A Novel AdpA Homologue Negatively Regulates Morphological Differentiation in Streptomyces xiamenensis 318

AdpA is a key regulator of secondary metabolism and morphological differentiation in Streptomyces species. However, AdpA had not been reported to negatively regulate morphological differentiation. Here, we characterized the regulatory role of AdpA_Sx in Streptomyces xiamenensis 318, which has a naturally streamlined genome. In this strain, AdpA_Sx negatively regulated cell growth and morphological differentiation by directly controlling genes associated with these functions. AdpA_Sx also bidirectionally controlled the biosynthesis of xiamenmycin and PTMs by directly regulating their gene clusters rather than through other regulators. Our findings provide additional evidence for the versatility of AdpA in regulating morphological differentiation and secondary metabolism in Streptomyces.

ABSTRACT

The pleiotropic transcriptional regulator AdpA positively controls morphological differentiation and regulates secondary metabolism in most Streptomyces species. Streptomyces xiamenensis 318 has a linear chromosome 5.96 Mb in size. How AdpA affects secondary metabolism and morphological differentiation in such a naturally minimized genomic background is unknown. Here, we demonstrated that AdpA_Sx, an AdpA orthologue in S. xiamenensis, negatively regulates cell growth and sporulation and bidirectionally regulates the biosynthesis of xiamenmycin and polycyclic tetramate macrolactams (PTMs) in S. xiamenensis 318. Overexpression of the adpA_Sx gene in S. xiamenensis 318 had negative effects on morphological differentiation and resulted in reduced transcription of putative ssgA, ftsZ, ftsH, amfC, whiB, wblA1, wblA2, wblE, and a gene encoding sporulation-associated protein (sxim_29740), whereas the transcription of putative bldD and bldA genes was upregulated. Overexpression of adpA_Sx led to significantly enhanced production of xiamenmycin but had detrimental effects on the production of PTMs. As expected, the transcriptional level of the xim gene cluster was upregulated, whereas the PTM gene cluster was downregulated. Moreover, AdpA_Sx negatively regulated the transcription of its own gene. Electrophoretic mobility shift assays revealed that AdpA_Sx can bind the promoter regions of structural genes of both the xim and PTM gene clusters as well as to the promoter regions of genes potentially involved in the cell growth and differentiation of S. xiamenensis 318. We report that an AdpA homologue has negative effects on morphological differentiation in S. xiamenensis 318, a finding confirmed when AdpA_Sx was introduced into the heterologous host Streptomyces lividans TK24.

IMPORTANCE AdpA is a key regulator of secondary metabolism and morphological differentiation in Streptomyces species. However, AdpA had not been reported to negatively regulate morphological differentiation. Here, we characterized the regulatory role of AdpA_Sx in Streptomyces xiamenensis 318, which has a naturally streamlined genome. In this strain, AdpA_Sx negatively regulated cell growth and morphological differentiation by directly controlling genes associated with these functions. AdpA_Sx also bidirectionally controlled the biosynthesis of xiamenmycin and PTMs by directly regulating their gene clusters rather than through other regulators. Our findings provide additional evidence for the versatility of AdpA in regulating morphological differentiation and secondary metabolism in Streptomyces.

INTRODUCTION

Streptomyces species are well known for two extraordinary traits, the ability to produce a wide variety of valuable secondary metabolites with diverse biological activities and a complex developmental life cycle that includes the formation of vegetative mycelium, aerial mycelium, and spores at different growth stages during development (1–3). The processes of morphological differentiation and antibiotic biosynthesis are tightly controlled via multiple levels of regulators, including cluster-situated, pleiotropic, and global regulators that respond to numerous physiological and environmental conditions (4–6).

AdpA, the most-studied AraC family pleiotropic transcriptional regulator in Streptomyces, had been reported to play a central role in regulation of secondary metabolism and morphological differentiation in most Streptomyces species. It was first shown to influence streptomycin production in Streptomyces griseus (7). In the A-factor regulatory cascade, the expression of adpA in S. griseus is dependent on the small signaling molecule named A-factor, which contains a γ-butyrolactone ring. This A-factor can bind to the repressor protein ArpA, resulting in dissociation of ArpA from the promoter region of adpA and activation of adpA (8). The induced AdpA then activates the transcription of various genes related to secondary metabolism and morphological differentiation.

AdpA is required for morphological differentiation in most Streptomyces species (9–13). In S. griseus, it functions as an activator to upregulate the transcription of direct target genes involved in spore formation, such as ssgA, which encodes a protein that strongly influences septum formation, σ^AdsA, an ECF sigma factor that is required for aerial mycelium formation, and the response regulator that controls expression of the adjacent SapB biosynthetic operon, amfTSBA (1, 3). In addition, AdpA can regulate chromosome replication by binding to a region close to OriC (14). However, AdpA has never been reported to negatively regulate morphological differentiation.

AdpA also activates the transcription of strR, the pathway-specific regulatory gene for streptomycin. Other examples of AdpA positively regulating antibiotic biosynthesis through the control of cluster-situated regulators (CSRs) have been found in Streptomyces clavuligerus (15), Streptomyces avermitilis (16), Streptomyces chattanoogensis (12, 17), and Streptomyces roseosporus (10). Since the goal of “one strain—many compounds” is attractive when performing genome mining in Streptomyces (18), the different effects of AdpA on the production of multiple secondary products in a single Streptomyces host have been studied with S. ansochromogenes. In S. ansochromogenes, AdpA-L can activate nikkomycin biosynthesis via the pathway-specific regulator SnG but represses oviedomycin biosynthesis via the CSRs OvmW and OvmZ (13, 19).

Streptomyces xiamenensis 318 was isolated from a mangrove sediment and represented a novel species of Streptomyces genus (20, 21). Whole-genome sequencing shown that S. xiamenensis 318 has a linear chromosome 5.96 Mb in size, which is considerably smaller than other reported complete genomes of the genus Streptomyces (21). The streamlined genome in S. xiamenensis 318 was characterized by fewer rRNA operons (rrn) and fewer genes in most gene functional categories as observed by comparing to the other five representative Streptomyces genome sequences (21). Based on the bioinformatics analysis, a total of 21 gene clusters in the genome of S. xiamenensis 318 were predicted to be involved in secondary metabolism. At the same time, S. xiamenensis 318 produces two types of molecules with different skeletons, xiamenmycin and the polycyclic tetramate macrolactams (PTMs) (21). Xiamenmycin is a prenylated benzopyran compound with anti-inflammatory and antifibrotic activity (22, 23), and PTMs are well known for their multiple biological activities (24). The biosynthesis of xiamenmycin and PTMs does not involve any CSRs. Whether the pleiotropic regulator AdpA_Sx, the S. xiamenensis 318 homologue of AdpA, can directly regulate the expression of the respective gene clusters to control the biosynthesis of xiamenmycin and PTMs is of great interest.

