Pyrrolizidine alkaloids. The biosynthesis of senecic acid

摘要

1972 671Pyrrolizidine Alkaloids. The Biosynthesis of Senecic AcidBy D. H. G. Crout," N. M. Davies, E. H. Smith, and D. Whitehouse, Department of Chemistry, University ofExeter, Exeter EX4 4QDFeeding experiments have shown that two five-carbon units derived from isoleucine are specifically incorporatedinto senecic acid (IV) in Senecio rnagnificus. The specific incorporation of the methyl carbon atom of methionineinto C-3 of serine in pea seedlings has been demonstrated. The significance of this result in relation to necic acidbio syn th esis i s discussed .THE pyrrolizidine alkaloids characteristic of Seneciospecies (fam. Compositae) have structures which arevariants of a basic type, exemplified by seneciphylline(I) and senecionine (11), in which a ten-carbon substi-( I I R ' R ~ = C H ~(XI) R'= Me,R2= HR' amp;OHCH,-C- $--Metuted adipic acid is esterified with a tetrahydro-l-hydroxy-7-hydroxymethylpyrrolizine base.Althoughthe corresponding seneciphyllic (111) and senecic (IV)acids have structures which can be formally derivedfrom isoprene units, feeding experiments with labelledacetate and mevalonate have shown that these acids arenot formed by the acetate-mevalonate pathway ofterpene biosynthesis.lS2 On the other hand, iso-leucine (V) and its biological precursor threonine (VI)are efficient and specific precursors of seneciphyllic acid(111) in Seyzecio douglasii DC.I With uniformly labelled14C-~-isoleucine as precursor, the five-carbon compo-D. H. G. Crout, M. H. Benn, H.Imaseki, andT. A. Geissman,Phvtocizemisti.v. 1966. 5. 1.nent comprising (C-4, -5, -6, -7, and -10) of seneciphyllicacid (111) was most heavily labelled, and it was suggestedthat isoleucine was incorporated according to Schemel(a). This conclusion was based on a consideration ofthe biosynthetic pathway leading to isoleucine, the salientfeatures of which are summarised in Scheme 1. Accord-ing to this pathway, uniformly labelled 14C threonine(VI) would be expected to label C-1, C-2, C-4, and C-5 ofisoleucine equally. If, therefore, isoleucine were to beincorporated into seneciphyllic acid (111) in the mannershown in Scheme 1 (b), u-14C-~-threonine (VI) would beexpected to produce equal labelling at C-6, C-7, and C-10of the necic acid.However, it was observed that theC-6,7 unit of seneciphyllic acid (111) derived from~-~~C-~-threonine contained more than four times asmuch activity as C-10. Accordingly, the route (a)in Scheme 1 was preferred to route ( b ) . It was assumedthat activity from zc-l*C-~-threonine appeared in C-10of seneciphyllic acid (111) by an indirect pathway.There is now evidence to suggest that this analysisrequires modification, since enzyme systems are knownwhich may operate on uniformly labelled threonine (VI)in vivo to introduce an inequality in the labelling of itsC-1,2 and C-3,4 components.The essential chemical transformations mediated by2 C. G. Gordon-Gray and F. D. Schlosser, J . South AfricanChenz. Inst.. 1970. 23. 13672 J.C.S. Perkin Ithe enzyme systems in question are summarised in of threonine by equilibration, via either of the twoScheme 2(a and b) .The operation of threonine dehydro- enzyme systems, with a pool of glycine which was con-genase and a-amino-oxobutyrate-CoA ligase (amino- siderably larger than the pool of acetyl CoA at the siteacetone synthetase) 3 9 4 results in the conversion of threo- of necic acid biosynthesis.nine into an equimolar mixture of acetyl CoA and glycine Threonine-cleaving enzyme systems have been studiedScheme 2(a). A similar result is obtained by the almost exclusively in micro-organisms. The occurrencesequential action of threonine aldolase, acetaldehyde of threonine dehydrogenase in higher plants has not beenMeC0,HC i t r i c - acidcycleCO HI 2y 25OC02HHMe*7bsol;/' Eisre'H;H* ./ =C IC02H *(mi R ' R ~ = CH,(IYI R'= Me,R2= HR' R 2 0 HMe - c - c ---C02 H1: IISCHEME 1dehydrogenase, and acetyl CoA synthetase Scheme recorded.However, it has been reported that threonine2(b). If it is assumed that in vivo either or both of aldolase is widely distributed in higher plants andthese systems are reversible to some degree, then the recently Senecio magni$cus F. Muell., the species useddistribution of radioactivity in threonine (VI), initially in the present studies, has been found to contain low butMe Me Mei I 1CH*OH i CO i i CO*SCoAI = = I - +CH-NHt CH.NH2 CH2*NH2I I IdCQZH C02H C 02HCH-NH, CNH,I IC02 H CHSCHEME 2significant levels of threonine aldolase activity.' Thisobservation increased the possibility that isoleucinemight be incorporated into necic acids of the seneciphyl-lic (111) type by route (b) (Scheme 1) rather than byroute (a).Incorporation experiments with specificallylabelled isoleucine have been carried out in order toresolve this question.These experiments were designed at the same time toshed light on the origin of the five-carbon (C-1, -2, -3, -8,and -9) unit of the necic acids (cf. Scheme 1). Althoughisoleucine was incorporated into this portion of seneci-phyllic acid (111) at only half the rate at which it wasincorporated into the other component, it was observedthat with ~-~~C-~-isoleucine as precursor, C-1, C-8, andC-9 were nearly equally 1abelled.l It appeared possibletherefore that the C-1, -2, -3, -8, and -9 unit of seneciphyl-MeIco,sco AReagents : i, threonine dehydrogenase ; ii, a-amino-oxobuty-rate-CoA ligase (' aminoacetone synthetase ') ; iii, threoninealdolase ; iv, acetaldehyde dehydrogenase ; v, acetyl CoAsynthetase.uniformly labelled, will depend on the extent to which thesystem moves towards equilibrium and on the relativesizes of the existing pools of acetyl CoA and glycine.Thus the observed distribution of activity in seneciphyl-lic acid (111) derived from u-14C-~-threonine might haveresulted from dilution of the label in the C-1,2 portionW. G.Laver, A. Neuberger, and J. J. Scott, J . Claem. SOG.,1959, 1483.4 (a) ?V. H. Elliot, Biochim.BioFhys. Acta, 1958, 29, 446;(b) A. Neuberger and G. H. Tait, Biochem. J., 1962, 84, 317;(c) G. Urata and S. Granick, J . Biol. Cheun., 1963, 238, 811.lic acid (111) might also be derived from isoleucink (V)as indicated in Scheme 3.There was an objection to this proposal in that C-2 ofseneciphyllic acid (111) appeared to be labelled when~-~~C-~-threonine was administered to Senecio dougZasii,lwhereas this position should not be labelled if threonine(VI) is incorporated via isoleucine (V) (Scheme 3).However, the activity of C-2 was not measured directlybut was obtained by difference from a number of inde-pendent activity determinations, a procedure recognised5 J. G. Morris, Biochem. J . , 1969, 115, 603.6 Tsung-ya Lo and Yu-Wei Tang, Sheng Wu Hua Hsueh YuSheizg Wu Wu Li Hsueh Pao, 1964, 4, 527 (Chem.Abs., 1965, 62,15,004).7 Dr. J. M. Turner, personal communication1972 tionas leading to an unreliable result .l This uncertainty hasbeen resolved by administering u-14C-~-threonine toSenecio magni$cus, which produces senecionine (11) as. u,*Me "Me/R'* R' OH 1; 1.* 2 1H bsol; ,c=c /cH2-c- CO,H c----IMe COZHSCHEME 3the major alkaloid.8 The labelledhydrolysed to senecic acid (IV) andScheme 4(a). As in the experimentsH(Y)senecionine wasretronecine (VII)with S. douglasii,lHobsol;-f"*.DH HMe OHI 1,CH2- C ~ - 2 : -MeOHSCHEME 4Itcagcnts: i, Ba(OH),; ii, LiAlH,; iii, HIO,; iv, PhMgBr;v, IhlnO,.~-~~C-~-threonine was found to be efficiently and selec-tively incorporated into the necic acid component of thealkaloid (see Table 1).The C-2 atom of the necic acidwas isolated as the carboxy-group of benzoic acid seeScheme 4(b), and was essentially inactive (see Table 2,experiment 1 ) .The labelled necic acid was further degraded bymethods illustrated in Scheme 5. As in the correspond-ing experiment with seneciphyllic acid,l C-8 was essenti-ally unlabelled, whereas C-1 was strongly labelled. Onthe other hand, the ratio of the activities of the C-6,7unit and C-10 was relatively close to the expected value673of 2 as compared with the value of 4.3 observed in thecase of seneciphyllic acid (see Table 2).With the removal of this objection to the propositionthat the C-1, -2, -3, -8, and -9 unit of the necic acids isderived from isoleucine, the incorporation of u-14C 1-L-isoleucine into senecionine (11) was examined. As ex-pected, this precursor was incorporated efficiently andselectively into the necic acid component (Table 1,experiments 2 and 3).Degradation of the senecionine(Scheme 5 ) showed that C-1 and C-8 were nearly equallylabelled, in confirmation of the result obtained withTABLE 1Incorporation of 14C-labelled precursors into senecionineand distribution of the label between senecic acid andretronecineO / 9 ; Actil-ity in/O Incorpor- senecic retronecine1 ~t-~~C-~-Threonine 0.25 98.5 amp; 3.9 0.6 0.062 ~-~~C-~-Isoleucine 0-44 99.0 amp; 3.6 0.98 amp; 0.053 ~-~~C-~-Isoleucine 0.48 96.4 f 3-9 0.80 i.0.085 2-14CI~~leucine 0.44 100 amp; 3.8 1.6 amp; 0.36 5-14CIsoleucine 0.42 96.6 amp; 3.5 2.0 amp; 0.17 6-14CIsoleucine 0.15 99.4 f 4.0 3.9 5 0.88 Me-14C-~- 0.00015Expt. Precursor ation * acid (IV) (YII)4 2-14CIsoleucine 0.27 95-2 f 3.S 0.0Methionine* Where mixtures of DL-isoleucine aiicl DL-alloisoleucine werefed, incorporations were calculated on the assumption, nowknown to be valid (D. H. G. Crout and N. M. Davies, unpub-lished observations), that only L-isoleucine is an effective pre-cursor of senecionine.seneciphyllic acid in S. douglasii. (The lack of function-ality at C-9 in senecic acid (IV) precluded direct determin-ation of the activity at this position.However, whereas in the experiments with S. douglasiithe labelling pattern indicated that the C-4, -5, -6, -7, and-10 unit of seneciphyllic acid (111) contained two-thirdsof the total activity and the remaining C, unit one-third,in the experiments with S.magnijcus the labellingpattern indicated that the activity was evenly distri-buted throughout the necic acid (Table 2, experiments2 and 3).We next carried out a series of incorporation experi-ments with specifically labelled isoleucine in order to de-termine the manner in which isoleucine is incorporatedinto the C-4, -5, -6, -7, and -10 component of senecic acid(IV) as discussed above, and to test the proposition thatisoleucine is a specific precursor of the C-1, -2, -3, -8, and-9 component. 