Biosynthesis of Furocoumarins: Further Studies on Ruta Graveolens

BIOSYNTHESIS OF FUROCOUMARINS 813

Biosynthesis of Furocoumarins: Further Studies on Ruta graveolens

F. DALL'ACQUA, A. CAPOZZI, S. MARCIANI, and G. CAPORALE

Institute of Pharmaceutical Chemistry of the University, Centro di Studio per la Chimica del Farmaco e dei Prodotti biologicamente attivi del Consiglio Nazionale delle Ricerche, Padua (Italy)

(Z. Naturforsch. 27 b, 813—817 [1972]; received March 20/April 24, 1972)

The Authors have studied the biosynthesis of some furocoumarins contained in Ruta graveolens. Labeled 5,7- and 7,8-dihydroxycoumarins have been administered to the herb, however practically no incorporation of radioactivity was observed in the isolated furocoumarins. Moreover, with the aim to study in a more detailed way the role of rutaretin and marmesin, trapping experiments have been performed administering to the herb labeled umbelliferone together with unlabeled rutaretin and marmesin respectively. A very effective incorporation was observed into marmesin, in lesser extent into rutaretin. In trapping experiments carried out using labeled rutaretin together with unlabeled marmesin and "vice versa", the results obtained clearly show that while marmesin was converted into rutaretin in a very effective way, rutaretin was transformed into marmesin only in a smaller extent. The Authors suggest the biogenetic origin of rutaretin and moreover formulate a new and more detailed aspect of the scheme for the biosynthetic pathway of furocoumarins in Ruta graveolens.

The biogenesis of furocoumarins has been studied (III); from this compound other naturally occur- in various plants [Ficus carica 2, Pimpinella ring furocoumarins, for instance bergapten (IV) magna3'4, Ammi majus, Angelica archangelica, and xanthotoxin (V) can derive by means of hy- Heracleum lanatum, Ruta graveolens 5-9, Pastinaca droxylation and O-alkylation reactions. sativa10]. The general scheme of the biosynthetic More recently we have found that not only hy- pathway appears to be well established: it involves droxylation and O-alkylation reactions, but also de- in a first stage the formation of a 7-hydroxycouma- methoxylation reactions can take place in plants: in rin-derivative (the biogenesis of coumarins has fact we have observed that a reciprocal conversion already been largely studied); successively the furan between psoralen and xanthotoxin can occur6; ring is formed on this compound through an iso- moreover, 4',5' - dihydropsoralen, 4',5' - dihydro- prenylation reaction leading to a 4',5'-dihydrofuro- bergapten and 4/,5'-dihydroxanthotoxin proved to coumarin-intermediate; finally this last is trans- be precursors not only of the corresponding furo- formed into furocoumarin-derivative. coumarins (psoralen, bergapten and xanthotoxin, It has been found that 7-hydroxycoumarin (I) respectively) but also of the other two, although (umbelliferone) is the most effective general pre- only to a small extent5. cursor at the first stage. Recently BROWN 7- 8 sug- Concerning Ruta graveolens, some doubt could gested that the only natural intermediate at the stage arise about the preceding biogenetic scheme, be- of 4',5'-dihydrofurocoumarin-derivative is marme- cause marmesin till now has not been found present sin (II), which is then transformed into psoralen in this plant, although feeding experiments have

OCH3

/ ykAA,

IV ovo v CH3 ^ II IH O'^V^O'^'O Fig. 1. General biosynthetic 0CH3 pathway proposed for psoralen V (III), bergapten(IV) and xanthotoxin (V).

Requests for reprints should be sent to Dr. F. DALL'ACQUA, Universitä di Padova, Istituto di Chimica Farmaceutica, Via Marzolo 5, 35100 Padova (Italy).

