Investigation of Early Tailoring Reactions in the Oxytetracycline Biosynthetic Pathway
2007; Elsevier BV; Volume: 282; Issue: 35 Linguagem: Inglês
10.1074/jbc.m703437200
ISSN1083-351X
AutoresWenjun Zhang, Kenji Watanabe, Clay C. C. Wang, Yi Tang,
Tópico(s)Fungal Biology and Applications
ResumoTetracyclines are aromatic polyketides biosynthesized by bacterial type II polyketide synthases. The amidated tetracycline backbone is biosynthesized by the minimal polyketide synthases and an amidotransferase homologue OxyD. Biosynthesis of the key intermediate 6-methylpretetramid requires two early tailoring steps, which are cyclization of the linearly fused tetracyclic scaffold and regioselective C-methylation of the aglycon. Using a heterologous host (CH999)/vector pair, we identified the minimum set of enzymes from the oxytetracycline biosynthetic pathway that is required to afford 6-methylpretetramid in vivo. Only two cyclases (OxyK and OxyN) are necessary to completely cyclize and aromatize the amidated tetracyclic aglycon. Formation of the last ring via C-1/C-18 aldol condensation does not require a dedicated fourth-ring cyclase, in contrast to the biosynthetic mechanism of other tetracyclic aromatic polyketides, such as daunorubicin and tetracenomycin. Acetyl-derived polyketides do not undergo spontaneous fourth-ring cyclization and form only anthracene carboxylic acids as demonstrated both in vivo and in vitro. OxyF was identified to be the C-6 C-methyltransferase that regioselectively methylates pretetramid to yield 6-methylpretetramid. Reconstitution of 6-methylpretetramid in a heterologous host sets the stage for a more systematic investigation of additional tetracycline downstream tailoring enzymes and is a key step toward the engineered biosynthesis of tetracycline analogs. Tetracyclines are aromatic polyketides biosynthesized by bacterial type II polyketide synthases. The amidated tetracycline backbone is biosynthesized by the minimal polyketide synthases and an amidotransferase homologue OxyD. Biosynthesis of the key intermediate 6-methylpretetramid requires two early tailoring steps, which are cyclization of the linearly fused tetracyclic scaffold and regioselective C-methylation of the aglycon. Using a heterologous host (CH999)/vector pair, we identified the minimum set of enzymes from the oxytetracycline biosynthetic pathway that is required to afford 6-methylpretetramid in vivo. Only two cyclases (OxyK and OxyN) are necessary to completely cyclize and aromatize the amidated tetracyclic aglycon. Formation of the last ring via C-1/C-18 aldol condensation does not require a dedicated fourth-ring cyclase, in contrast to the biosynthetic mechanism of other tetracyclic aromatic polyketides, such as daunorubicin and tetracenomycin. Acetyl-derived polyketides do not undergo spontaneous fourth-ring cyclization and form only anthracene carboxylic acids as demonstrated both in vivo and in vitro. OxyF was identified to be the C-6 C-methyltransferase that regioselectively methylates pretetramid to yield 6-methylpretetramid. Reconstitution of 6-methylpretetramid in a heterologous host sets the stage for a more systematic investigation of additional tetracycline downstream tailoring enzymes and is a key step toward the engineered biosynthesis of tetracycline analogs. Tetracyclines are aromatic polyketides biosynthesized by soil-borne Streptomyces bacteria (1Kim E.S. Bibb M.J. Butler M.J. Hopwood D.A. Sherman D.H. Gene (Amst.). 