BIOLOGICALL Y ACTIVE AGENTS IN NATURE 18

CHIMIA 52 (1998) Nr. 1/2 (Jannar/Fcb"mr)

Chimia 52 (1998) 18-28 methods. The first alkaloid to be investi- © Neue Sclnveizerische Chemische Gesellschaft gated was dioncophylline A [18][19], the ISSN 0009-4293 main secondary metabolite ofT. peltatum. It is easily accessible by standard isolation procedures, and was found to have a naph- The Alkaloids of thylisoquinoline basic structure mainly by NMR investigations (Fig. 4). It has three peltatum () [1] stereogenic units: two stereocenters (C( I), C(3»and astereogenic axis (C(7) ~C(I'». Its structural elucidation thus requires at Gerhard Bringrnanna)*, Guido Fran~oisb), Laurent Ake Assn, and least three pieces of stereochemical infor- U Jan Schlauer ) mation.

Dedicated to Professor Meinhart H. Zenk, University of Munich, on the occasion of his 2.1. The Relative Configuration at the 65th birthday. Stereocenters: through NMR From an NOE interaction of the axial proton at C(3) (Fig. 5) with the likewise Abstract. A great diversity of naphthylisoquinolines, presumably acetogenic biaryl axial Me group at C( 1), a relative cis-array alkaloids, has been obtained from the West-African Triphyophyllumpeltatum. By of these two spin systems and thus trans- the example of dioncophylline A, the main alkaloid from T. peltatum, we illustrate the orientation of the two Me groups was various methods of structural elucidation established in our laboratory. The structural deduced, establishing an (R,R)- (or an peculiarities of some of the minor alkaloids from the same species are discussed. A (5,5)-!) configuration [18] - the first re- perspective on the interesting biological activities of various alkaloids from T.peltatum quired stereochemical information. and related is given. The most promising lead is dioncophylline C, marked by pronounced antiplasmodial activity. Chemical syntheses, especially dimerization of 2.2. The Absolute Configuration at naphthylisoquinolines, and QSAR-guided modifications of the most active structures C(3) (and thus C(1)): through Total are featured as rewarding strategies in the search for improved drugs. A chemotaxo- Synthesis nomic summary is presented. The absolute configuration at C(3) was established by a stereochemically unam- 1. Introduction precursors, named korupensamines, into biguous total synthesis of the alkaloid micheUamines [13]. from alanine (Scheme 1) [19][20]. For the Dimeric naphthylisoquinoline alka- This paper will deal mainly with such regio- and stereoselective construction of loids, e.g. michellamine B (1, Fig. ]), are monomers, especially those from Tri- the biaryl axis, the 'lactone method' was most promising novel antiviral compounds phyophyllum peltatum, since they like- applied, an efficient synthetic procedure from African plants [2-4]. They occur in wise display intriguing structural, biosyn- developed in our group [21][22]: Prefix- one single species, koru- thetic, and pharmacological (in particular ation of the two molecular moieties via an pens is, which is endemic to a small region antimalarial) properties. T. peltatum is the ester bridge «(f 2) and subsequent in- at the Cameroonian/Nigerian border and most widespread species of the Dionco- tramolecularC,C-bond formation gave the has been discovered within an anti-HIV phyllaceae, a small family consist- lactone-bridged biaryl 3, which is config- screening program initiated by the US ing of only three species oflianas climbing uratively unstable - it rapidly isomerizes National Cancer Institute [5][6]. In collab- up high trees, by means of hooked at the axis. But out of this equilibrium of oration with Dr. Boyd and his NCI group, (cf Fig. ]9) [14]. Unique are the large thin helimericforms3aand3b, itcan be cleaved we have elucidated the stereostructure of seeds with large membranous wings (Fig. such that optionally either ofthe atropiso- michellamine B and related compounds 2), which during maturation exceed the [2][6-8] and have achieved first total syn- size ofthe fruits they are borne from [15]. theses of these challenging and attractive From these seeds, we have succeeded in target molecules [9-12]. One of the latest cultivating juvenile plants of T. peltatum achievements of our group is the detection (Fig. 3). At the end of the juvenile growth and characterization of dimerization en- period, just before it starts climbing, T. zymes which transform the monomeric peltatum forms leaves with stalked glands and reduced lamina that capture and digest invertebrates [16]; it is a 'part-time' car- *Correspondence: Prof. Dr. G. Bringmann nivorous plant [17]. Phytochemically, T. a) Institut fUrOrganische Chemie der Universitat peltatum proved to be a rich source of Am Hubland D-97074 Wi.irzburg novel alkaloids. Tel.: +49931 8885323 Fax: +49 931 8884755 E-Mail: [email protected] 2. Dioncophylline A as a Probe Alka- b) Prins Leopold Instituut voor Tropische loid for the Analytical Methodology Geneeskunde Nationalestraat 155 For the rapid and unambiguous attri- B-2000 Antwerpen o configuratively unstable C) Centre National de Floristique bution of the full stereostructures of such Universite d' Abidjan compounds, we have established and fur- Fig. I. Michellamille n, a dimeric naphthyliso- 22 B.P. 582 Abidjan 22, Ivory Coast ther developed a broad series of analytical quinoline with high anti-HIV activity BIOLOGICALL Y ACTIVE AGENTS IN NATURE 19

CHIMIA 52 (1998) Nr. 1/2 (Jan"ar/Fehn,"r)