Dieses Werk wurde im Jahr 2013 vom Verlag Zeitschrift für Naturforschung This work has been digitalized and published in 2013 by Verlag Zeitschrift in Zusammenarbeit mit der Max-Planck-Gesellschaft zur Förderung der für Naturforschung in cooperation with the Max Planck Society for the Wissenschaften e.V. digitalisiert und unter folgender Lizenz veröffentlicht: Advancement of Science under a Creative Commons Attribution Creative Commons Namensnennung 4.0 Lizenz. 4.0 International License. 814 F. DALL'ACQUA, A. CAPOZZI, S. MARCIANI, AND G. CAPORALE demonstrated that it may be a valid precursor for experiment was comprehended between the range furocoumarins (psoralen, bergapten and xantho- 10 —12 g. This was placed into 100 ml beakers containing a mixture formed by adding to the aqueous toxin) and especially for psoralen. On the other solution of the labeled compound to be administered, hand, another 4 ,5 -dihydrofurocoumarin-derivative an equal volume of an "aqueous nutritious solution" 17. has been isolated from Ruta graveolens, that is ruta- The system was illuminated with 500 W Osram HWL retin (VI), which is an 8-hydroxyderivative of mar- lamps; "metabolism time" lasted 7 days for feeding experiments with umbelliferone and its hydroxy- mesin n. derivatives, while for trapping experiments it lasted 54 hours. During the metabolic period "aqueous nutritious solution" was added to maintain constant the level of the liquid layer in the beakers. 3 OH VI Trapping experiments Fig. 2. Molecular structure of rutaretin (VI). Two different types of trapping experiments have been performed A) and B) : Very recently we have found that rutaretin (VI) A) 2.4 mg of tritiated umbelliferone was solubilized is well incorporated specifically into xanthotoxin in 40 ml of distilled water and successively 16 mg of unlabeled rutaretin was also dissolved in the same and also into psoralen and bergapten, even to a solution and this was administered to the herb as indi- 6 lesser extent . It may therefore be the natural cated in the section "feeding procedure". Analogous 4',5'-dihydrofurocoumarin-intermediate, at least for experiment was performed using unlabeled marmesin the biosynthesis of xanthotoxin. (15 mg) together with tritiated umbelliferone (2.6 mg); The studies reported in this paper have been per- B) 3 mg of tritiated rutaretin together with 10 mg of unlabeled marmesin were solubilized in 40 ml of formed to obtain further experimental information distilled water and this solution was then administered on the role of rutaretin, on its biogenetic origin and to Ruta graveolens as reported in the "feeding proce- on the possibility that marmesin also may be a phy- dure". Parallel experiment was carried out using 3 mg siological intermediate, perhaps present only in a of tritiated marmesin and 10 mg of unlabeled rutaretin, operating in the same experimental conditions. very small amount. The results obtained feeding Ruta graveolens with Isolation and purification of furocoumarins and of 5,7-dihydroxycoumarin (VII) and with 7,8-dihy- rutaretin and marmesin droxycoumarin (VIII) (daphnetin), as well as per- Isolation and purification of psoralen (III), bergap- forming some trapping experiments, in which marme- ten (IV) and xanthotoxin (V) were performed following a procedure described elsewhere 6. sin and rutaretin were involved, allowed us to for- Rutaretin (VI) and marmesin (II) were isolated mulate a more detailed scheme for the biosynthetic and purified in a way similar to that described for pathway of psoralen and xanthotoxin in Ruta gra- furocoumarins with little change in the procedure for veolens. preparing the "coumarinic extract"1,2'18 as follows: the etheral coumarinic extract was washed three times with an equal volume of saturated sodium bicarbonate Materials and Methods aqueous solution. After this treatment, while marmesin and furocoumarins remained in the etheral solution 7-hydroxycoumarin or umbelliferone (I) 12, 5,7-di- (A), rutaretin passed into the aqueous bicarbonate hydroxycoumarin (VII) 13 and 7,8-dihydroxycoumarin phase; this last phase was acidified with 2n HCl and (VIII) 14 have been prepared in this Institute. then extracted with ether (B). From the two etheral Rutaretin (VI) 11 (extracted from Ruta graveo- 15 extracts (A) and (B) the solvent was distilled and the lens) and marmesin (II) (extracted from Aegle mar- residues were chromatographed as follows: melos, Correa) have kindly been provided by Prof. Dr. a) thin layer silica gel preparative chromatography G. SCHNEIDER and Prof. Dr. A. CHATTERJEE respectively. 3H labeling procedure of all these substances has (Merck cat. 5717) using as solvent of development ethyl acetate : cyclohexane (3 : 1, v/v) ; been carried out according to the WILZBACH method 16. Chemical and radiochemical purification of labeled b) thin layer silica gel analytical chromatography compounds has been performed following a procedure (Merck cat. 5715) : solvent of development ethyl ace- described elsewhere 2. The specific radioactivity of these tate : cyclohexane (2 : 1, v/v) : at this stage chemical compounds is reported in Tables I, II and III. and radiochemical purity of the substances were tested. Spectrophotometrical measurements and radiochemical Feeding procedure determinations Cut ends of Ruta graveolens herb were employed; The different ethanolic solutions containing the generally the dry weight of this material for every various isolated compounds were examined spectro- BIOSYNTHESIS OF FUROCOUMARINS 815

