The Alkaloids Volume 4

A Review of the Literature Published Between July 1972 and June 1973

By J. E. Saxton

The Royal Society of Chemistry

Copyright © 1974 The Chemical Society
All rights reserved.
ISBN: 978-0-85186-287-3

Contents

Chapter 1 Biosynthesis By R. B. Herbert, 1,
Chapter 2 Pyrrolidine, Piperidine, and Pyridine Alkaloids By V. A. Snieckus, 50,
Chapter 3 Tropane Alkaloids By J.E. Saxton, 78,
Chapter 4 The Pyrrolizidine Alkaloids By J. E. Saxton, 84,
Chapter 5 The Indolizidine Group By J. E. Saxton, 100,
Chapter 6 The Quinolizidine Alkaloids By J. E. Saxton, 104,
Chapter 7 Quinoline, Quinazoline, Acridone, and Related Alkaloids By V. A. Snieckus, 117,
Chapter 8 β-Phenethylamines and the Isoquinoline Alkaloids By V. A. Snieckus, 128,
Chapter 9 The Aporphines By M. Shamma and S. S. Salgar, 197,
Chapter 10 Amaryllidaceae and Related Alkaloids By V. A. Snieckus, 266,
Chapter 11 Erythrina and Related Alkaloids By V. A. Snieckus, 273,
Chapter 12 Indole Alkaloids By J. A. Joule, 280,
Chapter 13 Lycopodium Alkaloids By V. A. Snieckus, 322,
Chapter 14 Chemistry of the Diterpenoid Alkaloids By S. W. Pelletier and S. W. Page, 323,
Chapter 15 Steroidal Alkaloids of the Apocynaceae, Buxaceae, Asclepiadaceae, and of the Salamandra–Phyllobates Group By F. Khuong-Huu and R. Goutarel, 346,
Chapter 16 Solanum and Veratrum Steroidal Alkaloids By R. B. Herbert, 383,
Chapter 17 Miscellaneous Alkaloids By V. A. Snieckus, 395,
Addendum to Chapter 9, 428,
Errata, 429,
Author Index, 430,

CHAPTER 1

Biosynthesis

BY R. B. HERBERT

1 Piperidine, Pyridine, and Pyrrolidine Alkaloids

Lycopodine. — Pelletierine (1) serves as a specific precursor for cemuine (2) and lycopodine (3). The reasonable hypothesis that these alkaloids were modified dimers of pelletierine had to be abandoned when it was discovered that pelletierine gave only one each of the two C8N units (heavy bonding) of (2) and (3).

The results were surprising as lysine (4), cadaverine (5), and Δ1-piperideine (6) had all been built equally into each of the two C8N units of lycopodine and cernuine. It followed, therefore, that if lycopodine is formed from two pelletierine-like units, these units must either be identical or, if different, be derived from Δ1-piperideine (6) with the same overall dilution by non-labelled pools. Two models were proposed which could account for the incorporation of pelletierine, one (Scheme l) in which it could substitute for one of the two identical units, i.e. (8), and the other where it was an obligatory intermediate and precursor for one of the C8N units (Scheme 2); it is a consequence of the equal labelling of both halves of lycopodine by Δ1-piperideine, lysine, and cadaverine that in this scheme the steady-state concentration of pelletierine and (10) be small compared with (9).

The failure of [2-14C]-2-allylpiperidine [as (7)] to label C-5 of lycopodine (3) specifically in Lycopodium tristachyum provided evidence against the first hypothesis. Evidence was then adduced which showed that pelletierine (1) was very probably a normal intermediate in lycopodine biosynthesis. Firstly, the presence of pelletierine (1) in L. tristachyum was demonstrated. [2-14C]-Δ1-Piperideine [as (6)] together with a large quantity of inactive pelletierine was administered to the plant; a similar experiment was carried out with [1,5-14C2]-cadaverine [as (5)]. Radioactive pelletierine was isolated at the end of each experiment.

