This Article Abstract Full Text (PDF) Other Versions of this Article: JB.01989-07v1 190/9/3293    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Google Scholar Articles by Bunet, R. Articles by Aigle, B. PubMed PubMed Citation Articles by Bunet, R. Articles by Aigle, B.

 Previous Article  |  Next Article 

Journal of Bacteriology, May 2008, p. 3293-3305, Vol. 190, No. 9
0021-9193/08/$08.00+0     doi:10.1128/JB.01989-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Regulation of the Synthesis of the Angucyclinone Antibiotic Alpomycin in Streptomyces ambofaciens by the Autoregulator Receptor AlpZ and Its Specific Ligand

Robert Bunet,¹ Marta V. Mendes,^1, Nicolas Rouhier,² Xiuhua Pang,^1, Laurence Hotel,¹ Pierre Leblond,¹ and Bertrand Aigle^1*

Laboratoire de Génétique et Microbiologie, UMR INRA-UHP 1128, IFR 110,¹ Interactions Arbres/Micro-organismes, UMR INRA-UHP 1136, IFR 110, Faculté des Sciences et Techniques, Nancy Université, Vandœuvre-lès-Nancy, France²

Received 21 December 2007/ Accepted 13 February 2008

	ABSTRACT

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Streptomyces ambofaciens produces an orange pigment and the antibiotic alpomycin, both of which are products of a type II polyketide synthase gene cluster identified in each of the terminal inverted repeats of the linear chromosome. Five regulatory genes encoding Streptomyces antibiotic regulatory proteins (alpV, previously shown to be an essential activator gene; alpT; and alpU) and TetR family receptors (alpZ and alpW) were detected in this cluster. Here, we demonstrate that AlpZ, which shows high similarity to -butyrolactone receptors, is at the top of a pathway-specific regulatory hierarchy that prevents synthesis of the alp polyketide products. Deletion of the two copies of alpZ resulted in the precocious production of both alpomycin and the orange pigment, suggesting a repressor role for AlpZ. Consistent with this, expression of the five alp-located regulatory genes and of two representative biosynthetic structural genes (alpA and alpR) was induced earlier in the alpZ deletion strain. Furthermore, recombinant AlpZ was shown to bind to specific DNA sequences within the promoter regions of alpZ, alpV, and alpXW, suggesting direct transcriptional control of these genes by AlpZ. Analysis of solvent extracts of S. ambofaciens cultures identified the existence of a factor which induces precocious production of alpomycin and pigment in the wild-type strain and which can disrupt the binding of AlpZ to its DNA targets. This activity is reminiscent of -butyrolactone-type molecules. However, the AlpZ-interacting molecule(s) was shown to be resistant to an alkali treatment capable of inactivating -butyrolactones, suggesting that the AlpZ ligand(s) does not possess a lactone functional group.

	INTRODUCTION

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Streptomycetes are gram-positive, filamentous, soil-dwelling bacteria that undergo a complex program of morphological differentiation. In addition to this singular multicellular morphogenesis, the members of the Streptomyces genus are well known for their ability to produce a variety of secondary metabolites with important uses in medicine and in agriculture. Recent advances in the sequencing of Streptomyces genomes, and more generally those of actinomycetes, have highlighted the underestimated potential of these organisms to synthesize secondary metabolites. For example, bioinformatic analysis of the Streptomyces coelicolor genome revealed the presence of 22 gene clusters with predicted roles in the production of secondary metabolites; only half a dozen of them had previously been identified experimentally, despite decades of study (4, 8). Similarly, 30 secondary metabolite gene clusters have now been identified using the Streptomyces avermitilis genome sequence (18). More recently, the genome sequences of non-Streptomyces actinomycetes such as the biotechnologically important Rhodococcus sp. RHA1 (27), the marine strain Salinispora tropica that is responsible for producing the potent anticancer agent salinosporamide A (48), or the erythromycin-producing bacterium Saccharopolyspora erythraea (31) added, in total, more than 70 secondary metabolite biosynthesis clusters to the ever-growing list. From this wealth of sequences, a genome-mining approach has already led, for instance, to the discovery of the tris-hydroxamate tetrapeptide siderophore coelichelin from S. coelicolor A(3)2 (22).

In Streptomyces ambofaciens ATCC 23877, a known producer of congocidin (11) and spiramycin (36), the sequencing project of the left (1,544 kb; accession number AM238663) and right (1,367 kb; AM238664) arms of the linear chromosome has unveiled 11 novel secondary metabolite gene clusters (http://www.weblgm.scbiol.uhp-nancy.fr/ambofaciens/) (10). With the exception of the des cluster identified in the core region of the chromosome that directs the synthesis of desferrioxamine B and E, only the coelichelin-encoding cch cluster (2) and clusters presumably involved in the synthesis of carotenoids were found to be redundant compared with clusters present in the phylogenetically close relative S. coelicolor. This highlights the powerful ability of different streptomycete strains to produce distinct arrays of secondary metabolites.

In the course of deciphering the role of the remaining cryptic clusters in S. ambofaciens, the function of the type II polyketide synthase (PKS) gene cluster alp, located in the terminal inverted repeats (TIRs) and thus present in two copies, has been unraveled. This cluster was shown to be responsible for the synthesis of at least two distinct compounds: an antibiotic of the angucyclinone class, alpomycin, and also a diffusible orange pigment that is thought to be a degradation or modification product of the antibiotic (34). Interestingly, the chimerical formation of the alp cluster that presumptively occurred through multiple horizontal gene transfer events is still distinguishable (34). Indeed, one part, which includes the minimal PKS genes responsible for assembling the polyketide chain and coding for a β-ketoacyl synthase (alpA), a chain length factor (alpB), and an acyl carrier protein (alpC), shows strong synteny with a part of the kinamycin cluster of Streptomyces murayamaensis (accession number AY228175). Consistent with this, preliminary structural determination has shown that alpomycin is related to kinamycin C and D (L. Song and G. Challis, personal communication). A second part is identical to that found in the mithramycin gene cluster carried by the pSLA2-L plasmid of Streptomyces rochei (28) or the chromosome of Streptomyces argillaceus (24). Finally, a third region that includes most of the alp regulatory genes is highly conserved with the regulatory module of the type I PKS tylosin cluster in Streptomyces fradiae (3, 34).

