Hyper-inducible expression system for ... - Semantic Scholar

Hyper-inducible expression system for streptomycetes §, ៮ Sachio Herai*†, Yoshiteru Hashimoto*†, Hiroki Higashibata*, Hideaki Maseda*, Haruo Ikeda‡, Satoshi Omura and Michihiko Kobayashi*¶

*Institute of Applied Biochemistry and Graduate School of Life and Environmental Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8572, Japan; ‡Kitasato Institute for Life Sciences, Kitasato University, Kanagawa 228-8555, Japan; and §The Kitasato Institute, Tokyo 108-8642, Japan ៮ mura, August 19, 2004 Contributed by Satoshi O

S

treptomycetes (1) are Gram-positive soil microorganisms that produce a wide variety of secondary metabolites, many of which have potent biological activities. They produce more than half of the known biologically active microbial products, including many commercially important antibiotics, immunosuppressive compounds, animal health products, and agrochemicals (2). They also produce various enzymes that are commercially and academically valuable (3). This vast reservoir of diverse products makes Streptomyces one of the most important industrial microbial genera. One of the major strategies for improving the productivity of useful products by means of genetic engineering is to attempt to overexpress critical genes by using vectors with constitutively expressed promoters (4–6). However, this strategy has the disadvantage that the constitutive expression of secondary metabolites or proteins may cause a reduction of the host cell growth rate, and concomitantly of the amounts of overall products. Therefore, it is desirable to develop a regulatory expression system that can suppress the expression of target genes until a culture of recombinant cells reaches high density, and then switches the expression ‘‘ON’’ at a selected time point. Although a few regulatory systems in streptomycetes have been reported (4, 7), they have a number of weak points, such as a low level of induced expression, a high level of basal expression, a narrow host range, and an expensive inducer. Thus, it is highly desirable to develop novel regulatory expression systems that overcome these problems. In the actinomycete Rhodococcus rhodochrous J1 (8–10), which is used for the industrial production of acrylamide (11, 12) and nicotinamide (13), nitrilase is strongly induced on the addition of ␧-caprolactam to the culture medium (14). The induced nitrilase (15), which catalyzes the hydrolysis of nitrile to acid and ammonia, corresponds to 35% of all soluble protein in R. rhodochrous J1 (16). In this strain, the expression of nitrilase encoded by nitA is regulated by a transcriptional positive regulator protein, NitR (17). NitR would form a complex with ␧-caprolactam, and binds to a specific site within the nitA promoter region for activation of nitrilase expression (17). This simple and strong induction mechanism was used for the construction of a hyper-inducible expression system for streptomycetes. Here, we describe the development of such a nitR-based www.pnas.org兾cgi兾doi兾10.1073兾pnas.0406058101

system. This unique expression system has been demonstrated to function in various streptomycete strains, and the induced expression level with the system is much higher than those with any other streptomycete expression systems so far known. Materials and Methods Strains, Plasmids, Chemicals, and Genetic Manipulation. Streptomyces lividans TK24, Streptomyces coelicolor A3(2) M145, Streptomyces avermitilis K139, and Streptomyces griseus IFO13350 were used as the host strains for the constructed vectors. Escherichia coli DH10B was used as the host for pUC19 and pBluescript II SK⫹ derivatives (18). Streptomyces plasmid pIJ487 (19) was used for subcloning and plasmid construction in streptomycetes. The synthetic oligonucleotides used for the vector construction and cloning experiments are described in Table 1, which is published as supporting information on the PNAS web site. Restriction endonucleases, DNA polymerase, and T4 DNA ligase were purchased from Toyobo (Osaka, Japan). ␧-Caprolactam was obtained from Nacalai Tesque (Kyoto, Japan). General genetic techniques for streptomycetes and E. coli have been described (3, 18). Plasmid Construction. Construction of trimmed fragments of the R. rhodochrous J1 nitrilase expression system. A set of plasmids contain-

