Role of sgcR3 in positive regulation of enediyne antibiotic C-1027 production of Streptomyces globisporus C-1027

Background C-1027, produced by Streptomyces globisporus C-1027, is one of the most potent antitumoral agents. The biosynthetic gene cluster of C-1027, previously cloned and sequenced, contains at least three putative regulatory genes, i.e. sgcR1, sgcR2 and sgcR3. The predicted gene products of these genes share sequence similarities to StrR, regulators of AraC/XylS family and TylR. The purpose of this study was to investigate the role of sgcR3 in C-1027 biosynthesis. Results Overexpression of sgcR3 in S. globisporus C-1027 resulted in a 30–40% increase in C-1027 production. Consistent with this, disruption of sgcR3 abolished C-1027 production. Complementation of the sgcR3-disrupted strain R3KO with intact sgcR3 gene could restore C-1027 production. The results from real time RT-PCR analysis in R3KO mutant and wild type strain indicated that not only transcripts of biosynthetic structural genes such as sgcA1 and sgcC4, but also putative regulatory genes, sgcR1 and sgcR2, were significantly decreased in R3KO mutant. The cross-complementation studies showed that sgcR1R2 could functionally complement sgcR3 disruption in trans. Purified N-terminal His10-tagged SgcR3 showed specific DNA-binding activity to the promoter region of sgcR1R2. Conclusion The role of SgcR3 has been proved to be a positive regulator of C-1027 biosynthesis in S. globisporus C-1027. SgcR3 occupies a higher level than SgcR1 and SgcR2 in the regulatory hierarchy that controls C-1027 production and activates the transcription of sgcR1 and sgcR2 by binding directly to the promoter region of sgcR1R2.


Background
C-1027, also called lidamycin, is a chromoprotein antitumor antibiotic produced by Streptomyces globisporus C-1027 [1]. As a member of the enediyne family characterized by two acetylenic groups conjugated to a double bond within a 9-or 10-membered ring, C-1027 is 1,000 times more potent than adriamycin, one of the most effective chemotherapeutic agents [2]. C-1027 is a complex consisting of a 1:1 non-covalently associated mixture of an apoprotein and a 9-membered enediyne chromophore. The chromophore of the enediyne family can undergo a rearrangement to form a transient benzenoid diradical species that can abstract hydrogen atoms from DNA to initiate a cascade leading to DNA breaks, ultimately leading to cell death [3,4]. This novel mode of action has attracted great interest in developing these compounds into therapeutic agents for cancer. A CD33 monoclonal antibody (mAB)-calicheamicin (CAL) conjugate (Mylotarg) and neocarzinostatin (NCS) conjugated with poly (styrene-co-maleic acid) (SMANCS) were approved in the USA [5] and in Japan [6], respectively. Recently, C-1027 has entered phase II clinical trial in China [7]. Appreciation of the immense pharmacological potential of enediynes has led to a demand for the economical production of C-1027 and its analogues at an industrial scale.
Control of secondary metabolite production in streptomycetes and related actinomycetes is a complex process involving multiple levels of regulation in response to environmental factors [For review, see [8,9]]. In most cases that have been studied in detail, the final checkpoint in production of a secondary metabolite is a pathway-specific transcriptional regulatory gene situated in the biosynthetic cluster. Remarkable progress has been made in dissecting the functions of the pathway-specific regulators. For example, ActII-ORF4 regulates transcription from the actinorhodin biosynthetic genes of S. coelicolor [10,11] and StrR controls the streptomycin biosynthetic cluster of S. griseus [12,13]. Recently, along with the tremendous increase in sequence information for secondary metabolic gene clusters, more and more clusters with multiple cluster-situated regulators were reported (e.g., [14][15][16][17]). The best studied multiple pathway-specific regulatory cascade involves remarkably five regulatory genes in tylosin biosynthetic gene cluster of S. fradiae, and a model for their regulation has been proposed [14,[18][19][20][21][22][23]. Deciphering the complexity of these pathway-specific regulatory networks is of great interest not only for better understanding of the antibiotic regulatory mechanism, but also for providing new strategy for targeted genetic engineering of antibiotic producing strains.
C-1027 nonpeptidic chromophore is a structure of an enediyne core, a deoxy aminosugar, a β-amino acid and a benzoxazolinate ( Fig. 1) [7]. The biosynthetic gene cluster for C-1027, which is the first cloned enediyne gene cluster, contains a total of 56 open reading frames (ORFs) in a region of 75 kbp [24,25]. Bioinformatic analysis and biochemical studies revealed a distinct iterative type I enediyne polyketide synthase (SgcE) and provided a convergent biosynthetic strategy for C-1027 from four biosynthetic building blocks [25]. Further cloning and characterization of biosynthetic gene clusters for four other enediynes (CAL [26], NCS [27], maduropeptin (MDP) [28] and dynemicin [29]) confirmed the unifying paradigm for enediyne biogenesis. In accordance with the complexity of the biosynthetic process, there are no fewer than three ORFs annotated as transcriptional regulators in each known enediyne antibiotic biosynthetic cluster. At least three putative regulatory genes (sgcR1, sgcR2 and sgcR3) associated with the C-1027 biosynthetic gene cluster of S. globisporus C-1027 were annotated in the earlier work by sequence analysis [25]. Furthermore, the biosynthetic gene clusters for two 9-membered enediynes produced by streptomycetes (C-1027 and NCS) show high similarity in the organization of genes around these regulatory genes ( Fig. 2A). Despite chromophore structural uniqueness, all homologues of three genes are located adjacent to the genes of enediyne PKSs (sgcE and ncsE) and the tailoring enzymes (E1 to E11), which are responsible for the biosynthesis of enediyne core. However, almost no cognitional knowledge was acquired about the transcriptional regulation of enediyne antibiotic production prior to the present work.
Hereby we investigated the role of sgcR3 in C-1027 biosynthesis, and provided an initial understanding of pathway-specific regulatory network of sgcR1, sgcR2 and sgcR3 in S. globisporus C-1027.

