HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 43, 33443-33448, October 27, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/43/33443    most recent M006766200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Marti, T. Articles by Khosla, C. Search for Related Content PubMed PubMed Citation Articles by Marti, T. Articles by Khosla, C. Social Bookmarking                      What's this?

Cloning, Nucleotide Sequence, and Heterologous Expression of the Biosynthetic Gene Cluster for R1128, a Non-steroidal Estrogen Receptor Antagonist

INSIGHTS INTO AN UNUSUAL PRIMING MECHANISM^*

Thomas Marti, Zhihao Hu, Nicola L. Pohl^§, Aseema N. Shah, and Chaitan Khosla^¶

From the Departments of Chemical Engineering and ^¶ Chemistry and Biochemistry, Stanford University, Stanford, California 94305-5025

Received for publication, July 28, 2000, and in revised form, August 6, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

R1128 substances are anthraquinone natural products that were previously reported as non-steroidal estrogen receptor antagonists with in vitro and in vivo potency approaching that of tamoxifen. From a biosynthetic viewpoint, these polyketides possess structurally interesting features such as an unusual primer unit that are absent in the well studied anthracyclic and tetracyclic natural products. The entire R1128 gene cluster was cloned and expressed in Streptomyces lividans, a genetically well developed heterologous host. In addition to R1128C, a novel optically active natural product, designated HU235, was isolated. Nucleotide sequence analysis of the biosynthetic gene cluster revealed genes encoding two ketosynthases, a chain length factor, an acyl transferase, three acetyl-CoA carboxylase subunits, two cyclases, two oxygenases, an amidase, and remarkably, two acyl carrier proteins. Feeding studies indicate that the unusual 4-methylvaleryl side chain of R1128C is derived from valine. Together with the absence of a dedicated ketoreductase, dehydratase, or enoylreductase within the R1128 gene cluster, this suggests a functional link between fatty acid biosynthesis and R1128 biosynthesis in the engineered host. Specifically, we propose that the R1128 synthase recruits four subunits from the endogenous fatty acid synthase during the biosynthesis of this family of pharmacologically significant natural products.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Non-steroidal estrogen receptor (ER)¹ antagonists such as tamoxifen are the first choice of anti-hormone treatment for breast cancer. However, ER antagonists have variable effects in different target tissues. For example, tamoxifen is an ER antagonist in breast tissue but an ER agonist in bone (1) and uterine (2) tissue. Raloxifene is also an ER antagonist in breast tissue; however, it exerts agonist activity in bone but not uterine tissue (3). Recent studies have suggested that this tissue-dependent variability may be caused by differential ligand activation of ER and ER at AP1 transcription activation sites (4). There is thus a significant need for new ER antagonists that can inhibit ER function at both ER- and AP1-dependent DNA binding sites.

The isolation, fermentation, structures, and properties of R1128 A-D (Fig. 1) were described in 1993 (5, 6). These molecules, which were identified as potent ER antagonists, had IC₅₀ values in the range of 0.1-0.3 µM. Although these activities were approximately 7-17-fold less than that of tamoxifen, their selectivity (50-fold better binding to ER as compared with the androgen receptor) and low toxicity (significant inhibition of tumor volume in a subrenal capsule assay at 100 mg/kg; no acute toxicity in mice or rats at 500 mg/kg) profiles were promising (7). Thus, the development of rational approaches to manipulate the structures of these natural products offers an opportunity to optimize their pharmacological properties. As a first step in this direction, we describe the cloning and nucleotide sequence of the R1128 biosynthetic gene cluster and its expression in a model heterologous host, Streptomyces lividans K4-114 (8). We also present evidence that highlights an unusual mechanism for priming this polyketide synthase (PKS).

