sign in sign up
<<
Home , Isopimpinellin

Medicinal Plant Research Editor - in - Chief Editorial Board Hippokrates Verlag E. Reinhard, Univ. Tiibingen H. Arnmon. Tubingen Pharrnazeutisches lnstitut W. Barz, Miinster Stuttgat? Auf der Moraensteile E. Reinhard. Tubinaen"- 0.Sticher, ~~rich Vol. 36 H. Wagner, Miinchen M. Zenk, Bochum August 1979 NO.4

Biosynthetic Studies on

Stewart A. Brown

Department of Chemistry, Trent University, Peterborough, Ontario, Canada.

Key Word Index: Biosynthesis; Coumarins; Furocoumarins.

Abstract ortho- or para-hydroxycinnamic acid. ' Radiotracer experiments later indica- ~oumarins,a group of naturally oc- ted that the furan ring of the furano- This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. curring phenylpropanoid lactones exert- coumarins, found chiefly in the Ruta- ing a wide range of physiological effects, ceae and Umbelliferae, is constructed are elaborated by many plants and in a on the nucleus following at- few microbial species. Early studies on tachment of a mevalonate-derived side- the biosynthetic reactions leading to the chain to , and cyclization simple coumarins in plants demonstra- to dihydrofuranocoumarin intermedia- ted derivation from the shikimic acid tes. These findings have been substantia- pathway via cinnamic acid and either ted and extended by the isolation and 300 Brown characterization of several of the en- these metabolites are potent hepato- zymes participating in the biosynthetic toxins and are among the most intense pathway. carcinogens yet discovered [2, 33. Seve- ral coumarins hydroxylated atcarbon-4,

, including the well-known Introduction '(IV), function as anticoagulants [4, 51. Dicoumarol results from the action of The coumarins are a class of second- Aspergillus and Penicillium spp. on ary plant and microbial products whose sweet .clover (MeZilotus sp.) hay, the members exhibit a considerable range of substrae probably being o-hydroxycin- physiological effects in animils. The . namic acid [6]. It is responsible for the least complex member of the class struc-' very destructive haemorrhagic sweet turally, known ,simply as coumarin (I), clover disease of cattle. A synthetic 4- is toxic to mammals [:I.], and at the other hydroxycoumarin (V) has been sold for end of the scale the highly substituted many years as an effective rat poison coumarin, novobiocin (11), is a com- under the name . mercial antibiotic. Another microbial Of particular interest to the phar- coumarin type is the aflatoxins (e. g. macologist and the phytochemist alike 111) elaborated by Aspergillus flavus; are the linear , or furo-

OH

81

OH 111 I OMe 11 This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. (a) : R5 = Re = H (b) : R6 = OMe, RE = H (c) Xanthotoxin: R, = H, R, = OMe (d) Isopimpinellin: R6 = R8 = OMe (e) 5-Hydroxyxanthotoxin: R, = OH. RE = OMe (f) 8-Hydroxybergapten: R, = OMe, R, = OH (g) Bergaptol: R, = OH, Re = H (h) : R, =, H, RE = OH Biosynthetic Studies on Coumarins 301 . . coumarins. One of these, xanthotoxin (VIc), has been recognized as poisonous to fish for many years, and this pro- perty is shared by some others in this . class. But more importantly several na- tural furanocoumarins, and especially psoralen (VIa) and xanthotoxin, are noted for their unique ability to photo-. . sensitize the skin to the near ultraviolet, leading to the development of erythema, vesicle formation, and pigmentation of the affected areas [7]. Of all the physio- logical effects of the coumarins this is the best understood at the molecular level, ,thanks largely to the work of MUSAJO, RODIGHIERO,DALL'ACQUA and the their collaborators in Italy [8] and PATHAK,FITZPATRICK, and their group in the United States [9].Studies on the mechanism of their action have shown that linear furanocoumarins can, under the influence of appropriate ultraviolet wavelengths, form cross- linkages between the two strands of the . DNA double helix, joining pyrimidine bases, (VIIa-d), and thereby resulting furanocoumarins, which are princ;pal- in a partial loss of template activity ly found in species of the two plant for RNA synthesis as well as inhibition families Umbelliferae and Rutaceae. of DNA replication. The American group have recently taken advantage of the reduced protein synthesis this im- Biosynthetic Routes to the simple plies in the epidermal cells to devise Coumarins an effective treatment for the stubborn and disfiguring skin disease psoriasis, Coumarins, to, the biochemist, are which is characterized by a proliferation lactones of phenylpropanoid acids, a of these epidermal cells. Oral doses of class which derives 'from the shikimic This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. xanthotoxin followed by controlled acid pathway via the protoaromatic exposure to ultraviolet radiation have amino acids phenylalanine and, in a few shown remarkable success in experimen- plants, tyrosine [12]. The entire carbon ' tal treatment of this disease [lo, 1I]. skeleton of the simple coumarins, i. e. The subject of the present review is those having no rings additional to the the biosynthesis of coumarins, and I benzopyrone nucleus, is phenylalanine- wish especially to discuss the linear derived [I31. In a number of early bio- 302 Brown