ing trimmed fragments of the nitrilase expression system was constructed as follows. The expression system was obtained from pNJ10 (15) carrying the entire region in a PstI fragment on pUC19 (Fig. 1a). The 5.4-kb PstI–PstI and the 3.5-kb PstI– EcoT22I fragments of pNJ10 were inserted into the PstI-digested site of pBluescript II SK⫹ (2.9 kb), in the direction that avoided possible nitrilase expression under the control of the lac promoter, to generate pSK⫹10 (8.3 kb) and pSK⫹20 (6.4 kb), respectively. The 2.6-kb PstI–NaeI fragment of pNJ10 was inserted into the PstI–EcoRV-digested site of pBluescript II SK⫹ to generate pSK⫹30 (5.5 kb). Subsequently, pSH10 (11.6 kb), pSH20 (9.7 kb), and pSH30 (8.8 kb) were constructed by inserting XbaI–HindIII fragments of pSK⫹10, pSK⫹20, and pSK⫹30 into the XbaI–HindIII-digested site of pIJ487 (6.2 kb), respectively. pSH40 (9.3 kb) and pSH50 (8.4 kb) were also constructed by inserting NheI–HindIII fragments from pSK⫹20 and pSK⫹30 into the XbaI–HindIII digested site of pIJ487, respectively. Construction of expression vector pSH19 containing the PnitA-NitR system.

The fragment carrying the replication and thiostreptonresistance gene (tsr) (20) regions of pIJ487 was used as the backbone of pSH19, the construction procedure being as follows (see also Scheme 1, which is published as supporting information on the PNAS web site). The 0.3-kb DNA fragment containing transcriptional terminator fd (fd-ter) (21) was amplified from pIJ487 as a template by PCR using a primer pair that included a flanking MluI site within the forward primer (fd-ter1; see Abbreviations: PnitA, niA promoter; fd-ter, transcriptional terminator fd; MCS, multicloning site. †S.H. ¶To

and Y.H. contributed equally to this work.

whom correspondence should be addressed. Fax: ⫹81-29-853-4605 (Institute).

© 2004 by The National Academy of Sciences of the USA

PNAS 兩 September 28, 2004 兩 vol. 101 兩 no. 39 兩 14031–14035

APPLIED BIOLOGICAL SCIENCES

Streptomycetes produce useful enzymes and a wide variety of secondary metabolites with potent biological activities (e.g., antibiotics, immunosuppressors, pesticides, etc.). Despite their importance in the pharmaceutical and agrochemical fields, there have been no reports for practical expression systems in streptomycetes. Here, we developed a ‘‘PnitA-NitR’’ system for regulatory gene expression in streptomycetes based on the expression mechanism of Rhodococcus rhodochrous J1 nitrilase, which is highly induced by an inexpensive and safe inducer, ␧-caprolactam. Heterologous protein expression experiments demonstrated that the system allowed suppressed basal expression and hyper-inducible expression, yielding target protein levels of as high as ⬇40% of all soluble protein. Furthermore, the system functioned in important streptomycete strains. Thus, the PnitA-NitR system should be a powerful tool for improving the productivity of various useful products in streptomycetes.

Fig. 1. Construction of a set of plasmids and nitrilase expression in each S. lividans TK24 transformant. (a) Genetic organization of the nitrilase expression system of R. rhodochrous J1. For clarity, only the restriction sites of the nitrilase expression system discussed in the text are shown. Various trimmed fragments are diagrammatically shown below the restriction map, and the construction procedures are described in Materials and Methods. (b) Nitrilase activity in cell-free extracts of S. lividans TK24 transformants containing each plasmid (pSH10 –50 and pIJ487). Each transformant was cultured in the presence or absence of ␧-caprolactam. ND, not detected at a lower limit of detection of 5 nmol兾min per mg of protein. (c) Coomassie-stained SDS兾PAGE showing nitrilase formation in S. lividans TK24 transformants. Lanes M were loaded with the following molecular mass standards: phosphorylase (94 kDa), BSA (67 kDa), ovalbumin (43 kDa), carbonic anhydrase (30 kDa), and soybean trypsin inhibitor (20 kDa). ⫹, 20 ␮g of cell-free extracts of S. lividans TK24 transformants cultured in the medium supplemented with ␧-caprolactam (0.1%, wt兾vol); ⫺, 20 ␮g of cell-free extracts of S. lividans TK24 transformants cultured in the absence of ␧-caprolactam. Arrows indicate the band corresponding to the nitrilase.

