Structure-Based Discovery of Ligands Targeted to the RNA Double Helix † Qi Chen, Richard H. Shafer,* and Irwin D. Kuntz* Department of Pharmaceutical Chemistry, School of Pharmacy, UniVersity of California, San Francisco, California 94143-0446 ReceiVed April 1, 1997; ReVised Manuscript ReceiVed July 3, 1997 X ABSTRACT: Ligands capable of specific recognition of RNA structures are of interest in terms of the principles of molecular recognition as well as potential chemotherapeutic applications. We have approached the problem of identifying small molecules with binding specificity for the RNA double helix through application of the DOCK program [Kuntz, I. D., Meng, E. C., and Shoichet, B. K. (1994) Acc. Chem. Res. 27, 117-123], a structure-based method for drug discovery. A series of lead compounds was generated through a database search for ligands with shape complementarity to the RNA deep major groove. Compounds were then evaluated with regard to their fit into the minor groove of B DNA. Those compounds predicted to have an optimal fit to the RNA groove and strong discrimination against DNA were examined experimentally. Of the 11 compounds tested, 3, all aminoglycosides, exhibited pronounced stabilization of RNA duplexes against thermal denaturation with only marginal effects on DNA duplexes. One compound, lividomycin, was examined further, and shown to facilitate the ethanol-induced B to A transition in calf thymus DNA. Fluorine NMR solvent isotope shift measurements on RNA duplexes containing 5-fluorouracil provided evidence that lividomycin binds in the RNA major groove. Taken together, these results indicate that lividomycin recognizes the general features of the A conformation of nucleic acids through deep groove binding, confirming the predictions of our DOCK analysis. This approach may be of general utility for identifying ligands possessing specificity for additional RNA structures as well as other nucleic acid structural motifs. Molecular recognition of nucleic acids by small molecules can occur by virtue of binding specificity at the level of primary, secondary, or tertiary structure. By far, the majority of studies to date have focused on the first of these possibilities, i.e., specificity of binding via readout of a particular base or base pair composition and/or sequence in DNA. Some compounds possess a composite specificity involving recognition of both primary and secondary struc- tural features, such as preferential binding to A-T base pairs in the DNA minor groove, with little or no binding to corresponding RNA sequences. Few ligands, however, exhibit the reverse secondary structure specificity, i.e., binding to RNA in one of its grooves but not to DNA. Because of the many diseases caused by RNA viruses, including AIDS, compounds capable of specific binding to RNA should be considered in developing effective chemo- therapy. Bacterial disease sources are also susceptible to this approach to drug treatment. The duplex RNA motif appears in many different contexts. First, a small number of viruses carry a double-stranded RNA genome. Second, some RNA viruses are based on single RNA strands but require synthesis of a complementary RNA strand as part of their life cycle. Third, retroviruses such as RNA tumor viruses and HIV contain single strands of RNA that are often folded into a variety of secondary and tertiary conformations, including local regions of duplex structure, some of which may be distorted due to base mismatches, bulges, etc. Finally, rRNA and mRNA from all sources involve similar motifs of secondary and tertiary folding. Thus, the op- portunities for developing therapeutic agents targeted to regular or distorted duplex RNA structures are manifold. For this approach to succeed, it is apparent that such agents should possess specificity for the RNA target in comparison to DNA, in order to avoid unwanted side effects. There are relatively few studies on small molecules capable of specific RNA binding. Wilson et al. (1) examined a wide range of DNA-binding compounds in a study aimed at delineating the effects of ligand structure on binding affinity to RNA and DNA duplexes. While many of these com- pounds did show significant RNA binding, none showed greater affinity to the RNA duplex than the DNA duplex. Subsequently, Wilson and co-workers synthesized a series of polycationic ligands, some of which demonstrated sig- nificant preference for DNA over RNA, as measured by thermal denaturation studies (2). Both charge and steric effects were observed to play a role in groove binding. Based on the positive charge on these ligands, along with the fact that the negative charge density is greatest in the major groove of A-like duplexes, it was inferred that binding occurred in the major (deep) groove of the RNA duplex. Here we describe an approach for discovery of compounds possessing specificity for the RNA double helix based on the unique geometry of its deep major groove. Using the DOCK methodology (3), we have identified several ami- noglycosides as candidate ligands, characterized by shape complementarity to the RNA groove. We show that one of these compounds not only binds preferentially to RNA over B-form DNA but also facilitates the B to A transition in calf thymus DNA. We also provide preliminary NMR evidence that this ligand binds in the targeted RNA major groove. † This work was supported by NIH Grants GM51650 (R.H.S.) and GM31497 (I.D.K.), National Institute of General Medical Science. The UCSF Computer Graphics Laboratory is supported by Grant RR01081, Division of Research Resources, NIH. * Authors to whom correspondence should be addressed. X Abstract published in AdVance ACS Abstracts, August 15, 1997. 11402 Biochemistry 1997, 36, 11402-11407 S0006-2960(97)00756-3 CCC: $14.00 © 1997 American Chemical Society