.. , synthetic experiments on these com- The ovtho-hydroxylations of cinnamic pounds in plants it was shown that and p-hydroxycinnamic acids are me- shikimic acid, phenylalanine, and trans- diated by different enzymes [22, 231, cinnamic acid are common precursors which are apparently seldom found to- of this benzopyrone nucleus [14-161. gether.in the same,species, so that coexi- Cinnamic acid, however, represents a stence of coumarin and the umbellifer- branch point in the elaboration of cou- ones is an exceptional fjhenomenon. marin itself and probably other cou- It should be noted that a number of marins lacking -/-oxygenation, on the the simple ,coumarins are artifacts of one hand, and the 7-hydroxycoumarins isolatiop, and do not occur to any very based on umbelliferone (VIII) on the significant extent in the free form in the intact plant cell. Thus coumarin [24], [25], and umbelliferone [26] HO have all been shown to occur in the form of glycosides of the corresponding cis- VIII o-hydroxycinnamic acids (e. g. IX, R= glucosyl). However, glucosidases are . other [17, 181, the latter group consti- also present which gain access to these tuting the vast majority of the class. substrates upon disruption of the cells, ortho-Hydroxylation of trans-cinnamic and the liberated aglycones (e. g. IX, R = acid leads to coumarin itself (Fig. I), H) lactonize spontaneously to the free

via a light-catalysed trans-cis isomeriza- ' coumarins [27]. The pungent dour of tion, and lactone ring formation which coumarin associated with new-mown can be formally represented as a dehy- 'hay originates in this way. dration [I9, 201. para-Hydroxylation of trans-cinnamic acid is a necessary prerequisite for synthesis of the 7-hy- ~coOH droxycoumarins, via ortho-hydroxyla- RI tion and lactonization as before (Fig. 1). IX

rCOOH > rCOOH, -. OH. -

CINNAMIC ACID 0-COUMARICACID COUMARIN This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

p-COUMARIC ACID Fig. I. Pathways from trans-cinnamic acid to coumarins. Biosynthetic Studies on Coumarins 303

In at least one case in the elaboration- pinella magna L. Coumarin itself was of a microbial coumarin - novobiocin not utilized, an observation consistent (11) by Streptomyces nivens - lactone with ,the idea that pathways to couma- ring closure is not by dehydration but rin and the oxygenated coumarins by an oxidative mechanism. KENNER diverge at the cinnamic acid.hydroxy- and his co-workers [28, 291 have ad- Iation stage. duced evidence based on tracer studies The furan ring is therefore construc- with I4C and 180 that the lactone ring ted on umbelliferone, the simplest of this compound originates directly naturally occurring hydroxycoumarin. from tyrosine, with the ring oxygen In the case of the linear furanocouma- deriving from the carboxyl oxygens of rins mentioned above the side chain is this amino acid. attached at position-6. In a second sub- Current evidence, admittedly incom- class, the angular furanocoumarins (X), plete, suggests that oxygenation of the which have a-more restricted distribu- benzene ring of simple plant coumarins tion, side-chain attachment is at position tends to occur befor the benzopyrone 8. A clue to the pathway of furan ring nucleus has been elaborated. Thus aescu- construction came from the existence letin (6, 7-dihydroxycoumarin) is for- of two substituted dihydrofuranocou- med from cafieic (3, 4-dihydroxycin- marins, (XI) in the linear namic) acid [30], and (6-me- series and columbianetin (XII) in the thoxy-7-hydroxycoumarin) from feru- angular. The isopentane skeleton of the lic (3-methoxy-4-hydroxycinnamic) side-chains in those compounds sug- acid [31, 321. But occasional instances gested that their origin is in an isopre- exist where a simple coumarin acts as n0i.d side-chain attached to carbons 6 or the precursor of another, more highly 8 of umbelliferone, respectively, and .oxygenated simple coumarin. arising ultimately from acetate via me- valonic acid (XIII). FLOSSand MOTHES[34] did indeed Biosynthetic Routes to the Furano- present evidence, from the feeding of coumarins

