J. CHEM. SOC. PERKlN TRANS. 1 1983 Studies in Terpenoid Biosynthesis, Part 28.' The Acetate and Mevalonate Labelling Patterns of the Steroid, Demethoxyviridin James R. Hanson.' Margaret A. O'Leary, and Harry J. Wadsworth School of Molecular Sciences, University of Sussex, Brighton, Sussex BN7 9QJ The enrichment and labelling patterns of demethoxyviridin, biosynthesized by Nodulisporium hinnuleum from l -13C-, 1,2-13C2-acetate, 2-13C- and 5-13C-mevalonate have been used to define the isoprene units in this metabolite and are consistent with a triterpenoid origin. The number and location of the hydrogen atoms originating from acetate and the 2-, 4-, and 5-positions of mevalonate, have been determined by a combination of 3H: 14C ratio and 2H n.m.r. studies.The biosynthesis of steroids may be divided2V3 into two phases. The first synthetic phase involves the assembly of the prenyl units and their cyclization to afford the C30 carbon skeleton of lanosterol and its relatives whilst the second phase involves the oxidative degradation of the C30 skeleton to afford the individual families of steroid. Acetate and mevalon- ate labelling patterns have provided considerable information in this context. There are a number of fungal steroidal anti- biotics. Amongst these the biosynthesis of fusidic acid and its relatives, has been studied in some detaiL4 Demethoxyviridin (1) which is produced by the fungus, Nodulisporium hinnul- eum, has an androstane structure but with an aromatic ring c. The formation of the fungal androstanes of the viridin group invites comparison with bacterial, plant, and mammalian steroid biosynthesis.Earlier degradative experiments on viridin (2) biosynthesized from 2-14Cmevalonate located labels at C-1, C-7, and C-15 consistent with its formation in a steroid-like manner. Some preliminary experiments have also been reported 'on the biosynthesis of a similar fungal meta- bolite, wortmannin (3) which does not have an aromatic ring c. Some studies on the incorporation of sterol precursors into viridin have also been reported.8 In this paper we describe some experiments which define the mevalonoid origin of the carbon skeleton and hydrogen atoms of demethoxyviridin(1h9As a prelude to the application of 13C labelling studies to the biosynthesis, the n.m.r.signals from demethoxyviridin were assigned. (see Table 1). The resonances were assigned using their multiplicities in the SFORD spectra combined with known substituent effects and simple models. The three triplets in demethoxyviridin at 6 28.2, 35.7, and 46.5 were assigned to C-15, C-16, and C-2 respectively. The resonances of C-15 and C-16 were relatively insensitive to changes in the substitution of ring A and are in agreement with the assign- ments of the comparable carbon atoms of indanone." The doublets assigned to C-11 and C-12 (6 128.4 and 126.2 res- pectively) also paralleled those of an indanone mode1,'O whilst the extra furanoid carbon atom (6 149.4) was assigned by comparison with furan. Amongst the singlets, the carbonyl signals at 6 205.5 and 172.5 were relatively insensitive to changes in the substitution pattern of ring A and were there- fore assigned to C-17 and C-7 respectively leaving the car- bony1 singlet at 6 190 to be assigned to C-3.The signal at 6 122.9 was assigned to C-4 on the basis of its upfield shift on reduction of C-3. The aromatic singlets 6 129.5, 126.3, and 156.9 (two resonances) were assigned to C-8, C-13, C-9, and C-14 respectively again by comparison with the appropriate signals in the spectrum of indanone. The remaining singlets (6 144.3 and 145.55) were assigned to C-5 and C-6 respec- tively, the former being more sensitive to structural variations on ring A. The preparation of the derivatives will be described in a paper on the chemistry of demethoxyviridin.