- Email: [email protected]
threonine kinases in Streptomyces coelicolor A3(2)
Gene 334 (2004) 53 – 61 www.elsevier.com/locate/gene
Phosphorylation of AfsR by multiple serine/threonine kinases in Streptomyces coelicolor A3(2) Reiko Sawai, Ayano Suzuki, Yuji Takano, Ping-Chin Lee 1, Sueharu Horinouchi * Department of Biotechnology, Graduate School of Agriculture and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo, Tokyo 113-8657, Japan Received 11 November 2003; received in revised form 22 January 2004; accepted 13 February 2004 Available online 5 May 2004 Received by T. Sekiya
Abstract AfsK, a protein serine/threonine kinase, autophosphorylates on serine and threonine residues and phosphorylates serine and threonine residues of AfsR, a transcriptional activator for afsS involved in secondary metabolism in Streptomyces coelicolor A3(2). pkaG encoding a 592-amino-acid protein and SCD10.09 (named afsL) encoding a 580-amino-acid protein, both of which encode an AfsK-like protein, were transcribed throughout growth. PkaG with a histidine-tag and the kinase catalytic domain of PkaG, produced in Escherichia coli, autophosphorylated dominantly on threonine and slightly on serine residues. In addition, these proteins phosphorylated AfsR on threonine and serine residues. The catalytic domain of AfsL also autophosphorylated and phosphorylated AfsR, on threonine and serine residues in both cases. AfsR was thus found to be phosphorylated by multiple kinases. Disruption of the chromosomal pkaG gene resulted in slightly reduced production of the pigmented antibiotic actinorhodin. These findings, together with the presence of about 40 AfsK homologues and at least five AfsR homologues in S. coelicolor A3(2), suggest that the regulatory networks via eukaryotic-type protein phosphorylation are more diverse and versatile than we have expected. D 2004 Elsevier B.V. All rights reserved. Keywords: Signal transduction; Antibiotic production; Transcriptional factor; Actinorhodin
1. Introduction A number of proteins in the Gram-positive, filamentous bacterial genus Streptomyces are phosphorylated on their serine/threonine and tyrosine residues in response to developmental phases (Umeyama et al., 2002; Horinouchi, 2003). AfsK, the first serine/threonine kinase found in Streptomyces coelicolor A3(2), autophosphorylates on serine and threonine residues and phosphorylates AfsR, a transcriptional factor with ATPase activity, which activates afsS encoding
Abbreviations: a.a.; amino acid(s); bp; base pair(s); PCR; polymerase chain reaction. * Corresponding author. Tel.: +81-3-5841-5123; fax: +81-3-58418021. E-mail address: [email protected] (S. Horinouchi). 1 Present address: School of Science and Technology, University of Malaysia Sabah, Kota Kinabalu, Sabah, Malaysia. 0378-1119/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.gene.2004.02.046
a small protein involved in secondary metabolism (Lee et al., 2002). Disruption of afsK, afsR, or afsS results in reduced production of actinorhodin, undecylprodigiosin, and a calcium-dependent antibiotic (Floriano and Bibb, 1996; Umeyama et al., 2002), showing that these genes comprise a linear signal transduction system for secondary metabolite formation. In addition to the linear signal relay, AfsR is phosphorylated on its serine and threonine residues by a kinase(s) other than AfsK, because AfsR is phosphorylated by a cell lysate prepared from an afsK null mutant (Matsumoto et al., 1994). Thus, multiple protein kinases, including AfsK, have been suggested to phosphorylate AfsR. By analogy to various protein kinases in eukaryotic cells, each of the kinases may autophosphorylate on its serine/threonine residues on sensing some environmental stimulus and transfer the signal by means of phosphorylating AfsR. We therefore started to identify an additional kinase(s) able to phosphorylate AfsR in S. coelicolor A3(2) and to determine its enzymatic properties. For this purpose, we used the
54
R. Sawai et al. / Gene 334 (2004) 53–61
genome sequence of this strain (http://www.sanger.ac.uk/ Projects/S_coelicolor) (Bentley et al., 2002). In this paper, we describe that AfsR is a substrate of multiple kinases. When we think of the possibility that these kinases sense some environmental signals and physiological conditions, AfsR may serve as an integrator of such signals. We also discuss a complex regulatory network involving multiple serine/threonine kinases for growth, morphological development, and secondary metabolism.
2. Materials and methods 2.1. Bacterial strains, plasmids, and media Escherichia coli JM109 and pUC19 used for DNA manipulation were purchased from Takara Shuzo. E. coli BL21(DE3) and pET32a(+) used for production of histidinetagged proteins were purchased from Novagen. Media and growth conditions for E. coli were described by Maniatis et al. (1982). S. coelicolor A3(2) was routinely cultured at 30 jC on Trypto – Soya broth (TSB) (Nissui), which was supplemented with 10 Ag/ml of thiostrepton or kanamycin when necessary. 2.2. General recombinant DNA techniques Restriction enzymes, T4 DNA ligase, and other DNAmodifying enzymes were purchased from Takara Shuzo. [a-32P]dCTP (110 TBq/mmol) for DNA labeling with the Takara BcaBest DNA labeling system and [g-32P]ATP (220 TBq/mmol) for end labeling at 5V ends with T4 polynucleotide kinase were purchased from Amersham Biotech. DNA was manipulated in E. coli (Maniatis et al., 1982) and in S. coelicolor A3(2) (Hopwood et al., 1985), as described. Nucleotide sequences were determined with the Thermo Sequence fluorescence-labeled primer cycle sequencing kit (Amersham) on an automated DNA sequencer. 2.3. Cloning and expression of pkaG and afsL The pkaG-coding sequence was amplified by polymerase chain reaction (PCR) with two primers, 5VGCCGAATTC CATCTGATGTCAGCGCTGGAGCCGGGAC-3V (The underline and the boldface letters indicate an EcoRI site and the start codon of pkaG, respectively) and 5V-GCCAAGCTT GGATCCCTCGAGTCAGCCGAGTTGCGCGAGCTGCGCG-3V (the underline and the boldface letters indicate a HindIII site and the stop codon of pkaG, respectively) and cloned between the EcoRI and HindIII sites of pUC19, resulting in pUC19pkaG. After no undesired alterations in PCR had been checked by nucleotide sequencing, the EcoRI – HindIII fragment was inserted between the EcoRI and HindIII sites of pET32DTRX, in which the thioredoxin-coding
region was removed from pET32a(+) by NdeI digestion (Umeyama and Horinouchi, 2001), resulting in pET32pkaG that would express histidine-tagged PkaG (H-PkaG). For production of PkaGDC, plasmid pET32-pkaGDC was constructed by digestion of pET32-pkaG with NotI and recirculization. H-PkaG and PkaGDC were produced in E. coli BL21(DE3) and purified with His-bind resin, as recommended by the manufacturer. Plasmid pET32-afsLDC used for production of the kinase domain of AfsL was constructed, as follows. A 5V region of the afsL-coding sequence was amplified by PCR with two primers, 5V-GCCGAATTCATGAATCAACACGGCCG-3V (the underline and the boldface letters indicate an EcoRI site and the start codon of afsL) and 5V-GCCAAGCTT GACGGTGCCCGGTGCAAG-3V (the underline indicates a HindIII site; a stop codon was included in the vector sequence) and inserted between the EcoRI and HindIII sites of pUC19, resulting in pUC19-afsLDC. After the afsLDC-coding sequence had been checked for errors in amplification, it was inserted between the EcoRI and HindIII sites of pET32DTRX to generate pET32-afsLDC. AfsLDC was purified with Hisbind resin. 2.4. Phosphorylation protocols and phosphoamino acid analysis For autophosphorylation of PkaG, PkaGDC, and AfsLDC, the standard reaction mixture, containing 15 pmol each of the kinases in 10 mM Tris – HCl (pH 7.2) – 10 mM MnCl2 – 10 mM MgCl2 – 0.1 mM ATP – 10 ACi (370 kBq) [g-32P]ATP – 1 mM dithiothreitol in a total volume of 20 Al, was incubated at 30 jC for 15 min. For phosphorylation of H-AfsR by these kinases, 30 pmol of H-AfsR was added. The reaction was terminated by boiling for 2 min after adding 4 Al of a dye solution. The stopped reaction mixture was fractionated by SDS-polyacrylamide gel electrophoresis. Phosphorylated proteins were transferred to a polyvinylidene difluoride (PVDF) membrane and detected by autoradiography. Phosphorylated a.a. residues were determined by phosphoamino acid analysis by cellulose thin-layer chromatography (Cooper et al., 1983; Kamps and Sefton, 1989). 2.5. S1 nuclease mapping of pkaG and afsL Methods for RNA isolation and S1 mapping were as described by Kelemen et al. (1998). 32P-labeled probes were prepared by PCR with a pair of 32P-labeled and non-labeled primers (see Fig. 2C and D). 2.6. Disruption of pkaG For disruption of the chromosomal pkaG, an about 800base pair (bp) region upstream of pkaG was amplified by P C R w i t h p r i m e r s , 5 V- G C C GAATTC T C G A C G -
R. Sawai et al. / Gene 334 (2004) 53–61
TACCGGCGCACCGTGCG-3V (the underline indicates an EcoRI site) and 5V-GCCCTGCAG CTTGGCGTGCGGCCCTCCCTAGA-3V (the underline indicates a PstI site). An about 800-bp region downstream of pkaG was also amplified with 5V-GCCCTGCAG CGGCAACTCATCAGCGTTCCGCA-3V (the underline indicates a PstI site) and 5V-GCCAAGCTT CGGCTGGTCTCGGTCAACT-3V (the underline indicates a HindIII site). These fragments, together with the kanamycin resistance gene from Tn5 (Beck et al., 1982), were cloned in pUC19, resulting in a plasmid used for disruption of the chromosomal pkaG gene (see Fig. 4A). After the plasmid had been linearized by digestion with DraI and introduced by transformation (Oh and Chater, 1997) into S. coelicolor A3(2) M130, candidates containing correct disruption as a result of double crossover were obtained. Among kanamycin (10 Ag/ml)-resistant colonies, three true disruptants were found by means of Southern hybridization using the upstream region of pkaG and the kanamycin resistance gene as 32P-labeled probes.