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Structures of the R1128 compounds.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Construction of Cosmid Vector pHU204-- SCP2* is a widely used low-copy vector for Streptomyces sp. (9). Although cosmid derivatives of SCP2* have been described earlier (10), their ability to replicate in Streptomyces sp. in a stable manner has not been tested. Plasmid pHU204 (Fig. 2) is a derivative of pRM5 (11). To make it suitable for library construction, a series of cloning steps were used to replace the 10.5-kb HindIII/EcoRI segment of pRM5 with the following cos site containing a 390-base pair sequence (NsiI and EcoRI sites (Fig. 2) are underlined): AAGCTAATTC ATAGATGCAT GTTTGACCGC TTATCATCGA TAAGCTCTGC TTTTTGTTGA CTTCCATTGT TCATTCCACG GACAAAAACA GAGAAAGGAA ACGACAGAGG CCAAAAAGCT CGCTTTCAGC ACCTGTCGTT TCCTTTCTTT TCAGAGGGTA TTTTAAATAA AAACATTAAG TTATGACGAA GAAGAACGGA AACGCCTTAA ACCGGAAAAT TTTCATAAAT AGCGAAAACC CGCGAGGTCG CCGCCCCGTA ACAAGGCGGA TCGCCGGAAA GGACCCGCAA ATGATAATAA TTATCAATTG CATACTATCG ACGGCACTGC TGCCTGCATG GGTCTAGAAA AAATCCGGAC CCACTAGTCC CAGATCTAAA GCTAGAATTC.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Physical map of a new shuttle cosmid vector for Streptomyces.

Library Construction, Transformation, and DNA Sequencing-- A cosmid library was constructed in pHU204 using genomic DNA from Streptomyces sp. R1128. The average insert size was greater than 30 kb. The entire library was transformed into S. lividans K4-114 without further screening. All recombinant DNA procedures and Streptomyces cloning procedures were according to standard protocols (9, 12). DNA sequencing was performed using standard automated procedures. The nucleotide sequence of the R1128 biosynthetic gene cluster has been deposited in GenBank^TM/EBI (accession number AF293442).

Isolation of Compounds-- Cultures of S. lividans were grown on R2YE agar medium (9) for 7 days and extracted with ethyl acetate. The extract was dried with MgSO₄, evaporated, and subjected to column chromatography (SiO₂, hexane/ethyl acetate, 1:1). The yellow fractions were dried and further separated on a preparative reverse-phase HPLC column ((25 × 150)-mm C₁₈-IP column (Beckman); isocratic run with 1% acetic acid in water/CH₃CN 55:45; flow rate, 20 ml/min; UV-detection with  of 250 and 350 nm). Three major UV-active peaks (t_R = 10.2, 22.5, and 28.3 min) were collected, dried, and analyzed.

For the preparation of ¹³C-labeled polyketides, cultures were grown on R2YE agar medium supplemented with 500 mg/liter sodium [1,2-¹³C]acetate for 7 days. Products were isolated as described above. Enrichment levels of 0.6% were observed. For supplementation experiments with amino acids, cultures of S. lividans were grown in liquid SMM (13) with 1% valine or leucine added at 30 °C for 8 days. Extraction of compounds as described above was followed by quantitative analysis by HPLC (Beckman (4 × 250)-mm C₁₈-IP reverse-phase column; gradient: 1% acetic acid in water/CH₃CN, 70:30 to 30:70 (45 min); flow rate, 1.5 ml/min; UV-detection with  of 250 and 350 nm). The retention times of the compounds were 22.1 min for HU235, 26.0 min for R1128B, and 30.1 min for R1128C.

