Phytochemistry, Vo\. 31, No. 11, pp. 40154018,1992 0031-9422/92 $5.00 + 0.00 Printed in Great Britain. (!'J 1992 Pergamon Press Lld (±)-DIONCOPHYLLACINE A, A NAPHTHYLISOQUINOLINE ALKALOID WITH A 4-METHOXY SUBSTITUENT FROM THE LEAVES OF TRIPHYOPHYLLUM PELTATUM* GERHARD BRINGMANNt, THOMAS ORTMANN, RAINER ZAGST, BERND ScH6NER, LAURENT AKE ASSIt and CHRISTIAN BURSCHKA§ Institute of Organic Chemistry, University of Wiirzburg, Am Hubland, 0-8700 Wiirzburg, Germany; ~Centre National de Floristique (Conservatoire et Jardin Botaniques), Universite d'Abidjan, 22 b.p. 582 Abidjan 22, Ivory Coast; §Institute of Inorganic Chemistry, University of Wiirzburg, Am Hubland, D-8700 Wiirzburg, Germany (Received 31 January 1992) Key Word Index-Triphyophyl/um peltaturn; Dioncophyllaceae; leaves; (± )-dioncophyllacine A; naphthylisoquino­ line alkaloids; biaryls, naturally occurring; structure elucidation. Abstract-The isolation and structure elucidation of rac-dioncophyllacine A from the leaves of Triphyophyllwn peltatum, is described. Unlike all other naphthylisoquinoline alkaloids, this fully dehydrogenated representative has an additional methoxy group at C-4, the position of which is deduced from NOE results. Dioncophyllacine A has a 7,1'­ site of the biaryl axis, as in dioncophylline A. Its constitution is confirmed by an X-ray structure analysis, which shows that the crystalline form of this new alkaloid is racemic. INTRODUCTION RESULTS AND DISCUSSION The genus Triphyophyllum with its only species The leaves of T. peltatum were extracted with di­ T. peltatum Airy Shaw belongs to the family of the chloromethane. The early fractions obtained on CC of the Dioncophyllaceae, a very small group of African medi­ extract yielded a nitrogen-containing crystalline solid cinal plants [2], the taxonomical position of which in the (130 mg). IH NMR spectroscopy reveals a strong struc­ plant kingdom is not yet completely established [3]. An tural relationship with dioncophylline A (la) for the eventual chemotaxonomical connection may be deduced resonances and multiplicity patterns of the protons in the from the occurrence of the acetogenic [4] naphthoquin­ naphthalene part. By contrast, the isoquinoline moiety one plumbagin [5], which also occurs i.a. in Droseraceae, shows distinctly different signals: the lack of the reson­ Plumbaginaceae, Nepenthaceae, Ebenaceae [6], and fi­ ances for protons at C-l, C-3 and C-4, as present in la, as nally in the Ancistrocladaceae [7, 8]. A distinctly more well as the chemical shifts of the signals of Me-l and Me-3 specific relationship can be recognized, especially to this ((i3.04 and 2.64, resp.), which moreover appear as singlet~, latter family, by the joint occurrence of naphthylisoquin­ clearly hint at a fully dehydrogenated isoquinoline oline alkaloids; structurally intriguing natural biaryls of moiety. And indeed, the I H NMR spectrum of the new apparently acetogenic origin [9, 10]. After early work on alkaloid is largely superimposable to that of 'O-methyl­ the chemical constituents of T. peltatum [5, 11-13], we tetradehydro-triphyophylline' (6), which was likewise re­ have been able to isolate and elucidate the structures of ported to occur in T. peltatum [12]. The constitution of the alkaloids dioncophylline A (la) [14-18], dioncopel­ this naphthylisoquinoline was confirmed by our first total tine A (2a) [19], dioncolactone A (3) [19], and di­ synthesis [22]. Yet, despite the synthetic availability of oncophylline B (4) [20] from the root bark of this plant. this compound as a reference substance, we could not Despite their structural differences, some joint features detect it in our plant material of T. pe/tatum, so far. The are common to all ofthese four alkaloids, e.g. the identical only distinct difference in the I H NMR spectra of 6 and tetrahydroisoquinoline part, with its R-configured stereo­ the new alkaloid is the presence of an additional methoxy centre at C-3. In this paper, we describe the isolation and group, obviously in an aromatic position, because one of structure elucidation of cUoncophyllacine A (5), which the aromatic protons (a singlet) found in 6 is missing in occurs as both atropenantiomers Sa and Sb. Some of the the new alkaloid. results described herein have recently been reported in a This is confirmed by the molecular formula, preliminary form [21]. C26H 27N04> as deduced from the mass spectrum (m/z =417 [M]+) and from combustion analysis. The exact position of this methoxy substituent can be determined by NOE spectroscopy (Fig. 1), revealing a close proximity ·Part 35 in the series 'Acetogenic lsoquinoline Alkaloids'. For of this additional functional group to Me-3 and H-5, Part 34 see ref. [1]. which clearly shows the methoxy group to be located at tAuthor to whom correspondence should be addressed. C-4. From these