INTRODUCTION

Top Abstract Introduction Materials & Methods Results & Discussion References

Streptomycetes are gram-positive bacteria characterized by their versatile ability to produce secondary metabolites and by their morphological complexity (1, 6). Both or either of these phenotypes are controlled in some Streptomyces species by low-molecular-weight compounds called butyrolactone autoregulators (24), and the 10 butyrolactone autoregulators isolated to date have been classified into three types according to structural differences in their C-2 side chains: (i) the virginiae butanolide (VB) type possesses a 6--hydroxy group, as exemplified by VB-A~E of Streptomyces virginiae (3, 14, 22, 27, 28), which controls virginiamycin production; (ii) the IM-2 type possesses a 6--hydroxy group, as exemplified by IM-2 of Streptomyces sp. strain FRI-5 (17, 24, 30), which controls the production of a blue pigment and nucleoside antibiotics; and (iii) the A-factor type possesses a 6-keto group, as exemplified by A-factor of Streptomyces griseus (8, 13, 18). Although the structural differences among these autoregulators are small, producer strains show a high degree of ligand specificity toward the corresponding autoregulator type, indicating the presence of receptor proteins of strict ligand specificity (5, 16, 20).

The VB-specific binding protein (BarA) of S. virginiae was purified, and the gene encoding it (barA) was cloned and characterized in our laboratory (21). The N-terminal region of BarA has been predicted to form a helix-turn-helix DNA-binding motif, and in vitro analyses using recombinant BarA have revealed that BarA binds to specific DNA sequences in the absence of VB and dissociates from the DNA by binding with VB (12), suggesting that BarA should function as a transcriptional regulator, the DNA-binding activity of which is controlled by VB. Although BarA was the only logical candidate as the mediator of VB signal because we detected no other VB binding protein during the purification of BarA, it was less clear whether the VB-BarA pathway was actually involved in the control of virginiamycin production.

In this study, to assess the in vivo function of BarA, two kinds of barA mutants were constructed by homologous recombination between the wild-type barA gene on the chromosome and the mutated barA gene on a plasmid. Phenotypic and biochemical analyses of the mutants provided the first in vivo evidence that the VB-BarA pathway participates not only in virginiamycin production but also in autoregulator biosynthesis.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results & Discussion References

Strains, growth conditions, and plasmids. S. virginiae (strain MAFF 10-06014; National Food Research Institute, Ministry of Agriculture, Forestry, and Fisheries, Tukuba, Japan) was grown at 28°C as described previously (10, 28). For genetic manipulation in Escherichia coli and Streptomyces, E. coli DH5 (4) and Streptomyces lividans TK21 (7) were used. Streptomyces strains were grown at 28°C in yeast extract-malt extract (YEME) liquid medium for preparation of protoplasts (7), in tryptic soy broth (TSB) (Oxoid, Basingstoke, Hampshire, United Kingdom) for preparation of total DNA, and on ISP no. 2 (Difco, Detroit, Mich.) for spore formation. S. lividans TK21 was obtained from D. A. Hopwood (John Innes Centre, Norwich, United Kingdom).

Chemicals. All chemicals were of reagent or high-performance liquid chromatography grade and were purchased from either Nacalai Tesque (Osaka, Japan), Takara Shuzo (Shiga, Japan) or Wako Pure Chemical Industrial (Osaka, Japan). Virginiamycin M₁ and S were purified as described previously (15). Authentic VB was synthesized as described previously (20).

Construction of vectors for gene replacement. A 6.2-kbp PstI fragment (12) ranging from 3.4 kbp upstream to 2.1 kbp downstream of barA was subcloned into pUC18 to generate pBA1 (see Fig. 1A). To delete an internal 99-bp EcoT14I fragment encoding from Lys⁵¹ to Ser⁸³ of BarA, pBA1 was digested with XbaI and EcoT14I and then a 2.8-kbp EcoT14I fragment containing the upstream fragment and the 5' 154-bp fragment of the barA coding sequence was inserted. The construct was confirmed to have the desired deletion by DNA sequencing. From the resulting plasmid, the EcoRI-HindIII fragment was subcloned into the EcoRI-HindIII-digested pIJ486 (26) to generate pBAD22. The plasmid pBAD22 was used to generate a barA mutant strain, NH1.