2-14C-, 5-14C-, and 6-14C-Isoleucinewere synthesised, as described later, and administeredto S.magni$cus plants hydroponically. The labelledalkaloid was degraded by the methods shown in Scheme5. (In order to optimise yields, more than one route tocertain degradation products was developed. The choiceof method for a particular degradation depended on thelocation of the label in the precursor.)Each precursor was efficiently and selectively incorpo-rated into the necic acid component of the alkaloid8 C. C . J. Culvenor, Austral. J. Chem., 1962, 15, 158674 J.C.S. Perkin I(Table 1, experiments 4-7). Moreover, 2-14Ciso-leucine labelled C-1 and C-10 of senecic acid (IV) equallyand exclusively and the label from 5-14Cisoleucine waslocated at C-9 and C-7, also equally and exclusively(Table 2, experiments 6 6 ) .6-14CIsoleucine gavesenecic acid in which 58 of the label was at C-8; theas the activity of C-7. The activity of C-9 was obtainedby subtracting the activity of the C-6,7 unit from the(statistically corrected) activity of the barium acetate(C-2, -3, -6, -7, -8, and -9) obtained by Kuhn-Rothoxidation of senecionine (11) (Scheme 5). The ratios ofthe activities of C-9 and C-1, and of C-7 and C-10 wereTABLE 2Distribution of activity in senecic acid (IV) derived from l4C-labe1led precursorsC-2,3,6,Expt. Precursor c- 1 c-2 C-6,7 c- 7 C- Pi c-9 c-10 7,8,9 C-2,3,6 C-7,8,91 u-14C-~-Threo- 15.4 f 0-6 0.59 f 18.8 f 0.7 0.10 amp; 0.01 12.1 f 0-52 U-l4C-L-ISO- 9.8 0.4 8-5 f 0.43 U-14C-L-ISO- 11.6 amp; 0.5 18.3 f 0.6 10.0 amp; 0.44 2-14CIsoleucine 54.5 f 2.4 43.6 f 2.25 2-14CIsoleucine 50.1 f 2.0 2.95 f 0.90 0.35 f 0.24 49.0 f 2.06 5-14CIsoleucine 49.7 f 1.8 49.4 amp; 0.51 amp; 0.17 48.1 f 98.0 f 0.94 f 98.0 fnine 16.3 amp; 0.6 0.10leucineleucine1.8 * 3.9 t 3.5 0.09 3.57 6-14CIsoleucine 57.5 3.3* Calculated on the assumption that the activity of the BaCO, (C-2,3,6) is equally distributed between C-2, C-3, and C-6.t Bydifference (C-7,8,9 - (C-7) + (C-8)).(Mec o 2x ii 1BCiCO,( c-I i vL I 2 8 1f HCHO L C X )MeSCHEME 5Reagents: i, Ba(OH),; ii, NaIO,; iii, NaIO; iv, LiAlH,; v, 0,; vi, dimedone; vii, HIO,; viii, OsO,-NaIO,; ix, HOAc;x, Kuhn-Roth oxidation; xi, HN,; xii, Ba(OH),; xiii, 2,4,6-trinitrotoluene.remainder was presumably at C-4.(The activity at thisposition could not be determined directly.)The direct incorporation of C-5 units derived fromisoleucine without rearrangement or degradation wasdemonstrated by feeding isoleucine doubly labelled withI4C at the 2- and 5-positions. The labelled senecionine(11) was degraded by methods given in Scheme 5. Sincethe activity of senecic acid (IV) derived from W4C-isoleucine was almost entirely at C-7 and C-9 (Table 2,experiment 6), the activity of the C-6,7 unit was takenthe same, within experimental error, as the correspond-ing ratio in the administered doubly-labelled isoleucine(Table 3).These experiments demonstrate conclusively that twoC-5 units derived from isoleucine are incorporated intosenecic acid (IV) in the manner indicated in Scheme 6.This pathway involves, a t some stage, the loss of thecarboxy-carbon atom from both participating isoleucinemolecules, a result that had been suggested by earlierfeeding experiments in S. doztglask1972 675The conclusion that isoleucine can serve as a specificprecursor of senecic acid introduced an apparent anomalyin that in experiments with S.douglasii, the label fromMe-14C-~-methionine was found to be incorporatedTABLE 3Percentage activities of degradation products of senecioninereversal, of the tetrahydrofolate-mediated pathway ofmethionine biosynthesis. Splittstoesser and Mazelis '2have shown that radioactivity from Md4C-~-methi-onine is effectively incorporated into serine in a numberof higher plants.However, the distribution of radio-activity in the labelled serine was not determined. Wehave confirmed the observation that , in pea seedlings,the label from Me-14C-~-methionine is incorporated (11) derived from 2, 5-14C,isoleucine (V) ~-Activity of C-5Activity of C-2= 2-52 I. 0-10)Carbon atoms Percentage activityc- 1 15.5 amp; 0.6c-10 14.0 amp; 0.6C-6,7 32.0 2. 1.2C-2,3,6,7,8,9 70.8 f 2.7Ratio C-7/C-10 2.29 amp; 0.13Ratio C-9/C-1 2.50 f 0.22*Me . -ISICH.NH* 1 I 461 *CO2H MeC0.C02H - MeCO.SCoA preferentially (73433) into the necic acid componentactivity of the seneciphyllic acid (111) was found to beI t therefore appears that a pathway exists in higherof seneciphylline (I).In addition, 22-26y0 of theat C-8.l I( X ) jlcLscherne 1 )*MeMeCH2*CH -CH( NH2)-COp (Y)plants whereby the methyl group of methionine can SCHEME 7Me (0.14) into serine. Degradation of the rigorouslyIMe-CH2-CH-Me ---- CH2-CH-MeI ICH.NH2 CH-NH2,t- -- .-f-bsol;I1 c o p I C02H'bsol;SCHEME 6enter metabolic pathways which lead to isoleucine as anintermediary product. The C-8 atom of the necic acids(111) and (IV) is derived from C-6 of isoleucine (V), whichin turn is derived from the methyl carbon atom ofpyruvate (cf. Scheme 1). It follows that the methylgroup of methionine can be converted in Senecio speciesinto the methyl group of pyruvate.Since it is well known that serine (XV) can be in-purified serine showed that the radioactivity was entirelylocated in the hydroxymethyl group (see Scheme 8).3 3HO*CH2*CH(NH21*C02H - HCHO (XIActivity =I00 Activity ~ 1 0 1SCHEME 8Rtagents: i, NaIO,; ii, dimedone.A similar selective incorporation of methionine methylcarbon atom into serine has been demonstrated in therat.13 These results lend support to the suggestionmade above concerning the route whereby the methylcarbon atom of methionine is incorporated into seneci-phyllic acid.By comparison with the results obtained with S.dozlgZasii,l methionine was found to be a very poor pre-cursor of senecionine in S.magnijcus (see Table 1). Ittherefore appears that there are large interspeciesdifferences in the degree to which methionine participatesin metabolic pathways having serine as an intermediate.If it is assumed that acetate enters isoleucine and thecorporated via pyruvate and acetyl CoA'into metabolites necic acids by the pathway outlined in Scheme 1, the(e.g. chloroplastidic terpenoids) directly derived from labelling pattern in necic acids derived from 14Cacetateacetate,S*lO it appeared possible that the methyl group can be predicted.Although the incorporation ofof methionine (XVI) enters the biosynthetic pathway l-14Cacetate into seneciphyllic acid accords with thatleading to isoleucine by conversion into the hydroxy- predicted,l the pattern of incorporation of activity frommethyl group of serine (XV) (see Scheme 7).