1182 J.C.S. Perkin ICarbon4 3 Nuclear Magnetic Resonance Assignments and Biosynthesis ofAflatoxin B, and SterigmatocystinBy Klaus G. R. Pachler, Pieter S. Steyn," Robert Vleggaar, and Philippus L. Wessels, National ChemicalResearch Laboratory, Council for Scientific and Industrial Research, P.O. Box 395, Pretoria 0001, Republicof South AfricaDe Buys Scott, Faculty of Agriculture, University of Pretoria, Pretoria 0002, Republic of South AfricaThe 13C n.m.r. spectra of two secondary fungal metabolites of related structure, aflatoxin 6, and sterigmatocystin,have been completely assigned. The distribution pattern of acetate units has been determined from lSC n.m.r.spectra of aflatoxin B, derived from singly and doubly labelled 13Cacetate. These results and other publisheddata on 13C-labelled sterigmatocystin agree with a biosynthetic scheme based on a C,, polyketide precursor.THE biosynthesis of aflatoxin B, (l), a potent hepato-carcinogen produced by Aspergillus f l a w s and A .parasiticus, and the related metabolite sterigmatocystin(2), produced by e.g.A . versicolor, has been a topic ofJ. S. E. Holker and J. G. Underwood, Chem. and Ind., 1964,1865.R. Thomas, in Biogenesis of Antibiotic Substances,' ed. 2.VanBk and 2. HoSfamp;lek, Academic Press, London, 1965, pp. 155-167.much conflicting speculation.1-5 Extensive l*C-labellingand degradation studies 3 9 5 indicated that the two meta-bolites are totally derived from acetate units, and thatM. Biollaz, G. Biichi, and G. Milne, J . Amer.Chem. SOC.,J. G. Heathcote, M. F. Dutton, and J. R. Hibbert, Chem.6 J. S. E. Holker and L. J. Mulheirn, Cliem. Comm., 1968,1970'92, 1035.and I n d . , 1973, 1027.15761976. I1183methionine contributes the methoxy methyl group. shifts for the various protons facilitated the assignmentTheir acetate origin has subsequently been confirmed of the C(4) and C(5) proton signals. The measuredby feeding experiments with l3C-1abelled acetate.(j-B shifts, extrapolated to an equimolar ratio (Table l),13C N.m.r. spectroscopy provides an efficient tool for indicate that complexation occurs preferentially at thestudying the ambiguities in the biosynthesis of aflatoxin carbonyl groups and the signal at 6 2.62 is thereforeB, and sterigmatocystin, as indicated in our preliminary attributed to the C(4) protons.TABLE 1Proton chemical shifts and proton-proton coupling constants of aflatoxin B, (1) and sterigmatocystin (2),Proton 8H '46913141516OMe6.42 (S)6.80 (D)4.75 (DT)8.46 (T)6.46 (T)3.95 ( S )Aflatoxin B, SterigmatocystinA r 7JE.H/Hz LIS (p.p.m.) b Proton 8H ' JH.HIHz4 6.74 (DD) J4.5 8.2, J4.6 1.05 7.84 (T) Js.6 8.26 6.80 (DD)7.23.01.1 11 6.42 ( S )AA'BB'J13.14 7*3 1.1 14 6.81 (D) Jl~,ls 7.0J14.15 = J14.16 = 2.5 2.5 15 4-78 (DT) Jls.lfi = J15.17 = 2.6J15.16 2.5 2.0 16 5.44 (T) J16.172.50.7 17 6.50 (T)1.0 OMe 3.98 (sja Relative to internal Me,Si; S = singlet, D = doublet, T = triplet. Eu(fod),-Induced shift extrapolated to an equimolar ratio.Aflatoxin B,175I I I I I 1 I I I I 1 I I I 1 I I I I200 150 100 50bcFIGURE 1 Natural-abundance proton-noise-decoupled 25.2 MHz 13C n.m.r.spectrum of aflatoxin B, (1); spectral width 5 500 Hz;pulse delay 9 s; 90" r.f. pulse; transients 5 K. Insert: result of an SPI experiment (see text)communi~ations.