Overexpression of sgcR3 increased the production of C-1027
Computer-assisted analysis of the sgcR3 gene product (395 aa) showed a high sequence similarity (33% identities and 47% positives) within the whole length of protein Structure of C-1027 chromophore Figure 1 Structure of C-1027 chromophore.
TylR of S. fradiae (Fig. 2B), a pathway-specific global activator of tyl cluster [20,23]. To investigate the function of sgcR3, the expression plasmid of sgcR3 associated with its native promoter, named pKCR3 (see Methods), was constructed based on the multi-copy pKC1139 [30] and then introduced into S. globisporus C-1027 by conjugation. Thereafter, the resultant sgcR3 overexpression strains were fermented by incubation in liquid medium FMC-1027-1 (see Methods). The antibacterial bioassay against Bacillus subtilis CMCC(B) 63501 (data not shown) and the HPLC analysis indicated that the pKCR3 led to a 30-40% increase in C-1027 production (Fig. 3c) in comparison to that in wild type strain (Fig. 3b), whereas C-1027 production level detected in the wild type strain with the parental vector pKC1139 had no difference. Therefore, the result suggested that the function of sgcR3 could be positive for C-1027 biosynthesis in S. globisporus C-1027.

Inactivation and complementation of sgcR3
In order to ascertain the contribution of sgcR3 to the regulation of C-1027 biosynthesis, a part of coding region of sgcR3 (507 bp) was replaced with a thiostrepton resistant gene (tsr) to create the sgcR3 disrupted strain S. globisporus R3KO (Fig. 4A). Successful disruption of the intended tar-Comparison of two 9-membered enediyne (C-1027 and NCS) biosynthetic gene clusters around the genes of enediyne PKS (sgcE and ncsE) (A) and amino acid sequence alignment for SgcR3 (B) get was confirmed by PCR using primers complementary to one end of tsr and to untouched DNA outside the disruption constructs (data not shown). Southern blot analyses authenticated the site-specific disruptions of sgcR3 using left arm for crossover and deleted part of sgcR3 gene as probes respectively (Fig. 4B, 4C). The antibacterial bioassay against B. subtilis (Fig. 4D, b) and HPLC analysis (Fig. 3d) showed that disruption of sgcR3 completely abolished C-1027 production.
To confirm that the disruption of sgcR3 was indeed responsible for the abolition of C-1027 production, the mutant was complemented with sgcR3 gene. Three sgcR3 expression plasmids (pKCR3, pSETR3 and pLR3) were introduced into R3KO mutant by conjugation respectively. pSETR3 and pLR3, both based on the plasmid pSET152 [30] integrating into the ΦC31 attB site on the chromosome, had a copy of sgcR3 controlled by its native promoter and a strong constitutive promoter ermE*p respectively. The resultant strains with pKCR3 (Fig. 4D, c) and pSETR3 (Fig. 4D, d) restored the C-1027 production and showed dose proportionality as expected. The strain containing pLR3 in which sgcR3 was controlled by ermE*p showed less production of C-1027 (Fig. 4D, e) compared with the strain containing pSETR3. No production of C-1027 was detected for the R3KO mutants transformed with pKC1139 and pSET152 (data not shown). These results, fully consistent with those obtained upon overexpression of sgcR3 gene, confirmed the positive regulatory role of sgcR3 in C-1027 biosynthesis.