* Such a 2-14Cacetate into seneciphyllic, senecic, and retronecic1*2conversion would constitute a reversal, or an effective acids shows that considerable randomisation of acetate* For further examples of incorporation studies which reflectthe probable operation of this pathway, see refs. 9 and 11,J.P. Kutney, J. F. Beck, V. R. Nelson, K. L. Stuart, and A. K.Maier, and P. Simchen, Experientia, 1970, 26, 820.synthetica, 1968, 2, 184.10 S. p. J. shah, L. J. R ~ ~ ~ ~ ~ , and T. W. ~ ~ ~ d ~ i ~ ~ i ~ ~ h ~ J.,11 S. p. J. Shah and L. J. Rogers, Biochem. J . , 1969, 114, 395.1967, 103, 52p; 1967, 105, 13p; 1968, 108, 17p.6, 39.Chem., 1956, 221, 885.Base* J . A ~ w . Chew. S0C.j 19702 927 2174) and D* Groger? W. 12 W. E. Splittstoesser and M. Mazelis, Phytochemistyy, 1967,0 L. J. Rogers, S. P. J. Shah, and T.W. Goodwin, Photo- 13 R. L. Kisliuk, W. Sakami, and M. V. Patwardhan, J . Biol676 J.C.S. Perkin Iradiochemical yield of 2-14Cisoleucine from diet hylacetamid02-~~Cmalonate. The product was shown byamino-acid analysis to consist of DL-isoleucine (60 amp; 3)and DL-alloisoleucine (40 amp; 3). This material wasused for feeding in experiment 4 (see Tables 1 and 2).Method (b) (Scheme 9), although longer than method(a) gave a much improved yield of isoleucine (22).In the radiochemical synthesis, a 17.5y0 radiochemicalyield of 2-14Cisoleucine was obtained from diethyl2-14Cmalonate. This material consisted (amino-acidanalysis), of equal amounts, within experimental error,of DL-isoleucine and DL-alloisoleucine. The final Schmidtreaction Scheme 9 ( b ) , which differed from the publishedmethyl carbon atom takes place.The main cause ofrandomisation is probably the obligatory passage ofacetate through the citric acid cycle (Scheme 1) and theinvolvement of acetate in associated anabolic andanaplerotic pathways. Compartmentation of inter-mediary metabolites, for which there is now compellinge~idence,~J~ is also probably involved.Equilibration of 2-l4Cacetate in the citric acid cycleleads to distribution of activity over all four carbonatoms of the C, intermediates, with the terminal positionseach carrying one sixth of the activity and the centralatoms one-third. The label from l-14Cacetate, how-ever, is distributed only in the terminal positions of the( a ) MeCHiCHMeBr + AcNH-CH(C02Et I 2 L MeCHt.CHMe-C(COgt lt.NHAc (Yl(m1...(6) MeCH2-CHMeBr + CH2(CO#t12 MeCH,-CHMeoCH(CVt Iz i v + MeCH2.CHMe*CH(CO2H lz v, (Y)(m) (XIXI( d ) EtCH:C(CO2Etl2amp; EtMeCH*CH(C0gt12 i v r " - (Yl(XXTI (XY-Ixr)SCHEME 9Reagents: i, KOBut-dimethyl sulphoxide; ii, HC1, iii, NaOEt; iv, KOH; v, HN,; vi, EtNgI; vii, MeMgI.C, intermediates and is completely eliminated, as CO,,after two turns of the cycle. The opportunity forrandomisation of acetate carboxy-carbon atom is there-fore much less than for the methyl carbon atom. Theseconsiderations afford a possible explanation for thedifference in the specificity of the incorporation of theacetate carboxy- and acetate methyl carbon atom intosenecic acid and related metabolites.The specific andrandom incorporations of 1-14C- and 2-14C-acetate, re-spectively, have recently been observed in the course ofinvestigations into the biosynthesis of ~0niine.l~Synthesis of S$eciJ;cally Labelled I~oleucine.-2-~~C-Isoleucine was prepared by the two routes (a) and (b)illustrated in Scheme 9. Condensation of diethylacetamidomalonate (XVII) with 2-bromobutane inthe presence of base gave only poor yields of isoleucine(V); the product invariably contained a large excess ofglycine. In ethanol with sodium ethoxide, yields ofisoleucine in the range l-6 were obtained. Thesynthesis was improved to give a yield of 9 withpotassium t-butoxide in dimethyl sulphoxide. A radio-chemical synthesis under the latter conditions gave a 5* A five-step procedure designed to allow for the preparationof isoleucine labelled in the 6-position has been described butdoes not appear to have been put into effect (G.B. Ceresia,G. L. Jenkins, and E. F. Degering, J . Amer. Pharm. Assoc., 1951,40, 341).procedure16 in that it was carried out under homo-geneous conditions, therefore proceeded without anappreciable degree of asymmetric induction. The2-14Cisoleucine obtained by this method was used inexperiment 5 (Tables 1 and 2).5-14C - and 6-14C-Isoleucine were prepared by con-jugate addition of 2-14Cethylmagnesium iodide and14Cmethylmagne~ium iodide respectively to the appro-priate diethyl alkylidenemalonate Scheme 9(c) and (41,followed by hydrolysis and Schmidt degradation of theresulting substituted malonic acid.A 30-33y0 chemicalyield of isoleucine was obtained by both methods. Theradiochemical yield of 6-14Cisoleucine was lower as thelabelled methyl iodide was used in excess."EXPERIMENTALAll m.p.s are corrected. Radioactivity was iiieasured in aPackard 2000 series Tri Carb liquid scintillation spectrometer.Samples were counted in dioxan-based scintillation solutionsNE 220 and NE 250 (Nuclear Enterprises Ltd.) and inB.D.H. dioxan scintillation solution. Coloured samplesl4 -4. Oaks and R. G. S. Bidwell, Ann. Rev. Plant Physiol.,l5 E. Leete, J . Amer. Chem. SOC., 1970, 92, 3835.l6 S. Takagi and K. Hayashi, Chein. and Pharm. Bull. (Japan),1970, 21, 43.1959, 7, 96, 1831972.