DNA manipulation. DNA manipulations in E. coli and Streptomyces were performed by the methods of Sambrook et al. (23) and Hopwood et al. (7), respectively. Protoplast formation and transformation of S. virginiae were performed by the methods of Kawachi et al. (9).

Southern blot analysis. Three micrograms of digested DNA were loaded on each lane, electrophoresed on a 1% agarose gel, and transferred to Hybond-N+ (Amersham, Little Chalfont, Buckinghamshire, United Kingdom) according to the manufacturer's recommendations. Membranes were prehybridized for 1 h at 65°C and hybridized for 18 h. The probe used was a 0.9-kbp AgeI fragment containing the barA gene and labeled with [-³²P]dCTP by using a Random Primer DNA Labeling Kit, version 2 (Takara Shuzo). Membranes were then washed thoroughly in 0.1× SSC (1× SSC consists of 0.015 M sodium citrate and 0.15 M NaCl [pH 7.7]) containing 0.1% sodium dodecyl sulfate (SDS) at 65°C. Autoradiography was performed by using Kodak X-Omat films at 80°C, with intensifying screens, for 1 to 5 h.

Western blot analysis. SDS-polyacrylamide gel electrophoresis (SDS-PAGE) was performed on 10 to 20% polyacrylamide precast gradient gels (Daiichi Pure Chemicals, Tokyo, Japan). Crude extracts containing 50 µg of protein were loaded. After transfer to an Immobilon-P^SQ transfer membrane (Millipore, Bedford, Mass.), the proteins were immunodetected with rabbit antiserum raised against the recombinant BarA protein expressed in E. coli (21), using an enhanced chemiluminescense (ECL) kit from Amersham according to the manufacturer's instructions. Marker proteins for SDS-PAGE were purchased from New England Biolabs (Beverly, Mass.). Densitometric analysis was performed with a Shimadzu densitometer (model CS-9300PC).

Determination of VB and virginiamycin. The amounts of virginiamycins produced were determined by a bioassay using Bacillus subtilis PCI219 as an indicator strain (29). Virginiamycin is a mixture of two chemically different compounds, virginiamycins M₁ and S, showing synergistic antibiotic activity. Because the ratio between them will affect apparent antibiotic activity, we also analyzed the ratio between virginiamycins M₁ and S by C₁₈ reverse-phase high-performance liquid chromatography as described previously (25) and confirmed that the ratio was unchanged for strain NH1 and the wild-type strain.

VB binding assay. VB binding activity was routinely assayed by the ammonium sulfate precipitation method (11) with [³H]VB-C₇ (54.6 Ci/mmol) in the presence and absence of 2,000-fold cold VB-C₆.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials & Methods Results & Discussion References

Construction of barA mutant strain NH1. To determine the in vivo function of BarA in S. virginiae, wild-type barA in the chromosome was replaced with a mutated barA by using pBAD22 (Fig. 1A). The modified barA gene on pBAD22 (for details, see Materials and Methods) lost a 99-bp EcoT14I fragment, which resulted in the in-frame deletion of 33 amino acid residues containing the second helix region of the estimated helix-turn-helix motif (Fig. 1B). After S. virginiae MAFF 10-06014 was transformed using pBAD22, pBAD22-integrated strains from the first crossover, as by route a in Fig. 1A, were selected among thiostrepton-resistant transformants. After single colony isolation in the presence of thiostrepton (5 µg/ml), Southern blot hybridization was performed to confirm the integration of pBAD22 (Fig. 1C, lane 2), and no plasmid form of pBAD22 was detected (data not shown). To facilitate the second crossover, the plasmid-integrated strain was put through two rounds of cultivation on liquid TSB medium lacking thiostrepton. As expected, two types of thiostrepton-sensitive colonies were obtained: namely, barA mutants (in which the wild-type barA gene was replaced with the altered sequence) and regenerated wild-type strains (Fig. 1C, lanes 3 and 4). One of these mutants, designated strain NH1, was used for further investigation.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Gene replacement of S. virginiae barA gene with mutated barA by homologous recombination. (A) Restriction maps of pBAD22 and the chromosomal barA region of wild-type S. virginiae and of the pBAD22-integrated strain and barA mutant strain NH1. A single crossover between pBAD22 and a homologous DNA in the chromosome (such as via route a) gave the pBAD22-integrated strain. A loss of the plasmid sequence by the second crossover generated strain NH1. Filled bars represent regions from S. virginiae DNA. Thick arrows indicate the location and orientation of barA, and open bars represent the estimated helix-turn-helix motif inside barA. tsr is the thiostrepton resistance gene. Abbreviations: ER, EcoRI; E, EcoT14I; H, HindIII; P, PstI; S, SacII; X, XbaI. (B) Schematic representation of wild-type BarA and the mutated BarA of strain NH1. Filled bars indicate the barA coding region. Amino acid residues for the estimated helix-turn-helix motif and those encoded by the 99-bp EcoT14I fragment (small capitals) are shown below the filled bars. The numbers above the filled bars indicate the total number of amino acid residues (a.a.) for the wild-type BarA and mutated BarA. (C) Hybridization pattern of DNA. SacII-digested total DNAs from the S. virginiae wild-type strain (lane 1), pBAD22-integrated strain (lane 2), barA mutant strain NH1 (lane 3), and wild-type segregant (lane 4) were used. A 0.9-kbp AgeI fragment containing the barA gene was used as a probe. A single 2.4-kbp SacII band, rather than two SacII bands (0.5 and 2.0 kbp), was detected in strain NH1, because of the deletion of the EcoT14I fragment containing a SacII site. M, marker DNAs.