(1) R1 = H, R2= OH (3) (2) R' =OMe, R2 = OH (4) R' = H, R2 = OAC (6) R' = R2= H (9) R1 = H, RZ= a-OMe,P-H 0 R (5) (7) R = 0 (8) R = OAc,H Preliminary feeding experiments with 2-'4Cmevalonic acid established that the optimum time to feed mevalonate to achieve a good incorporation into demethoxyviridin by sur-face cultures of Nodulisporium hinnuleum, was 10-1 1 days after inoculation, with the fungus being harvested 20 days after inoculation (see Figure). Sodium l-13C- and 1,2-13C2-acetate and 2-I3C- and 5-13C-mevalonatewere then fed separately to N. hinnuleum. The 'jc-enrichment and 13C-'3C coupling patterns of the resultant samples of demethoxyviridin are shown in (10) and (1 1) whist the numerical values are given in Table 2.The coupling and enrichment patterns are consistent with a triterpenoid biosynthesis.2 There was sufficient enrichment from the I-13Cacetate experiment which was carried out as a 'single-dose ' feeding for couplings (J55 Hz)to be observed between the adjacent enriched centres C-11 and C-12 and C-8-C-14 arising from multiply labelled precursors. These coupling patterns show that the aromatic ring has been formed without skeletal rearrangement. These experiments also de- fined the origin of the ' extra ' carbon atom attached to C-4, constituting part of the furan ring. Although it was likely that 868 J. CHEM.SOC. PERKIN TRANS. 1 1983 Table 1.I3CN.m.r. signals of demethoxyviridin and some derivatives (determined in CDC13at 25 MHz, p.p.m. from Me) Compound Carbon no. r-(1) a (4) (5) (6) A (7) = (8) 3 (9) ‘ 1 70.86 72.68 147.88 32.64 26.15 62.19 83.55 2 46.53 42.65 131.29 36.22 28.33 28.58 39.93 3 189.96 187.6 179.23 191.48 64.13 73.1 1 189.18 4 122.92 122.92 121.59 121.83 120.98 120.25 121.83 5 144.34 142.09 145.37 144.28 144.70 141.85 144.03 6 145.55 146.34 145.73 146.04 144.34 144.88 146.70 7 172.55 172.73 172.86 172.49 173.52 173.03 173.28 8 129.54 130.2 130.69 129.41 130.51 130.44 131.05 9 156.29 153.38 151.92 156.71 157.69 154.29 154.35 10 41.56 40.47 40.71 36.34 37.01 40.53 42.77 11 128.44 126.99 124.25 125.83 124.74 126.87 124.68 12 126.2 127.41 127.55 126.5 126.94 126.87 127.35 13 136.33 137.60 137.72 136.39 137.06 137.30 137.36 14 156.29 158.23 159.08 157.26 158.35 158.17 158.96 15 28.15 28.58 28.45 27.85 32.94 36.34 28.45 16 35.17 36.28 36.34 35.67 36.46 32.34 36.4 17 205.5 205.92 205.92 205.14 206.41 206.35 206.22 19 25.24 26.21 41.07 30.03 32.40 26.34 31.06 Furan-c 149.37 148.28 148.22 149.68 148.16 148.46 148.34 “In DMSO.bOAc,21.30, 169.15. ‘OAc, 21.17, 171.09. OAc, 21.00, 21.42, 169.46, 170.68. ‘OMe 61.3. 2 4 6 8 10 12 14 16 16 20 Day of mevalonate feed Figure. Variation of mevalonate incorporation with time of feeding to N.hinnuleum this carbon atom was part of a prenyl unit, nevertheless it could have arisen from the C-l pool at a late stage in the biosynthesis. The fungus produces ergosterol which lacks both the substituents at this centre. Furthermore, there are chemi- cal analogies for the formation of a furan ring at C-4 by the intramolecular condensation of a C-6 ester. The retention of the coupling between C-4 and the carbon atom of the furan ring showed that this bond had remained intact throughout the biosynthesis from acetate. The stereochemistry of labelling of the cyclic triterpenoids at this centre is known.’ The 1,2- 13C2acetate results show that the extra carbon atom at C-4 originates from the 3’-position rather than from C-2 of mevalonate.It thus corresponds to the 4P-methyl group of a protolanosterol/lanosterol precursor. This contrasts with fusidic acid biosynthesis by the fungus Fusidium coccirterrm in which the other methyl group derived from C-2 of mevalon-0@Q 0@0 (10 1 (11 1 (10) = Coupling derived from 1 ,2-13Clacetate,0 = enrichment from l-13Cacetate, (1 1) =: Enrichment from 2J3C-MVA, = enrichment from 5-13C-MVA ate, is retained.” In sterol biosynthesis, the normal order of demethylation involves loss of the 4a-methyl group first. The 2-13Cmevalonate results are in accordance with the earlier carbon-14 work on viridin but strictly define only three out of the five isoprene units. The 5-t3Cmevalonate results served to distinguish all five isoprene units which go to form demethoxyviridin thus providing further evidence for excluding a diterpenoid precursor of the cleistanthene type (cf.ref. 6). Having defined the mevalonoid origin of the carbon skele- ton, we turned our attention to the number and location of the mevalonoid hydrogen atoms. The number of mevalonoid hydrogen atoms that were incorporated, were established by conventional 3H : I4C ratio studies and are tabulated (see Table 3). The location of the labels was established by ’H n.m.r. spectroscopy. Whereas demethoxyviridin is poorly soluble or unstable in suitable n.m.r. solvents over the time scale re- quired for ’H spectroscopy, the acetate is conveniently soluble in chloroform. The ‘H n.m.r.signals of demethoxy- viridin acetate (4) were assigned from the 220 MHz spe-t rum (see Table 4). The signals assigned to the 2-H and the 15-and 16-H were identified by decoupling experiments based on the readily identified I-H signal. 2-’H3Acetate and 2-’H2- and 5-’H2-mevalonate were then fed to the fungus. The demethoxyviridin was isolated and I. CHEM. SOC. PERKIN TRANS. I 1983 Table 2. Enrichment * and coupling patterns (J/Hz) of de- methoxyviridin (1) Precursor - Carbon atom T 1 -'TI-Acetate 1,2-1'C2-Acetate A I2-wJ-MVA 5-"c-MVA 1 0.99 2.02 1 .o 2 1.75 39 1.01 1.37 3 1.14 39 I .02 0.91 4 2.02 65 1.22 0.93 5 1.1 37 1.11 1.06 6 1.57 37 0.62 1.18 7 0.73 1.67 1.01 8 1.58 t 0.94 1.02 9 0.95 53 1.03 0.96 10 1.97 33 1.03 0.99 11 1.43 53 0.95 1.33 12 1.61 t 58 I .09 1.41 13 0.95 58 0.91 0.94 14 1.53 t 0.95 0.99 15 1.08 2.0 0.93 16 1.91 39 0.99 1.34 17 1.08 39 0.96 1.o 19 0.95 33 1.o 0.96 Furan-c 1.02 65 0.96 1.06 * Enrichment = Enriched suectrurn ueak heinhtlnormalisinn factor Natural abundance spectrum peak height/normalising factor Normalising factor = Sum of unlabelled peak heights Number of peaks t J 55 Hz.Table 3. Incorporation of 'HJmevalonates into demethoxyviridin (1) Mevalonate 4(R)-4-5-'H2,2-2(R)"'H,2-''CJ-'H,2-14C-"C-3H : I4C Ratio 2.76: 1 4.2: 1 4.5: 1 1.65: 1 as fed Amount '"fed 30pC 21 pC 30pC 30pC 'H : I4C Ratio 0.61 : 1 0.72:1 0.12: 1 1.3: 1 in (1) Incorp.I4C 3.0 0.48 3.3 3.09 Atom ratio in 0.67 : 3 0.51 :3 0.08 : 3 4.72: 3 (1) Fed as the (3RS)-mixtures. converted into its acetate (4). The results are tabulated (Table 5). These results conform to the pattern anticipated for a steroid. The 2-'H3acetate results show that the 19-methyl group retains all three deuterium labels thus excluding a cyclopropanoid precursor related to cycloartenol from the biosynthesis, i.e. the biosynthesis follows a mammalian rather than:a plant scheme. Subsequent to our work lanosterol rather than cycloartenol has been shown to be the precursor of the sterols of the fungus Uromyces phaseoli. An interesting feature of the 2-'H3acetate experiment is that the signals arising via the 2-methylene group of mevalonate show a small drop in integral compared to those derived via the 3'-methyl group.Although this may reflect the action of prenyl isomerase, it might also arise from isotope effects later in the biosynthesis. The I-H and 15-H signals were equally labelled from 2- ZHzmevalonate. The 2-(R)-2-3H,2-14Cmevalonateresults show that 0.67 atom/molecule of demethoxyviridin were in- corporated. Incorporation studies with 1 a,2~-~H~dehydr- Table 4. 