Introduction

The γ-butyrolactone autoregulators found in the genus Streptomyces control the production of secondary metabolites and/or morphological differentiation. The effectiveness of these autoregulators, which are active at nanomolar concentrations, as well as the presence of the specific receptor proteins as mediators of autoregulator signaling, implies that these γ-butyrolactone autoregulators should be regarded as Streptomyces hormones. From in vitro studies these autoregulator receptors are dimeric DNA-binding proteins that, in the absence of autoregulators, recognize and bind to the specific DNA sequences located in the promoter region of target genes (Kinoshita et al. 1999; Kitani et al. 1999; Takano et al. 2001). Autoregulator-binding to the corresponding receptors causes the receptor to dissociate from the DNA, which in turn allows transcription of the target genes.

From the in vivo study of actinomycetes, the disruption of autoregulator receptor genes from Kitasatospora setae and Streptomyces lavendulae FRI-5 resulted in the overproduction of secondary metabolites (bafilomycin in K. setae, and nucleoside antibiotics in S. lavendulae FRI-5), while no apparent effect was observed on growth or morphological differentiation, indicating that the corresponding autoregulator receptors only act as negative regulators on the biosynthesis of secondary metabolites. Such non-involvement of autoregulator receptors in morphological differentiation has also been reported for S. virginiae (Nakano et al. 1998) and S. coelicolor A3(2) (Takano et al. 2001); but not for S. griseus and S. natalensis, in which disruption of the autoregulator receptor gene resulted in 10- and 4.6-fold higher productions of antibiotic, as well as earlier sporulation and 10-fold more abundant spore production, respectively (Onaka et al. 1995; Lee et al. 2005).

In the present study, a gene encoding an autoregulator receptor from S. ambofaciens producing spiramycin, a commercially important macrolide antibiotic consisting of a 16-membered lactone ring and three amino-sugar residues, was cloned by PCR using primers designed for the two highly conserved regions of Streptomyces autoregulator receptors, along with genomic Southern hybridization. The in vivo function of the autoregulator receptor in S. ambofaciens was also identified by phenotypic comparison between the receptor gene-disruptants and the wild-type.

Materials and methods

Bacterial strains, plasmids, and growth conditions

Streptomyces ambofaciens ATCC15154 was grown at 28°C in a HT medium (g/l, 1 g yeast extract, 1 g beef extract, 2 g NZ-amine A, 10 g dextrin, 20 g agar) for preparation of spores. E. coli strain XL10-Gold (Stratagene, UK) was used as the general cloning host. pUC19 was used to construct a genomic library and for DNA sequencing. For the seed culture the preparation, spores of S. ambofaciens (5 × 106) were inoculated into 20 ml of vegetative medium (Ford et al. 1990), and then incubated at 28°C for 24 h on a rotary shaker (180 rpm). The main cultivation was performed by inoculating 2.1 ml of seed culture into 70 ml vegetative medium at 28°C for 132 h.

DNA manipulation and sequencing

The degenerate primers F (5′-CGCGGATCCSGCNGCGGCCNNGGTSTTCGA-3′) and R (5′-CGCGGATCCGTGGAANTASANSGCNCCCTT-3′) were used to clone an autoregulator receptor gene from S. ambofaciens (Fig. 1). Southern blot hybridization was performed using a DIG High Prime DNA Labeling and Detection Starter Kit II (Roche Co., Penzberg, Germany) according to the protocol supplied by the manufacturer. DNA sequencing was performed on a fluorescence DNA sequencer (ALFred; Amersham Pharmacia Biotech, USA) on double-stranded templates derived from different clones in pUC19, using the dideoxy-chain termination method with a thermosequenase cycle sequencing kit (Amersham Pharmacia Biotech, USA) and Cy5-labelled primers. A partial genomic library was constructed with size-fractionated EcoRI fragments (4.5 kb) and pUC19, using E. coli XL10-Gold as a host, and screened by colony hybridization with a 100-bp PCR fragment.