photometrically both recording their u.v. spectra using It is well known that the biogenetic route starting a 124 double beam model Perkin-Elmer instrument and from umbelliferone and leading to furocoumarins successively determining their absorbance at X max involves 4',5'-dihydrointermediates; in the case of by means of a CF 4 Optica single beam spectrophotometer. Radiochemical measurements were per- Ruta graveolens recently it has been shown 6- 8 that, formed by a liquid scintillator spectrophotometer at this stage, marmesin (II) and rutaretin (VI) are model 3375 Packard using a dioxane base scintillator: involved. However while rutaretin is present in a naphthalene g 120, 2,5-diphenyl-oxazole g 4, 2,2'-p- valuable extent in this herb11 marmesin has not phenylen bis (5-phenyl-oxazole) g 0.075 in dioxane been till now identified. up to 1000 ml of solution. The apparatus efficiency in experimental conditions used was comprehended within Therefore with the aim: 1) to investigate if mar- the range 25 — 37% for tritium. mesin really is a physiological intermediate in Ruta graveolens, deriving from umbelliferone and 2) to Results and Discussion obtain further informations on the role of rutaretin and on its biogenetic origin, we have performed the The possible role of precursors of two dihydroxy- following trapping experiment (A) : we have ad- coumarins, namely 5,7-(VII) and 7,8-dihydroxy- ministered to Ruta graveolens a) umbelliferone-3H coumarin (VIII), in the biosynthesis of furocouma- together with unlabeled marmesin and b) umbelli- rins contained in Ruta graveolens was tested by ferone-3H together with unlabeled rutaretin. After feeding experiments. the metabolism time, marmesin and respectively rutaretin were isolated and their acquired radioactivity determined. The results obtained are reported in Table II. T o OH Recovered marmesin has acquired very much ef- vn vm fectively a good amount of radioactivity; isolated Fig. 3. Molecular structure of 5,7-(VII) and 7,8-dihydroxy- rutaretin is also radioactive, but in a lower degree coumarin (VIII). than that of marmesin. The data are in a good agreement with the hypothesis indicating that mar- The two 3H-labeled dihydroxycoumarins, as well mesin really derives from umbelliferone through an as umbelliferone-3H for comparison, were ad- isoprenylation reaction6,8; also rutaretin seems ministered to Ruta graveolens; after the metabolism derive from umbelliferone, however this trans- time, the three most important furocoumarins (pso- formation occurs with a smaller effectiveness. There- ralen, bergapten and xanthotoxin), were isolated fore the results obtained in these experiments could and their acquired radioactivity was determined. indicate that while marmesin derives directly from The results obtained are reported in Table I; they umbelliferone (isoprenylation reaction only is in- clearly show that only umbelliferone (I) was well volved), rutaretin should originate from the same incorporated in all furocoumarins. The very small umbelliferone, probably involving marmesin through value of incorporation of 5,7-dihydroxycoumarin an hydroxylation reaction. (VII) (possible precursor of bergapten) and of 7,8-dihydroxycoumarin (VIII) (possible precursor With the aim to verify this hypothesis we have for xanthotoxin) demonstrate that these substances performed a second trapping experiment (B), ad- have no importance as precursors of furocoumarins. ministering to the herb: a) marmesin-3H together

administered Specific Psoralen Bergapten Xanthotoxin compound activity Specific dilution51 Specific dilutiona Specific dilution8, [dpm/^moles activity activity activity X106] [dpm/^moles] [dpm/^umoles [dpm/«moles]