Fig. 2. Seeds of Triphyophyllum peltatum (Dioncophyllaceae), showing Fig. 3. A juvenile plant ofT. peltatum, several weeks after germination the large circular wing meric products is obtained in high selec- possible conditions to avoid isomeriza- tivity [20] - a stereochemically efficient tion at the axis. From the positive first and mechanistically remarkable reaction. Cotton effect in the experimental CD spec- One of the products ultimately obtained is trum of 5, a likewise 'positive chirality' fully identical with natural dioncophylline was deduced, thus establishing P-config- A, whereas the other one is diastereomeric uration at the biaryl axis. - when starting from D-alanine, while L- alanine leads to the wrong enantiomeric 2.4. Further Support for the Axial series [19]. This shows dioncophylline A Configuration: Quantum Chemical CD to have (R)-configuration at C(3) (and Calculations thus also at C(l)) - the second stereoinfor- By these three pieces of stereoin- mation achieved. formation, the full absolute stereostruc- ture 4 of dioncophylline A as shown in 2.3. The Absolute Axial Configuration: Scheme 2 had become evident. Possibly Fig. 4. Constitution ofdioncophylline A by select- by Experimental CD Spectroscopy the same compound, but named 'triphyo- ed HMBC and NOE interactions and chemical While we had selectively produced phylline', had previously been isolated by shifts dioncophylline A and its stereoisomers, a French group [24][25], who had correct- its axial configuration was still unclear at ly assigned the constitution and the rela- ~H~N~Me this point. This missing third stereoinfor- tive configuration at C(l) and C(3), but 1 H 3 mation was acquired by CD spectroscopy, had erroneously assumed the absolute con- Me H for which the molecule first had to be figuration at the stereocenters to be (5,5), modified by dehydrogenation to give 5 so while the chirality of the biaryl axis had that, by enlargement of the isoquinoline been disregarded. Additional severe in- u chromophor, the Exciton Chirality meth- consistencies in the literature [24-26] ne- ~ trans od [23] became applicable (Scheme 2). cessitated a new naming for 4, subse- Fig. 5. Relative configuration at C( J) VS. C(2) of This had to be done under the mildest quently termed dioncophylline A [18]. dioncophylline A by NOE BIOLOGICALLY ACTIVE AGENTS IN NATURE 20 CHIMIA 52 (1998) Nr. 1/2 (JonunrlFebl1lnr)

Scheme 1. Total Synthesis Establishing the Absolute Configuration at C(3) (and C(1))

e R 04- H02C~Me I • 04- Br:::"" R N .... ;::,... I R RN.... : Bn "\ MeO M ~(- Bn NH2 Wo Me HO Me 2 D-Ala o MeO CI 0

Me Me

MeO Me MeO Me (rapid) 3a o - 7-epi- atropisomer- Dioncophylline A selective 1~fast cleavage ~ Me Me MeO - Me Me 3b MeO o Bn = CH Ph 2 (slow) Dioncophylline A m.p.214 o,[a]D-14.9 natural m.p.215°,[a]D-14

Scheme 2. Absolute Configuration at the Axis by Scheme 3. An Atropisomer-Dif.ferentiating Reaction for the Chemical Analysis of the Relative Exciton Chirality CD Configuration at the Axis

7-epi-6 (Dioncophylline A, full absolute stereostructure) 7-epi-4 4 PdlC ~ toluene, 1200

Me

MeO

MeO

Forthis reason, we kept looking for further further developed in our group [27-29]. confirmation of our revised structure 4. As shown in Fig. 6, the theoretical CD Therefore, any new method that we in- spectrum (dotted) for the dehydrogena- troduced into naphthylisoquinoline chem- tion product of dioncophylline A matches istry was first tested on this important very well with the experimental one (full alkaloid, exemplarily. line), in particular in the region of the For a first additional support of the crucial couplet. This unambiguously con- absolute stereostructure, we have applied firms the absolute axial P-configuration the quantum chemical calculation and thus of dioncophylline A [30] - stereochemi- prediction of CD spectra, established and cal information no. 4. BIOLOGICALLY ACTIVE AGENTS IN NATURE 21

CHI MIA 52 (1998) Nr, 1/2 (Jnnunr/Fcbruor)

2.5. Relative Axial vs. Centrochirality: 200 positive first a Chemical Method / Cotton effect Furthermore, we have developed a 120 chemical method for the analysis of the .. relative configuration at the biaryl axis as 40 compared to the stereocenters [31]. Ofthe L1£ -40 two atropo-diastereomeric synthetic pre- -- experimental cursors 6 and 7 to dioncophylline A -epi-6 -120 calculated (4) and its atropisomer (7-epi-4) (Scheme A 3), specifically only the syn-compound 7- -200 epi-6 can be bridged by a succinate unit to 200 230 260 290 320 give 7, whereas the anti-diastereomer 6 Fig. 6. Absolute axial configuration through does not give a similar ansa-compound. quantum-chemical CD calculation This strictly atropisomer-differentiating MeO reaction (here stereochemical information no. 5) again confirms the relative axial MeO configuration attributed above. MeO Me Me H

2.6. Relative Axial vs. Central Chirali- MeO ty: by Long-Range NOE Interaction Fig. 8. X-ray structure analysis: constitution Even without any chemical modifica- and full relative configuration tion, such a syn/anti-relationship can be analyzed by NMR: Despite the long dis- tances between the diagnostically deci- Dioncophylline A (4) sive nuclei, the crucial long-range NOE interactions that specifically confirm the configuration at the axis relative to the centers can be observed (Fig. 7)- the sixth independent stereochemical information MeO [32].