In this work, we characterized a novel role for AdpA_Sx in controlling the morphological differentiation of S. xiamenensis 318. Furthermore, we determined that AdpA_Sx positively controls the biosynthesis of xiamenmycin and negatively regulates PTMs by directly regulating the expression of their gene clusters rather than through other regulators.

RESULTS

Overexpression of adpA_Sx impairs cell growth and morphological differentiation.

To explore the role of the AdpA orthologue AdpA_Sx (sxim_38050) in S. xiamenensis 318, we constructed the mutant strain ΔadpA, which has an in-frame deletion of adpA_Sx, and strain OEadpA, which overexpresses adpA_Sx. PCR analysis was used to verify the deletion of adpA_Sx from the S. xiamenensis 318 chromosome or the integration of adpA_Sx under the control of ermE*p into the S. xiamenensis 318 chromosome for ΔadpA and OEadpA strains, respectively (see Fig. S1a and b in the supplemental material). Quantitative real-time PCR (qRT-PCR) results revealed that the transcription level of adpA_Sx in the OEadpA strain increased 362% by 24 h and 575% by 48 h compared to levels in its parental strain, S. xiamenensis 318 (Fig. S1c), indicating that adpA_Sx was successfully overexpressed under the control of the strong promoter ermE*p.

Disruption of adpA results in a “bald” phenotype in many Streptomyces species (10, 11, 13). To define the role of AdpA_Sx in morphological differentiation, we inoculated the wild-type, ΔadpA, and OEadpA strains on mannitol-soybean powder (MS) agar medium using mycelium, as our preliminary observations indicated that the OEadpA strain had poor sporulation. Further investigation revealed that the OEadpA strain had a markedly sporulation-defective phenotype and that the ΔadpA strain showed slightly earlier formation of aerial hyphae and spores than the wild-type strain (Fig. 1a). Scanning electron microscopy (SEM) was used to analyze the effect of adpA_Sx overexpression in greater detail. The wild-type strain and the ΔadpA mutant strain had abundant spore chains, whereas the OEadpA strain had vegetative mycelium without sporulation septa (Fig. 1b). These findings suggest that AdpA_Sx negatively affects the formation of aerial mycelium and sporulation.

FIG 1.

Effects of AdpA_Sx on cell growth and morphological differentiation. (a) Phenotype of S. xiamenensis 318 (WT), OEadpA, and ΔadpA strains grown on MS agar. (b) Scanning electron micrographs of WT, OEadpA, and ΔadpA strains after growth on MS agar for 7 days. (c) Phenotypes of S. lividans TK24, S. lividans TK24-AdpA_Sx, S. xiamenensis 318 (WT), and S. xiamenensis 318-AdpA_Sl (WT-AdpA_Sl) grown on MS agar. (d) The growth curves of WT, OEadpA, and ΔadpA strains grown on TSB medium and determined using intracellular protein content.

To further investigate the effects of AdpA_Sx on morphological differentiation of Streptomyces, we introduced AdpA_Sx, driven by the strong promoter ermE*p, into the heterologous host Streptomyces lividans TK24, generating strain S. lividans TK24-AdpA_Sx. S. lividans TK24-AdpA_Sx exhibited delayed formation of aerial mycelium on MS medium in the early growth stages, although it produced abundant spores in the late stages (Fig. 1c; see also Fig S2). As a control, AdpA_Sl, from S. lividans TK24, driven by the strong promoter ermE*p, was introduced into S. xiamenensis 318. The derivative strain S. xiamenensis 318-AdpA_Sl showed no significant difference in sporulation on MS medium plates in comparison with that of the parent strain (Fig. 1c). These results suggest that AdpA_Sx has a negative role in the normal development of Streptomyces.

For S. xiamenensis, in comparison with the wild-type strain, cell growth of the OEadpA strain was delayed significantly (see Fig. S3), whereas the ΔadpA strain showed no significant difference in growth on tryptic soybean broth (TSB) medium. Growth was also measured by determining the total intracellular protein amount of the strains following cell disruption by sonication. Compared to that of the wild-type strain, the growth of the OEadpA strain was delayed for 48 h, whereas growth of the ΔadpA strain was slightly enhanced (Fig. 1d). These findings indicate that AdpA_Sx negatively controls the cell growth of S. xiamenensis 318.

Phylogenetic analysis of AdpA_Sx from S. xiamenensis 318.

To elucidate the negative effect of AdpA_Sx on morphological differentiation in S. xiamenensis 318, a multiple sequence alignment of AdpA homologues was performed. The result revealed that the amino acid sequence of AdpA_Sx was highly similar to that of the well-studied AdpA (approximately 73%) except SVA742 (16%), which regulates secondary metabolite biosynthesis and morphological differentiation in S. avermitilis (25). Two conserved domains were detected in the alignment, a type 1 glutamine amidotransferase (GATase1) domain in the N terminus and a helix-turn-helix (HTH) DNA-binding domain, of the AraC family regulators, in the C terminus (Fig. 2a). Additionally, some amino acids were conserved in the other five AdpA proteins but not in AdpA_Sx, and AdpA_Sx also contained a linker sequence not present in the other proteins (Fig. 2a). In S. griseus and Streptomyces coelicolor, adpA contains the codon TTA, the main target of bldA, which encodes the only tRNA for the rare leucine codon UUA and which thereby regulates morphological differentiation (26, 27). Leucine²⁴⁰ of AdpA_Sx is also encoded by TTA (Fig. 2a), indicating that the translation level of AdpA_Sx could be affected by bldA (9).

FIG 2.

Amino acid alignment and phylogenetic analysis of AdpA_Sx. (a) Amino acid alignment of AdpA_Sx with well-characterized AdpA proteins from five different Streptomyces species. The type 1 glutamine amidotransferase (GATase1) domain in the N terminus and the helix-turn-helix (HTH) DNA-binding domain of the AraC family of regulators in the C terminus are marked. *, amino acids conserved in the other five AdpA proteins but mutated in AdpA_Sx. A linker in AdpA_Sx is marked. ▲, location of the Leu residue encoded by the UUA codon; AdpA_Sg, AdpA in S. griseus; AdpA_Sch, AdpA in S. chattanoogensis; AdpA_Sc, AdpA in S. coelicolor A3(2); AdpA-L, AdpA in S. ansochromogenes; AdpA_Sl, AdpA in S. lividans TK24; AdpA_Sx, AdpA in S. xiamenensis 318. (b) Phylogenetic analysis of the novel AdpA_Sx with NCBI BLASTP hits and known AdpA homologues. AdpA_Sx (red star) was phylogenetically distinct from the clade containing well-characterized AdpA proteins. For protein accession numbers, see Table S1 in the supplemental material.