Spectral Data of HU235-- []_D = 33.0° (C, 0.63 in CH₃OH); ¹H NMR (500 MHz, (CD₃)₂SO):  = 16.10 (s, 1 H-O-C-13)), 10.23 (s, 1 H-O-C-9), 9.81 (s, 1 H-O-C-11), 4.84 (s, 1 H-O-C-3), 2.95 (2 d, J = 16.3, 2 H-C-4), 2.58 (dd, J = 8.1, 3.0, 1 H-C-16), 1.75 (m, 1 H-C-18), 1.64 (m, 1 H-C-17)), 1.57 (m, 1 H-C-17), 1.25 (s, 3 H-C-2), 0.95 (d, J = 6.4, 3 H-C-19), 0.91 (d, J = 6.6, 3 H-C-19). ¹³C NMR (125 MHz, (CD₃)₂SO, (coupling constants of ¹³C-enriched material, 100 MHz, (CD₃)₂SO)):  = 206.4 (C-15, J_{C, C} = 39.9), 165.3 (C-13, J_{C, C} = 63.0), 161.3 (C-9, J_{C, C} = 65.0), 159.1 (C-11, J_{C, C} = 62.5), 140.8 (C-7, J_{C, C} = 58.2), 136.2 (C-5, J_{C, C} = 64.9), 115.9 (C-6, J_{C, C} = 64.9), 106.9 (C-14, J_{C, C} = 63.0), 106.4 (C-12, J_{C, C} = 62.5), 101.7 (C-8, J_{C, C} = 58.2), 101.1 (C-10, J_{C, C} = 65.0), 71.2 (C-3, J_{C, C} = 36.8), 55.1 (C-16, J_{C, C} = 39.9), 41.2 (C-4, J_{C, C} = 36.8), 34.4 (C-17), 28.0 (C-2), 26.5 (C-18), 22.5 (C-19), 21.8 (C-19); IR (KBr): ~ = 3448m, 2956m, 2553w, 1638s, 1597s, 1458w, 1420w, 1384w, 1163m, 1095w, 917m, 848m.

Acetylation of HU235-- HU235 (5 mg) was stirred in 1 ml of pyridine/Ac₂O (3:2) for 2 h. The mixture was poured into 2 N HCl and extracted with ether. Two products were isolated after chromatographic separation (hexane/ethyl acetate, 3:2). Diacetate of HU235: ¹H NMR (400 MHz, CDCl₃):  = 7.33 (d, J = 2.1, 1 H), 7.02 (s, 1 H), 6.87 (d, J = 2.1, 1 H), 3.21 (d, J = 16.5, 1 H), 3.00 (d, J = 16.5, 1 H), 2.60 (dd, J = 8.3, 4.0, 1 H), 2.38 (s, 3 H), 2.33 (s, 3 H), 1.78 (m, 1 H), 1.66-1.53 (2 m, 2 H), 1.37 (s, 3 H), 0.99 (d, J = 6.4, 3 H), 0.93 (d, J = 6.5, 3 H); MS (FAB⁺): m/z (%): 459 (15) [M  H+2 Na⁺], 437 (100) [M + Na⁺]; 415 (20) [MH⁺]; HRMS (FAB⁺) calculated for C₂₃H₂₆O₇ Na⁺ 437.1576, found 437.1589. HU235Ac₃: ¹H-NMR (400 MHz, CDCl₃):  = 7.57 and 7.55 (2 s, 1 H), 7.48 and 7.45 (2 d, J = 2.1, 1 H), 6.95 and 6.93 (2 d, J = 2.1, 1 H), 3.31 (m, 1 H), 3.15 (m, 1 H), 2.58 (m, 1 H), 2.47 and 2.45 (2 s, 3 H), 2.40 and 2.38 (2 s, 3 H), 2.33 and 2.31 (2 d, 3 H), 1.65-1.51, 1.78 (m, 1 H), 1.66-1.53, 1.42-1.34 (2 m, 3 H), 1.36 and 1.34 (2 s, 3 H), 0.96-0.84 (m, 6 H); MS (FAB⁺): m/z (%): 479 (100) [M + Na⁺]; HRMS (FAB⁺): calculated for C₂₅H₂₈O₈ Na⁺ 479.1682, found 479.1695.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Construction of a New Cosmid Shuttle Vector for Streptomyces-- A minimal prerequisite for biosynthetic engineering of natural products is the availability of cloned genes, which are typically clustered and coordinately regulated in microorganisms. An ideal strategy for cloning a desired gene cluster would involve its direct identification through heterologous expression in a biologically well studied host. Building on the pioneering studies of Malpartida and Hopwood (14), recent studies have successfully demonstrated the ability of the model organism, S. lividans, to produce heterologous aromatic polyketides (15, 16). However, because the site-specific integrative plasmids used in such efforts could not be readily recovered from transformants of interest, genomic libraries had to be pre-screened for candidate cosmids carrying putative polyketide genes prior to their transfer into S. lividans. There is thus a need for new self-replicating shuttle vectors with carrying capacity for large DNA fragments. We developed a vector pHU204 that has such properties. Plasmid pHU204 (12.16 kb, Fig. 2) is a derivative of pRM5 (11). It includes the SCP2* replicon and the colE1 replicon, the thiostrepton (tsr) and carbenicillin (bla) drug resistance genes, and a cos site. Whereas its carrying capacity as a cosmid cloning vector in Escherichia coli is limited to a 40-kb insert, the SCP2* replicon has been used to construct up to 70 kb plasmids in S. lividans.²

