Vinca alkaloids have been widely used in cancer chemotherapy for over 30 years. 1 Extensive chemistry research has led to several total syntheses, 2 and numerous derivatives have been evaluated, with the aim of improving the therapeutic potency of this class. 3 However, only four drugs are currently available worldwide, namely vinblastine (1a), vincristine (1b), a semisynthetic amide related to vinblastine, vindesine 4 (1c, Scheme 1), and vinorelbine (2, Navelbine).Vinorelbine (2) was obtained in two steps: (1) biomimetic coupling of the two precursor monomers, 5 catharanthine and vindoline, to form 3′,4′-anhydrovinblastine (3), and (2) C′ ring contraction of this intermediate 6 (Scheme 2).In our search for new and more potent vinorelbine derivatives, we were interested in an original chemical approach, which conceivably could induce dramatic changes in the skeleton of the molecule. We decided to investigate the reactivity of these highly functionalized compounds in superacid media. Superacids are able to induce modifications at nonactivated bonds, 7 and in these unusual conditions, indolines and indoles were found stable enough to react with various electrophiles. 8 The effects and reactivity of various electrophiles were investigated. Among these, chloromethanes (CH 2 Cl 2 , CHCl 3 , and CCl 4 ) act as superelectrophiles in HF-SbF 5 , the resulting cations CH 2 Cl + , CHCl 2 + , and CCl 3 + , respectively, exhibiting an extremely reactive hydride-abstracting power. 9Vinorelbine (2) was treated with CCl 4 in HF-SbF 5 at -40 °C. After workup, the main isolated product was identified as

The clinical interest of Vinca alkaloids was clearly identified as early as 1965 and so this class of compounds has been used as anticancer agents for more than 30 years. Today, two natural compounds, vinblastine and vincristine and two semi-synthetic derivatives, vindesine and vinorelbine, have been registered and thus Vinca alkaloids can be considered to represent a chemical class of definite utility in cancer chemotherapy. Today, relatively few groups actively research in the chemistry of Vinca alkaloids. However, using superacidic chemistry, a new family of such compounds was synthesised and vinflunine, a difluorinated derivative, was selected for clinical testing. A consideration of the pharmacological data relating to these new derivatives appears to reveal a lack of any marked correlation between in vitro and in vivo results. Furthermore, structure/activity relationships have failed to assist the chemist in the rational design. Such rational design of new derivatives is limited by the fact that the Vinca binding site(s) on tubulin and the exact mechanism(s) of action of Vinca alkaloids remain unclear. Nevertheless, the preclinical evaluations of the new derivative vinflunine have already suggested that certain in vitro assays, in addition to in vivo experiments, could be proposed to select more rationally newer generation Vincas. Moreover, recent studies have demonstrated that certain newly identified properties, such as antiangiogenic activities, could enlarge the therapeutic usage of natural and semi-synthetic Vinca alkaloids. Thus, Vinca alkaloids remain a drug family with a continuing interest for future anticancer therapy.

The ability of a class of C-20' modified vinca alkaloid congeners to induce tubulin spiral formation was investigated relative to their ability to inhibit microtubule assembly, their cytotoxicity against a leukemic cell line, L1210, and their measured and calculated partition coefficients. These studies were prompted by the observation that the energetics of vinca alkaloid-induced tubulin spiral polymers, or spiraling potential, is inversely related to their clinical dosage and are aimed at the long-term goal of developing the ability to predict the cytotoxic and antineoplastic properties of antimitotic drugs. We demonstrate here that vinca-induced tubulin-spiraling potential is significantly correlated with cytotoxicity against L1210 cells. This is consistent with the size of spirals formed being proportional to the relaxation time for polymer redistribution, the lifetime of cell retention, and effects on microtubule ends and dynamics. Spiraling potential also correlates with calculated but not measured partition coefficients. Surprisingly, spiraling potential does not correlate with the ability to inhibit microtubule formation with purified tubulin or microtubule protein. For the set of C-20' modified compounds studied, the largest inhibitory effects on spiraling potential and cytotoxicity are caused by multiple sites of halogen (-F, -Cl) substitution with the introduction of increased rigidity in the ring. This suggests the C-20' position interacts with a hydrogen bond acceptor or an electrophilic region on the protein that electrostatically disfavors halogen substitutions. These studies are discussed in terms of the cellular mode of action of antimitotic drugs, particularly the importance of microtubule dynamics during mitosis and the factors that regulate those dynamics.