Carbamate prodrug concept for hydroxamate HDAC inhibitors.

Reversible acetylation of histones and other proteins has emerged over the last 10 years as an important mechanism for cell proliferation and has been identified as a valuable target for anticancer drug design. Acetylation is executed and maintained by the histone acetyltransferases and reversed by their counterparts, histone deacetylases (HDACs). The first HDAC inhibitors have already been approved for therapeutic use, and many additional clinical studies are currently under way. Herein, we describe virtual screening efforts that identified several novel HDAC6 inhibitors with cellular isoform selectivity. In particular, a carbamate-protected hydroxamic acid exhibited improved effects with respect to protein hyperacetylation compared with the parent hydroxamate, possibly because of improved cell permeability. The carbamate structure therefore represents a potential prodrug concept for hydroxamic acidcontaining HDAC inhibitors. HDACs are zinc-dependent amidohydrolases, and 11 human subtypes are known. Among the dozens of nonhistone substrates of HDACs, tubulin has attracted a lot of attention because it is a validated target for established anticancer drugs, such as taxanes, the vinca-alkaloids and their derivatives. Tubulin is deacetylated by a single zinc-dependent HDAC subtype, HDAC6, and by the NAD-dependent histone deacetylase Sirt2. HDAC6-specific inhibitors and nonselective HDAC inhibitors synergize in cytotoxicity with the proteasome inhibitor bortezomib (Velcade), which makes HDAC6 an interesting target for inhibitor development. So far, reports of HDAC6selective inhibitors are limited compared to nonselective or class I-selective inhibitors. Aiming at novel selective HDAC6 inhibitors, we were particularly interested in whether structure-based virtual screening using a HDAC6 homology model could deliver such compounds. We recently reported a homology model for HDAC6 used for docking studies of hydroxamates. The model was generated by exploiting multiple solved crystal structures of related HDACs as templates. Based on a substructure search for hydroxamates and related zinc binding groups, we identified 252 potential compounds by virtually screening the Maybridge compound collection comprising 55000 molecules. The docking and subsequent visual inspection of the docking poses showed that, among the compounds with the highest docking scores, several hydroxamate and hydrazide derivatives could be identified. Finally, the five top-ranked compounds (1–5) were selected for in vitro testing.