The Alkaloids Volume 11

A Review of the Literature Published between July 1979 and June 1980

By M. F. Grundon

The Royal Society of Chemistry

Copyright © 1981 The Royal Society of Chemistry
All rights reserved.
ISBN: 978-0-85186-347-4

Contents

Chapter 1 Biosynthesis By R. B. Herbert, 1,
Chapter 2 Pyrrolidine, Piperidine, and Pyridine Alkaloids By A. R. Pinder, 29,
Chapter 3 Tropane Alkaloids By G. Fodor and R. Dharanipragada, 36,
Chapter 4 Pyrrolizidine Alkaloids By D. J. Robins, 44,
Chapter 5 Indolizidine Alkaloids By J. A. Lamberton, 59,
Chapter 6 Ouinolizidine Alkaloids By M. F. Grundon, 63,
Chapter 7 Ouinoline, Ouinazoline, and Acridone Alkaloids By M. F. Grundon, 71,
Chapter 8 β-Phenylethylamines and the lsoquinoline Alkaloids By K. W. Bentley, 78,
Chapter 9 Aporphinoid Alkaloids By M. Shamma, 117,
Chapter 10 Amaryllidaceae Alkaloids By M. F. Grundon, 131,
Chapter 11 Erythrina and Related Alkaloids By A. S. Chawla and A H. Jackson, 137,
Chapter 12 Indole Alkaloids By J. E. Saxton, 145,
Chapter 13 Lycopodium Alkaloids By W. A. Ayer, 199,
Chapter 14 Diterpenoid Alkaloids By S. W. Pelletier and S. W. Page, 203,
Chapter 15 Steroidal Alkaloids By D. M. Harrison, 225,
Chapter 16 Miscellaneous Alkaliods, 238,
Author Index, 245,

CHAPTER 1

Biosynthesis

BY R. B. HERBERT

References to earlier Reports in this series are included in the text. Two additional comprehensive reviews are also cited.1·2

1 Pyrrolidine, Pyridine, and Piperidine Alkaloids

Hygrine. — The alkaloid hygrine (3) is an intermediate in the formation of tropane bases. Biosynthesis is from acetic acid, plausibly via acetoacetic acid (cf. Vol. 10, p. 12), and from ornithine (1), very reasonably in the manner shown in Scheme 1. Acetoacetate has been confirmed as an intact precursor for hygrine in experiments with [3-14C]- and [4-14C]-acetoacetic acid in Nicandra physaloides. Labelling in (3) was, respectively, of C-2′ and C-3′, which confirms the suspected orientation of acetoacetate in hygrine (see Scheme 1).

Nicotine. — The pyrrolidine ring of nicotine (6) derives from ornithine (1), label from, e.g., C-2 appearing equally spread over C-2′ and C-5′. This symmetrical incorporation of the precursor amino-acid is accounted for by the intermediacy of the symmetrical diamine putrescine (4), which is supported by other evidence too.1·2 The symmetrical incorporation of ornithine into nicotine (6) and into nornicotine (7) has been confirmed by the results4 of experiments with [2,3-13C2]ornithine [as (1)], thus reinforcing earlier 14C and 13C results (cf. Vol. 8, p. 5; Vol. 10, p. 14 ). Equal labelling of C-2′, C-3′ and of C-4′, C-5′ was observed.

The 13C-labelled ornithine was prepared so that individual molecules bore both labels. Consequently the 13C n.m.r. resonances appeared as doublets. 13C Likewise, the 13C signals due to precursor label in the derived alkaloid appeared as (low-intensity) doublets on either side of the natural-abundance singlets, and were thus separate from them. This allowed an estimation of the extent to which label was incorporated into each site, even though that extent was low (0.05 — 0.07% ). Clearly, this novel application of 13C-labelling will find use elsewhere, since incorporation of a precursor can be measured at high dilution (ca 4000 times).

Enzymic and tracer evidence indicates strongly that the conversion of putrescine (4) into nicotine (6), in Nicotiana tabacum, involves first N-methylation, to give (5), and then oxidation, to give (2) as an intermediate. Enzymes responsible for these steps have been identified in N. tabacum, being called, respectively, putrescine N-methyltransferase and N-methylputrescine oxidase (cf. Vol. 4, p. 7). The levels of these two enzymes in four closely related genotypes of N. tabacum have been found to be proportional to the levels of nicotine (6), thus strongly supporting a role for such enzymes in normal nicotine biosynthesis. A further enzyme, quinolinic acid phosphoribosyltransferase, has been identified as involved in nicotine biosynthesis, being responsive to demand for the nicotinic acid that is required for biosynthesis.

