Microbiological hydroxylation of 17-norkauran-16-one andent-17-norkauran-16-one with the fungusRhizopus nigricans

摘要

1202 J.C.S. Perkin IMicrobiological Hydroxylation of 17-Norkauran-I 6-one and enf-I 7-Nor-kauran-I 6-one with the Fungus Rhizopus nigricansBy Robert McCrindle and James K. Turnbull, Department of Chemistry, University of Guelph, Guelph,Ontario, CanadaAllan B. Anderson, Department of Botany and Genetics, University of Guelph, Guelph, Ontario, CanadaIncubation of 17-norkauran-16-one (1 b) and ent-17-norkauran-16-one (2b) with Rhizopus nigricans gavemixtures of mono- and di-hydroxy-derivatives (8a-d) and (5a-d), respectively. A modification of an earlierrepresentation of the relevant enzymic sites i s used to explain the patterns of hydroxylation observed.IN attempts to prepare oxygenated derivatives ofkaurene (la), ent-kaurene (2a), and phyllocladene (3a),we have incubated the derived 17-nor-l6-ketones,(lb)-(3b), with various fungal species.Transform-ations were observed with several of these, althoughonly in the case of AsPergiZZus niger was there con-vincing t.1.c. evidence for the production of derivativesfrom all three ketones.to these products. Two other organisms appearedpromising: Calonectria decora, converting (2b) into amore polar compound; and Rhizopus nigricans, pro-ducing several polar derivatives from both (lb) and(2b). Unfortunately, the product from C. decora issimply the related endo-alcohol (4) which was formed in22 yield during 5 days of incubation. However, theStructures have been assignedA. B. Anderson, R. McCrindle, and J. K. Turnbull, Canad.R. A. Appleton, P.A. Gunn, and R. McCrindle, J. Chem.J. Chem., in the press.SOC. (C), 1970, 1148.products from R. nigyicans are of the type sought andwe now describe the assignment of their structures anddiscuss the mode of their formation.Incubation of ent-17-Norkaurnn-16-one (2b) .-Theketone (2b) was added in dimethylformamide (DMF)to a submerged culture of R. nigricans which had beengrowing for 4 days in a full nutrient medium3 underaerobic conditions. After 5 days the mycelium wasfiltered off and the filtrate extracted with ethyl acetate.T.1.c. comparison of this extract with that of a controlculture, to which only DMF had been added, showedthe presence in the former of appreciable quantities offour transformation products in addition to substrate(2b).Extraction of the ground mycelium yieldedsubstrate but no transformationProducts. The extractfrom the filtrate when chromatographed over silica gelgave (2b) and then, in later fractions, the four trans-J. C. Galbraith and J. E. Smith, Brit. Mycol. SOC. Trans.,1969, 52, 2371975 1203formation products.of increasing chromatographic polarity on silica gel.These will be discussed in order(1) a ; R = CH2b; R = O( 2 ) a; R = CH2b ; R = O flR H( 3 ) a;R=CH2b; R =OHO '-(5) a; R'=R2=R3=H (6) a ; R'=R2=Acb; R2=OH,R'= $:Hd; R'=OH,R2=R3=Hb; R'=H,R2= ACC; R3=OH,R'=R2=H C; R'=AC,R~=HAcO-' HO(7) (8) a; R' =R2 = R 3 =H6; R2=OH,R1=R3=Hc; R' = OH,R2= R3=Hd; R3=OH,R'=R2=H19) (10)The least polar product (5a) (2 based on substratenot recovered) was readily identified by direct com-t An Australian group are also studying the hydroxylation oftetracyclic diterpenoids by fungi.We thank Professor P. R.Jeff eries for discussions of unpublished results.parison with the known4 keto-alcohol which we hadobtained earlier from incubation of (2b) with A . niger.The major transformation product (5b) (15 yield)was eluted next together with a small proportion of thethird product (5c). The former was purified by pre-parative t.1.c. and formulated as a keto-diol on thebasis of its t.1.c. polarity and i.r. and mass spectra.Evidence for the location of the hydroxy-groups camefrom its lH n.m.r. spectrum. Two broad multiplets(each 1H) represent the X parts of two ABX systemsand can be ascribed to two axial carbinol protons eachof which has only two neighbouring protons in a vicinalrelationship.Thus the hydroxy-groups are equatorialand situated at C-1, C-3, or C-7. Further, the positionand multiplicity of the upfield carbinol proton resonanceand the chemical shift pattern of the three singletsarising from the tertiary