Within the five regulatory genes that form the alp regulatory subcluster, alpT, alpU, and alpV are predicted by comparison with database sequences to encode proteins from the Streptomyces antibiotic regulatory protein (SARP) family (49). Among them, alpV was previously shown to be essential for the production of both the orange pigment and alpomycin (1). It has been postulated that the AlpV protein occupies the critical position in the alp regulatory cascade, leading to the activation of transcription of the biosynthetic genes, although direct binding to promoter regions within the cluster has not yet been demonstrated. The two other deduced regulatory genes, alpW and alpZ, encode proteins from the TetR transcriptional regulator family (37). The AlpW protein shows sequence similarity to proteins acting as transcriptional repressors, such as TylQ from S. fradiae, which is involved in tylosin production (41), and BarB from Streptomyces virginiae, which regulates virginiamycin (26). The deduced product of alpZ shares homology with -butyrolactone autoregulator receptors that act as transcriptional factors to regulate antibiotic production and sometimes morphological differentiation in streptomycetes (for reviews, see references 17 and 44).

Several -butyrolactone receptors have been identified in various Streptomyces spp. and also in one non-Streptomyces actinomycete. (9). Typically, the autoregulator receptor binds as a homodimer to conserved DNA sequences (exemplified by the BarA responsive elements) within the promoter region of target genes to repress their transcription (20). A binding consensus sequence for autoregulator receptors has more recently been defined as autoregulatory elements (ARE) (13). Upon binding of their cognate ligand, namely, the -butyrolactones, the DNA-binding ability of the receptor is abolished, which in turn allows expression of the target gene(s). The -butyrolactone receptors most likely occupy a high regulatory level within the pathway-specific cascade by globally repressing the synthesis of antibiotics until proper physiological conditions are reached (i.e., when a threshold concentration of -butyrolactone is attained).

Several previous studies characterizing signaling systems involving these receptors and their specific -butyrolactone interactive ligand(s) have been reported. For example, the model A-factor-ArpA system that controls the activation of the AdpA regulon (that includes a number of genes required for secondary metabolism and differentiation) in Streptomyces griseus (for review, see reference 17), the SCB1-ScbR system that exerts influence on actinorhodin and undecylprodigiosin production in S. coelicolor and directly controls the expression of a pathway-specific transcriptional regulator from the cryptic type I polyketide biosynthetic gene cluster (35, 45, 46), the virginiae butanolides-BarA system of S. virginiae (29), the IM-2-FarA system of Streptomyces lavendulae (21), and the recently published TylP system and its cognate unknown ligand controlling tylosin synthesis in S. fradiae (6).

In this study we address the role of AlpZ, a putative -butyrolactone receptor, in the alp regulatory cascade. We report AlpZ target promoters within the biosynthetic cluster and investigate whether activation of the alp cluster is influenced by -butyrolactone-like molecules produced in S. ambofaciens. The putative role of AlpZ as a sensor protein is discussed.

	MATERIALS AND METHODS

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Bacterial strains, plasmids, and growth conditions. Strains, plasmids, and cosmids used in this study are listed in Table 1. Streptomyces strains were cultivated on or in R2, SFM, HT, or SMMS medium (19) and manipulated as described previously (34). Pigment and antibiotic production were assessed on/in R2 medium as described previously (1, 34). Escherichia coli strains were cultivated in Luria-Bertani (LB) and SOB liquid medium (39).

PCR and RT-PCR. Amplification of DNA fragments by PCR was performed with Taq DNA polymerase (NEB) or high-fidelity DNA polymerase Phusion (Finnzymes), according to the manufacturer's instructions. The method used for reverse transcription-PCR (RT-PCR) analysis was as previously described (1, 34). The cDNAs were obtained after reverse transcription of 4 µg of DNase I-treated total RNA with SuperScript III reverse transcriptase (Invitrogen) and high-GC-content random hexamer primers (Oligo Spiking; Eurogentec). Primer pairs hrdB-F/hrdB-R, KSI-F/KSI-R, KSII-F/KSII-R, alpV-1/alpV-2, alpT-1/alpT-2, alpU-1/alpU-2 and alpW-1/alpW-2 (Table 2) were used to amplify cDNAs for an hrdB-like gene, alpA, alpR, alpV, alpT, alpU, and alpW, respectively. Two different sets of primers were used for transcriptional analyses of alpZ: alpZ-1/alpZ-2 was used in analysis of the wild type, and alpZscar-1/alpZscar-2 (amplification of the 81-bp scar; see below) was used in a double alpZ deletion mutant (the alpZ strain).

(a) Replacement of alpZ by the selectable aac(3)IV-oriT cassette in cosmid F19. The primer pair alpZ-rep1/alpZ-rep2 (Table 2) was used to amplify the aac(3)IV-oriT cassette from pIJ773 (15). The replacement of alpZ coding sequence (only the start and stop codons were retained in the mutant) by the amplified cassette was mediated by the recombination RED system harbored by pIJ790 in E. coli BW25113 (15). The targeted F19 cosmid was designated F19alpZ::aac(3)IV-oriT (Table 1). Gene replacement was confirmed by Southern blotting and PCR analysis using the flanking primers CZ1 (124 nucleotides upstream from the start codon of alpZ) and CZ2 (149 nucleotides downstream the stop codon of alpZ), giving a product of 984 bp in the wild type and 1,658 bp after replacement with the aac(3)IV-oriT cassette.

(b) In-frame deletion of alpZ in the F19 cosmid mediated by Flp recombinase. Excision of the aac(3)IV-oriT cassette in F19alpZ::aac(3)IV-oriT through the recombination between the two Flp recognition target sites flanking the cassette, leaving an 81-bp scar, was carried out as described previously (1, 15). The deletion of alpZ in the generated F19alpZ::scar was confirmed by restriction and PCR analysis using primers CZ1 and CZ2 (354-bp expected fragment).

(c) Construction of F19::aadA-oriT-alpZ::scar. To render the F19alpZ::scar cosmid proficient for conjugation, the neo resistance gene of the cosmid part was replaced by an aadA-oriT cassette as described previously (1), yielding F19::aadA-oriT-alpZ::scar.

