J. CHEM. SOC. PERKIN TRANS. 1 1991 29 1 Stages in the Biosynthesis of the Epoxy Lactol Side Chain of Nic-I, Insect Antifeedant Steroid of Nicandra physaloides Warren Andrews-Smith, Harjit K. Gill, Roland W. Smith and Donald A. Whiting* Department of Chemistry, The University, Nottingham NG7 2RD, UK Isotopic labelling experiments with Nicandra physaloides plants show that the insect antifeedant steroid Nic-I 1 is formed from 24(28)-methylenecholesteroI 2a; in the double bond isomerisation to 24-methylcholesta-5,24(25) -dien-3P-ol 3a, the 25-(pro-R) methyl becomes C-27, the pro-Z methyl in 3a. Hydroxylations lead through the diol 4a to the trio1 6a, and hence to the lactol 7a, with retention of the 26-hydrogen. The Solanaceae is a large plant family containing around 90 genera, some of which are economically important e.g.Solanurn (potato), Lycopersicon (tomato) and Capsicum (pepper); a number of species show interesting biological activities, e.g., Withania. A large group of plant steroids, the withanolides,' have been isolated from members of the Solanaceae, all based on the 24-methylcholestane skeleton and these are characterised by extensive oxidative modifications. A distinctive subset has been found in Nicandra physaloides (the Peruvian shoo-fly plant), including the insecticidal and insect antifeedant sterol Nic-1 1.* Nic-1 displays a number of interesting structural features and in a previous paper we have elucidated the biogenetic origins of the aromatic ring-D. We now turn our attention to the unusual epoxy lactol side chain.? This unit also appears in Nic-3 and in modified form, in other nicandrenoids.Structural comparisons within the group suggest that side chain modifications may either precede ring A/B oxidative development or be partly independent of it, as in a metabolic grid with enzymes of low substrate specificity. In this paper, we report experiments which confirm this view and which define stages in the oxidative development of the epoxy lactol function. Selectivity in the functionalisation of the diastereotopic C-25 methyl groups of the 24-methylenecholesterol precursor is also revealed. a; R=H b; R=Ac C; R=TBDMS RO A reasonable hypothetical sequence of oxidative development is shown in Scheme 1, starting from 24(28)-methylenecholesterol 2a, a known metabolite of Solanaceae.' Isomerisation to the 24(25)-isomer 3a could be followed by either 22a- or 26-hydroxyla tion, forming 4a or 5a, respectively.A second hydroxylation then would lead to the 22a, 26-diol6a. Oxidation at the primary alcohol site would then afford the unsaturated lactol 7a, finally epoxidised to the Nic-l side chain 8a. Further 28 II I bsol; 6 A L/ Scheme 1 Biosynthetic pathways to the Nic-1 side chain in N. physaloides oxidation of the lactol7a to lactone oxidation level may provide the pathway to the withanolide type 9a, in other solanaceous plants. It is possible, although perhaps less likely, that 9a+7a reduction could take place under certain circumstances. A grid involving either alcohols 4a or 5a is possible.To test these notions we set out to synthesise several of these intermediates, suitably radiolabelled for biogenetic experiments. Synthesis of Labelled Precursors.