-were oxidised to CO, by the Van Slyke meth0d.l' TheCO, was absorbed in a solution of 2-aminoethanol in 2-meth-oxyethanol (1 : 11 v/v; 6 cm3). For counting, a 5 cm3aliquot portion was added to a solution (7 cm3) of 2,5-di-phenyloxazole (PPO) (8.25 g) in scintillation-grade toluene(1 dm3). Barium acetate and amino-acids were also countedin the latter system. Alternatively, amino-acids were dis-solved in hyamine hydroxide (1 mol dm-3) in methanol (1cm3) and the resulting solution was added to toluene-PPOscintillator (1 1 cm3). Radioactive samples were normallyrecrystallised to constant activity and counted in duplicate.In a few cases, where only limited amounts of material wereavailable, samples were rigorously purified by methodswhich were known to give radiochemically pure materialand were then counted in duplicate.For radioautography,Kodires X-ray film was used. Alternatively, paper chroma-tograms were scanned by cutting the paper into narrowstrips, moistening each strip with NE 220 scintillator (1 om3)and counting the strips in the scintillation counter. Forpaper chromatography of the amino-acids the systemsbutanol-acetic acid-water (37 : 9 : 25) and butanol-pyrid-ine-water (1 : 1 : 1) were used. For preparative t.1.c. ofamino-acids, Kieselgel P F 2 5 4 (Merck) was used. Electro-phoresis was carried out at 100 V cm-1 for 1 h at pH 1-9 inacetic acid-formic acid-water (6 : 2 : 40). For the prepara-tion of 5,5-dimethylcyclohexane-1,3-dione (dimedone) deri-vatives of aldehydes, a 0.4y0 solution of dimedone was used.All radiochemicals were purchased from the RadiochemicalCentre, Amersham, Bucks. The quoted errors in percentageactivities are overall standard errors computed from thecalculated statistical errors and weighing errors in the usualway.Feedi~zg ,Vethods.--Senecio rnagnificus plants were grownfrom seed in a standard compost.Immediately before thefeeding experiments, 3-6 plants, (2-6 months old) wereremoved from the compost and their roots were washedfirst with tap water and then with deionised water. Theplants were placed in nutrient solution Phostrogen(Phostrogen Ltd.) (0.6 g dm-3) and the solutions werecontinuously aerated for the duration of the experiment.Radioactive precursors, dissolved in deionised water, wereadded directly to the nutrient solution.After 8-10 days,the plants were removed from the solution and their rootswere washed with deionised water. The plants were re-peatedly macerated with methanol in a Waring blendor untilthe filtered methanolic extract was colourless. In a typicalwork-up, the combined methanolic extracts were evap-orated, the residue was taken up in H2S04 (1 mol dm-3, 30cm3), and the acidic solution was washed with chloroform(6 x 30 cm3). The acidic solution was stirred with zincdust for 90 min (to reduce N-oxides) and filtered. Thefiltrate was washed with chloroform (3 x 30 cm3), madestrongly alkaline with conc. ammonia, and extracted withchloroform (6 x 30 om3). The combined extracts weredried (Na,SO,) and evaporated to give the crude alkaloid.This was purified by dissolving it in H2S04 (1 mol dmW3, 30cm3), washing the acidic solution with chloroform (4 x 30om3), making the residual solution strongly basic with conc.ammonia, and again extracting with chloroform (6 x 30cm3). The extracts were dried (Na,SO,) and evaporatedto give substantially pure senecionine (11) (0.2-0.5 basedon the dry weight of the plant material).The labelledsenecionine was diluted with inactive senecionine (200 mg)and applied to a column of acid-washed, activated alumina(12 g). Senecionine was eluted with chloroform and re-677crystallised (chloroform-methanol) to constant activity.Uptake of radioactivity by the plants was usually greaterthan 95 after 8-10 days.In the experiment with doublylabelled isoleucine (Table 3), in which the precursor had arelatively low specific activity, the uptake was 87.Hydrolysis of Senecionine.-In a typical experiment,senecionine (11) (200 mg, 0.6 mmol) in water (4 cm3) wasboiled under reflux with barium hydroxide octahydrate(190 mg, 0.6 mmol) for 2-5 h. The solution was allowed tocool and was treated with solid CO,. The resulting slurrywas filtered; the filtrate was acidified (Congo Red) withdilute HCl and extracted continuously with ether for 48 h.The ether extract was dried (Na,SO,) and evaporated. Theresidue was recrystallised ethyl acetate-light petroleum(b.p. 60-80deg;) to give senecic acid (IV) (78 mg), m.p. 142-144". The residual acidic solution was passed through acolumn of Dowex 1-X8 ion-exchange resin (OH-, 10 g) andthe eluate was collected until it was no longer alkaline(litmus) and evaporated to dryness.The residue was ex-tracted thrice with boiling acetone. The extracts werefiltered and evaporated. The residue was neutralised withdil. HC1, filtered to remove polypyrrolic material, and evap-orated. The residue in methanol (3 cm3) was boiled brieflywith activated charcoal and the solution was filtered andevaporated. The residue was crystallised (acetone) to giveretronecine (VII) hydrochloride (41 mg), m.p. 164".Degradation for C-1 of Senecic Acid (IV) .