7-hydroxyeoumarin 11.24 81,500 138 24,900 451 7,190 1,562 5.7-dihydroxycoumarin 3.62 106 34,200 320 11,600 149 24,300 7.8-dihydroxycoumarin 5.42 320 16,900 417 12,997 166 32,650

Table I. Conversion of oxygenated coumarins into furocoumarins. a Dilution is defined as the ratio between the specific activity of compound administered and that of isolated furocoumarins. 816 F. DALL'ACQUA, A. CAPOZZI, S. MARCIANI, AND G. CAPORALE

Labeled compound administered Trapping compounds administered and recovered Marmesin Rutaretin Specific activity Specific activity dilution3. Specific activity dilution3 [dpm/^moles] [dpm/^moles] [dpm/^moles]

Umbelliferone 11.24 XlO6 3.188x105 35 1.289 X105 87

Table II. Results obtained from trapping experiment A). a Dilution is defined as the ratio between the specific activity of compound administered and that of isolated substances.

Labeled compound administered Trapping compounds administered and recovered Marmesin Rutaretin Specific activity Specific activity dilution3 Specific activity dilution3 [dpm/^moles] [dpm/^moles] [dpm/^moles]

Rutaretin 16.15X106 3.6 XlO4 449 Marmesin 5.00 XlO6 4.82 x 105 10

Table III. Results obtained from trapping experiment B). a Dilution is defined as the ratio between the specific activity of compound administered and that of isolated substances.

with unlabeled rutaretin and b) rutaretin-3H together CH3 HO-^C II I] HO^OCX) O^O with unlabeled marmesin. CH3 After the metabolism time rutaretin (a) and marmesin (b) were isolated and counted for their radioactivity. The results are reported in Table III. We can see the very high radioactivity acquired by rutaretin in the experiment a) indicating a very CH3 much efficient and rapid transformation of marme- HO^C sin into rutaretin. By contrast experiment b) shows CH3 OCH3 that rutaretin is transformed into marmesin only