Table 1 Amino acid identity between AfsK and putative serine/threonine kinases in S. coelicolor A3(2) Name
No. of a. a. residues
Identity with AfsKa 1 – 300 a.a.
301 – 500 a.a.
AfsK PkaG AfsL (SCD10.09) SC66T3.32c SCE7.11 SC1G2.06c SCL11.07 SCD69.08 SCD69.01/SCD65.24 PksC SC7A12.07 PkaI SCH69.18 SCD35.14 SCI11.13 PkaA PkaH PkaF SCGD3.21c PkaD PkaB PkaE SCD63.30 SCK13.03 PkaJ SC5H1.01/SCF8.01c SC2A6.02c SCD63.08 SCC24.21 SC2H4.01/SC9G1.09 RamC
799 592 580 783 720 686 493 626 632 556 745 380 673 586 550 543 717 667 522 599 417 487 576 670 548 538 452 979 1349 1261 930
– 50.3 47.9 47.8 45.4 44.2 41.5 39.6 39.1 34.7 33.3 32.3 32.0 31.8 31.3 30.7 30.6 30.4 30.3 30.0 29.8 29.7 29.4 28.3 27.8 27.4 27.1 24.2 20.8 20.6 18.0
– 25.3 20.4 23.8 23.7 19.5 23.3 27.4 8.1 20.5 24.6 20.6 20.3 52.3 27.8 24.5 27.1 17.9 23.6 26.0 22.6 24.0 25.7 24.6 23.1 29.2 20.9 26.5 31.6 19.6 20.5
a
The N-terminal regions of 300 a.a. containing the catalytic domain (1 – 300 a.a.) and the next 200 a.a. (301 – 500 a.a.) are compared with those of AfsK.
55
3. Results
3.1. Putative serine/threonine kinases in S. coelicolor A3(2) The presence of more than 40 proteins having characteristic sequence motifs, termed Hanks’ consensus sequences (Hanks et al., 1988), in S. coelicolor A3(2) have been predicted. Table 1 lists 30 such proteins, together with amino acid (a.a.) identity to the kinase catalytic domain of AfsK. All these kinases have a catalytic domain of serine/threonine or tyrosine kinases at their N-terminal portions, and the identity of their Cterminal parts is low. We chose three genes, pkaG (encoding a 592 a.a. protein; having a catalytic domain with 50% a.a. identity to that of AfsK), SCD10.09 (580 a.a.; 48% identity; later named afsL), and SCD66T3.32c (783 a.a.; 48% identity), to determine whether these genes are actually transcribed and whether these putative serine/threonine kinases autophosphorylate and phosphorylate AfsR. These genes show the highest similarity in amino acid sequence to AfsK. Although pkaG was previously cloned and its nucleotide sequence was determined (Hirakata et al., 1998; Ogawara et al., 1999), no further study has been reported. We succeeded in production of proteins PkaG and SCD10.09 in E. coli, but failed to produce protein SCD66T3.32c because of its instability in E. coli. Therefore, we analyzed PkaG and SCD10.09 (AfsL) in this study. The a.a. sequences of the kinase catalytic domains of these proteins are shown in Fig. 1. Hanks’ consensus sequences, numbered I to XI (Hanks et al., 1988), are well conserved. 3.2. Transcription of pkaG and afsL We first examined the transcription of the two genes by S1 nuclease mapping. RNA was prepared from cells grown at 30 jC for various periods in TSB liquid medium. The 32P-labeled probes were prepared by PCR with a pair of 32P-labeled and non-labeled primers, shown in Fig. 2C and D. High-resolution S1 mapping with the RNA prepared from the cells grown for 12 h identified a single transcriptional start point for each of the two genes (Fig. 2A and B). The transcriptional start points of afsL and pkaG were 34 and 36 nucleotides, respectively, upstream of the translational start codons. In front of the transcriptional start point of pkaG, a TATGCT sequence, very similar to a typical 10 sequence TATAAT found in Streptomyces and other bacteria (Strohl, 1992), is present. However, no sequence similar to a 35 consensus sequence, TTGACA for many bacteria and TTGACR (R: A or G) for housekeeping genes in Streptomyces (Strohl, 1992), is present at an appropriate position. In front of the transcriptional start point of afsL, no sequences similar to 10 and 35 consensus sequences are present.
56 R. Sawai et al. / Gene 334 (2004) 53–61 Fig. 1. Amino acid alignment of the kinase catalytic domains of AfsK and the two kinases, PkaG and AfsL, used in this study. Conserved subdomains I to XI are taken from Hanks et al. (1988). The a.a. residues well conserved in eukaryotic serine/threonine kinases (Hanks et al., 1988) are shown under the alignment. The a.a. identical between two of the three proteins are shown in white letters in black boxes.
R. Sawai et al. / Gene 334 (2004) 53–61
57
We also determined the course of transcription of these genes by low-resolution S1 mapping (Fig. 2E). hrdB encoding a major sigma factor, which is transcribed throughout growth, was used to monitor the quantity and quality of the RNA used, as described previously (Umeyama and Horinouchi, 2001). Both genes were transcribed throughout growth, just like hrdB. Under these conditions, actinorhodin and undecylprodigiosin were detectable at 36 h. The courses of transcription of these genes are the same as that of afsK, which is known to regulate actinorhodin production by phosphorylating AfsR (Lee et al., 2002). 3.3. Characterization of pkaG
Fig. 2. Transcription of pkaG and afsL. RNA isolation from mycelium grown at 30 jC for 12 h on solid medium and S1 mapping were described previously (Umeyama and Horinouchi, 2001). (A and B) High-resolution S1 mapping for determination of the transcriptional start points of pkaG and afsL with 32P-labeled probes prepared by PCR with primers shown in C and D. The positions of the S1-protected fragments are shown by arrowheads, and the transcriptional start points are assigned to the residues as shown after corrections for DNA/RNA hybrids (Sollner-Webb and Reeder, 1979). (C and D) Nucleotide sequences of the promoter regions of pkaG and afsL. The transcriptional start points are shown by arrows. The primers used for preparing 32P-labeled probes are also shown. (E) Low-resolution S1 mapping of pkaG and afsL. The RNA and 32P-labeled probes were the same as those used for high-resolution S1 mapping.
Plasmid pET32-pkaG would direct the synthesis of a protein Met-His6-S tag-PkaG (Met-1 to Gly-592) (H-PkaG; 632 a.a., 69.5 kDa). The histidine tag and S tag are derived from the pET32 vector. E. coli BL21(DE3) harboring pET32-pkaG was grown and the cell extract was analyzed by SDS-polyacrylamide gel electrophoresis. The amounts of H-PkaG produced both in the soluble and in the insoluble fractions were only small (data not shown), probably because H-PkaG was unstable and readily digested by E. coli proteases. This is also true for AfsK (Umeyama and Horinouchi, 2001). We purified H-PkaG from the soluble fraction with His-bind resin and assayed its ability to autophosphorylate, as described previously (Umeyama and Horinouchi, 2001). H-PkaG thus purified showed distinct activity to autophosphorylate dominantly on threonine and slightly on serine residues (Fig. 3C). Our attempt to phosphorylate histidine-tagged AfsR (H-AfsR) (Umeyama and Horinouchi, 2001) was rather difficult because of the instability and weak activity of PkaG. The kinase catalytic domain (Met-1 to Arg-311) of AfsK can autophosphorylate and phosphorylate AfsR in the same way as AfsK with the full length (Umeyama and Horinouchi, 2001). We then produced PkaGDC, in which a Cterminal portion from Gly-338 to Gly-592 was deleted. PkaGDC, with a structure of Met-His6-S tag-PkaG (Met-1 to Gly-337)-15 a.a. (406 a.a., 45 kDa; the C-terminal 15 a.a. is derived from the vector sequence), was produced stably and purified from the soluble fraction of E. coli harboring pET32-pkaGDC, giving a single major protein band on SDS-polyacrylamide gel electrophoresis (Fig. 3A). Incubation of PkaGDC with [g-32P]ATP yielded an autophosphorylated form (Fig. 3B) on threonine and serine residues, as determined by phosphoamino acid analysis by cellulose thin-layer chromatography (Fig. 3C). Addition of H-AfsR of 110 kDa to the incubation mixture yielded a phosphorylated form of H-AfsR at the position corresponding to 110 kDa, in addition to the autophosphorylated form of PkaGDC (Fig. 3B). H-AfsR was purified as described previously (Umeyama and Horinouchi, 2001) and gave a single major band on SDS-polyacrylamide gel electrophoresis. Phosphoamino acid analysis showed that
58
R. Sawai et al. / Gene 334 (2004) 53–61
Fig. 3. Characterization of PkaG and AfsL. (A) SDS-polyacrylamide gel electrophoresis of the cell lysates prepared from E. coli harboring pET32-pkaGDC or pET32-afsLDC and the purified PkaGDC and AfsLDC used for phosphorylation assays. The soluble fraction (lane 1) and insoluble fraction (lane 2) from E. coli harboring pET32-pkaGDC are shown, together with PkaGDC purified with His-bind resin (lane 3) and molecular size markers (lane M). The soluble fraction (lane 4) of E. coli harboring pET32-afsLDC and purified AfsLDC (lane 5) are also shown. A considerable amount of PkaGDC of 45 kDa was produced as inclusion bodies in the insoluble fraction. AfsLDC of about 50 kDa was purified as a mixture of two proteins, the smaller one of which was perhaps generated by digestion with an E. coli protease. (B) Autoradiograms showing autophosphorylation of PkaGDC and AfsLDC and phosphorylation of AfsR by them. Incubation of PkaGDC or AfsLDC with [g-32P]ATP yielded an autophosphorylated form of 45 or 50 kDa, respectively. Addition of H-AfsR to the reaction mixture yielded a phosphorylated form of H-AfsR of 110 kDa. Incubation of H-AfsR alone in the presence of [g-32P]ATP, as a negative control, yielded no phosphorylated form. (C) Phosphoamino acid analyses of the autophosphorylated forms of PkaGDC and AfsLDC and the H-AfsR protein phosphorylated by them. The autophosphorylated form of PkaGDC contains phosphorylated threonine and serine residues and the H-AfsR phosphorylated by PkaGDC also contains phosphorylated threonine and serine residues (left panel). PkaG with the full length, purified from E. coli harboring pET32-pkaG, also autophosphorylates on threonine and serine residues. AfsLDC was found to autophosphorylate on threonine and serine residues and to phosphorylate threonine and serine residues of H-AfsR (right panel).