A phylogenetic tree was constructed to show the relationship between AdpA_Sx and well-studied AdpA proteins (Fig. 2b). The phylogenetic tree was divided into three clades, and AdpA_Sx was found in a separate clade (clade II) rather than with the well-studied AdpA proteins in clade I or with other homologs of AdpA (clade III). In clade II, the subclade II-b containing AdpA_Sx and its homologues was phylogenetically distinct from subclade II-a containing SVA742, which regulates secondary metabolite biosynthesis and morphological differentiation in S. avermitilis (25, 28, 29), and AdpA homologs from other genera (30). A phylogenetic analysis indicated that AdpA_Sx may play a special role in morphological differentiation or secondary metabolism in S. xiamenensis 318.

Target genes of AdpA_Sx potentially involved in cell growth and development.

Our results showed that AdpA_Sx negatively affects cell growth and morphological differentiation in S. xiamenensis 318. Electrophoretic mobility shift assays (EMSAs) were therefore performed on several putative AdpA_Sx targets involved in cell growth and morphological differentiation: ssgA (sxim_03860), encoding a putative membrane protein homologous to ssgB of Streptomyces coelicolor (31); ftsZ (sxim_10220), encoding a putative cell division protein; ftsH (sxim_23050), encoding a putative cell division GTPase; amfC (sxim_30820), encoding an aerial mycelium-associated protein (32); sxim_18210 (whiB), sxim_19730 (wblA1), sxim_24890 (wblA2), and sxim_39910 (wblE), encoding WhiB-family proteins that regulate morphological differentiation (33, 34); sxim_29740, encoding a sporulation-associated protein; bldA, encoding a UUA-reading tRNA-leucine (21); bldD (sxim_03450), encoding a pleiotropic negative regulator of morphological and physiological development; and bldG (sxim_41700), encoding an anti-antisigma factor homolog to BldG, which controls key developmental processes in S. coelicolor and S. griseus (35, 36).

His6-tagged AdpA_Sx was overexpressed and purified from Escherichia coli for EMSAs (see Fig. S4). EMSA results indicated that AdpA_Sx binds to the putative promoter regions, i.e., regions of 150 to 300 bp upstream of target genes, except for the region upstream of bldG (Fig. 3a). Subsequently, the transcription levels of these target genes were measured by qRT-PCR following strain growth on glucose-yeast extract-malt extract (GYM) medium for 24 h or 48 h. Transcription levels of ssgA, ftsZ, ftsH, amfC, whiB, whiA1, whiA2, whiE, and sxim_29740 were lower in the OEadpA strain than in the wild-type strain (Fig. 3b). However, transcription of the regulatory gene bldD was enhanced in the OEadpA strain (Fig. 3b).

FIG 3.

Predicted target genes involved in cell growth and morphological differentiation. (a) EMSAs with AdpA_Sx-His₆ and the promoter regions of ssgA, ftsZ, ftsH, amfC, whiB, wblA1, wblA2, wblE, sxim_29740, bldA, bldD, and bldG. (b) qRT-PCR analysis of transcripts of identified AdpA_Sx target genes (ssgA, ftsZ, ftsH, amfC, whiB, wblA1, wblA2, wblE, sxim_29740, and bldD) in S. xiamenensis 318 and the OEadpA strain. (c) AdpA_Sx positively regulates activity of the promoter bldAp. WT and the ΔadpA mutant were transformed with pDR3-bldAp (bldAp-xylE) and cultured in GYM medium. Cells were collected and disrupted by sonication at 24 h and 48 h, and XylE was measured.

The transcription of bldA could not be measured by qRT-PCR, because the tRNA-leucine is too short. Therefore, the reporter gene xylE was selected to evaluate the activity of the bldA promoter in the wild-type and ΔadpA strains. Notably, the XylE activity in the ΔadpA strain was reduced by 70% at 24 h and by 74% at 48 h compared to activity in the wild-type strain (Fig. 3c), suggesting that bldA is activated by AdpA_Sx. Taken together, these findings indicate that AdpA_Sx negatively controls cell growth and morphological differentiation by directly repressing the transcription levels of genes ssgA, ftsZ, ftsH, amfC, whiB, wblA1, wblA2, wblE, and sxim_29740 and by activating transcription of bldA and bldD.

Overexpression of adpA_Sx increases the production of xiamenmycin while decreasing PTM production.

The effects of overexpression or deletion of AdpA_Sx on the biosynthesis of xiamenmycin and PTMs were further investigated. During growth on GYM medium, the titer of xiamenmycin produced by the OEadpA strain reached 43.3 mg/liter, 435% higher than levels produced by the wild-type strain (Fig. 4a). Although the production of xiamenmycin decreased in the ΔadpA strain by 29% compared to wild-type levels, production was still detectable (Fig. 4a), indicating that AdpA_Sx is not required for xiamenmycin biosynthesis. This contrasts with other results showing that the ability to synthesize some products was lost in adpA deletion mutants of multiple Streptomyces species (10–13, 15, 37).

FIG 4.

Effect of adpA_Sx overexpression on the production of xiamenmycin and PTMs. (a) The yields of xiamenmycin in WT, OEadpA, ΔadpA, and Δika strains. (b) Growth curves of WT, OEadpA, and ΔadpA strains in GYM medium. (c) The production of PTM compounds 1, 4, 6, and 9 in WT, OEadpA, and ΔadpA strains. (d) HPLC analysis of PTMs in WT and OEadpA strains. Absorbance at 325 nm was monitored. For more information, see the supplemental material.

As mentioned above, the cell growth of the OEadpA strain was remarkably delayed during growth on TSB medium. To verify that the increased production of xiamenmycin by the OEadpA strain was not due to altered cell growth, the growth of the wild-type strain and the OEadpA and ΔadpA mutant strains was further evaluated on GYM medium. Cell growth of the OEadpA strain was decreased compared to that of the wild-type strain, indicating that the improvement in xiamenmycin production did not result from greater cell growth (Fig. 4b). Moreover, the titer of xiamenmycin produced by the OEadpA strain was further enhanced in GYM-plus medium, reaching 115.7 mg/liter.

Interestingly, PTM production by the OEadpA strain was markedly decreased compared to production by the wild-type strain (Fig. 4c and d). To determine whether the enhancement of xiamenmycin production resulted from the decreased levels of PTMs, a Δika strain, which lacks the PTM biosynthesis gene cluster, was constructed by homologous recombination and confirmed by PCR analysis (see Fig. S5). The production of xiamenmycin by the Δika strain had no obvious change compared to wild-type strain production (Fig. 4a), demonstrating that the improvement of xiamenmycin was not caused by the decrease in PTMs.

As shown in Fig. S5c, compounds 1 (capsimycin G), 4 (capsimycin), 6 (capsimycin B), and 9 (ikarugamycin), which are major PTM compounds, were isolated and identified in S. xiamenensis 318 (38). The production of compounds 1 and 9 by the OEadpA strain decreased significantly, by 80% and 96%, respectively, compared with the wild-type strain levels (Fig. 4c). Even more notable was the complete absence of compounds 4 and 6 in the OEadpA strain (Fig. 4c and d). In the ΔadpA strain, the production of compounds 4, 6, and 9 increased by 41%, 50%, and 35%, respectively, compared to wild-type levels (Fig. 4d). However, the production of compound 1 in the ΔadpA strain was decreased. These results indicated that AdpA_Sx is a bidirectional regulator, positively regulating xiamenmycin biosynthesis and negatively regulating PTM biosynthesis.

