From anhydrovinblastine synthesis

The synthesis of Catharine (250), " to which catharinine was initially believed to be closely related, has in fact been achieved by a process which involves the fission of ring D of the velbanamine component of leurosine (249). This conversion was first reported as a result of the accidental over-oxidation that occurred in the preparation of leurosine from anhydrovinblastine by means of t-butyl hydroperoxide in the presence of trifluoroacetic acid. The by-product in this reaction was initially regarded as the 21-lactam related to leurosine, but it has now been recognised as Catharine, and can be prepared equally well by oxidation in the absence of acid (Scheme 41) a radical mechanism appears to be involved. In view of this facile conversion under oxidising conditions, the status of Catharine as a bona fide natural product is open to question. Indeed, the status of leurosine itself as an alkaloid has been questioned, in view of the ease with which anhydrovinblastine is oxidised to leurosine, even in the absence of specific oxidising agents. For example, anhydrovinblastine is oxidised to leurosine if not stored in an inert atmosphere, and the conversion is even more rapid in solution, particularly in the presence of adsorbents such as silica or alumina. A conversion of 40% has been observed after only 72 hours at room temperature. In view of these results it is perhaps not surprising that anhydrovinblastine has not been isolated from any Catharanthus species examined to date. 

There are two especially relevant cases where Cp2TiCl-promoted epoxide deoxygenations have been demonstrated to conform to the requirements of selectivity, mildness, and wide functional group tolerance desirable in natural product synthesis the chemical correlation between cryptophycin-23 and cryptophycin-45 and the synthesis of anhydrovinblastine, the key intermediate in the synthesis of the anticancer drug Navelbine, from leurosine (Scheme 14) [91-93]. 

In 1975, Potier and collaborators proposed that, in planta, the dimeric vinblastine type alkaloids resulted from the coupling of catharanthine and vindoline and, in light of this hypothesis, they reported for the first time the chemical synthesis of a dimer with the natural configuration through a modified Polonovski reaction [18, 19]. This reaction resulted in the formation of an iminium dimer which, after reduction with NaBH4, yielded a-3 ,4 -anhydrovinblastine, Fig. (2), later proved to be the first dimeric biosynthetic precursor of vinblastine in the plant. The group of Potier investigated possible modifications of anhydrovinblastine and produced vinorelbine, Fig. (1), which was the first active derivative with an altered cleavamine (catharanthine) moiety [20, 21]. 

It would thus appear that the presence of an a-acetoxy-group at C-15 severely inhibits the fission of the 16,21-bond in the coupling reaction, since the isovinblastine O-acetate (258) was obtained in yields of only 6 and 4%, respectively, from (257) and (260). The effect of a /3 -acetoxy-group is less well defined Honma and Ban " report the formation of anhydrovinblastine (255), but only as the minor product of the reaction, whereas Kutney and Worth report the formation of (253) and (254), but in unspecified yield. For the synthesis of vinblastine derivatives the absence of a C-15 substituent, as in catharanthine and dihydro-catharanthine, seems preferable for example, catharanthine N-oxide was... 

---