Polypeptides encoded by mutated barA in strain NH1. To confirm that the mutated barA gene was expressed in strain NH1, we performed a Western blot analysis with an antibody raised against recombinant BarA (Fig. 2). A protein band of 25 kDa was detected in strain NH1 (Fig. 2, lanes 3 to 6). Although the detected band had a lower electrophoretic mobility than the molecular mass (21.5 kDa) predicted from the DNA sequence, it was concluded to be the mutated BarA protein, because BarA protein tends to migrate more slowly on SDS-PAGE than its actual molecular mass (21), as evident from the behavior of wild-type BarA (lanes 1 and 2, apparent molecular mass, 28 kDa; actual molecular mass, 25.0 kDa).

View larger version (83K): [in this window] [in a new window]   Fig. 2.   Western blot analysis of the mutated BarA protein from strain NH1. Crude cell extracts containing 50 µg of protein were loaded and analyzed by anti-BarA antibody. Lane 1, purified recombinant BarA from E. coli; lane 2, wild-type strain harvested after 12 h of cultivation; lanes 3 to 6, strain NH1 harvested at 6, 8, 10, and 12 h of cultivation. The top arrow to the right of the gel indicates the position of wild-type BarA, and the bottom arrow indicates the position of mutated BarA.

Several phenotypes of the strain NH1. Virginiamycin production by strain NH1 was examined by bioassay using B. subtilis (Fig. 3). While a wild-type segregant as well as the wild-type parental strain began to produce virginiamycin after 12 h of cultivation, virginiamycin production by strain NH1 began much earlier (6 h of cultivation), suggesting that the BarA protein acts as repressor in virginiamycin production. No difference in growth was observed between the NH1 strain and the wild-type strain (data not shown). In the wild-type strain, it has been shown that artificial addition of VB at various times prior to the natural production of VB (after 4 to 10 h of cultivation) induces earlier production of virginiamycin (31). Therefore, it is conceivable that repression by native BarA is released by VB binding, which would lead to virginiamycin production in the wild-type strain, while the mutated BarA could not exert repression either due to the lack of helix-turn-helix motif or the small amount of protein. Virginiamycin is a mixture of two chemically distinct compounds, virginiamycins M₁ and S, showing strong synergistic antibiotic activity. Because the two components were produced by strain NH1 with the same ratio to that by the wild-type strain (data not shown), BarA can be concluded to coordinately control the two biosynthesis pathways. However, the amount of virginiamycin produced by strain NH1 was about 10% of that produced by the wild-type strain, which agreed well with our previous observation that VB addition at the beginning of cultivation reduced virginiamycin production to the same extent in the wild-type strain (31). Therefore, it seems that derepression or lack of repression by BarA from the beginning of cultivation caused a lower production of virginiamycin, although the underlying mechanism requires further study.