MeO 2.7. Absolute Configuration at Both Stereocenters: by Chemical Degrada- tion 7-epi-Dioncophylline A 8 For an additional proof of the absolute (7-epi-4) configuration at the stereocenters, we have Fig. 9. Full absolute stereostructure through developed a ruthenium-mediated oxida- Fig. 7. Long-range NOE: relative axial YS. cen- anomalous X-ray diffraction of a bromo deriva- tive degradation reaction, which neatly tral chirality tive cuts out the centers by transforming them into 3-aminobutyric acid (ABA) and ala- Scheme 4. Oxidative Degradation: Absolute Configurations at C(l) and C(3) nine (Ala) [33] (Scheme 4). From the (R)- configuration of the two amino acids, di- Me ~Me oncophylline A is clearly confirmed to be H02C I NH2 (R)-configurated at both stereocenters pro- RuCI3 viding stereochemical informations no. 7 ---'''- + (R)-ABA and 8. This useful reaction can be per- Me MeO formed down to submilligram quantities. It is also applicable to N-methylated and MeO - Me N,N-dimethylated cationic analogs, and Dioncophylline A (4) D-Ala even to f3-carboline alkaloids [34][35]. / We perform it routinely for several other research groups [36-38]. I GC-MSD analysis I 2.8. Constitution and All Relative Configurations: by a 'Normal' Crystal- Structure Analysis It was only after all the aforementioned figurations of dioncophylline A and thus dibenzyldioncophylline A (8) allowed a stereochemical investigations that we hap- provides two further pieces of stereochem- confirmation of the full absolute stereo- pened to obtain crystals suited for an x- ical information. structure of dioncophylline A (here with ray structure analysis [39] - but only from benzyl abbreviated as R) through anoma- CH2CI2, which turned out to be an integral 2.9. The Full Absolute Stereostructure: lous X-ray diffraction [40]. This corre- part of the crystal (Fig. 8). This crystal- by Anomalous X-Ray Diffraction sponds to three pieces of stereochemical structure analysis fully confirms the con- More recently, another crystal-struc- information in one experiment. stitution and all of the relative con- ture analysis (Fig. 9) on 5-bromo-O,N- BIOLOGICALL Y ACTIVE AGENTS IN NATURE 22

CHIMIA 52 (1998) Nr. 1/2 (JanllarIFcbruar)

Scheme 5. Structures and Synthesis of Three Further, Differently Oxygenated Alkaloids ofT. peltatum

Me - - HO

o A.maxem = 442 nm 5'-o.Oemethyl- Oioncopeltine A (10) Oioncolactone A (11) dioncophylline A (9) = tpdll

~Me

Br~N'Bn o Me

o 12

Itotal synthesis I

5'- o.Oemethyl-8-o.methyl- dioncophylline A (13) Dioncophyllacine A (16) Dioncophylline B (17)

OMe Me

5'· o.Oemethyl-8- o.methyl- 7-epi-dioncophylline A (14)

Oioncophyllacine B (18)

Fig. 12. 7,6 '-coupled alkaloids, only knownfrom T. peltatum

tine A (10) is the as yet only known 5'-0. Oemethyl-8- o.methyl- naphthylisoquinoline with an additional 7-epi-dioncophyllidine A (15) oxygen function at the methyl group of the naphthalene moiety; its structure was con- Fig. 10. Further minor 'A-type' alkaloids.from Fig. 11. Dioncophyllacine A: fully dehydroge- firmed by X-ray diffraction analysis. Still both atropisomeric series nated and axially chiral- but racemic further oxidized is the strongly fluores- cent alkaloid dioncolactone A (11), which 2.10. Conclusive Remark 3. Further Related 'A-Type' (i.e., 7,1'- is closely related to the synthetic lactone- All of the methods described above Coupled) Alkaloids of T. pellalum bridged biaryl precursor 3 to dioncophyl- full y confirmed the assignment of the com- line A (cf Scheme 1). It was thus an easy plete absolute stereostructure of dionco- Three such new alkaloids 9-11 with a and rewarding task to synthesize all of phylline A as 4. This broad analytical 7,1 '-linkage of the aryl moieties ('A-type') these three compounds at a blow - despite arsenal was now available to be applied to like dioncophylline A (4), but with a free, their additional free C(5')-OH group - by the numerous other alkaloids Iikewise pro- not O-methylated OH function at C(5') intramolecular biaryl coupling of the es- duced by T. peltatum. [41][42] are seen in Scheme 5. Dioncopel- ter-type prefixed precursor 12 and subse- BIOLOGICALLY ACTIVE AGENTS IN NATURE 23

CHIMIA 52 (1998) Nr. 1/2 (Januar/Fcbruar)

Scheme 6. Dioncophylline C, the Only 5,1 '-Coupled Alkaloidfrom T. peltatum have a configuratively stable biaryl axis either, it is the only fully achiral naphthyl- isoquinoline alkaloid.

HO OMe MeO O~ MeO o~ 4.2. A fe-Type' (i.e., 5,1'-Coupled) Alkaloid oM There is only one 5,I'-coupled naph- thylisoquinolinefrom T. peltatum: Dionco- Me -- Me -~Pdll 0 Me phylline C (19) [47] (Scheme 6). It is also RRN, :;.-'I the only one without an oxygen function - Bn next to the biaryl axis. Nonetheless, we HO W yO Me have synthesized the molecule via our Dioncophylline C (19) 'lactone procedure', by intramolecular 20 I 21 coupling of the ester 21 to a helimeric (main atropisomer) bridged biaryl 20 [2][48], ring opening, and eventual reductive elimination of the useful intermediate 'bridgehead' oxygen function. Me course disposes of axial chirality [44]. Still, itis not optically active, since itis the 4.3. The 'D-Type' (i.e., 7,8'-Coupled) first naphthylisoquinoline alkaloid that is Alkaloids MeO racemic! As seen by an X-ray structure We have recently detected a fourth analysis, both atropo-enantiomers are coupling type, which is based on a 7,S'- present also in the crystal. linkage, as found in the three new alka- loids 22-24 (Fig. 13) [2][49]. Oneofthem, Dioncophylline D (22) dioncophyllinol D (24), is the first alka- 4. Alkaloids with Other Coupling o configuratively unstable (rapid rotation) loid of this type to have a free OH group at Positions C(4) - and is thus the first monomeric naphthylisoquinoline to have a maximum Me 4.1. 'B-Type' (i.e., 7,6'-Coupled) number of stereocenters. By D-NMR, 24 Alkaloids was shown to be a mixture of two rapidly A representative of the less frequently interconverting atropo-diastereomers. This occurring 7,6'-coupling type, as found in isomerization can be 'frozen' at ca. J 90 K dioncophylline B (17) [45], is seen in Fig. to give two sets of signals for the two 12. Here the coupling position allows a atropo-diastereomeric species. 8-O-Methyldioncophylline D (23) free rotation around the axis, so that there is no stable axial configuration, only the 4.4. Triphyophyllum Alkaloids - a stereocenters C(l) and C(3). Similarto 16, Broad Series of Biaryl Alkaloid Struc- dioncophyllacine B (18) has an additional tures OMe group at C( 4) and a fully dehydroge- Thus, T. peltatum from Ivory Coast is nated isoquinoline moiety and thus has no a rich source of a plethora of meanwhile stereocenters [2][46]. Since it does not nearly 20 different naphthylisoquinoline