the phosphorylated residues of AfsR were threonine and serine (Fig. 3C). We therefore conclude that PkaGDC, and PkaG itself, too, autophosphorylate on threonine and serine residues and phosphorylate threonine and serine residues of AfsR. Although PkaGDC was produced mostly in the soluble fraction, the autophosphorylating activity varied depending on the protein preparation, which hampered us to determine the kinetic parameters for autophosphorylation and phosphorylation of AfsR. This was also true for AfsLDC (see below). 3.4. Characterization of afsl Plasmid pET32-afsLDC would direct the synthesis of a protein Met-His6-S tag-AfsL (Met-1 to Val-338)-13 a.a. (AfsLDC; 399 a.a., 44 kDa). The C-terminal 13 a.a. is derived form the vector sequence. AfsLDC gave two protein bands on SDS-polyacrylamide gel electrophoresis (Fig. 3A). Since the histidine-tag was attached to the N-terminal end of the fused protein, the smaller protein was supposedly generated by a cleavage at its C-terminal region by a protease of E. coli. Incubation of AfsLDC in the presence of [g-32P]ATP yielded an autophosphorylated form (Fig. 3B) on threonine and serine residues (Fig. 3C). The two proteins in the AfsLDC sample autophosphorylated, giving an apparent broad band. In addition, AfsLDC phosphorylated H-AfsR dominantly on threonine and very slightly on serine residues (Fig. 3B and C). These findings show that
AfsLDC and AfsL, too, autophosphorylate and phosphorylate AfsR. 3.5. Influence of pkaG on actinorhodin biosynthesis Disruption of afsK reduces production of the pigment antibiotics actinorhodin and undecylprodigiosin (Matsumoto et al., 1994; Umeyama et al., 2002). We therefore disrupted the chromosomal pkaG gene to examine its possible effect on the pigment production. For this purpose, the upstream and downstream regions of pkaG, together with the kanamycin resistance gene from Tn5, were cloned in pUC19, resulting in a plasmid used for disruption of the chromosomal pkaG gene (Fig. 4A). By using this plasmid, the pkaG-coding region was replaced by the kanamycin-resistance gene. Correct replacement was checked by Southern hybridization using the upstream region of pkaG and the kanamycin resistance gene as 32 P-labeled probes (data not shown). We obtained three pkaG-disruptants. The pkaG-disruptants, DpkaG, of S. coelicolor A3(2) M130 were grown at 30 jC on TSB medium. The three disruptants grew normally and developed aerial hyphae and spores in the same time course as the parental strain, when they were grown by spreading spores and by spreading a lump of mycelium. Until 4 days of growth, however, most colonies, but not all, of mutant DpkaG
R. Sawai et al. / Gene 334 (2004) 53–61
59
Fig. 4. Disruption of the chromosomal pkaG gene and actinorhodin production by the pkaG-disruptant. (A) The pUC19 plasmid, in which the kanamycin resistance gene is sandwiched between the upstream and downstream regions of pkaG, was used for disruption of the chromosomal pkaG gene of S. coelicolor A3(2) M130. (B) The pkaG-disruptant, DpkaG, was grown at 30 jC for 3 days on TSB agar, together with the strain M130 and mutant DafsK (lower). This photograph is a representative that shows reduced actinorhodin production by DpkaG and DafsK. (C) The parent M130, mutant DpkaG, DpkaG harboring pKPK1, and DpkaG harboring the vector pKUM20 were grown at 30 jC on TSB agar for the indicated period.
produced an apparently smaller amount of actinorhodin, which was visible as blue pigmentation, than those of strain M130. All three pkaG-disruptants showed the same phenotype. The degree of reduction of the yield of actinorhodin in DpkaG was lower than that in DafsK, which was used as a control in this assay. A photograph representing this tendency is shown in Fig. 4B. After 5 days, the amounts of actinorhodin produced by strains M130 and DpkaG were apparently the same. A slight
decrease in the amount of actinorhodin and a slight delay in actinorhodin production by mutant DpkaG was also observed in TSB liquid medium, when assayed as described previously (Horinouchi and Beppu, 1984) (data not shown). The reduced production of actinorhodin by mutant DpkaG resulted solely from the pkaG null mutation, because introduction of pKPK1 into DpkaG restored the actinorhodin production (Fig. 4C). Plasmid pKPK1
60
R. Sawai et al. / Gene 334 (2004) 53–61
contained the pkaG promoter and the coding sequence (from nucleotide position 372, with respect to the transcriptional start point of pkaG, to the termination codon of pkaG) between the EcoRI and HindIII sites of a lowcopy-number vector pKUM20 (Yamazaki et al., 2000). In 2.5 days of growth, the parental strain M130 and mutant DpkaG harboring pKPK1 produced actinorhodin in detectable amounts, whereas neither mutant DpkaG nor mutant DpkaG harboring the vector pKUM20 produced a detectable amount of actinorhodin. In 4 days of growth, actinorhodin production by the parent and mutant DpkaG harboring pKPK1 was almost the same, whereas that by mutant DpkaG harboring no plasmid or pKUM20 was detectably reduced under these growth conditions. Actinorhodin production in mutant DpkaG by pKPK1 was also restored on YMPD medium (Umeyama et al., 1999) (data not shown). It is thus apparent that pkaG-disruption influences the production of actinorhodin, to a slight but detectable extent. The time course and amount of undecylprodigiosin produced by mutant DpkaG and the parent were almost the same (data not shown), when undecylprodigiosin in mycelium grown on TSB agar medium and in TSB liquid medium was quantified as described previously (Horinouchi and Beppu, 1984). The degree of reduction of the yield of actinorhodin is lower than those observed for mutant DafsK. Perhaps, AfsK contributes to AfsR phosphorylation to a greater extent than any other kinases. Phosphorylated AfsR then controls secondary metabolism via afsS and some other genes (Lee et al., 2002).