(d) In-frame deletion of alpZ in S. ambofaciens. F19alpZ::aac(3)IV-oriT was first introduced in S. ambofaciens by intergeneric conjugation via the E. coli donor ET12567 containing pUZ8002, and exconjugants resistant to apramycin and sensitive to kanamycin (Apra^r and Kan^s, respectively) were retained. F19::aadA-oriT-alpZ::scar was then introduced in the Apra^r and Kan^s strains, and exconjugants resistant to spectinomycin were selected. Subsequent screening for the loss of both spectinomycin and apramycin resistance indicating the allelic exchange of the resistance marker with the unmarked, in-frame deletion was carried out. Replacement of alpZ in both chromosomal arms was obtained by natural homogenotization between the TIRs of the chromosome during fermentation of S. ambofaciens Spec^s and Apra^s exconjugants. Double alpZ deletion mutants (named alpZ) were confirmed by Southern blotting, PCR with CZ1/CZ2 primers (Table 2), and pulsed-field gel electrophoresis analysis.

(ii) In-frame deletion of alpIABCD. Using the DM1 strain in which alpIABCD biosynthetic genes are replaced by an aac(3)IV-oriT cassette (34), the same strategy as described above was used to remove the selectable cassette, leaving an 81-bp scar and yielding a alpID strain.

Complementation of the alpZ mutant. Using F19 cosmid as a template and a high-fidelity enzyme, a PCR product spanning from 528 bp upstream of the alpZ start codon to 15 bp downstream of the stop codon of alpZ was obtained with primers alpZ-compl-F/alpZ-compl-R (Table 2). After terminal dATP addition, the PCR product was first ligated with pGEM-Teasy vector (Promega). After transformation of E. coli DH5 with the ligation mixture, two white E. coli DH5 independent clones were confirmed by endonuclease restrictions to contain pGEM-Teasy including the insert. The insert part of the newly generated vectors from the two clones was further verified by sequencing to confirm the integrity of alpZ. The insert of one of them presented the wild-type sequence, and this vector was named pGEM-alpZ1, while in the other one, a single nucleotide substitution (G341 to A with respect to the first nucleotide of the start codon) was detected in the alpZ DNA coding sequence, resulting in a single amino acid change (Gly114 to Asp) in AlpZ. This construct, designated pGEM-alpZ2, was later used to analyze the effect of the mutation. The gel-purified XbaI/EcoRI restriction fragments of pGEM-alpZ1 and pGEM-alpZ2 which include alpZ were ligated into pSET152 vector (5) previously digested with the same enzymes, yielding pSET152-alpZ1 and pSET152-alpZ2, respectively. These integrative shuttle vectors were then introduced into the S. ambofaciens alpZ deletion strain by means of intergeneric conjugal transfer. For comparison, pSET152 was introduced into the wild-type and alpZ deletion strains.

Overexpression of alpZ. (i) Overexpression of alpZ using the thiostrepton inducible promoter tipAp. Using cosmid F19 as a template, the alpZ coding sequence was amplified by PCR with primers ExpZ1/ExpZ2 (Table 2), which include an NdeI and BamHI site, respectively. The PCR product (729 bp) was first cloned into pGEM-Teasy (Promega), yielding pGEM-alpZexp1. This vector was digested with NdeI and BamHI enzymes, and the fragment corresponding to the alpZ coding sequence (716 bp) was gel purified and ligated into pIJ8600 (43), previously digested with the same enzymes, yielding pIJ8600-alpZ.

(ii) Overexpression using the strong, constitutive promoter ermEp*. Using cosmid F19 as template, the alpZ coding sequence was amplified by PCR with primers ExpZ1/alpZ-compl-R (Table 2), which includes an NdeI and XbaI site in their respective sequences. The PCR product (743 bp) was first cloned into pGEM-Teasy (Promega), yielding pGEM-alpZexp2. This vector was digested with NdeI and XbaI enzymes, and the fragment corresponding to the alpZ coding sequence (731 bp) was gel purified and cloned into pIB139 (50) in which a typical Streptomyces ribosomal binding site sequence (AAAGGAGG) was previously inserted between the BamHI and NdeI sites of the MCS region, yielding pIB139-alpZ. The conjugative and integrative vectors pIJ8600-alpZ and pIB139-alpZ were checked for the integrity of the alpZ sequence and then introduced by conjugation into wild-type S. ambofaciens.

Antibiotic production assays and reverse-phase HPLC. Bioassays were carried out as previously described with Bacillus subtilis ATCC 6633 as an indicator strain using plugs from S. ambofaciens cultures grown in R2 agar or supernatant from cultures grown R2 liquid medium (34). High performance liquid chromatography (HPLC) for the detection of alpomycin was performed as previously described (34) with a Lichrosphere RP18 column (150- by 2-mm inner diameter; 5-µm particle size [Merck]).

