Bioinformatic Searching for Optimal RNA Targets of Dimeric Compounds Informs Design of a MicroRNA-27a Inhibitor

Various studies have shown that selective molecular recognition of RNA targets by small molecules in cells, although challenging, is indeed possible. One facile strategy to enhance selectivity and potency is binding two or more sites within an RNA simultaneously with a single molecule. To simplify the identification of targets amenable to such a strategy, we informatically mined all human microRNA (miRNA) precursors to identify those with two proximal noncanonically paired sites. We selected oncogenic microRNA-27a (miR-27a) for further study as a lead molecule binds its Drosha site and a nearby internal loop, affording a homodimer that potently and specifically inhibits miR-27a processing in both breast cancer and prostate cancer cells. This reduction of mature miR-27a ameliorates an oncogenic cellular phenotype with nanomolar activity. Collectively, these studies demonstrate that synergistic bioinformatic and experimental approaches can define targets that may be more amenable to small molecule targeting than others.

Automated summary

Studies have shown that selective molecular recognition of RNA targets by small molecules in cells is possible, with one strategy being the simultaneous binding of two or more sites within RNA with a single molecule. By informatically mining human miRNA precursors, targets amenable to small molecule targeting can be identified, with oncogenic miRNA-27a being a lead molecule that can inhibit miRNA processing in cancer cells. This demonstrates that a synergistic approach involving bioinformatics and experimental methods can define suitable targets for small molecule targeting.