1994; 141: 141-142Crossref PubMed Scopus (39) Google Scholar, 2Rawlings B.J. Nat. Prod. Rep. 1999; 16: 425-484Crossref PubMed Scopus (155) Google Scholar). It is well known that the carbon skeleton of an aromatic polyketide is assembled through stepwise decarboxylative condensation of malonate equivalents catalyzed by the minimal PKS, 2The abbreviations used are:PKSpolyketide synthaseLCliquid chromatographMSmass spectroscopyHPLChigh performance liquid chromatographyEAethyl acetateDMAC3,8-dihydroxy-methylanthraquinone carboxylic acidHR-ESIMShigh resolution electrospray ionization-MS. which consists of a ketosynthase/chain length factor heterodimer (KS-CLF or KSα-KSβ), an acyl-carrier protein (ACP), and a malonyl-CoA:ACP acyltransferase (3Hertweck C. Luzhetskyy A. Rebets Y. Bechthold A. Nat. Prod. Rep. 2007; 24: 162-190Crossref PubMed Scopus (405) Google Scholar). Dedicated tailoring enzymes transform the highly reactive poly-β-ketone backbone into fused, richly substituted compounds (4Rix U. Fischer C. Remsing L.L. Rohr J. Nat. Prod. Rep. 2002; 19: 542-580Crossref PubMed Scopus (227) Google Scholar). The biosynthesis of tetracyclines has been studied using blocked mutants of the chlorotetracycline producer Streptomyces aureofaciens (5McCormick J.R.D. Jensen E.R. J. Am. Chem. Soc. 1965; 87: 1794-1795Crossref PubMed Scopus (16) Google Scholar, 6McCormick J.R.D. Jensen E.R. J. Am. Chem. Soc. 1968; 90: 7126-7127Crossref PubMed Scopus (22) Google Scholar, 7McCormick J.R.D. Jensen E.R. J. Am. Chem. Soc. 1969; 91: 206-208Crossref PubMed Scopus (9) Google Scholar, 8McCormick J.R.D. Jensen E.R. Arnold N.H. Corey H.S. Joachim U.H. Johnson S. Miller P.A. Sjolande N.O. J. Am. Chem. Soc. 1968; 90: 7127-7128Crossref PubMed Scopus (15) Google Scholar, 9McCormick J.R.D. Jensen E.R. Johnson S. Sjolande N.O. J. Am. Chem. Soc. 1968; 90: 2201-2202Crossref PubMed Scopus (16) Google Scholar, 10McCormick J.R.D. Joachim U.H. Jensen E.R. Johnson S. Sjolande N.O. J. Am. Chem. Soc. 1965; 87: 1793-1794Crossref PubMed Scopus (20) Google Scholar, 11McCormick J.R.D. Sjolander N.O. Johnson S. J. Am. Chem. Soc. 1963; 85: 1692-1694Crossref Scopus (18) Google Scholar). However, the underlying enzymology of several key tailoring steps that give rise to its structural features remain unresolved, including cyclization of the tetracyclic scaffold and C-methylation of C-6 to afford the key intermediate 6-methylpretetramid (Fig. 1). The oxytetracycline (oxy) biosynthetic gene cluster from Streptomyces rimosus has been completely sequenced, hence allowing a thorough investigation of the biochemical basis of these features (12Butler M.J. Binnie C. Hunter I.S. Sugden D.A. Warren M. Dev. Ind. Microbiol. 1990; 31: 41-50Google Scholar, 13Hunter I.S. Fierro F. Martin J.F. Microbial Secondary Metabolites: Biosynthesis Genetics and Regulation. Research Signpost, Lucknow, India2002: 141-166Google Scholar, 14McDowall K.J. Doyle D. Butler M.J. Binnie C. Warren M. Hunter I.S. Noack D. Krugel H. Baumberg S. Genetics and Product Formation in Streptomyces. Plenum Press, New York1990: 105-116Google Scholar, 15Petkovic H. Cullum J. Hranueli D. Hunter I.S. Peric-Concha N. Pigac J. Thamchaipenet A. Vujaklija D. Long P.F. Microbiol. Mol. Biol. Rev. 2006; 70: 704-728Crossref PubMed Scopus (86) Google Scholar, 16Zhang W.J. Ames B.D. Tsai S.C. Tang Y. Appl. Environ. Microbiol. 2006; 72: 2573-2580Crossref PubMed Scopus (98) Google Scholar). polyketide synthase liquid chromatograph mass spectroscopy high performance liquid chromatography ethyl acetate 3,8-dihydroxy-methylanthraquinone carboxylic acid high resolution electrospray ionization-MS. During tetracycline biosynthesis, the C-9 reduced decaketide backbone first undergoes C-7/C-12 intramolecular aldol condensation to fix the regioselectivity of the D ring followed by three sequential cyclization reactions to form a fully aromatized, tetracyclic intermediate pretetramid (Fig. 1). Biosynthesis of aureolic acids such as mithramycin presumably follows identical cyclization patterns, as inferred from the structure of premithramycinone, an early intermediate during mithramycin biosynthesis (17Rohr J. Weissbach U. Beninga C. Kunzel E. Siems K. Bindseil K.U. Lombo F. Prado L. Brana A.F. Mendez C. Salas J.A. Chem. Commun. 