Cloning and Heterologous Expression of the R1128 Gene Cluster-- Anticipating that the R1128 gene cluster was under 40 kb in size, a cosmid library was constructed in pHU204 using genomic DNA derived from Streptomyces sp. R1128, the R1128 producer. The entire library was transformed without prior screening into S. lividans K4-114. Several pigmented colonies were identified; one of them, K4-114/pHU235 was cultivated on a larger scale (500 ml on semi-solid R5 agar medium). After neutral extraction and pre-purification using column chromatography, two major UV-active peaks were collected from reverse-phase HPLC, dried, and analyzed. Mass-, ¹H and ¹³C NMR spectra revealed that one of the compounds was identical to R1128C. The purified yield of the compound was 1 mg/liter. The second compound, designated HU235, was produced in considerably greater quantities (16 mg/liter). A combination of ¹H, ¹³C, ¹H,¹H COSY, HMBC, HMQC, ROESY-NMR, and IR spectroscopy was used to deduce that HU235 was a novel tetrahydroanthracenone derivative, 3,6,8,9-tetrahydroxy-2-isobutyl-3-methyl-1,2,3,4-tetrahydro-1-anthracenone (Fig. 3; note that the carbon atom numbering in the figure is intended to facilitate tracing of the polyketide backbone, and not as per IUPAC recommendations). Mass spectrometry was performed with the acetylated derivatives. Although nuclear Overhauser effect NMR measurements did not allow unambiguous determination of the relative configurations of the two chiral centers of the molecule, optical rotation analysis revealed that the compound HU235 was optically active ([]_D = 33.0°). This suggests that an enzyme-catalyzed reaction might be involved in the final steps of HU235 formation (see below).

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Structure and acetate labeling pattern for HU235. Bold bonds indicate multicarbon units derived from the incorporation of doubly labeled acetate.

Because the carbon chain backbone of HU235 is identical to that of R1128C, the polyketide origin of this product was established by the feeding of sodium [1,2-¹³C]acetate (Fig. 3). The carbon skeleton comprised of C-2 to C-14 was labeled as expected. (C-1 is eliminated via decarboxylation.) Interestingly, the incorporation efficiency of labeled acetate into the C-15-C-16 multicarbon unit was comparable with that in the remaining multicarbon units of HU235, suggesting that the 4-methylvaleryl moiety is contemporaneously synthesized with the rest of the polyketide skeleton. If so, this raises the question as to the mechanism of reduction of the C-17 carbon atom (discussed below).

Nucleotide Sequence Analysis of the Biosynthetic Gene Cluster-- Cosmid pHU235 was initially sequenced to about 3-4 × coverage using a shotgun approach. Based on this data a 17-kb region of the cosmid was identified that appeared to contain all the biosynthetic genes; accurate sequence of this region was obtained through a combination of subcloning and oligonucleotide walking. The results of sequence analysis of this segment of pHU235 are graphically presented in Fig. 4 and summarized in Table 1. 

View larger version (7K): [in this window] [in a new window]   Fig. 4.   Organization of the R1128 biosynthetic gene cluster. For details regarding individual genes and their products, see Table 1 and text.