~~~ A prerequisite for any biosyntheticstudy with 13C-labelled precursors is the unambiguousassignment of the resonances in the natural-abundance13C n.m.r. spectrum of the compound under study.The proton n.m.r. spectrum of aflatoxin B, (I), withthe exception of the C(4) and C(5) protons, which giverise to an AA'BB' pattern at 6 2.62 and 3.39, has beenpreviously assigned lo (Table 1). The Eu(fod),-inducedM. Tanabe, T. Hamasaki, and H.Seto, Chem. Comm., 1970,1539. ' D. P. H. Hsieh, J. N. Seiber, C. A. Reece, D. L. Fitzell, S. L.Yang, J. I. Dalezios, G. N. la Mar, D. L. Budd, and E. Motell,Tetrahedron, 1975, 31, 661.P. S. Steyn, R. Vleggaar, P. K. Wessels, and D. B. Scott,J.C.S. Chem. Coiutm., 1975, 193.The 13C n.m.r. data of aflatoxin B, (I) derived fromcoupled nuclear Overhauser enhanced (n.0.e.) , protoa-noise-decoupled (p.n.d.), and proton off-resonance de-coupled spectra are given in Table 2. The p.n.d. naturalabundance I3C n.m.r. spectrum is shown in Figure 1.The proton-bearing carbon signals have been assignedby correlating the residual splittings in off-resonancedecoupled spectra with the known proton chemicalshifts-ll The magnitudes of the observed directly9 K.G. R. Pachler, P. S. Steyn, R. Vleggaar, and P. L. Wes-sels, J.C.S. Chem. Comm., 1975, 355.10 T. Asao, G. Biichi, M. M. Abdel-Kader, S. B. Chang, E. L.Wick, and G. N. Wogan, J. Amer. Chem. SOG., 1965, 87, 882.11 K. G. R. Pachler, J. Magnetic Resonance, 1972, 7 , 442J.C.S. Perkin Ibonded 13C-H coupling constants (Table 2) support theseassignments. Identical assignments for these proton-bearing carbon atoms, based on comparisons with modelcompounds, were recently reported by Hsieh et a1.'TABLE 213C Chemical shifts, directly bonded ( ~ J I s ~ , ~ ) and long-rangeJIB^,^) coupling constants of aflatoxin B, (1) andcoupling constants ( 1 J ~ 8 c c . ~ ~ c ) of 1, 2-13Cacetate-enrichedaflatoxin B, (J in Hz)Carbonatom ac (I 'J1'C,H "J1'C,H lfiC,'sC1 155.2 (S) S2 117.4 (S) 603 201.3 (Stt) 6,3 404 35.1 (T) 131 406 177.1 (Stt) 7,3 607 104.0 (Sd) 6 648 161.6 (Sq) 4 719 90.9 (D) 165 7110 165.8 (Sdd) 7,3 6111 107.9 (Sdd) 9,3 6112 153.0 (S) 645 29.0 (T) 134 S13 113.6 (Dddd) 187 9,6,3 3314 47.9 (Dq) 147 6 3315 102.7 (Dddd) 182 13,5,2 7516 145.4 (Ddt) 196 12,4 75OCH, 56.6 (Q) 146 S0 Relative to internal Me,Si.Capital letters refer to thepattern resulting from directly bonded protons and smallletters to long-range 1-H coupling. S = singlet, D or d =doublet, T or t = triplet, Q or q = quartet.In the assignment of the quaternary carbon atomsignals extensive use was made of techniques based onlong-range (more than one bond) 13C-H couplings. Theresult of an SPI (selective population inversion) experi-ment l2 in which a x-pulse of 0.085 s is applied at 3 Hz tohigher field of the C(9) proton resonance is shown in aninsert in Figure 1.The two affected resonances S 104.0(Sd) and 107.9 (Sdd) are assigned to C(7) and C ( l l ) , asthree-bond 13C-H couplings are usually larger than two-bond couplings in aromatic systems.13 The smallerobserved intensity changes of the resonance at 6 107.9are due to the additional splitting of this resonance bythe C(14) proton as well as the larger coupling (9 Hz)to the C(9) proton, which resulted in only a partialpopulation transfer of the affected transition. Theassignment of C(7) ( 6 104.0) and C(l1) (6 107.9) signals iscompatible with the observed long-range multiplicityof these signals and was confirmed by data obtained forZ-Wlacetate-derived