ALKALOIDS 255 



Similarly, the participation of methylamine as such is not essential. The actual reactant 

 could be glycine followed by a decarboxylation. Thus the Robinson proposals are not 

 tied to specific compounds. Rather, the pyrrolidine ring may be thought of as derived 

 from any one of a group of related compounds (i. e. ornithine, glutamic acid, succinic 

 acid) which have structural similarities and which are now known to be biochemically in- 

 terrelated as well. The first specific derivative of say, ornithine which is irrevocably 

 committed to forming an alkaloid rather than some other compound remains unknown. 

 In fact, it would seem that the crucial problem in alkaloid biosynthesis is the identifica- 

 tion of the point in the pathways of metabolism where an intermediate is formed whose 

 subsequent transformations are directed solely toward an alkaloid with no links to other 

 classes of compounds. 



The types of compounds suggested by Robinson do occur in plants, and the required 

 reactions take place so readily that large structures similar to alkaloid skeletons may be 

 formed in vitro by mere mixing of the supposed precursors. Qn the other hand, Wenkert 

 (43, 44) believes that alkaloid formation may be more closely related to carbohydrate and 

 acetate metabolism than to amino acid metabolism. 



Experiments have been carried out where supposed alkaloid precursors were labelled 

 with isotopic tracers and fed to plants. Isolated alkaloids were then found to have the 

 label just where it would be predicted by the theory. On the whole, then, the pathways in- 

 dicated serve, at least, as helpful guides. What biochemical evidence is available sug- 

 gests that in the main they give a picture close to the truth. However, specific aspects 

 of alkaloid biosynthesis can be discussed in more detail and biochemical evidence for or 

 against steps presented. There are reviews on alkaloid biosynthesis by Battersby (45) 

 Marion (46), Mothes (47) and Franck (48). Mothes has also reviewed methods for in- 

 vestigating alkaloid biosynthesis (49, 50). 



The clearest generalization concerning alkaloid precursors concerns the methyl 

 groups found in many alkaloids, both as N -methyl and 0-methyl groups. Tracer experi- 

 ments have, almost without exception, shown that, as might be expected, these are de- 

 rived from the methyl group of methionine (51, 52, 53). Enzyme systems from various 

 plants have been shown to catalyze the transfer of methyl groups from S-adenosylmethi- 

 onine to form gramine, hordenine and trigonelline (54, 55). Byerrum et al. (56) have 

 found, however, that glycine -2 -C^"* was at least as good a donor for the methyl group of 

 nicotine. Formate does not appear to be an efficient precursor in this system although 

 it apparently works well as a methyl donor for ephedrine formation (57). Choline and 

 betaine have also been found to act as methyl donors in some cases (46). 



The participation of C3, C4 and C5 compounds at various points in the scheme is 

 evident, and these participants are most likely interrelated — for example, as acetone, 

 acetoacetic acid and acetonedicarboxylic acid. These three compounds, although readily 

 derived from plant components by well-known enzymatic reactions, are found, if at all, 

 in very small amounts. Kaczkowski et al. (58) have shown by tracer experiments that 

 the "C3 unit" of hyoscyamine is derived from two molecules of acetate with loss of one 

 carboxyl group. Tuppy and Faltaous (59) have described an enzyme system which can 

 utilize acetoacetic or acetonedicarboxylic acid for synthesis of hygrine and other alka- 

 loids. 



Quite good tracer evidence exists for the formation of the pyrrolidine ring from 

 ornithine (keeping in mind that the ornithine itself probably does not immediately cyclize). 

 This generalization apparently holds whether we are dealing with an isolated ring as in 

 nicotine or a more complex structure as in hyoscyamine (60). The degradation data of 

 Leete and Siegfried (61) showed equal activity in the a-carbons of the pyrrolidine ring of 

 nicotine when ornithine-2-C" was fed. This indicates the existence of a symmetrical in- 

 termediate such as putrescine or succindialdehyde. However only one bridgehead carbon 

 in hyoscyamine received tracer from ornithine -2 -C*^ (60). Mann and Smithies (62) have 