supporting information) and a BglII site within the reverse primer (fd-ter2). The amplified blunt-end fragment was inserted into the HincII-digested pUC19 to generate pSH011 (3 kb). The 1-kb nitR fragment was amplified from pNJ10 by using a primer pair that included a flanking BamHI site within the forward primer (nitR-1) and an SpeI site within the reverse primer (nitR-2). The amplified blunt-end fragment was inserted into SmaI-digested pSH011 to generate pSH012 (4 kb). The 0.1-kb nitA promoter (PnitA) fragment (corresponding to the ⫺89 ⬇ ⫹14 region from the transcription start point) was amplified from pNJ10 by using a primer pair that contained a flanking BamHI site within the forward primer (PnitA-1) and a BglII site within the reverse primer (PnitA-2). The amplified fragment was digested with BamHI–BglII and then inserted into pSH012 previously digested with BglII–BamHI to generate pSH014 (4.1 kb), which cannot be digested with BamHI or BglII. On the other hand, the 0.1-kb PnitA fragment was amplified from pNJ10 by using a primer pair that included a flanking XhoI restriction site within the forward primer (PnitA-3) and BglII兾MluI兾SpeI兾 EcoT22I restriction sites within the reverse primer (PnitA-4). The amplified blunt-end fragment was inserted into HincII digested pUC19 to generate pSH021 (2.8 kb). The 0.3-kb DNA fragment containing fd-ter was amplified from pIJ487 by using a primer pair that included flanking PstI兾Bsp1407I sites within the forward primer (fd-ter 3) and an XhoI site within the reverse primer (fd-ter 4). The amplified fragment was inserted into 14032 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0406058101

pSH021 previously digested with PstI–XhoI to generate pSH022 (3.1 kb). The multicloning site (MCS) was amplified from pUC19 by PCR using a primer pair that included a flanking BglII restriction site at both ends (MCS-FR and MCS-RV), and the amplified MCS (0.1 kb) was inserted into BglII-digested pSH022 to generate pSH019 (3.2 kb). pSH014 was digested with MluI– SpeI, and the fragment obtained (1.4 kb) was inserted into the MluI–SpeI-digested pSH019 to generate pSH031 (4.6 kb). Finally, pSH031 was digested with Bsp1407I and EcoT22I, and the resultant fragment (1.9 kb) containing the PnitA-NitR system was ligated with the Acc65I–PstI-digested backbone fragment (3.8 kb) from pIJ487 containing the replication region (ori and rep) and a selectable marker gene (tsr) to generate pSH19 (5.7 kb) (Fig. 2b). Construction of pSH19 recombinants. nitA (15) was amplified by PCR using a primer pair that contained a flanking HindIII site within the forward primer (SDnitA-FR) and a PstI site within the reverse primer (nitA-RV). The amplified fragment was inserted into HindIII–PstI-digested pSH19 to generate nitA::pSH19. In like manner, inhA (22, 23) and xylE (24, 25) were also amplified by PCR using in each case a primer pair which contained a flanking HindIII site within the forward primer (SDinhA-FR or SDxylE-FR) and a BamHI site within the reverse primer (inhA-RV or xylE-RV), respectively. Each amplified fragment was inserted into HindIII–BamHI-digested pSH19 to generate inhA::pSH19 and xylE::pSH19, respectively. Herai et al.

Culture Conditions, Preparation of Cell-Free Extracts, and Plasmid Stability. Streptomycete transformants were grown in baffled

flasks containing yeast extract–malt extract (YEME) liquid medium for 120 h at 28°C on a rotary shaker. The inducer was added at 96 h. Cell-free extracts were prepared by the described method (15). The stability testing of pSH19 in Streptomyces was performed as follows. S. lividans TK24 transformant harboring pSH19 was grown in the absence of thiostrepton for up to 20 generations. Aliquots from the culture were plated on YEME agar plate with or without thiostrepton, and the plasmid stability was expressed as a percent value from the number of colonies obtained with antibiotic selection relative to the number of colonies obtained in the absence of selection. Enzyme Assays. Nitrilase (15, 17), isonitrile hydratase (22, 23), and catechol 2,3-dioxygenase (24, 25) activities were assayed by described methods. One unit of enzyme activity was defined as the amount of enzyme that catalyzed the formation of 1 ␮mol of product per min from the substrate under the standard experimental conditions. Protein concentrations were determined by the method of Bradford, with BSA as a standard protein. Specific activity is expressed as units per mg of protein.