-Scheme 2 shows the formation of 28- ''C-24(28)-methylenecholesterol 2a and its t A preliminary communication has been p~blished.~ 10 11 i, vi 13 12 viiiI 145H3 cH H 14 15 + 145H3 16 Scheme 2 Reagents: i, Ac,O, Py; ii, Br,; iii, CrO,, AcOH; iv, Zn, AcOH; v, HO-; vi, SOCl,; vii, (Me,CH),Cd; viii, ''CH,=PPh3; ix, I,, PhH 24(25)-isomer 3a. A classical degradation of cholesterol 10a gave 3P-hydroxychol-5-enic acid 1la. Treatment of the acid chloride 12b with diisopropylcadmium provided 24-ketocholes- terol 13a which yielded the required product 14a on reaction with ''C-methylenetriphenylphosphorane.5a Isomerisation of the side-chain double bond was effected with iodine in refluxing benzene to give 15a;the 23-ene 16a was also formed to a minor (20) extent and could be separated by HPLC.Scheme 3 outlines the route to C3H-lactone 23a and C3H- trio124a. A well known cleavage of stigmasterol 17a afforded the aldehyde 18b,7 which was then elaborated to the lactone 9a using literature methods.8 The C-22 stereochemistry was controlled by the stereoselective epoxidation of 19b. The reduction of the epoxide with aluminium amalgam was prone to give the product of carbonyl reduction as a contaminant. Heating lactone 9a at reflux in tetrahydrofuran (THF) with tritiated water and diazabicyclononane gave the C3H-compound 23a, which could then be reduced to the C3H-triol 24a.Scheme 4 displays the approach to the C3H-3,26-diol 3a. A recent and efficient route' to the aldehyde 18c from pregnen- olone 25 was utilised. Condensation of the aldehyde with the anion derived from deprotonation of 2,3-dimethylbut-2-enoic acid gave the lactone 28c, epimeric at C-22 with 9a.lo Reduction of the lactone with sodium 3Hborohydride in THF-methanol J. CHEM. SOC. PERKIN TRANS. 1 1991 .*...JCHO i-iv { H 17 H H 20 19 viiI 0g amp; vii H H 21 22 ix, x I *I xi, xiii H 23 xiiiI H 24 Scheme 3 Reagents: i, Ac,O, Py; ii, PhIBr,; iii, 0,; iv, Zn, AcOH; v, Ph,P=CHCOMe; vi, H,O,, HO-; vii, Al-Hg; viii, MeCHBrCOBr; ix, (EtO),P; x, NaH; xi, HO; xii, ,HHO, DBN; xiii, LiAIH,.* denotes 3 H site. gave the C3H-diol 29. Protection of the primary alcohol, mesylation, reduction with lithium aluminium hydride and deprotection afforded the desired C3H-24(25)-en-26-ol 31a. Inconveniently, almost equal quantities of the isomeric alcohols 32a and 33a were formed, but the mixture could be separated by HPLC to give the required compound. Results and Discussion The five sterols 14a, 15a, 23a, 24a and 31a were administered in Tween 2@-water-2-methoxyethanol to cut stems of 7-week old J. CHEM. SOC. PERKIN TRANS. 1 1991 {irH 25 26 iii!sctio~ iv H 18 27 28 29 i, vii1 @ OTBDMS -H 31 30 + 3332 Scheme 4 Reagents: i, TBDMSCl; ii, CH,=PPh,; iii, dicyclohexyl- borane; iv, PCC; v, Me,C==C(Me)CO,Et, BuLi-HMPA; vi, ,HNaBH,, MeOH, THF; vii, MsCI; viii, LiAIH, Nicandra physaloides plants.After 4 days, the plants were dried and Nic- 1 was isolated chromatographically and recrystallised to constant activity. The outcome of the experiments is shown in Table 1; absolute and specific incorporations are given and are based