Fourteen genes involved in R1128 biosynthesis were identified. Their putative functions were assigned based on homology with aromatic PKS genes of known function (17, 18). As in the case of daunorubicin biosynthesis (19), zhuC and zhuH encode an acyltransferase and a primer unit-specific ketosynthase, respectively. Homologs of these genes are found in aromatic PKS gene clusters that utilize primers derived from sources other than malonyl-CoA. The polyketide backbones of R1128A, R1128B, and R1128C are presumably primed by acetyl-CoA, propionyl-CoA, and isobutyryl-CoA, respectively. Together with an acyl carrier protein (ACP; encoded by either zhuG or zhuN), a ketoreductase, a dehydratase, and an enoylreductase, ZhuC and ZhuH are predicted to generate an ACP-bound intermediate (butyryl-ACP, valeryl-ACP, or 4-methylvaleryl-ACP, respectively) (Fig. 5). This intermediate is then transferred onto the ketosynthase-chain length factor (KS-CLF) heterodimer, encoded by zhuA and zhuB, respectively. Together with an ACP (encoded by either zhuG or zhuN) and a malonyl-CoA:ACP malonyltransferase (MAT), which is recruited from the endogenous fatty acid synthase, the KS-CLF heterodimer catalyzes seven additional extension cycles to produce the full-length polyketide chain shown in Fig. 5 (20). This reactive intermediate subsequently undergoes stagewise cyclization and aromatization, catalyzed by the zhuI and zhuJ gene products, to yield the bicyclic intermediate shown in Fig. 5 (21, 22). The final cyclization leading to the formation of the anthracycline core of the R1128 compounds is presumably non-catalytic, a feature that is precedented in the biosynthesis of aloesaponarin II (11, 23). Two homologous cytochrome P450 type oxygenases are encoded by zhuK and zhuM. Homologs of these enzymes are present in several other aromatic polyketide biosynthetic pathways. For example, urdE encodes a gene responsible for the hydroxylation of tetracenomycin to 6-hydroxytetracenomycin (24); a similar reaction would be required for anthraquinone formation in the R1128 biosynthetic pathway (Fig. 5). The reason for the presence of two such enzymes in the R1128 gene cluster is unknown, and perhaps reflects redundancy in function.

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Proposed biosynthetic pathway for R1128 substances and HU235.

In addition to the above biosynthetic enzymes, the R1128 gene cluster also encodes two other unusual enzymes. First, it contains a homolog (zhuL) of a class of known amidases, such as an enantioselective amidase from a Rhodococcus strain that hydrolyzes 2-arylpropionamides (25). The role of ZhuL in R1128 biosynthesis is unclear, but this family of amidases also shares similarity to the GatA subunit of glutamyl-tRNA^Gln amidotransferases (e.g. the Deinococcus radiodurans enzyme (26)), which is responsible for amide hydrolysis as well as transfer of the resulting amine to glutamyl-tRNA. Perhaps the conserved aminotransferase activity in ZhuL allows it to function as a valine deaminase, an enzyme that is predicted to be involved in conversion of valine into isobutyryl-CoA required for R1128C biosynthesis (see below). Second, the gene cluster encodes three polypeptides, ZhuD, ZhuE, and ZhuF that are homologous to acetyl-CoA carboxylase subunits. ZhuD is predicted to be a biotin carboxylase, ZhuE is predicted to be a biotin carboxyl carrier protein, whereas ZhuF is homologous to carboxyl transferase subunits. Recent studies on jadomycin biosynthesis have revealed the presence of zhuD and zhuE homologs (but not a zhuF homolog) in the jadomycin gene cluster (27). Genetic inactivation of these subunits leads to significant reduction in jadomycin biosynthesis, underscoring the role of this multisubunit enzyme in the biosynthesis of malonyl-CoA precursors for polyketide biosynthesis.

Perhaps the most interesting feature of the R1128 gene cluster is the presence of not one but two ACP genes. This feature is absent in biochemically characterized PKS pathways such as actinorhodin (28) or daunorubicin (29, 30), but is present in the frenolicin PKS (31) (see also GenBank 228 /EBI accession number AAC18105). As in R1128A, the frenolicin polyketide backbone is presumably derived by transfer of a butyryl unit onto the KS-CLF heterodimer (32). The mechanistic implications of a dual ACP aromatic PKS system are discussed below. We simply note here that ZhuN is more similar to FrnN (58% identity with FrnN versus 40% identity with FrnJ) (31), whereas ZhuG is more similar to FrnJ (46% identity with FrnJ versus 40% identity with FrnN, GenBank 228 /EBI accession number AAC18105).

