|Indole Alkaloid Biosynthesis: the Pathways leading to Ajmalicine and Ajmaline
Leif Barleben1 and Joachim Stöckigt1,2
in memoriam of Professor Asima Chatterjee
(1) Institute of Pharmacy, Department of Pharmaceutical Biology, Johannes Gutenberg-University, Staudinger Weg 5, D-55128 Mainz, Germany
(2) College of Pharmaceutical Sciences, Department of Traditional Chinese Medicine and Natural Drug Research, Building of College of Pharmaceutical Sciences, Zijingang Campus, Zhejiang University, Hangzhou 310058, P.R. China
Originating from India, and dating back to the origin of Ayurvedic medicine, Rauvolfia serpentina (Sanskrit: Sarpagandha) has been a medicinal plant of significant importance, whose root drug has gone on to become therapeutically applied worldwide. In the foreword of the monograph “RAUVOLFIA SERPENTINA – Sarpagandha Botany and Agronomy”1 Asima Chatterjee stated: “This particular plant attracted the world attention because of its high therapeutic efficiency in the treatment of hypertension, insomnia and mental disorders.”
Rauvolfia’s therapeutic importance is based on the occurrence of pharmacologically active natural products, namely the monoterpenoid indole alkaloids. Examples of this alkaloid group include reserpine, yohimbine, sarpagine, raucaffricine, ajmalicine and ajmaline, whose chemical formulae demonstrate their structural complexity and diversity (Figure 1). Enzymatic biosynthesis of two alkaloids in particular has been the subject to the most of extensive research: the heteroyohimbine-type ajmalicine and the ajmalan-type ajmaline. Whereas the former alkaloid also occurs in the plant Catharanthus roseus, ajmaline is a typical constituent of the genus Rauvolfia, thus the principle ajmalicine and ajmaline synthesizing enzymes have been obtained from C. roseus and R. serpentina respectively. With particular emphasis on ajmaline biosynthesis, the current knowledge pertaining to both of these pathways is summarized in this chapter
Figure 1 here
Plant Cell Suspension Cultures
The biological prerequisite for investigation of both pathways has not been the plant itself but plant cell suspension cultures, which were first established in the laboratories of Zenk at Bochum and Munich Universities. Development of this cell culture methodology was particularly important as it now provides at the lab scale “unlimited” amounts of cell material for use in the laboratory, namely the generation of 10-20 kg of cells in approximately four weeks. This was found to be especially important for Rauvolfia plants, due to the difficulties relating to their cultivation under laboratory or green-house conditions. Compared to the cell culture method, their slow growth characteristics make it impossible to generate the necessary quantities of physiologically active material in a short space of time. For these reasons, application of plant in vitro culture techniques can be regarded as one of the early breakthroughs in natural plant product biosynthesis. Though, in many ways representing the optimal system for natural product biosynthesis, the limitation of the cell cultured cells do not undergo differentiation, thus the natural product biosynthesis that would otherwise be coupled to differentiation processes may not be expressed. A prominent example of such a case is the biosynthesis of the high valuable anticancer bis-indole alkaloids vinblastine (VBL) and vincristine (VCR), which have never been detected in de-differentiated C. roseus cell suspension cultures. The reason for this is that only few late enzymes of the formation of the vindoline part are suppressed.
From Alkaloid to Enzyme
The basis for unraveling a biosynthetic pathway at the enzyme level is the functional expression of the enzyme in the cell culture system and the availability of the enzyme substrate and the enzyme product – both the appropriate alkaloids. The initiation of a broad phytochemical screening resulting in an overwhelming number of isolated alkaloids from cell suspensions of different plant origin has always represented the first step in our biosynthetic research strategies. Extensive phytochemical analysis of cell suspensions and organ cultures (so-called “hairy roots”) of Catharanthus2 and Rauvolfia3 and, in particular, the isolation and structural identification of the single alkaloids, has been of great help in searching for distinct enzymes of the heteroyohimbine alkaloid and ajmaline pathways in cell-free systems, which typically yield crude or only pre-purified protein preparations.
The Route to Ajmalicine
The availability of large alkaloid collections and purified enzyme preparations from the relevant plants has allowed extensive substrate studies to be carried out, in which single enzymes can be “placed” into a biosynthetic scheme based on their substrate specificity. The relatively short pathway linking tryptamine, secologanin and ajmalicine isomers was established using this technique about three decades ago4. This pathway is illustrated in Figure 2. It included the detection of the enzyme strictosidine synthase (STR), which catalyzes the stereo-selective Pictet-Spengler reaction between secologanin and tryptamine, followed by strictosidine glucosidase (SG) for generation of the highly reactive strictosidine aglycone. Through isolated, indirectly identified and structurally suggested intermediates (illustrated in brackets), 4,21-dehydrocorynantheine aldehyde and 4,21-dehydrogeissoschizine are generated first. The stable five-ring system of the heteroyohimbine type alkaloids, the cathenamine isomers, is formed enzymatically by a Michael-addition at the stage of dehydrogeissoschizine.