on sterol uptake. It can be seen that both 24- methylenecholesterol 14a and its 24(25) double bond isomer 15a are incorporated into Nic-1, at levels expected in such experiments. The triol 24a shows a similar specific incorporation, but the putative intermediate alcohol 31a was relatively poorly accepted.This suggests that the pathway from 3-6 pre-ferentially uses intermediate 4 rather than 5, with the last perhaps taking on a minor role. Clearly a competitive experiment between 4 and 5 is desirable to confirm this issue, but a viable synthesis of 22a-alcohol4 is required for such a test. So far we have not been able to achieve this: the closest parallels are afforded by the recent syntheses of 22R-hydroxylanosterol 34 35 32 7 Scheme 5 Stereochemistry of the 2-3 conversion in N. physaloides 6 4.9pH 'OH C-Me, 6 1.34, 1.22, 1.20 64.92 #J 5.0 4.0 3.0 2.0 1.o 0.0 p.p.m. Fig. 1 'H NMR of Nic-1 from N. physaloides fed with 3'-CD,-MVA and 22R-hydroxydesmosterol,' which employed an arsenic ylid to form the 22-23 bond.However our attempts to extend this work to the 24-methyl series were fruitless. The lactone 23a was a distinctly poorer precursor than the triol 24a, although structurally closer to Nic-1. This suggests that the predominant pathway to the lactol7 from the sterol 6 proceeds by way of a C-26 aldehyde, in preference to reduction of the lactone 9 (although the latter may constitute a minor route). This view is reinforced by the outcome of administration of 3'-C2Hmevalonolactone 34 to N. physaloides. Examination 294 J. CHEM. SOC. PERKIN TRANS. 1 1991 Table 1 Weight of dry leaf/g Weight of Nic-1 /mg Specific activity of Nic-1/dpm mol-' Absolute Incorporation () Specific Incorporation (I Sterol 14a 32.9 68 7.748 x lo4 0.234 0.056 15a 31.4 66 8.825 x lo4 0.158 0.068 23a 31.0 109 2.18 x 10' 0.168 0.015 24a 30.0 95 5.45 x lo8 0.50 0.057 31a 40.0 15 2.09 x 109 0.033 0.012 of the 2HNMR spectrum of the resulting Nic-I, see Fig.1, shows signals arising from (i), methyl groups C-19, C-21 as expected; (ii), 18-H, see ref. 3 and (iii), 26-H (6 4.9). This indicates that C- 26 of Nic-1 was derived from mevalonic acid C-6 with a substantial degree of retention of hydrogen, thus excluding lactone 9 from the dominant pathway.* Scheme 1 thus summarises the present knowledge of the oxidative development of the Nic-1 side chain, from 24-methylenecholesterol 2-3+6 (chiefly via 4-743). A further stereochemical inference may be drawn from these experiments, as in Scheme 5. Mevalonic acid C-6 (3'-methyl) provides C-27 of desmosterol 35.It has been shown l3 that, in the biosynthesis of 24-methylenecholesterol 2a in Physalis peruviana cell cultures,? C-27 of 35 becomes C-26 (the pro-R methyl) of 2.Since we have correlated C-26 of Nic-17 and the C- 27 of sterol 3,with mevalonic acid C-6, it follows that thepro-R methyl of sterol 2 becomes the pro-Z methyl (C-27) of sterol 3, implying p-face loss of 25-H. Net C-24 methylation of desmosterol 35 proceeds with retention of double bond geometry. Experimental For experimental generalisations see ref. 3. All Jvalues are in Hz. Preparation of 28-''C-24-Methylenecholesterol.