Results Investigation of the R. rhodochrous J1 Nitrilase Expression System in Streptomycete. To examine whether the Rhodococcus nitrilase

Effect of an Inducer on the R. rhodochrous J1 Nitrilase Expression System. We investigated the effect of the concentration of

␧-caprolactam as an inducer on the induction level of the J1 nitrilase expression system. ␧-Caprolactam was added after 96-h cultivation of S. lividans TK24 harboring pSH40, and then the incubation was continued for a further 24 h. The specific activity of the expressed nitrilase increased (i.e., 0.64, 1.17, and 3.02 units per mg of protein) with increasing concentration of ␧-caprolactam (i.e., 0.01, 0.05, and 0.1% wt兾vol, respectively). This observation suggests that the J1 nitrilase expression system is dosedependent. We also observed that ␧-caprolactam had no influence on streptomycetes [e.g., S. lividans TK24, S. coelicolor A3(2), S. avermitilis K139, and S. griseus IFO13350] as to morphological differentiation or secondary metabolism during culture on solid media (data not shown). Herai et al.

Fig. 2. Hyper-inducible expression system. (a) Schematic model of the PnitA-NitR system. NitR interacts with ␧-caprolactam (inducer), and the activated form of NitR causes the production of a large amount of the target protein. To suppress transcriptional read-through from upstream of PnitA, fd-ter is located just upstream of each PnitA. (b) Physical structure of plasmid pSH19. Unique restriction endonuclease sites in the multicloning-site (MCS) of the plasmid are indicated in bold type.

Construction and Evaluation of a PnitA-NitR Expression System. The

above results suggested that only three components (PnitA, NitR, and an inducer) are required for the inducible expression in streptomycetes, allowing the construction of an expression system (PnitA-NitR system) in plasmid pSH19 (Fig. 2). In this system, PnitA, which is located just upstream of a MCS, is used as a possible strong inducible promoter for the expression of the target protein. NitR is also expressed by PnitA even in the absence of an inducer, although its level is very low. For the purpose of minimizing transcriptional read-through from upstream of PnitA, fd-ter (21) is located just upstream of each PnitA in the constructed PnitA-NitR system. When the PnitA-NitR system is induced, the activated form of NitR enhances transcription from both PnitA(s), i.e., those upstream of MCS and nitR. As a result, the NitR level increases, which brings about very strong expression of the target protein, synergistically. pSH19 includes the pIJ101-derived replicational region (26, 27), which gives a multicopy plasmid (⬇300 per cell) and exhibits a broad host range. The stability of pSH19 was 67.4% under the nonselective condition, whereas the vector was completely maintained under the selective pressure. The marker gene used for recombinant selection is the thiostrepton-resistance gene (tsr) (20), and diverse restriction sites are available within the MCS to facilitate gene cloning. To evaluate the PnitA-NitR system as an inducible expression system, three genes [i.e., nitA (15, 17), inhA (22, 23), and xylE (24, 25)] were used for heterologous protein expression. inhA and xylE are derived from Pseudomonas sp. and encode isonitrile hydratase and catechol-2,3-dioxygenase, respectively, whose products can be assayed easily. These three genes were amplified by PCR and individually cloned into pSH19 to generate PNAS 兩 September 28, 2004 兩 vol. 101 兩 no. 39 兩 14033


expression mechanism functions in genus Streptomyces and identify the DNA region required for the inducible expression, we constructed a set of plasmids containing trimmed fragments of a 5.4-kb PstI segment from pNJ10 (15) that includes the entire prototype nitrilase expression system (Fig. 1a). These plasmids were used to transform S. lividans TK24, which makes no detectable nitrilase, and the resulting transformants were cultured in yeast extract–malt extract (YEME) medium with or without ␧-caprolactam (0.1%, wt兾vol). SDS兾PAGE and enzyme assay with benzonitrile as a substrate for each cell-free extract revealed that the nitrilase expression mechanism functioned inducibly in the S. lividans TK24 transformant carrying pSH10, pSH20, or pSH40, and that the induced expression level was extremely high, nitrilase making up ⬇20% of all soluble protein in the supernatants of the cell-free extracts (Fig. 1 b and c). These findings suggest that the transcriptional activity caused by the nitrilase gene promoter (PnitA) of genus Rhodococcus remains very strong, even in genus Streptomyces. The leaky basal expression is possibly caused by transcriptional read-through from upstream of PnitA. The observation that the nitrilase expression mechanism worked in the streptomycete suggests that no transactive element important for the system, other than what is included in pSH40 (more strictly, PnitA and nitR), is located elsewhere on the R. rhodochrous J1 genome.