4. Discussion S. coelicolor A3(2) contains about 40 genes that encode possible serine/threonine or tyrosine kinases, of which AfsK, PkaG, and AfsL are found to phosphorylate AfsR in vitro, a transcriptional activator for afsS controlling secondary metabolism in an as yet unknown way. Because the catalytic domains of PkaG and AfsL are considerably similar to each other and to that of AfsK, we assume that these kinases and several other kinases listed in Table 1 are also capable of phosphorylating AfsR both in vitro and in vivo. If each of the kinases senses its own signal among a plethora of environmental cues, as is known for eukaryotic protein kinases, AfsR makes an integrator of the signals. As an example, PkaF (the counterpart of Streptomyces toyocaensis StoPK-1), which has a kinase domain similar to that of AfsK, was reported to sense oxidative stress (Neu et al., 2002). AfsR as a phosphorylated form appears to control not only afsS but some other gene(s) that influences the yields of the pigmented antibiotics actinorhodin and undecylprodigiosin (Lee et al., 2002). The additional gene(s) activated by the phosphorylated AfsR
may help the host to grow and develop in response to the respective environmental stimuli by adjusting physiological conditions. The yields of secondary metabolites are the sum of the outcomes of multiple regulatory systems, one of which is the AfsR regulatory pathway. Therefore, when mutant DpkaG is grown under a certain growth condition where pkaG receives the specific signal, the yields of the pigments may be distinctly different from those in strain M130. The conditions are conceivably difficult to be reproduced on a Petri dish in the laboratory, when we think of the growth of Streptomyces in soil in the complex ecosystem. The difference in the degree of reduction of the pigment production between mutants DpkaG and DafsK perhaps reflects the difference of the environmental conditions that are sensed by these two kinases. Although we have not generated afsL-disruptants, we expect that afsL-disruption also influences, more or less, the production of actinorhodin, since AfsL can phosphorylate AfsR. Thus, multiple protein kinases able to phosphorylate AfsR serve as tuners to determine the yields of secondary metabolites, depending on the growth conditions. In addition to the presence of multiple protein kinases, several genes that encode proteins showing end-to-end similarity to AfsR of 993 a.a. are found in the database of the S. coelicolor A3(2) genome. These are cdaR locating in the calcium-dependent antibiotic (CDA) biosynthetic gene cluster and encoding a 638 a.a. protein, SC4G2-07 encoding a 636 a.a. protein, SC6A11-09 encoding a 761 a.a. protein, SCD72A-02 encoding a 1114 a.a. protein, and SCC75A-05c encoding a 1334 a.a. protein. The five proteins all contain an OmpR-type DNA-binding motif (Wietzorrek and Bibb, 1997) at their N-terminal ends, and the three (CdaR, SC4G2-07, and SC6A11-09) contain distinct A- and B-type ATP-binding consensus sequences required for ATPase activity and for transcriptional activation (Lee et al., 2002). Although the functions of these proteins have not been studied, it is possible that these putative DNA-binding proteins are also activated by phosphorylation by protein serine/threonine kinases. Because of the end-to-end similarity to AfsR, these may be phosphorylated by AfsK, PkaG, and AfsL. The present study shows the presence of multiple kinases able to phosphorylate AfsR. The presence of about 40 eukaryotic-type protein kinases and at least five AfsR-like proteins in S. coelicolor A3(2) suggests that the signal transduction networks via eukaryotic-type protein phosphorylation for growth, morphogenesis, and secondary metabolism are diverse and versatile. Accumulating evidence has shown that complex eukaryotic-type signal transduction systems function in a variety of Streptomyces spp. (Umeyama et al., 2002; Horinouchi, 2003). Conceivably, the mycelial, multicellular growth in soil in the ecosystem has forced Streptomyces spp. to retain the signal transduction systems via eukaryotic-type protein kinases for the growth as multicellular filamentous hyphae, by which
R. Sawai et al. / Gene 334 (2004) 53–61
distant cells in an individual hypha simultaneously respond to environmental cues (Kennelly and Potts, 1996; Zhang, 1996). It is also noted that the AfsK-AfsR system in Streptomyces griseus controls aerial mycelium formation in response to glucose, but not secondary metabolism to a detectable extent (Umeyama et al., 1999), showing that a certain kinase and its target protein control secondary metabolism in one strain and morphological development in another strain. Furthermore, during evolution, some specific regulatory systems have been incorporated into the regulatory pathway via protein kinases; afsR2, a probable target of AfsR in Streptomyces lividans, seems to have acquired a regulatory system that responds to carbon sources (Vo¨gtli et al., 1994; Kim et al., 2001). These events during evolution may have increased the diversity of the regulatory mechanisms via eukaryotic-type kinases. Acknowledgements P.-C. Lee was supported by the Japan Society for the Promotion of Science. This work was supported by the Asahi Glass Foundation and by the Bio Design Program of the Ministry of Agriculture, Forestry, and Fisheries of Japan (BDP-03-VI-2-8). References Beck, E., Ludwig, G., Auerswald, E.A., Reiss, B., Schaller, H., 1982. Nucleotide sequence and exact localization of the neomycin phosphotransferase gene from transposon Tn5. Gene 19, 327 – 336. Bentley, S.D., et al., 2002. Complete genome sequence of the model actinomycete Streptomyces coelicolor A3(2). Nature 417, 141 – 147. Cooper, J.A., Sefton, B.M., Hunter, T., 1983. Detection and quantification of phosphotyrosine in proteins. Methods Enzymol. 99, 387 – 402. Floriano, B., Bibb, M., 1996. afsR is a pleiotropic but conditionally required regulatory gene for antibiotic production in Streptomyces coelicolor A3(2). Mol. Microbiol. 21, 385 – 396. Hanks, S.K., Quinn, A.M., Hunter, T., 1988. The protein kinase family: conserved features and deduced phylogeny of the catalytic domains. Science 241, 42 – 52. Hirakata, T., Kieser, H., Hopwood, D.A., Urabe, H., Ogawara, H., 1998. Putative protein serine/threonine kinase genes are located in several positions on the chromosome of Streptomyces coelicolor A3(2). FEMS Microbiol. Lett. 159, 1 – 5. Hopwood, D.A., Bibb, M.J., Chater, K.F., Kieser, T., Bruton, C.J., Kieser, H.M., Lydiate, D.J., Smith, C.P., Ward, J.M., Schrempf, H., 1985. Genetic Manipulation of Streptomyces: a Laboratory Manual The John Innes Foundation, Norwich, UK. Horinouchi, S., 2003. AfsR as an integrator of signals that are sensed by multiple serine/threonine kinases in Streptomyces coelicolor A3(2). J. Ind. Microbiol. Biotech. 30, 462 – 467. Horinouchi, S., Beppu, T., 1984. Production in large quantities of actinorhodin and undecylprodigiosin induced by afsB in Streptomyces lividans. Agric. Biol. Chem. 48, 2131 – 2133. Kamps, M.P., Sefton, B.M., 1989. Acid and base hydrolysis of phos-
61
phoproteins bound to immobilon facilitates analysis of phosphoamino acids in gel-fractionated proteins. Anal. Biochem. 176, 22 – 27. Kelemen, G.H., Brian, P., Fla¨rdh, K., Chamberlin, L., Chater, K.F., Buttner, M.J., 1998. Developmental regulation of transcription of whiE, a locus specifying the polyketide spore pigment in Streptomyces coelicolor A3(2). J. Bacteriol. 180, 2515 – 2521. Kennelly, P.J., Potts, M., 1996. Fancy meeting you here! A fresh look at ‘‘prokaryotic’’ protein phosphorylation. J. Bacteriol. 178, 4759 – 4764. Kim, E.-S., Hong, H.-J., Choi, C.-Y., Cohen, S.N., 2001. Modulation of actinorhodin biosynthesis in Streptomyces lividans by glucose repression of afsR2 gene transcription. J. Bacteriol. 183, 2198 – 2203. Lee, P.-C., Umeyama, T., Horinouchi, S., 2002. afsS is a target of AfsR, a transcriptional factor with ATPase activity that globally controls secondary metabolism in Streptomyces coelicolor A3(2). Mol. Microbiol. 43, 1413 – 1430. Maniatis, T., Fritsch, E.F., Sambrook, J., 1982. Molecular Cloning: A Laboratory Manual Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Matsumoto, A., Hong, S.-K., Ishizuka, H., Horinouchi, S., Beppu, T., 1994. Phosphorylation of the AfsR protein involved in secondary metabolism in Streptomyces species by a eukaryotic-type protein kinase. Gene 146, 47 – 56. Neu, J.M., MacMillan, S.V., Nodwell, J.R., Wright, G.D., 2002. StoPK-1, a serine/threonine protein kinase from the glycopeptide antibiotic producer Streptomyces toyocaensis NRRL 15009, affects oxidative stress response. Mol. Microbiol. 44, 417 – 430. Ogawara, H., Aoyagi, N., Watanabe, M., Urabe, H., 1999. Sequences and evolutionary analyses of eukaryotic-type protein kinases from Streptomyces coelicolor A3(2). Microbiology 145, 3343 – 3352. Oh, S.H., Chater, K.F., 1997. Denaturation of circular or linear DNA facilitates targeted integrative transformation of Streptomyces coelicolor A3(2): possible relevance to other organisms. J. Bacteriol. 179, 122 – 127. Strohl, W.R., 1992. Compilation and analysis of DNA sequences associated with apparent streptomycete promoters. Nucleic Acids Res. 20, 961 – 974. Sollner-Webb, B., Reeder, R.H., 1979. The nucleotide sequence of the initiation and termination sites for ribosomal RNA transcription in X. laevis. Cell 18, 485 – 499. Umeyama, T., Horinouchi, S., 2001. Autophosphorylation of a bacterial serine/threonine kinase, AfsK, is inhibited by KbpA, an AfsK-binding protein. J. Bacteriol. 183, 5506 – 5512. Umeyama, T., Lee, P.-C., Ueda, K., Horinouchi, S., 1999. An AfsK/AfsR system involved in the response of aerial mycelium formation to glucose in Streptomyces griseus. Microbiology 145, 2281 – 2292. Umeyama, T., Lee, P.-C., Horinouchi, S., 2002. Protein serine/threonine kinases in signal transduction for secondary metabolism and morphogenesis in Streptomyces. Appl. Microbiol. Biotechnol. 59, 419 – 425. Vo¨gtli, M., Chang, P.C., Cohen, S.N., 1994. afsR2: a previously undetected gene encoding a 63-amino-acid protein that stimulates antibiotic production in Streptomyces lividans. Mol. Microbiol. 14, 643 – 653. Wietzorrek, A., Bibb, M., 1997. A novel family of proteins that regulates antibiotic production in streptomycetes appears to contain an OmpRlike DNA-binding fold. Mol. Microbiol. 25, 1181 – 1184. Yamazaki, H., Ohnishi, Y., Horinouchi, S., 2000. An A-factor-dependent extracytoplasmic function sigma factor (jAdsA) that is essential for morphological development in Streptomyces griseus. J. Bacteriol. 182, 4596 – 4605. Zhang, C.-C., 1996. Bacterial signalling involving eukaryotic-type protein kinases. Mol. Microbiol. 20, 9 – 15.