Dioncophyllinol D (24)

Fig. 13. Dioncophyllinol D (24) and two further 7,8 '-coupled alkaloids 22 and 23 HO OMe

quent stereoselective ring opemng and deoxygenation [42]. Fig. 10 shows three further related' A- type' alkaloids, 13-15, again all of them with an 'OH-OMe' -substitution pattern HO Me on the naphthalene, including, for the first time, representatives 14 and 15 of the Dioncophylline C (19) Dioncopeltine A (10) other atropo-diastereomeric series, and the IGSO = 0.014 llg/ml IGso = 0.021 llg/ml first dihydroisoquinoline 15 within this (0.015 llg/ml) (0.038 llg/ml) class of compounds [2][43]. Even further dehydrogenated and with an unprecedented additional OMe func- tion at C(4) is dioncophyllacine A (16, Fig. 14. Antimalarial activities ofnaphthylisoquinolines against P. falciparulll (and P. berghei) in Fig. 11). It has no stereocenters, but of vitro: the 'top two' BIOLOGICALLY ACTIVE AGENTS IN NATURE 24 CHIMIA 52 (1998) Nr. 1/2 (JnnunrIFcbruar)

alkaloids. Most of them have the normal 5.2. Antimalarial Activity of Dionco- 7,1 '-coupling type as realized in dionco- phylline C and other Naphthylisoquin- phylline A, but there are also the less olines frequent types B, C, and D. All of these Of the greatest pharmacological sig- structures are unique and intriguing, nice nificance is the distinct anti plasmodial textbook examples of compounds with activity against Plasmodium falciparum, both stereocenters and -axes, with atrop- the pathogenic agent of the pernicious N-Methyl- isomerization barriers ranging from un- malaria tropica [55][56]. Already the ex- dioncophylline A (25) measurably low for bridged or low-substi- tracts of T. peltatum exhibit very good tuted species, via rotations that can be activities, also against chloroquine-resis- (lGso = 13.6 Jlg/ml) frozen at low temperatures, up to atropiso- tant strains. Among the pure isolated com- mers that are stable to above 150°. pounds tested, in particulardioncophylline C (19) and dioncopeltine A (10, Fig. 14) show excellentICso values [57]. Likewise 5. Naphthylisoquinolines as Promising good in vitro activity was found against Leads in Pharmacology and Plant the related parasite, Plasmodium berghei. Protection First in vivo experiments on the cura- tive potential of naphthylisoquinoline al- The high diversity in the chemical struc- kaloids in this rodent system showed P. ture of naphthylisoquinolines is paralleled berghei-infected OFI mice to be healed N,N-Dimethyl- by their various biological activities, by treatment with dioncophylline C, their dioncophyllinium A Iodide (26) among them activity against plant-patho- parasitaemia was reduced to zero from genic fungi [2][50] and antifeedant and day four on. Without any noticeable side (lC 0.63 Jlg/ml) so = growth-retarding activity against herbivo- effects, the animals survived for months, rous insects like Spodoptera littoralis whereas all of the control animals, which [51][52]. Even more importantly, some of had been infected but not treated with the OMe our alkaloids are strong agents in connec- alkaloid, had died soon after infection Me OMe tion with the widespread tropical diseases [58]. schistosomiasis and, in particular, malar- A most promising further perspective ia. for this new antimalarial lead structure is given by the fact that naphthylisoquino- 5.1. Molluscicidal Activity of Dionco- lines do not only act efficiently against the phylline A more susceptible blood forms of Plasmo- In cooperation with Prof. Hostettmann, dium spp., but also against the more per- Me good molluscicidal activities against sistent liver forms [59]. Mea Biomphalaria glabrata, the intermediate host of bilharzia, were found in the first 5.3 Dimerization may Improve Antima- (M'p}-Jozimine A (27) line for dioncophylline A (4). It kills all of larial Activity of Naphthylisof]uinolines the snails within 24 h at a concentration of In a first attempt to further improve the (ICso = 0.075 Jlg/ml !) 20 ppm [53]. By derivatization of the antimalarial activity of naphthylisoquino- natural product, the activity was increased lines, we have prepared a broad series of Fig. 15. In vitro activities of dioncophylline A significantly [54]. We expect further im- structural analogs, e.g., of dioncophylline analogs against P. falciparum provement along these lines in the future. A (4), one of the moderately active mono-

MeO OMe

Mea OMe

MeO OMe

••Me

Me

Me OMe MeO OMe

Jozimine B (28) Ancistrocladine (29) Isojozimine B (30)

(IGSO = 0.90 Jlg/ml) (IGSO = 18.4 /lg/ml) (/GSO= 0.54 Jlg/ml)

Fig. 16. Increase of antimalarial activity of ancistrocladine by dimerization BIOLOGICALLY ACTIVE AGENTS IN NATURE 25

