Studies in terpenoid biosynthesis. Part XVI. Formation of the sesquiterpenoid trichothecin

摘要

1212 J.C.S. Perkin IStudies in Terpenoid Biosynthesis. Part XV1.l Formation of the Ses-quiterpenoid TrichothecinBy Roger Evans, James R. Hanson,' and Tim Marten, The School of Molecular Sciences, University of Sussex,The biosynthetic sequence trichodermol (2) + 12.1 3-epoxy-4~,8-dihydroxytrichothec-9-ene (3) __ttrichothecolone (4) ,-+ trichothecin (5) in Tricbotbecium roseum has been established. Earlier evidencesuggesting the intervention of crotocin (6) has not been substantiated.Brighton BN1 9QJTRICHODIENE (1) has been shown to be a precursor oftrichothecolone (4) in Trichothecium ~oseum.~~ Althougha number of plausible intermediates have been isolated,there is no biosynthetic information on the interveningstages. The introduction of the C-8 carbonyl group oftrichothecolone and the order of the late stages in itsbiosynthesis are the subjects of this paper.We haveconsidered a number of schemes. One attractive hypo-( 2 ) R = H,( 3 ) a j R = ac-OH,H( 4 ) R = 0b;R = ~ - O H , HI o2 C-CH f C H M ~(5) ( 6 )thesis was that the C-8 and C-11 oxygen functions wererelated and that they arose by rearrangement of aA9-8,11-epidioxide, itself formed from a ring A diene.4There are precedents for the formation of dienes in therelated cuprenene series, and this biogenetic scheme mightalso be reflected in the structure of helicobasidin. Asecond hypothesis involved the rearrangement of a7,s-epoxide. A 7,8-epoxide, crotocin (6) ,5 co-occurswith trichothecin in T. yoseum.6 A third proposal in-volved simple hydroxylation a t C-8 followed by oxid-ation.'The introduction of the C-8 carbonyl group prior tothe formation of the 2,ll-ether linkage of the 12,13-epoxytrichothecene skeleton is implicit in a mechanisminvolving an 8,l l-epidioxide.Trichodermol (2), a1 Part XV, J. R. Hanson, T. Marten, and R. Nyfeler, J.C.S.Perkin I , 1976, 876.2 Y . Machida and S. Nozoe, Tetrahedron Letters, 1972, 1969.3 R. Evans, A. M. Holtom, and J. R. Hanson, J.C.S. Chem.Cotnm., 1973, 465; R. Evans and J. R. Hanson. J.C.S. Perkin I ,1976, 326.Cf. R. J. Conca and W. Bergmann, J . Org. Chem., 1953, 18,1104.J. Gyimesi and A. Melera, Tetrahedron Letters, 1967, 1665.metabolite of Trichoderma species,* lacks a C-8 carbonylgroup. It was shown to be present in Trichotheciumyoseum by dilution analysis of a fermentation to which2-1*Cmevalonic acid had been added (0.07 incorpor-ation).2,10-3H;4,8, 14-14CTrichodermol (2), preparedbiosynthetically from (4R)-4-3H;2-14Cmevalonic acidby using Trichodermn sfiorzdosum, was incorporated intotrichothecin in 3.4 yield by Trichothecium roseurn.This excluded oxygenation at C-8 prior to formation ofthe trichothecene skeleton, and showed that, once theskeleton is formed, oxidation a t C-4 precedes that atC-8. The retention of the two (4R)-4-3Hmevalonatelabels was in accord with the folding of farnesyl pyro-phosphate which has been proposed10 for these com-pounds.In the formation of a 7,8-epoxide, a C-5 and a C-2mevalonoid hydrogen atom are removed from C-7 andC-8, respectively.During the subsequent isomerizationof the epoxide to form the carbonyl group, a C-8 hydrogenatom, originating from C-2 of mevalonate, wouldmigrate to C-7. On this basis trichothecin mightTABLE 1Incorporation of 2-3H, ;2J4C- and Eb3H, ;2-l4C-mevalonates into the metabolites of T . roseurnWH, ;2-14CMevalonic acidDeoxyrosenonolactoneRosenonolactoneTrichothecinTrichothecolone6-3H, ;2-14CMevalonic