Flanking the zhuI locus are three regulatory genes and three genes encoding membrane-bound proteins that are presumably involved in transcriptional control and product export, respectively. Although their nucleotide sequence has not been determined in an error-free form, sufficient coverage was obtained to deduce their functions based on homology (data not shown). No open reading frame could be detected within approximately 1 kb of the zhuA gene, suggesting that zhuA represents the other end of the R1128 gene cluster.

Establishing the Biosynthetic Origin of the Branched-chain Alkyl Moiety of R1128C and HU235-- The structures of both HU235 and R1128C suggest the formation of an early 4-methylvaleryl intermediate. Similar branched-chain moieties are present in several Streptomyces-derived fatty acids and polyketides and are derived from the catabolism of branched-chain amino acids (33). To probe the origin of this unusual starter unit, S. lividans K4-114/pHU235 was grown for 8 days in defined SMM liquid medium (13) in the presence of leucine or valine. The titers of HU235 and R1128C, quantified via HPLC, are shown in Table 2. Although overall production levels are lower in SMM medium than in R2YE medium (reported above), productivity is dramatically enhanced upon valine supplementation, confirming that the 4-methylvaleryl moiety of R1128C is derived from valine. Oxidation of valine to yield isobutyryl-CoA is known to involve elimination of the carboxyl group as carbon dioxide (33). Consistent with this notion, [1-¹³C]valine did not result in enhancement of any peak in the ¹³C NMR spectrum of HU235.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

A number of bacterial aromatic PKS gene clusters have been cloned and sequenced thus far (17). Genetic, biochemical, and chemical studies on these systems have shed light on a variety of mechanistic aspects of these multienzyme systems. The core PKS is known to be comprised of four components: a KS-CLF heterodimer, an ACP, and an MAT (20). Of these, the KS, CLF, and ACP are typically encoded by open reading frames within the dedicated polyketide gene cluster, whereas the MAT is recruited from the endogenous fatty acid synthase. Together these four enzymes catalyze the iterative decarboxylative condensation of malonyl-CoA-derived building blocks into an enzyme-bound poly--ketone backbone of defined length. Although direct physical evidence for the association of the KS-CLF, ACP, and MAT is lacking, kinetic studies on this "minimal" PKS system have yielded evidence for the role of protein-protein interactions in chain elongation (34, 35). Sequence analysis suggests that ZhuB (the KS), ZhuA (the CLF), ZhuN (the ACP), and the MAT from the Streptomyces fatty acid synthase are the constituents of the R1128 minimal PKS (Fig. 5). Although the precise difference (if any) between the physiological roles of ZhuG and ZhuN remains to be established, inclusion of ZhuN rather than ZhuG in the minimal PKS is based on its closer evolutionary relationship to well characterized ACPs from minimal PKSs such as the actinorhodin, frenolicin, or daunorubicin PKSs.