The presence and synthesis of pelletierine from lycopodine precursors was thus demonstrated but its obligatory participation in lycopodine biosynthesis was not proven. Secondly, therefore, the ability was examined of inactive pelletierine to dilute radioactivity from one of the C8N units following simultaneous administration of labelled Δ1-piperideine or cadaverine. Difficulties normally experienced with this type of experiment in plants were obviated since labelling of the C8N unit not derived from pelletierine served as an internal standard. It was found that pelletierine very effectively diluted activity in both experiments from the C8N unit (heavy bonding) of (3), consonant with its being an obligatory intermediate in lycopodine biosynthesis. All the experimental evidence is accommodated in Scheme 2, which also allows for the non-intact incorporation of β-hydroxybutyrate and acetoacetate.

Dioscorine. — Full details of the mode of incorporation of acetate into dioscorine (11) have been published. The results are consistent with either variant, a6 and b,4 of a pathway involving condensation of four acetate units with a lysine-derived unit, plausibly Δ1-piperideine (Scheme 3); as indicated in path b, pelletierine may be involved. Administration of [2-14C]lysine [as (4)] to the tropical yam, Dioscorea hispida, however, gave dioscorine (11) with little radioactivity; whilst [6-14C]-Δ1-piperideine [as (6)] was better used, the labelling pattern was essentially the same as that from [1-14C]acetate, arising presumably by catabolism of the radioactive Δ1-piperideine to acetate. These poor incorporations were rationalized by suggesting that, at the time of feeding, some compound, derivable from lysine, was not being actively synthesized but that it was available for condensation with acetate. This hypothetical compound could not be pelletierine for [1-14C]acetate incorporation would result in labelling of C-10 and C-12 but not C-5, and in fact almost equal labelling of these positions is observed.

It is worth noting that incorporation of acetate into dioscorine was only achieved with considerable difficulty and it seems possible that the administered lysine and Δ1-piperideine are not reaching the site of alkaloid synthesis, in which case pelletierine (1) may yet prove to be a precursor for dioscorine (11).

Anabasine. — Study of the biosynthesis of the piperidine ring of alkaloids like anabasine (14), N-methylpelletierine (16), and sedamine (17) has proved most interesting. Two pathways (Scheme 4, paths a and b) have been proposed which are consistent with the experimental results, in particular the incorporation of L-lysine without intervention of symmetrical intermediates or loss of hydrogen from C-2 or C-6. N-Methyl-Δ1-piperideine (13) is implicated as an intermediate in path a, and its role in the biosynthesis of anabasine (14) in Nicotiana glauca and N. tabacum has been examined by administration of [2-14C]-labelled material. Similar results were obtained for both plants. Thus efficient incorporations into (-)-anabasine (14) and (-)-N-methylanabasine (15) were recorded and without randomization of the label. However, the specific activity of the N-methylanabasine, unlike that of the anabasine, was the same as the specific activity of the precursor, i.e., no dilution by inactive alkaloid had occurred in the plant. The failure of these plants to produce N-methylanabasine normally was confirmed by g.l.c. analysis of crude alkaloid extracts and also by the failure to dilute it out of the plants after administration of DL-[2-14C]lysine. The formation of N-methylanabasine (15) from N-methyl-Δ1-piperideine (13) in Nicotiana, therefore, is an aberrant one. The synthesis of unnatural alkaloids by Nicotiana had been reported earlier and this then is a further example. The configuration of the N-methylanabasine at C-2′ was found to be the same as that of natural anabasine and nicotine.

It follows that N-methyl-Δ1-piperideine is unlikely to be a normal intermediate in anabasine biosynthesis and its incorporation into nicotine must be the result of the action of a non-specific demethylating enzyme on (13) or (15). The biosynthesis of anabasine (14) probably proceeds therefore by path b, Scheme 4, and the conversion of lysine into (12) without loss of hydrogen from C-2 is accounted for by a reasonable mechanism involving pyridoxal, as shown in Scheme 5.

The idea that ε-N-methyl-lysine (Scheme 4, path a) is a precursor for N-methylpiperidine alkaloids like sedamine (17) must apparently also be abandoned: although ε-N-methyl-lysine has been detected in Sedum acre plants, administration of [methyl-14C]methionine together with [3H]lysine to these plants gave ε-N-methyl-lysine with a 3H : 14C ratio quite different from that of sedamine (17), which indicated therefore that ε-N-methyl-lysine was not a precursor of this alkaloid.