methyl groups are almostidentical with those of (5a). This suggests the presenceof an equatorial hydroxy-group at C-3 and that theother substituent is positioned in such a way that itdoes not affect the chemical shifts of the methyl groups.Therefore, the possibility of a C-1 substituent can bediscarded and the second hydroxy-group assigned toC-7. Confirmatory evidence for this comes from thelocation (7 7.32) of the resonance attributed to H-15a,the 7a-hydroxy-group inducing a pronounced down-field shift from its normal value cJ z 7.85 for (2b)l.Indeed, when the IH n.m.r.spectrum of (5b) was run inpyridine this doublet was further shifted downfield by0.37 p.p.m. Incubation of (5a) with Rlzixopus nigricans tgives a compound to which structure (5b) has beenassigned.' Direct comparison of these samples con-firmed that they were identical.The next component (5c) was eluted from the columnin fractions which in addition contained either themajor product (5b) or the most polar product (5d).Preparative t.1.c. furnished (5c) (7 yield), which hasmany spectral features in common with (5b). Thustheir mass and i.r. spectra are very similar, as are thechemical shift patterns of the resonances attributableto the three tertiary methyl groups. In addition, the1H n.m.r.spectrum of (5c) contains a multiplet whichin shift position and shape is almost identical with thatascribed to the axial carbinol proton at C-3 in (5b). Aresonance at slightly lower field is attributable to anequatorial carbinol proton. This evidence was takento imply that (5c) is an isomer of (5b), an axial hydroxy-group in the former being present in lieu of the 7a-hydroxy-group in the latter. The spectral propertiesof the di- and mono-acetates (6a-c) derived from(5c) supported this conclusion and allowed assignmentP. R. Jefferies and R. W. Retallack, Austral. J . Chem., 1968,21, 1311; D.A. H. Taylor, J . Chem. SOC. (C), 1967, 1360.6 L. M. Jackmann and S. Sternhell, 'Applications of NuclearMagnetic Resonance Spectroscopy in Organic Chemistry,'Pergamon, Oxford, 1969, p. 237.Cf. B. P. Hatton, C. C. Howard, and R. A. W. Johnstone,J . C . S . Chem. Comm., 1973, 744.J. P. Beilby, E. L. Ghisalberti, P. R. Jefferies, M. A. Sefton,and P. N. Sheppard, Tetruhedvon Letters, 1973,2689.8 P. R. Jefferies, personal communication1204 J.C.S. Perkin Iof the axial hydroxy-group to C-7. The shifts inducedby Eu(dpm), on the methyl resonances in the lH n.m.r.spectra of the two monoacetates (6b) and (6c) wereparticularly valuable. For the less polar monoacetate(6b) the induced shifts were in a normalised ratioDwhich is entirely consistent l p D with the presence of anequatorial hydroxy-group at C-3.In the case of themore polar monoacetate (6c) the shifts induced in theresonances of the two C-4 methyl groups were similarin magnitude whereas those of the C-10 and acetylmethyl groups were larger and smaller, respectively.Only a hydroxy-group at C-7 could provide a reasonableexplanation for these observations. Since the alter-native 3-ax,7-eq-diol formulation for (512) is irreconcilablewith the spectral data, inter-relation of (5b) and (5c)by oxidation of both to the corresponding trione wouldestablish the structure of (5c) firmly. This has beenaccomplished for the enantiomeric diols (8b and c) (seebelow).That the most polar transformation product is alsoa keto-diol (5d) (7 yield) was apparent from itsspectral characteristics.However, the ill-defined natureof its carbinol proton resonances precluded their use indetermining the location of the hydroxy-groups. ThelH n.m.r. spectrum of the diacetate (7) was moreinformative. In this the analogous resonances wereattributable to the X parts of ABX systems, the observedmultiplicities indicating that both carbinol protons areaxial and each is coupled to two vicinal protons, oneaxial and one equatorial. Thus the hydroxy-functionsin (5d) are equatorial and at C-1, C-3, or C-7. Since thisfinal transformation product is not identical with (5b) itmust be either the l-eq,3-eq- or the l-eq,'l-eq-diol. That itis the former was shown by comparison with data for asample of (5d) prepared by Jefferies and his co-workers.Incubation of 17-Norkauran-16-one (lb) .