expression of the nitrilase in S. coelicolor and S. avermitilis and isonitrile hydratase in S. griseus were observed (Fig. 3b). These findings demonstrated that the PnitA-NitR system is very strongly induced even in other strains, suggesting its wide usability in streptomycetes. The nonexpression of isonitrile hydratase in S. coelicolor and S. avermitilis, and nitrilase in S. griseus, might be caused by misfolding and subsequent degradation of the expressed proteins.

Fig. 3. Expression analysis of the PnitA-NitR system by SDS兾PAGE. ⫹, 20 ␮g of cell-free extracts of each transformant cultured in the medium supplemented with ␧-caprolactam (0.1%, wt兾vol); ⫺, 20 ␮g of cell-free extracts of each transformant cultured in the absence of ␧-caprolactam. Arrows indicate the band corresponding to each expressed heterologous protein. (a) Expression analysis of the PnitA-NitR system in S. lividans TK24. In each induced culture, an enormous amount of the target protein was formed. Specific activity of each expressed enzyme is shown below the SDS兾PAGE panel. (b) Hyper-production of heterologous proteins in S. coelicolor A3(2), S. avermitilis K139, and S. griseus IFO13350.

nitA::pSH19, inhA::pSH19, and xylE::pSH19, respectively, as described in Materials and Methods. On each upstream primer used for PCRs, the Shine–Dalgarno (SD)-sequence region (5⬘AGCAACGGAGGTACGGAC-3⬘) of nitA (15) was designed to be a well recognized ribosome-binding-site in streptomycetes. Each of these three recombinant plasmids was individually introduced into S. lividans TK24, and the resulting transformants were cultured in yeast extract–malt extract (YEME) medium with or without ␧-caprolactam (0.1%, wt兾vol). Analysis by enzyme assay and SDS兾PAGE revealed that the target proteins were exceptionally highly expressed, constituting up to ⬇40% of all soluble protein in each case (Fig. 3a). No enzyme activity was detected in noninduced transformants, demonstrating that the basal expression is strictly suppressed. NitR production was found to be much lower than that of the target protein, in part because of the absence of an optimal SD sequence just upstream of the nitR start codon. To investigate whether the PnitA-NitR system is functional in other streptomycete strains, it was also examined in S. coelicolor A3(2), S. avermitillis K139, and S. griseus IFO13350. These strains were individually transformed with nitA::pSH19 and inhA::pSH19, and the resulting transformants were treated by the usual procedure. As a result, very high levels of induced 14034 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0406058101

Discussion The nitrilase gene system of an actinomycete, R. rhodochrous J1, was engineered as a regulatory expression system for streptomycetes. Using a prototype system, we showed the following: (i) a Rhodococcus-derived promoter (PnitA) can function well in streptomycetes; (ii) the system functions with only very simple components (PnitA, NitR, and an inducer), and requires no other trans-active elements; (iii) there is little, if any, basal expression, and there is a high induction ratio; and (iv) the system shows dose-dependence, with an inducer, ␧-caprolactam, that has no obvious influence on the physiological conditions of streptomycetes. These features of this nitrilase expression system are very promising for the design of a new system. Based on these findings, we constructed an expression vector, pSH19, containing a unique regulatory expression system (PnitA-NitR system). Heterologous protein expression experiments using pSH19 revealed that the PnitA-NitR system is hyperinduced on the addition of ␧-caprolactam as an inducer, and is able to express target proteins at levels of as high as ⬇40% of all soluble protein (i.e., 396 mg of protein per liter of culture medium). To the best of our knowledge, this surprising induction level is much higher than those obtained with any other preexisting regulatory systems for streptomycetes (e.g., the tipA promoter system expresses only 38 mg of protein per liter of culture medium; ref. 4). On the other hand, the noninduced basal expression level was so low that it was undetectable. Although we observed that the expression system is induced by isovaleronitrile as well as ␧-caprolactam in streptomycetes, isovaleronitrile is highly toxic and comparatively expensive and unsuitable for large-scale use. ␧-Caprolactam is a very inexpensive and safe inducer, and it is very practical for large-scale industrial use. We also found that the PnitA-NitR system is functional in various streptomycetes such as S. lividans TK24, S. coelicolor A3(2), S. avermitilis K139, and S. griseus IFO13350, demonstrating the wide usability of this system. These organisms have been very well studied, especially S. coelicolor (28, 29) and S. avermitilis (30, 31), for each of which the whole genome sequence has already been determined and extensive genomic information involving biosynthetic pathways for useful products or other metabolic pathways has been revealed. The combined application of the genomic information and regulated expression systems such as that described here could contribute to the systematic improvement of productivity in streptomycetes, for example, by enhancing the predicted rate-limiting steps of biosynthetic pathways (32–38). In the cases that the tsr gene in pSH19 is not ideal as a selection marker [e.g., the host strains are resistant to thiostrepton (3) or the addition of thiostrepton can induce a regulon of genes via the tipA system (3, 7)], an alternative selection marker that could be substituted for tsr would be preferable. The PnitA-NitR regulatory expression system is also potentially important for research use. For example, streptomycetes show remarkable morphological differentiation, and the developmental process is controlled through complex regulatory cascades involving multiple gene functions (39, 40). In such research, a regulatory expression system is indispensable for providing easily controlled and conditional induction of target gene expression. Moreover, it could be valuable for ‘‘knock-in’’ studies, Herai et al.