CHIMIA 52 (1998) NT. 1/2 (JouumlFebmar)

OH OMe

MeO OH .---.,.OH

OH Me Dioncophylline C (19) => ICSO(P. falcip.) = 0.014 !tglml (no anti-HIV activity) I Alignment I

OH

Me

Me Dioncopeltine A (10)

OH Jozimine C (31)

ICso(P. falcip.) = 0.44 !tg/ml Fig. 18. Optimum alignment for the two most active antimalarial naphthylisoquinolines ECso(HIV-1) = 271lg/ml Unfortunately, this effect was not ob- then prepared). For this purpose, we have Fig. 17. Antimalarial and anti-HIV activities of served for dioncophylline C (19), still the started to perform quantitative structure- mono- and dime ric dioncophylline C best monomer. Its dimer, namedjozimine activity relationship (QSAR) investiga- C (31) (Fig. 17) [3][48][62], has a lower tions, using the CoMFA technique [64]. activity than 19. This was not entirely For a comparison of thejoint structural mers, which, however, is available in larg- unexpected, because the structurally close- properties of naphthylisoquinolines, the er quantities. Fig. 15 shows three exam- ly related dimers of antimalarial korupens- existence of different coupling types and ples, 25-27, that reveal, for instance, N- amines, the michellamines, e.g., 1, are atropo-diastereomers has to be taken into methylation to decrease activity, while a virtually inactive towards P. falciparum consideration. A most efficient alignment second N-methylation, which establishes [63]. On the other hand, this structural is attained by a combined match for the a positive charge on the nitrogen, increas- relationship to michellamines may account isoquinoline and the naphthalene part, such es the antiplasmodial activity vs. the natu- for the appreciable anti-HIV activity dis- that the assumedly pharmacophoric func- ral product [35]. played by 31 - it is, actually, the most tions, in particular the free phenolic OH An unexpectedly positive result was active unnatural dimer of a natural mono- groups of the naphthalene parts, respec- obtained through chemical coupling of meric naphthylisoquinoline ever prepared. tively the basic N-atoms are close to each dioncophylline A (4) to give its unnatural other. As shown in Fig. 18, we have thus dimer 27, named jozimine A [60], which is 5.4. Quantitative Structure-Activity achieved a compact stereochemical match, ca. 20 times more active than the mono- Relationship (QSAR) Investigations e.g., of the two as yet best candidates, meric parent compound 4. It is now one of A most promising further strategy of dioncophylline C (19) and dioncopeltine the three most active naphthylisoquino- improving antimalarial activity is to try to A (10), which have so similar biological lines at all, after dioncophylline C (19) and predict structures with higher activities activities and indeed look quite similar dioncopeltine A (10)! from structure-activity correlations, even when aligned in this way [65]. Also for ancistrocladine (29) (Fig. 16), though neither the molecular target nor the With this alignment and a training set an alkaloid of the related (v.i.) Ancistro- mode of action are known yet. It is, there- of 34 compounds, we have found a very c1adaceae family, an increase of antima- fore, rewarding to find out what the struc- good correlation (q2 = 0.76) between cal- larial activity by dimerization was accom- tural requirements for high activities are. culated and experimental values, which plished [3]: Jozimine B (28), a constitu- The most active naphthylisoquinolines so now allows us to predict ICso values for tionally unsymmetric dimer of 29, has a far, dioncophylline C (19) and dioncopel- unknown or not yet measured structural 20-fold increased activity, and iso-joz- tine A (10), however, look quite different analogs. Based on these first QSAR inves- imine B (30), an isoquinoline-isoquino- at first sight (cf Fig. 14). From an align- tigations, we have started to predict and line-coupled dimer, is even more active by ment of these and the other molecules, synthesize new and hopefully still more a factor of 34 [61]. better analogs might be predicted (and active compounds. BIOLOGICALLY ACTIVE AGENTS IN NATURE 26

CHIMIA 52 (1998) Nr. In (JanuarlFebruar)

Scheme 7. Assumedly Joint Biosynthetic Pathways to Acetogenic Naphthylisoquinolines, Tetralones, and Naphthoquinones

in , HO 0 HO 0 HO 0 Drosophyllaceae, ~OH M~ Dioncophyllaceae, ~Me W~-~Me Ancistrocladaceae, OH o o Nepenthaceae, Isoshinanolone (32) Plumbagin (33) Droserone (34) Plumbaginaceae t OH OH

- ~ Me COOH ~Me MeO 5 < 35 HO COSCoA reduction Dioncopeltine A (10) Me / reductive coupling + ~ 0W0 0 amination O-methylation only in Dioncophyllaceae, oxygenation o 0 Ancistrocladaceae -< /' etc. COSEnz 36 Me ?' I R/S XW~ ....-:N x = OH, H OH Me MeO 37 MeO Ancistrocladisine (38)