No differences were observed in the enzymic oxidation of putrescine and N-methylputrescine by plant extracts of a cultivar of N. tabacum that had a high nicotine content and one with a high content of nornicotine (7). Thus a high nornicotine (7) content cannot be attributed to direct oxidation of putrescine, and this supports evidence which shows nornicotine (7) to be a demethylation product of nicotine.

Pyrrolizidine Alkaloids. — The necic acid component of senecionine (8) derives from two molecules of isoleucine, radioactivity from precursor amino-acid being equally incorporated into both halves of the necic acid fragment, as shown in Scheme 2 (cf. Vol. 9, p. 4). It has now been shown that biotransformation of isoleucine into the necic acid involves loss of half of a tritium label from C-4 in each of the two amino-acid fragments. Removal of a proton is, therefore, stereospecific, and oxidation at C-4 does not proceed beyond the two-electron level; i.e., a higher intermediate oxidation level, corresponding to a ketone, is excluded. Further results indicate that for each molecule of isoleucine it is the 4-pro-S proton [see (14)] which is lost.

Threonine (10) and 2-oxobutanoic acid (11) are sequential precursors for isoleucine (14), the isoleucine skeleton being derived from (11) by a 1,2-ethyl shift within (12). Further results have been obtained for pyrrolizidine alkaloids with (13), which is a logical precursor for (11) by simple transamination. Both enantiomers of (13) were found to be comparably good precursors for the necic acid fragments of retrorsine (9), and it was found (but not rigorously proved) that biosynthesis was accompanied by loss of the 3-pro-S proton of (13) in each of the two isoleucine fragments. This, taken with the isoleucine results above, indicates that migration of the ethyl group within (12) proceeds with retention of configuration, which is in accord with predictions based on orbital-symmetry considerations.

Lythraceae Alkaloids. — Results showing that lysine (15) is a precursor for decodine (22) and decinine (23) in Decodon verticillatus, which were published in preliminary form (cf. Vol. 1, p. 6), are now available in a full paper. Label from either C-2 or C-6 of the amino-acid was found to be spread equally over C-5 and C-9 of the alkaloids, indicating that ring A derived from this amino-acid and that incorporation was via a symmetrical intermediate. Cadaverine (16), formed by decarboxylation of (15), is a logical candidate for this symmetrical intermediate. In support, [1,5-14C2]cadaverine gave radioactive decodine (22). Labelling was significantly and appropriately of C-5 and C-9. In the biosynthesis of other piperidine alkaloids, Δ1-piperideine (17) is an intermediate after lysine and cadaverine. Material that was labelled on C-6 has been found also to be a precursor for decodine (22) with specific labelling of C-9, as expected.

A key stage in the biosynthesis of piperidine alkaloids is reached with the formation of Δ1-piperideine. For the elaboration of diverse alkaloids, this intermediate undergoes condensation with a variety of nucleophiles, commonly a β-keto-acid. (A similar situation is found for pyrrolidine alkaloid biosynthesis; see, e.g., Scheme l). Existing evidence on Lythraceae alkaloid biosynthesis, taken up again below, indicated that condensation occurred in this case between Δ1-piperideine (17) and acetoacetic acid to give pelletierine (26), further elaboration yielding alkaloids like (22). In the event, however, labelled pelletierine was found not to be a precursor for (22) or (23). Negative evidence is always difficult to interpret, but is here made persuasive by the fact that other precursors that were fed concurrently were incorporated. Conclusive support for these results depended on others outlined below.

Both decodine (22) and cryogenine (25) (in another plant) were known (cf. Vol. 4, p. 13) to incorporate label from [3-14C]phenylalanine [as (20)] into C-1 and C-13. Other results indicated that label from [1-14C]phenylalanine was located at C-11 but not C-3. This is what strongly intimated that pelletierine (26) could be an alkaloid precursor, with its side-chain (rather than that of phenylalanine) providing C-4, C-3, and (possibly) C-2. Results of a complete set of feeding experiments with variously labelled samples of phenylalanine and accompanying careful degradation of (22) and (23) has shown that C-3 does derive from C-1 of phenylalanine, as well as confirming the derivation of the other atoms from this aromatic amino-acid. The negative results that were obtained with pelletierine (26) fall into place.

