Virtual and experimental high-throughput screening (HTS) in search of novel inosine 5′-monophosphate dehydrogenase II (IMPDH II) inhibitors

IMPDH (Inosine 5'-monophosphate dehydrogenase) catalyzes a rate-limiting step in the de novo biosynthesis of guanine nucleotides. IMPDH inhibition in sensitive cell types (e.g., lymphocytes) blocks proliferation (by blocking RNA and DNA synthesis as a result of decreased cellular levels of guanine nucleotides). This makes it an interesting target for cancer and autoimmune disorders. Currently available IMPDH inhibitors such as mycophenolic acid (MPA, uncompetitive inhibitor) and nucleoside analogs (e.g., ribavirin, competitive inhibitor after intracellular activation by phosphorylation) have unfavorable tolerability profiles which limit their use. Hence, the quest for novel IMPDH inhibitors continues. In the present study, a ligand-based virtual screening using IMPDH inhibitor pharmacophore models was performed on in-house compound collection. A total of 50,000 virtual hits were selected for primary screen using in vitro IMPDH II inhibition up to 10 mu M. The list of 2,500 hits (with > 70 % inhibition) was further subjected to hit confirmation for the determination of IC50 values. The hits obtained were further clustered using maximum common substructure based formalism resulting in 90 classes and 7 singletons. A thorough inspection of these yielded 7 interesting classes in terms of mini-SAR with IC50 values ranging from 0.163 mu M to little over 25 mu M. The average ligand efficiency was found to be 0.3 for the best class. The classes thus discovered represent structurally novel chemotypes which can be taken up for further development.