1998; : 437-438Crossref Scopus (30) Google Scholar). The biochemical basis of the C-1/C-18 cyclization reaction, which affords the last rings in both tetracycline and mithramycin, is not well understood. Combinatorial biosynthesis using mithramycin and tetracenomycin biosynthetic genes indicated that MtmX may function as a fourth ring cyclase (18Kunzel E. Wohlert S.E. Beninga C. Haag S. Decker H. Hutchinson C.R. Blanco G. Mendez C. Salas J.A. Rohr J. Chem.-Eur. J. 1997; 3: 1675-1678Crossref Scopus (31) Google Scholar), although no direct biochemical evidence is available. The oxy gene cluster encodes an MtmX homologue, OxyI, which may be involved in A ring cyclization. Compounds in the anthracycline family, such as daunorubicin, undergo identical cyclization reactions to fix the first three rings (Fig. 1) (19Hautala A. Torkkell S. Raty K. Kunnari T. Kantola J. Mantsala P. Hakala J. Ylihonko K. J. Antibiot. (Tokyo). 2003; 56: 143-153Crossref PubMed Scopus (15) Google Scholar). Formation of the fourth, non-aromatic ring to yield alkaviketone during daunorubicin biosynthesis is catalyzed by the aklanonic acid methyl ester cyclase DnrD via distinct C-2/C-19 connectivity (20Kendrew S.G. Katayama K. Deutsch E. Madduri K. Hutchinson C.R. Biochemistry. 1999; 38: 4794-4799Crossref PubMed Scopus (30) Google Scholar, 21Kantola J. Kunnari T. Hautala A. Hakala J. Ylihonko K. Mantsala P. Microbiology. 2000; 146: 155-163Crossref PubMed Scopus (38) Google Scholar). The tetracenomycin family of compounds is constructed from fully aromatized intermediate tetracenomycin F1, but the regioselectivities are significantly different, starting with the first cyclization, which occurs between C-9 and C-14. The fourth ring cyclization takes place between C-2/C-19 and is catalyzed by a dedicated cyclase TcmI (22Thompson T.B. Katayama K. Watanabe K. Hutchinson C.R. Rayment I. J. Biol. Chem. 2004; 279: 37956-37963Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar, 23Hutchinson C.R. Chem. Rev. 1997; 97: 2525-2536Crossref PubMed Scopus (121) Google Scholar). Another important structural feature of tetracyclines is a methyl group installed on C-6 of second ring. The methylation modification occurs relatively early during tetracycline biosynthesis as evidenced by isolation of the intermediate 6-methylpretetramid (Fig. 1) (8McCormick J.R.D. Jensen E.R. Arnold N.H. Corey H.S. Joachim U.H. Johnson S. Miller P.A. Sjolande N.O. J. Am. Chem. Soc. 1968; 90: 7127-7128Crossref PubMed Scopus (15) Google Scholar). The C-methylation reaction is likely catalyzed by one of two methyltransferases in the oxy gene cluster (OxyF and OxyT), both containing putative S-adenosylmethionine binding domains. C-Methyltransferases are found infrequently among bacterial type II PKSs, including MtmMII in the mithramycin gene cluster (24Lozano M.J. Remsing L.L. Quiros L.M. Brana A.F. Fernandez E. Sanchez C. Mendez C. Rohr J. Salas J.A. J. Biol. Chem. 2000; 275: 3065-3074Abstract Full Text Full Text PDF PubMed Scopus (63) Google Scholar), RemG and RemH from the resistomycin gene cluster (25Jakobi K. Hertweck C. J. Am. Chem. Soc. 2004; 126: 2298-2299Crossref PubMed Scopus (64) Google Scholar), and the recently identified BenF and NapB2 from the benastatin (26Xu Z. Schenk A. Hertweck C. J. Am. Chem. Soc. 2007; 129: 6022-6030Crossref PubMed Scopus (58) Google Scholar) and napyradiomycin (27Winter J.M. Moffitt M.C. Zazopoulos E. McAlpine J.B. Dorrestein P.C. Moore B.S. J. Biol. Chem. 2007; 282: 16362-16368Abstract Full Text Full Text