aflatoxin B, (see later).The remaining resonance in the 90-120 p.p.m.region(6 117.4) of the spectrum of aflatoxin B, arisesfromeitherC(2) or C(6). The corresponding carbon atoms incoumarin, C(3) and C(4), resonate at 8 116.3 and 142.8,respectively l4 whereas in styrene the amp;-carbon atomresonates at 6 135.8 and the p-carbon at 6 112.3.16 Simi-larly, in cyclopent-2-enone the corresponding carbonatoms resonate at 6 132.9 and 164.2, respectively.16l2 K. G. R. Pacliler and P. L. Wessels, J . Magnetic Resonance,1973, 12, 337.13 G. C. Levy and G. L. Nelson, ' Carbon-13 Nuclear MagneticResonance for Organic Chemists,' Wiley-Interscience, New York,1972, p.100, and references cited therein.On the basis of these chemical shifts the peak at 6 117.4should be assigned to C(2) in aflatoxin B,. Additionalevidence for this assignment was provided by data ob-tained for aflatoxin B, derived from p3Cacetate (seelater).15( 3 )16JC(1) of acetateo C(2)of acetateOH 0 OMe18(51Selective decoupling of the C(4) protons results in thecollapse of the multiplets at 6 201.3 and 177.1 to tripletswith couplings of 3 and 7 Hz, respectively. Conversely,decoupling the C(5) protons simplified the same multi-plets to triplets with 6 and 3 Hz couplings, respectively.The effect of these selective proton decouplings on theappearance of the C(2) resonance could not be observedas a result of the overlap of the C(2) resonance and thehigh frequency leg of the C(13) resonance.The reson-ance at 6 201.3, characteristic of a ketone carbonylcarbon atom, is assigned to C(3). The consequentialassignment of the signal at 6 177.1 to C(6) is corroboratedby the data obtained for 1-l3Cacetate-derived aflatoxinB, (see later).Selective decoupling of the C(13) proton removes a6 Hz, a 2 Hz, and a 4 Hz splitting from the resonancesdue to C(14), C(15), and C(16), respectively. Theresonance at 6 165.8 (Sdd) changes to a doublet ( J 3 Hz)and is therefore attributed to C( 10).l4 R. D. Lapper, Tetrahedron Letters, 1974, 4293.l5 J. B. Stothers, ' Carbon-13 NMR Spectroscopy,' Academicl6 Reference 15, p. 193.Press, New York, 1972, p. 2041976 1185The resonances at 6 153.0 and 155.2 exhibit no long- has only one proton three bonds away, it follows that therange coupling and appear as singlets in the coupled resonance at 6 161.6(Sq) is due to C(8).Additionaln.0.e. spectrum, and the resonance at 6 161.6 is a multi- support for this assignment is obtained from the study ofIAftatoxin B1Hi3C02Na( C 1 1753 6 10 8 112 16 213 11 15 9 17 14 4 57 II I J?LI3C Hg C02NaI16 2 117 9 14 4CHi3C02 NaL5 1 3 6 108 112 13 15I I l l , I l l 1 I I l t I t200 150 100 50 s6 ,FIGURE 2 Proton-noise-decoupled 26.2 MHz I3C n.m.r. spectra of aflatoxin B, derived from (a) l-13Cacetate; spectral width 6 600Hz, pulse delay 6 s, 80" r.f. pulse, transients 40 K; (b) 2-x3Cacetate; spectral width 5 500 H; pulse delay 1 s 80" r.f.pulse,transients 32 K; (c) 1,2-I3Cacetate; spectral width 5 500 Hz, pulse delay 6 s, 80Or.f. pulse, transients 28 Kplet with at least quartet-like pattern. These resonances sterigmatocystin (see later). The assignments of C(l)are ascribed to C(I), C(8), and C(12). As C(l) is four (6 155.2) and C(12) (6 153.0) are based on the results of abonds removed from the nearest protons and as C(12) Eu(fod),-induced shift experiment. As complexatio1186 J.C.S. Perkin Ioccurs at the ketone