acidDeoxyrosenonolactoneRosenonolactoneTrichothecinTrichothecolone { $13H : 14CRatio16.56 : 114.06 : 110.41 : 17.23 : 17.10: 116.13 : 116.32 : 116.07 : 113.92 : 1Atom Incorpora-ratio tion8.00: 4 3.86.92: 4 * 6.13.09: 3 * 1.33.03 : 3 *8.09: 4 0.817.97: 4 1.195.17: 3 0.1310.62 : 1 3.95 : 37 .9 : 1 2.89: 3* Calculated on the basis that deoxyrosenonolactone in-corporated 8 labels.incorporate four labels from each of the 2-3H2- and5-3H,-mevalonates. Our earlier results implied thatthis was the case.9 However we have been led torepeat our earlier experiments, with different results(see Table 1). In order to take account of the problems6 B. Achilladelis and J. R. Hanson, Phytochemistry, 1969, 8,765.7 Y . Machida and S. Nozoe, Tetrahedron, 1972,20,6113.8 W. 0. Godtfredsen and S. Vangedal, Acta Chem. Scand.,9 B. A. Achilladelis, P. M. Adams, and J. R. Hanson, J.C.S.10 J. R. Hanson, T. Marten, and M. Siverns, J.C.S. Perkin I ,1965, 19, 1088.Perkin I . 1972, 1425.1974, 10331976 1213which might arise through the action of prenyl isomerase,the atom ratios of the 2-3H2mevalonate experiment arenormalized to those of the diterpene, deoxyrosenono-lactone. This metabolite was also produced by thefungus and has been shownll to incorporate eight2-3HJmevalonate labels.The trichothecin (5) in-corporated three 2-3H2- and five 5-3H,-mevalonoidhydrogen labels. The trichothecin (5) from both feed-ing experiments was subjected to hydrolysis and ex-change in methanE2Hol containing 2~-sodium deuter-oxide. While there was no loss of label by exchangefrom the 2-3HJmevalonate-labelled trichothecolone (4),there was a loss of two 5-3HJ-mevalonoid labels. Themass spectrum of the deuteriated products showed thattwo deuterium atoms had been introduced, and then.m.r.spectrum lacked the resonances at 6 2.31 and2.89 (J 16 Hz) which have been assigned12 to the7a- and 7p-protons. The signal due to the otherpotentially enolisable proton (C-11) at 6 3.85 was un-changed. Thus there are two 5-3H2-mevalonoid labelsat C-7, excluding a A7-olefin and epoxide from the bio-synthesis. The earlier result may have been due to acombination of easy exchange at C-7 and contaminationof the trichothecin with rosenonolactone. Rosenono-lactone not only incorporates mevalonate more efficientlythan trichothecin but also co-crystallizes with it.The third route to trichothecolone involves hydroxyl-ation of trichodermol at C-8 followed by oxidation.Reduction of trichothecolone with sodium borohydrideafforded a separable mixture of the 8a- and 8P-alcohols(3a and b).Previous workers had only obtained oneisomer. In view of the similarity of the 7,8-couplingconstants, the stereochemistry of these alcohols wasassigned on the basis of solvent shifts in the 1H n.m.r.spectrum (see Table 2). The methyl group resonancesTABLE 2lH N.m.r. solvent shifts for the epimeric 12,13-epoxy-4p, 8-dihydroxytrichothec-9-enes8a-Alcohol 8P-Alcohol-7 rProton CDCI, C,D,N CDCI, C,D,N t13-H 2.84,3.11 2.90,3.12 2.83,3.11 2.74,2.8114-H 0.85 1.16 0.82 0.9215-H 1.07 1.37 0.91 1.04showed larger solvent shifts in the 8~-alcohol (3a), andthe p-oriented C-13 epoxide protons were shielded to agreater extent in the 8p-alcohol (3b).4,8,14-14CTri-chothecolone was reduced with sodium borohydride andan inactive sample was also reduced with pHsodiumborohydride. The epimers were separated in each case.The labelled alcohols were then mixed in a suit-able 3H : 14C ratio to afford 8p-3H;4,8,14-14C12,13-epoxy-4~,8a-dihydroxytrichothec-9-ene (3a) and 