Phosphorylation of AfsR by multiple serine/threonine kinases in Streptomyces coelicolor A3(2) Reiko Sawai, Ayano Suzuki, Yuji Takano, Ping-Chin Lee 1, Sueharu Horinouchi * Department of Biotechnology, Graduate School of Agriculture and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo, Tokyo 113-8657, Japan Received 11 November 2003; received in revised form 22 January 2004; accepted 13 February 2004 Available online 5 May 2004 Received by T. Sekiya
Abstract AfsK, a protein serine/threonine kinase, autophosphorylates on serine and threonine residues and phosphorylates serine and threonine residues of AfsR, a transcriptional activator for afsS involved in secondary metabolism in Streptomyces coelicolor A3(2). pkaG encoding a 592-amino-acid protein and SCD10.09 (named afsL) encoding a 580-amino-acid protein, both of which encode an AfsK-like protein, were transcribed throughout growth. PkaG with a histidine-tag and the kinase catalytic domain of PkaG, produced in Escherichia coli, autophosphorylated dominantly on threonine and slightly on serine residues. In addition, these proteins phosphorylated AfsR on threonine and serine residues. The catalytic domain of AfsL also autophosphorylated and phosphorylated AfsR, on threonine and serine residues in both cases. AfsR was thus found to be phosphorylated by multiple kinases. Disruption of the chromosomal pkaG gene resulted in slightly reduced production of the pigmented antibiotic actinorhodin. These findings, together with the presence of about 40 AfsK homologues and at least five AfsR homologues in S. coelicolor A3(2), suggest that the regulatory networks via eukaryotic-type protein phosphorylation are more diverse and versatile than we have expected. D 2004 Elsevier B.V. All rights reserved. Keywords: Signal transduction; Antibiotic production; Transcriptional factor; Actinorhodin
1. Introduction A number of proteins in the Gram-positive, filamentous bacterial genus Streptomyces are phosphorylated on their serine/threonine and tyrosine residues in response to developmental phases (Umeyama et al., 2002; Horinouchi, 2003). AfsK, the first serine/threonine kinase found in Streptomyces coelicolor A3(2), autophosphorylates on serine and threonine residues and phosphorylates AfsR, a transcriptional factor with ATPase activity, which activates afsS encoding
Abbreviations: a.a.; amino acid(s); bp; base pair(s); PCR; polymerase chain reaction. * Corresponding author. Tel.: +81-3-5841-5123; fax: +81-3-58418021. E-mail address: [email protected] (S. Horinouchi). 1 Present address: School of Science and Technology, University of Malaysia Sabah, Kota Kinabalu, Sabah, Malaysia. 0378-1119/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.gene.2004.02.046
a small protein involved in secondary metabolism (Lee et al., 2002). Disruption of afsK, afsR, or afsS results in reduced production of actinorhodin, undecylprodigiosin, and a calcium-dependent antibiotic (Floriano and Bibb, 1996; Umeyama et al., 2002), showing that these genes comprise a linear signal transduction system for secondary metabolite formation. In addition to the linear signal relay, AfsR is phosphorylated on its serine and threonine residues by a kinase(s) other than AfsK, because AfsR is phosphorylated by a cell lysate prepared from an afsK null mutant (Matsumoto et al., 1994). Thus, multiple protein kinases, including AfsK, have been suggested to phosphorylate AfsR. By analogy to various protein kinases in eukaryotic cells, each of the kinases may autophosphorylate on its serine/threonine residues on sensing some environmental stimulus and transfer the signal by means of phosphorylating AfsR. We therefore started to identify an additional kinase(s) able to phosphorylate AfsR in S. coelicolor A3(2) and to determine its enzymatic properties. For this purpose, we used the
54
R. Sawai et al. / Gene 334 (2004) 53–61
genome sequence of this strain (http://www.sanger.ac.uk/ Projects/S_coelicolor) (Bentley et al., 2002). In this paper, we describe that AfsR is a substrate of multiple kinases. When we think of the possibility that these kinases sense some environmental signals and physiological conditions, AfsR may serve as an integrator of such signals. We also discuss a complex regulatory network involving multiple serine/threonine kinases for growth, morphological development, and secondary metabolism.
2. Materials and methods 2.1. Bacterial strains, plasmids, and media Escherichia coli JM109 and pUC19 used for DNA manipulation were purchased from Takara Shuzo. E. coli BL21(DE3) and pET32a(+) used for production of histidinetagged proteins were purchased from Novagen. Media and growth conditions for E. coli were described by Maniatis et al. (1982). S. coelicolor A3(2) was routinely cultured at 30 jC on Trypto – Soya broth (TSB) (Nissui), which was supplemented with 10 Ag/ml of thiostrepton or kanamycin when necessary. 2.2. General recombinant DNA techniques Restriction enzymes, T4 DNA ligase, and other DNAmodifying enzymes were purchased from Takara Shuzo. [a-32P]dCTP (110 TBq/mmol) for DNA labeling with the Takara BcaBest DNA labeling system and [g-32P]ATP (220 TBq/mmol) for end labeling at 5V ends with T4 polynucleotide kinase were purchased from Amersham Biotech. DNA was manipulated in E. coli (Maniatis et al., 1982) and in S. coelicolor A3(2) (Hopwood et al., 1985), as described. Nucleotide sequences were determined with the Thermo Sequence fluorescence-labeled primer cycle sequencing kit (Amersham) on an automated DNA sequencer. 2.3. Cloning and expression of pkaG and afsL The pkaG-coding sequence was amplified by polymerase chain reaction (PCR) with two primers, 5VGCCGAATTC CATCTGATGTCAGCGCTGGAGCCGGGAC-3V (The underline and the boldface letters indicate an EcoRI site and the start codon of pkaG, respectively) and 5V-GCCAAGCTT GGATCCCTCGAGTCAGCCGAGTTGCGCGAGCTGCGCG-3V (the underline and the boldface letters indicate a HindIII site and the stop codon of pkaG, respectively) and cloned between the EcoRI and HindIII sites of pUC19, resulting in pUC19pkaG. After no undesired alterations in PCR had been checked by nucleotide sequencing, the EcoRI – HindIII fragment was inserted between the EcoRI and HindIII sites of pET32DTRX, in which the thioredoxin-coding
region was removed from pET32a(+) by NdeI digestion (Umeyama and Horinouchi, 2001), resulting in pET32pkaG that would express histidine-tagged PkaG (H-PkaG). For production of PkaGDC, plasmid pET32-pkaGDC was constructed by digestion of pET32-pkaG with NotI and recirculization. H-PkaG and PkaGDC were produced in E. coli BL21(DE3) and purified with His-bind resin, as recommended by the manufacturer. Plasmid pET32-afsLDC used for production of the kinase domain of AfsL was constructed, as follows. A 5V region of the afsL-coding sequence was amplified by PCR with two primers, 5V-GCCGAATTCATGAATCAACACGGCCG-3V (the underline and the boldface letters indicate an EcoRI site and the start codon of afsL) and 5V-GCCAAGCTT GACGGTGCCCGGTGCAAG-3V (the underline indicates a HindIII site; a stop codon was included in the vector sequence) and inserted between the EcoRI and HindIII sites of pUC19, resulting in pUC19-afsLDC. After the afsLDC-coding sequence had been checked for errors in amplification, it was inserted between the EcoRI and HindIII sites of pET32DTRX to generate pET32-afsLDC. AfsLDC was purified with Hisbind resin. 2.4. Phosphorylation protocols and phosphoamino acid analysis For autophosphorylation of PkaG, PkaGDC, and AfsLDC, the standard reaction mixture, containing 15 pmol each of the kinases in 10 mM Tris – HCl (pH 7.2) – 10 mM MnCl2 – 10 mM MgCl2 – 0.1 mM ATP – 10 ACi (370 kBq) [g-32P]ATP – 1 mM dithiothreitol in a total volume of 20 Al, was incubated at 30 jC for 15 min. For phosphorylation of H-AfsR by these kinases, 30 pmol of H-AfsR was added. The reaction was terminated by boiling for 2 min after adding 4 Al of a dye solution. The stopped reaction mixture was fractionated by SDS-polyacrylamide gel electrophoresis. Phosphorylated proteins were transferred to a polyvinylidene difluoride (PVDF) membrane and detected by autoradiography. Phosphorylated a.a. residues were determined by phosphoamino acid analysis by cellulose thin-layer chromatography (Cooper et al., 1983; Kamps and Sefton, 1989). 2.5. S1 nuclease mapping of pkaG and afsL Methods for RNA isolation and S1 mapping were as described by Kelemen et al. (1998). 32P-labeled probes were prepared by PCR with a pair of 32P-labeled and non-labeled primers (see Fig. 2C and D). 2.6. Disruption of pkaG For disruption of the chromosomal pkaG, an about 800base pair (bp) region upstream of pkaG was amplified by P C R w i t h p r i m e r s , 5 V- G C C GAATTC T C G A C G -
R. Sawai et al. / Gene 334 (2004) 53–61
TACCGGCGCACCGTGCG-3V (the underline indicates an EcoRI site) and 5V-GCCCTGCAG CTTGGCGTGCGGCCCTCCCTAGA-3V (the underline indicates a PstI site). An about 800-bp region downstream of pkaG was also amplified with 5V-GCCCTGCAG CGGCAACTCATCAGCGTTCCGCA-3V (the underline indicates a PstI site) and 5V-GCCAAGCTT CGGCTGGTCTCGGTCAACT-3V (the underline indicates a HindIII site). These fragments, together with the kanamycin resistance gene from Tn5 (Beck et al., 1982), were cloned in pUC19, resulting in a plasmid used for disruption of the chromosomal pkaG gene (see Fig. 4A). After the plasmid had been linearized by digestion with DraI and introduced by transformation (Oh and Chater, 1997) into S. coelicolor A3(2) M130, candidates containing correct disruption as a result of double crossover were obtained. Among kanamycin (10 Ag/ml)-resistant colonies, three true disruptants were found by means of Southern hybridization using the upstream region of pkaG and the kanamycin resistance gene as 32P-labeled probes.