6. The Chemotaxonomic 'Neighbor- further related families Drosophyllaceae, A. korupensis produce typical Ancistro- hood' of T. peltatum Droseraceae, Nepenthaceae (all carnivo- cladaceae-type alkaloids, pure Dionco- rous), and Plumbaginaceae (subfamily phyllaceae-type alkaloids, and, in addi- Another approach to look for new struc- Plumbaginoideae) likewise do produce tion, mixed, hybrid-type alkaloids with a tural analogs, is to investigate related naphthalene-related metabolites [66], but great structural di versi ty, including dimers plants. As mentioned earlier, the family apparently no naphthylisoquinolines. (like michellamines, cf Fig. 1), naphtha- Dioncophyllaceae is extremely small: lene-free quaternary isoquinolinium salts Besides Triphyophyllum pel tatum there are only the very rare further species 7. The Bottleneck Reaction in Naph- Habropetalum dawei from Sierra Leone thylisoquinoline Biosynthesis: The and Liberia, and Dioncophyllum thollonii Reductive Introduction of Nitrogen from Gabon and Congo [15]. At first sight, these two closest relatives of T.peitatum, All these metabolites are likely de- which have recently become available to rived from a common polyketide precur- us, seem entirely different: just abundant sor 36, the ability to produce naphthyliso- amounts of the isocyclic natural products quinolines obviously depending on a isoshinanolone (32), plumbagin (33), and unique reductive amination step to give droserone (34), which are apparently de- the dihydroisoquinoline 37 (Scheme 7). rived from the naphthalene part 35 of For Dioncophyllaceae, this step yields naphthylisoquinoline alkaloids, were exclusively (3R)-configurated alkaloids found (Scheme 7). Only by applying that lack an oxygen function at C(6) (e.g., HPLC-MS trace analytical techniques, we dioncopeltine A (10) Scheme 7) - the so- detected low quantities of dioncophylline called Dioncophyllaceae-type [2], where- A (4) and related compounds in these as for Ancistrocladaceae, the situation is species. more complex: while Asian Ancistro- Besides the Dioncophy llaceae, the only c1adaceae contai n on Iy 'Ancistrocladace- other plant family known to produce ae-type' alkaloids [2] like ancistrocladisine naphthylisoquinolines, are the closely re- (38) (Scheme 7), which are characterized lated Ancistrocladaceae. This small palaeo- by their (S)-configuration at C(3) and an tropical family consists of a single oxygen function at C(6), West-African Fig. 19. Unique morphological characteristics (Ancistrocladus) with ca. 22 species. The Ancistrocladaceae likeA. abbreviatus and ofT. peltatum BIOLOGICALL Y ACTIVE AGENTS IN NATURE 27

CHIMIA 52 (1998) Nr. 1/2 (Januar/Fcbruar)

Scheme 8. Biologically Active and Biogenetically Unprecedented Alkaloids ofT. peltatum

2 x 6 HOAc

Dioncophylline A (4) / (molluscicidal) Me Me

MeO HO OMe

Me Dioncophylline 8 (17) (fungicidal)

HO Me dimerization Dioncophylline C (19) .. Jozimine C (31) (antimalarial) (anti-HIV)

[67], and naphthy lisoquinoline gl ycosides to this most peculiar and productive Afri- Received: November 26, 1997 [37]. can plant, which will certainly yield most interesting further results in the near fu- [I] 'Acetogenic Isoquinoline Alkaloids', Part ture. 106; for Part 105, see ref. [48]; 'Antiproto- 8. Conclusion lOal Activity of Naphthylisoquinoline Al- We are grateful to our skillful and extremely kaloids' , Part II; for Part 10, see ref. [48]. Triphyophyllum peltatum does not pro- engaged coworkers (their names can be seen [2] G. Bringmann, F. Pokorny, in 'The Alka- duce dimers, glycosides, isoquinolinium from the literature citations), to T. Pabst, R. loids', Ed. G.A. Cordell, Academic Press, salts, (S)-configurated or 6-oxygenated al- Bruckner, M. Blank, J. Hinrichs, M. Lehrmann, New York, 1995, Vol. 46., p. 127-271. [3] G. Bringmann, Bull. Soc. Chirn. Belg.1996, kaloids as Ancistrocladaceae do, still it and Dr. D. Feineis for engaged help with the manuscript, and to H. Bringrnann, to whom we 105,601. occupies an outstanding position and is a owe the photos. We would further like to express [4] G. Bringmann, in 'Phytopharmaka in For- particularly rewarding plant species, and our sincere gratitude to all our African and Asian schung und klinischer Anwendung', Eds. thus the main topic of this paper: partners who have given us all sorts of support N. Rietbrock and D. Loew, Steinkopff, 1) It has characteristic morphological fea- and cooperative reliability, particularly Dr. A.M. Darmstadt, 1995, p. I L3-128. tures (Fig. 19): the carnivorous organs, Louis (Gabon) and K. Durnbuya (Sierra Leone). [5] K.P. Manfredi, J.W. Blunt, J.H. Cardellina the clawed leaves, and the bizarre seeds, We finally thank the Deutsche Forschungsge- 11, J.B. McMahon, L.L. Pannell. C.M. meinschaft (SFB 251 'Okologie, Physiologie und for which the species bears the name Cragg, M.R. Boyd, J. Med. Chern. 1991, Biochemie pflanzlicher und tierischer Leistun- 30,2067. 'peltatum' (= shielded). gen unter Stress', Graduiertenkolleg 'Magne- [6] M.R. Boyd, Y.F. Hallock, J.H. Cardellina 2) It contains most peculiar chemical con- tische Kernresonanz in vivo und in vitro', and II, K.P. Manfredi, J.W. BLunt, J.B. McMa- stituents, i.a. dioncophyllines A (4), B Normalverfahren), the UNDP/WorldBanklWHO hon, R.W. Buckheit Jr., G. Bringmann, M. (17), and C (19) of likewise unprece- Special Programme for Research and Training in Schaffer, G.M. Cragg, D.W. Thomas, J.G. dented biosynthetic origin from ace- Tropical Diseases (TDR), the Fonds der Chemi- Jato, J. Med. Chern. 1994, 37, L740. tate units (Scheme 8). schen Industrie, the BASF AG, the Kneippwerke AG, and the Max-Buchner-Stiftung for generous [7] G. Bringmann, R. Zagst, M. Schaffer, Y.F. 3) It has promising biological properties, financial support. Hallock, J.H. Cardellina II, M.R. Boyd, e.g., molluscicidal activity for dionco- Angew. Chern. 1993, 105, 1242; ibid., Int. phylline A (4), fungicidal activity for Ed. Eng/. 1993,32, 1190. dioncophylline B (17), and, in particu- [8] G. Bringmann, K.-P. Gulden, Y.F. Hal- lar, excellent antimalarial activity for lock, K.P. Manfredi, J.H. Cardell ina II, dioncophylline C (19). Moreover,joz- M.R. Boyd, B. Kramer, J. Fleischhauer, imine C (31), the apparently unnatural Tetrahedron 1994, 50, 7807. [9] G. Bringmann, R. Gatz, P.A. Keller, R. dimer of 19, shows a very high anti- Walter, P. Henschel, M. Schaffer, M. HIV activity, an additional novel di- StabLein, T.R. Kelly, M.R. Boyd, Hetero- mension of these Triphyophyllum al- cycles 1994, 39, 503. kaloids. [10] G. Bringmann, S. Harmsen, J. Holenz, T. Thus, we have true hits in four impor- Geuder, R. Gatz, P.A. Keller, R. Walter, tant fields of indications. All this, along Y.F. Hallock, J.H. Cardellina II, M.R. Boyd, with a lot of scientific satisfaction, we owe Tetrahedron 1994, 50, 9643. BIOLOGICALLY ACTIVE AGENTS IN NATURE 28