permitting the functional analysis of interesting genes through rearrangement of the regulatory networks.

1. Hopwood, D. A. (1999) Microbiology 145, 2183–2202. 2. Demain, A. L. & Fang, A. (2000) Adv. Biochem. Eng. Biotechnol. 69, 1–39. 3. Kieser, T., Bibb, M. J., Buttner, M. J., Chater, K. F. & Hopwood, D. A. (2000) Practical Streptomyces Genetics: A Laboratory Manual (The John Innes Foundation, Norwich, U.K.). 4. Schmitt-John, T. & Engels, J. W. (1992) Appl. Microbiol. Biotechnol. 36, 493–498. 5. Anne, J. & Van Mellaert, L. (1993) FEMS Microbiol. Lett. 114, 121–128. 6. Binnie, C., Cossar, J. D. & Stewart, D. I. (1997) Trends Biotechnol. 15, 315–320. 7. Takano, E., White, J., Thompson, M. J. & Bibb, M. J. (1995) Gene 166, 133–137. 8. Komeda, H., Kobayashi, M. & Shimizu, S. (1996) Proc. Natl. Acad. Sci. USA 93, 4267–4272. 9. Komeda, H., Kobayashi, M. & Shimizu, S. (1996) J. Biol. Chem. 271, 15796– 15802. 10. Komeda, H., Kobayashi, M. & Shimizu, S. (1997) Proc. Natl. Acad. Sci. USA 94, 36–41. 11. Kobayashi, M., Nagasawa, T. & Yamada, H. (1992) Trends Biotechnol. 10, 402–408. 12. Kobayashi, M. & Shimizu, S. (1998) Nat. Biotechnol. 16, 733–736. 13. Yamada, H., Shimizu, S. & Kobayashi, M. (2001) Chem. Rec. 1, 152–161. 14. Nagasawa, T., Nakamura, T. & Yamada, H. (1990) Arch. Microbiol. 155, 13–17. 15. Kobayashi, M., Komeda, H., Yanaka, N., Nagasawa, T. & Yamada, H. (1992) J. Biol. Chem. 267, 20746–20751. 16. Nagasawa, T., Kobayashi, M. & Yamada, H. (1988) Arch. Microbiol. 150, 89–94. 17. Komeda, H., Hori, Y., Kobayashi, M. & Shimizu, S. (1996) Proc. Natl. Acad. Sci. USA 93, 10572–10577. 18. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Moleclular Cloning: A Laboratory Manual (Cold Spring Harbor Lab. Press, Plainview, NY), 2nd Ed. 19. Ward, J. M., Janssen, G. R., Kieser, T., Bibb, M. J., Buttner, M. J. & Bibb, M. J. (1986) Mol. Gen. Genet. 203, 468–478. 20. Bibb, M. J., Bibb, M. J., Ward, J. M. & Cohen, S. N. (1985) Mol. Gen. Genet. 199, 26–36. 21. Gentz, R., Langner, A., Chang, A. C. Y., Cohen, S. N. & Bujard, H. (1981) Proc. Natl. Acad. Sci. USA 78, 4936–4940. 22. Goda, M., Hashimoto, Y., Shimizu, S. & Kobayashi, M. (2001) J. Biol. Chem. 276, 23480–23485.