Gene expression analysis in R3KO mutant
To investigate the role of sgcR3 gene in transcriptional regulation of C-1027 biosynthetic gene cluster, the gene expression analysis was conducted by quantitative real time RT-PCR. The relative level of the transcripts of two other putative regulatory genes, sgcR1 and sgcR2, and two biochemically characterized structural genes, sgcA1 and sgcC4, were analysed together with sgcR3. The deduced product of sgcR1 displays 44% end-to-end identity to StrR, a well-characterized pathway-specific transcriptional activator for streptomycin biosynthesis in S. griseus [12]. SgcR2 shares high sequence identity (>40% along the whole length) to AraC/XylS family transcriptional regulators. SgcA1 and SgcC4 were reported to catalyze the first step in the biosynthesis of the deoxy aminosugar and the β-amino acid moieties of C-1027 chromophore respectively [31,32]. Total RNA from the wild type strain and R3KO mutant was extracted under which condition the wild type strain commenced C-1027 production at about 48 h growth on S5 agar. The cDNA was synthesized and then used as template in quantitative PCR. As expected, sgcR3 transcripts were almost undetectable in R3KO mutant while readily detectable in wild type strain. Tran-Determination of C-1027 production in sgcR3 overexpres-sion strain and disruption strain R3KO scripts of the other four genes described above were also readily detected in wild type strain, but were significant lower in the R3KO mutant (13-22% to their counterparts in wild type strain) (Fig. 5). The results indicated that at the onset of the production of C-1027, sgcR3 directly or indirectly controlled the expression of not only the structural genes that responsible for the biosynthesis of C-1027 chromophore but also the other two putative regulators situated in the same gene cluster.

In trans complementation of R3KO mutant with sgcR1R2
The sgcR1 and sgcR2 were two adjacent genes transcribed in the same direction with a gap of only 25 bp, suggesting that they were transcriptionally coupled within an operon. Confirmation that sgcR1 and sgcR2 were controlled by sgcR3 came from in trans complementation of R3KO mutant with sgcR1R2 (sgcR1 and sgcR2 genes). The amplified DNA fragment of sgcR1R2 associated with its native promoter was cloned into multi-copy pKC1139 directly or under control of ermE*p to give pKCR1R2 and pKCER1R2. These two plasmids were introduced into sgcR3 mutant after conjugal transfer from Escherichia coli. C-1027 production was partially restored when sgcR1R2 was overexpressed under the control of either the native promoter ( Fig. 6c) or ermE*p (Fig. 6d). C-1027 production was not detected in the R3KO mutants in which pKC1139 and pSET152 were introduced (Fig. 6e, 6f). The expression of sgcR1R2 functionally complemented the disruption of sgcR3, together with results of the gene expression analysis, verified that sgcR3 occupied the higher level than sgcR1R2 did in the regulatory cascade for C-1027 biosynthesis in S. globisporus C-1027.