8a-3H; 4 , 8,l 4-14C 12,13-epoxy-4p, 8 p-dih ydrox ytrichot hec-9-ene (3b). The alcohols were then incubated with T.11 B. Achilladelis and J. R. Hanson, J . Chem. SOC. (C), 1969,l2 B. K. Tidd, J . Chem. SOC. (C), 1967, 218.2010.roseum. Trichothecolone was isolated from both in-cubations. It showed a 6.6 and 5.6 incorporation,respectively but in neither instance was tritium retained.The comparable efficiency of these incorporations wassurprising.However only 12,13-epoxy-4~,8a-dihydroxy-trichothec-9-ene (3a) has been isolated from the fungusand it is likely that this is the genuine intermediate.The efficiency of incorporation of the other epimer maybe because it is a substrate for the 8a-hydroxylase, i.e.for the preceding step in the biosynthesis. Trichothe-colone, which co-occurs with trichothecin in T. roseam,was converted into trichothecin by the fungus in 27yield. Hence esterification is probably one of the laststages in the biosynthesis.EXPERIMENTALGeneral experimental details have been describedpreviously.l3 The strains of Trichothecium roseum used inthis work were CMI50660 and IF0 6157 and they werecultured as described previ0us1y.l~Dilution Analysis of T.roseum for TyichodermoZ.-2-14CMevalonic acid (250 pCi) in ethanol (10 ml) wasdistributed amongst six 5-day old surface cultures (750 ml)of Tyichothecium roseurn. After a further twelve daysgrowth, the mycelium was filtered and the broth wasextracted with chloroform. The extract was dried andevaporated to give a gum (0.89 g) which was diluted withunlabelled trichodermol (25 mg). The gum was chromato-graphed on silica to afford trichodermol (19 mg), m.p.117-118', 22 590 disint. min-l mg-l (0.07 incorporation).Incorporation of Trichodermol.-2, 10-3H;4, 8,14-14CTri-chodermol (100 mg) (3H : 14C 4.8 : 1; 8.103 disint. min-lmg-1) in ethanol (5 ml) was distributed amongst five 5-dayold surface cultures (750 ml) of T.Yoseum. After a further4 weeks growth, the mycelium was filtered and the brothwas extracted with chloroform. Evaporation of thesolvent gave a gum which was chromatographed on silica toafford trichothecin (90 mg), m.p. 117-118" (3H : 14C,4.7 : 1; 300 disint. min-l mg-1; 3.4 incorporation).Incor9oration of 2-3H,;2-14C- and J~-~H,; 2-14C-MevaZonicAcids.-The mevalonates (50 pCi W) in ethanol (0.5 ml)were added separately to 3-day old shake cultures (100 ml)of T . roseum. After a further ten days, the mycelium wasfiltered and the metabolites were recovered as above andpurified by preparative layer chromatography. Thetrichothecin and deoxyrosenonolactone fractions werediluted with inactive material (20 mg each) and crystallizedto constant activity.Deuterium Exchange Reactions.-The trichothecin (30 mg)from the Z-3H,; 2-14CI- and 5-3H2;2-14C-mevalonate experi-ments was taken up in methan2Hol (2 ml) and 2~-sodiumdeuteroxide (1 ml) was added.The solutions were left atroom temperature for 3 h. The methanol was removedin vacuo and the residue was extracted with ethyl acetate.The trichothecolone was purified by preparative layerchromatography on silica in ethyl acetate-light petroleum(1 : 1). The trichothecolone (i) (23 mg) fromthe Z-3H,; 2-14CI-mevalonate experiment showed no change in 3H : 14C ratio.The results are given in Table 1.l8 J. R. Hanson, J. Hawker, and A. F. White, J.C.S. Perkin I ,14 B. Achilladelis and J. R. Hanson, Phytochemistry, 1968, 7,1972,1892.6891214 J.C.S. Perkin IThe trichothecolone (21 mg) from the 5-3H2;2-14C-mevalonate experiment showed a drop in 3H: atomratio of 1.20.The exchange was repeated on this materialfor a further 24 h. The trichothecolone (ii) (9 mg) which wasrecovered had 3H : 14C 7.9 : 1. The mass spectrum showedYnle 248 Calc. for Cl,Hl,2H,0,: (Mf - H20), 2481.Reduction of TrichothecoZone.