PDF PubMed Scopus (137) Google Scholar) gene clusters, respectively. All these C-methyltransferases modify the respective aromatic aglycons with unique regiospecificity. Therefore, identification and reconstitution of the dedicated 6-methyltransferase in the oxytetracycline biosynthetic pathway will add to the molecular toolbox of tailoring enzymes. Gene disruption through homologous recombination followed by structure elucidation of the metabolites produced by the mutants has been widely used to access enzyme functions in polyketide biosynthetic pathways. However, gene inactivation has been demonstrated to be of limited use in studying oxytetracycline biosynthesis in S. rimosus. Genomic deletions of oxyK (28Petkovic H. Thamchaipenet A. Zhou L.H. Hranueli D. Raspor P. Waterman P.G. Hunter I.S. J. Biol. Chem. 1999; 274: 32829-32834Abstract Full Text Full Text PDF PubMed Google Scholar) and oxyS (29Peric-Concha N.A. Borovicka B. Long P.F. Hrnaueli D. Waterman P.G. Hunter I.S. J. Biol. Chem. 2005; 280: 37455-37460Abstract Full Text Full Text PDF PubMed Scopus (35) Google Scholar) in the oxy gene cluster led to the recovery of truncated polyketides only (30Deseo M.A. Hunter S. Waterman P.G. J. Antibiot. (Tokyo). 2005; 58: 822-827Crossref PubMed Scopus (7) Google Scholar), and the resulting phenotypes cannot be directly associated with the functions of the deleted genes. The exact causes for biosynthesis of these truncated compounds in S. rimosus mutants are not understood. Purification of target enzymes followed by in vitro biochemical characterizations are complicated by the inaccessibility of the substrates for the cyclization and methylation reactions of interest here. As a result of these technical difficulties, investigation of these early tailoring reactions can be performed through the systematic and incremental reconstitution of the pathway in a heterologous host. Using a heterologous Streptomyces host and shuttle vectors containing different combinations of oxy genes, we recently reported the biosynthesis of novel amidated polyketides by coexpressing the oxy minimal PKS and OxyD (16Zhang W.J. Ames B.D. Tsai S.C. Tang Y. Appl. Environ. Microbiol. 2006; 72: 2573-2580Crossref PubMed Scopus (98) Google Scholar, 31Zhang W. Watanabe K. Wang C.C. Tang Y. J. Nat. Prod. 2006; 69: 1633-1636Crossref PubMed Scopus (19) Google Scholar). These studies verified that the amidotransferase homologue OxyD is a key enzyme in the biosynthesis of the amidated tetracycline backbone. The main goal of this work is to elucidate the enzymes involved in formation of the tetracyclic scaffold and C-methylation of the aromatic aglycon. We compiled the minimal set of enzymes required to reconstitute 6-methylpretetramid biosynthesis in the heterologous host. These results will enable additional investigation of oxytetracycline downstream tailoring steps and lead to the rational and combinatorial biosynthesis of tetracycline analogs. Bacterial Strains and General Techniques for DNA Manipulation—Streptomyces coelicolor strain CH999 (32McDaniel R. Ebert-Khosla S. Hopwood D.A. Khosla C. Science. 1993; 262: 1546-1550Crossref PubMed Scopus (393) Google Scholar) was used as a host for transformation of shuttle vectors. Protoplast preparation and polyethylene glycol-assisted transformation were performed as described by Hopwood et al. (33Hopwood D. 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. Genetic Manipulation of Streptomyces: A Laboratory Manual. The John Innes Foundation, Norwich, UK1985Google Scholar). Escherichia coli XL-1 Blue (Stratagene) was used for the manipulation of plasmid DNA. PCR was performed using Platinum Pfx DNA polymerase (Invitrogen). PCR products were first cloned into a pCR-Blunt vector (Invitrogen) followed by DNA sequencing. Unmethylated DNA was obtained using the methylase-deficient strain GM2163 (New England Biolabs). T4 DNA ligase (Invitrogen) was used for ligation of restriction fragments. The cosmid clone pYT264, which harbors the entire oxy gene cluster (16Zhang W.J. Ames B.D. Tsai S.C. Tang Y. Appl. Environ. Microbiol. 