Consideration of possible biosynthe- tic pathways to the furanocoumarins raises the question of how the fused furan ring is constructed, and the order in which the three rings are formed. The This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. (a) Isobergapten: R, = OMe, R, = H 'first question has not yet been fully ans- (b) Sphondin: R, = H, R,, = OMe wered, but the second was answered in (c) Pimpinellin: R5 = R,, = OMe early studies by FLOSSand MOTHES[3 31, who showed by 14C tracer experiments that umbelliferone (7-hydroxycouma- rin) is an efficient precursor of the fura-

nocoumarins of the umbellifer, Pim- XI 304 Brown

XI1 XI11 "C-labelled mevalonic acid to P. mag- furanocoumarins [42]. The use of trapp- na, of a specific incorporation of car- ing techniques [42] has shown &at DMS, bons 4 and 5 of this compound into car- which apparently .accumulates in quan- bons 5' and 4', respectively, of the furan tity only in Chloroxylon swietenia DC., . ring of pimpinellin, an angular furano- a species of mahogany [40]. is formed coumarin, but this precursor was utili- from umbelliferone in zed with very low efficiency. However, L. of the Rutaceae as well. DMS and despite later questions raised about the osthenol are thus presumed to be inter- suggested role of mevalonate as an mediates in all species forming the res- intermediate in this pathway [35, 361, pective subclasses of furanocoumarins. the studies of KUTNEY'Sgroup 137-391 The reaction by which the dimethyl- on cell cultures of Thamnosma montana allyl side-chain is attached to C-6 of TORR.and FRBM.have amply confirmed umbelliferone to form DMS, the com- this role. mitted step in linear The specific incorporation of 14C biosynthesis, has been the subject of in- from mevalonate into the furan ring vestigation at the enzyme level in the suggests involvement of. an isoprenoid author's laboratory. The participating unit in the biosynthetic pathway, and enzyme, a particulate dimethylallyl its condensation at some stage with the (prenyl) transferase, was demonstrated benzopyrone nucleus. Several simple in extracts of R. graveolens cell cultures coumarins are known which have iso- [MI, as well as in fresh leaves of the prenoid side chains [40], but two which same species. It mediates the reaction are of particular interest in the present shown in Fig. 2. It does not, however, context are demerhylsuberosin (DMS, transfer a prenyl group from dimethyl- XIV) and osthenol (XV), which have allyl pyrophosphate to C-8 of umbelli- dimethylallyl side-chains attached ortho ferone to form osthenol. It is also in- to the hydroxyl group of imbelli-ferone, active against pyrophosphates of two at positions 6 and 8, respectively. Tra- or three isoprenoid units - geranyl and cer experiments have . demonstrated farnesyl. DHILLONand BROWN[45]

that DMS is a precursor of the linear later established that this transferase is This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. furanocoumarins [41-43] and that a chloroplast enzyme, a point of interest ostheriol is a precursor .of the angular in the light of earlier demonstrations

I4n2' , PRENYLASE

Fig. 2. Formation of demethylsuberosin by a prenylase from Ruta graveolens. Biosynthetic Studies on Coumarins 305