- I4C-Methyltriphenylphosphonium iodide (2.22 x lo8 dpm) in dry ethanol (1 cm3) was syringed into a Reactivial and the solvent was removed under nitrogen.Unlabelled salt (0.2 g) was added and the mixture was suspended in THF (0.73 cm3). Butyllithium 1.5 mol dm-3 (0.4 cm3); was added with stirring; after 5 min a red solution formed, to which was added 24-ketocholesterol (0.1 g) in dry THF (0.73 cm3). The reaction mixture was stirred overnight and then it was diluted with ether and washed with brine. Evaporation of the solvent afforded a crystalline solid which was purified on a silica column (ethyl acetate-hexane, 1 :9-3 :9), and then recrystallized to yield the title compound (0.41 g, 4.55 x lo7dpm). Preparation of 28-'4C-24-Methylcholesta-5,24-dien-3P-ol.-28-'4C-24-Methylenecholesterol (0.025 g, 2.78 x lo7 dpm) and iodine (2 mg) were heated at reflux in benzene (1.25 cm3) for 3 d.The mixture was evaporated and the residue was separated on a silica column using ethyl acetate (10-30) in hexane, to afford the title compound (0.025 g, 1.9 x lo7 dpm) as white * Veleiro et d.'*have isolated labelled withaferin A from Acnistus breuijlorus fed with 2-14C-MVA. Degradation revealed 14C at C-26 (lactone carbonyl), in discord with results reported here, but only 1.6 of the radioactivity of withaferin A was located at C-26; the authors consider that the biosynthesis of this withanolide involves side-chain degradation and resynthesis, with loss of label. t In ref. 4, we quoted the original conclusions of Seo et d.;l 30 however these were reversed in a later paper,'3b following reassignment of NMR signals.crystals, identified by TLC comparison with unlabelled material l4 made by the same method. Preparation of (22R)-3P-Hydroxyergosta-5,24-dien-22,26-o1ide.-Sodium hydride (60 dispersion in oil; 24 mg); was added to a stirred solution of phosphonate (0.44 g) (from bromide 22)8in dry THF (30 cm3) under nitrogen. The solution was refluxed for I h, cooled and diluted with water (30 cm3). The mixture was extracted with ether. The extracts were evaporated and the residue was chromatographed on silica ethyl acetate (1amp;30) in hexane to yield the 3P-acetoxylactone 9b, (0.25 g, 76), m.p. 232-233 "C from ethanol (lit.,* m.p. 233 "C).This sample was heated at reflux in 4 methanolic potassium hydroxide (40 cm3) for 30 min. After acidification of the mixture, the product was isolated by ether extraction, to yield the 3p- hydroxy lactone 9a (0.24 g), m.p. 219-222 "C from ethanol (Found: C, 78.3; H, 10.1. Calc. for CZ8H42O3: C, 78.9; H, 9.9); vmax/cm-' 3450 and 1697. Preparation of 2H- and H-(22R)-3P-Hydroxyergosta-5,24-dien-22,26-olide.-The hydroxy lactone 9a (25 mg) was dissolved in dry THF (1 cm3) with diazabicyclononane (15 mg) and deuterium oxide (0.2 g) in a Reactivial and the solution was heated at 60 "C for 72 h. The product was diluted with ether and washed with water. Evaporation of the solvent gave the 2H,hydroxy lactone (23 mg), m/z 430(5, MD,), 429(15, MD,), 428(22, MD2), 427(20, MD) and 426( 12, M).A similar preparation using tritiated water (0.05 cm3; 5.55 x 10" dpm) in THF (0.45 cm3) gave the 'HI hydroxy lactone (24 mg, 2.38 x lo8 dpm). 