It may be concluded then that of the two pathways outlined in Scheme 4, path a is not followed in the biosynthesis of piperidine alkaloids. Path b is in accord with many of the results obtained for piperidine alkaloid biosynthesis but cannot be regarded as a general hypothesis as it stands, for cernuine (2) and decodine are derived from lysine via a symmetrical intermediate, reasonably cadaverine. It has been suggested that cadaverine (5) may be a normal intermediate common to the biosynthesis of piperidine alkaloids. The degree of dissociation of the enzyme–cadaverine complex formed on decarboxylation of lysine would decide whether the incorporation of lysine into a piperidine alkaloid occurs in non-symmetrical or symmetrical manner. Incorporation of cadaverine into N-methyl-pelletierine (16) has been found to proceed in stereospecific fashion with retention of the pro-R and loss of the pro-S hydrogen at C-1.

By using these ideas the pathway illustrated in Schemes 4 (path b) and 5 can be made general for the biosynthesis of all piperidine alkaloids with the implication of cadaverine as a pyridoxal-linked intermediate by an extension of Scheme 5. The importance of this sequence relative to that outlined in Scheme 5 would decide whether lysine is symmetrically or unsymmetrically incorporated into the piperidine alkaloids.

Nicotine. — These results for the piperidine alkaloids are in marked contrast to those observed with pyrrolidine alkaloids like hygrine (22) and nicotine (21) where the pathways from ornithine proceed via N-methylated derivatives. The route to hygrine (22) and its relatives must necessarily involve non-symmetrical intermediates whereas that to nicotine must involve at least one symmetrical intermediate, i.e. putrescine (18); the proposed biosynthetic pathway to nicotine (21) is illustrated in Scheme 6.

The pathway to nicotine has received further support in the isolation of two enzymes from tobacco roots, one of which, putrescine N-methyltransferase, will catalyse the formation of N-methylputrescine (19) from putrescine (18). The other enzyme which was isolated was N-methylputrescine oxidase. It catalysed the conversion of N-methylputrescine (19) into N-methylpyrrolinium salt (20). This enzyme oxidized putrescine and cadaverine at a rate 40% that found for N-methylputrescine whilst other amines were unaffected, thus showing reasonable specificity for N-methylputrescine.

It has long been known that nicotinic acid [as (23)] is a precursor of the pyridine ring of nicotine (21) and that the pyrrolidine ring becomes attached to the position from which the carboxy-group is lost. The mechanism by which nicotinic acid and the pyrrolidine ring become linked has not been established, however. Germane to a consideration of the mechanism of this reaction are results obtained from feeding experiments with variously tritiated/deuteriated nicotinic acids in excised root cultures of N. tabacum. It was found that [2-3H]-, [4-2H]-, and [5-3H]-nicotinic acids [as (23)] were incorporated approximately ten times more efficiently than [6-3H]nicotinic acid. An explanation for the loss of tritium from C-6 was offered by suggesting that 6-hydroxynicotinic acid was an intermediate between nicotinic acid and nicotine. This was shown to be unlikely when 6-hydroxy[15N]nicotinic acid failed to be incorporated. It was also shown that [6-3H]nicotinic acid did not undergo loss of tritium while present in the culture medium.

It was suggested then that 1,6-dihydronicotinic acid might be an intermediate and that the results could be accounted for by the mechanism shown in Scheme 7; it is necessary to postulate, as is likely, that hydrogen introduction and removal are stereospecific otherwise tritium would be preferentially retained by a primary isotope effect.

The above results warranted further examination which has now been carried out and the results provide confirmation of those obtained earlier. Thus [2-3H]nicotinic acid was found to be a much more efficient precursor (8 — 35 times) for nicotine (21) than [6-3H]nicotinic acid, and the label from [2-3H]-nicotinic acid was essentially all located at C-2 of nicotine (cf. ref. 19). Surprisingly, degradation of the nicotine, obtained after administration of the [6-3H]nicotinic acid, revealed that only 40 — 58% of the residual activity was located at C-6, the remainder being located on the other carbons of the pyridine ring. It was also confirmed by administering doubly labelled nicotinic acid that little of the tritium was lost from C-2 or C-6 of nicotinic acid over the period of the other feedings.