carbonyl group the signal due toC(4) which is a @-carbon atom relative to the complex-ation centre, shifts upfield by 5 p.p.m.l', addition ofEu(fod), (ca. 1 : 4 molar ratio) resulted in a downfieldshift of 4 p.p.m. for the signal at 8 155.2 C(l), whereasno shift was observed for the signal at 8 153.0 C(l2).The assignments of the quaternary carbon signals differfrom those reported recently by Hsieh et aL7 In theirstudy, use of model compounds led to an interchange ofthe C(2) and C(7) assignments and a wrong assignmentfor C(1). Furthermore, C(8), C(lO), and C(12) were notassigned individually.assignment of the resonance at 8 107.9 to C(11) is confirmedby the 13C--13C coupling of 44 Hz between C(14) at 6 47.9and C(11).The 13C-13C coupling of 21 Hz, seen clearlyas satellites on the C(4) resonance and weakly on theresonance at 6 117.4 (Figure 2b, insert), lends furthersupport to the assignment of the latter resonance toC(2). The geminal C-C coupling constant in acetone is16.5 Hz.lgThe labelling pattern of aflatoxin B, derived from thesingly 13C-labelled acetates is consonant with thereported 14C-labelling patterns (3).3The p.n.d.13C n.m.r. spectrum of aflatoxin B, enrichedSt erigmatocyst in6OH 0 OMc181 103 78 17 5 14 2 616 1112 4 918 15I 1 I I I I I I (13) I I I I 1 I150 100 506,FIGURE 3 Natural-abundance proton-noise-decoupled 25.2 MHz l3C n.1n.r. spectrum of sterigmatocystin (2) ; spectral width5 000 Hz; pulse delay 2 s; 75" r.f. pulse; transients 4 IAdditional evidence corroborating our assignments wasprovided by the spectra of 1-13C- and 2-13C-acetate-derived aflatoxin B,.Cultures of A . fEavus, strain NRRL 3251, were grownby the replacement technique l 8 and supplemented withsodium 1--, 2-13C-, or 1,2-13C-acetate (all 90).Aflatoxin B, was in each case enriched approximatelyeight-fold above the natural 13C abundance, as determinedby mass spectrometry.The 1-13Cacetate-derived aflatoxin B, spectrum(Figure 2a) shows nine enriched carbon signals.Thecoupling of 34 Hz between C(5) and C(6) provided ad-ditional evidence for the assignment of C(6) (6 177.1).7Similarly the 2-13Cacetate-derived aflatoxin B, spectrum(Figure 2b) shows seven enriched carbon signals. Thewith 1,2-13Cacetate (Figure 2c) showed satelliteresonances due to 13C-13C spin-spin couplings. Theobserved spin-spin coupling data (Table 2) indicatedthat C(2)-C(6), C(3)-C(4), C(7)-C(12), C(8)-C(9), C(10)-C(11), C(13)-C(14), and C(15)-C(16) originated fromseven intact acetate units, arranged as shown in (4).Both the C(l) and C(5) resonances are enhanced in the13C spectrum of the sample derived from l-13Cacetate,but show no one-bond 13C--13C coupling in the doubly lab-elled derivative.These carbon atoms were thereforederived from two separate acetate units, each of which17 D. H. Williams, Pure Appl. Chena., 1974, 40, 25.18 D. P. H. Hsieh and R. I. Mateles, APPZ. Micvobiol., 1971, 22,lS Reference 15, p. 374.791976 1187lost one acetate-derived methyl carbon atom in the bio-synthesis of aflatoxin B,.The natural-abundance 13C n.m.r. spectrum of sterig-matocystin (2) is shown in Figure 3. The 13C n.m.r.data of (2) derived from coupled, p.n.d., and off-reson-ance decoupled spectra are given in Table 3.Methoxyand tertiary carbon signals have been assigned by cor-relating residual splittings in off -resonance proton-decoupled spectra with the known proton chemicalshifts (Table 1). The proton chemical shifts are inagreement with reported values 2o with the exception ofthose of the C(4) and C(6) protons. The assignment ofthe C(4) (6 6.74) andC(6) (6 6.80) protonsignals is based onthe results of recent nuclear Overhauser effect experi-ments on sterigmatocystin and related compounds.21Directly bonded C-H couplings support the assignmentcoupled n.0.e. spectrum (Figure 4) the resonance centredaround 6 111.0 shows two long-range couplings of 7 Hz,one of which is removed by exchanging the OH protonwith deuterium.This clearly distinguishes C (4) fromC(6), which is five bonds away from the hydroxy-proton.Quaternary carbon signals have been assigned fromlong-range couplings, on the assumption that three-bond C-H couplings in aromatic systems are larger thantwo-bond couplings, and by comparison with chemicalshifts in related compounds dihydrosterigmatocystin,O-acetyldihydrosterigmatocystin, and aflatoxin B, (l).The following evidence also supports the assignmentsdifferent from those of Set0 et aZ.m C(7), C(8), C(lO), andThe resonance at 6 180.9 is assigned to C(l) on theSelective decoupling of thec (W1.basis of the chemical shift.TABLE 313C Chemical shifts of stcrigniatocystin and derivatives, and directly bonded (lJIk,J and long-range ( 1J13C,H) couplingconstants ( J in Hz) of sterigmatocystinCarbon atom 6 C 1J'3C,H 'lJI",H 1)3c,13C b1 180.9 ( S ) 583 162.1 (Sddd) 10,5,29 Y 108.8 (S)165 7 70164 59167 8 5812,3 S4 111.0 (Dt)5 135.4 (D)6 105.7 (Dd)7 154.7 (Sdd)8 153.7 (S)9 106.4 (S)10 164.3 (Sdd)11 90.4 (D)12 163.0 (Sqd)13 106.7 rt 0.314 113.1 (Dddd15 47.9 (Dq)16 102.4 (Dddd17 145.1 (Ddt)OCH, 66.6 (Q)CH,COCH,CO8,4165 727,3187 9,6,3 34146 6 34181 13,5,2 76197 12,5 76146S ) *Dihydrosterig-matocystin6C181.0108.8162.1111.0135.3106.6154.7154.3105.3165.989.7163.2(1 05.6)113.444.231.467.756.6O-Acetyl-dihydros terig-matocystin60173.8115.9150.1118.6133.1115.0155.7 f153.6 t104.8165.089.7163.1107.7113.144.331.567.656.721.3168.7Chemical shifts relative to internal Me,Si.* This resonance overlaps with that of C(6).Capital letters refer to the pattern resulting from directly bonded protons and smallt May be inter-letters to long range W-H coupling.changed.S = singlet, D or d = doublet, T or t = triplet, Q or q = quartet.From ref. 22.The position was back-extrapolated from a Eu(fod), experiment.of aromatic, olefinic, and aliphatic carbon atoms. Theabsence of the resonances at 6 102.4 and 145.1 in dihydro-sterigmatocystin (Table 3) further supports the assign-ment of these resonances to C(16) and C(17), respectivelyin sterigmatocystin and confirms that the previousassignment of C(16) and C(14) should be reversed.The assignments of the proton-bearing carbon signalsagree largely with those of Set0 et except for thecrucial interchange of the C(4) and C(6) resonances.Neither data from the off-resonance proton-decoupledspectrum, because of the small difference in the protonchemical shifts, nor the acetylation shifts observed forO-acetyldihydrosterigmatocystin can be used to dis-tinguish between these two carbon atoms.In the2o E. Bullock, J. C. Roberts, and J. G. Underwood, J . Chew.SOC., 1962, 4179.21 P. S. Steyn and R. 