Binding of SgcR3 to the sgcR1R2 promoter region
For further investigation of the function of sgcR3, its product was therefore expressed as an N-terminal His 10 fusion protein in E. coli (see Methods). Subsequent SDS-PAGE analysis revealed overproduction of a clone-specific protein of the expected size of His 10 -SgcR3 (45 kDa). This His 10 -tagged SgcR3 protein was purified from the soluble fraction of cell lysate by nickel affinity chromatography and was estimated by SDS-PAGE to be about 90% pure (Fig. 7A, lane 9).
To be a transcriptional activator of C-1027 biosynthesis, SgcR3 was speculated that it may act as a positive regulator by binding at or near the promoter region of biosynthetic genes or regulatory genes and thereby activating their transcription. EMSA were carried out to verify whether SgcR3 indeed had DNA-binding activity, using the purified His 10 -tagged SgcR3 and selected DNA fragments from the biosynthetic gene cluster of C-1027. Eight intergenic regions of interest are chosen for EMSA, including upstream region of sgcA1, sgcB1, sgcC1, sgcD2, sgcK, cagA, sgcR3 and sgcR1R2 (Fig. 7B). The results showed that the recombinant SgcR3 protein had binding activity to the Transcriptional analyses of different genes in S. globisporus C-1027 and R3KO mutant Figure 5 Transcriptional analyses of different genes in S. globisporus C-1027 and R3KO mutant. The relative abundance of sgcR1, sgcR2, sgcA1, sgcC4 and sgcR3 transcripts in mycelial patches of wild type strain and R3KO mutant grown on S5 agar plates for 48 h was determined using quantitative real time RT-PCR analysis. Data are from 2 biological samples with 2 determinations each. The values were normalized using values obtained for hrdB mRNA and represented as the mean ± SD. The amounts of each particular transcript in wild type strain were expressed as 1.
Determination of C-1027 production in R3KO mutant com-plemented with sgcR1R2 455 bp upstream fragment of the sgcR1R2, but not for any other of the eight DNA fragments investigated. Further EMSA carried out using different concentration of purified recombinant SgcR3 showed that the shift band emerged along with the increase of the protein amount. Shifting of the labelled probe was not observed when the corresponding unlabelled probes were added in excess to binding reaction (Fig. 7C). Specific binding of SgcR3 to the upstream fragment of the sgcR1R2 in vitro, together with the results of gene expression analysis and sgcR1R2 crosscomplementation in R3KO mutant, indicated that SgcR3 activates the transcription sgcR1R2 directly by binding to its promoter region.