Table 1 Amino acid identity between AfsK and putative serine/threonine kinases in S. coelicolor A3(2) Name
No. of a. a. residues
Identity with AfsKa 1 – 300 a.a.
301 – 500 a.a.
AfsK PkaG AfsL (SCD10.09) SC66T3.32c SCE7.11 SC1G2.06c SCL11.07 SCD69.08 SCD69.01/SCD65.24 PksC SC7A12.07 PkaI SCH69.18 SCD35.14 SCI11.13 PkaA PkaH PkaF SCGD3.21c PkaD PkaB PkaE SCD63.30 SCK13.03 PkaJ SC5H1.01/SCF8.01c SC2A6.02c SCD63.08 SCC24.21 SC2H4.01/SC9G1.09 RamC
799 592 580 783 720 686 493 626 632 556 745 380 673 586 550 543 717 667 522 599 417 487 576 670 548 538 452 979 1349 1261 930
– 50.3 47.9 47.8 45.4 44.2 41.5 39.6 39.1 34.7 33.3 32.3 32.0 31.8 31.3 30.7 30.6 30.4 30.3 30.0 29.8 29.7 29.4 28.3 27.8 27.4 27.1 24.2 20.8 20.6 18.0
– 25.3 20.4 23.8 23.7 19.5 23.3 27.4 8.1 20.5 24.6 20.6 20.3 52.3 27.8 24.5 27.1 17.9 23.6 26.0 22.6 24.0 25.7 24.6 23.1 29.2 20.9 26.5 31.6 19.6 20.5
a
The N-terminal regions of 300 a.a. containing the catalytic domain (1 – 300 a.a.) and the next 200 a.a. (301 – 500 a.a.) are compared with those of AfsK.
55
3. Results
3.1. Putative serine/threonine kinases in S. coelicolor A3(2) The presence of more than 40 proteins having characteristic sequence motifs, termed Hanks’ consensus sequences (Hanks et al., 1988), in S. coelicolor A3(2) have been predicted. Table 1 lists 30 such proteins, together with amino acid (a.a.) identity to the kinase catalytic domain of AfsK. All these kinases have a catalytic domain of serine/threonine or tyrosine kinases at their N-terminal portions, and the identity of their Cterminal parts is low. We chose three genes, pkaG (encoding a 592 a.a. protein; having a catalytic domain with 50% a.a. identity to that of AfsK), SCD10.09 (580 a.a.; 48% identity; later named afsL), and SCD66T3.32c (783 a.a.; 48% identity), to determine whether these genes are actually transcribed and whether these putative serine/threonine kinases autophosphorylate and phosphorylate AfsR. These genes show the highest similarity in amino acid sequence to AfsK. Although pkaG was previously cloned and its nucleotide sequence was determined (Hirakata et al., 1998; Ogawara et al., 1999), no further study has been reported. We succeeded in production of proteins PkaG and SCD10.09 in E. coli, but failed to produce protein SCD66T3.32c because of its instability in E. coli. Therefore, we analyzed PkaG and SCD10.09 (AfsL) in this study. The a.a. sequences of the kinase catalytic domains of these proteins are shown in Fig. 1. Hanks’ consensus sequences, numbered I to XI (Hanks et al., 1988), are well conserved. 3.2. Transcription of pkaG and afsL We first examined the transcription of the two genes by S1 nuclease mapping. RNA was prepared from cells grown at 30 jC for various periods in TSB liquid medium. The 32P-labeled probes were prepared by PCR with a pair of 32P-labeled and non-labeled primers, shown in Fig. 2C and D. High-resolution S1 mapping with the RNA prepared from the cells grown for 12 h identified a single transcriptional start point for each of the two genes (Fig. 2A and B). The transcriptional start points of afsL and pkaG were 34 and 36 nucleotides, respectively, upstream of the translational start codons. In front of the transcriptional start point of pkaG, a TATGCT sequence, very similar to a typical 10 sequence TATAAT found in Streptomyces and other bacteria (Strohl, 1992), is present. However, no sequence similar to a 35 consensus sequence, TTGACA for many bacteria and TTGACR (R: A or G) for housekeeping genes in Streptomyces (Strohl, 1992), is present at an appropriate position. In front of the transcriptional start point of afsL, no sequences similar to 10 and 35 consensus sequences are present.
56 R. Sawai et al. / Gene 334 (2004) 53–61 Fig. 1. Amino acid alignment of the kinase catalytic domains of AfsK and the two kinases, PkaG and AfsL, used in this study. Conserved subdomains I to XI are taken from Hanks et al. (1988). The a.a. residues well conserved in eukaryotic serine/threonine kinases (Hanks et al., 1988) are shown under the alignment. The a.a. identical between two of the three proteins are shown in white letters in black boxes.
R. Sawai et al. / Gene 334 (2004) 53–61
57
We also determined the course of transcription of these genes by low-resolution S1 mapping (Fig. 2E). hrdB encoding a major sigma factor, which is transcribed throughout growth, was used to monitor the quantity and quality of the RNA used, as described previously (Umeyama and Horinouchi, 2001). Both genes were transcribed throughout growth, just like hrdB. Under these conditions, actinorhodin and undecylprodigiosin were detectable at 36 h. The courses of transcription of these genes are the same as that of afsK, which is known to regulate actinorhodin production by phosphorylating AfsR (Lee et al., 2002). 3.3. Characterization of pkaG
Fig. 2. Transcription of pkaG and afsL. RNA isolation from mycelium grown at 30 jC for 12 h on solid medium and S1 mapping were described previously (Umeyama and Horinouchi, 2001). (A and B) High-resolution S1 mapping for determination of the transcriptional start points of pkaG and afsL with 32P-labeled probes prepared by PCR with primers shown in C and D. The positions of the S1-protected fragments are shown by arrowheads, and the transcriptional start points are assigned to the residues as shown after corrections for DNA/RNA hybrids (Sollner-Webb and Reeder, 1979). (C and D) Nucleotide sequences of the promoter regions of pkaG and afsL. The transcriptional start points are shown by arrows. The primers used for preparing 32P-labeled probes are also shown. (E) Low-resolution S1 mapping of pkaG and afsL. The RNA and 32P-labeled probes were the same as those used for high-resolution S1 mapping.
Plasmid pET32-pkaG would direct the synthesis of a protein Met-His6-S tag-PkaG (Met-1 to Gly-592) (H-PkaG; 632 a.a., 69.5 kDa). The histidine tag and S tag are derived from the pET32 vector. E. coli BL21(DE3) harboring pET32-pkaG was grown and the cell extract was analyzed by SDS-polyacrylamide gel electrophoresis. The amounts of H-PkaG produced both in the soluble and in the insoluble fractions were only small (data not shown), probably because H-PkaG was unstable and readily digested by E. coli proteases. This is also true for AfsK (Umeyama and Horinouchi, 2001). We purified H-PkaG from the soluble fraction with His-bind resin and assayed its ability to autophosphorylate, as described previously (Umeyama and Horinouchi, 2001). H-PkaG thus purified showed distinct activity to autophosphorylate dominantly on threonine and slightly on serine residues (Fig. 3C). Our attempt to phosphorylate histidine-tagged AfsR (H-AfsR) (Umeyama and Horinouchi, 2001) was rather difficult because of the instability and weak activity of PkaG. The kinase catalytic domain (Met-1 to Arg-311) of AfsK can autophosphorylate and phosphorylate AfsR in the same way as AfsK with the full length (Umeyama and Horinouchi, 2001). We then produced PkaGDC, in which a Cterminal portion from Gly-338 to Gly-592 was deleted. PkaGDC, with a structure of Met-His6-S tag-PkaG (Met-1 to Gly-337)-15 a.a. (406 a.a., 45 kDa; the C-terminal 15 a.a. is derived from the vector sequence), was produced stably and purified from the soluble fraction of E. coli harboring pET32-pkaGDC, giving a single major protein band on SDS-polyacrylamide gel electrophoresis (Fig. 3A). Incubation of PkaGDC with [g-32P]ATP yielded an autophosphorylated form (Fig. 3B) on threonine and serine residues, as determined by phosphoamino acid analysis by cellulose thin-layer chromatography (Fig. 3C). Addition of H-AfsR of 110 kDa to the incubation mixture yielded a phosphorylated form of H-AfsR at the position corresponding to 110 kDa, in addition to the autophosphorylated form of PkaGDC (Fig. 3B). H-AfsR was purified as described previously (Umeyama and Horinouchi, 2001) and gave a single major band on SDS-polyacrylamide gel electrophoresis. Phosphoamino acid analysis showed that
58
R. Sawai et al. / Gene 334 (2004) 53–61
Fig. 3. Characterization of PkaG and AfsL. (A) SDS-polyacrylamide gel electrophoresis of the cell lysates prepared from E. coli harboring pET32-pkaGDC or pET32-afsLDC and the purified PkaGDC and AfsLDC used for phosphorylation assays. The soluble fraction (lane 1) and insoluble fraction (lane 2) from E. coli harboring pET32-pkaGDC are shown, together with PkaGDC purified with His-bind resin (lane 3) and molecular size markers (lane M). The soluble fraction (lane 4) of E. coli harboring pET32-afsLDC and purified AfsLDC (lane 5) are also shown. A considerable amount of PkaGDC of 45 kDa was produced as inclusion bodies in the insoluble fraction. AfsLDC of about 50 kDa was purified as a mixture of two proteins, the smaller one of which was perhaps generated by digestion with an E. coli protease. (B) Autoradiograms showing autophosphorylation of PkaGDC and AfsLDC and phosphorylation of AfsR by them. Incubation of PkaGDC or AfsLDC with [g-32P]ATP yielded an autophosphorylated form of 45 or 50 kDa, respectively. Addition of H-AfsR to the reaction mixture yielded a phosphorylated form of H-AfsR of 110 kDa. Incubation of H-AfsR alone in the presence of [g-32P]ATP, as a negative control, yielded no phosphorylated form. (C) Phosphoamino acid analyses of the autophosphorylated forms of PkaGDC and AfsLDC and the H-AfsR protein phosphorylated by them. The autophosphorylated form of PkaGDC contains phosphorylated threonine and serine residues and the H-AfsR phosphorylated by PkaGDC also contains phosphorylated threonine and serine residues (left panel). PkaG with the full length, purified from E. coli harboring pET32-pkaG, also autophosphorylates on threonine and serine residues. AfsLDC was found to autophosphorylate on threonine and serine residues and to phosphorylate threonine and serine residues of H-AfsR (right panel).