Fig. 1
figure 1

Alignment of overall amino acid sequences (a), and phylogenetic tree (b) of BarA, ArpA, FarA, ScbR, SngR, TarA, and SaaR. The alignment and the phylogenetic tree were created by Genetyx software (GENETYX CO., Tokyo, Japan). Identical amino acids are indicated by black boxes. Inverted arrows represent the regions of designed primers. BarA, VB-specific receptor (BarA) of S. virginiae; ArpA, A-factor receptor (ArpA) of S. griseus; FarA, IM-2 specific receptor (FarA) of S. lavendulae FRI-5; ScbR, SCB1-receptor (ScbR) of S. coelicolor A3(2); SngR, receptor of S. natalensis; tarA, receptor of S. tendae ; SaaR, receptor of S. ambofaciens (this study)

Full size image

Construction of a saaR disrupted strain

To construct ΔsaaR, the cloned 4.5-kb fragment was digested with Eco81I to remove a 714-bp saaR fragment (Fig. 2), and then ligated. The entire 3.8-kb insert containing ΔsaaR was recovered and inserted into the EcoRI sites of pKC1132, a homologous recombination vector containing oriT of RK2 and an apramycin resistance gene for selection in actinomycetes and E. coli (Bierman et al. 1992), to give pMK201 (Fig. 3a). E. coli ET12567/pUZ8002 transformed with pMK201 was conjugated with S. ambofaciens. Exconjugants, in which the plasmid pMK201 had integrated at the saaR locus by a single crossover via homologous recombination, were selected with apramycin. After three rounds of incubation at 28°C on HT medium in the absence of apramycin, putative saaR-disrupted strains formed from the second crossover were detected by their apramycin sensitivity. Because all the saaR-disrupted strains showed identical behavior, such as morphology and growth, one of the strains was chosen for detailed analyses (strain MK1) (Fig. 3a).

Fig. 2
figure 2

Nucleotide and deduced amino acid (one-letter notation) sequences of saaR in the cloned 4.5-kb EcoRI fragment. An asterisk denotes a translational stop codon. Location of putative −10 and −35 sequences are boxed. A putative 26-bp receptor-binding sequence is indicated by the light gray box. A probable ribosome-binding sequence, ggagg, is present six nucleotides upstream of the putative ATG start codon (under dotted line). The regions amplified by the initial PCR with the designed primers are shown by the underline

Full size image
Fig. 3
figure 3

Gene replacement of the S. ambofaciens saaR gene with deleted saaR by homologous recombination. (a) Schematic representation of the strategy used for the disruption of saaR. The light gray arrows indicate the location and orientation of saaR. (b) Southern hybridization analysis of PstI-digested chromosomal DNA from a S. ambofaciens wild-type strain (lane 1) and a saaR-disruptant, strain MK1 (lane 2). The probe used was the 0.8-kb SalI-Eco81I fragment

Full size image

Morphological assessment and analysis of spiramycin

To analyze morphological differentiation, 106 spores each of the wild-type strain and the saaR-disrupted strain KM101 were grown on HT, MS (Hobbs et al. 1989) and ISP2 medium, respectively, and were cultivated at 28°C for 7 days. After making the spore suspension, spore counting was done by plating the serially diluted spore suspension. Values represented the average from three independent experiments, and the error rate was less than 5%. Growth in the liquid culture was measured as the OD600. To assess spiramycin production in the culture broths, the seed culture and the main culture were performed as described above. The broth of the main culture was centrifuged at 13,000g for 10 min at 4°C to remove mycelia, and the supernatant was used for bioassay on nutrient agar (g/l, 3 g beef extract, 5 g peptone, 15 g agar, pH 7.0), with Micrococcus luteus as a test organism. The clear-zone was measured after incubation for 20 h at 30°C. Commercial spiramycin (Sigma) was used as a standard.

Nucleotide sequence accession number

The nucleotide sequence data reported in this paper will appear in the DDBJ/EMBL/GenBank nucleotide sequence databases with the accession number AB362564.

Results and discussion

Cloning and sequencing of saaR

To search for a receptor gene from S. ambofaciens, primers were designed for the two highly conserved amino acid s4 autoregulator receptors [BarA of S. virginiae (Okamoto et al. 1995), FarA of S. lavendulae (Waki et al. 1997), ScbR of S. coelicolor A3(2) (Takano et al. 2001), and ArpA of S. griseus (Onaka et al. 1995)] with codon usage data derived from 64 Streptomyces genes (Fig. 1) (Wright and Bibb 1992). A 100-bp PCR product clearly encoding the targeted region of a putative autoregulator receptor was obtained, and a 4.5-kb EcoRI fragment was cloned using the PCR fragment as a probe, yielding pMK101, as described in the ‘Materials and methods’ (Fig. 3a).

