Because MASS can detect ligands with a wide range of K^s (10 nM to 1 mM) [12, 13, 30], we have developed a drug design approach utilizing the structure-activity relationships (SAR) of weak ligand-target interactions to build ligands that show increased binding affinity. We refer to this as SAR by MS . In this approach, a panel of motifs (small, rigid molecules with molecular mass less than 300 Da) is initially screened against one or more targets. MASS identifies ligands that bind the target and, if two ligands bind the same target at different sites, a ternary complex is observed. Next, simple derivatives of the most interesting motifs are synthesized to provide information about the target-binding site; and these compounds are screened in another round of MASS to further probe the individual and collective affinities of the compounds. For example, if the addition of a chemical group changes a pair of ligands from concurrent to competitive binders, it implies that the additional moiety sterically hinders the binding of the other compound to the target and that the two compounds must share relatively close binding sites. Lastly, cumulative information is used to guide the linking of motifs into a single structure with higher affinity for the target.
The SAR by MS method was used to identify a new class of ligands that bind to the 1061 region of bacterial 23S rRNA which interacts with the L11 protein and is the binding site for the antibiotic thiostrepton . Even though there is a crystal structure for the protein-RNA interaction , traditional structure-based rational drug design approaches are difficult to perform since the interaction between the protein and RNA is complex. Thus, it is an ideal target for the SAR by MS strategy.
A screen of compound libraries containing compounds from commercial sources and RNA-directed combinatorial libraries, revealed two classes of motifs that showed interesting SAR toward the 1061 region of bacterial 23S rRNA. The first class consists of d-amino acids (series A). A positively charged side-group improves binding relative to uncharged and unsubstituted side-chains. The second class consists of the quinoxalin-2,3-diones (series B). Substitutions of the carboxyl groups of the quinoxaline-2,3-dione are well tolerated, with large pendant groups being preferred. Because A and B are structurally different, it was hypothesized that the ligands bind at distinct sites on the target RNA. To further examine the spatial relationships between the different motifs that bind the RNA, MS competition experiments were conducted with the different ligand classes. Ligand A and Ligand B1 were shown to bind the RNA concurrently, as evidenced by the formation of a ternary complex between A, B1 and the RNA (Fig. 10.8). In contrast, Ligand A and Ligand B2 were shown to be competitive binders, as evidenced by the lack of binding of B2 in the presence of A and the lack of a ternary complex (Fig. 10.8).
Based on these competition experiments, it was postulated that the furan portion of A is separated from the carboxyl functional group of B1 by approximately three atoms. To test this hypothesis, several fused compounds were made with different linkages between the furan of A and the carboxyl functional group of B1. The compounds were tested for affinity to the RNA target as well as for their ability to inhibit bacterial transcription/translation in cell-free functional assay (Fig. 10.8b, c). The fused compounds all bound tighter to the target RNA than the parent motifs. The Kd measured by mass spectrometry for the fused compounds were in the range of 6-50 mM, compared with >100 mM for the parent motifs. The rigid biaryl-linked compound AB in Fig. 10.8a shows 20-fold enhanced affinity for the RNA target relative to the motif ligands. More importantly, this compound shows similar activity in a functional assay (IC50 = 14 mM) whereas the motifs are not active in the functional assay . Thus, this compound may bind to the target RNA in a manner that interferes with ribosomal function. While still a relative newcomer to the drug discovery tool kit, SAR by MS appears to be a promising method for ligand-based lead discovery of specific, high affinity ligands which have the potential to have significant therapeutic activities. As with many newly emerging methodologies, time will tell which approaches have ''staying power'' in the drug discovery arena and provide the most value for lead identification and optimization.
Was this article helpful?