the phosphorylated residues of AfsR were threonine and serine (Fig. 3C). We therefore conclude that PkaGDC, and PkaG itself, too, autophosphorylate on threonine and serine residues and phosphorylate threonine and serine residues of AfsR. Although PkaGDC was produced mostly in the soluble fraction, the autophosphorylating activity varied depending on the protein preparation, which hampered us to determine the kinetic parameters for autophosphorylation and phosphorylation of AfsR. This was also true for AfsLDC (see below). 3.4. Characterization of afsl Plasmid pET32-afsLDC would direct the synthesis of a protein Met-His6-S tag-AfsL (Met-1 to Val-338)-13 a.a. (AfsLDC; 399 a.a., 44 kDa). The C-terminal 13 a.a. is derived form the vector sequence. AfsLDC gave two protein bands on SDS-polyacrylamide gel electrophoresis (Fig. 3A). Since the histidine-tag was attached to the N-terminal end of the fused protein, the smaller protein was supposedly generated by a cleavage at its C-terminal region by a protease of E. coli. Incubation of AfsLDC in the presence of [g-32P]ATP yielded an autophosphorylated form (Fig. 3B) on threonine and serine residues (Fig. 3C). The two proteins in the AfsLDC sample autophosphorylated, giving an apparent broad band. In addition, AfsLDC phosphorylated H-AfsR dominantly on threonine and very slightly on serine residues (Fig. 3B and C). These findings show that
AfsLDC and AfsL, too, autophosphorylate and phosphorylate AfsR. 3.5. Influence of pkaG on actinorhodin biosynthesis Disruption of afsK reduces production of the pigment antibiotics actinorhodin and undecylprodigiosin (Matsumoto et al., 1994; Umeyama et al., 2002). We therefore disrupted the chromosomal pkaG gene to examine its possible effect on the pigment production. For this purpose, the upstream and downstream regions of pkaG, together with the kanamycin resistance gene from Tn5, were cloned in pUC19, resulting in a plasmid used for disruption of the chromosomal pkaG gene (Fig. 4A). By using this plasmid, the pkaG-coding region was replaced by the kanamycin-resistance gene. Correct replacement was checked by Southern hybridization using the upstream region of pkaG and the kanamycin resistance gene as 32 P-labeled probes (data not shown). We obtained three pkaG-disruptants. The pkaG-disruptants, DpkaG, of S. coelicolor A3(2) M130 were grown at 30 jC on TSB medium. The three disruptants grew normally and developed aerial hyphae and spores in the same time course as the parental strain, when they were grown by spreading spores and by spreading a lump of mycelium. Until 4 days of growth, however, most colonies, but not all, of mutant DpkaG
R. Sawai et al. / Gene 334 (2004) 53–61
59
Fig. 4. Disruption of the chromosomal pkaG gene and actinorhodin production by the pkaG-disruptant. (A) The pUC19 plasmid, in which the kanamycin resistance gene is sandwiched between the upstream and downstream regions of pkaG, was used for disruption of the chromosomal pkaG gene of S. coelicolor A3(2) M130. (B) The pkaG-disruptant, DpkaG, was grown at 30 jC for 3 days on TSB agar, together with the strain M130 and mutant DafsK (lower). This photograph is a representative that shows reduced actinorhodin production by DpkaG and DafsK. (C) The parent M130, mutant DpkaG, DpkaG harboring pKPK1, and DpkaG harboring the vector pKUM20 were grown at 30 jC on TSB agar for the indicated period.
produced an apparently smaller amount of actinorhodin, which was visible as blue pigmentation, than those of strain M130. All three pkaG-disruptants showed the same phenotype. The degree of reduction of the yield of actinorhodin in DpkaG was lower than that in DafsK, which was used as a control in this assay. A photograph representing this tendency is shown in Fig. 4B. After 5 days, the amounts of actinorhodin produced by strains M130 and DpkaG were apparently the same. A slight
decrease in the amount of actinorhodin and a slight delay in actinorhodin production by mutant DpkaG was also observed in TSB liquid medium, when assayed as described previously (Horinouchi and Beppu, 1984) (data not shown). The reduced production of actinorhodin by mutant DpkaG resulted solely from the pkaG null mutation, because introduction of pKPK1 into DpkaG restored the actinorhodin production (Fig. 4C). Plasmid pKPK1
60
R. Sawai et al. / Gene 334 (2004) 53–61
contained the pkaG promoter and the coding sequence (from nucleotide position 372, with respect to the transcriptional start point of pkaG, to the termination codon of pkaG) between the EcoRI and HindIII sites of a lowcopy-number vector pKUM20 (Yamazaki et al., 2000). In 2.5 days of growth, the parental strain M130 and mutant DpkaG harboring pKPK1 produced actinorhodin in detectable amounts, whereas neither mutant DpkaG nor mutant DpkaG harboring the vector pKUM20 produced a detectable amount of actinorhodin. In 4 days of growth, actinorhodin production by the parent and mutant DpkaG harboring pKPK1 was almost the same, whereas that by mutant DpkaG harboring no plasmid or pKUM20 was detectably reduced under these growth conditions. Actinorhodin production in mutant DpkaG by pKPK1 was also restored on YMPD medium (Umeyama et al., 1999) (data not shown). It is thus apparent that pkaG-disruption influences the production of actinorhodin, to a slight but detectable extent. The time course and amount of undecylprodigiosin produced by mutant DpkaG and the parent were almost the same (data not shown), when undecylprodigiosin in mycelium grown on TSB agar medium and in TSB liquid medium was quantified as described previously (Horinouchi and Beppu, 1984). The degree of reduction of the yield of actinorhodin is lower than those observed for mutant DafsK. Perhaps, AfsK contributes to AfsR phosphorylation to a greater extent than any other kinases. Phosphorylated AfsR then controls secondary metabolism via afsS and some other genes (Lee et al., 2002).