3 p,22,26- Trihydroxyergosta-5,24-diene.-The hydroxy lac-tone 9a (20 mg) was heated at reflux in dry THF (2 cm3) with lithium aluminium hydride (20 mg) for 1 h. The cooled mixture was treated with ethyl acetate (10 cm3) and the emulsion was washed into a Soxhlet thimble. Continuous extraction with chloroform and evaporation of the solvent gave the title compound, (18 mg, 89), m.p. 127-129 "C (Found: C, 78.06; H, 10.32. C28H4603 requires C, 78.14; H, 10.70); 6, 0.72 (3 H, s, 18-H), 1.02 (3 H, s, 19-H), 1.71 and 1.81(6 H, s, 27-H and 28-H), 3.50 (1 H, 3a-H), 3.76 (2 H, m, 26-H), 4.36 (1 H, m, 22-H) and 5.38 (1 H, m, 6-H).3H-3P,22,26-Trihydroxyergosta-5,24-diene.-3H Hy-droxy lactone (20 mg) was reduced as in the preceding experi- ment, to yield the C3H-triol (19 mg), 1.84 x lo8 dpm. 3p-(t-Butyldimethylsily1oxy)ergosta- 5,24-diene-22,26- o1ide.- Butyllithium (1.6 mol dm-3 0.9 cm3); was added to di-isopropylamine (0.3 cm3) in dry THF (1 cm3), and the resulting solution was stirred for 15 min and then cooled to -50 "C. A mixture of 2,3-dimethylbut-2-enoate (100 mg) and hexamethyl- phosphoramide (0.05 cm3) was added and the resulting yellow solution was stirred at -50 "C for 3 h. 3~-(t-Butyldimethylsilyl-oxy)-23,24-dinorchol-5-enaldehyde(210 mg) in dry THF (2.5 cm3) was added and the resulting solution was maintained at J.CHEM. SOC. PERKIN TRANS. 1 1991 -20 "C for 15 h. The product was partitioned between ether and water and the organic layers were dried and evaporated. Chromatography of the residue on silica (ethyl acetate-hexane 1:19) gave the title compound (1 10 mg, 4373, m.p. 203-204 "C from benzene; Found: C, 75.55; H, 10.5; m/z 483.328. CJ4H,,O3Si requires C, 75.50; H, 10.44; M -C4H9, 438.329); v,,,/cm-' 1690; 8,0.05 (6 H, s, 2 x Me), 0.62 (3 H, s, 18-Me), 0.83 (9 H, s, Bu'), 0.92 (3 H, s, 19-Me), 0.98 (3 H, d, 21- Me), 1.72 (3 H, s, 27-H), 1.87 (3 H, s, 28-H), 2.65 (2 H, m, 23-H), 3.48 (1 H, m, 3-H), 4.41 (1 H, br d, 22-H) and 5.30 (1 H, m, 6-H). 3~-t-Butyldimethylsilyloxy)-22,26-dihydroxyergosta-5,24-diene.-The above lactone (1.4 g) was dissolved in dry THF (20 cm3) and methanol (3.5 cm3); sodium borohydride (0.9 g) was added and the resulting mixture was stirred at ambient temperature for 3 h.The mixture was diluted with water and extracted with ether. The extracts were dried and evaporated. Chromatography of the residue on silica (ethyl acetate-hexane, 1 :4) gave the title diol (1.3 g, 92), m.p. 195 OC from ethanol (Found: C, 74.7; H, 11.35; M, 544.431. C,,H6,03Si requires C, 74.94; H, 11.1 1; M +,544.43 1); ?jH0.06 (6 H, s, 2 x Me), 0.70 (3 H, s, 18-Me), 0.89 (9 H, s, Bu'), 0.96 (3 H, d, 21- Me), 1.01 (3 H, s, 19-Me), 1.71 and 1.81 (each 3 H, s, 27-H and 28-H), 2.75 (2 H, m, 23-H), 3.48 (1 H, m, 3-H), 3.73 (2 H, br, 22-H and 26-H), 4.33 (1 H, d, J 11.2,26-H) and 5.31 (1 H, m, 6-H).This compound, with acetic anhydride-pyridine, formed a diacetate, 1.99 and 2.06 (each 3 H, s, Ac). In a radiochemical experiment, the lactone (200 mg was reduced with sodium C3Hborohydride (7 mg, 250 mCi), to yield the C3H-diol (200 mg, 5.0 x lo1, dpm mol-'). 