Heterologous expression and purification of recombinant and mutant AlpZ. The alpZ coding sequence was isolated from pIJ8600-alpZ (Table 1) after a NdeI/BamHI digest and inserted into the expression vector pET-12a (Novagen) digested with the same enzymes, resulting in pET-alpZ (Table 1). The mutated alpZ coding sequence [alpZ(G341A)] was amplified using high-fidelity DNA polymerase from pGEM-alpZ2 (see above and Table 1) with the primer pair ExpZ1/ExpZ2 (Table 2). After terminal dATP addition, the PCR product was ligated with pGEM-Teasy (Promega) and checked by sequencing. After an NdeI/BamHI digest, the alpZ(G341A) fragment was inserted into pET-12a digested with the same enzymes, yielding pET-alpZ(G341A) (Table 1). The alpZ coding sequence contains rare translated arginine codons (AGA and AGG) in E. coli. For heterologous expression, competent E. coli BL21(DE3) cells (Novagen) were electroporated with pET-alpZ or pET-alpZ(G341A) along with the pSBET vector (40) that contains an arginine tRNA-encoding gene able to efficiently recognize AGA and AGG codons. E. coli BL21(DE3)/pSBET/pET-alpZ [or with pET-alpZ(G341A)] was grown in LB medium containing ampicillin (50 µg/ml) and kanamycin (50 µg/ml) at 37°C and 250 rpm until an optical density at 600 nm of 0.6 was reached. IPTG (isopropyl-β-D-thiogalactopyranoside; 0.1 mM) was added for induction. After a further 4-h incubation at 37°C, cells were collected, resuspended in TE buffer (30 mM Tris-HCl, 1 mM EDTA, pH 8) containing NaCl 200 mM, and then disrupted by sonication. After centrifugation (30 min at 16,000 rpm at 4°C), aliquots of crude protein extract containing AlpZ and AlpZ(G114D) were kept for use in gel retardation experiments. The soluble protein extract fraction containing AlpZ was precipitated with 40% (wt/vol) ammonium sulfate. The pellet containing the partially purified recombinant AlpZ was resuspended in TE buffer-NaCl and applied to a gel filtration column, Ultrogel AcA44 (Biosepra), previously equilibrated with the same buffer. Fractions containing AlpZ were combined and filtered through a YM10 membrane (cutoff of 10 kDa; Millipore) to remove salts. The residual extract was diluted in TE buffer, and proteins were separated on a DEAE Sepharose column (Pharmacia Biotech). Fractions corresponding to purified AlpZ were pooled and reconcentrated using a YM10 membrane. SDS-PAGE confirmed the purity of a predominantly single band with an apparent migration at 26 kDa.

Gel retardation assays. Labeling of the probes (PCR products or oligonucleotides), gel retardation assays, and chemiluminescence detection were carried out as described in the DIG Gel Shift kit, 2nd generation, following the manufacturer's instructions (Roche). Primer pairs AREU-1/AREU-2, AREXW-1/AREXW-2, AREV-1/AREV-2, and AREZ-1/AREZ-2 were used to amplify DNA probes alpUp (296 bp), alpXWp (219 bp), alpVp (386 bp), and alpZp (392 bp), respectively (Table 2; see also Fig. 4B). Primer pairs d-AREV-1/AREV-2, d-AREV-2/AREV-1, d-AREZ-1/AREZ-2, d-AREZ-2/AREZ-1, d-AREXW-1/AREXW-1 and d-AREXW-2/AREXW-2 were used in PCR amplification to obtain d-AREV1 (185 bp), d-AREV2 (201 bp), d-AREZ1 (213 bp), d-AREZ2 (180 bp), d-AREXW1 (172 bp), and d-AREXW2 (47 bp) probes, respectively (Table 2). The alpU-ARE, alpV-ARE, alpXW-ARE, and alpZ-ARE probes (40 bp including the ARE sites) were generated by annealing the complementary synthetic deoxyoligonucleotides, ZAREU1 and ZAREU2, ZAREV1 and ZAREV2, ZAREXW1 and ZAREXW2, and ZAREZ1 and ZAREZ2, respectively (Table 2). Probes (20 fmol) end labeled with 3' digoxigenin (DIG)-11-ddUTP were incubated at 30°C for 15 min with purified AlpZ protein (8 ng) or crude protein extract in binding buffer containing poly(dI-dC), according to the manufacturer's protocols (Roche). When necessary, S. ambofaciens extract (1 µl) or natural or synthetic -butyrolactones was added directly after the 15-min incubation period, and incubation was continued at 30°C for a further 10 min. Reaction mixtures were analyzed by native PAGE using 5% or 8% acrylamide buffered and run in 0.5x TBE (Tris-borate-EDTA) buffer (39). DNA was then transferred onto a positively charged nylon membrane (Amersham Hybond-N⁺) by electroblotting (400 mA for 30 min in 0.5x TBE buffer) using a Mini Trans-Blot Electrophoretic Transfer Cell (Bio-Rad). Light emission was recorded with a Fluor-S MultImager (Bio-Rad).

View larger version (45K): [in this window] [in a new window]   FIG. 4. DNA-binding activity of AlpZ. (A) Alignment of the putative AlpZ binding sequences. The consensus corresponds to the ARE consensus defined in Folcher et al. (13). The distance between the ARE sites and the start codon of the downstream gene is indicated. The letters W, Y, and K stand for A or T, C or T, and G or T, respectively. (B) Probes used in gel shift assays. The genetic organization of the alp regulatory subcluster is shown. Filled ovals symbolize the location of the ARE sequences. Sizes (bp) and names of the probes are indicated. (C) Gel shift assay using PCR probes including the respective alp-ARE motives. DIG-labeled probes (20 fmol) were incubated in the presence (+) or absence of purified recombinant AlpZ (rAlpZ; 8 ng). In competition experiments, excess of unlabeled probes (100x, 2 pmol) was added in the reaction mixture. (D) Gel retardation assay using synthetic short alp-ARE probes. DIG-labeled probes (20 fmol) were incubated in the presence (+) or absence of purified recombinant rAlpZ (8 ng). In competition experiments, excess of unlabeled probes (100x, 2 pmol) was added to the reaction mixture. In the assay using the alpU-ARE probe, increasing amounts of rAlpZ were tested: 1x, 8 ng; 5x, 40 ng; 10x, 80 ng. Open and filled triangles denote the migration of the probe alone and of probe-rAlpZ complex, respectively.

Antibiotic induction assays. Solvent extracts (1 µl) (see above) were applied directly on confluent lawns of S. ambofaciens or S. coelicolor spores spread on R2 or SMMS plates, respectively. Plates were incubated for 20 h to 40 h at 30°C.