[22, 231 that hydroxylases mediating Analogous results were obtained in the committed step in the elaboration the angular series with columbianetin, of coumarin and the 7-oxygenated cou- in two other umbellifers, A. angelica marins are also chloroplast enzymes. and H. lanatum, both of whi& form The prenyl transferase was successfully angular as well as linear furanocou- solubilized, and purified over 300-fold marins [49]. . . by ammonium sulphate fractionation The nature of the pathways between followed by column chromatography DMS and marmesin, and between osthe- on DEAE-cellulose and hydroxylapa- no1 and columbianetin, is still not defi- tite [45]. nitely es~blished. It was' suggested The next intermediates in the biosyn- quite early in the biosynthetic studies thetic pathways to the linear and angu- that the intermediates in question are lar furanocoumarins, the hydroxyis- coumarins in which the double bonds propyldihydrofuranocoumarins, were of the prenyl side chains have been actually the first to be identified. As epoxidized to yield ;tructures XVI (R= mentioned above, these are marmesin H) and XVII [47, 491.' A coumarin of (XI) in the linear series and columbia- this type in the linear series is aculeatin netin (XII) in the angular. A glucoside (XV1,R =Me) which occurs in Toddalia of marmesin accumulates in the umbel- aculeata PERS.[4O]. Its 7-h~droxyl lifer Ammi majus L. [46]. This plant group is methylated, but where such is was the subject of tracer studies by not the caseit is easy to envisage the sort BROWN et al. [47], who found after ad- of reaction shown in Fig. 3, involving an ministration of [14C] umbelliferone that attack of the ~henolgroup on the epox- the specific radioactivity of the agly- ide ring, with a cyclization, and the for- cone was higher than that of the linear mation of a substituted dihydrofurano- furanocoumarins formed in the same coumarin. There is no experimental experiment. This indication of an inter- evidence pertaining to this step, and it mediate status for marmesin in the bio- is still possible that diols such as XVIII, synthetic pathway was confirmed by formed by hydrolysis of the analogous trapping experiments in.,R, graveolens, epoxides, participate. and in addition tritiated marmesin was Another important lacuna in our efficiently utilized as a linear furano- knowledge of these pathways is the coumarin precursor in both species. The nature of the mechanism whereby the . marmesin used in th&e species was the (+)-isomer; the enantiomer is inactive as a precursor in R. .graveolens, Ange- This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. lica archangelica L., and Heracleum lanatum MICHX.1481.

Fig. 3. Hypothetical cyclization reaction leading to marmesin.formation. 306 Brown

with furanocoumarins. It is true .that - CAPORALE'Sgroup [54, 551 have repor- ted 8-hydroxymarmesin (mtaretin, XX) to be an intermediate between marmesin isopropyl side-chain is eliminated from and 8-methoxypsoralen (xanthotoxin, . marmesin and columbianetin, and the 6c) in R. graveolens, bat this reaction double bond iitroduced to form the appears to be exceptional, as attempts to furan ring. The only evidence thus far demonstrate it in other species have fai- . is negative. In the linear series, despite led [56]. On the other hand, several tra' one early claim to the contrary [50], 4', cer studies [47, 54, 57, 581 have provi- 5'-dihydropsoralen (XIX) is almost cer- ded evidence that psoralen (VIa), the tainly not an intermediate [51, 521. unsubstituted linear furanocoumarin, is - KUTNEYand his co-workers [37] dedu- converted in vivo to the mono- and di- ced from experiments with specifically methoxypsoralens, and as a few species tritiated mevalonic acid that furan ring a~cum5latethe corresponding hydxoxy- formation cannot proceed via any [40, 591 it was postulated that mechanism that involves loss of all the psoralen is first hydroxylated, and that protons from either the 4' R- or 5'-phi- a transmethylation reaction completes tions of this intermediate. A plausible . the biosynthetic pathway to furanocou- mechanism, in which a carbonium ion is marins [47]. generated at C-4', followed by a 1, 3- This hypothesis has now been partial- elimination, as shown in Fig, 4, has ly substantiated by experiments in this been proposed by Birch and his associa- laboratory at the enzyme level. Exten- tes [53], the 3-carbon side-chain being sive work both in vivo and with cell- converted to acetone in the process. To free systems has been done in this labo- date there is no in vivo experimental ratory on the transmethylation reac- support for this hypothesis. tions which constitute the final step in It was noted above that oxygenation the synthesis of such coumarins as of the benzene ring of simple coumarins bergapten (VIb), xanthotoxin (VIc), can often occur before. the benzopyrone and isopimpinellin (VId) in the linear nucleus has been elaborated. The oppo- series, and which are also presumed to site situation seems generally to obtain' be involved in formation of isobergap- This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