3 P,26- Di-( t-butyldimethylsilyloxy)-22-hydroxyergosta-5,24-diene.-The diol (410 mg) from the previous experiment was dissolved in dry DMF (10 cm3) with imidazole (150 mg) and t- butyldimethylsilyl chloride (160 mg). The mixture was kept at ambient temperature for 12 h and then diluted with water. The mixture was extracted with ether and the extracts were washed, dried and evaporated to yield the title alcohol (490 mg, 98), pure by TLC (ethyl acetate-hexane, 1 :4) (Found: m/z 658.516.C40H7403Si2requires M+, 658.517). A parallel preparation with C3H-labelled diol afforded the 3H-alcohol. 3 P-26-( t-Butyldimethylsilyloxy)-22-(methylsulphonyloxy)erg-osta-5,24-diene.-The alcohol (700 mg) from the previous experiment in dry pyridine (10 cm3) was cooled to 0 "C and methanesulphonyl chloride (0.63 cm') was added. The mixture was stirred at 0 OC for 10 min and then at ambient temperature for 5 min and then diluted with water. The product was extracted with ether. The extracts, after being washed, dried and evaporated, yielded the title sulphonate, (790 mg, 97) as a colourless oil, pure by TLC, 8H 0.05 and 0.08 (each 6 H, s, 2 x Me), 0.67 (3 H, s, 18-Me), 0.88 and 0.89 (each 9 H, s, Bu'), 0.95 (3 H, s, 19-Me), 1.0 (3 H, d, 21-Me), 1.69 and 1.72 (each 3 H, s, 27- and 28-H3), 2.94 (3 H, s, MeSO,), 3.48 (1 H, m, 3-H), 4.13 (1 H, m, 26-H), 4.26 (1 H, d, J 11,26-H), 4.84 (1 H, m, 22-H), 5.34 (1 H, m, 6-H); required number only of peaks in the 13CNMR spectrum.The C3H-methanesulphonate was similarly prepared. 3P,26-Dihydroxyergosta-5,24-diene.-The methanesulphon-ate (500mg) from the previous experiment was dissolved in dry THF (10 cm3) and lithium aluminium hydride (0.15 g) was added. The mixture was heated at reflux for 5 h and then it was cooled, quenched with water and filtered. The filtrate was extracted with ether and the precipitate was washed with the same solvent.The combined ethereal solutions were dried and evaporated. The residue showed a spot on TLC at Rf0.95 and the corresponding fraction was isolated from a silica column (ethyl acetate-hexane, 1 :19). This product was desilylated using tetrabutylammonium fluoride in THF, to yield material showing one TLC spot (ethyl acetate-hexane, 1:1) which was separated by HPLC on a p-Porasil column (300 x 7 mm), (ethyl acetate-hexane, 1 :4). The first compound that eluted was (E)-3P,26-dihydroxyergosta-5,22,24-triene(Found: mjz 41 2.334, 394.322. C28H4402 requires M+, 412.334; M -H20 394.322); 13~0.55 (3 H, s, 18-Me), 1.0 (3 H, s, 19-Me), 1.67 and 1.77 (each 3 H, s, 27- and 28-H), 3.47 (1 H, m, 3-H), 4.25 (2 H, s, 26-H2), 5.34 (1 H, m, 6-H), 5.52 (1 H, dd, J8.8, 15.4,22-H) and 6.44 (1 H, d, J 15.4, 23-H).The second compound that eluted was the desired 38,26- dihydroxyergosta-5,24-diene(Found: mlz 414.350, 396.342 and 314.258. C2sH4602 requires M', 414.350; M -H20, 396.339; M -C6H120, 314.261); 8H 0.68 (3 H, S, 18-Me), 0.96 (3 H, d, J 7,21-Me), 1.0 (3 H, s, 19-Me) 1.67 and 1.79 (each 3.H, s, 27-and 28.H),3.53(1H,m,3-H),4.11(2H,s,26-H),and5.35(1H,m,6-H). In a radiochemical preparation from 3Hmethanes-ulphonate, the ,H-diol was obtained (8.5 mg, 4.54 x lo', dpm mol-'). The third compound that eluted was (E)-3P,26-dihydroxy- ergosta-5,20(22),24-triene(Found: m/z 412.334 and 394.322. C28H4402 requires M', 412.334, M -H20, 394.324); gH0.56 (3 H, s, 18-Me), 1.0 (3 H, s, 19-Me), 1.68,1.69 and 1.77 (each 3 H, s, 21,27- and 28-H), 2.85 (2 H, d, J7, 23-H), 3.53 (1 H, m, 3-H), 4.13 (2 H, s, 26-H), 5.06 (1 H, t, J 7,22-H) and 5.36 (1 H, m, 6-H); 6c 18.05 (21-Me).Administration Experiments with Nicandra physaloides Plants.