2006; 72: 2573-2580Crossref PubMed Scopus (98) Google Scholar), was used as the template for the amplification of individual oxy genes. Construction of Plasmids—Primers used to amplify the individual genes are listed in supplemental Table S4. Multicistronic cassettes were constructed using the compatible XbaI/SpeI/NheI cohesive ends for most of the genes. Different combinations of genes were introduced into pWJ35 (16Zhang W.J. Ames B.D. Tsai S.C. Tang Y. Appl. Environ. Microbiol. 2006; 72: 2573-2580Crossref PubMed Scopus (98) Google Scholar) or pRM5 (32McDaniel R. Ebert-Khosla S. Hopwood D.A. Khosla C. Science. 1993; 262: 1546-1550Crossref PubMed Scopus (393) Google Scholar) to yield the constructs shown in Table 1.TABLE 1Plasmid constructions and resulting polyketide productsPlasmidGenesMajor productMinor productsReferencepYT319oxyABCSEK15SEK15b(24Lozano M.J. Remsing L.L. Quiros L.M. Brana A.F. Fernandez E. Sanchez C. Mendez C. Rohr J. Salas J.A. J. Biol. Chem. 2000; 275: 3065-3074Abstract Full Text Full Text PDF PubMed Scopus (63) Google Scholar)pWJ85oxyABCDSEK15, WJ85WJ85b(43Bililign T. Hyun C.G. Williams J.S. Czisny A.M. Thorson J.S. Chem. Biol. 2004; 11: 959-969Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar)pWJ35oxyABCDJWJ35, RM20bRM20, RM20c(24Lozano M.J. Remsing L.L. Quiros L.M. Brana A.F. Fernandez E. Sanchez C. Mendez C. Rohr J. Salas J.A. J. Biol. Chem. 2000; 275: 3065-3074Abstract Full Text Full Text PDF PubMed Scopus (63) Google Scholar)pWJ83oxyABCDJKN12-5, DMACThis workpWJ196oxyABCDJK/dpsY12-5, DMACThis workpWJ83coxyABCJKN5DMACThis workpWJ90oxyABCDJKNI12-5, DMACThis workpWJ120oxyABCDJKNT12-5, DMACThis workpWJ119oxyABCDJKNF62-5This workpWJ123actI-III/actVII/actIV/oxyFDMACAloesaponarin IIThis workpWJ180pdmABCD/oxyFTW95a,bThis workpWJ190tcmKLMN/oxyFTcmF2RM80This work Open table in a new tab Culture Conditions, Extractions, and Small-scale Analysis—Strains were grown on solid R5 plates with 50 mg/liter thiostrepton at 30 °C for 7-10 days. For LC/MS and analytical HPLC analysis, a well pigmented plate was chopped into fine pieces and extracted with 50 ml of ethyl acetate (EA)/methanol (MeOH)/acetic acid (89%/10%/1%). Extracts were dried over anhydrous Na2SO4. Solvent was removed in vacuo, and the residue was dissolved in 0.5 ml of dimethyl sulfoxide (Me2SO). The polyketide products were separated by reverse-phase HPLC and detected at 280 and 410 nm using an analytical C18 column (Varian Pursuit 5 μm, 250 mm × 4.6 mm) with a linear gradient of 5% acetonitrile (CH3CN) in water (0.1% trifluoroacetic acid) to 95% CH3CN in water (0.1% trifluoroacetic acid) over 30 min with a flow rate of 1 ml/min. HPLC retention times (tR) were as follows: WJ83T1 (1), 26.4 min; WJ83Q2 (2a/b), 17.8 min; WJ83Q3 (3), 29.0 min; WJ83Q1 (4), 16.2 min; desmethylaklanonic acid (5), 22.8 min; WJ119 (6), 22.5 min; DMAC, 20.5 min. Spectroscopic Analysis—High resolution mass spectrometry was performed at the Vincent Coates Foundation Mass Spectrometry Laboratory at Stanford University using a Micromass Q-Tof hybrid quadrupole-time of flight LC-MS. NMR spectra were obtained on Bruker DRX-500 and DRX-600 spectrometers at the NMR facility in the Department of Chemistry and Biochemistry at UCLA. 1H and 13C chemical shifts were referenced to the solvent peak (Me2SO-d6) δ 2.49 and 39.5 ppm, respectively. Standard parameters were used for one- and two-dimensional NMR experiments, which included 