-T.1.c.trans-formation products obtained similarly from (lb) sug-gested that appreciable quantities of four diterpenoidswere present in the mixture and that three of them,(Sa-c), were enantiomeric with products from (2b)while the fourth contained a new substitution pattern.All four were isolated by chromatography, the first(l), third (21), and fourth (8) (order of increasingchromatographic polarity) being identified as @a),(8b), and (8c), respectively by comparison of theirphysical and spectral characteristics with those of theirenantiomers (Sa--c). In addition, (8b) and (8c) werecorrelated by oxidation of both to the trione (9).The other component, second in polarity, was theketo-diol (8d) (15).Its i.r., mass, and lH n.m.r.spectra pointed strongly to its formulation as a keto-D. G. Buckley, G. H. Green, E. Ritchie, and W. C. Taylor,Chem. and Ind., 1971, 298.M. Hayano, M. Gut, R. I. Dorfman, 0. K. Sebek, and D. H.Peterson, J. Amev. Chcm. Soc., 1968,80, 2336; E. J. Corey, G. A.Gregoriou, and D. H. Peterson, ibid., p. 2338.(a) J. W. Browne, W. A. Denny, Sir E. R. H. Jones, G. D.Meakins, Y. Morisawa, A. Pendlebury, and J. Pragnell, J.C.S.Perkin I, 1973, 1493; (b) V. E. M. Chambers, W. A. Denny,J. M. Evans, Sir E. R. H. Jones, A. Kasal, G. D. Meakins, and J.Pragnell, ibid., p. 1500.l1 G. Ferguson and W. C. Marsh, unpublished result.diol, with one hydroxy-group secondary (equatorial)and the other tertiary.The tertiary nature of one ofthese groups was confirmed by acetylation, only amonoacetate (10) being formed even under forcingconditions. The lH n.m.r. spectrum of (10) was runin the presence of Eu(dpm), in the hope that it mightindicate the location of the tertiary hydroxy-group.However, the addition of the shift reagent, even insubstantial quantities, did not induce shifts largeenough to be of any diagnostic value, presumablybecause the hydroxy-group is too hindered to complexsignificantly with the reagent. Although no inform-ation had been acquired about the location of thetertiary substituent, the Eu(dpm),-induced shifts in thelH n.m.r. spectrum of the parent diol (8d) could now beexamined in the knowledge that only complexing withthe secondary alcohol would be important in deter-mining their magnitude.Indeed, the normalised ratioof the shifts induced in the methyl resonances agreesclosely with that expected l v 9 for an equatorial hydroxy-group at C-3. The lH n.m.r. spectra of (8d) and (10)allowed assignment of the other hydroxy-group to C-9a.In both, the resonance of H-15a appears far downfield(about 0.8 p.p.m.) from its position in related compoundswhich lack oxygenation at C-7 and C-9. The resonanceis easily assigned on the basis of its multiplicity. Innorkauran-16-ones this proton shows a large geminalcoupling, since it is adjacent to a carbonyl group, andin addition a long-range coupling to H-14P. Thus thetertiary hydroxy-group must be at C-9 and have thea-orientation expected on mechanistic grounds, ahydroxy-group normally having lo the same orientationas the hydrogen atom which suffers microbial displace-ment.Thus the diol can be formulated as (€id), anassignment which has been substantiated l1 by anX-ray structural analysis of (IO), by direct methods.Mechanism.-The outcome of the microbial oxidationreported here for (lb) and (2b) can be readily rationalisedon the basis of an extension of a mechanism proposed l 2 ~ 3by Jones, Meakins, and their co-workers for the hydroxyl-ation of steroidal substrates by R. nigricans. Theynoted that certain steroidal ketones are dihydroxylatedand considered12 that both hydroxy-groups may beintroduced in a concerted manner by a single enzyme-substrate complex.They suggested that the simplestexplanation would involve an enzyme which has threeactive sites, each with binding and hydroxylatingcapabilities, located in a triangular arrangement, corre-sponding to C-3, C-11, and C-16 of the steroid nucleus see(lla). This predicts the location of hydroxylation inmany, but not all, cases although it fails to explain theorientation (a or p).If one refines Jones' representation in a three dimen-sional sense, practically all the relevant results 12-16 in13 Sir