'H N.m.r. signals of demethoxyviridin acetate (4) (inCDCI, at 220 MHz) CouplingProton Signal (6) Multiplicity constants (Hz) 1 5.5 dd 11.5, 6 2 3.22 dd 19, 6 2 3.00 dd 19, 11.5 I1 8.0 s 12 8.0 S 15 3.62 d, t 20, 5.5 15 3.86 d, t 20, 5.5 16 2.78 t 5.5 16 2.78 t 5.5 19 1.8 s 21 8.24 S OAc 2.34 S Table 5.2H N.m.r. signals of biosynthetically enriched demethoxyviridin acetate (4) (in p.p.m.) ZH 'H N.m.r. Rel. N.m.r. Precursor signal (6) integral signal (6) Assignment zH3Acetatea 1.76 3.72 3.4 0.9 1 1.80 3.74 19-H 15-H 5.48 0.91 5.5 1-H 8.3 1 8.32 Furan-H 2-'HZ-MVA 3.72 5.52 1.o 1.08 3.74 5.5 15-H 1-H 5-'H+MVA 2.8br 4.1 2.78, 2-H, 3.0,. 3.22 16-H 8.08 1.5 8.0 11-H, 12-H a Determined at 30.3 MHz. Determined at 60.6 MHz. Deter-mined at 55.3 MHz. oxydemethoxyviridin show that, as expected, hydroxylation at C-1s has proceeded with retention of c~nfiguration.'~ Since the lp-H-position of steroids is labelled by a pru-2(R)-meva- lonoid hydrogen atom, i.e.the hydrogen atom which is replaced by a hydroxy-group in demethoxyviridin, the pro-2(R)-mevalonoid label in demethoxyviridin must be located at 15-H.Although the 15a- and 1Sp-hydrogen resonances are too close for a confident distinction to be made (6 3.74 and 3.86) only the more shielded 1 5a-resonance appears to be labelled in accordance with inversion at this centre. Inversion at this cen- tre has been shown to accompany the loss of a substituent from C-14 in sterol biosynthesi~.'~ The aromatic signals at 6 8.08 (1 1- and 12-H) bore approximately 1.5 labels compared to the 2-H and 16-H signals which bore a total of 4 labels from 5-ZHzmevalonate.This loss of hydrogen from C-1 1 and C-12 is in accordance with the loss of one hydrogen atom from the two farnesyl pyrophosphate residues which form squalene.' Efforts to separate these signals by using different n.m.r. solvents were not successful. In those situations, e.g. demeth-oxyviridin in dimethyl sulphoxide or trifluoroacetic acid, where a separation was obtained in the 'H n.m.r. spectrum, the 2H n.m.r. signal was either too broad or the sample decomposed during the time required to accumulate the data. On one occassion ergosterol was obtained from a 5-'Hzmevalonate feed. This showed 'H n.m.r. signals at 6 5.59 (6-H), 5.26 (25-H) and a group of signals at 6 1.25, 1.48, 1.72, 1.85, 2.0 and 2.12 attributable to the labels at 2-H, 11-H, 12-H, and 16-H.The following conclusions may be drawn from these re- sults. The carbon-1 3 enrichment and coupling pattcrns define 870 J. CHEM. soc. PERKIN TRANS. I 1983 Table 6.Feeding experiments to Nodulisporium hinnuleum Precursor l-13CAcetate 1,2-13CzAcetate2-"Cz-MVA5-13C-MVA2(R)2-3H,2-14C-MVA4(R)4-3H,2-'4C-MVA5-'H2,2-Wz-MVA2-2H3Acetate2-'HZ-MVA5-'HZ-MVA(5-'HMVA a Diluted with 800 mg unlabelled acetate. Ergosterol. Quantity mg (pC I4C) 1 g (21.6) 400"(23)250 (1 1.7) 100 (12.2) 30 pC30 pC30 pC2.5 g (11.7)1.67 g (12)300 (12) 490 (31) Mass of VOl. of fermentat ion Period of incubation demethoxy-viridin litres (day--day) isolated (mg) Incorp. "C 2 10-20 335 0.64 2 2 13, 15, 17-22 10-20 184 45 0.40 0.23 2 10-20 185 I .01 2 10-20 360 3.0 2 10-20 430 3.33 2 t 0-20 370 3.09 5 10-20 344 0.32 4 10-20 370 0.77 4 10-20 1 570 2.3 5 9-19 670 1.2 30 0.1 m.p.220-225 "C (decomp.), alDZo-67" (c 0.6) (Found: C, 69.2; H, 4.5. C21H1606 requires C, 69.2; H, 4.4); vmx-1 740, 1 700, 1 660, 1 575, and 1 525 crn-'; h,,,,. 310 (E 12 0o0) and 238 nm (E 32 600); for n.m.r. see Table 4. Acknowledgements We thank Mrs. N. Dransfield and Mrs. C. Perring for growing the cultures, S.E.R.C. for financial assistance and I.C.I. Pharmaceuticals Division for the gift of a strain of Noduli-sporiurn hinnuelum. References I Previous part, A. P. W. Bradshaw, J.R. Hanson, and I. H. Sadler, J. Chem. SOC.,Perkin Trans., I, 1982, 2787. 