In most cases the minimal PKS defined above is primed via decarboxylation of a malonyl unit to yield an acetyl-S-KS intermediate that is subsequently elongated. However, some bacterial aromatic PKSs are primed by alternative primer units. For example, daunorubicin is derived from a propionyl primer unit. Such systems possess two additional PKS components, a priming KS dedicated solely to the first condensation cycle and an acyltransferase. In the daunorubicin PKS, these proteins are responsible for converting malonyl-S-ACP into -ketopentanoyl-S-ACP, which is subsequently transferred onto the active site of the KS-CLF of the minimal PKS. By analogy, ZhuC, ZhuH, and malonated ZhuG produce either an acetoacetyl-S-ACP, -ketopentanoyl-S-ACP, or 2-oxo-4-methylpentanoyl-S-ACP (Fig. 5). However, unlike the daunorubicin example, these ACP-bound intermediates must be reduced into butyryl-S-ACP, valeryl-S-ACP, or 4-methylvaleryl-S-ACP, respectively, before transfer to the KS-CLF heterodimer. A similar situation is also encountered in the context of bacterial fatty acid biosynthesis (36), where FabH (a ZhuH homolog) catalyzes the first chain extension step, followed by the successive action of a ketoreductase, a dehydratase, and an enoylreductase to reduce the -ketothioester. Because homologs of the ketoreductase, dehydratase, and enoylreductase are absent from the R1128 gene cluster, we anticipate that the R1128 synthase recruits these enzymes from the endogenous fatty acid synthase (Fig. 5). This is analogous to the well established recruitment of the MAT by the minimal PKS (20), and may also occur in the case of frenolicin biosynthesis, as suggested by the DNA sequence of the gene cluster and the structure of the polyketide product. Perhaps this explains the need for two ACPs rather than one in systems that interact with two different sets of ketosynthases, acyltransferases, and auxiliary enzymes. If so, then this subclass of PKSs might present an attractive model system for studying the role of ACP-mediated protein-protein interactions in multienzyme assemblies (37). Moreover, it would represent yet another example of physiological cross-talk between polyketide and fatty acid biosynthesis, as originally proposed in the context of actinorhodin biosynthesis in Streptomyces coelicolor (38).

Upon synthesis of the full-length polyketide chain through the action of ZhuA, ZhuB, ZhuC, ZhuG, ZhuH, ZhuN, and four components of the fatty acid synthase, this highly reactive intermediate must undergo controlled cyclization and oxidation into the R1128 series of products. The first regiocontrolled cyclization occurs via an aldol condensation between the C-7 carbonyl atom and the acidic C-12 methylene atom of the nascent polyketide chain and is followed by enolization and the elimination of a water molecule to generate an aromatic ring system. These reactions are presumably catalyzed by the ZhuI aromatase/cyclase (Fig. 5). Both monodomain and didomain members of this protein family are known (39). Didomain orthologs are encountered in PKSs that produce compounds in which the C-9 carbonyl undergoes reduction, whereas monodomain subunits are involved in the biosynthesis of unreduced polyketides. The fact that ZhuI is a monodomain protein is consistent with the lack or reduction of the C-9 carbonyl atom.

The second aldol condensation between C-5 and C-14 is presumably catalyzed by ZhuJ, whose homologs include DpsY from the daunorubicin pathway (Fig. 5). Once this bicyclic aromatic system is formed, two fates are possible (Fig. 5). On one hand, the intermediate can undergo cyclization between C-2 and C-15, followed by C-6 oxygenation and decarboxylation, to afford the R1128 series of compounds. On the other hand, decarboxylative elimination of C-1, followed by cyclization between C-3 and C-16 can yield the novel compound HU235 reported here. The precise sequence of reactions in either path cannot be predicted, although ZhuK and/or ZhuM are presumably involved in the pathway to R1128. It might therefore be suspected that HU235 is a shunt product because of inadequate activity of these enzymes in vivo. However, the optical activity of HU235 suggests an enzymatic route for this unusual metabolite. Further studies will be required to decipher the biochemical course of these divergent pathways.

Recent studies on bacterial aromatic polyketide biosynthesis have demonstrated considerable promise with respect to altering four aspects of product structure: primer unit incorporation, chain length, degree and regiochemistry of reduction, and regiochemistry of cyclization (40-42). Most of these features are potentially attractive targets for manipulation in the R1128 family of ER antagonists. For example, the incorporation of an additional ketide unit into the polyketide backbone could lead to the introduction of an aliphatic ketone functionality in place of the variable alkyl group. Likewise, elimination of the C-9 hydroxyl has been accomplished in similar systems by co-expression of a suitable ketoreductase with regiospecificity toward the C-9 carbonyl of the nascent polyketide backbone. Finally, of particular interest is the unusual and variable functional group in the R1128 anthraquinone series, which is presumably derived from alternative primer unit incorporation in the polyketide backbone. Recent studies have localized the observed specificity of the daunorubicin PKS toward propionyl primer units to the DpsC subunit, whose homolog in the R1128 system is ZhuH (42). Thus, a better understanding of the relaxed substrate specificity of the ZhuH ketosynthase could lead to new opportunities for the rational design of R1128 analogs.