'ileggaar, J.C.S. Perkin I, 1974, 3250.C(5) proton changes the resonances at 6 162.1 and 154.7to a doublet of doublets (J 5 and 2 Hz) and a broad un-resolved singlet, respectively. Deuterium exchange ofthe hydroxy-proton removes the 5 Hz coupling from theresonance at 6 162.1.This confirms the assignments ofC(3) (6 162.1) and C(7) (6 154.7). The complex pattern at6 163.0 is assigned to C(12) since decoupling of the meth-oxy-protons changes the signal to a simple doublet( J 2 Hz). This assignment is also supported by Eu(fod),-induced shift experiments. The methoxy carbon atomexperiences the largest shift indicating that complexat ionoccurs preferentially at the methoxy-group. The secondbiggest shift is shown by the C(12) resonance. Theassignment of C(8) (6 153.7) and C(10) (6 164.3) follows22 H. Seto, L. W. Carp, and M. Tanabe, Tetrahedron Letters,1974,44911188 J.C.S. Perkin Ifrom a comparison with the corresponding carbon signalsof aflatoxin B, (1) at 6 153.0 and 165.8, respectively.Assignments for the three quaternary carbon atomsC(2), C(9), and C(13), previously not assigned, weremade as follows.The resonance at 6 106.4 is shiftedleast by Eu(fod), and is therefore assigned to C(9).Acetylation of the C(3) hydroxy-group in dihydrosterig-matocystin caused the resonance at 6 108.8 C(2) toshift to lower field (6 115.9) with no significant effect onC(9) and C(13). This identifies the C(2) signal. Thecystin have been reported previously.22 The p.n.d.spectrum showed only one singlet, which we assign toC(7). Our 13C assignments (Table 3) and the reassignedone-bond 13C--13C couplings from reference 22 allow onlyone arrangement of intact acetate units in sterigmato-cystin, as shown in (5).The 13C assignments of sterig-matocystin by Set0 et aLZ2 lead to the incorrect arrange-ment of intact acetate units, viz. the location of only twointact acetate units in ring 3, and furthermore to adifferent folding of the original polyketide chain.1 1 1 l 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ,115 110 105 ?OObcFIGURE 4 Part of the natural-abundance coupled n.0.e. 25.2 MHz 13C spectrum of sterigmatocystin (2) : total spectral width5 000 Hz; 75" r.f. pulse; transients 17 K; decoupler on time 2 sremaining resonance at 6 105.7 in the spectrum ofsterigmatocystin is therefore assigned to C( 13).Great caution should be exercised when model com-pounds are used for the assignment of carbon signals incomplex organic compounds.The effects of methoxy-substitution on aromatic carbon chemical shifts 23 havebeen used for the assignment of carbons 2-7 in sterig-matocystin.22 The shift values as determined from ourvalues for sterigmatocystin and those reported for 6-methoxysterigmatocystin, with the expected values inbrackets, are (in p.p.m.): C(3) -9.0 ( - 6 4 , C(4) -2.0(+2.4), C(5) -14.8 (-15.0), C(6) +34.3 (+31.3),andC(7)-10.9 (-8.8). Although these values are in closeagreement, with the possible exception of those forC(4), a completely wrong assignment was made.=Similarly, even if the observed chemical shift of themethoxy-bearing carbon atom in aflatoxin B, C(8),S 161.61 is used to assign the corresponding carbon signalin sterigmatocystin C(12), an interchange of the C(3)(6 162.1) and C(12) (6 163.0) assignments might result.13C N.m.r.data on 1,2-13Cacetate-derived sterigmato-On the basis of their extensive 1-labelling studies onaflatoxin B, (l), Biollaz et aZ.3 postulated the currentlyaccepted biogenesis by which a single polyacetate chaingives rise to the C,, polyhydroxynapthacene endo-peroxide (6) which rearranges through a pyran inter-mediate to the bisdihydrofuran unit as in versicolorinA (7). The consecutive oxidative loss of two separateacetate-derived methyl carbon atoms can then lead firstto sterigmatocystin and secondly to aflatoxin B,.Heathcote et aZ.4 preferred an alternative route where aC4 unit is linked to a preformed anthraquinone molecule;however, feeding experiments showed that C4 units werevery poorly incorporated into ailatoxin B1.4The two possibilities for the expected arrangement ofacetate units in aflatoxin B, and sterigmatocystin, ifbiosynthesised according to Biollaz3 are shown in (8)and (9) and in (10) and (ll), respectively.From ourdata it was clear that the observed arrangements ofacetate units in aflatoxin B, (4) and sterigmatocystin23 R. H. Levin, J. Y. Lallemand, and J. D. Roberts, J . Org.Chem., 1973, 38, 19831976 1189(5) differed from those postulated by Biollaz et aL3An important finding is that ring 3 contains threeacetate units and must therefore be derived from theouter ring of an aromatic precursor. A C,, naphthaceneprecursor3 is thus no longer tenable.Our results onaflatoxin B, and sterigmatocystin are in agreement withthe biosynthesis from a single C,, polyketide precursorwhich is folded in only one mode leading to averufinand sterigmatocystin (Scheme). The proposed C,pathway (Scheme) is supported by recent experimentswhich established that averufin * and sterigmatocystincan be converted very efficiently by cultures of A .9arasiticus into aflatoxin B,. The precursor-productrelationship between acetate and averufin has recentlybeen demonstrated by using 13Cacetate.a6 The uniquehead-to-head linkage of the two acetate units in (1) and(2) for the coupling of the dihydrofuran ring and thearomatic system can be adequately accommodated by amechanism proposed by Th0mas.~.27obtained from solutions in CDCl, with a Varian XL-100-15FT spectrometer equipped with a 16 K Varian 620i computerand a gated gyrocode decoupler; 12 mm sample tubes wereused, except for 13C-enriched aflatoxin B, (5 mm tubes). ol OH0 0 0 0 0 0IIV0 0OHII ref. 27vaverufinIII7 1f----- -i1 ref. 2 4OH 0 OH OH 0 OH OH 0 OH I( 6 ) ( 7) II1I versicolorin AOH 0 OMe OH 0 OMeAflatoxin G, (12) and parasitic01 (13) 28 are mostprobably derived biogenetically in a similar fashion, asshown in the Scheme.EXPERIMENTAL100 niMz lH.N.m.r. spectra (solvent CDCl,) were recordedwith a Varian HA-100 spectrometer (Me@ as lock signaland internal reference). 25.2 MHz 13C N.m.r. spectra were24 M. T. Lin and D. P. H. Hsieh, J . Amer. Chem. Soc., 1973,95,1668.25 D. P. H. Hsieh, M. T. Lin, and R. C. Yao, Biochem. Bio$~hys.Res. Comm., 1973, 52, 992.26 D. L. Fitzell, D. P. H. Hsieh, R. C. Yao, and G. N. la Mar, J .Agric. Food Chem., 1976, 23,442.t1VIIVOH 0 OCH,(21(131 ( 121SCHEME Proposed biogenetic pathwaySpectral width, pulse angle, pulse delay, and number oftransients are specified on the Figures. Acquisition timeswere chosen to obtain the maximum number (8192) of datapoints. SPI experiments were performed as describedpreviously.12euro;ill932 Received, 3rd October, 1976127 R. Thomas, personal communication to M. 0. Moss, in' Phytochemical Ecology,' ed. J. B. Harborne, Academic Press,London, 1972, p. 140.a* R. D. Stubblefield, 0. L. Shotwell, G. M. Shannon, D.Weisleder, and W. K. Rohwedder, J . Agric. Food Chenz., 1970, 18,391
展开▼
