Novel acridine-triazenes as prototype combilexins: synthesis, DNA binding, and biological activity.

A series of bifunctional ligands has been developed as prototype DNA-binding combilexins using a DNA template-directed approach. These novel agents contain a 1,3-diaryltriazene linker moiety, present in the established DNA minor groove-binder berenil [1,3-bis(4'-amidinophenyl)-triazene], which is attached to an intercalating acridine chromophore by a functionalized thiazole residue. This 9-arylacridine is predicted to confer rotational freedom to the hybrid molecule and thus facilitate bifunctional interaction with double-stranded DNA through a combination of 'classical' intercalation and minor groove-binding processes. The noncovalent DNA-binding properties of these acridine-triazene combilexins, together with the component molecular fragments, have been examined by fluorescence quenching and thermal denaturation studies with calf thymus DNA and two oligonucleotides, [poly(dA-dT)]2 and [poly(dG-dC)]2. In addition, the binding behaviors of these acridine compounds are compared to those of proflavine (3,6-diaminoacridine) and its 9-phenyl derivative. The results indicate that the hybrid agents (i) are more DNA-affinic than either molecular component, (ii) retain the AT-preferential binding properties of the parent difunctionalized 1,3-diaryltriazene residues, despite weak GC-preferential behavior associated with the acridine chromophore, and (iii) have a reduced binding affinity at pH 7 that reflects the protonation status of the acridine. In contrast, the more basic proflavines show much greater binding affinity and a marked preference for GC-rich DNA sequences. In vitro cytotoxicity data with L1210 mouse leukemia and A2780 human colon cancer cell lines show that the conjugate molecules are approximately 10-40-fold more potent than the acridine or triazene subunits and have activities that compare favorably with those of other reported synthetic combilexins. Intercalative binding modes with a model d(GATACGATAC).d(GTATCGTATC) target duplex have been investigated using molecular modeling techniques. These studies provide a rational basis for the binding properties and suggest that the prototype combilexins can bind in a bimodal manner that induces little distortion of the host DNA duplex. Energy-minimized models for the possible dual interactions are discussed.

Design and synthesis of RNA-specific groove-binding cations: implications for antiviral drug design.

As as initial step in the design of structure-specific RNA-interactive molecules as potential antiviral agents, we have focused on the synthesis of molecules that exhibit strong and preferential binding to duplex RNA. A series of polycationic ligands have been synthesized, and the degree of preferential binding to RNA has initially been determined by thermal denaturation (delta Tm) with both RNA [poly(A).poly(U)] and DNA [poly(dA).poly(dT)] polymers at a variety of pH values. Seven compounds from the series exhibit a substantial degree of RNA-selective binding. The relatively high delta Tm values obtained suggest a specific mode of interaction between these ligands and the RNA helix. By contrast, the much lower delta Tm values with poly(dA).poly(dT) DNA reflect a more nonspecific interaction mode. A viscometric titration study with poly(A).poly(U) confirms that they do not bind by intercalation. The results, combined with the known structure and electronegative potential of duplex RNA, suggest that these molecules bind in the major groove via specific electrostatic and/or hydrogen-bonded interactions.

Design and analysis of RNA structure-specific agents as potential antivirals.

A number of pathogenic RNA viruses, such as HIV-1, have extensive folded RNA conformations with imperfect A-form duplexes that are essential for virus function, and could serve as targets for structure-specific antiviral drugs. A method for the discovery of such drugs involves evaluation of the interactions with RNA of a wide variety of compounds that are known to bind to nucleic acids by different mechanisms. This approach has been initiated by using corresponding sequence RNA and DNA polymers as initial test systems for analysis of RNA binding strength and selectivity. Compounds that bind exclusively in the minor groove in AT sequences of DNA do not have significant interactions with RNA. Polycations, however, can show significant RNA affinity and binding selectivity, probably through complex formation in the RNA major groove. Some intercalators and a group of diphenylfuran cations have strong interactions with RNA that are very dependent on compound structure. RNA hairpin model systems for the RRE binding site of HIV-1 Rev protein were constructed for more detailed investigations. The diphenylfuran cations bind strongly to RRE and selectively inhibit Rev binding. CD, NMR, and fluorescence binding studies indicate that the active compounds bind in the internal loop region of RRE (with binding constants > 10(7)M-1), and cause a conformational change in the RNA. None of the standard nucleic acid binding modes appears to fit the results for complexes of the active compounds with RRE, and it is proposed that the diphenylfuran system threads through the internal loop region of RRE. Such a model allows contacts of the furan cationic substituents with both grooves of RRE in addition to the intercalation interactions with the bases.

Design and analysis of molecular motifs for specific recognition of RNA.

Selective targeting of RNA has become a recent priority in drug design strategies due to the emergence of retroviruses, the need for new antibiotics to counter drug resistance, and our increased awareness of the essential role RNA and RNA structures play in the progression of disease. Most organic compounds known to specifically target RNA are complex, naturally occurring antibiotics that are difficult to synthesize or derivatize and modification of these compounds to optimize interactions with structurally unique RNAs is difficult. The de novo design of synthetically accessible analogues is one possible alternative; however, little is known about the RNA recognition principles on which to design new compounds and limited information on RNA structure in general is available. To contribute to the growing body of knowledge on RNA recognition principles, we have prepared two series of polycationic RNA-binding agents, one with a linear scaffold, the other with a macrocyclic scaffold. We evaluated these compounds for their ability to bind to DNA and RNA, as well as to a specific RNA, the regulatory sequence, RRE, derived from HIV-1, by using thermal melting, circular dichroism, and electrophoresis gel shift methods. Out results suggest that cationic charge centers of high pKa that are displayed along a scaffold of limited flexibility bind preferentially to RNA, most likely within the major groove. Related derivatives that bind more strongly to DNA more closely mimic classical DNA minor-groove binding agents. Several of the macrocyclic polycations expand on a new binding motif where purine bases in duplex RNA are complexed within the macrocyclic cavity, enhancing base-pair opening processes and ultimately destabilizing the RNA duplex. The results in this report should prove a helpful addition to the growing information on molecular motifs that specifically bind to RNA.