CHIMIA 52 (1998) NT. 1/2 (JnnunT/FcbrunT)

[ 11] T.R. Kelly, A. Garcia, F. Lang, J.J. Walsh, [32] G. Bringmann, D. Koppler, D. Scheutzow, [53] G. Bringmann, 1. Holenz, L. Ake Assi, C. K.V. Bhaskar, M.R. Boyd, R. Gatz, P.A. A. Porzel, Magn. Reson. Chem. 1997,35, Zhao, K. Hostettmann, Planta Med. 1996, Keller, R. Walter, G. Bringmann, Tetrahe- 297. 62,556. dron Lett. 1994,35, 7621. [33] G. Bringmann, T. Geuder, M. Riibenacker, [54] G. Bringmann, 1. Holenz, L. Ake Assi, K. [12] More recent syntheses: T.R. Hoye, M. R. Zagst, Phytochemistry 1991, 30, 2067. Hostettmann, Planta Med. 1998, in press. Cheng, L. Mi, O.P. Priest, Tetrahedron [34] G. Bringmann, R. God, M. Schaffer, Phy- [55] G. Fran~ois, G. Bringmann, J.D. Phillip- Lett. 1994, 47, 8747; G. Bringmann, R. tochemistry 1996, 43, 1393. son, L. Ake Assi, C. Dochez, M. Riiben- Gatz, S. Harmsen, 1. Holenz, R. Walter, [35] G. Bringmann, M. Ochse, M. Schaffer, R. acker, C. Schneider, M. W6ry, D.C. War- Liebigs Ann. 1996,2045; P.D. Hobbs, V. God, R. Walter, G. Fran~ois, Planta Med., hurst, G.c. Kirby, Phytochemistry 1994, Upender, 1. Liu, DJ. Pollart, D.W. Tho- In press. 35, 1461. mas, M.I. Dawson, J. Chem. Soc., Chem. [36] P. Chau, I.R. Czuba, M.A. Rizzacasa, G. [56] G. Fran~ois, G. Bringmann, C. Dochez, C. Commun. 1996,923; P.O. Hobbs, V. Up- Bringmann, K.-P. Gulden, M. Schaffer, J. Schneider, G. Timperman, L. Ak6 Assi, J. ender, M.1. Dawson, Synlett 1997, 965; G. Org. Chem. 1996,61,7101. Ethnopharmacol. 1995, 46, 115. Bringmann, R. Gatz, P.A. Keller, R. Walter, [37] Y.F. Hallock, J .H. Cardellina 1I, M. [57] G. Fran~ois, G. Timperman, 1. Holenz, L. M.R. Boyd, F. Lang, A. Garcia, J.J. Walsh, Schaffer, M. Stahl, G. Bringmann, G. Ake Assi, T. Geuder, L. Maes, J. Dubois, I. Tellitu, K.V. Bhaskar, T.R. Kelly, 10rg. Fran~ois, M.R. Boyd, Tetrahedron 1997, M. Hanocq, G. Bringmann, Ann. Trop. Chem. 1998, in press. 53,8121. Med. Parasitol. 1996, 90, J 15. [13] 1. Schlauer, B. Wiesen, M. Riickert, L. Ak6 [38] N.H. Anh, A. Porzel, H. Ripperger, G. [58] G. Fran~ois, G. Timperman, W. Eling, L. Assi, R.D. Haller, S. Bar, K.-U. Frahlich, Bringmann, M. Schaffer, R. God, T.V. Ake Assi, J. Holenz, G. Bringmann, Allti- G. Bringmann, Arch. Biochem. Biophys. Sung, G. Adam, Phytochemistry 1997, 45, microh. Agellls Chemother. 1997,4/,2533. 1998,350,87. 1287. [59] G. Fran~ois, G. Timperman, T. Steenack- [14] R. Schmid, Bot. Jahrb. Syst. 1964,83, 1. [39] G. Bringmann, R. Zagst, B. Schaner, H. ers, L. Ake Assi, J. Holenz, G. Bringmann, [15] H.K. Airy Shaw, Kew Bull. 1951,3,327. Busse, M. Hemmerling, C. Burschka,Acta Parasitology Res. 1997,83,673. [16] S. Green, T.L. Green, Y. Heslop-Harrison, Crystallogr. 1991, C47, 1703. [60] G. Bringmann, W. Saeb, D. Koppler, G. Bot. J. Linn. Soc. 1979, 78, 99. [40] G. Bringmann, W. Saeb, K. Peters, E.-M. Fran~ois, Tetrahedron 1996, 52, 13409. [17] G. Bringmann, M. Wenzel, H. Bringmann, Peters, Phytochemistry 1997, 45, 1283. [61] G. Bringmann, W. Saeb, G. Fran~ois, in 1. Schlauer, L. Ak6 Assi, Der Palmengar- [41] G. Bringmann, M. Riibenacker, P. Vogt, H. prep. ten 1996, 60/2, 32. Busse, L. Ak6 Assi, K. Peters, H.G. von [62] G. Bringmann, in 'Chemistry, Biological [18] G. Bringmann, M. Riibenacker, J.R. Jan- Schnering, Phytochemistry 1991, 30, 1691. and Pharmacological Properties of African sen, D. Scheutzow, L. Ak6 Assi, Tetrahe- [42] G. Bringmann, W. Saeb, R. God, M. Medicinal Plants', Eds. K. Hostettmann, F. dron Lett. 1990,31,639. Schaffer, G. Fran~ois, K. Peters, E.