2 R. B. Clayton, Q. Rev. Chem. SOC.,1965, 19, 168, 201 ; L. J. Mulheirn and P. J. Ramm, Chem. SOC.Rev., 1972, 1, 259. 3 H. L. Holland, Chem. SOC.Rev., 1981, 10,435. 4 R. C. Ebersole, W. 0.Godtfredsen, S. Vangedal, and E. Caspi, J. Am, Chem. SOC.,1974, 96, 6499 and refs. therein. 5 D. C. Aldridge, A. J. Geddes, B. Sheldrick, and W. B. Turner, J. Chem. SOC.,Perkin Trans. I, 1975, 943. 6 J. F. Grove, J. Chem. SOC.C, 1969, 549. 7 J. MacMillan, T. J. Simpson, and S. K. Yeboah, J. Chem. SOC., Chem. Commun., 1972, 1063. 8 W. S. Golder and T. R. Watson, J. Chem. SOC.,Perkin Trans. I,1980,422. 9 For a preliminary communication of part of this work see J. R. Hanson and H.J. Wadsworth, J. Chem. SOC.,Chem. Commun., 1979, 360. 10 L. F. Johnson and W. C. Jankowski, ' Carbon-13 NMR Spectra,' John Wiley and Son, New York, 1972, p. 338. 11 T. Komeno, S. Ishihara, K. Takigawa, H. Itani, and H. Iwakura, Chem. Pharm. Bull., 1969, 17, 2586. 12 E. Caspi and L. J. Mulheirn, J. Am. Chem. SOC.,1970, 92,404. 13 S. K. Bansai and H. W. Knoche, Phytochemistry, 1980, 19, 1240. 14 J. R. Hanson and M. A. O'Leary, unpublished work. 15 G. F. Gibbons, L. J. Goad, and T. W. Goodwin, Chem. Commun., 1968, 1458. 16 J. R. Hanson and A. F. White, J. Chem. SOC.C,1969, 981. Received 9th August 1982 ;Paper 211 380 five isoprene units in the structure of demethoxyviridin con- sistent with a triterpenoid origin. The 'extra 'carbon atom of the furan ring at C-4 comes from the 3'-position of mevalonic acid and represents the 4-methyl group of a protolanosterol/ lanosterol precursor. This is consistent with the normal mammalian order of removal of the methyl groups from C-4 of lanosterol in which the initial oxidation and decarboxylation involves the C-4a-methyl group.The enrichment and coupling patterns of the aromatic ring c together with the induced 13C-'3Ccouplings, J11,12 show that it is formed with- and JBSl4, out skeletal rearrangement of a steroid. The hydrogen label- ling patterns are also consistent with a lanostane triterpenoid biosynthesis and exclude a cycloartenol intermediate. The apparent loss of a hydrogen atom from C-15 suggests that the loss of the 14a-methyl group may follow a similar pathway to that found in other steroid biosyntheses.Experimental General experimental details have been described pre-viousIy.1*16 General Fermentation Conditions.-Nodulisporium hinnul-euin (obtained from I.C.I. Pharmaceuticals Division, strain ACC 3 199) was grown on surface culture in Roux bottles (200 ml in each bottle) containing the following medium (quantities in g 1-I) glucose, 50; tartaric acid, 4; potassium carbonate, 0.6; ammonium dihydrogen phosphate, 0.6; magnesium carbonate, 0.4; ammonium sulphate, 0.25; zinc sulphate, 0.1 ;ferrous sul- phate, 0.1. The ages of the cultures at the time of feeding and harvest are given in Table 6. The precursors were adminis- tered in ethanol or water. The mycelium was filtered off, dried, and extracted in a Soxhlet funnel with chloroform for 5 h. The extract was concentrated under reduced pressure to yield a dark gum which was triturated with light petroleum. The solid product was filtered and recrystallized from acetone (charcoal) to afford demethoxyviridin (100-200 mg 1-I) as needles, m.p. 150-152 "C, (lit.,5 145-160 "C) aID2O-85" (c 0.4) which was identified by comparison (t.1.c. and 'Hn.m.r.) with an authentic sample. The broth contained very little demethoxyviridin. The acetate of demethoxyviridin, prepared with acetic anhydride in pyridine, crystallized from acetone as needles,
展开▼