1H,13C, nuclear Overhauser effect, heteronuclear multiple quantum correlation (1H,13C), and heteronuclear multiple-bond correlation (1H,13C). Isolation of WJ83T1 (1) from S. coelicolor CH999/pWJ83—One hundred R5 plates (3 liter) streaked with CH999/pWJ83 were incubated at 30 °C for 10 days. The plates were chopped into fine pieces and extracted with 3 liter of EA/MeOH/acetic acid (94%/5%/1%). The solvent was removed in vacuo, and the residue was partially dissolved in 100 ml of MeOH. The methanol-insoluble fraction collected by filtration was dissolved in 5 ml Me2SO, filtered, and purified with a preparative reverse-phase HPLC column (SunFire Prep C18 OBD column, 5 μm, 19 × 50 mm). A 20-40% CH3CN in water (10 mm triethylamine) gradient was used over 30 min with a flow rate of 3 ml/min to give purple fractions containing 1 with a retention time of 13.8 min. The eluent was extracted with ethyl acetate and dried in vacuo to give pure solid 1 (approximate yield of 55 mg/liter). WJ83T1 (1): HR-ESIMS m/z = 364.0450 (C19H10 NO7 [M-H]-, 364.0457 calculated); UV (CH3CN/H2O/trifluoroacetic acid) λmax, 278 and 500 nm; see supplemental Table S1 for NMR spectral data. Isolation of WJ83Q1-3 from S. coelicolor CH999/pWJ83—From the same extract as above, the methanol soluble fraction was dried in vacuo and redissolved in 5 ml of EA. The residue was first chromatographed on a normal-phase silica gel (150 g) column and eluted with ethyl acetate in 1% acetic acid to give two impure fractions containing 3 and 4 (Rf values if 0.72 and 0.3, respectively). Further purification of 3 and 4 were achieved on a preparative reverse-phase HPLC column (XTerra Prep MS C18 OBD column, 5 μm, 19 × 50 mm), and a 30-80% CH3CN in water (0.1% trifluoroacetic acid) gradient was used over 45 min with a flow rate of 3 ml/min. 4 and 3 crystallized as pure yellow/orange needles after elution from the column, with retention times of 8.6 and 17.8 min and approximate yields of 5 and 2 mg/liter, respectively. The MeOH-soluble fraction of the crude extract was also directly applied to the preparative reverse-phase HPLC column to give pure 2 (approximate yield of 8 mg/liter with retention time of 11.2 min at the same gradient as above). WJ83Q2 (2): HR-ESIMS m/z = 338.0662 (C18H12NO6 [M-CO2-H]-, 338.0665 calculated); UV (CH3CN/H2O/trifluoroacetic acid) λmax, 228, 240, 257, 289, and 430 nm. WJ83Q3 (3): HR-ESIMS m/z = 319.0610 (C19H11O5 [M-H]-, 319.0607 calculated); UV (CH3CN/H2O/trifluoroacetic acid) λmax, 240, 267, 284, and 412 nm. WJ83Q1 (4): HR-ESIMS m/z = 296.0565 (C16H10NO5 [M-H]-, 296.0559 calculated); UV (CH3CN/H2O/trifluoroacetic acid) λmax: 228, 258, 290, and 430 nm; see supplemental Tables S2 and S3 for NMR spectral data. Isolation of WJ119 (6) from S. coelicolor CH999/pWJ119—Seventy R5 plates (2 liter) streaked with the CH999/pWJ119 were incubated at 30 °C for 8 days. The plates were extracted with 2 liter of EA/MeOH/acetic acid (94%/5%/1%). The solvent was removed in vacuo, and the residue was first chromatographed on a normal-phase silica gel (100 g) column and eluted with EA in 1% acetic acid to give impure yellow fractions containing 6 (Rf value 0.38). Further purification of 6 was achieved on a preparative reverse-phase HPLC column (SunFire Prep C18 OBD column, 5 μm, 19 × 50 mm). A 5-95% CH3CN in water (10 mm triethylamine) gradient was used over 30 min with a flow rate of 3 ml/min to give yellow fractions containing 6 with a retention time of 12.4 min. The pH of the eluent was adjusted to ∼2 with 6 n HCl, extracted with EA, and dried in vacuo to give pure solid 6 (approximate yield of 58 mg/liter). WJ119 (6): HR-ESIMS m/z = 380.0780 (C20H14NO7 [M-H]-, 380.0770 calculated); UV (CH3CN/H2O/trifluoroacetic acid) λmax, 220, 268, 300, and 435 nm; see supplemental Table S1 for NMR spectral data. Enzyme Purification and in Vitro Assays—Gris ARO/CYC, OxyN, OxyI, and holo-OxyC were expressed and purified from E. coli strains BL21(DE3)/pWJ184, BL21(DE3)/pWJ181, BL21(DE3)/pWJ189, and BAP1/pWJ66, respectively (see supplemental data for details). ActIII ketoreductase and malonyl-CoA:ACP acyltransferase were expressed in E. coli and purified as described previously (34Korman T.P. Hill J.A. Vu T.N. Tsai S.C. Biochemistry. 