	RESULTS

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Precocious production of orange pigment and alpomycin by alpZ disruption. To unravel the role of the potential -butyrolactone receptor protein encoded by alpZ in the regulation of the biosynthesis of alpomycin and orange pigment, the two copies of alpZ (one in each alp cluster present in each TIR) were deleted as described in Materials and Methods. The genomic DNA isolated from two independent mutants was verified by PCR and Southern analysis for the presence of the mutation in both TIRs and by pulsed-field gel electrophoresis to rule out any large genomic rearrangements (data not shown). The resulting strains lacking alpZ, designated the alpZ1 and alpZ2 strains, showed growth and morphological characteristics identical to those of the parent strain on SFM, HT, R2YE and R2 plates (data not shown) or in R2 liquid medium (Fig. 1D). The production of both the orange pigment and the antibacterial activity previously associated with the alp cluster was explored in more detail according to the appearance of these compounds during growth in/on R2 medium, a suitable medium for their production (34). Plugs from agar plates or supernatant from liquid-grown cultures of the mutant and parent strains were assessed during a time course for their ability to inhibit the growth of B. subtilis. The two mutant strains behaved identically, and only the alpZ1 strain is depicted in the figures. As shown in Fig. 1A, orange pigment production was readily detected in the mutant strain after 19 h of growth on agar plates, while the wild-type strain commenced production only after 36 h. In addition, plugs taken at 17 to 19 h from agar-grown mutant cultures exhibited inhibitory effects on the growth of B. subtilis whereas the same inhibitory effect was not observed for the parent strain until 35 to 36 h of growth (Fig. 1B). To ensure that the absence of alpZ in the alpZ strains was responsible for the observed precocious production of both orange pigment and alpomycin, complementation experiments were carried out by reintroducing a copy of alpZ under the control of its own promoter using plasmid pSET152-alpZ1 (see Materials and Methods). As controls, both the parent and the alpZ strains were transformed with the empty pSET152 vector. Compared to the wild-type strain harboring pSET152, the alpZ strains bearing the vector alone precociously synthesized pigment (data not shown) and antibiotic (Fig. 1C), in a manner similar to that of alpZ strains. The complemented alpZ strain containing pSET152-alpZ1 showed a production of orange pigment (data not shown) and bioactivity (Fig. 1C) over a time course comparable to that of the wild-type strain containing the pSET152 vector, confirming the involvement of alpZ in the regulation of orange pigment and alpomycin production. The precocious production of the antibiotic and pigment in the alpZ mutant strains was also confirmed during growth in R2 liquid cultures (Fig. 1D). Whereas production of both compounds in the S. ambofaciens wild-type strain occurred in the late transition phase or early stationary phase (at 21 h for the antibiotic and by 25 h at the latest for the pigment), that of the mutant strains occurred significantly earlier during the mid-exponential phase (i.e., at 14 h) (Fig. 1E). Intriguingly, a second round of antibacterial activity production was detected in the alpZ strains after 36 h on agar-grown cultures (Fig. 1B) and around 25 h in liquid cultures (Fig. 1E). The presence of alpomycin in these early and late antibiotic production periods was confirmed by HPLC analyses (data not shown). The decrease in detectable antibacterial activity between 31 to 35 h in agar (Fig. 1B) and 19 to 21 h in liquid-grown cultures (Fig. 1E) presumably results from both a stop in the production of alpomycin (which is consistent with the reduction in transcript level of the biosynthetic gene, alpA (see below and also Fig. 3A), and a concomitant degradation or modification of precociously produced alpomycin (e.g., in orange pigment).

View larger version (39K): [in this window] [in a new window]   FIG. 1. Precocious production of orange pigment and antibiotic by the alpZ deletion strain. (A) Pigment production on R2 agar for the wild-type and alpZ knockout (alpZ1) strains after 19 h of growth at 30°C. (B) Antibiotic activity of wild type and the alpZ1 strain against B. subtilis ATCC 6633 (Bio) and orange pigment production (Pig) was followed through a time course on R2 agar; +, pigmented; –, not pigmented. Agar plugs of mycelia taken at various times were placed on an LB plate spread with B. subtilis spores. The antibiotic activity is visualized by the absence of B. subtilis growth corresponding to the dark halo surrounding the plug. (C) The pSET152 derivative pSET152-alpZ1 was used to complement the alpZ mutant. Antibiotic activity (see above) of the complemented strain (alpZ1/pSET152-alpZ1) was compared to that of the control wild type (WT/pSET152) and alpZ mutant strains (alpZ1/pSET152) carrying the empty pSET152 vector. (D) Growth curves of the wild-type and the alpZ1 strains in R2 liquid cultures. At time points 1 to 10 (indicated by the arrow), total RNAs were isolated for use in the transcriptional studies shown in Fig. 3. (E) The antibiotic activity and orange pigment production at these ten time points are shown. The orange pigment was directly visualized in the spectrophotometer cuvettes. WT, wild type.

View larger version (49K): [in this window] [in a new window]   FIG. 3. Comparative transcriptional analyses of alp genes by RT-PCR. (A) Transcripts of hrdB (major and essential sigma factor gene; internal control), alpA (β-ketoacyl synthase gene involved in both orange pigment and alpomycin biosynthesis), alpR (the second alp-located β-ketoacyl synthase gene not involved in orange pigment and alpomycin biosynthesis), alpV (essential SARP activator gene), alpZ (repressor gene), alpW (deduced repressor gene), alpT (SARP gene), and alpU (SARP gene) from wild type (WT) and alpZ mutant (alpZ1) strains were analyzed by RT-PCR with 25 cycles of PCR on cDNA generated from RNA isolated from late exponential phase to early stationary phase (time points 1 to 10) (Fig. 1D). Assessed bioactivity (Bio) against B. subtilis at time points 1 to 10 (Fig. 1E) are indicated. +, detected; –, not detected. (B) To detect low-level transcript of alpV in the wild type at 17 h, 28 cycles of PCR were applied.

View larger version (34K): [in this window] [in a new window]   FIG. 2. Effect of the alpZ overexpression on pigment and antibiotic production. (A) The wild-type strain harboring pIJ8600 (control) or pIJ8600-alpZ was streaked on R2 agar in the absence (–Tsr) or presence of 25 µg/ml thiostrepton (+Tsr) to induce alpZ expression. Plates were photographed from below after 4 days of growth at 30°C. (B) The wild-type strain harboring pIB139 (control) or pIB139-alpZ (constitutive expression of alpZ) was streaked on R2 agar. Plates were photographed from below after 3 days of growth at 30°C (panel 1). Antibiotic production was tested on B. subtilis with agar plugs of mycelia taken from the above plates (panel 2).

In contrast to the four other alp regulatory genes, transcripts of alpZ were readily detectable even at early time points in the wild-type strain before the onset of alpomycin and orange pigment. This is consistent with the idea that the product of alpZ is needed during the early stages of growth to repress the production of alpomycin.