AdpA_Sx activates xim structural genes and represses PTM biosynthetic genes directly.

Next, we wanted to find out whether AdpA_Sx regulates xiamenmycin and PTM biosynthetic structural genes directly or through other regulators. In general, the genes required for antibiotic biosynthesis are clustered together and cotranscribed. For the xim gene cluster, the lengths of the intergenic DNA segments between sxim_01900 and ximA, ximA and ximB, ximB and ximC, ximC and ximD, and ximD and ximE are 160 bp, 224 bp, 65 bp, 8 bp, and 1 bp, respectively, indicating that ximB, ximC, ximD, and ximE could be cotranscribed (see Fig. S6). The sequence 5′-TGGCGCGAAC-3′, observed in the region upstream of ximA (Fig. S6), was a potential AdpA binding site based on its similarity to the consensus sequence 5′-TGGCSNGWWY-3′ (where S is G or C, W is A or T, Y is T or C, and N is any nucleotide), derived from the binding of AdpA to over 500 operator regions in S. griseus (39). Thus, we cloned the putative promoter regions of ximA (PrximA) and ximB (PrximB) as probes, and then the promoters were incubated with His6-tagged AdpA_Sx for EMSAs in vitro. EMSA results showed that the complex AdpA_Sx-P_ximA was formed in a protein concentration-dependent manner, confirming that AdpA_Sx specifically bound to the promoter region of ximA (Fig. 5a). The formation of two complexes for AdpA_Sx-P_ximB, as indicated by two shifted bands, suggested that AdpA_Sx also specifically bound to the promoter region of ximB.

FIG 5.

AdpA_Sx directly activates xiamenmycin biosynthetic genes and represses PTM biosynthetic genes. EMSAs with AdpA_Sx and probes P_ximA and P_ximB (a) and probes P_ikaA and P_ikaD (b). Transcriptional levels of xiamenmycin biosynthetic genes (c) and PTM biosynthetic genes (d) in WT and OEadpA strains grown in GYM medium.

For the PTM gene cluster, the lengths of the intergenic DNA segments between sxim_40850 and ikaA, ikaA and ikaB, ikaB and ikaC, and sxim_40680 and the gene ikaD are 373 bp, 4 bp, 1 bp, and 231 bp, respectively (Fig. S6), suggesting that ikaB and ikaC are likely cotranscribed with ikaA. Therefore, the promoters of ikA and ikaD were cloned and used as probes with His6-tagged AdpA_Sx. As shown in Fig. 5b, increasing concentrations of AdpA_Sx resulted in increased shifting of the PikaA and PikaD probes, suggesting that AdpA_Sx can specifically bind the promoter regions of ikaA and ikaD. These findings suggest that AdpA_Sx regulates xiamenmycin production and PTM production through directly controlling the transcription of structural genes in the cluster.

Additionally, we performed qRT-PCR to assess the effect of AdpA_Sx on the transcription levels of xiamenmycin and PTM biosynthetic genes. The transcription levels of ximA, ximB, ximC, ximD, and ximE in the OEadpA strain were increased by 160%, 295%, 398%, 248%, and 364%, respectively, at 24 h, and by 205%, 406%, 849%, 439%, and 555% at 48 h during growth on GYM medium (Fig. 5c). With regard to PTM biosynthesis, the transcription levels of the structural genes ikaA, ikaB, ikaC, and ikaD in the OEadpA strain were decreased by 67%, 34%, 56%, and 41% at 24 h and by 82%, 76%, 72%, and 72% at 48 h during growth on GYM medium (Fig. 5d). These results demonstrated that overexpression of adpA_Sx increased the transcription of xiamenmycin biosynthetic genes and repressed the transcription of PTM biosynthetic genes, consistent with the enhanced production of xiamenmycin and the decline in production of PTMs in OEadpA.

AdpA_Sx is negatively autoregulated.

Members of the AraC family of transcriptional factors are generally autoregulated (25, 40–42). To investigate whether AdpA_Sx regulates the expression of its own gene, we selected a 290-bp putative promoter region of adpA_Sx containing a predicted AdpA binding site as probe for use in EMSAs (Fig. 6a). When incubated with increasing concentrations of AdpA_Sx, this PadpA probe showed increased shifting, revealing that AdpA_Sx binds to its own promoter (Fig. 6b). To determine whether AdpA_Sx is a positive or negative autoregulator, XylE activity was selected to evaluate the promoter activity of adpA_Sx in the wild-type strain and the ΔadpA strain. XylE activity was enhanced by 106% at 24 h and by 93% at 48 h in the ΔadpA strain (Fig. 6c), suggesting that AdpA_Sx regulates itself negatively. In conclusion, AdpA_Sx, as with other AdpA homologues, functions as a repressor of its own gene.

FIG 6.

AdpA_Sx is a negative autoregulator. (a) Putative AdpA_Sx binding sites and EMSA probes in the promoter region of adpA_Sx. Probe PadpA is highlighted in gray. A possible AdpA_Sx binding site is marked and shown in red lettering. (b) EMSAs of AdpA_Sx-His₆ binding to the promoter region of adpA_Sx. (c) AdpA_Sx negatively regulates adpAp activity. WT and the ΔadpA mutant were transformed with pDR3-adpAp (adpAp-xylE) and cultured in GYM medium. Cells were collected and disrupted by sonication at 24 h and 48 h, and 2,3-dioxygenase activity was measured.

DISCUSSION

In general, the biosynthesis of secondary metabolites and morphological differentiation in Streptomyces are regulated by complex regulatory networks that sense and respond to numerous environmental conditions. Previous studies have revealed that the pleiotropic transcriptional regulator AdpA controls the biosynthesis of secondary metabolites by regulating the expression of CSRs and that it positively regulates morphological differentiation (12, 13). In this study, we investigated the regulatory roles of an AdpA homologue in a naturally minimized genomic background (Fig. 7).

FIG 7.

Proposed model of the AdpA_Sx-mediated regulatory network in S. xiamenensis 318. Solid lines indicate that direct regulation was confirmed experimentally. Dashed lines indicate that bldA would affect the translation levels of the TTA codon-containing genes adpA_Sx and ximC.

In previous studies, the deletion of adpA consistently resulted in sporulation-defective phenotypes (10–13, 15). Remarkably, a sporulation-defective phenotype was observed with our OEadpA strain, in which the adpA gene was overexpressed rather than deleted. The negative effect of AdpA_Sx on morphological differentiation was further confirmed in the heterologous host S. lividans TK24. Subsequent EMSA analysis and qRT-PCR showed that AdpA_Sx can repress the transcription levels of the ssgA, amfC, whiB, wblA1, wblA2, wblE, and sxim_29740 genes, different from the study on AdpA in S. griseus. We noticed that AdpA_Sx differs from other AdpA proteins in several conserved amino acids, the presence of a linker region, and the varied position of AdpA binding site in the promoter region of target genes (see Fig. S7 in the supplemental material). Such differences probably resulted in the unique effects on morphological differentiation.