Discussion
The original sequence analysis of the C-1027 biosynthetic gene cluster identified several ORFs whose gene products may have a potential regulatory function [25]. We focused our initial study on the sgcR3 gene situated at the right end of the cluster. Overexpression studies with additional copies of sgcR3 expressed under the control of its native promoter in wild type strain indicated a positive effect on C-1027 production. The results obtained in the gene disruption experiment clearly demonstrated the essential positive role of sgcR3 in regulation of C-1027 biosynthesis.
The results obtained in sgcR3 inactivation experiments were proved by complementation of the R3KO mutant using different strategies to express sgcR3 in trans. The results showed that expression of sgcR3 under the control of its native promoter either introduced by a multi-copy plasmid or integrated into the ΦC31 attB site on the chromosome fully restored C-1027 production. Unexpectedly, the complementation of sgcR3 under strong constitutive promoter ermE*p produced less C-1027 than under its native promoter, suggesting that the promoter region of sgcR3 was intricately regulated for its timing or the amount of expression which was important for the C-1027 production. One possibility is that there is a positive feedback mechanism controlling the expression of sgcR3, e.g., SgcR1 and/or SgcR2 can activate the expression of sgcR3 in return.
Analysis of gene expression in the mutant and wild type strain suggested that sgcR3 control C-1027 production through transcriptional regulation of biosynthetic genes. It also helped to establish a hierarchy among the three regulators of the C-1027 gene cluster. The expression level of sgcR1 and sgcR2 was significantly lower in R3KO mutant than in wild type strain, implying that sgcR3 occupied a higher rung than sgcR1 and sgcR2 did in the hierarchy of C-1027 regulatory genes. Only TylR among SgcR3 orthologues was characterized by gene disruption, in vivo complementation and gene expression experiments [14,23].
Overexpression of TylR was experimentally proved to increase tylosin yield by 60-70% [23]. According to these studies, TylR occupies the lowest level in the genetic hierarchy that controls tylosin production in S. fradiae, but that was probably not the case of SgcR3 for C-1027 production in S. globisporus C-1027.
Additional evidence for a correlation between these regulators of biosynthesis was observed through the study of cross-complementation experiment. The sgcR1R2 functionally complemented R3KO mutant under either its native promoter or strong constitutive promoter ermE*p. Furthermore, the recombinant SgcR3 protein bound specifically to the promoter region of sgcR1R2, but not that of sgcR3 and some structural genes detected. Therefore, it Gel mobility-shift assays of His 10 -SgcR3 with sgcR1R2 pro-moter region was very likely that SgcR3 activated the transcription of sgcR1 and sgcR2 by directly binding to their promoter region, to control the expression of biosynthetic structural genes indirectly. On the other hand, although the recombinant SgcR3 can bind to sgcR1R2 promoter region DNA fragment without further macromolecular factor in vitro, our results do not completely rule out the possibility that other protein(s) may be required for activating the transcription of sgcR1R2.
With few except that no regulatory gene present in the biosynthetic gene cluster, e.g., erythromycin cluster of Saccharopolyspora erythraea [33], most much-studied antibiotic gene clusters contain at least one pathway-specific regulator. However, the biosynthesis of more complex molecules may need more regulatory gene products involving a regulatory cascade to affect a positive or negative regulation. Some particularly interesting examples are the tylosin biosynthetic gene cluster of S. fradiae [14,18,19,[21][22][23] and the rapamycin biosynthetic gene cluster of S. hygroscopicus [16] which contain, remarkably, no fewer than five putative regulatory genes. Further analysis of other ORFs in C-1027 gene cluster revealed that additional three unknown genes might have regulatory role in C-1027 biosynthesis. The sgcE1 encodes a protein homologous (43% end-to-end identity) to a transcriptional regulator of HxlR family (GenBank accession no. ABX37987). The sgcR encodes a protein demonstrating some homology (20% end-to-end identity) to a transcriptional regulator protein (GenBank accession no. EDS60418) which belongs to XRE (Xenobiotic Response Element) family. The deduced product of sgcM was also found to be highly similar to a putative DNA-binding protein of S. coelicolor A3(2) with a helix-turn-helix motif (GenBank accession no. NP_630506.1). Both sgcE1 and sgcM have a highly homologous counterpart in NCS biosynthetic gene cluster of S. carzinostaticus. This is not surprising due to the complicated biosynthesis of enediyne chromophore, which involves multiple moieties and a convergent biosynthetic approach used to piece together the final product.
This work is the first step in deciphering the regulatory factors involved in the biosynthesis of C-1027, and a primary model for pathway-specific regulation of C-1027 production is shown in Fig. 8. Therefore, precise roles for sgcR3, sgcR1, sgcR2 and other putative regulatory genes and their complex interaction remain to be defined. The data presented in this work set the stage for subsequent studies to delineate the complexity of the regulation of C-1027 biosynthesis, as well as for designing strategies for the construction of strains with enhanced C-1027 production.