2004; 43: 14529-14538Crossref PubMed Scopus (61) Google Scholar, 35Summers R.G. Ali A. Shen B. Wessel W.A. Hutchinson C.R. Biochemistry. 1995; 34: 9389-9402Crossref PubMed Scopus (118) Google Scholar). OxyAB was purified from CH999/pWJ131. The in vitro assays converting malonyl-CoA into aromatic polyketides were performed in reaction buffer (100 mm NaH2PO4 (pH 7.4), 2 mm MgCl2, and 10% glycerol). Each reaction mixture contained 2 mm malonyl-CoA, 300 nm malonyl-CoA:ACP acyltransferase, 20 μm holo-OxyC, and 10 μm OxyAB in a final volume of 100 μl. All other enzymes were added to final concentrations between 20 and 100 μm. NADPH (2 mm final concentration) was added as needed. Reactions were initiated by adding the substrate malonyl-CoA and incubated at 30 °C for 2 h. The reaction mixture was extracted with EA/acetic acid (99%/1%), and the organic phase was dried in a SpeedVac and redissolved in 20 μl of Me2SO and analyzed with HPLC. We constructed a series of plasmids to identify the minimal set of enzymes required to afford the key intermediate 6-methylpretetramid. The genes of interest were systematically introduced into the E. coli/S. coelicolor shuttle vector as shown in Table 1. The plasmids were transformed into the engineered S. coelicolor host CH999 (32McDaniel R. Ebert-Khosla S. Hopwood D.A. Khosla C. Science. 1993; 262: 1546-1550Crossref PubMed Scopus (393) Google Scholar), and the production of polyketides by the host/vector pairs were assayed by LC/MS. Functional Study of OxyN—Our previous study had identified that OxyABCDJ were essential and sufficient to produce the amidated decaketide tetracycline backbone with good yield in CH999 (16Zhang W.J. Ames B.D. Tsai S.C. Tang Y. Appl. Environ. Microbiol. 2006; 72: 2573-2580Crossref PubMed Scopus (98) Google Scholar). Furthermore, OxyK was identified by Petkovic et al. (28Petkovic H. Thamchaipenet A. Zhou L.H. Hranueli D. Raspor P. Waterman P.G. Hunter I.S. J. Biol. Chem. 1999; 274: 32829-32834Abstract Full Text Full Text PDF PubMed Google Scholar) and confirmed in our laboratory (36Hertweck C. Xiang L. Kalaitzis J.A. Cheng Q. Palzer M. Moore B.S. Chem. Biol. 2004; 11: 461-468Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar) to be the bifunctional cyclase/dehydratase that is responsible for formation of the D ring through C-7/C-12 cyclization (Fig. 2). OxyN showed strong sequence similarity to DpsY (37Lomovskaya N. Doi-Katayama Y. Filippini S. Nastro C. Fonstein L. Gallo M. Colombo A.L. Hutchinson C.R. J. Bacteriol. 1998; 180: 2379-2386Crossref PubMed Google Scholar) and MtmY (38Prado L. Lombo F. Brana A.F. Mendez C. Rohr J. Salas J.A. Mol. Gen. Genet. 1999; 261: 216-225Crossref PubMed Scopus (55) Google Scholar), which catalyze cyclization of the second ring during daunorubicin and mithramycin biosynthesis, respectively. We tested the function of OxyN through coexpression with the oxy minimal PKS, OxyJ, OxyK, and OxyD in plasmid pWJ83. CH999/pWJ83 grown on R5 agar developed rust-colored, non-diffusible pigmentation in the mycelia. Several new biosynthetic products were detected from the CH999/pWJ83 extract and were partitioned with methanol and purified separately. The methanol-insoluble fractions contained the most abundant compound produced by this strain. WJ83T1 (1) was isolated at ∼55 mg/liter after reversed-phase HPLC purification. The UV spectrum of the compound in dilute acid showed absorption maxima at 278 nm and a broad peak at 500 nm that is indicative of an extended chromophore. The compound exhibited a bathochromic shift and developed a deep purple color with the addition of NaOH (λmax = 566 nm). HR-ESIMS indicated the molecular formula of C19H11NO7 for 1, consistent with the molecular composition of an amidated decaketide (Fig. 2). The 13C NMR spectrum (supplemental Table S1) showed 19 signals composed of 5 aromatic C-H residues and 14 quaternary carbons, suggesting a completely aromatized scaffold. Two downfield signals (δC-6 = 183.1 and δC-13 = 184.5 ppm) were indicative of conjugated carbonyls. The 1H NMR spectrum contained two broad signals corresponding to the free amide protons (δ = 7.87 and 9.67 ppm). Strong 1H,3C heteronuclear multiple-bond correlations of H-8/C-12, H-4/C-14, H-4/C-16, H-2/C-16, and H-2/C-18 together with the nuclear Overhauser effect coupling of H-2/H-4 established the polyhydroxyl naphthacenedione structure of 1. Accumulation of the linear, tetracyclic 1 as a major product in CH999/pWJ83 demonstrates that only two cyclases (OxyK and OxyN) were necessary to cyclize all four rings in tetracycline. Oxidation of ring C to a quinone was likely a spontaneous, nonenzymatic modification that is widely observed for aromatic polyketides (32McDaniel R. Ebert-Khosla S. Hopwood D.A. Khosla C. Science. 1993; 262: 1546-1550Crossref PubMed Scopus (393) Google Scholar, 39Tang Y. Lee T.S. Khosla C. PLoS Biol. 2004; 2: 227-238Crossref Scopus (60) Google Scholar, 40McDaniel R. Ebert-Khosla S. Hopwood D.A. Khosla C. Nature. 1995; 375: 549-554Crossref PubMed Scopus (250) Google Scholar, 41Yu T.W. Shen Y. McDaniel R. Floss H.G. Khosla C. Hopwood D. Moore B.S. J. Am. Chem. Soc. 1998; 120: 7749-7759Crossref Scopus (89) Google Scholar). LC-MS analysis indicated the likely presence of trace amount of the acetyl-primed version of 1 (m/z 364.1) in the extract of CH999/pWJ83. Numerous anthraquinone compounds were recovered from the methanol-soluble fraction of the CH999/pWJ83 extract. Reverse-phase HPLC purification yielded an orange-pigmented metabolite WJ83Q2 (2a) at ∼8 mg/liter (Fig. 2). WJ83Q2 has strong UV absorptions at 228, 257, and 430 nm and has a molecular formula of C19H14NO8. The 1H spectrum showed four aromatic proton signals, two methylene singlets and two broad 1H signals at δH = 7.12 and 7.52 ppm identified as amide protons. 2a is the amidated version of the well known metabolite aklanonic acid, which is an intermediate in the anthracycline biosynthetic pathways (Fig. 1) (42Eckardt K. Tresselt D. Schumann G. Ihn W. Wagner C. J. Antibiot. (Tokyo). 1985; 38: 1034-1039Crossref PubMed Scopus (38) Google Scholar). NMR analysis in Me2SO-d6 also showed both the C-17/C-18 keto (2a) and enol (2b) isoforms were present with an equilibrium ratio of 2 to 3. Furthermore, after 2 days at room temperature in Me2SO-d6, 2a and 2b were completely decarboxylated to 2c and 2d as indicated by NMR and LC-MS analysis (supplemental Table S2). The acetyl-primed version of 2, desmethylaklanonic acid (5), was also purified from the methanol-soluble fraction (∼4.5 mg/liter). Two minor orange-pigmented metabolites, WJ83Q3 (3) and WJ83Q1 (4), were isolated from CH999/pWJ83 (Fig. 3). Compound 3 (2 mg/liter) had a molecular composition of C19H12O5, which indicated that it may be an acetyl-primed decaketide that has undergone loss of one CO2. NMR spectra (supplemental Table S3) confirmed regiospecific cyclization of D, C, and B rings and a unique O-15/C-19 cyclization to yield a fused γ-pyrone. The 4H-anthra[1,2-b]-pyran-4,7,12-trione aglycon is present
Referência(s)