23. Goda, M., Hashimoto, Y., Takase, M., Herai, S., Iwahara, Y., Higashibata, H. & Kobayashi, M. (2002) J. Biol. Chem. 277, 45860–45865. 24. Worsey, M. J. & Williams, P. A. (1975) J. Bacteriol. 124, 7–13. 25. Nakai, C., Kagamiyama, H., Nozaki, M., Nakazawa, T., Inouye, S., Ebina, Y. & Nakazawa, A. (1983) J. Biol. Chem. 258, 2923–2928. 26. Kiser, T., Hopwood, D. A., Wright, H. M. & Thompson, C. J. (1982) Mol. Gen. Genet. 185, 223–228. 27. Kendall, K. J. & Cohen, S. N. (1988) J. Bacteriol. 170, 4634–4651. 28. Hopwood, D. A., Chater, K. F. & Bibb, M. J. (1995) Biotechnology 28, 65–102. 29. Bentley, S. D., Chater, K. F., Cerdeno-Tarraga, A. M., Challis, G. L., Thomson, N. R., James, K. D., Harris, D. E., Quail, M. A., Kieser, H., Harper, D., et al. (2002) Nature 417, 141–147. 30. Omura, S., Ikeda, H., Ishikawa, J., Hanamoto, A., Takahashi, C., Shinose, M., Takahashi, Y., Horikawa, H., Nakazawa, H., Osonoe, T., et al. (2001) Proc. Natl. Acad. Sci. USA 98, 12215–12220. 31. Ikeda, H., Ishikawa, J., Hanamoto, A., Shinose, M., Kikuchi, H., Shiba, T., Sakaki, Y., Hattori, M. & Omura, S. (2003) Nat. Biotechnol. 21, 526–531. 32. Alvarez, M. A., Fu, H., Khosla, C., Hopwood, D. A. & Bailey, J. E. (1996) Nat. Biotechnol. 14, 335–338. 33. Sezonov, G., Blanc, V., Bamas-Jacques, N., Friedmann, A., Pernodet, J. L. & Guerineau, M. (1997) Nat. Biotechnol. 15, 349–353. 34. Madduri, K., Kennedy, J., Rivola, G., Inventi-Solari, A., Filippini, S., Zanuso, G., Colombo, A. L., Gewain, K. M., Occi, J. L., MacNeil, D. J., et al. (1998) Nat. Biotechnol. 16, 69–74. 35. Cropp, T. A., Wilson, D. J. & Reynolds, K. A. (2000) Nat. Biotechnol. 18, 980–983. 36. Butler, A. R., Bate, N., Kiehl, D. E., Kirst, H. A. & Cundliffe, E. (2002) Nat. Biotechnol. 20, 713–716. 37. Cane, D. E., Walsh, C. T. & Khosla, C. (1998) Science 282, 63–68. 38. Tang, L., Shah, S., Chung, L., Carney, L., Katz, L., Khosla, C. & Julien, B. (2000) Science 287, 640–642. 39. Horinouchi, S. & Beppu, T. (1994) Mol. Microbiol. 12, 859–864. 40. Chater, K. F. & Horinouchi, S. (2003) Mol. Microbiol. 48, 9–15.


We dedicate this manuscript to Prof. Keith F. Chater on his retirement from the John Innes Centre. We thank him for his helpful suggestions and critical reading of this manuscript. We are also indebted to Dr. Takeshi Fujii (National Institute for Agro-Environmental Sciences) and Dr. Kenji Ueda (Nihon University) for providing the Streptomyces strains

and plasmids. This work was supported in part by the 21st Century Center of Excellence (COE) Program of the Ministry of Education, Culture, Sports, Science, and Technology (MEXT), a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, and Culture of Japan, the Industrial Technology Research Grant Program in ’02 of the New Energy and Industrial Technology Development Organization (NEDO) of Japan, the National Project on Protein Structural and Functional Analyses, and by a Research Grant (A) of the University Research Projects.

Herai et al.

PNAS 兩 September 28, 2004 兩 vol. 101 兩 no. 39 兩 14035