Another β-ketoacyl synthase present within the alp cluster and encoded by alpR has previously been shown to be dispensable for the synthesis of either the orange pigment or alpomycin (34). However, analysis of alpR expression by RT-PCR showed a pattern of transcription similar to that of alpA in both the wild-type and the alpZ mutant backgrounds (Fig. 3A). It thus seems that expression of alpR (and presumptively that of its operonic partner alpQ, encoding a chain length factor subunit) is efficiently regulated by AlpZ, perhaps through regulation of AlpV. The AlpZ regulon may thus include another product whose structure and role have yet to be properly defined.

DNA-binding by recombinant AlpZ. Candidate ARE-like sequences to which -butyrolactone receptor proteins typically bind (13) are present in the promoter regions of alpZ, alpV, and alpU (1). A similar ARE-like sequence was also detected in the promoter region of the putative alpX-alpW operon (Fig. 4A). In sum, the possible AlpZ binding sites are therefore located 60 bp upstream of the alpZ coding sequence, 90 bp upstream of alpV, 40 bp upstream of alpX/alpW, and 205 bp upstream of alpU, which is found divergent from alpT (Fig. 4B). To determine whether the -butyrolactone receptor-like protein AlpZ truly targets these promoter regions, gel mobility shift assays were performed using DNA fragments encompassing the ARE-like sequences (ranging from 219 bp to 392 bp and designated alpZp, alpVp, alpXWp, and alpUp) in the presence of purified recombinant AlpZ protein (Fig. 4B) (see Materials and Methods). Shifts were clearly observed with the probes alpVp, alpZp, and alpXWp but never with alpUp, demonstrating that AlpZ directly binds to its own promoter region and that of alpV and alpXW but was unable to bind to the promoter region of alpU (Fig. 4C). The ARE-like sequence upstream of alpU is, however, the least conserved compared to the consensus sequence. Competition with unlabeled probe DNA showed a release of the shift, verifying the specificity of the observed AlpZ interaction with the promoter regions of alpV, alpZ, and alpXW (Fig. 4C). In addition, use of labeled DNA fragments containing only one-half of the predicted alpVp, alpZp, or alpXWp ARE sequences in the gel shift assays never resulted in a shift, indicating that binding of AlpZ to these promoter regions depends upon the presence of the complete ARE motifs (data not shown). To confirm this hypothesis, shorter probes (40-bp deoxyoligonucleotides) bearing one or the other of the four alp-ARE sequences (26 bp) were used in further mobility shift assays. Formation of an AlpZ-DNA complex was observed for the alpZ-ARE, alpV-ARE, and alpXW-ARE probes but not with alpU-ARE even when increased concentrations of purified AlpZ were used (Fig. 4D). Altogether, these results demonstrate that AlpZ binds to a specific DNA sequence that resembles the ARE motifs recognized by typical -butyrolactone receptors.

Influence of solvent-extractable material on DNA-binding activity of AlpZ. In the context of AlpZ belonging to the autoregulator receptor family, the derepression of the alpomycin production might be alleviated by -butyrolactone-type molecules. Several members of the autoregulator receptor family have been shown to respond to small signaling molecules (the -butyrolactones) that are extractable from the aqueous fermentation broth by the use of organic solvents. In addition, these molecules were able in gel retardation experiments to disrupt binding of the respective -butyrolactone receptor from their cognate DNA targets.

To determine whether similar signaling molecules were produced by S. ambofaciens, samples of supernatant from R2 liquid-grown cultures taken over a time course (Fig. 5A) were extracted with ethyl acetate, and the processed extracts were assayed by means of gel retardation experiments for their ability to disrupt DNA-bound AlpZ from the probe alpXWp (Fig. 5B). Loss of the formation of a shifted probe complex indicated the presence of a factor capable of disrupting the AlpZ-DNA interaction. This activity was detectable in cultures from early transition phase (at 17 h), 2 h before the onset of alpomycin production, up to about 24 h into stationary phase. Under these culture conditions, the absence of detectable amounts of the disrupting activity in late-stationary phase coincided with the cessation of alpomycin production. Control extracts from uninoculated R2 medium showed no activity in the assay. Extracts that disrupted the binding of AlpZ to the probe alpXWp also disrupted binding of AlpZ with the probes alpZp and alpVp and with the short probes alpZ-ARE, alpV-ARE, and alpXW-ARE (data not shown). AlpZ-DNA complex formation was also disrupted using ethyl acetate extracts obtained from the S. ambofaciens mutant strain alpID (Table 1 and Materials and Methods) (data not shown). This strain lacks the minimal PKS (alpABC) involved in the synthesis of the core polyketide compound, thus indicating that the active factor is not derived from the alpomycin-biosynthetic pathway.

View larger version (31K): [in this window] [in a new window]   FIG. 5. Production of autoregulator molecules by S. ambofaciens. (A) Growth curve of wild-type S. ambofaciens in R2 medium. Growth was followed via optical density (OD) measurement at 600 nm. Production of alpomycin, tested with the standard bioassay (see Materials and Methods), is shown by the arrow. Time points 1 to 10 at which a supernatant sample was extracted with ethyl acetate are indicated by black squares. (B) Gel shift assay to address the presence of AlpZ-disrupting material in S. ambofaciens extracts. The DIG-labeled probe alpXWp (20 fmol) was incubated with 8 ng of recombinant AlpZ (rAlpZ) before addition of extracted material from time points 1 to 10 (as described in panel A). R2 denotes the (negative) control (noninoculated R2 medium extracted under the same conditions). Time points at which alpomycin was detected by bioassay are depicted by the arrows. Open and filled triangles denote the migration of the probe alone and of probe-rAlpZ complex, respectively.

View larger version (94K): [in this window] [in a new window]   FIG. 6. Induction of orange pigment and antibiotic production by the AlpZ-interactive ligand(s). Extracts (1 µl) from supernatant culture of the S. ambofaciens mutant alpID strain, which lacks the biosynthetic genes responsible for alpomycin synthesis (34), and from noninoculated R2 medium (negative control) were applied (at the x sign) on a confluent lawn of S. ambofaciens wild-type spores. After 20 h of incubation at 30°C, an agar plug of mycelia (symbolized by the open circle) taken in the vicinity of the dropped extracts was tested for the production of bioactivity against B. subtilis.