It is interesting that AdpA_Sx functions as both an activator and repressor. We noted all the possible AdpA binding sites and predicted −35 and −10 regions in the promoter regions of genes involved in cell growth and morphological differentiation (Fig. S7). The possible AdpA binding sites is located upstream of the −35 region of the promoter regions of bldA and bldD, and this may enhance RNA polymerase binding to the promoter region. In the case of other genes, most of the other possible AdpA binding sites are located on −10 or −35 regions or the region between the −10 and −35 regions, and this may hinder RNA polymerase binding to the promoter region. Therefore, the location of AdpA binding sites at target genes makes sense in terms of its function as an activator or repressor.

The bldA gene encodes the only tRNA for the rare leucine codon UUA and plays a crucial role in translational regulation of morphological differentiation (9). Mutation of bldA not only affects the translation level of TTA-containing genes involved in morphological differentiation but also impairs the translation of AdpA, which positively regulates morphological differentiation (9). Our results confirmed that AdpA_Sx positively regulates the transcription level of bldA. Notably, a TTA codon was found in AdpA_Sx, indicating that bldA could affect the expression of adpA_Sx. However, the target genes of AdpA_Sx in S. xiamenensis 318, i.e., ssgA, amfC, whiB, wblA1, wblA2, wblE, sxim_29740, and bldD, have no TTA codon, which suggests that bldA does not directly influence the translation level of these genes involved in morphological differentiation. Nevertheless, as we found that AdpA_Sx negatively regulates these TTA codon-free genes, by inference, bldA could indirectly downregulate the transcription of these genes by its positive effects on the translation of AdpA_Sx. bldA also affects secondary metabolism in the genus Streptomyces via structural genes containing the TTA codon in several biosynthetic gene clusters, such as the lincomycin and moenomycin gene clusters (11, 43). Two TTA codons were found in ximC of the xiamenmycin gene cluster (Fig. S6), and we also showed that AdpA_Sx positively regulates the transcription level of bldA. Therefore, the increased transcription of bldA due to overexpression of adpA_Sx could contribute to the increased biosynthesis of xiamenmycin.

Although we observed that AdpA_Sx positively affects the production of xiamenmycin, there was little effect on the production of xiamenmycin by exogenous feeding with the γ-butyrolactone analogue 1,4-butyrolactone (44). Whether the A-factor cascade regulon exists in S. xiamenensis 318 has been investigated. afsA and arpA were found frequently as paired neighbors on Streptomyces chromosomes (45). sxim_17800, encoding the putative A-factor biosynthesis protein AfsA homolog, is flanked by sxim_17790, which encodes the SARP family regulator AfsS (see Fig. S8a). No gene homolog of the A-factor receptor protein ArpA was detected in S. xiamenensis 318 by BLASTp analysis (Fig. S8b). Moreover, there is no conserved ArpA binding site sequence in the promoter region of AdpA_Sx, suggesting that the expression of adpA_Sx would not be affected by ArpA (Fig. S8c). These findings may indicate that ArpA was lost from the streamlined genome of S. xiamenensis 318. Additionally, we detected sxim_43910, a putative homolog of absB, the gene encoding RNase III, which cleaves AdpA_Sc mRNA in S. coelicolor (46). Therefore, the expression of AdpA_Sx, as with the S. coelicolor and Streptomyces ghanaensis homologs (11, 46), could be affected by AbsB.

The PTM biosynthetic pathway has been found in phylogenetically diverse bacteria (24). Many PTM gene clusters are silent and can be activated by a “plug-and-play” substitution of stronger promoters (47). AdpA homologs are widely conserved pleiotropic transcriptional regulators, and manipulation of the expression level of adpA will have profound effects on the physiological status of Streptomyces. The changed profiles of PTMs in the OEadpA and ΔadpA strains suggest that AdpA overexpression or deletion leads to the appearance of structurally different PTM compounds or to the improvement of PTM production (see Fig. S9 and S10).

We analyzed the possible AdpA binding sites of all the genes tested (Fig. S7). To identify the AdpA_Sx regulon in S. xiamenensis 318, all the possible AdpA_Sx binding sites were used to scan the genome using the PREDetector programs (http://predetector.hedera22.com/). In addition to identified target genes of AdpA_Sx by EMSAs, a total of 121 putative AdpA_Sx targets were identified with the score of the reliability of the predicted targets set at 7 (see Excel file in the supplemental material). These target genes could be classified into functional groups, including carbohydrate metabolism (target 9), nucleic acids metabolism (7), secondary metabolism (1), transporter (6), transcription and translation (3), redox reaction (8), regulation function (12), and unclassified or unknown function proteins (75).

In conclusion, we characterized the regulatory role of AdpA_Sx in S. xiamenensis 318, a species with a streamlined genome. AdpA_Sx acts as a global regulator, negatively affecting morphological differentiation and bidirectionally regulating the biosynthesis of xiamenmycin and PTMs. Our findings provide additional evidence for the versatility of AdpA in regulating morphological differentiation and secondary metabolism.

MATERIALS AND METHODS

Strains, plasmids, and media.