in a limited extent. Fig. 4. Proposed biosynthetic pathway for psoralen (III) and The data obtained from the trapping experiment xanthotoxin (V) in the case of Ruta graveolens. (B) strongly support the above formulated hypo- In fact umbelliferone proved to be for Ruta gra- thesis indicating that rutaretin from a biogenetic veolens also the whole physiological precursor; it point of view, derives from a hydroxylation reaction cannot undergo reactions of hydroxylation or on marmesin. O-methylation (in fact 5,7-dihydroxycoumarin, Conclusion 7,8-dihydroxycoumarin, 5-methoxy-7-hydroxycoumarin and 7-hydroxy-8-methoxycoumarin proved The results now obtained, together with other data not to be precursors of furocoumarins); therefore acquired in previous studies on the biosynthesis of the — OH group present in the — 8 — position of furocoumarins contained in Ruta graveolens, allow rutaretin must be introduced only after the dihydro- the formulation of the following biogenetic scheme furan ring had formed. From this derives the essen- for two furocoumarins contained in this herb, tial role of marmesin as intermediate. psoralen and xanthotoxin. The fact that marmesin until now has not been This scheme is analogous to that proposed by found present in Ruta graveolens would be justified BROWN (see above), modified for applying it to the by its rapid transformation, so that an accumulation situation of Ruta graveolens. In this herb in fact an in detectable amounts would not be possible. This important role as 4',5'-dihydrofurocoumarin-inter- hypothesis is supported by the very efficient trans- mediate is played by rutaretin, while marmesin ap- formation into rutaretin now observed. Moreover pears to be a very rapidly transient, although essen- we point out that, in feeding experiments performed tial intermediate. always in Ruta graveolens, marmesin proved to be BIOSYNTHESIS OF FUROCOUMARINS 817 an efficient precursor specifically for psoralen; we could take place on marmesin, giving a 5-hydroxy- must conclude therefore that marmesin, may be intermediate, transformed then into bergapten. No transformed not only into psoralen, but also in an experimental support, however, exists for the last efficacious way into rutaretin. These two combined hypothesis. metabolic transformations could justify a very short Very recently we have determined the amount of life of this intermediate. Concerning bergapten, no psoralen, bergapten and xanthotoxin normally pre- detailed information is available on its biogenetic sent in Ruta graveolens, cultured in a garden near pathway in Ruta graveolens. Although 5-methoxy-7- Padua 19. The ratio between the amounts of the three hydroxycoumarin proved to be a good precursor for furocoumarins was as follows: xanthotoxin 10: bergapten in Pastinaca sativa10 and in Ficus bergapten 5: psoralen 1. The small amount of carica2, the same compound 6, as well as 5,7-dihy- psoralen in respect to that of xanthotoxin is in good droxycoumarin, has been ineffective in Ruta gra- agreement with the observed rapid transformation of veolens. For bergapten also, therefore, umbelliferone marmesin into rutaretin, providing therefore a is the whole precursor at 7-hydroxycoumarin stage. further support for the above proposed biogenetic The successive steps have not been clarified; it may scheme. be formed by methoxylation of psoralen according to the hypothesis of BROWN, supported by the fact We are indebted to Prof. GIOVANNI RODIGHIERO that a transformation of psoralen into bergapten has and Prof. LUIGI MUSAJO for helpful discussion on this been ascertained; alternatively, a 5-hydroxylation research.

1 G. CAPORALE, A. BRECCIA, and G. RODIGHIERO, Prepn. 9 W. STECK and S. A. BROWN, Canad. J. Biochem. 48, 872 Bio-Med. Appl. Labeled Mol., Proc. Sympos., pag. 103, [1970]. Venice 1964. 10 S. A. BROWN, Phytochem. 9, 2471 [1970]. 2 G. CAPORALE, F. DALL'ACQUA, S. MARCIANI, and A. CA- 11 G. SCHNEIDER, H. MÜLLER, and P. PFAENDER, Arch. Phar- POZZI, Z. Naturforsch. 25 b, 700 [1970]. maz. 300, 73 [1967]. 3 H. G. FLOSS and U. MOTHES, Phytochem. 5, 161 [1966]. 12 H. V. PECHMANN, Chem. Ber. 17, 929 [1884]. 4 H. G. FLOSS and H. PAIKERT, Phytochem. 8, 589 [1969]. 13 R. G. HEYES and A. ROBERTSON, J. diem. Soc. [London] 5 G. CAPORALE, F. DALL'ACQUA, S. MARCIANI MAGNO, and 1936, 1831. A. CAPOZZI, Herba Hungarica 1971, in press. 14 H. V. PECHMANN, Chem. Ber. 17, 934 [1884]. 6 G. CAPORALE, F. DALL'ACQUA, A. CAPOZZI, S. MARCIANI, 15 A. CHATTERJEE and S. S. MITRA, J. Amer. chem. Soc. 71, and R. CROCCO, Z. Naturforsch. 26 b, 1256 [1971]. 606 [1949]. 7 W. STECK, M. EL-DAKHAKHNY, and S. A. BROWN, Tetra- 16 K. E. WILZBACH, J. Amer. chem. Soc. 79, 1013 [1957]. hedron Letters [London] 54, 4805 [1969]. 17 E. J. HEWITT, Plant Physiology, vol. 3, pag. 101, Academic 8 S. A. BROWN, M. EL-DAKHAKHNY, and W. STECK, Canad. Press, New York 1963. J. Biochem. 48, 863 [1970]. 18 E. SPÄTH, Chem. Ber. 70 A, 83 [1937]. 19 G. CAPORALE, unpublished data.