New strategies towards synthesis of doxorubicin analogs

Most of the new drug candidates in the anthracycline class of antitumor antibiotics are a result of synthetic efforts involving modification of both the aglycone and sugar moieties. In such an approach, formation of a glycosidic bond is an important step that often becomes a limiting factor in the preparation of certain target structures and can also affect the efficiency of synthetic processes for obtaining analogs of anthracycline antibiotics of clinical interest. We have developed a general approach to de novo glycosylation of anthracycline aglycones leading to doxorubicin analogs with L-lyxo-and L-arabino- monosaccharides. Such glycosylation procedures are also effective in preparation of daunorubicin congeners. Specifically, we have explored the use of 1-O-silylated 3-azido-2,3,6-trideoxy-hexopyranoses as stable glycosyl donors and have successfully demonstrated the practical use of the 3-azido group to generate an amino function during the last steps of synthesis to allow easy generation of doxorubicin analogs. We have also shown that other known glycosyl donors can be conveniently generated from 1-O-silylated-hexopyranoses and can be used effectively to take advantage of a particular glycosidation and deprotection strategy. We describe two standard glycosylation procedures that were designed to attain the desired level of α-glycoside stereoselectivity and overall efficiency. Different glycosylation procedures were selected depending on the sugar synthon configuration and sensitive C-14 substitution in the target anthracycline aglycone. These achievements are exemplified by the use of previously unreported 3-azidosugar synthons of L-lyxo- and L-arabino- configuration protected by a 4-O-chloroacetyl group or an acid-labile triethylsilyl (TES) protecting group.