-Senecionine (11)(250 mg) in tetrahydrofuran (150 cm3) was heated underreflux with lithium aluminium hydride (800 mg) for 14 h.Water (50 cm3) was cautiously added to the mixture and thetetrahydrof uran was removed under reduced pressure.Theresidual slurry was filtered (Kieselguhr) and the filtrate waspassed through a column of Dowex 50W-X8 ion-exchangeresin (Hf; 20 g). The eluate was evaporated to give5-hydro~ymethyl-2~3-dirnethylhept-5-ene- 1,2-diol (VIII) asan oil (130 mg) . This was dissolved in O*O5~-periodic acid(50 cm3) and left in the dark for 2 h. The solution was passedthrough a column of Dowex 1-X8 ion-exchange resin(HC03-; 10 g) and the eluate was added to dimedone solu-tion (100 cm3). After 24 h, the derivative (X) (120 mg) offormaldehyde was filtered off and purified by preparativet.1.c. Kieselgel GF254; benzene-methanol (9 : l). Thederivative was extracted with chloroform and recrystallised(ethanol-water) to give needles, m.p.193".Degradation for C-8 of Senecic Acid (IV).-Senecic acid(IV) (108 mg, 0.5 mmol) in water (15 cm3) was treated withsodium periodate (300 mg, 1.4 mmol) and the solution wasset aside in the dark for 48 h. Sulphur dioxide was passedinto the solution until the iodine colour was just discharged.NaOH (1 mol dm-3, 20 cm3) was added, followed by iodine-potassium iodide I, (5 g) and potassium iodide (10 g) inwater (50 cm3) dropwise until the yellow colour persistedfor 1 min. The mixture was kept overnight; the pre-cipitate was filtered off and recrystallised (methanol-water)to give iodoform (110 mg), m.p. 119".Degradation for C-6,7 and C-10 of Senecic Acid (IV).-Thefiltrate after removal of the dimedone derivative of formal-dehyde obtained by the periodate oxidation of the trio1(VIII) (180 mg) as before, was steam-distilled until the distil-late no longer gave any turbidity when tested with 2,4-di-nitrophenylhydrazine (12.6 mmol dm-3) in H2S0, (1 molThe distillate was continuously extracted withl7 D.D. Van Slyke, J. Plazin, and J. R. Weisiger, J. Biol.Chew., 1951, 191, 299; D. D. Van Slyke, R. Steele, and J. Plazin,ibid., 1951, 192, 769678 J.C.S. Perkk Iether for 24 h; the extract was dried (MgSO,) and evap-orated to give 5-hydroxymethyl-3-methylhept-6-en-2-one(XII) (88 mg, 0.56 mmol) as an oil. This was stirred withether-water (1 : 1 v/v; 30 cm3). Sodium periodate (250mg, 1.17 mmol) was added, followed by osmic acid solution(4 cm3; osmium tetroxide 4 mg ~ m - ~ ) and stirring was con-tinued for 4 h.The ether was removed under reducedpressure and the aqueous residue was passed through acolumn of Dowex 1-X8 ion-exchange resin (HC0,- ; 10 g) .The eluate and washings (45 cm3) were treated with acetatebuffer (pH 4.6, 20 cm3) followed by dimedone solution (35cm3). After 3 days the precipitate was filtered off to givethe mixture of dimedone derivatives of acetaldehyde andformaldehyde (1 2 1 mg) . The mixture in acetic acid (2 cm3)was heated on a steam-bath under reflux for 6 h. Theproduct was poured into water (12 cm3), left overnight, andfiltered. The solid was washed with water (5 cm3) andNaOH (1 mol dm-3, 10 cm3) in small portions.Theinsoluble residue was washed with water and dried to givethe anhydro-derivative of acetaldehyde dimedone (21 mg) .The alkaline washings were neutralised with acetic acid,treated with acetate buffer (pH 4.6, 25 cm3), and left over-night, to give the dimedone derivative of formaldehyde(79 mg). Both derivatives were purified by preparativet.1.c. as described before to give the anhydro-derivative ofacetaldehyde (XI) (1 1 mg), m.p. 177-1 78' (ethanol-water) ,and the dimedone derivative of formaldehye (X) (53 mg),m.p. 193-194' (ethanol-water).Alternative Degradation of 5-Hydroxymethyl-3-methylhept-5-en-2-one (XII).-The filtrate from the removal of thedimedone-formaldehyde derivative (cf. preceding section)was continuously extracted with ether for 18 h. Theextract was dried (MgSO,) and evaporated, and the residuewas triturated with acetone and left at 3" for 1-5 h.Themixture was filtered and the filtrate was evaporated togive the ketol (XII) (182 mg from 497 mg senecionine).The ketol was ozonised in 50 aqueous acetic acid untilno more ozone was consumed. The solution was stirredwith zinc dust for 0.5 h and steam distilled until the distillateno longer gave a positive test with 2,kdinitrophenyl-hydrazine reagent. The distillate was brought to pH 4.5by the addition of sodium hydrogen carbonate and wastreated with dimedone solution (50 cm3). After 2 days, thesolution was filtered to give the dimedone derivative ofacetaldehyde (16 mg), m.p. 145-147' (homogeneous byt.1.c.benzene-methanol (95 : 5). The derivative wasdiluted with inactive material (32 mg) and purified by pre-parative t.1.c. Kieselgel GFZs4, benzene-methanol (95 : 5 ) Jto give acetaldehyde dimedone (IX) (32 mg), m.p. 142-143". The residue from the steam distillation was filteredand extracted with ether for 17 h. The extract was dried(MgSO,) and evaporated to leave an oil (165 mg) containingl-hydroxy-4-methylhexane-2,5-dione (XIII) , which wasoxidised with periodic acid (0-05 mol dm-3; 50 cm3) asbefore to give, after purification by preparative t.1.c. andrecrystallisation as before, the dimedone derivative offormaldehyde (24 mg), m.p. 192-193'.Degradation for C-2 of Senecic Acid.