Fig. 4. Mechanism proposed by BIRCHet al. 1531 for the conversio; of marmesin to psoralen. Biosynthetic Studies on Cournarins 307 ten (Xa), sphondin (Xb), and pimpinel- psoralen (xanthotoxol, VIh) ,produces lin (Xc) in the angular. Isopimpinellin, xanthotoxin. All four O-methylations 5,8-dimethoxypsoralen, appears to arise are done in the presence of S-adenosyl- in vivo by: the two pathways, de- L-methionine by a soluble methyltrans- picted in Fig. 5, although the one via ferase system in cell-free extracts of R. xanthotoxin and 5-liydroxyxanthotoxin graveolens which, in contrast to the (VIe) has been shown by tracer experi- enzyme mentioned earlier, does not ap- ments to predominate in R. graveolens pear to be confined to the chloroplast [601. [61]. Conditions of stability, molecular In cell-free extracts, O-methylation weight differences and differing effects of both 5-hydroxyxanthotoxin and of various divalent cations on activity 8-hydroxybergapten (VIf) leads to iso- suggested the existence of at least two pimpinellin, whereas O-methylation of O-methyltransferases in the system, 5-hydroxypsoralen (bergaptol, VIg) acting respectively at the 5- and 8-hy- yields bergapten, and that of S-hydroxy- droxyls.

OH ' . OMe OMe

BERGAPTOL BERGAPTEN 'H 8-HB

OMe

OMe

PSORACEN ISOPlMPlNELLlN This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

OH OMe OMe

XANTHOTOXOL XANTHOTOXIN Fig. 5. Biosynthetic routes to isopimpinellin in Rutd. 308 Brown

nmol PRODUCT FORMED/h AT 30° C

I I Xonthotoxol (8-OH)

I 1 5-OH Xonthotoxin

Bergaptol (5-OH)

. 5-OH Xonthotoxin

I , I Xanthotoxol (8-OH) This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

1 , I 8-OH Bergapten

Both

Fig. 6. Results of mixed substrate experiments utilizing an extract of Ruta graveolens acting on 5- and 8-hydroxylated linear furanocoumarins. Biosynthetic Studies on Coumarins