Figure 2 here
During the final reaction catalyzed by NADPH-dependent cathenamine reductase(s), the three isomers tetrahydroalstonine, 19-epi-ajmalicine and ajmalicine are generated. Using cell-free and stable isotope labeling, this pathway has been mechanistically investigated in detail. One of the authors of this chapter (J.S.) fondly recalls the extensive discussions held with Asima Chatterjee during her visit to Munich University 27 years ago relating to this precise topic.
An in-depth description of the route to the ajmaline isomers involving such labeling experiments were published in the journal Biochemistry5.
The Key Role of the Glucoside Strictosidine
An important finding during elucidation of the ajmalicine biosynthesis was the detection of the initial condensation product between tryptamine and secologanin, leading to the glucoalkaloid strictosidine6,7. It is somewhat unusual that a glucoside is located at the beginning of a biosynthetic pathway. In order to achieve sufficient water-solubility, for storage in high concentrations within the vacuole, glucosides are typically known to be synthesized at the end of metabolic routes in plants. In the alkaloidal structure of strictosidine, the glucosidic residue acts as a protective group. Following enzymatic removal of the glucose unit by strictosidine glucosidase, the resulting aglycone becomes highly active and functionalized so that it can enter a number of different pathways; these eventually lead to synthesis of >2000 members of the monoterpenoid indole alkaloid family, as well as a number of quinoline alkaloids8. The biosynthetic significance of strictosidine has been investigated by way of feeding experiments and part of its universal role in alkaloid biosynthesis is illustrated in Figure 3.
The biosynthetic Pathway to Ajmaline
Due to the significant structural complexity of ajmaline (the alkaloid consists of a six-membered ring system with nine chiral carbon atoms) its biogenetic formation was expected to be significantly more complicated. In fact, following many years of enzymatic studies ajmaline was found to be synthesized in Rauvolfia by way of a ten-step reaction scheme (Figure 4); the individual transformations are catalyzed by a number of novel and functionally diverse enzymes belonging to a variety of enzyme families9. Of the clearly characterized 10 steps, six of the relevant enzymes have been purified and partially sequenced, leading to the successful implementation of a “reverse genetics” approach in which the corresponding cDNAs have been heterologously and functionally cloned. In addition to enabling the determination of the primary structure of the catalytic enzymes, a number of these have been overexpressed in Escherichia coli and purified as C- or N-terminal His6-tagged fusion proteins in milligram quantities by nickel-nitrilo-triacetic acid matrix-based affinity chromatography9. Recombinant strictosidine synthase (STR1), strictosidine glucosidase (SG), polyneuridine aldehyde esterase (PNAE), and vinorine synthase (VS) have been synthesized in the most significant quantities. Whereas the sarpagine bridge enzyme (SBE), vinorine hydroxylase (VH), cytochrome P450 reductase (CPR), vomilenine reductase (VR), di-hydrovomilenine reductase (DHVR), acetyl-norajmaline esterase (AAE) and norajmaline methyltransferase (NAMT) have been well characterized, AAE and the CPR alone have been functionally cloned, albeit with low expression rates. It is also worth mentioning that a number of side routes branch from this major pathway in Rauvolfia, resulting in a plant alkaloidal network which is one of the biggest metabolomes in natural product biosynthesis, connecting nearly 30 monoterpenoid indole alkaloids. Figure 4 illustrates the summary of a part of this Rauvolfia network, which represents the complete ajmaline biosynthetic pathway.
Approximately half of the enzymes on the branch which leads to ajmaline synthesis have so far been cloned; the availability of the corresponding genes (cDNAs) is such that synthesis of alkaloids by more efficient systems such as prokaryotic organisms (Escherichia coli) may become a distinct reality in future. In contrast to biosynthesis in microorganisms, genes within the plant cell are assumed to be not clustered. However, in the future, clustering of the relevant plant alkaloid biosynthesis enzyme cDNAs may permit more efficient synthesis in organisms such as E. coli compared to the slowly growing differentiated plants or cell suspension cultures. An increasing number of examples have been documented in which this type of metabolic engineering has been carried out successfully: most recently, hydrocortisone biosynthesis has been successfully achieved by the integration and expression of thirteen genes in yeast cells10. Characterization of the genes of a plant biosynthetic pathway such as that described above encourage the application of this methodology to the future synthesis of plant secondary products.