E. R. H. Jones, Pure Appl. Chem., 1973, a, 39.14 A. M. Bell, V. E. M. Chambers, Sir E. R. H. Jones, G. D.Meakins, W. E. Muller, and J. Pragnell, J.C.S.Perkin I, 1974,312.16 2. Prochazka, Abh. Deut. Akad.Wiss. Berlin, Kl. Med., 1968,no. 2, 131.16 W. Charney and H. L. Herzog, ' Microbial Transformationsof Steroids,' Academic Press, New York. 19671975 1205the steroid field and those reported here are explained.The only refinement necessary is to define a referenceH( l l a ) (12)norml (n)mode(Ilbl Act; Bcc or cop1anar;CPn(13) reversed ( r ) mode (14) capsized normal(cnl mode(15) capsized reversed(cr 1 modeplane as that of the steroidal ring system of a substrateand to require that relative to this plane site A is below,site B is below (or coplanar), and site C is above ( l l b ) .The four most obvious binding modes 12*13 for a steroidare drawn (12)-(15). When one analyses the resultsfrom the steroid field it appears that the uncapsizedmodes, (12) and (13), are generally preferred for binding.In binding, site A favours oxygen atoms below theplane and hydroxylates from below (a); site B hassimilar binding requirements but can hydroxylate a(axial or equatorial) or p (equatorial only); site Cprefers to bind oxygen functions above the plane andhydroxylates p.Thus in the case of 5a-androstane-derived substrates 12aa 3-ketone interacts (16) with site A (n mode) and aP-hydroxy-group is inserted at C-16 (site C) and ana-hydroxy-group at C-11 (site B).On the other hand,a 2-ketone interacts (17) with site C (Y mode) anda-hydroxy-groups are inserted at C-16 (site A) and a tC-6 (site B). With appropriately disubstituted sub-strates,12* interactions with the binding sites will beoptimised.Thus the 17P-hydroxy-3-ketone (18) suffersn mode; p-01 at site C, C(3)=0 a t site A mainly C-11hydroxylation while its 3p-OH,C(17)=0 isomer (19) ishydroxylated Y mode; p-01 a t site C, C(17)=0 at site Aat C-7. In cases where the functionality is so disposedthat binding through two sites simultaneously is notpossible, then only one will be utilised, the resultingcompetition among different available modes of bindinggiving rise to mixtures 126 of products. Difficultieswere experienced in accounting for some of the minorproducts. Thus the hydroxylation of C-17 derivativesat the 5a- and 6cc-positions is not readily rationalisedby the model presented here unless one assumes thatan adsorptive mode such as (20) * is involved cf.the site A-site C distance in (21) and (22), hydroxyl-ation proceeding at site A.Further, it seems likelythat during insertion of the first hydroxy-group intoring B or ring c steroidal monoketones, modes of inter-action other than (12)-(15) are involved in whichthe ketone complexes with either site A or site C ratherthan the expected site B. Two final points should benoted. First, in mono-oxygenated substrates the twoHO-’ ,w(18)(20)0 (19)0(211(23)(241hydroxy-groups may be introduced by a sequential,rather than a simultaneous, process and second, themodel of the binding-hydroxylating sites outlined above* Such a mode would also explain the considerable amount oflla-hydroxylation,12a at the expense of the more usual C-3hydroxylation, in 6a-hydroxy-5a-androstan-17-one, the 6a-hydroxy-group being suitably oriented for interaction with siteA C(17)=0 at site C and hydroxylation at site B1206 J.C.S.Perkin Imay represent the sum of the locations and propertiesof the active sites on two or more enzyme units.Based on this general mechanism the followingenzyme-substrate complexes * would explain the pro-duction of (8a-d) from (lb) and (5a-d) from (2b).Interaction (21) (Y mode; ring c boat?) of the ketone of(lb) with site A would result in insertion of a p-OH atC-3 by the participation of the suitably disposedhydroxylation site C. This accounts for the formationof (8a). The two functional groups in (8a) then causebinding (21) (Y mode) at the same, or a new, activeregion and allow hydroxylation (site B) to proceed atthe 7p- (equatorial) or 7a-position as observed withsteroids. However, with this substrate, a molecularmodel indicates that more efficient interaction of theC(16)=0 with site A could be achieved by starting withthis Y