Strains, media and growth conditions
Hypothetical schematic regulatory hierarchy of C-1027 bio-synthesis in S. globisporus C-1027 CaCO 3 (pH 7.0)) for C-1027 production, in Trypticase Soy Broth (TSB, BD, Sparks, USA) for isolation of genomic or plasmid DNA, and maintained as mycelial fragments or spores in 20% (v/v) glycerol at -70°C. The antibiotics apramycin (Am) and thiostrepton (Th) were added at final concentrations of 50 and 30 μg ml -1 to solid medium, and at 10 and 5 μg ml -1 to liquid medium, respectively. Strains used and constructed in this study are listed in Table 1.

DNA manipulations
Standard genetic techniques with E. coli and in vitro DNA manipulations were as described by Sambrook & Russell [36]. Plasmids used and constructed in this study are listed in Table 1. Recombinant DNA techniques in Streptomyces species were performed as described by Kieser et al. [34]. PCR reactions for amplification of indicated DNA products and for verification of gene deletion in S. globisporus were carried out using Pfu DNA polymerase (Trans-Gen Biotech, Beijing, CN). The primers for PCR amplification are shown in Table 2. Total S. globisporus DNA was isolated using the Kirby mix procedure [34]. Southern blot analysis was performed on Hybond-N + nylon membrane (Amersham Biosciences, Buckinghamshire, UK) with a fluorescein-labelled probe by using the Gene Images Random Prime Labelling Module and CDP-Star Detection Kit (Amersham Biosciences).

Construction of expression plasmids
Three plasmids for sgcR3 expression were constructed as follows. The sgcR3 with its promoter region (2,539 bp) was amplified by PCR and then cloned into the E. coli/ Streptomyces shuttle vector pKC1139 [30] to give pKCR3. The fragment was also ligated into an integrative vector pSET152 [30] to give pSETR3. The sgcR3 coding region (1,188 bp) amplified by PCR was introduced to pL646 [37], displacing atrAc gene under the control of a strong constitutive promoter ermE*p, to give pLR3.
Similarly, sgcR1R2 (2,461 bp) with its promoter region were amplified by PCR and cloned into pKC1139 vector to yield pKCR1R2. This fragment was also cloned into pKC1139 under the control of ermE*p, resulting in plasmid pKCER1R2.

Disruption of sgcR3
The disruption construct consists of a thiostrepton resistant gene (tsr), sandwiched between two PCR products ("arms") that each contains sequence from sgcR3 plus flanking DNA. The arms (which were authenticated by sequence analysis) were of approximately equal size (1.4 kbp).
The primers for sgcR3 disruption introduced restriction sites into the arms (EcoRI and BglII in the upstream arm, BglII and HindIII in the downstream arm), and thus   allowed fusion at the BglII sites by ligation into pUC18. Then, the tsr fragment (a 1 kbp BclI restriction fragment from pIJ680 [34]) was introduced into the BglII site and thereby displaced 507 bp of sgcR3. Disrupted sgcR3 plus flanking DNA (approximate 3.8 kbp in total) was ligated into suicide plasmid pOJ260 [30] to give pOJR3KO. This plasmid was introduced by transformation into E. coli ET12567/pUZ8002 and then transferred into S. globisporus C-1027 by conjugation. Double-crossover exconjugants were selected on MS agar containing Th and Am (Th r , Am s ). Deletions within sgcR3 were confirmed by PCR and Southern blot hybridization.

Gene expression analysis by real time reverse transcriptase PCR (RT-PCR)
RNA was isolated from S. globisporus mycelia scraped from cellophane laid on the surface of S5 agar plates, treated with DNaseI (Promega, WI, USA) and quantitated as described previously [37,38]. The first strand synthesis of cDNA was performed with SuperScript III First-strand Synthesis System (Inivitrogen, CA, USA) using 2 μg total RNA and the random hexamers as primers following the manufacturer's instructions. Oligonucleotides were designed to amplify fragments of about 100-150 bp from the target genes (  [39].