Though this kind of metabolic engineering is in itself a significant scientific challenge, synthesis of natural products lends itself primarily to production of a single compound, most likely the end product of a biogenetic route, which is of high commercial value and which cannot be generated in sufficient amounts by classical methods. On the other hand, modification of cDNAs, which leads to enzyme substrate acceptance which can be modulated to become broader, could lead to the synthesis of ‘libraries’ of a significant number of novel natural products. In order to change enzyme activities and substrate specificities, diverse approaches such as “directed evolution” might be applied, generating a broad spectrum of mutated enzymes. From the resulting mutant pool, proteins with the desired properties need to then be detected and selected; rational structure-based re-design of enzymes may also be a more challenging but appealing way in which to succeed. In this case, however, knowledge of the three-dimensional structure and properties of the enzyme substrate binding sites is indispensable. To date, only a handful of three-dimensional structures of enzymes of secondary metabolism have been determined, this has been due to the limited quantity of proteins obtained from such biosynthetic pathways, which has hindered the potential for crystallization. This situation has however improved thanks to heterologous expression, resulting in synthesis of milligram quantities of proteins11, rapid X-ray analysis developments and advanced software pipelines for fast structure solution (such as Auto-Rickshaw12).
Crystallization and 3D-Structure of Rauvolfia Enzymes
Systematic protein expression in E. coli followed by purification by Ni-NTA and hanging drop crystallization, diffracting crystals have been obtained for the four Rauvolfia enzymes STR1, SG, PNAE and VS as illustrated in Figure 5.
Structure determination of vinorine synthase and strictosidine synthase are of special interest for a number of reasons: VS belongs to the BAHD enzyme family, whose members exhibit important functions relating to the biosynthesis of well-known plant secondary metabolites such as morphine, taxol, vindoline -a precursor of the “dimeric” alkaloids mentioned at the beginning of this article-, or VS itself in ajmaline biosynthesis.
Crystallization and structure determination of VS were fascinating from several points of view. Firstly, extensive crystallization trials lead to the finding that unusual crystallization conditions yielded the best results. Small protein amounts and relatively high temperature were found to be the optimal conditions for crystallization of VS. Best diffraction crystals were grown with enzyme concentrations of 2 mg ml-1 only at temperatures above 30 °C (optimum at 32 °C). Secondly, there were no 3D-structures known from the BAHD enzyme family, so that eventually a number of selenomethionine mutants needed to be prepared. Finally, successful VS-structure elucidation (using the MAD approach) became the first example in the BAHD protein family, allowing simplified subsequent structural determination by the molecular replacement technique. Moreover, an extensive set of mutation experiments which previously had resulted in a number of VS-mutants before could be elegantly explained on the basis of the 3D-structure of the synthase, providing deep insight into the mechanism of this acetyl CoA-dependent enzyme13.
The structure of the second enzyme, strictosidine synthase (STR1), represented the first example of a six-bladed β-propeller fold from higher plants. More interestingly, along with several ligand complexes containing the substrates tryptamine, secologanin and the product strictosidine, the 3D-analysis provided new clues relating to the nature of the active center of the enzyme and its molecular mechanism14. The STR1 structure was indispensable in understanding the mechanism of the first enzyme known to catalyze a Pictet-Spengler reaction (Pictet-Spenglerase). However, in the light of generating novel STR1 mutants with new substrate specificities which would be useful for generating large alkaloid libraries, the structure of those ligand complexes proved to be of great help for rational re-design of the synthase. For this reason, STR1 mutants have been systematically generated by site-directed mutagenesis, particularly for creating a larger binding pocket in order to generate novel strictosidine derivatives with an extended substitution pattern at the aromatic ring. As a first example, the wild-type STR1 does not accept 5-methyl- and 5-methoxytryptamine. Determination of the catalytic center of STR1 with the bound strictosidine molecule clearly indicated that the amino acid Valine208 is in close proximity to the 5 position of tryptamine, leading to steric hindrance and rejection by STR1 of 5-methyl- and 5-methoxy-substituted tryptamines. The novel mutant Val208Ala was therefore designed which exhibited conversion of these tryptamines and opened the way to rational-designed novel strictosidine derivatives for the first time. In this context, it is important to note that alkaloids with specific substitutions at positions 5 and 6 belong to highly valuable drugs such as vinblastine, vincristine, reserpine and quinine all of which harbor a methoxy-group at these positions. Because strictosidine can be efficiently used to generate large numbers of novel indole alkaloids by a biomimetic chemo-enzymatic strategy, such an approach would allow establishment of new alkaloid libraries of optimized or novel biological potential15.