All strains and plasmids used in this study are listed in Table 1. S. xiamenensis 318 and its derivatives were grown at 30°C with appropriate antibiotics. MS agar medium (2% soybean powder, 2% mannitol, and 2% agar) was used for sporulation and phenotypic observation. Tryptic soybean broth (3% TSB) was used as seed medium and for the growth of mycelia for the purposes of extracting DNA. GYM medium (0.4% yeast extract, 1% malt extract, and 0.4% glucose) and GYM-plus medium (1.5% glucose, 1% yeast extract, 1.5% malt extract, 0.15 g/liter p-hydroxybenzoic acid [PHBA], and 0.1 g/liter l-threonine) were used as the fermentation media. Escherichia coli DH5α and E. coli Rosetta (DE3) (Novagen) were used as the cloning host and expression host, respectively. E. coli ET12567(pUZ8002) was used to propagate nonmethylated DNA for transformation into S. xiamenensis 318.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Characteristics	Source or reference
Strains
E. coli
DH5a	General cloning host	TaKaRa
Rosetta (DE3)	Host for expression of AdpASx	Novagen
ET12567(pUZ8002)	Donor strain for conjugation between E. coli and Streptomyces, dam dcm hsdS/pUZ8002	49
Streptomyces
S. xiamenensis 318	Wild-type xiamenmycin-producing strain	20
S. lividans TK24	str-6 SLP2– SLP3–	49
OEadpA mutant	Wild-type strain carrying pIB139-adpASx	This work
ΔadpA mutant	Wild-type strain with deletion of adpASx	This work
Δika mutant	Wild-type strain with deletions of ikaA, ikaB, and ikaC	This work
WT::pDR3(adpAp)	Wild-type strain carrying pDR3-adpAp	This work
ΔadpA::pDR3(adpAp) mutant	ΔadpA strain carrying pDR3-adpAp	This work
WT::pDR3(bldAp) mutant	Wild-type strain carrying pDR3-bldAp	This work
ΔadpA::pDR3(bldAp) mutant	ΔadpA strain carrying pDR3-bldAp	This work
WT-AdpASl	Wild-type strain carrying pIB139-adpASl	This work
S. lividans TK24-AdpASx	S. lividans TK24 carrying pIB139-adpASx	This work
Plasmids
pIB139	Integrative vector based on ΦC31 integrase, ermE*p	53
pIB139-adpASx	pIB139 derivative for overexpression of adpASx	This work
pIB139-adpASl	pIB139 derivative for overexpression of adpASl	This work
pJTU1278	bla, tsr, lacZ, oriT, oripIJ101, oriColE1	57
pDadpA	pJTU1278 derivative for adpASx deletion	This work
pDika	pJTU1278 derivative for ikaA, ikaB, and ikaC deletion	This work
pDR3	Double-reporter vector containing a xylE-neo cassette and multicloning site	58
pDR3-adpAp	Insertion of adpAp upstream of xylE in pDR3	This work
pDR3-bldAp	Insertion of bldAp upstream of xylE in pDR3	This work
pET-24b	kan, PT7, His-tag	Novagen
pET-24b-adpASx	pET-24b derivative carrying adpASx	This work

E. coli strains were grown at 37°C in Luria-Bertani (LB) medium or on LB plates. When necessary, the media were supplemented with antibiotics (100 μg/ml for ampicillin, 50 μg/ml for each of apramycin and kanamycin, and 25 μg/ml for chloramphenicol).

DNA manipulation and sequence analysis.

All primers used in this study are listed in Table 2. DNA manipulations were performed according to standard procedures for Streptomyces and E. coli (48, 49).

TABLE 2.

Primers used in this study

Namea	Sequence (5′→3′)b
adpAU-F	GAGCTCCACCGCGGTGGCGGCCGCTCTAGAACCGTCGCGATCATCCCGGACGATG
adpAU-R	CCTCTCGGTCCTGCACGTTTCCCCC
adpAD-F	GGGGGAAACGTGCAGGACCGAGAGGACGGGAGCACCACCCACAGGTACAC
adpAD-R	CCCCTCGAGGTCGACGGTATCGATAAGCTTGCTCTGGCTGACGACCAGCGGAGAG
adpAcheck-F	CTGTTTCGTCAGCCACAAGC
adpAcheck-R	CACACCATCCGGTCGTTCAT
ikaU-F	GCTCTAGATGGGGATCATCGAGCCGGGT
IkaU-R	CGGGATCCGCCCAACACGCTGGCCCTTT
ikaDH-F	CGGGATCCCGTGAGAATCCTTGGGTGAT
IkaDH-R	CCCAAGCTTGAGCGTCACCGTCAGGGTCA
ikacheck-F	TATCCGCTTCACCCTCTACGA
ikacheck-R	GCATTCCTGCGGCTGATGAT
apra-F	GGGGTACCTGGTTCATGTGCAGCTCCATC
apra-R	GGGGTACCTGAGCTCAGCCAATCGACTG
adpA-F	GGAATTCCATATGATGCGACGACCGAGGGAACTGAG
adpA-R	CGGAATTCTCAGGCCACCGGCCGGCCCCGCT
adpAsl-F	CAATCGTGCCGGTTGGTAGGATCCACATATGATGAGCCACGACTCCACCGCCGCGC
adpAsl-R	CAGGAAACAGCTATGACATGATTACGAATTCTCACGGCGCGCTGCGCTGGCCCGGG
P1	CCGGCACGGTGATCCGCTGG
P2	TCCGAGACAATCGGAATCGC
P3	GCTCTAGACTTGGGCTGCAGGTCGACTC
P4	AGGATCGTGGCATCACCGAA
P5	TAGGGCAGCAGATAGCCCGT
adpA-24b-F	CGGAATTCGATGCGACGACCGAGGGAACTGAGG
adpA-24b-R	CCGCTCGAGGGCCACCGGCCGGCCCCGCTGCCCC
Q_adpA-F	GCGACAACCCACTGGATGTA
Q_adpA-R	GGACGATGTGCAGACACAGA
PrximA-F	CCCGGTGCCGAGCTTGAGCGGGC
PrximA-R	CCGGGCAACGTTGCGCCATTATG
PrximB-F	ACACGGTTCGGCAAACACCGTTAG
PrximB-R	CTGACCGGGCGAGGGTTATTGTTG
PrikaA-F	CATAAGGCAGGCTTCAACCG
PrikaA-R	CGTGAGAATCCTTGGGTGAT
PrikaD-F	GGAGTTCCCCGGCTATGTGC
PrikaD-R	GGGCGGTGTTGCCTTTCGTG
Q_ikaA-F	TCCACCTTCACCTTCATC
Q_ikaA-R	GTTCTGCCAGTTGAGGAA
Q_ikaB-F	GTACAACGCCAAGGTGGAGA
Q_ikaB-R	CGCCAGGTGTCGTTGAGATA
Q_ikaC-F	TACACGGAGGAGATGGTC
Q_ikaC-R	TGGATGGTCTGGTGGTAG
Q_ikaD-F	CCACCAGATCGCCAACTTCA
Q_ikaD-R	AAGGAGAAGCCGCTGAGGAT
PrssgA-F	ACCATACACAGCTCAGCGCCCCAGA
PrssgA-R	AAGCGCCAGGGCCTTTCGCTCAGTG
PrftsZ-F	CGATGGATACCTGGTTCGCCGCACC
PrftsZ-R	GTCGAAGGCCTCTCGCCTCGATTAC
PrftsH-F	GTCTATTCACACGGGGCGGTGCG
PrftsH-R	CTACCAGGGCTGAGCGGACCGCGC
PramfC-F	GCGGAATTTTCCCCTCATCCCCATTTC
PramfC-R	AGCACACACAGTCCTTGGCCAACG
PrwblA1-F	GGTGGGCGAGCGGCTGCCCGCACAG
PrwblA1-R	GACCATTCTCAGCTGAGGACGTTAC
PrwblA2-F	GCCCATACATCGGACAAAACAAGAG
PrwblA2-R	GACCGGTGCCGTCCTCTCCCGAATC
PrwblE-F	GCGGAACAGCTAAAAGTCCTTA
PrwblE-R	GTCTGCTCCATCTCCTGCACG
PrwhiB-F	GTGCGCGCCCCTCGTCCTCTTTG
PrwhiB-R	GGTCAGGAGCACGCATCAGTCACGC
Pr29740-F	CCCCTTCGTCACCCATCGTGCGACG
Pr29740-R	TTAACGGTCAGCAGGCACCGAAC
PrbldA-F	CCTGGGATCCGTCGCACTGGTGG
PrbldA-R	ACAGAATCCCCAAGCCGTACCAAG
PrbldD-F	CGCCTTTGGCGCGTTCGACGATC
PrbldD-R	GTGTCTCCCCGGACGGCGGTTTTG
PrbldG-F	ACCCGGTTCGGGCGGGCGGGGG
PrbldG-R	ATAACACTCCCCGGAATTGCTT
Q_ssgA-F	GAGTCCTCACTGCCTGTAC
Q_ssgA-R	CCACTCGACGGTTTCTTC
Q_ftsZ-F	CGTCAAGTCCGTGATGTC
Q_ftsZ-R	TCGTTGATCTCGAACAGAC
Q_ftsH-F	ATCTGAACAACGTCCTCAAC
Q_ftsH-R	CTTCTCCTTGTCGCTCATG
Q_amfC-F	CCGACGACGAACTGCATGA
Q_amfC-R	CGAGCAGATCATCTACCTGGG
Q_wblA1-F	AACCGACTGGAGCAAGAAC
Q_wblA1-R	TCCGAACTCGATCCTGTG
Q_wblA2-F	GTATGACCGAACGGGAAC
Q_wblA2-R	CGTAACCTCCGTCACTCTC
Q_wblE-F	ATGGACTGGCGTCACAAC
Q_wblE-R	CCTTGGCTTCCTCGATCTG
Q_whiB-F	GTACTCCAGGCATTCGGAGC
Q_whiB-R	ACGAAGCGGACGAAGAGATG
Q_29740-F	GTCCTTCTACGCCTACGAC
Q_29740-R	GATTGCCCGATTCCAGTTC
Q_bldD-F	TACACCGCCACCATCCAGAG
Q_bldD-R	GGGACTGGTCGTAGATCACC
PrbldA-xylE-F	GCTGAAAGGAGGAACTATATCCGCGGGATCCCCTGGGATCCGTCGCACTGGTGGTT
PrbldA-xylE-R	ATGACGTCACCTCTTCAACTCAGATACTAGTACAGAATCCCCAAGCCGTACCAAGC
PradpA-F	CAGCCAACTTCCCCACTGCTGAAA
PradpA-R	CCTCTCGGTCCTGCACGTTTCCCCC
PradpA-xylE-F	GCTGAAAGGAGGAACTATATCCGCGGGATCCCAGCCAACTTCCCCACTGCTGAAA
PradpA-xylE-R	ATGACGTCACCTCTTCAACTCAGATACTAGTCCTCTCGGTCCTGCACGTTTCCCCC