The nature of the nucleophile which condenses with Δ1-piperideine (17) needs to be reconsidered. Very plausibly, this could be (18), which is formed as shown in Scheme 3 from phenylalanine (20) via cinnamic acid (19) and malonyl-CoA. A further unit of this type is found in alkaloids such as lythrumine (24). An outline biosynthetic route to Lythraceae alkaloids is given in Scheme 3.

Quinolizidine Alkaloids. — Biosynthesis of quinolizidine alkaloids, e.g. sparteine (28), is from lysine (15) by way of cadaverine (16), as shown in Scheme 4. Three cadaverine units (as indicated by the thickened bonds) are required for the construction of alkaloids such as sparteine (28). Although something has been discerned about the biosynthetic relationships of various quinolizidine alkaloids, the nature of early intermediates beyond cadaverine has remained quite elusive. Exciting new results obtained with crude enzyme preparations from cell suspension cultures of Lupinus polyphyllus indicate why this is so.

The crude enzyme preparation was found to catalyse the conversion of cadaverine (16) mainly into 17-oxosparteine (27) in the presence of pyruvic acid. The pyruvic acid served as a receptor for the amino-groups of (16) in a transamination reaction, having manifestly a close relationship to alkaloid formation. Diamine oxidase activity might have been expected to account for the transformation of cadaverine into alkaloid, but the use of an inhibitor for this enzyme, so far from resulting in a decrease in alkaloid formation, actually led to an increase. A diamine oxidase is thus not involved in alkaloid formation.

Δ1-Piperideine (17) has been shown to be a precursor of quinolizidine alkaloids in whole plants (cf. Vol. 8, p. 3). However, neither it nor its self-condensation products could be detected as products in the enzymic reaction. [This conclusion is not completely unambiguous, albeit reasonably safe, because the products of the reaction of diamine oxidase, the first of which is (17), were simply compared with those of the alkaloid synthase reaction by g.l.c., and the products of the two reactions were found to be different]. It seems likely at this stage that (17) is not normally implicated in quinolizidine biosynthesis but can be substituted for an enzyme-generated intermediate via its open form (32) (see Scheme 5). Since no intermediates earlier than (27) could be detected, it is suggested that biosynthesis in vitro and in vivo proceeds by a series of enzyme-linked intermediates (see Scheme 5), none of which is desorbed from the enzyme or enzyme-complex until (27) is liberated. However, in some plants, biosynthesis must stop with the liberation of a compound (31), having the lupinine skeleton (29), to allow for the formation of alkaloids of this type.

In some incubations, sparteine (28) was also formed, and it is suggested to derive from (27); sparteine (28) is known to be a precursor for other alkaloids in whole plants. Experiments with 14CO2 in whole plants suggest that sparteine (28) and lupanine (30), a closely related alkaloid, have a separate genesis, possibly with a dehydrosparteine [formed from (27)] as a common intermediate (cf. Vol. 2, p. 26). This is supported by unpublished observations with incubations of the crude alkaloid synthase.

Quinolizidine alkaloids are associated with plant chloroplasts, which suggests that chloroplasts might be involved in alkaloid biosynthesis. This has been substantiated by further results obtained with chloroplasts from L. polyphyllus.

Incubation of a crude chloroplast preparation with cadaverine yielded lupanine (30) as the main alkaloid. The production of 17-oxosparteine (27), at a lower level, was also observed. Instead of lupanine (30), chloroplasts that had been treated with digitonine synthesized sparteine (28) and a small quantity of another base corresponding to a dehydro-oxosparteine. Enzyme that had been solubilized from the chloroplasts produced 17-oxosparteine (27), which was also a lupanine precursor in chloroplasts. Correlation is thus given to the enzyme work discussed above. (Alkaloid synthesis in chloroplasts again does not involve a diamine oxidase). 17-0xosparteine is therefore to be assigned a key role in alkaloid biosynthesis in plants (see Schemes 4 and 5). In addition, it is clear that intact membranes are important for the normal biosynthesis of quinolizidine alkaloids.