mode, (21) (ring c chair), and rotating it ca.45"around the C-3,C-16 axis in such a way that H-9 rotatesdownwards and back toward hydroxylation site B.Thus formation of the three products (8b-d) isexplained.In formation of the substrate-enzyme complex with(2b) two possible niodes of interaction, Y mode (22)(ring c boat?) and cn mode (23) (ring c chair) presentthemselves, both of which predict formation of (5a) andby further attack (5d). Hydroxylation a t C-7 in theformation of (5b and c) requires absorption of thesubstrate in a tilted CY mode (24), an interaction whichappears acceptable on the basis of the active siterepresentation outlined above when one compares (24)with (19) using space-filling molecular models.EXPERIMENTALFor general experimental directions see ref.1. Molecularrotations were measured for solutions in methanol, unlessotherwise stated. Coupling constants reported are observedvalues.Incubation of ent- 17-iiorkauran- 16-one with Calonectriadecora.-Cultures (2 x 2-25 1) of C . decora were grown inMedium B l7 lacking beef extract (but containing allessential nutrients) for 5 days.(2b) (707 mg) in DMF (50 ml) was added and the ferment-ation was stirred vigorously for 7 days.The myceliumwas filtered off and the filtrate extracted with ethyl acetate(2 x 1.2 1 per 1.5 1 of filtrate) after the addition of sodiumchloride (500 g). Evaporation of the combined ethylacetate fractions gave an oil (1.4 g) which was chromato-graphed on silica gel (60 g). A control culture (2.5 l),containing DMF (25 ml) alone, was treated similarly.Elution with ether-light petroleum (1 : 4) gave first thestarting ketone (2b) and in later fractions ent- 17-norkauran-16a-01 (4) (27 mg, 22y0), m.p. 160-162' (from methanol-water) (lit.,2 162-163') ; vOH 3630 cm-l; T 9.17, 9.14, and8.95 (all s, quaternary CMe) and 5-70 (lH, m, W+ 30 Hz,H-16). Elution with more polar solvents, including ethylacetate-methanol mixtures, gave only compounds whichwere shown to be present in the control extract by analyticalt.1.c. and n.m.r.The mycelium was extracted with ethyl acetate and the* A referee has suggested that in some of these (see text) inter-action of the substrate with the enzyme sites could be optimised ifring c were to adopt a boat conformation.ent-17-Norkauran- 16-oneresulting extract chromatographed on silica gel (30 g) togive the starting ketone (2b).All fractions containing (2b)were combined (total 584 mg).Incubation of ent- 17-Norkauran-16-one with Rhizopusnigricans.-After 4 days of growth, cultures (4 x 2.5 1) ofRhizopus nigricans in a full nutrient medium3 were in-oculated with ent-17-norkauran- 16-one (2b) (2.0 g total) inDMF (50 ml). The fermentation was continued for 5 daysand then the mycelium was filtered off.After addition ofsodium chloride (1 kg) the filtrate was extracted with ethylacetate (2 x 1.5 1 per 2.5 1 of filtrate). Evaporation of theextracts gave an oil (8.3 g) which was chromatographedover silica gel (300 g). .4 control fermentation (2.5 1)containing DMF (7 ml) was treated in the same way.Elution of the column with ether-light petroleum (1 : 4 and1 : 1) gave fractions that contained the starting nor-ketone(2b) (600 mg), along with non-diterpenoid compounds.Later fractions contained the following transformationproducts.ent-3~-Hydroxy-l'i-norlzauran- 16-one (5a) .-Elution withpure ether gave a fraction (150 mg) containing a compoundnot present in the control extract.The major componentof this fraction, a non-terpenoid fungal metabolite, wasremoved by distillation under vacuum at 50" t o leave asolid (75 mg) which was subjected to careful preparativet.1.c. (pure chloroform, then ethyl acetate-light petroleum,1 : 3). The resulting keto-alcohol (5a) (17 mg, 1.7y0), oncrystallisation from methanol-water and then lightpetroleum had m.p. 179-183' (lit.,l 179-181') ; mixedm.p. 180-183".ent-3p, 7p-Dihydroxy- 17-norkauran- 16-one (5b) .