^aF, forward primer; R, reverse primer.

^bRestriction endonuclease sites are underlined.

Multiple sequence alignment and phylogenetic analysis.

A homologous sequence database search was executed using BLASTp (https://blast.ncbi.nlm.nih.gov/Blast.cgi?PROGRAM=blastp&PAGE_TYPE=BlastSearch&LINK_LOC=blasthome). Multiple sequence alignment of AdpA homologues was performed using online tools available from the European Bioinformatics Institute (https://www.ebi.ac.uk/Tools/msa/muscle/) with default parameters and then analyzed by ESPript 3.0 (http://espript.ibcp.fr/). Phylogenetic analyses were conducted in MEGA7 (50). The evolutionary history was inferred by using the maximum likelihood method based on the JTT matrix-based model (51). The phylogenetic tree was further edited by EvolView (52).

Construction of deletion mutants and overexpression mutants.

For in-frame gene deletion of either adpA (sxim_38050) or the PTM gene cluster, left and right flanking fragments were amplified from the genome of S. xiamenensis 318 with primer pairs adpAU-F/R (1,828 bp) and adpAD-F/R (1,735 bp) for adpA or ikaU-F/R (1,571 bp) and ikaDH-F/R (1,631 bp) for the PTM gene cluster. The two fragments, ikaU and ikaDH, were digested by XbaI/BamHI and BamHI/HindIII, respectively, and ligated into XbaI/HindIII-digested pJTU1278, generating plasmid ika-pJTU1278. With plasmid pIB139 as the template, an aac(3)IV cassette was cloned using primer pair apar-F/R, the purified PCR product was digested by KpnI and ligated into corresponding sites of the plasmid ika-pJTU1278, resulting in plasmid pDika. For the construction of pDadpA, the insert fragment for replacing the ika genes on pDika was removed by restriction enzyme digestion using XbaI/HindIII, and then the 10.1-kb plasmid backbone fragment (pJTU1278A) from pDika was recovered. The purified DNA fragments adpAU and adpAD were ligated with the above-mentioned XbaI/HindIII plasmid fragment by one-step assembly, generating plasmid pDadpA. pDadpA and pDika were sequenced and then transferred separately into the parental strain S. xiamenensis 318 by intergeneric conjugation. Second crossover events were expected after a round of nonselective growth of the initial apramycin-resistance exconjugants. The marker-free Δika and ΔadpA deletion mutants were screened for by a diagnostic PCR using the corresponding pair primers ikaCheck-F/R or adpACheck-F/R, and the removal of the whole PTM gene cluster or adpA was confirmed by DNA sequencing of the mutant gene locus in the Δika and ΔadpA strains.

For gene overexpression, the gene adpA_Sx was amplified from the genome of S. xiamenensis 318 with respective primers adpA-F/-R. The PCR product were purified and digested by NdeI and EcoRI and ligated into the corresponding sites of the integrative expression plasmid pIB139 (53), generating plasmid pIB139-adpA_Sx. The gene adpA_Sl was amplified from the genome of S. lividans TK24 with primers adpAsl-F/-R. The PCR product was purified and ligated with the plasmid pIB139 digested by NdeI/EcoRI using one-step assembly, generating plasmid pIB139-adpA_Sl. pIB139-adpA_Sx, which carries adpA_Sx under the control of the strong promoter ermE*p, was confirmed by sequencing, introduced into E. coli ET12567(pUZ8002), and then transferred into the parental strain S. xiamenensis 318 (wild type) and S. lividans TK24 by intergeneric conjugation. Plasmid pIB139-adpA_Sl was also introduced into E. coli ET12567(pUZ8002) and then transferred into the parental strain S. xiamenensis 318 (wild type) by intergeneric conjugation. The OEAadpA overexpression strain, S. lividans TK24-AdpA_Sx, and S. xiamenensis 318-AdpA_Sl were obtained by apramycin resistance screening and OEAadpA confirmed by PCR analysis with the primers P1/P2, P3/adpA-24b-R, and P4/P5.

Scanning electron microscopy.