View larger version (11K): [in this window] [in a new window]   Fig. 3.   Time course of virginiamycin production in two strains of S. virginiae. Wild-type S. virginiae () and barA mutant strain NH1 () were studied. The amount of VB produced by the wild-type strain is also shown (). Strain NH1 did not produce any VB.

Construction of barA mutant strain NH2. We constructed another type of barA mutant by using pGM122 (for details, see Materials and Methods) in which a 1.0-kbp neomycin resistance gene was inserted 239 bp downstream of the helix-turn-helix motif of barA (Fig. 4A). As in the case of strain NH1, a pGM122-integrated strain derived from the first crossover was selected from the colonies resistant to both neomycin (200 µg/ml) and thiostrepton (5 µg/ml). After cultivating the plasmid-integrated strain for two rounds on liquid TSB medium plus only neomycin (200 µg/ml), we selected colonies that were both neomycin resistant and thiostrepton sensitive to obtain barA mutants derived from the second crossover. Wild-type segregants were obtained by cultivating the plasmid-integrated strain in the absence of antibiotics and selecting the neomycin-thiostrepton-sensitive strains. The genomic structure of the representative strains from each crossover step was analyzed by Southern blot hybridization (Fig. 4B), and one of the neomycin-resistant, thiostrepton-sensitive strains, NH2, was found to have the expected increase in length of the BamHI fragment from 2.8 to 3.7 kbp (Fig. 4B, lane 3). With respect to the morphological phenotypes on solid media, strain NH2 was identical to the wild-type strain, further confirming that the VB-BarA pathway does not participate in morphological control in S. virginiae.

View larger version (16K): [in this window] [in a new window]   View larger version (32K): [in this window] [in a new window]   Fig. 4.   Gene replacement of S. virginiae barA gene with mutated barA by homologous recombination. (A) Restriction maps of pGM122 and the chromosomal barA region of wild-type S. virginiae and of the pGM122-integrated strain and barA mutant strain NH2. A single crossover between pGM122 and a homologous DNA in the chromosome gave the pGM122-integrated strain. A loss of the plasmid sequence by the second crossover generated strain NH2. Filled bars represent regions from S. virginiae DNA. Thick arrows indicate the location and orientation of barA, and open bars represent the estimated helix-turn-helix motif inside barA. Open bars containing an arrow indicate the location and orientation of the neomycin resistance gene. tsr is the thiostrepton resistance gene. Abbreviations: B, BamHI; EN, EcoNI; S, SalI; Sm, SmaI. (B) Hybridization pattern of DNA. BamHI-digested total DNAs from the S. virginiae wild-type strain (lane 1), pGM122-integrated strain (lane 2), barA mutant strain NH2 (lane 3), and wild-type segregant (lane 4) were used. The AgeI fragment containing the barA gene was used as a probe. M, marker DNAs.

Several phenotypes of strain NH2. While the wild-type segregant and the wild-type parental strain produced virginiamycin in similar amounts, neither strain NH2 nor the plasmid-integrated strain produced virginiamycin (Table 1), indicating that this version of mutated barA was dominant over the wild-type barA and that the presence of the mutated barA caused the inhibition of virginiamycin production. Although the mutated BarA protein was scarcely visible in crude extracts of strain NH2 by anti-BarA antibody, probably due to the deletion of a major epitope in the mutated BarA (unpublished data), the dominant negative phenotype of strain NH2 suggested that sufficient mutated BarA should be present in the cell to allow complete repression of virginiamycin production, even in the presence of VB. The results of the VB binding assay revealed that strain NH2 was deficient in VB binding activity (Table 1), which may suggest that the C-terminal deletion of BarA severely impaired VB binding activity. In marked contrast to the virginiamycin production and VB binding activity, strain NH2 retained its ability to produce VB, while strain NH1 completely lost this ability (Table 1). Because the main difference between the two constructs in the two strains is that the intact helix-turn-helix motif is present in the former and absent in the latter, this motif of BarA can be considered important in the control of VB biosynthesis. Assuming that the major function of BarA is exerted by binding to specific DNA sequences in order to repress target genes in the absence of VB, as indicated from our previous in vitro study (12), VB biosynthesis seems to require an unidentified ene (gene X), the repression of which by intact BarA may be essential for VB biosynthesis. Alternatively, if BarA is assumed as a dual-function regulator acting both as a repressor and an activator, BarA may act as an activator for VB biosynthesis.