In conclusion, crystallization and 3D-structure determination of the enzymes involved in the biosynthesis of natural products provide not only mechanistic insights into enzyme function, but also represent the basic requirements for a rational re-design and modulation of the activities of the single enzymes.
1. B.N. Sahu, Rauvolfias, Botany and Agronomy (Today and Tomorrow´s Printers and Publishers, New Dehli, 1979), Vol. I, p.v.
2. J. Stöckigt and H. J. Soll, Planta Medica, 40, 22-30 (1980).
3. Y. Sheludko, I. Gerasimenko and J. Stöckigt, Planta Medica, 68, 435-439 (2002).
4. J. Stöckigt, Indole and Biogenetically Related Alkaloids. (Academic Press Inc. London, New York, 1980), pp. 113-141.
5. J. Stöckigt, T. Hemscheidt, G. Höfle, P. Heinstein and V. Formacek, Biochemistry, 22, 3448-3452 (1983).
6. J. Stöckigt and M.H. Zenk, FEBS Letters, 79, 233-237 (1977).
7. J. Stöckigt and M.H. Zenk, J. Chem. Soc. Chem. Comm., 646-648 (1977).
8. T. Kutchan, Phytochemistry, 32,493-506 (1993).
9. M. Ruppert, X. Ma and J. Stöckigt, Current Organic Chemistry, 9, 1431-1444 (2005).
10. F.M. Szczebara, C. Chandelier, C. Villeret, A. Masurel, S. Bourot, C. Duport, S. Blanchard, A. Groisillier, E. Testet, P. Costaglioli, G. Cauet, E. Degryse, D. Balbuena, J. Winter, T. Achstetter, R. Spagnoli, D. Pompon and B. Dumas, Nat. Biotechnol., 21, 143-149 (2003).
11. J. Stöckigt, S. Panjikar, M. Ruppert, L. Barleben, X. Ma, E. Loris and M. Hill, Phytochem. Rev., 6, 15-34 (2007).
12. S. Panjikar, V. Parthasarathy, V.S. Lamzin, M.S. Weiss and P.A. Tucker, Acta Crystallogr., Sect. D., 61, 449-457 (2005).
13. X. Ma, J. Koepke, S. Panjikar, G. Fritzsch and J. Stöckigt, J. Biol. Chem., 280, 13576-13583 (2005).
14. X. Ma, S. Panjikar, J. Koepke, E. Loris and J. Stöckigt, Plant Cell, 18, 907-920 (2006).
15. E. Loris, S. Panjikar, M. Ruppert, L. Barleben, M. Unger, H. Schübel, and J. Stöckigt, Chem. Biol., 14, 979-985 (2007).
Figure 1. Typical examples of monoterpenoid indole alkaloids from Rauvolfia indicating some therapeutic applications, structural diversity and complexity.
Figure 2. The biosynthetic pathway on which the heteroyohimbine alkaloids ajmalicine, 19-epi-ajmalicine and tetrahydroalstonine are formed. Elucidation of the route was predominantly achieved at the enzymatic level in cell suspension cultures derived from Catharanthus roseus. Major enzymes are strictosidine synthase (STR) and strictosidine glucosidase (SG), which catalyze the formation of the highly unstable aglycon of strictosidine in its ring-closed and ring-opened form.
Figure 3. The central role of the enzymes strictosidine synthase (STR1) and strictosidine glucosidase (SG) in the biosynthesis of the entire family of monoterpenoid indole alkaloids and representative higher plant chinoline alkaloids. The indoles consist of approximately 2000 members of which only some structurally diverse members are shown.
Figure 4. The 10-step biosynthetic pathway leading to the antiarrhythmic alkaloid ajmaline in Rauvolfia serpentina cell suspension cultures. Abbreviations: STR1 (strictosidine synthase), SG (strictosidine glucosidase), SBE (sarpagan bridge enzyme), PNAE (polyneuridine aldehyde esterase), VS (vinorine synthase), VH (vinorine hydroxylase), CPR (cytochrome P450 reductase), VR (vomilenine reductase), DHVR (dihydrovomilenine reductase), AAE (acetylajmalan esterase), NAMT (norajmaline methyltransferase); with the exception of CPR in yeast and AAE in tobacco, enzymes marked with an asterisk have been heterologously expressed in Escherichia coli. The first four enzymes STR1, SG, PNAE and VS recently were crystallized and their 3D-structures determined.
Figure 5. Crystal and structure gallery of Rauvolfia enzymes catalyzing key reactions in the 10-step biosynthetic pathway from tryptamine and secologanin leading to the antiarrhythmic ajmaline (STR1, strictosidine synthase; SG, strictosidine glucosidase; PNAE, polyneuridine aldehyde esterase; VS, vinorine synthase).