NOE experiments, the coupling position 4015 4016 G. BRINGMANN et al. o 20 3 5b in the isoquinoline part becomes evident: from the con­ compound, due to restricted rotation at the biaryl axis, tinuous array of the NOE-interacting protons-Me-3, whereas 3 and 4 have no stable configuration at the axis. OMe-4, H-5, H-6-the naphthalene substituent must be With respect to the fa<:t that the related alkaloids la and located at C-7, as in dioncophylline A (la). This is la have been found to be stereochemically homogeneous, underlined by the signal of the 8-methoxy group, whose i.e. not or not noticeably accompanied by the correspond­ aryl-induced highfieJd shift (<53.26 ppm) hints at a direct ing atropisomers Ib and 2b, it was interesting to investig­ proximity to the naphthalene substituent. And the posi­ ate whether the new alkaloid would have the same tion of this methoxy group, for its part, is demonstrated absolute configuration at the axis as la and 2a* and by its nuclear Overhauser interaction with Me-3. would thus be represented by structure 5a, or whether it As to the naphthalene moiety, the presence of three would have the opposite configuration or whether it neighbouring aromatic protons (H-6' to H-8') indicates would be a mixture of both isomers, e.g. the racemate. that the isoquinoline moiety must be connected with the Crystalline 5 is optically inactive ([iX]D 0°), i.e. a first, methyl-containing aromatic ring of the naphthalene. This though not conclusive, hint at the latter possibility (cf. the is underlined again by the highfield shift (<52.21) of this stereochemically homogeneous alkaloid ancistrotecto­ methyl group, which thus must be in ortho-position with rine, which is optically inactive! [23]). Furthermore, the respect to the biaryl axis. The normal chemical shift CD spectrum of 5 is nearly fiat, showing only very weak, (<54.02) for the methoxy group at C-4' excludes the non-characteristic peaks, possibly arising from impurities isoquinoline substituent from being located between or from a slight excess of one of the two enantiomers. OMe-4' and Me-2'. Hence, the new alkaloid is based on Hence, at least in the crystals investigated, the new the very same 7,1 '-coupling type as dioncophylline A (la) alkaloid seems to be a racemic mixture of both and can thus be identified as 4-methoxy-8-0-methyl- atropenantiomers 5a and 5b. 1,2,3,4-tetradehydro-dioncophylline A (5), subsequently This was unambiguously confirmed by an X-ray struc­ named dioncophyllacine A, the first 4-oxygenated naph­ ture analysis performed on crystals grown in MeOH, thylisoquinoline alkaloid ever isolated. which clearly shows both atropisomers to occur in the In contrast to 1-4, this new alkaloid 5 possesses no crystal (Fig. 2). Simultaneously, this crystal structure stereocentre at either C-l or C-3. Nonetheless, 5 is a chiral analysis fully confirms the constitution of this new alkal- A /,,2.64 ppm CH 3 3.04 ppm 3.98 3.26 ppm 4.02 ppm 2.21 ppm Fig. 1. Selected chemical shifts (A, b values) and nuclear Overhauser effects (B. in %) of dioncophyllacine A (5): evidence for the constitution. (±) Dioncophyllacine A from Triphyophyllum peltatum 4017 CH" e h"......... v--.,""~~A..I22 6 oxygen on the isoquinoline nucleus; a hint at early stages of an oxidative biotransformation of the alkaloids in the plant, possibly the beginning of a catabolic sequence, complementary to the structures of 2 and 3, where the methyl group of the naphthalene moiety has been oxi- e dized. 26 EXPERIMENTAL General. Mps: uncorr. Optical rotation: 25°,10 cm cell, CHCI3 (filtered through basic A1 20 3); IR: KBr; IH NMR (400 MHz): CDCI3 with TMS as into standard; ElMS: 70 eV. Analyses (C, H and N) were performed by the Institute of Inorganic Chemistry, University of Wiirzburg. CC: silica gel (60-200 mesh, Merck) by addition of 7.5% aq. NH3. TLC: precoated silica gel 60 F 2S. plates (Merck), deactivated with gas NH3• Spots were visualized under UV light and by Dragendortr's reagent. The X-ray Fig. 2. Schakal plot [24] of the structure of Satb in the crystal, as measurement was carried out by a Nonius CAD-4 diffrac­ determined by X-ray diffraction (., = inversion centre). tometer. Plant material. Leaves of T. peltatum were collected in West oid with its unprecedented additional methoxy group at Ivory Coast in July 1990 and identified by one of us (L.A.A.). A C-4. voucher specimen is deposited at the Conservatoire et Jardin It has to be noted that the racemic nature of the new Botaniques de I'Universite d'Abidjan, Republique de Cote alkaloid in the crystals investigated does not necessarily d'Ivoire. mean that Sa and 5b also form a 1 : 1 mixture in the plant. Extraction and isolation. Leaves (ca 1 kg) were extracted with It may well be that these two atropisomers occur in CH2CI 2 and 7% NH40H in a Soxhlet apparatus.