Whilst it is clear that more experiments are needed to establish the mechanism whereby (20) and nicotinic acid [as (23)] become linked, it seems as if a pathway which involves complete loss of tritium from C-6 is the main one involved in nicotine biosynthesis. The residual tritium from C-6 of nicotinic acid could arise by hydrogen shifts not connected with formation of the nicotine skeleton, supported by variable efficiencies of incorporation of [2-3H]- compared with [6-3H]-nicotinic acid, under different experimental conditions.

Recent criticism of a degradation procedure for nicotine obtained after 14CO2 feeding experiments has been challenged by the original authors. They cite their original control experiments, now supported by additional experiments, which unlike the other work demonstrates that formaldehyde obtained by oxidation of NN-dimethylglycine derived from nicotine does not include label from the N-methyl groups. The original discrepancy between 14CO2 and tracer feeding experiments still stands therefore.

Miscellaneous Pyridine Alkaloids. — Preliminary results on the biosynthetic pathway to proferrorosamine A (24), a metabolite of Pseudomonas roseus fluorescens, have been obtained. An exceptionally efficient incorporation of picolinic acid (25) suggested that it was an immediate precursor. DL-[3,4-14C2]Glutamic acid was also incorporated with activity confined to the picolinic acid moiety.

Nicotinic acid and nicotinamide have been found to be precursors for N-methyl-5-carboxamide-2-pyridone (26), a new alkaloid found in young greenhouse-grown Trewia nudiflora. The mechanism of hydroxylation of this and related alkaloids, e.g. ricinine (27), is unknown. In the conversion of nicotinic acid into the related derivative (28) in the bacterium Pseudomonas fluorescens, the oxygen originates from water and not molecular oxygen. Consequently the reaction cannot be mediated by a mixed function oxidase and it represents a so-far unique example of such a hydroxylation in an aromatic system.

Hydrolysis of the complex ester alkaloids present in Tripterygium wilfordii yields either wilfordic acid (29) or hydroxywilfordic acid (30), for which [6-14C]-nicotinic acid and [carbonyl-14C]nicotinamide adenine nucleotide (NAD) serve as efficient precursors. Interestingly, nicotinic acid is more efficiently incorporated into the root alkaloids than NAD, the reverse being observed in the alkaloids of leaves and stems.

Coniine. — Unlike most of the piperidine alkaloids, coniine (31) is derived in Nature from acetate and not lysine. Full details of the fascinating discovery of its mode of biosynthesis have been published. Results additional to those already reviewed are as follows. [1-14C]Hexanoic acid was incorporated into coniine (31), with activity confined to C-4. This is an example of a fairly rarely observed elongation of a medium length fatty acid. A rapid equilibration of coniine (31) and γ-coniceine (32) is known to occur in hemlock and it was observed here that [2′-14C]coniine was incorporated into its biological precursor γ-coniceine with the natural (+)-isomer much more efficiently converted than (-)-coniine.

An enzyme has been isolated from hemlock leaves which catalyses transamination between alanine and 5-oxo-octanal (33), a precursor for coniine, a finding which strengthens the position of 5-oxo-octanal as an intermediate in coniine biosynthesis. The products of the enzymic transformation were γ-coniceine (32) and pyruvic acid. This reaction was shown to be irreversible in vivo when (-)-[15N,1′-14C]coniine [as (31)] and similarly labelled γ-coniceine were re-isolated after 9 days in hemlock without change in isotope ratio. The coniceine efficiently labelled a new alkaloid, conhydrinone (34), discovered during the course of these experiments.

(Continues…)Excerpted from The Alkaloids Volume 4 by J. E. Saxton. Copyright © 1974 The Chemical Society. Excerpted by permission of The Royal Society of Chemistry.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.