Slaframine. — Slaframine (37) is produced by the phytopathogen Rhizoctonia leguminicola. It has been known for some time that (37) derives in part from lysine via pipecolic acid (33), which is incorporated intact; the earliest bicyclic intermediate identified is (38) (cf. Vol. 5, p. 9 and ref. 2). New results have shown that the two skeletal carbons in (37), and also in the metabolite (36), not accounted for by pipecolic acid, derive from malonate (and acetate). The labelling of (37) by, in particular, [2-2H2]acetate was deduced to be of C-2 on the basis of mass spectral evidence (which is not entirely convincing). The acyl-CoA derivative (34) has been suggested as an intermediate in the biosynthesis of (37) and also of (36). It is to be noted that condensation between malonyl-CoA and pipecolic acid (33) to give (34) must be simultaneous with decarboxylation of malonyl-CoA, since two deuterium atoms of acetate are retained at C-2 in (37) (later intermediates with a double-bond to C-2 are also excluded by these results).

The lactam (39), formed by cyclization of (34), is not a slaframine precursor, indicating that cyclization is of the derived aldehyde (35). The earlier conclusion, i.e. that the alcohol corresponding to (38) is an intermediate in slaframine biosynthesis, has been confirmed.14

Caerulomycin A. — Caerulomycin A (40) is one of a unique group of metabolites from Streptomyces caeruleus. Study of the biosynthesis of (40) has shown that it derives in part from lysine (15) and acetate. Labelling of the unsubstituted pyridyl ring in (40) by [l-13C]-and [l,2-13C2]-acetate was consistent with biosynthesis through lysine and an earlier symmetrical intermediate like (2S,6S)-2,6-diaminopimelic acid (41) (superposition of two equal amounts of unsymmetrical labelling). Lysine was a highly efficient precursor, and the conversion of (41) into lysine is irreversible, so lysine has been deduced to be a direct precursor for caerulomycin A. The pyridine ring in proferrorosamine A (42) also derives from lysine, in this case via picolinic acid (43) (cf. Vol. 4, p. 9; Vol. 5, p. 11), and this compound may plausibly be an intermediate in the production of caerulomycin A too.

C-2 of the substituted pyridyl ring in (40) derives from lysine, and C-3 and C-4 from C-2 and C-1 of acetate, respectively, but the origin of the remainder is obscure (the O-methyl group derives from methionine). Sources likely to provide C2 and C1 units gave negative results; a source which will provide a C3 unit (dihydroxyacetone phosphate?) seems likely.

2 Phenethylamine and Isoquinoline Alkaloids

Normacromerine. — Normacromerine (44) has previously been shown to derive from tyrosine and tyramine (cf. Vol. 9, p. 7; Vol. 10, p. 15). New results have shown that N-methyltyramine is an efficient and specific precursor too.

Nornuciferine-I. — Nornuciferine-I (54) is one example of an aporphine alkaloid amongst many. A key step in the biosynthesis of aporphines is ring-closure within a benzylisoquinoline [as (46)] by an oxidative coupling reaction which involves the phenolic functions on the two rings. Roemerine (51) is like (54) in having ring D devoid of oxygen. Loss of oxygen occurs during roemerine biosynthesis by rearrangement of the dienol (50). This intermediate derives from N-methylcoclaurine (47) by way of (49). A similar biogenesis can be expected for (54), and this has turned out to be the case. Tyrosine, norcoclaurine (45), coclaurine (46), and N-methylcoclaurine (47) were satisfactorily incorporated into (54) in Croton sparsiflorus [(48) was a very poor precursor, showing that N-methylation does not precede O-methylation]. The specificity of the incorporation of N-methylcoclaurine (label on C-3) was established by degradation (N-methyl label also retained), and it was also shown that the (S)-isomer of (47), with the same stereochemistry as (54), was a much better precursor for (54) than its enantiomer. In accord with roemerine biosynthesis, N-methylcrot-sparine (52) and (53) were also found to be precursors for (54), one of the C-10 epimeric alcohols (53) being used with significantly greater efficiency than the other. The detail of nornuciferine-1 biosynthesis is completed by the observation that (47) and (52) are natural constituents of C. sparsiflorus. (For work on the biosynthesis of other alkaloids in this plant, see Vol. 6, p. 19).

(Continues…)Excerpted from The Alkaloids Volume 11 by M. F. Grundon. Copyright © 1981 The Royal Society of Chemistry. 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.