Some New Aspects of Actinomycin D-Nucleic Acid Binding

complexes have yielded many insights into how small molecules affect the structure of DNA (Berman and Young 1981). To date, the most widely used model sys-tems have been 2:1 complexes between dinucleoside phosphates and small planar molecules. With two im-portant exceptions (Seeman et al. 1975; Takusagawa et al. 1982), all the self-complementary sequences have formed intercalative complexes, thus providing a rich data base for the understanding of the structural details of this mode of interaction. DNA is the cellular target for actinomycin D (AMD) (Fig. 1), a widely used biochemical tool (Goldberg and Friedman 1971; Kersten and Kersten 1974) with specific chemotherapeutic a tivity (Farber 1966). However, its exact mode of binding is not as well understood as those of more simple molecules. For ex-ample, solution studies of complexes between AMD and a variety of oligonucleotides have indicated that its planar chromophore ing intercalates between base pairs in the double helix with the preferred binding sequence being GC (Patel 1974; Krugh and Chen 1975). The results of these studies also suggest that the conforma-tion of the phosphodiester linkage is different from that observed in solutions of planar drug-oligomer com-plexes. Our original objective in undertaking the crystal-structure analysis of d(GpC)-actinomycin was to delineate these conformational differences. Instead, we discovered a unique and unexpected structure that we have called pseudointercalated. This mode of bindin