-28-'4C-24-Methylenecholesterol (9 mg, 9.99 x lo6 dpm) was dissolved 2-methoxyethanol (2.5 cm3) and Tween 20 (40 mg); water (88 cm3) was added in portions with sonication, to give a clear solution. Aqueous solutions of 28-14C-24- methylcholesta-5,24-dien-3~-ol(l5mg, 1.299 x lo7 dpm), C3H- 3P-hydroxyergosta-5,24-dien-22,26-olide(9.0 mg, 8.94 x lo7 dpm), 3H-3P22,26-trihydroxyergosta-5,24-diene(10.0 mg, 9.7 x lo7 dpm) and 26-3H-3P,26-dihydroxyergosta-5,24-diene (8 mg, 6.49 x 10' ' dpm) were prepared in the same way.Each solution was administered to the cut stems of N. physaloides plants, 67 weeks old; the solutions were taken up within a few hours. After 4 d, the plants were air dried and extracted with ether for 3 d at ambient temperature. Nic-1 was isolated from the evaporated extracts by chromatography on a silica column (ether-chloroform, 1 :l), and recrystallised from benzene to constant activity. In each case, some unmetabolised starting sterol was recovered from the column and incorporation figures were corrected appropriately. References 1 E. Glotter, I. Kirson and A. Abraham in Bioorganic Chemistry, ed. E. E. van Tamelen, Academic Press, New York, 1978, vol. 11, p. 57; I. Kirson and E. Glotter, J.Natf. Prod., 1981, 44, 633; H. K. Gill, Ph.D. Thesis, University of Nottingham, 1986. 2 M. J. Begley, L. Crombie, P. J. Ham and D. A. Whiting, J. Chem. Sue., Perkin Trans. I, 1976,296 and 304. 3 H. K. Gill, R. W. Smith and D. A. Whiting, J. Chem. SOC.,Perkin Trans. I, 1990,2989. 4 H. K. Gill, R. W. Smith and D. A. Whiting, J. Cheni. Soc., Chem. Commun., 1986,1459. 5 (a) W. J. S. Locksley, H. H. Rees and T. W. Goodwin, Phyrochemistry, 1976,15,937. (b) P. L. C. Lu, M. M. El-Olemy and S. J. Stohs, Lloydia, 1974, 37, 593. 6 E. S. Wallis and E. Fernholz, J. Am. Chem. Soc., 1935, 57, 1504; S. Kuwada and M. Yago, J. Pharm. Sac. (Japan),1937,57,963. 7 M. Fryberg, A. C. Oelschlager and A. M. Unrau, Terrahedron, 1971, 27, 1261. 8 E. Glotter, M. Zviely and I. Kirson, J. Chem. Res., 1982, (M),373; J. CHEM. SOC. PERKIN TRANS. 1 1991 (S), 32; G. R. Weihe and T. C. Mcmorris, J. Org. Chem., 1978, 43, 13 (a)S. Seo, A. Uomori, Y. Yoshimura and K. Takeda, J. Am. Chem. 3942. SOC.,1983,105,6343.(b)J. Chem. SOC.,Chem. Commun., 1984,1174. 9 M. M. Midland and Y. C. Kwon, Tetrahedron Lett., 1985,26, 5017. 14 W. J. S. Locksley, D. P. Roberts, H. H. Rees and T. W. Goodwin, 10 N. Ikekawa, A. Kajikawa, H. Saito, M. Hirayama and M. Ishiguro, Tetrahedron Lett., 1 974, 3773. Heterocycles, 1981, 15, 823. 11 A. Almain, G.Ourisson and B. Luu, Synthesis, 1987,696, 1002. Paper 0/03410J 12 A. S. Veleiro, G. Burton and E. G. Gros, Phytochemistry, 1985, 24, Received 26th July 1990 2263. Accepted 5th September 1990
展开▼