When the cloned 4.5-kb fragment was sequenced, three complete open reading frames (ORFs) (orf1 to orf3) were identified. In addition, it appeared that the fragment cloned from S. ambofaciens ATCC15154 identified to that of S. ambofaciens ATCC23877, including a putative γ-butyrolactone-binding protein (accession number AAR30170), by means of sequence matching in GenBank detected by BLAST. The orf1 consisting of 714 bp is predicted to encode a 237-amino acid protein, which shows high similarity to several autoregulator receptor proteins such as FarA of S. lavendulae FRI-5 (50% identity, 63% similarity), ScbR of S. coelicolor A3(2) (48% identity, 66% similarity), ArpA of S. griseus (43% identity, 62% similarity), and BarA of S. virginiae (40% identity, 61% similarity). The orf2 and orf3 product identified to a putative acetyltransferase (accession number CAI78116) and a SimX2-like protein (accession number AAR30169) of S. ambofaciens ATCC23877, respectively. In addition, the Orf1 has a well-conserved helix-turn-helix (HTH) DNA binding motif (Fig. 1), and its estimated pI value is 5.3, which agreed well with pI values of ∼5 for all Streptomyces autoregulator receptors (Choi et al. 2004). Therefore, this orf1 was named saaR (S treptomyces a mbofaciens γ-butyrolactone-autoregulator receptor), and predicted to encode a real autoregulator receptor protein.

In Streptomyces, autoregulator receptor proteins usually bind to the promoter region of the receptor gene, and regulate their own transcription depending on the presence of autoregulators (Kinoshita et al. 1999; Kitani et al. 1999; Takano et al. 2001). To estimate whether similar regulation might operate in saaR, a putative receptor-binding sequence was identified in the 5′-upstream region of saaR using the consensus-binding sequence of Thompson et al. (TNANAWACNNACYNNNCGGTTTKTTT) (Folcher et al. 2001). A 26-bp sequence (GAAAATACGGACTCCCTGGTTTTGTT) was found at 61–86 bp upstream of the saaR initiation codon, which was localized between the putative −10 and −35, namely, typical transcriptional promoter sequences (Fig. 2), suggesting that the transcription of saaR is likely to be autoregulated via SaaR protein.

Disruption of the saaR gene and phenotypic analyses

To determine the in vivo function of SaaR in S. ambofaciens, the chromosomal saaR gene was disrupted as described in the ‘Materials and methods’, resulting in a saaR-disrupted strain (ΔsaaR, strain MK1) (Fig. 3a). As shown in Fig. 3b, it was confirmed by Southern blot analysis that the PstI fragment of strain MK1 was 0.7-kb shorter than that of the wild-type strain because a 737-bp fragment, including the whole saaR (714 bp), was removed by Eco81I and then self-ligated from S. ambofaciens chromosome.

To observe the morphological differentiation, the morphological characteristics of the wild-type strain and the strain MK1 were carefully compared on solid media in order to clarify whether saaR is involved in the morphological differentiation of S. ambofaciens. When the strains were grown on HT agar medium, the number of strain MK1 spores decreased 4-fold in comparison to the wild-type strain (HT agar 6.4 × 107 spores/cm2 for the wild-type strain; 1.6 × 107 spores/cm2 for MK1) (Fig. 4). A similar result was also obtained from the MS and ISP2 agar (MS agar 6.9 × 107 spores/cm2 and ISP2 agar 2.9 × 107 spores/cm2 for the wild-type strain; MS agar 1.8 × 107 spores/cm2 and ISP2 agar 6.6 × 106 spores/cm2 for MK1). In addition, although the growth of strain MK1 in liquid medium was more rapid than that of the wild-type strain, saaR deletion from the S. ambofaciens chromosome resulted in complete loss of spiramycin production (Fig. 5). Therefore, these results indicate that saaR plays an important role as a rare positive regulator, controlling both sporulation and antibiotic production in S. ambofaciens.

Fig. 4
figure 4

Effect of saaR-disruption on morphological differentiation. Spores (1 × 106) of each strain were plated on HT agar, followed by incubation for 7 days at 28°C

Full size image
Fig. 5
figure 5

Growth curves (a) and natamycin production (b) in liquid culture of a wild-type strain (solid circles) and a saaR-disruptant (strain MK1; open circles). Three independent saaR-disruptants showed identical patterns, and the representative data from strains MK1 are shown

Full size image