-Elutionof the column with pure ether gave, in later fractions, awhite solid (175 mg) which showed one major and oneminor component on t.l.c., both of which were not presentin the control extract. Preparative t.1.c. (methanol-ethylacetate, 1 : 19) separated these compounds to give theketo-diol (5b) and a small amount of the keto-diol (5c) (seebelow).The keto-diol (5b) (175 mg, 15) on crystallis-ation from methanol-water had m.p. 219-221' (reported,s223-224'); a, -32'; vOH 3630, vco 1740 cm-l; T (CDCl,)9-20, 8.99, and 8.89 (all s, quaternary CMe), 7-32 lH, dd,H-l5a(exo), JSm 18, J15a,13 2 Hz, and 6-22-6-90 (2H,complex m, H-3 and -7); T (pyridine) 9.00, 8.92, and8.83 (all s, quaternary CMe), 6-95 (lH, dd, H-l5a), and6-3-6-8 (2H, complex m, H-3 and -7) ; m/e 306 (goyo, Mf),291 lo, (M - 15)+, 288 25, (M - 18)+, 273 (30), and121 (100); mixed m.p.8 221-223".ent-3@,7a-Dihydroxy- 17-norkauran- 16-one (5c) .-Theinitial fractions eluted with ether-ethyl acetate (1 : 1)contained (52). This was separated from the majorcontaminant (5b) by preparative t.1.c.(methanol-ethylacetate, 1 : 19). On crystallisation from methanol-waterand then ether, the keto-diol (5c) (75 mg, 7) had m.p.208-210", a, - 8 O ; vOH 3620, vco 1747 cm-1; T (CDCl,)9.20, 9.01, and 8.90 (all s, quaternary CMe), 7.85 (d, part ofH-15a signal, J15a,13 1-5 Hz), 6-70 (lH, m, H-3, Wt 17 Hz),6-30 (lH, m, H-7, W, 8 Hz); T (pyridine) 8.94, 8.92, and8-74 (all s, quaternary CMe), 7-55 (d, downfield part ofH-16a signal), 6.48 (lH, m, H-3), and 6-10 (lH, m, H-7);m/e 306 (lo, M+), 291 5, ( M - 15)+, 288 15, ( M - IS)+,274 (13), 255 ( l l ) , and 121 (100) (Found: C, '74.2; H, 9.8.CI,H~,,O~ requires C, 74.45; H, 9.85).ent-l~,3~-Dihydroxy-17-norkauran-16-one (5d) .-Later17 J.W. Blunt, I. M. Clark, J. M. Evans, Sir E. R. H. Jones,G. D. Meakins, and J. T. Pinhey, J . Chew. SOC. ( C ) , 1971, 11361975 1207fractions eluted with ether-ethyl acetate (1 : 1) containedthe keto-diol (5d) along with some (5c). These wereseparated by preparative t.1.c. (methanol-ethyl acetate,1 : 19). On crystallisation from methanol-water the keto-diol (5d) (70 mg, 7) had n1.p. 246-251' (reported,8246246"); a, -37'; VOH 3630 cm-l; T (CDCl,) 9-22,9.05, and 8.84 (all s, quaternary CMe), 64-6.9 (2H,complex m, H-1 and -3); 7 (pyridine) 8-99, 8-85, and 8.67(all s, quaternary CMe), 6-4-6.8 (ZH, complex ni, H-1and -3); nt/e 306 (17, M+), 288 r31, (M - lS)-', 273(22), 270 (39), 255 (44), 234 (70), 233 (loo), 232 (90): 191(70), 177 (63), 174 (60), and 163 (47).Extraction of the mycelium with ethyl acetate gave anoil (1.2 g) which was combined with the impure (2b)previously isolated.Chromatography over silica gel andelution with ether-light petroleum (1 : 8) gave the startingketone (2b) (965 mg).A cetylation of the Keto-diol (5c) .-The keto-diol (5c)(65 mg) was treated with acetic anhydride ( 1 ml) andpyridine (4 ml) for 1 h at 0 "C. Methanol (2 ml) was thenadded and the solvent evaporated off under vacuum togive an oil, which was subjected to preparative t.1.c.(chloroform). The top band, consisting of diacetate, andthe middle band, two monoacetates, were retained, and thelarge quantity of starting material in the bottom band waseluted and acetylated as before.After four of these cyclesand a fifth cycle of 2 11 reaction time, the products wereisolated by preparative t.1.c.The least polar band contained ent-3pJ7a-dincetoxy- 17-norkauran-l6-one (6a) ( 18 mg), which on crystallisationfrom ether-light petroleum and then ether had m.p.176-177"; a, -6'; vco 1740 cm-l; T 9.23, 9-15, and8.88 (all s, quaternary CMe), 7.97 (6H, s, 2 x OAc), 5.47(lH, m, H-3, W* 17 Hz), and 5-10 (IH, m, H-7, W+ 8 Hz){Found : I+, 390.2410. C2.H3@5 requires M , 390.2406).The band of intermediate polarity gave a solid (35 mg)consisting of two compounds which were separated bypreparative t.1.c. (ethanol-chloroform 1 : 100, run threetimes). The less polar was ent-Sa-acetoxy-3P-hydroxy- 17-norkauran- 1 6-one (6b) ( 13 mg), which on crystallisationfrom ether had m.p.216-217"; V ~ H 3620, vco 1738 cm-l;T 9.18 (s, 19-H,), 9-07 (s, 18-H,), 8-87 (s, 2O-H,), 7.93 (s,OAc), 6.71 (IH, m, H-3, Wi 17 Hz), and 5.07 (lH, m,H-7, Wt 8 Hz) (Found: M+, 348-2303. C,1H3204 requiresM , 348.2301). In the presence of Eu(dpm), (28 mg),(6b) (10 mg) has T 6-50 (s, OAc), 4.25 (s, 2O-H,), -0-70 (s,18-H,), and -1.25 (s, 19-H,). The induced shifts are inthe ratio, 19-H, : IS-H, : 20-H, : OAc, of 10 : 9.4 : 4.4 : 1.4.The more polar