View larger version (59K): [in this window] [in a new window]   FIG. 7. Effect of a single amino acid substitution (Gly114 to Asp) in AlpZ. (A) Orange pigment production in R2 liquid cultures of the wild type and the alpZ1 strains carrying pSET152 (5) or pSET152-alpZ2 (Table 2) which contains the mutated sequence of alpZ. Fermentation flasks were photographed after 5 days of growth at 30°C under 250 rpm constant shaking. (B) Gel retardation assay comparing the DNA-binding activity and sensitivity to the AlpZ-interactive ligand of the native AlpZ and the mutated AlpZ(G114D) (AlpZ*). The DIG-labeled probe alpXWp (see Fig. 4) was incubated with a limiting amount of AlpZ and AlpZ(G114D) (i.e., the minimal quantity of protein required to fully retard the probe) before addition (+) of ligand extract. Lane 1, standard extract; lane 2, 10-fold dilution; lane 3, 20-fold dilution. Open and filled triangles denote the migration of the probe alone and of probe-recombinant AlpZ complex, respectively. WT, wild type.

Properties of the AlpZ-interactive ligand(s). In order to assess the nature of the AlpZ-interactive ligand, we examined the ability of several known natural and synthetic -butyrolactone molecules belonging to the three defined subclasses of -butyrolactones to prevent shifting of DNA probes by AlpZ. The -butyrolactones tested included A-factor from S. griseus, virginiae butanolides (and derivatives) from S. virginiae, SCB1 from S. coelicolor, and IM-2 of S. lavendulae (kindly provided by T. Nihira and S. Kitani). None of them was able to significantly disrupt DNA-bound AlpZ (data not shown), suggesting either a strong specificity of AlpZ for its cognate ligand or a molecule with a significantly different structure.

Members of the -butyrolactone family have previously been shown to be resistant to proteases and to heat and acidic treatments but sensitive to treatment with alkali to pH 12. The instability in alkali is believed to result from opening of the lactone ring, rendering the molecules inactive (14, 47). In this study, the active compound in the S. ambofaciens cultures was surprisingly found to be resistant to all treatments, including alkali (Fig. 8). The solvent extract from noninoculated R2 medium that received the same treatments failed to disrupt binding of AlpZ from DNA probes (data not shown). In addition, a control was done with extract from the S. coelicolor M145 strain treated in the same way. After the alkali treatment, the S. coelicolor M145 extract failed to induce pigmented antibiotic production in the producing strain compared to the untreated extract, proving that the alkali treatment efficiently inactivated -butyrolactone molecules (data not shown). It therefore appears that the AlpZ-interacting ligand(s) present in S. ambofaciens is structurally different from typical -butyrolactones and lacks the lactone ring. This could readily explain the absence of the complementation effect by the various investigated -butyrolactone molecules.

View larger version (58K): [in this window] [in a new window]   FIG. 8. AlpZ-interactive ligand properties. Gel shift experiments using the DIG-labeled probe alpXWp (20 fmol) and recombinant AlpZ (rAlpZ; 8 ng) were carried out with ligand extracts exposed to various treatments. NT, not treated; prot K, proteinase K; protease, proteases from S. griseus; heat, 100°C for 30 min; acidic, pH 1.2 for 30 min; alkali, pH 12 for 10 min. R2 corresponds to extract from noninoculated R2 medium.

	DISCUSSION

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

View larger version (26K): [in this window] [in a new window]   FIG. 9. Regulation model of the alp regulatory cascade. In the early stages of growth (panel 1), the product of alpZ represses the transcription of both alpW and alpV. The absence of AlpV activator prevents expression of the alp biosynthetic genes. When AlpZ-interactive ligands are produced (panel 2), AlpZ is dissociated from its DNA targets, which correspond to specific sequences in the promoter region of alpV and alpW, allowing transcription of the latter genes. In turn, AlpV triggers the expression of the biosynthetic genes leading to the production of alpomycin and orange pigment. Derepression of alpW transcription could result in accumulation of the autoregulator repressor AlpW that has been shown in preliminary gel shift assays to bind within the promoter region of alpV. Thus, AlpW could help to block alpV transcription in the late stages of growth, thereby stopping production of alpomycin and orange pigment (panel 3). Dilution or degradation of AlpZ ligands may also restore the original repression state by AlpZ.

Nature of AlpZ by comparison with CprB crystal structure and analysis of the effect of Gly114 replacement by Asp in AlpZ(G114D). Study of AlpZ(G114D), which was still able to bind DNA but was insensitive to its cognate ligand, strengthened the role of AlpZ as a repressor. On the basis of the crystal structure analysis of the related repressor CprB of S. coelicolor (30), the overall tertiary structure of AlpZ(G114D) was not modified compared to AlpZ (analysis using the DeepView program of the Swiss-PdbViewer; http://www.expasy.org/spdbv/). CprB was shown to be composed of two domains, a DNA-binding domain (residues 1 to 52) and a regulatory domain (residues 77 to 215). According to the amino acid sequence alignment between available -butyrolactone receptor-like proteins (including CprB) and AlpZ, Gly114 is located in the regulatory domain and, more precisely, between helices 6 and 7 (that are involved in the ligand binding pocket formation) in a weakly conserved region among these related proteins. Nevertheless, hydrophobic residues (which are mostly not well conserved) within the ligand binding pocket seem to play an important role in the recognition of the ligand in terms of specificity. The mutation may thus impair either the recognition or the proper binding of the ligand by AlpZ(G114D). As the shift of helix 6 upon ligand binding is predicted to induce the relocation of helix 4 from the DNA-binding domain (30), it is also conceivable that the substitution may affect the normal conformational change leading to the relocation of the DNA-binding domain and subsequently the dissociation of the repressor from DNA. In both hypotheses, the nonconservative mutation of Gly114 (hydrophobic) to Asp (polar) renders impossible the release of AlpZ(G114D) to its DNA targets, which, in turn, permanently blocks alpomycin production.