-Sodium borohydride (100mg) was added to a solution of trichothecolone (200 mg) inmethanol (3 ml) over 2 h. The solution was acidified withdilute hydrochloric acid. The products were recovered inethyl acetate and purified by preparative layer chromato-graphy on silica in ethyl acetate. Elution of the band atR p 0.7 gave 12,13-epoxy-4P, 8a-dihydroxytrichothec-9-ene(140 mg) , which crystallized from chloroform as needles,m.p.155", a, -57' (c 0.15) {lit.,' m.p. 190-191" (fromacetone-hexane), a, -53") (Found: C, 67.6; H, 8.5.Calc. for C1,H2,0,: C, 67.6; H, 8.3y0), vmX. 3 500, 3 400,1340, and 1070 cin-l, 6 0.85 (3 H, s ) , 1.07 (3 H, s), 1.87(3 H, s), 1.95-2.22 (3 H, m), 2.52 (1 H, d, J 7 Kz), 2.84and 3.11 (2 H, ABq, J 4 Hz), 3.59 (1 H, d, J 6 Hz), 3.82(1 H, d, J 6 Hz), 4.13 (1 H, d, J 6 Hz), 4.35 (1 H, dd, J 4and 7 Hz), and 5.57 (1 H, d, J 6 Hz). Elution of the banda t RF 0.6 gave 12,13-efioxy-4p,8$-dihydroxytrichothec-9-ene(20 mg) , which crystallized as needles from ethyl acetate-light petroleum; m.p. 157-158", a, +35.4" (c 0.26)(Found: C, 67.9; H, 8.0), vmL 3500, 3400, 1340,1070, and 960 cm-l, 6 0.82 (3 H, s), 0.91 (3 H, s), 1.83(3 H, s), 1.9-2.2 (3 H, m), 2.58 (1 H, dd, J 7 and 16 Hz),2.83 and 3.11 (2 H, ABq, J 4 Hz), 3.54 (1 H, d, J 6 Hz),3.85 (1 H, d, J 6 Hz), 4.10 (1 H, d, J 7 Hz), 4.27 (1 H, m),and 5.50 (1 H, d, J 6 Hz).Incubation of 8fb3H;4, 8,142 12,13-E$10xy-4p, 8a-di-hydroxytrichothec-9-ene with T . roseurn.-The labelled4P,8a-diols were mixed and recrystallized to give a sample(15 mg) (3H : 14C 10.03 : 1; 2 x 104 disint.min-1 mg-1 lac)which was dissolved in ethanol (0.3 ml) and distributedamongst three 3-day old surface cultures (750 ml) ofT. roseurn. After a further 5 days, the broth was extractedand the metabolites were isolated to afford trichothecin(20 mg), m.p. 117-118" (3H : 14C 0.07 : 1; 1.0 x lo3 disint.min-1 rng-l; 6.6 incorporation).Incubation of 8w3H;4,8, 14-14C 12,13-Epoxy-4@,8$-di-hydroxytrichothec-9-ene with T.roseurn.-The labelled4@,8fl-diols were mixed and recrystallized to give a sample(1 mg) (3H : 14C 18.6 : 1; 1.8 x lo4 disint. min-l mg-l 14C)which was dissolved in ethanol (0.1 ml) and added to a3-day old surface culture of T. j'oseum. After a further 5days, the broth mas extracted and the metabolites wereisolated to afford trichothecin (20 mg), m.p. 117-118O(3H : 14C 0.12 : 1; 51 disint. min-l rng-l; 5.6 incorpor-ation).In a blank experiment, the mixed C-8 alcohols from asodium borohydride reduction of trichothecin were dis-solved in ethanol and shaken in the T. roseurn culturemedium for 5 days. There was no detectable (g.1.c.)formation of trichothecin. (We thank Mr. B. Dockerill forthis experiment .)Incorporation of Trichothecolone into Trichothecin.-4,8, 14-14CTrichothec~lone (14.5 mg; 7 385 disint. min-lmg-1) (prepared biosynthetically from 2-14Cmevalonicacid) in ethanol (3 ml) was distributed amongst threesurface cultures (750 ml) of T. roseurn 10 days after inocu-lation. After a further 14 days, the mycelium was filteredand the metabolites were isolated. The trichothecin (91mg), m.p. 115-116", showed 779 disint. min-l mg'l (6.7incorporation).We thank the S.R.C. for financial support and Mrs. A.Ward for growing the fermentations.5/2230 Received, 17th November, 1976