ent-3~-acetoxy-7a-hydroxy-17-norkaztran-16-one (6c) (15 mg), on crystallisation from ether, had m.p.303-205"; v o ~ 3620, vco 1740 cm-l; T 9.12 (s, 19- andI8-H3), 8-87 (s, 2O-H,), 7.95 (s, OAc), 6.27 (lH, m, H-7,Wg 8 Hz), and 5.45 (lH, m, H-3, Wt 17 Hz) (Found: M+,348-2303.C,1H,,O4 requires M , 348.2301). In theprcsence of Eu(dpm), (27 mg), (6c) (15 mg) has T 7.20 and7.07 (both s, quaternary 4-Me2), 6-30 (OAc), and 5.88 (s,2O-H,). The induced shifts are in the ratio 20-H, : 4-Me : 4-Me : OAc, of 10 : 6.9 : 6.5 : 5.6.The most polar band of the original t.1.c. plates consistedof the starting keto-diol (5c) (5 mg).A cetylation of the Keto-diol (5d) .-The keto-diol (5d)(60 mg) was treated with acetic anhydride-pyridine (1 : 1 ;5 ml) overnight. Removal of solvent and then preparativet.1.c. (ethyl acetate-light petroleum, 1 : 2) gave ent-lp,3P-dincetoxy- 17-novkauran- 16-one (7) (70 mg), m.p. 202-204"(from ether); vco 1740s cn1-l; 7 9.12 (s, 19- and 18-H3),8.67 (s, 20-H,), 8.00 and 7-97 (both s, 2 x OAc), and 5-32and 5.21 (both m, H-1 and -3, LVt 16 Hz) (Found: M',390-2410.C2,H,,05 requires 144, 390.2406).Incubation of 17-Norkauran- 16-one ( 1 b) with Rhizopusnigricans.-A culture (2 x 2-5 1) of R. nigrzcans in theusual medium was grown for 3 days and then inoculatedwith 17-norkauran-16-one (lb) (843 mg) in DMF (20 ml).The fermentation was continued with vigorous aerationfor 5 days. The mycelium was then filtered off and afterthe addition of sodium chloride (500 g) the filtrate wasextracted with ethyl acetate (2 x 1.5 1 per 2-5 1). Theextracts were combined and the solvent removed to give anoil. A control culture (2.5 1) containing DMF (10 ml) onlywas treated similarly. The extracts were compared byt.1.c. which showed the presence of a t least four additionalcompounds in the fermentation extract.The fermentation extract (2.7 g) was chromatographedover silica gel (70 g).Elution with ether-light petroleum(1 : 4) gave in early fractions the starting ketone (Ib)contaminated with fungal metabolites. Later fractions ofthis eluent gave an oil (100 mg) which contained a furthercompound not present in the control.3~-Hydroxy-17-norkauran-16-one (8a) .-Preparative t.1.c.(chloroform, run twice), separated the keto-alcohol (8a)(10 mg, 1.3) from this oil. On crystallisation frommethanol-water the keto-alcohol had m.p. 181- 183" ;4, +33O {cf. enantiomer (5a), m.p. 180-183'; a, -45");vOH 3620, vco 1740 cm-l; n.m.r. and mass spectra as (5a)(Found : I+, 290-2247. Cl,H3,O2 requires N , 290-2246).Elution with pure ether gave a solid consisting of fourcomponents, three of which were not present in the controlextract. These were separated by repeated preparativet.1.c.(light petroleum-ethyl acetate, 1 : 3) to give fourbands. The top band gave a fungal metabolite (30 mg)which was also present in the control extract.3p, 9a-Dihydroxy-17-norkauvan- 16-one (8d) .-Band twooverlapped partially with band three b u t the top portiongave pure keto-diol (8d). The bottom portion was re-chromatographed. The keto-did (8d) (130 mg, 15) fromall fractions was combined and on crystallisation fromfirst ether-methanol and then methanol-water had m.p.224-226'; a $17"; VOH 3618, vco 1742 cm-1; T (CDC1,)9.17 (s, 19-H,), 8.98 (s, l8-H,), 8.87 (s, 20-H,), 7.24 (lH, dd,H-15a, Jsem 19, J15a,14@ 2.5 Hz), and 6.80 (IH, ni, H-3,15'4 16 Hz); T (pyridine) 8.91 and 8.75 (6H) (both s,quaternary CMe), 7.52 (lH, dd, H-14P, Jge,, 13, J14~.~5a2.5 Hz), and 6.80 (lH, dd, H-15a); m/e 306 (27, M+),288 loo, (M - IS)+, 270 30, (N - 36)+, 255 (25), 245(60), 227 (67), 165 (33), and 123 (100) (Found: C, 74.3;H, 10.0.C13@3 requires C, 74.45; H, 9.85). In thepresence of Eu(dpm), (28 mg), (8d) (30 mg) had z 7.40 (s,20-H,), 6-00 (s, 18-H,), and 5-89 (s, 19-H,). The inducedshifts are in the ratio 10 : 9.1 : 4.2 for 19-H, : IS-H, : 20-H,.3/3,7~-Dilzyd~oxy-l7-norkauran-l6-one (8b).