Conclusive evidence for two enzymes on the lactone ring carbons. T'heexisten: came from two further studies. As ce of these yet unexplored types, to- shown in Fig. 6, additive effects on gether with the still considerable. gaps activity were noted in the simultaneous in our knowledge of the biosynthesis of presence of a 5- and an 8-hydroxyfura- furanocoumarins, serves as a continuing nocoumarin, in the two diff erent cases. In reminder of the scope that remains for contrast, the total activity in the presen- biosynthetic research on this remarkable ce of either the two 5-hydroxylated or group of secondary plant products. the two 8-hydroxylated substrates was References lower than the higher of the two acti- vities observed when only one of the 1. Hazelton, L. W., T. W. Tusing, B. R. Zeit- two substrates was present. Kinetic lin, R. Thiesen Jun. and H. K. Murer: J. Pharmacol. Exp. Therap. 118, 348, (1956). theory therefore strongly indicated that 2. Ciegler, A:, R. W. Detroy and E. B. Lille- different enzymes were acting at the 5- hoj: Microb. Toxins 6, 409 (1971). and the 8-hydroxyls, but suggested that - 3. Detroy, R. W., E. B. Lillehoj and A. Cieg- no more than two enzymes were atting. . ler: Microb. Toxins 6, 3 (1971). In a second approach, an affinity system 4. Perone, V. B.: Microb. Toxins 8, 67 (1972). 5. Scheel, L D.: Microb. Toxins 8, 47 (1972). was developed with S-adenosyl-L-ho- 6. Bellis, D. M., M. S. Spring and J. R. Sto- mocysteine as ligand which selectively ker: Biochem. J. 103, 202 (1967). retarded 0-methyltransferases [62]. In 7. Musajo, L. and G. Rodighiero: Experientia a purification sequence incorporating 18, 153 (1962). this step activity ratios against the .8. Musajo, L. and G. Rodighiero: Photophy- siology 7, 115 (1972). 5- and 8-hydroxylated substrates fluc- 9. Pathak, M. A., D. A. KrZmer and T. B. tuated to a degree explicable only on the Fitzpatrick: in M. A. Pathak, L. C. Harber, basis of discrete enzymes [61]. Both the M. Seiji and A. Kukita (eds): Sunlight and mixed substrate studies (Fig. 6) and Man. Normal and Abnormal Photobiologic an examination of the divalent cation Responses, p. 335, University of Tokyo Press, 1974. effects indicate the absence of more than 10. Parrish, J. A., T. B. Fitzpatridt, L. Tanen- two 0-methyltransferases, but this can- baum and M. A. Pathak: New Engl. J. not yet be regarded as conclusively Med. 291, 1207 (1974). established. 11. Wolff, K., H. Honigsmann, F. Gschnait and It is of interest that thetransferase K. Konrad: Deut. Med. Wochschr. 100,2471 acting at position-5 methylates a hy- (1975). 12. Neish, A. C.: in Harborne, J. B. (ed.): Bio- droxyl ortho to the C, side-chain, the chemistry of Phenolic Compounds, p. 295, first instance of this pattern observed. London, 1964, Academic Press. Numerous other 0-methyltransferases, 13. Brown, S. A.: in Billek, G. (ed.): Biosyn- This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. including the one acting at C-8 of fura- thesis of Aromatic Compounds. Procee- nocoumarins, methylate meta or para dings of the 2nd Meeting of the Federation to the side-thain. of European Biochemical Societies, Vol. 3,. , Space has not permitted a considera- p. 15, Oxford, 1966, Pergamon Press. 14. Weygand, F. and H. Wendt: 2. Natur- . tion of studies done on other coumarins, forsch. 14 b, 421 (1959). nor is it possible yet to do more than 15. Kosuge, T. and E. E: Conn: J. BioI. Chem. speculate on the origin of still other 234, 2133 (1959). classes such as those bearing side-chains 16. Brown, S. A., G. H. N. Towers and D. Brown