A new type of small signaling molecule produced by S. ambofaciens. Since the material extracted from S. ambofaciens before and during alpomycin production was able to induce antibiotic production in young wild-type S. ambofaciens mycelia and to disrupt DNA-binding ability of AlpZ, a -butyrolactone autoregulator receptor homolog, it seemed evident, a priori, that this material could belong to the -butyrolactone group of small signaling molecules that are widespread in streptomycetes. However, in this study, several different natural and synthetic -butyrolactone molecules belonging to the three described classes (according to the length and branching of the C-2 acyl side chain and stereochemistry at C-6) were tested in gel shift assays for their ability to release AlpZ from its DNA target, but none of them could significantly dissociate DNA-bound AlpZ. This may indicate a strict specificity of the AlpZ-ligand(s) for AlpZ but could also reflect a molecular structure that is distinct from that of the other butyrolactones. Indeed, typical -butyrolactones lose activity following alkali treatment, but the S. ambofaciens extractable material was insensitive to this treatment, suggesting that the alp autoinducer lacks the lactone group characteristic of the -butyrolactones (although we cannot totally exclude the possibility that a small but sufficient amount of active material can survive this treatment and interact with AlpZ). To our knowledge, our results support the first demonstration that a member of the so-called -butyrolactone autoregulator receptor family could interact with molecules different from -butyrolactones. This finding should extend our overview on the diversity of autoregulatory systems. Further studies aimed at purifying and characterizing the AlpZ ligand(s) are currently under way.

	ACKNOWLEDGMENTS

 
R.B. and M.V.M. were supported by a grant of the European Community under the ActinoGEN 6th framework program (FP6-5224).

We thank Takuya Nihira and Shigeru Kitani (Osaka University, Japan) for inviting B.A. to work with them and for providing us with -butyrolactones used in this work. We thank Gregory Challis, Lijiang Song, and Christophe Corre (University of Warwick, United Kingdom) for alpomycin structure determination. We also thank Jean-Michel Girardet (University Henri Poincaré, France) for help with HPLC analyses and Peter Leadlay (University of Cambridge, United Kingdom) for the kind gift of pIB139. We are grateful to Andrew Hesketh (John Innes Center, United Kingdom) for critical reading of the manuscript.

	FOOTNOTES

 
* Corresponding author. Mailing address: Laboratoire de Génétique et Microbiologie, Faculté des Sciences et Techniques, Nancy Université, Boulevard des Aiguillettes, BP239, 54506 Vandœuvre-lès-Nancy, France. Phone: 33 3 83 68 42 05. Fax: 33 3 83 68 44 99. E-mail: Bertrand.Aigle{at}scbiol.uhp-nancy.fr

Published ahead of print on 22 February 2008.

Present address: IBMC-Instituto de Biologia Molecular e Celular, Universidade do Porto. Rua do Campo Alegre, 823 Porto, Portugal.

Present address: Department of Microbiology and Immunology, Center for Pulmonary and Infectious Disease Control, University of Texas Health Center, 11937 US Highway 271, Tyler, TX.

	REFERENCES

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Aigle, B., X. Pang, B. Decaris, and P. Leblond. 2005. Involvement of AlpV, a new member of the Streptomyces antibiotic regulatory protein family, in regulation of the duplicated type II polyketide synthase alp gene cluster in Streptomyces ambofaciens. J. Bacteriol. 187:2491-2500.[Abstract/Free Full Text]
2. Barona-Gomez, F., S. Lautru, F. X. Francou, P. Leblond, J. L. Pernodet, and G. L. Challis. 2006. Multiple biosynthetic and uptake systems mediate siderophore-dependent iron acquisition in Streptomyces coelicolor A3(2) and Streptomyces ambofaciens ATCC 23877. Microbiology 152:3355-3366.[Abstract/Free Full Text]
3. Bate, N., A. R. Butler, A. R. Gandecha, and E. Cundliffe. 1999. Multiple regulatory genes in the tylosin biosynthetic cluster of Streptomyces fradiae. Chem. Biol. 6:617-624.[CrossRef][Medline]
4. Bentley, S. D., K. F. Chater, A. M. Cerdeno-Tarraga, G. L. Challis, N. R. Thomson, K. D. James, D. E. Harris, M. A. Quail, H. Kieser, D. Harper, A. Bateman, S. Brown, G. Chandra, C. W. Chen, M. Collins, A. Cronin, A. Fraser, A. Goble, J. Hidalgo, T. Hornsby, S. Howarth, C. H. Huang, T. Kieser, L. Larke, L. Murphy, K. Oliver, S. O'Neil, E. Rabbinowitsch, M. A. Rajandream, K. Rutherford, S. Rutter, K. Seeger, D. Saunders, S. Sharp, R. Squares, S. Squares, K. Taylor, T. Warren, A. Wietzorrek, J. Woodward, B. G. Barrell, J. Parkhill, and D. A. Hopwood. 2002. Complete genome sequence of the model actinomycete Streptomyces coelicolor A3(2). Nature 417:141-147.[CrossRef][Medline]
5. Bierman, M., R. Logan, K. O'Brien, E. T. Seno, R. N. Rao, and B. E. Schoner. 1992. Plasmid cloning vectors for the conjugal transfer of DNA from Escherichia coli to Streptomyces spp. Gene. 116:43-49.[CrossRef][Medline]
6. Bignell, D. R., N. Bate, and E. Cundliffe. 2007. Regulation of tylosin production: role of a TylP-interactive ligand. Mol. Microbiol. 63:838-847.[Medline]
7. Buttner, M. J., K. F. Chater, and M. J. Bibb. 1990. Cloning, disruption, and transcriptional analysis of three RNA polymerase sigma factor genes of Streptomyces coelicolor A3(2). J. Bacteriol. 172:3367-3378.[Abstract/Free Full Text]
8. Challis, G. L., and D. A. Hopwood. 2003. Synergy and contingency as driving forces for the evolution of multiple secondary metabolite production by Streptomyces species. Proc. Natl. Acad. Sci. USA 100(Suppl. 2):14555-14561.[Abstract/Free Full Text]
9. Choi, S. U., C. K. Lee, Y. I. Hwang, H. Kinoshita, and T. Nihira. 2004. Cloning and functional analysis by gene disruption