4. Discussion S. coelicolor A3(2) contains about 40 genes that encode possible serine/threonine or tyrosine kinases, of which AfsK, PkaG, and AfsL are found to phosphorylate AfsR in vitro, a transcriptional activator for afsS controlling secondary metabolism in an as yet unknown way. Because the catalytic domains of PkaG and AfsL are considerably similar to each other and to that of AfsK, we assume that these kinases and several other kinases listed in Table 1 are also capable of phosphorylating AfsR both in vitro and in vivo. If each of the kinases senses its own signal among a plethora of environmental cues, as is known for eukaryotic protein kinases, AfsR makes an integrator of the signals. As an example, PkaF (the counterpart of Streptomyces toyocaensis StoPK-1), which has a kinase domain similar to that of AfsK, was reported to sense oxidative stress (Neu et al., 2002). AfsR as a phosphorylated form appears to control not only afsS but some other gene(s) that influences the yields of the pigmented antibiotics actinorhodin and undecylprodigiosin (Lee et al., 2002). The additional gene(s) activated by the phosphorylated AfsR
may help the host to grow and develop in response to the respective environmental stimuli by adjusting physiological conditions. The yields of secondary metabolites are the sum of the outcomes of multiple regulatory systems, one of which is the AfsR regulatory pathway. Therefore, when mutant DpkaG is grown under a certain growth condition where pkaG receives the specific signal, the yields of the pigments may be distinctly different from those in strain M130. The conditions are conceivably difficult to be reproduced on a Petri dish in the laboratory, when we think of the growth of Streptomyces in soil in the complex ecosystem. The difference in the degree of reduction of the pigment production between mutants DpkaG and DafsK perhaps reflects the difference of the environmental conditions that are sensed by these two kinases. Although we have not generated afsL-disruptants, we expect that afsL-disruption also influences, more or less, the production of actinorhodin, since AfsL can phosphorylate AfsR. Thus, multiple protein kinases able to phosphorylate AfsR serve as tuners to determine the yields of secondary metabolites, depending on the growth conditions. In addition to the presence of multiple protein kinases, several genes that encode proteins showing end-to-end similarity to AfsR of 993 a.a. are found in the database of the S. coelicolor A3(2) genome. These are cdaR locating in the calcium-dependent antibiotic (CDA) biosynthetic gene cluster and encoding a 638 a.a. protein, SC4G2-07 encoding a 636 a.a. protein, SC6A11-09 encoding a 761 a.a. protein, SCD72A-02 encoding a 1114 a.a. protein, and SCC75A-05c encoding a 1334 a.a. protein. The five proteins all contain an OmpR-type DNA-binding motif (Wietzorrek and Bibb, 1997) at their N-terminal ends, and the three (CdaR, SC4G2-07, and SC6A11-09) contain distinct A- and B-type ATP-binding consensus sequences required for ATPase activity and for transcriptional activation (Lee et al., 2002). Although the functions of these proteins have not been studied, it is possible that these putative DNA-binding proteins are also activated by phosphorylation by protein serine/threonine kinases. Because of the end-to-end similarity to AfsR, these may be phosphorylated by AfsK, PkaG, and AfsL. The present study shows the presence of multiple kinases able to phosphorylate AfsR. The presence of about 40 eukaryotic-type protein kinases and at least five AfsR-like proteins in S. coelicolor A3(2) suggests that the signal transduction networks via eukaryotic-type protein phosphorylation for growth, morphogenesis, and secondary metabolism are diverse and versatile. Accumulating evidence has shown that complex eukaryotic-type signal transduction systems function in a variety of Streptomyces spp. (Umeyama et al., 2002; Horinouchi, 2003). Conceivably, the mycelial, multicellular growth in soil in the ecosystem has forced Streptomyces spp. to retain the signal transduction systems via eukaryotic-type protein kinases for the growth as multicellular filamentous hyphae, by which
R. Sawai et al. / Gene 334 (2004) 53–61
distant cells in an individual hypha simultaneously respond to environmental cues (Kennelly and Potts, 1996; Zhang, 1996). It is also noted that the AfsK-AfsR system in Streptomyces griseus controls aerial mycelium formation in response to glucose, but not secondary metabolism to a detectable extent (Umeyama et al., 1999), showing that a certain kinase and its target protein control secondary metabolism in one strain and morphological development in another strain. Furthermore, during evolution, some specific regulatory systems have been incorporated into the regulatory pathway via protein kinases; afsR2, a probable target of AfsR in Streptomyces lividans, seems to have acquired a regulatory system that responds to carbon sources (Vo¨gtli et al., 1994; Kim et al., 2001). These events during evolution may have increased the diversity of the regulatory mechanisms via eukaryotic-type kinases. Acknowledgements P.-C. Lee was supported by the Japan Society for the Promotion of Science. This work was supported by the Asahi Glass Foundation and by the Bio Design Program of the Ministry of Agriculture, Forestry, and Fisheries of Japan (BDP-03-VI-2-8). References Beck, E., Ludwig, G., Auerswald, E.A., Reiss, B., Schaller, H., 1982. Nucleotide sequence and exact localization of the neomycin phosphotransferase gene from transposon Tn5. Gene 19, 327 – 336. Bentley, S.D., et al., 2002. Complete genome sequence of the model actinomycete Streptomyces coelicolor A3(2). Nature 417, 141 – 147. Cooper, J.A., Sefton, B.M., Hunter, T., 1983. Detection and quantification of phosphotyrosine in proteins. Methods Enzymol. 99, 387 – 402. Floriano, B., Bibb, M., 1996. afsR is a pleiotropic but conditionally required regulatory gene for antibiotic production in Streptomyces coelicolor A3(2). Mol. Microbiol. 21, 385 – 396. Hanks, S.K., Quinn, A.M., Hunter, T., 1988. The protein kinase family: conserved features and deduced phylogeny of the catalytic domains. Science 241, 42 – 52. Hirakata, T., Kieser, H., Hopwood, D.A., Urabe, H., Ogawara, H., 1998. Putative protein serine/threonine kinase genes are located in several positions on the chromosome of Streptomyces coelicolor A3(2). FEMS Microbiol. Lett. 159, 1 – 5. Hopwood, D.A., Bibb, M.J., Chater, K.F., Kieser, T., Bruton, C.J., Kieser, H.M., Lydiate, D.J., Smith, C.P., Ward, J.M., Schrempf, H., 1985. Genetic Manipulation of Streptomyces: a Laboratory Manual The John Innes Foundation, Norwich, UK. Horinouchi, S., 2003. AfsR as an integrator of signals that are sensed by multiple serine/threonine kinases in Streptomyces coelicolor A3(2). J. Ind. Microbiol. Biotech. 30, 462 – 467. Horinouchi, S., Beppu, T., 1984. Production in large quantities of actinorhodin and undecylprodigiosin induced by afsB in Streptomyces lividans. Agric. Biol. Chem. 48, 2131 – 2133. Kamps, M.P., Sefton, B.M., 1989. Acid and base hydrolysis of phos-
61
phoproteins bound to immobilon facilitates analysis of phosphoamino acids in gel-fractionated proteins. Anal. Biochem. 176, 22 – 27. Kelemen, G.H., Brian, P., Fla¨rdh, K., Chamberlin, L., Chater, K.F., Buttner, M.J., 1998. Developmental regulation of transcription of whiE, a locus specifying the polyketide spore pigment in Streptomyces coelicolor A3(2). J. Bacteriol. 180, 2515 – 2521. Kennelly, P.J., Potts, M., 1996. Fancy meeting you here! A fresh look at ‘‘prokaryotic’’ protein phosphorylation. J. Bacteriol. 178, 4759 – 4764. Kim, E.-S., Hong, H.-J., Choi, C.-Y., Cohen, S.N., 2001. Modulation of actinorhodin biosynthesis in Streptomyces lividans by glucose repression of afsR2 gene transcription. J. Bacteriol. 183, 2198 – 2203. Lee, P.-C., Umeyama, T., Horinouchi, S., 2002. afsS is a target of AfsR, a transcriptional factor with ATPase activity that globally controls secondary metabolism in Streptomyces coelicolor A3(2). Mol. Microbiol. 43, 1413 – 1430. Maniatis, T., Fritsch, E.F., Sambrook, J., 1982. Molecular Cloning: A Laboratory Manual Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Matsumoto, A., Hong, S.-K., Ishizuka, H., Horinouchi, S., Beppu, T., 1994. Phosphorylation of the AfsR protein involved in secondary metabolism in Streptomyces species by a eukaryotic-type protein kinase. Gene 146, 47 – 56. Neu, J.M., MacMillan, S.V., Nodwell, J.R., Wright, G.D., 2002. StoPK-1, a serine/threonine protein kinase from the glycopeptide antibiotic producer Streptomyces toyocaensis NRRL 15009, affects oxidative stress response. Mol. Microbiol. 44, 417 – 430. Ogawara, H., Aoyagi, N., Watanabe, M., Urabe, H., 1999. Sequences and evolutionary analyses of eukaryotic-type protein kinases from Streptomyces coelicolor A3(2). Microbiology 145, 3343 – 3352. Oh, S.H., Chater, K.F., 1997. Denaturation of circular or linear DNA facilitates targeted integrative transformation of Streptomyces coelicolor A3(2): possible relevance to other organisms. J. Bacteriol. 179, 122 – 127. Strohl, W.R., 1992. Compilation and analysis of DNA sequences associated with apparent streptomycete promoters. Nucleic Acids Res. 20, 961 – 974. Sollner-Webb, B., Reeder, R.H., 1979. The nucleotide sequence of the initiation and termination sites for ribosomal RNA transcription in X. laevis. Cell 18, 485 – 499. Umeyama, T., Horinouchi, S., 2001. Autophosphorylation of a bacterial serine/threonine kinase, AfsK, is inhibited by KbpA, an AfsK-binding protein. J. Bacteriol. 183, 5506 – 5512. Umeyama, T., Lee, P.-C., Ueda, K., Horinouchi, S., 1999. An AfsK/AfsR system involved in the response of aerial mycelium formation to glucose in Streptomyces griseus. Microbiology 145, 2281 – 2292. Umeyama, T., Lee, P.-C., Horinouchi, S., 2002. Protein serine/threonine kinases in signal transduction for secondary metabolism and morphogenesis in Streptomyces. Appl. Microbiol. Biotechnol. 59, 419 – 425. Vo¨gtli, M., Chang, P.C., Cohen, S.N., 1994. afsR2: a previously undetected gene encoding a 63-amino-acid protein that stimulates antibiotic production in Streptomyces lividans. Mol. Microbiol. 14, 643 – 653. Wietzorrek, A., Bibb, M., 1997. A novel family of proteins that regulates antibiotic production in streptomycetes appears to contain an OmpRlike DNA-binding fold. Mol. Microbiol. 25, 1181 – 1184. Yamazaki, H., Ohnishi, Y., Horinouchi, S., 2000. An A-factor-dependent extracytoplasmic function sigma factor (jAdsA) that is essential for morphological development in Streptomyces griseus. J. Bacteriol. 182, 4596 – 4605. Zhang, C.-C., 1996. Bacterial signalling involving eukaryotic-type protein kinases. Mol. Microbiol. 20, 9 – 15.