S. xiamenensis 318, S. lividans, and their derivative strains were inoculated onto cellophane-covered MS agar plates and incubated at 30°C for 7 days. To obtain bacterial cells, equivalent areas of the cellophane were excised and soaked with FA solution (50% ethyl alcohol, 5% glacial acetic acid, and 10% formaldehyde) at 4°C overnight. The cells were washed twice with water after removal of the supernatant and then subjected to gradient dehydration using ethanol solutions ranging from 30% to 100%. The samples were dried naturally, coated with gold, and then observed with a Sirion 200 field-emission scanning electron microscope.

Measurement of cell growth.

To assess cell growth, 1-ml samples of culture broth from independent flasks were taken at the various time points using trimmed tips. Cell growth was determined based on the intracellular protein concentration. The standard Bradford assay was used to determine the concentration of protein released from the mycelium, which was disrupted by sonication (37).

Production analysis of xiamenmycin and PTMs.

Mycelium from S. xiamenensis 318 and its derivatives was inoculated in TSB medium and grown for 30 h except for the OEadpA strain, which was grown for 72 h as seed culture, and then 10 ml seed broth was transferred into 100 ml GYM medium or GYM-plus medium. After 7 days of fermentation, each culture was then extracted three times at room temperature overnight with equal volumes of the solvent ethyl acetate. The supernatants were combined and concentrated under vacuum at 38°C to remove the organic phase. Each crude extract was subsequently dissolved in the same volume (5 ml) of high-performance liquid chromatography (HPLC)-grade methanol. The samples were diluted 5-fold, analyzed by HPLC using an Agilent XDB-C₁₈ column (4.6 mm by 150 mm, 5 μm), and monitored by UV detection at 254 nm and 325 nm. The solvent system of H₂O (A) and acetonitrile (B) was used as the mobile phase in the following linear gradient: 0 min, 25% B; 5 min, 25% B; 30 min, 90% B, at a flow rate of 1 ml/min.

RNA extraction and RT-PCR analysis.

Total mRNA was isolated from S. xiamenensis 318 and OEadpA strains after 24 h and 48 h growth in GYM liquid medium using TRIeasy reagent according to the manufacturer’s protocol (Yeasen Inc., China). The RNA concentration was determined using a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). Isolated RNA was treated with DNase I, and reverse transcription was achieved using a cDNA synthesis kit (TaKaRa, Japan). The transcriptional levels of genes were determined by qRT-PCR using a StepOnePlus real-time PCR system (Applied Biosystems, Foster City, CA, USA) with the SYBR/ROX qPCR Master Mix (NEB, USA). The following PCR conditions were used: 95°C for 1 min, 40 cycles of 95°C for 15 s and 60°C for 30 s, and an extension at 72°C for 10 min. The housekeeping gene hrdB from S. xiamenensis 318 was used as the internal control to normalize samples, with quantification by the 2^−ΔΔCT method (54).

Heterologous expression and purification of AdpA-His₆.

For heterologous expression of AdpA, adpA was amplified by PCR from the genome of S. xiamenensis 318 with primers adpA-24b-F/-R. The purified PCR product was digested by EcoRI/XhoI and cloned into pET-24b, digested by EcoRI/XhoI, generating the C-terminal His-tag fusion plasmid pET-24b-adpA_Sx. The plasmid construct was confirmed by sequencing. pET-24b-adpA_Sx was introduced into E. coli Rosetta (DE3) and then grown at 37°C in LB medium supplemented with kanamycin (50 µg · ml⁻¹) and chloramphenicol (34 µg · ml⁻¹). When an optical density of 0.6 to 0.8 at 600 nm was reached, the culture was induced with 0.4 mM isopropyl β-d-1-thiogalactopyranoside at 25°C for 15 h to express AdpA-His₆. Cells were collected by centrifugation at 8,228 × g at 4°C and then resuspended in equilibration buffer (50 mM phosphate buffer pH 7.4, 150 mM NaCl, 10% [vol/vol] glycerol, 0.1 mM dithiothreitol [DTT]). The AdpA-His₆ protein was released from cells by homogenization, applied to His-Tag purification resin, and eluted in elute buffer (50 mM phosphate buffer [pH 7.4], 150 mM NaCl, and 400 mM imidazole [pH 7.4]). The eluent was dialyzed in equilibration buffer, and then the solution was concentrated in concentration buffer containing 50 mM phosphate buffer (pH 7.4), 150 mM NaCl, 50% (vol/vol) glycerol, and 0.1 mM DTT. The quality of the purified proteins was judged by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and protein concentrations were determined with the Bradford assay.

Electrophoretic mobility shift assays.

EMSAs were performed as described previously (55). The putative promoter regions were amplified by PCR from the genome of S. xiamenensis 318 with respective primers listed in Table 2. The PCR products were purified using a PCR purification kit (Omega Biotech), and the concentration of these promoter probes was determined with a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). The probes were incubated individually with various concentrations of His6-tagged AdpA in binding buffer containing 10 mM phosphate buffer (pH 7.4), 1 mM EDTA, 100 mM KCl, 5% (vol/vol) glycerol, 0.1 mM DTT, and 0.01 mg/ml bovine serum albumin at 25°C for 30 min in 20 μl of reaction mixture. After incubating, the samples were fractionated on 6% native PAGE gels in ice-cold 0.5× Tris-acetate-EDTA (TAE) buffer at 100 V for 120 min and stained with ethidium bromide. The band shift was directly recorded and analyzed by using a Tanon 3500 gel imaging analysis system.

Catechol 2,3-dioxygenase activity assay.

The promoter regions of adpA and bldA were cloned following PCR amplification using primer pairs PradpA-xylE-F/-R and PrbldA-xylE-F/-R, respectively. The promoters were then ligated to the BamHI/SpeI-treated reporter plasmid pDR3, generating plasmids pDR3-adpAP and pDR3-bldAP, which were then sequenced. pDR3-adpAP and pDR3-bldAP were transferred separately into S. xiamenensis 318 and the ΔadpA strain. Mutant strains were obtained by apramycin resistance screening and confirmed by PCR. The colorless compound catechol can be converted into the yellow product 2-hydroxymuconic semialdehyde by catechol 2,3-dioxygenase (XylE). The catechol 2,3-dioxygenase activity assays were performed as described previously with minor modifications (56). Streptomyces mycelium was harvested, washed twice using cold deionized water, and suspended in cold sample buffer (100 mM phosphate buffer [pH 7.5], 20 mM Na₂EDTA [pH 8.0], 10% [vol/vol] acetone). Triton X-100 (1%) was added after cells were lysed by sonication, and the samples were centrifuged at 12,000 rpm for 5 min. Then, 10 µl supernatant was added to each well of the 96-well plates containing 200 µl assay buffer (10 mM phosphate buffer [pH 7.5], 0.2 mM catechol). The change at A₃₇₅ was recorded by a microplate reader (Synergy2; BioTek). The slope of the linear part of the A₃₇₅ output was used to calculate the specific activity as follows: mU catechol dioxygenase [nmol · min⁻¹] = 30.03 × ΔA₃₇₅/t [min].