-M. Chinyanganya, M. Maillard, andJ.-L. Wol- [19] G. Bringmann, J.R. Jansen, H. Reuscher, Peters, P. Proksch, K. Hostettmann, L. Ak6 fender, University of Zimbabwe Publica- M. Riibenacker, K. Peters, H.G. von Schne- Assi, Phytochemistry, submitted. tions, Harare, Zimbabwe, 1996, p. 1-19. ring, Tetrahedron Lett. 1990,31,643. [43] G. Bringmann, M. Riibenacker, W. Koch, [63] Y.F. Hallock, K.P. Manfredi, J.W. Blunt, [20] G. Bringmann,J.R. Jansen, Synthesis 1991, D. Koppler, T. Ortmann, M. Schaffer, L. 1.H. Cardellina II, M. Schaffer, K.-P. Gul- 825. Ak6 Assi, Phytochemistry 1994, 36,1057. den, G. Bringmann, A.Y. Lee, 1. Clardy, G. [21] G. Bringmann, R. Walter, R. Weirich, An- [44] G. Bringmann, T. Ortmann, R. Zagst, B. Fran~ois, M.R. Boyd, J. Org. Chem. 1994, gew. Chern. 1990,102, 1006; ibid., Int. Ed. Schaner, L. Ake Assi, C. Burschka, Phyto- 59,6349. Eng!. 1990,29,977. chemistry 1992, 31, 4015. [64] R.D. Cramer, D.E. Patterson, 1.D. Bunce, [22] G. Bringmann, O. Schupp, S. Afr. J. Chern. [45] G. Bringmann, M. Riibenacker, T. Geuder, 1. Am. Chem. Soc. 1988,110,5959. 1994, 47, 83. L. Ak6 Assi, Phytochemistry 1991, 30, [65] G. Bringmann, D. Vitt, S. Schmitt, unpub- [23] N. Harada, K. Nakanishi, 'Circular Dichro- 3845. lished results. ic Spectroscopy - Exciton Coupling in Or- [46] G. Bringmann, T. Ortmann, M. Riiben- [66] R. Hegnauer, in 'Chemotaxonomie der ganic Stereochemistry', Oxford University acker, L. Ak6 Assi, Planta Med. (SuppJ. I) Pflanzen', Birkhauser, Basel, 1989, Vol. 8, Press, Oxford, 1983. 1992,58,701. p.388. [24] 1. Bruneton, A. Bouquet, A. Fournet, A. [47] G. Bringmann, M. Riibenacker, R. Wei- [67] Y.F. Hallock,J.H. CardellinaJJ, T. Kornek, Cave, Phytochemistry 1976, 15,817. rich, L. Ak6 Assi, Phytochemistry 1992, K.-P. Gulden, G. Bringmann, M.R. Boyd, [25] M. Lavault, 1.Bruneton, Planta Med. (Sup- 31,4019. Tetrahedron Lett. 1995, 36, 4753. pl.) 1980, 17. [48] G. Bringmann, 1. Holenz, R. Weirich, C. [26] For a more detailed discussion, see: G. Funke, M. Ri.ibenacker, R.I. Gulakowski, Bringmann, in 'The Alkaloids' ,Ed. A. Bros- M. Boyd, G. Fran~ois, Tetrahedron 1998, si, Academic Press, New York, 1986, Vol. 54,497. 29, p. 141-184. [49] G. Bringmann, M. Wenzel, M. Riickert, K. [27] G. Bringmann, K.-P. Gulden, H. Busse, J. Wolf, S. Busemann, M. Schaffer, L. Ak6 Fleischhauer, B. Kramer, E. Zobel, Tetra- Assi,HeterocycJes 1998, in press; G. Bring- hedron 1993, 49, 3305. mann, M. Wenzel, M. Riibenacker, M. [28] G. Bringmann, M. Stahl, K.-P. Gulden, Schaffer, M. RUckert, L. Ak6 Assi, Phy- Tetrahedron 1997, 53, 2817. tochemistry 1998, in press. [29] G. Bringmann, S. Busemann, in 'Natural [50] G. Bringmann, M. Riibenacker, E. Am- Product Analysis', Eds. P. Schreier, M. mermann, G. Lorenz, L. Ake Assi (BASF Herderich, H.U. Humpf, and W. Schwab, AG); D.O.S. DE 41 17080 A I (disclosure Vieweg, Braunschweig, 1998, in press. 26.11.92); EP 0515 856 Al (disclosure [30] J. Fleischhauer, A. Koslowski, B. Kramer, 02.12.1992). E. Zobel, G. Bringmann, K.-P. Gulden, T. [51] G. Bringmann, S. Gramatzki, C. Grimm, P. Ortmann, B. Peter, Z. Naturforsch. 1993, Proksch, Phytochemistry 1992, 3 I, 3821. 48h,140. [52] G. Bringmann, J. Holenz, B. Wiesen, B.W. [31] G. Bringmann, I.R. Jansen, H. Busse, Lie- Nugroho, P. Proksch, 1. Nat. Prod. 1997, bigs Ann. Chem. 1991, 803. 60,342.