-The bottomportion of band three gave pure keto-diol (Sb). The topportion, which contained (8d) as well, was rechromato-graphed.The keto-diol (8b) (180 mg, 21y0), on crystallis-ation from methanol-water and then methanol, had m .p.225-226"; a,, +34.5' {cf. enantiomer (5b), m.p. 219-221, 223-224", oilD -32'); v o ~ 3630, vco 1740 cm-l;T (pyridine) 9.00, 8-92, and 8-83 (all s, quaternary CMe),6-95 (lH, dd, H-15@, Jse, 18, J158.13 1.5 Hz), and 6-3-48(2H, complex m, H-3 and -7); nz/e 306 (SO, Mt), 291lo, (ill - 15)+, 288 25, (-42 - 18)+j, 273 (35), and 121208 J.C.S. Perkin I(100) (Found: C, 74.25; H, 9-95. C,,HsoOs requires C,74.45; H, 9.85).3p, 7a-Dihydroxy- 17-norkauran- 18-one (8c) .-Band fourcontained a small amount of impure material that showedthe same chromatographic behaviour as the keto-diol (5c).Elution of the chromatography column with ether-ethylacetate (1 : 1) gave more of this material.This fractionwas combined with band four and subjected to preparativet.1.c. (light petroleum+thyl acetate, 1 : 3) to give an oil(100 mg) which was homogeneous by t.1.c. Vacuumsublimation of this separated a fungal metabolite from lessvolatile diterpenoids (80 mg). Preparative t.1.c. (methanol-chloroform, 1 : 19) of the latter gave a glass which couldnot be induced to crystallise. This material, consistingmainly (ca. 85) of the keto-diol (€412) together with otherditerpenoids (ca. 16), had VOH 3620, vco 1747 cm-1;T (CDCl,) 9-20, 9-01, and 8.90 (all s, quaternary CMe), 7.85(d, part of H-15P signal, J158.13 1.5 Hz), 6-70 (lH, m, H-3,W+ 17 Hz), and 6.30 (lH, m, H-7, W+ 8 Hz) attributable to(8c), and T 8.75 and 8.70 due to impurities cf.n.m.r. of(SC) . This impure diol (23 mg)was treated with acetic anhydride-pyridine (1 : 1.5 ml)overnight. Removal of the solvent followed by preparativet.1.c. (ethyl acetate-light petroleum, 1 : 2, run twice) gavethe 3p,'Ia-diacetate (25 mg), which on crystallisation fromether-light petroleum (twice) had m.p. 176-176'; a,+7' (CHCl,) {cf. enantiomer (6c), m.p. 175-177'; a,-6'); i.r., n.m.r., and mass spectra identical with thoseof (6c) (Found: M+, 390.2410. C,,H,,05 requires M,390.2406).The mycelium was dried and extracted with ether togive an oil (1.3 g) which was combined with the impurestarting material obtained from the filtrate (above).Thismaterial was then chromatographed over silica gel (50 g).Elution with ether-light petroleum (1 : 4), followed bypreparative t.1.c. (ethyl acetate-light petroleum, 1 : 12),gave the starting norketone (lb) (72 mg) .A cetyhtion of the Keto-diol (8d) .-The keto-diol (8d)The yield of (8c) was 8.(46 mg) was treated with acetic anhydride-pyridine ( 5 ml;1 : 1) for 3 h. Evaporation of solvent followed by pre-parative t.1.c. (methanol-chloroform, 1 : 30) gave 3p-a.cetoxy-9~-hydroxy-l7-norkauvan-16-one (10) (41 mg), m.p.291-295' (from methanol) ; a, + 17.5O (CHCI,); VOH3620, vco 1737 cm-l; T 9.11 (s, 19- and l8-H,), 8.75 (s,20-HJ, 7.80 (s, OAC), 7.48 (lH, dd, H-l4P, Jsem 13, J1,fi,lso!2.5 Hz), 7.20 (IH, dd, H-15a, Jm 19, J 1 5 a r , l ~ 2.5 Hz),330 3, (M - IS)', 289 (62), 288 loo, (M - 60)+, 270and 5.50 (lH, m, H-3, Wt 17 Hz); m/e 348 (a, M+),(32), 255 (48), 245 (70), 227 (go), and 165 (53) (Found:C, 72.1; H, 9.4.Oxidation of the Keto-diol (8b) .-The keto-diol (8b)(62 mg) in acetone (15 ml) was treated with Jones reagent(0.2 ml) for 15 min. Addition of water followed by ex-traction with ether gave the crude trione. On crystallis-ation from ether 17-norkaurane-3,7,16-trione (9) (46 mg).had m.p. 140-142'; a, +So (CHCl,); vco 1745 and1710 cm-l; 7 8.89 (6H) and 8-63 (both s, quaternary CMe)and 6.88 (lH, dd, H-l5P, Jgm 18, J I 5 f i . 1 3 1.5 Hz); m/e 302(54, M+), 260 (75), 217 (45), 164 (70), and 136 (100)(Found: C, 75.2; H, 8.8. C,,H,BO, requires C, 75.45; H,8.65).Oxidation of the Keto-diol (8c) .-Impure keto-diol (8c)(31 mg) was oxidised with Jones reagent as for (8b). Thecrude product was subjected to preparative t.1.c. (ethylacetate-light petroleum, 2 : 3) to give the trione (9) (19mg), m.p. 140-142' (from ether), identical (n.m.r. andi.r. and mass spectra) with a sample of (9) prepared from(8b) ; mixed m.p. 140-142'.C,,H,,O, requires C, 72.4; H, 9.25).We thank the National Research Council of Canada foran Operating Grant (to R. McC.) and a Scholarship (toJ. K. T.). We are also indebted to Dr. T. Anthonsen,Trondheim, Norway, for the high resolution mass spectraand to Dr. G. L. Barron, University of Guelph, for gifts-offungal cultures.4/2667 Received, 9th December, 1974