Wright: Can. J. Biochem. Physiol. 38, 143 42. Brown, S. A. and W. Steck: Phyto&emistry (1960). 12, 1315 (1973). 17. Brown, S. A.: Z. Naturforsch. 15 b, 768 43. Innocenti, G., F. Dall'Acqua, A. Guiotto (1960). and G. Caporale: Atti 1st. Veneto Sci., Lett. 18. Brown, S. A.: Phytochemistry 2, 137 (1963). Arti. C1. Sci. Mat. Nat. 135, 37.(1977). . 19. Haskins, F. A., L. G. Williams and H. J. 44. Ellis, B. E. and S. A. Brown: Can. J. Bio- Gorz: Plant Physiol. 39, 777 (1964). chem. 52, 734 (1974). 20. Edwards, K. G. and J. R. Stoker: Phyto- 45. Dhillon, D. S. and S. A. Brown: Arch. Bio- chem. Biophys. 177, 74 (1976). chemistry 6, 655 (1967). 46. Starkowsky, N. A. and N. Badrani J. Org. . 21. Bown, S. A., G. H. N. Towers and D. Chem. 23, 1818 (1958). Chen: Phytochemistry 3, 469 (1964). 47. Brown, .S. A., M. El-Dakhakhny and W. 2. 22. Kindl, H.: Hoppe-Seylers Physiol. Chem. Steck: Can. J. Biochem. 48, 863 (1970). 352, 78 (1971). . 48. Steck, W. and S. A. Brown: Can. J. Bio- 23. Gestetner, B. and E. E. Conn: Arch. Bio- chem. 49, 1213 (1971). chem. Biophys. 163, 617 (1974). 49. Steck, W. and S. A. Brown: Can. J. Bio- '24. Haskins; F. A. and H. J. Gorz: Crop Sci. 1, chem. .48, 872 (1970). 320 (1961). 50. Caporale, G., P. Dall'Acqua, S. Marciani 25. Brown, S. A.: Can. J. Biochem. 43, 199 and A. Capozzi: 2. Naturforsch. 25 b, 700 (1965). (1970). 26. Austin, D. J. and M. B. Meyers: Phytoche- 51. Brown,. S. A.: Can. J. Biochem. 51, 965 mistry 4, 245 (1965). (1973). 27. Kosuge, T. and E. E. Conn: J. Biol. Chem. 52. Marciani, S., F. Dall'Acqua, G. Innocenti 236, 1617 (1961). and G. Caporale: Atti Inst. Veneto Sci., 28. Chambers, K., G. W. Kenner, M. J. T. Ro- Lett. Arti. CI. Sci. Mat. Nat. 132, 275 binson and B. R. Webster: Proc. Chem. . (1974). SOC.291 (1960). 53. Birch, A. J., M. Maung and A. Pelter: 29. Bunton, C. A,, G. W. Kenner, M. J. T. Ro- Australian J. Chem. 22, 1923 (1969). ' 54. Caporale, G., F. Dall'Acqua, A. Capozzi binson and B. R. Webster: Tetrahedron 19, and S. Marciani: 2. Naturforsch. 26b, 1256 1001 (1963). (1971). 30. Sato, M.: Phytochemistry 6, 1363 (1967). 55. Dall'Acqua, F., A. Capozzi, S. Marciani 31. Stedt, W.: Can. J. Biochem. 45, 889 (1967). and G. Caporale: 2. Naturforsch. 27 b, 813 32. Fritig, B., L. Hirth and G. Ourisson: Phy- (1972). tochemistry 9, 1963 (1970). 56. Dall'Acqua, F., G. Innocenti and G. Capo- 33. Floss, Ha G. and K. Mothes: Z. Natur- rale: Planta Med. 27, 343 (1975). forsch. 19 b, 770 (1964). 57. Austin, D. J. and S. A. Brown: Phytoche- 34. Floss, H. G. and K. Mothes: Phytoche- mistry 12, 1657 (1973). mistry 5, 161 (1966). 58. Caporale, G., F. Dall'Acqua, A. Capozzi 35. Caporale, G., A. Breccia and G. Rodighie- and S. Marciani: Ann. Chim. (Rome) 62, ro: Prepar. Bio-Med. Appl. Labeled Mol., 536 (1972). Proc. Symp., p. 103, Venice, 1964. . ' 59. Nielsen, B. E.: Dansk Tidsskr. Farm. 44, 36. Brown, S. A.: Phytochemistry 9, 2471 111 (1972). (1970).

60. Brown, S. A. and S. Sampathkumar: Can. This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited. 37. Kutney, J. P., P. J. Salisb,ury and A. K. J. Biochem. 55, 686 (1977). Verma: Tetrahedron 29, 2673 (1973). 61. Thompson, H. J., S. K. Sharma and S. A. 38. Kutney, J. P., A. K. Verma and R. N. Brown: Arch. Biochem. Biophys. 188, 272. Young: Tetrahedron 29, 2645 (1973).. 62. Sharma, S. K. and S. A. Brown: J. Chro- 39. Kutney, J. I?., A. K. Verma anh R. N. matogr. 157, 427 (1978). Young: Tetrahedron 29, 2661 (1973). 40. Murray, R. D. H.: Progr. Chem. Org. Nat. Address: Prof. S. A. Brown, Prod. 35, 199 (1978). Department of Chemistry, 41. Games, D. E. and D. H. James: Phytoche- Trent University, mistry 11, 868 (1972). Peterborough, Ont. (Canada) K9J 7B8