-Magnesium turnings(70 mg) were stirred with bromobenzene (370 mg) in ether(2 cm3) until the Grignard reagent was formed.The ketol(XII) (80 mg) obtained by steam distillation of the productfrom the periodate oxidation of the triol (VIII), as de-scribed before in ether (30 cm3) was added to the solutionl8 F. Wild, ' Estimation of Organic Compounds,' CambridgeUniversity Press, Cambridge, 1953, p. 220.of the Grignard reagent and the solution was stirred undernitrogen for 2 days. Hydrochloric acid (5 ; 30 cm3) wasadded and the mixture was extracted with ether (5 x 50cm3). The extracts were dried (MgSO,) and evaporated.The residual oil was dissolved in ether (30 cm3) and thesolution was washed with sodium hydroxide solution (0.25rnol dm-3; 4 x 10 cm3) and water (2 x 10 cm3).Theethereal solution was dried (MgSO,) and evaporated to givean oil (120 mg). This was boiled under reflux with potas-sium permanganate (600 mg) in water (60 cm3) for 4 h.The solution was cooled and ethanol (15 cm3) was added todecompose the excess of permanganate. The mixture wasfiltered (Kieselguhr) and the filtrate was acidified (dil. HC1)and extracted with ether (6 x 60 cm3). The extracts weredried (MgSO,) and evaporated to give a white solid. Thiswas dissolved in ether (25 cm3) and the solution was ex-tracted with sodium hydroxide solution (1 mol dm-3;4 x 2 cm3). The alkaline extracts were acidified (conc.HCl) and extracted with ether (6 x 60 an3). The etherealextracts were dried (MgSO,) and evaporated, and the residualsolid was purified by preparative t.1.c.Kieselgel GF,,, ;benzene-methanol-acetic acid (45 : S : 4) to give benzoicacid (16 mg), m.p. 120'.Ozonolysis of 5-Hydroxy~nethyl-2,3-dirnethylhept-5-ene- 1,2-diol (VIII).-The triol (193 mg), obtained as before, wasozonised and the product acetaldehyde was similarly iso-lated as the dimedone derivative. The product waspurified by preparative t .l.c. and recrystallisation to givethe derivative (IX) (61 mg), m.p. 141-142".Kuhn-Roth Oxidation of Senecionine.-Senecionine (11)was oxidised by the standard procedure.18 The acetic acidproduced (2-6-2.8 mol. equiv.) was titrated with bariumhydroxide. The solution of the barium salt was evaporatedto dryness and the residue was recrystallised (water-ethanol) to give barium acetate monohydrate as silkyneedles (36-38 yield).Schmidt Degradation of Barium Acetate.-This was carriedout as previously described.l Methylamine was isolated as5-methylamino-2,4-dinitrotoluene (XIV) .The derivativewas purified by preparative t.1.c. Kieselgel GF254; benzene-methanol (95 : 5) and crystallisation (ethanol). Frombarium acetate (30 mg) there was obtained the derivative(XIV) (15 mg), m.p. 170-172". The evolved CO, wascollected as barium carbonate.Feeding Experiments witlz Pea Seedlings.-The roots wereremoved from 24 ten-day-old pea seedlings and the seed-lings were placed in tubes each containing Ale-14C-~-methionine (1 cm3 ; 0.0041 6 mCi) . After 24 h the seedlingswere washed with deionised water and the amino-acids wereextracted by maceration in a Waring blendor with 75aqueous ethanol (94 of the administered radioactivitywas absorbed by the seedlings). The extraction procedurewas repeated twice.The filtered extracts were applied toa column of Dowex 5OW-XS ion-exchange resin (Hf ; 15 g),the column was washed with deionised water and thenwith ammonia solution (1.5 rnol dm-3) to elute the amino-acids. The ammoniacal solution was evaporated to drynessunder reduced pressure ; the residue was dissolved in waterand applied to a column of Amberlite CG-4B ion-exchangeresin (OH- ; 15 8). The neutral and basic amino-acids wereeluted with water and the acidic amino-acids were elutedwith dilute HC1 (1 mol dm-3).19 The eluate containing theneutral and basic amino-acids was concentrated and19 R.Consden, A. H. Gordon, and A. J. P. Martin, Biochern.J., 1948, 42, 4431972 679chromatographed on a column of Amberlite CG-120 ion-exchange resin (140 cm) .20 The amino-acids were elutedwith pyridine-formic acid buffer (pH 3.1); the eluate wascollected in 10 cm3 fractions. Serine (9 mg) was eluted infractions 54-75 together with two other amino-acids, oneof which was indicated by amino-acid analysis to be homo-serine.21 The amino-acids were isolated by lyophilisation ;the mixture was diluted with inactive serine (540 mg) andrecrystallised to constant activity.Periodate Oxidation of Serine (XV).-Serine (XV) (76 mg)in water (57 cm3) containing sodium periodate (620 mg) wasset aside for 17 h in the dark. Sodium arsenite was added,followed by dimedone solution (7 7 cm3).The precipitatewas recrystallised (ethanol-water) to give the dimedonederivative of formaldehyde (1 72 mg) which was recrystallisedto constant activity.Synthesis of Specijically Labelled Isoleucine.-The radio-chemical syntheses were carried out without purification ofintermediates, following trial experiments in which inter-mediates were satisfactorily characterised by spectroscopicmethods.Syathesis of 2-14CIsoleucine.-(a). Diethyl acetamido-2-14Cmalonate (XVII) (0.2 mCi; 6.16 mCi mmol-l) indimethyl sulphoxide (5 cm3) was added to a solution ofinactive diethyl acetamidomalonate (0.5 g) in dimethylsulphoside (15 cm3). The resulting solution was added topotassium t-butoxide prepared from potassium (0.25 g) andt-butyl alcohol.2-Bromobutane (1.5 g) was added and themixture was heated at 60" for 14 h. Water (80 cm3) wasadded and the solution was extracted with ether (5 x 50cm3). The extracts were dried (Na,SO,) and evaporated.The residue was boiled with conc. HCl (20 cm3) for 4 h andthe solution was evaporated to dryness. The residue wasdissolved in water (20 cm3) and again evaporated to dryness.The residue was dissolved in water (10 cm3) and applied toa column of Dowex 50W-X8 ion-exchange resin (Hf; 10 g).The column was washed with water (100 cm3) and theamino-acids were eluted with gix)nia solution (0.5 moldm-3 ; 150 cm3). The eluate WL- -zzporated to dryness andthe product was purified by preparative t.1.c.butanol-acetic acid-water (4 : 1 : l). The isoleucine band wasextracted with water and purified by passage throughDowex 50'cV-X8, as before. Evaporation of the ammoniacaleluate gave the product as a white solid (30 mg) in 5radiochemical yield. Amino-acid analysis showed that themixture consisted of DL-isoleucine (60 f 3) and DL-alloisoleucine (40 f 3). A radiochemical and chemicalpurity of 100 was indicated by autoradiography and paperchromatography.(b) To a warm solution of sodium (33.6 mg, 1.46 mmol) inethanol ( 1 - 5 cm3) was added with stirring dieth~l2-~~C-malonate (0.1 mCi, 5 mCi mmol-l) diluted with inactivematerial (232 mg, 1.45 mmol), followed after 5 min by2-bromobutane (200 mg, 1.46 mmol).The mixture wasstirred and heated under reflux for 26 h and was then lefta t room temperature for 17 h. Water (20 cm3) was addedand the solution was extracted with ether (4 x 8.5 cm3).The extracts were dried (MgSO,), filtered, and evaporated toL L L L ~ ~ L ie diethyl ~-butyl2-~~Cmalonate (XVIII) as a pale?aperllow liquid (218 mg). The crude ester was heated withrutassium hydroxide (218 mg) in water (2 cm3) for 5 h.The product was cooled and acidified (Congo Red) with conc.HCl, care being taken to ensure that the temperature never2o J. Liebster, amp;I. Dobiasova, J. Kopoldova, and J. Ekl, CoZZ.Czech. Chew. Comm., 1961, 26, 1700.rose above 10" during acidification. The acidic solutionwas extracted with ether (4 x 8 cm3); the extracts weredried (MgSO,) and evaporated to give the crude 2-butyl-malonic acid (XIX) as a viscous oil (121 mg).The crudeacid was dissolved in conc. H,SO, (2 cm3) with cooling, themixture was allowed to warm to room temperature, andsodium azide (70 mg, 1-08 mmol) was added. The mixturewas warmed rapidly to 60" and maintained a t that tem-perature for 3 h. Three portions of sodium azide (each 30.5mg, 0.47 mmol) were added during this time, one after eachhour. The mixture was maintained a t 60" for 1 h afterthe addition of the last portion of sodium azide, and themixture was cooled and poured into ice-water (15 cm3).The solution was washed with ether (3 x 5 cm3) and wasbrought to pH 3 with barium acetate. The solution wasfiltered (Kieselguhr), the precipitate was washed with water(20 cm3), and the combined filtrate and washings wereconcentrated to 15 cm3 under reduced pressure.The amino-acids were purified by passage through a column of Dowex50bsol;V-X8 as described before, followed by preparative t.1.c.chloroform-methanol-ammonia (40 : 40 : 20) and a secondpassage through Dowex 50W-X8, to give 2-14Cisoleucine(41 mg, 17.5 radiochemical yield). Amino-acid analysisindicated that the product contained DL-isoleucine (53 amp;3) and DL-alloisoleucine (47 amp; 3). A radiochemicalpurity of 98 was indicated by radioautography of paperchromatograms and electrophoretograms and by paperchromatography with liquid scintillation scanning.5-14CI~~Ze~cine.-To 2-14Cethyl iodide (0.5 mCi;4.28 mCi mmol-l) cooled in liquid nitrogen was added asolution of inactive ethyl iodide (404 mg, 2.58 mmol) inether (3 cm3). The mixture was added to magnesiumturnings (75 mg, 3.1 mmol).When the reaction had begun,ether (5 om3) was added and the mixture was stirred for20 min, after which time nearly all of the magnesium hadreacted. Diethyl ethylidenemalonate (XX) (500 mg, 2.69mmol) in ether (5 cm3) was added over 15 min ; the mixturewas stirred for 1 h and acidified with HCl(1 mmol dm-3), andthe ethereal layer was separated. The aqueous layer wasextracted with ether (2 x 10 om3). The combined extractswere dried (MgSO,) and evaporated to give the crude2-14CCdiethyl s-butylmalonate (XVIII) as an oil (550 mg).The ester was converted into isoleucine by method (b) forthe preparation of 2-14Cisoleucine. 5-14C Isoleucine wasobtained in 22.6 radiochemical yield. A radiochemicalpurity of 100 was indicated by radioautography and of 99 by dilution analysis.6-14CI~~leucine.-Inactive methyl iodide (800 mg, 5.6mmol), in ether (0.5 cm3) was added to 14Cmethyl iodide(0-05 mCi, 51.5 mCi mmol-l). The solution was added tomagnesium turnings (60 mg, 2.5 mmol), reaction was initi-ated by the addition of a crystal of iodine, and ether (2 cm3)was added. The mixture was stirred until all the magnes-ium had been consumed and diethyl propylidenemalonate(XXI) (500 mg, 2.5 mmol) in ether (2 cm3) was added drop-wise. The solution was stirred for 1 h ; water (20 cm3) wasadded, followed by conc. HC1 (3 cm3), and the mixture wasextracted with ether (5 x 40 cm3). The extracts weredried (MgSO,) and evaporated to give the crude ester(XVIII) as a red oil (450 mg). This was converted intoisoleucine as before. The amino-acid was purified by pas-sage through Dowex 5OW-XS ion-exchange resin, preparative41 A. I. Virtanen, A. Berg, and S. Kari, Acta Chena. Scand.,1953, 7, 1423; L. A. Larson and H. Beevers, Plant PhysioZ., 1965,40, 424; J. M. Lawrence and D. R. Grant, ibid., 1955, 38, 561680 J.C.S. Perkin It.1.c. butanol-acetic acid-water (4 : 1 : l), and a second We are grateful to the Chief Botanist, Arid Zone Researchpassage through Dowex 50W-X8 to give 6-Wiso- Institute, Alice Springs, Australia, and his staff, for suppliesleucine (21 mg) in 6 radiochemical yield. A radiochemical The receipt of S.R.C. student-purity of lOOyo was indicated by dilution analysis and of 99 by radioautography and paper chromatography withliquid scintillation scanning. 1/1402 Received, 9th Augztst, 19711of plant material and seeds.ships (to N. M. D., E. H. s., and D. W.) is acknowledged