C-1027 production and analysis
For C-1027 production, S. globisporus strains were grown in liquid medium FMC-1027-1 by a two-stage fermentation. The spore suspensions of the different strains were adjusted to the same concentration for inoculation. The seed inoculum was prepared by inoculating 100 ml of FMC-1027-1 with an aliquot of the spore suspension and incubating the mixture at 28°C and 250 rpm for 2 days.
The seed culture (5%) was added to a fresh 100 ml of FMC-1027-1, continuing to incubate at 28°C and 250 rpm for 5 days. To obtain statistically significant results, each strain was represented by a triplicate sample set. Dry weight of mycelia was measured in cultures taken at different time points in the fermentation course and the pattern of growth curves were monitored consistently among the relevant strains. The production of C-1027 was analyzed using the fermentation supernatants of relevant strains with the same growth curves.
C-1027 production was detected by assaying its antibacterial activity against B. subtilis [35]. The fermentation supernatant (180 μl) was added to stainless steel cylinders placed on F403 agar plate containing B. subtilis spores (0.4% v/v). C-1027 production was estimated by measuring the sizes of the inhibition zones after incubated at 37°C for 12 h.
Isolation and high-pressure liquid chromatography (HPLC) analysis of C-1027 chromophore were carried out mostly following Liu et al. [25]. Briefly, (NH 4 ) 2 SO 4 was added to the 250 ml fermentation supernatant of relevant strains to 100% saturation and then adjusted to pH 4.0 with 0.1 M HCl. The precipitated C-1027 chromoprotein was dissolved in 15 ml 0.1 M potassium phosphate (pH 8.0). The supernatant was then extracted with 50 ml ethyl acetate (EtOAc), concentrated in vacuum, and re-dissolved in 250 μl methanol. 25 μl cleared sample was subjected to HPLC on a Kromasil C-18 column (5 μm, 150 × 4.6 mm, Bohus, SE), eluted isocratically with 20 mM potassium phosphate (pH 6.86)/CH 3 CN (50:50, v/v) at a flow rate of 1.0 ml/min and detected by monitoring UV absorbance at 350 nm. The C-1027 enediyne chromophore standard for HPLC analysis was confirmed by ESI-MS.

Expression and purification of His 10 -tagged SgcR3
The sgcR3 coding sequence was PCR-amplified from S. globisporus C-1027 genome DNA containing an NdeI and BamHI restriction sites, and then ligated into pET-16b (Novagen, Madison, USA), authenticated by sequencing, and then transformed into the E. coli BL21 (DE3). For production of His 10 -tagged SgcR3, cultures (800 ml; OD 600 = 0.6) were induced with IPTG (0.05 mM final), incubated at 28°C for 6 h, harvested by centrifugation. The cell suspension was sonicated for 60 × 10 s with 10 s intervals between each treatment in 30 ml lysis buffer (50 mM NaH 2 PO 4 , pH 8.0, 300 mM NaCl, 10 mM imidazole, 2 mg lysozyme ml -1 ). Cellular debris was removed by centrifugation (12,000 rpm for 10 min). His 10 -tagged SgcR3 was then affinity purified using HisTrap™ FF crude (Amersham Biosciences) according to the manufacturer's directions and fractions eluted from the column were analysed on SDS-12% w/v polyacrylamide gels. Those fractions containing recombinant protein were pooled, dialysed overnight at 4°C against dialysis buffer (25 mM Tris/HCl (pH 7.5), 10% (w/v) glycerol, 2 mM DTT) and stored at -70°C. The BCA™ Protein Assay Kit (Pierce Biotechnology, Rockfold, USA) was used for protein quantification with bovine serum albumin as the standard.