	ACKNOWLEDGEMENTS

We are grateful to Fujisawa Research Laboratories for generously providing us with Streptomyces sp. R1128. We thank Dr. John R. Carney of Kosan Biosciences, Inc. for obtaining the two-dimensional NMR spectra of HU235.

	FOOTNOTES

^* This research was supported in part by Grant CA77248 from the National Institutes of Health.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank^TM/EMBL Data Bank with accession number(s) AF293442.

Recipient of postdoctoral fellowships from the Swiss National Science Foundation, the Roche Research Foundation, and the Novartis Foundation.

^§ Recipient of National Institutes of Health Postdoctoral Fellowship 1 F32 GM19540.

To whom correspondence should be addressed. Tel./Fax: 650-723-6538; E-mail: ck@chemeng.stanford.edu.

Published, JBC Papers in Press, August 7, 2000, DOI 10.1074/jbc.M006766200

² Z. Hu, unpublished results.

	ABBREVIATIONS

The abbreviations used are: ER, estrogen receptor; s, singlet; d, doublet; m, multiplet; kb, kilobases; HPLC, high pressure liquid chromatography; ACP, acyl carrier protein; PKS, polyketide synthase; KS-CLF, ketosynthase-chain length factor; MAT, malonyl-CoA:ACP malonyltransferase.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				W. Zhang, Y. Li, and Y. Tang Engineered biosynthesis of bacterial aromatic polyketides in Escherichia coli PNAS, December 30, 2008; 105(52): 20683 - 20688. [Abstract] [Full Text] [PDF]

				C. P. Ridley, H. Y. Lee, and C. Khosla Chemical Ecology Special Feature: Evolution of polyketide synthases in bacteria PNAS, March 25, 2008; 105(12): 4595 - 4600. [Abstract] [Full Text] [PDF]

				X. Zhang, L. B. Alemany, H.-P. Fiedler, M. Goodfellow, and R. J. Parry Biosynthetic Investigations of Lactonamycin and Lactonamycin Z: Cloning of the Biosynthetic Gene Clusters and Discovery of an Unusual Starter Unit Antimicrob. Agents Chemother., February 1, 2008; 52(2): 574 - 585. [Abstract] [Full Text] [PDF]

				W. Zhang, B. D. Ames, S.-C. Tsai, and Y. Tang Engineered Biosynthesis of a Novel Amidated Polyketide, Using the Malonamyl-Specific Initiation Module from the Oxytetracycline Polyketide Synthase Appl. Envir. Microbiol., April 1, 2006; 72(4): 2573 - 2580. [Abstract] [Full Text] [PDF]

				A. S. Rahman, J. Hothersall, J. Crosby, T. J. Simpson, and C. M. Thomas Tandemly Duplicated Acyl Carrier Proteins, Which Increase Polyketide Antibiotic Production, Can Apparently Function Either in Parallel or in Series J. Biol. Chem., February 25, 2005; 280(8): 6399 - 6408. [Abstract] [Full Text] [PDF]

				K. Ichinose, M. Ozawa, K. Itou, K. Kunieda, and Y. Ebizuka Cloning, sequencing and heterologous expression of the medermycin biosynthetic gene cluster of Streptomyces sp. AM-7161: towards comparative analysis of the benzoisochromanequinone gene clusters Microbiology, July 1, 2003; 149(7): 1633 - 1645. [Abstract] [Full Text] [PDF]

				D. Baitsch, C. Sandu, R. Brandsch, and G. L. Igloi Gene Cluster on pAO1 of Arthrobacter nicotinovorans Involved in Degradation of the Plant Alkaloid Nicotine: Cloning, Purification, and Characterization of 2,6-Dihydroxypyridine 3-Hydroxylase J. Bacteriol., September 15, 2001; 183(18): 5262 - 5267. [Abstract] [Full Text] [PDF]

				B. A. Pfeifer and C. Khosla Biosynthesis of Polyketides in Heterologous Hosts Microbiol. Mol. Biol. Rev., March 1, 2001; 65(1): 106 - 118. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 275/43/33443    most recent M006766200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar