136B from WT genomic DNA, and the 3×FLAG fragment was amplified with primers LXR60A and LXR60B from plasmid pIJ10500 (50). The resulting 453-bp hspR, 446-bp dnaK1 promoter, and 100-bp FLAG fragments were cloned into pSET152 to generate pSET152-hspR-3FLAG, which was then transformed into WT, ΔhspR, and ΔphoP strains to generate recombinant strains WT/hspR-3FLAG, ΔhspR/hspR-3FLAG, and ΔphoP/hspR-3FLAG strains, respectively, for expression of C-terminally 3×FLAG-tagged HspR.

Scanning electron microscopy.

Mycelia and spores of S. avermitilis WT and ΔhspR grown on YMS plates for 2 or 4 days were observed by SEM. Samples were prepared and examined as described in our previous study (51).

Production and analysis of avermectins.

Avermectin yield from S. avermitilis fermentation culture was analyzed by HPLC as described previously (52).

Heterologous expression and purification of His₆-HspR.

For expression of S. avermitilis HspR in E. coli, the 579-bp hspR ORF was generated by PCR using primers WQ27 and WQ28. The PCR fragment was digested with EcoRI/XhoI and ligated into pET-28a(+), generating pET28-hspR for overexpression of N-terminal His₆-tagged HspR recombinant protein. Expression vector pET28-hspR was confirmed by sequencing and transformed into E. coli BL21(DE3). His₆-HspR expression was induced by 3-h treatment with 0.2 mM isopropyl-β-d-thiogalactopyranoside (IPTG) at 37°C. Cell lysates were prepared by sonication on ice in lysis buffer (53) and centrifugation. Soluble His₆-HspR was purified on an Ni-nitrilotriacetic acid (NTA) column (Bio-Works; Sweden) and eluted by lysis buffer with 250 mM imidazole. The purified protein was dialyzed in binding buffer for EMSAs (54) to eliminate imidazole and stored at −80°C until use.

Electrophoretic mobility shift assays.

Promoter probes of tested genes were amplified by PCR with corresponding primers listed in Table 2, and labeled at the 3′ end with nonradioactive digoxigenin-1-ddUTP. EMSAs were performed as described previously (52). Specificity of the HspR-probe interaction was tested by adding ∼300-fold excess of unlabeled nonspecific probe (hrdBp) or respective specific probe to the binding reaction system.

Reverse transcription and real-time quantitative PCR analysis.

Total RNAs were extracted from S. avermitilis cultures grown in FM-I or FM-II or on YMS plates for various durations. RNA isolation, removal of genomic DNA, and subsequent RT-qPCR analysis (using primers listed in Table 2) were performed as described previously (52). Relative expression levels of tested genes were normalized with 16S rRNA (housekeeping gene) as the internal control. Experiments were performed in triplicates.

Western blotting.

Total protein was prepared at various time points from ΔhspR/hspR-3FLAG mycelia grown in FM-I. Western blotting was performed as described in our previous study (54), using 1:3,300-diluted mouse anti-FLAG monoclonal antibody (MAb) (Sigma; USA).

ChIP-qPCR.

To prepare samples for ChIP-qPCR, S. avermitilis WT, WT/hspR-3FLAG, ΔhspR/hspR-3FLAG, and ΔphoP/hspR-3FLAG strains were cultured in FM-II, and mycelia were harvested at various time points and treated as described previously (29). Immunoprecipitation reactions and ChIP-qPCR data analysis were performed as described previously (32, 48). Relative protein enrichment of each site was normalized based on input chromosomal DNA, and binding level was shown as a percentage of input DNA. Experiments were performed in triplicates.

DNase I footprinting.

To determine HspR binding sites, 5′ 6-carboxyfluorescein (FAM)-labeled DNA fragments corresponding to aveA1 and dnaK1 promoter regions were PCR-synthesized using primers listed in Table 2 and then purified. DNase I footprinting assays were performed as described previously (51, 55), and data were analyzed using GeneMarker software program v2.2.0.

GST pulldown.

pET28-hspR (for His₆-HspR expression) and pGEX-PhoP (for GST-PhoP expression) (38) were cotransformed into E. coli BL21(DE3). Bacteria containing pET28-hspR and pGEX-4T-1 (for GST tag expression) were used as negative controls. Following 0.4 mM IPTG induction for 3 h at 37°C, cells containing both His₆- and GST-tagged proteins were sonicated in lysis buffer (53) on ice and centrifuged. One milliliter cell lysate and 50 μl 50% glutathione-Sepharose beads (equilibrated with lysis buffer) were incubated overnight at 4°C. Beads were washed four times with phosphate-buffered saline (PBS) (56), and the resulting pellet was boiled with 50 μl protein loading buffer. Eluted bound proteins were separated by electrophoresis on SDS-PAGE gels and identified by Western blotting using anti-His₆ or anti-GST antibody (Tiangen; China).