Investigation of the histone deacetylase inhibitor potential of phorbazole analogues

Histone deacetylases (HDACs) are enzymes that remove acyl groups from histones and other proteins. This process leads to condensation of DNA and subsequent downregulation in expression of specific genes altering the activity of essential cellular pathways and changing the cell's behaviour. As a result, HDACs have emerged as potential therapeutic targets to treat different forms of cancer. In this study we synthesized and investigated compounds built around a phorbazole scaffold to characterize their histone deacetylase inhibitor (HDACi) activity. Eighteen phorbazole analogues were tested against class I and II HDACs in cell lysate, seven showed moderate activity, of which, three had IC₅₀ values below 50 μM. The five best candidates were evaluated more extensively against HDAC1-11 (except for HDAC10). The best candidate, compound 9, initially appeared to reduce the activity of HDAC9 and HDAC11 by more than 50 % at 10 μM concentration. The binding mode of compound 9 to HDACs was explored via computational docking, where two poses stood out. These were explored further by molecular dynamic simulation. We found that 9 likely binds HDAC9 with the pyrrole group buried in the active site and forming H-bonds with the backbone oxygen of one of the glycine residues lining the cavity. But neither of the poses explored offered convincing arguments to describe the mode of action of the phorbazole, especially when comparing 9 to other analogues tested in this study. Additional experiments found that 9 interfered with the cell lysate assays by inhibiting luciferase in a dose-dependent manner (IC₅₀ < 1 μM) and by exhibiting autofluorescence when tested on purified HDAC proteins, thereby confounding the obtained results during both the pan-HDAC screening and the single point HDAC inhibition assay. To address this, we employed fluorophores with excitation and emission wavelengths outside of the emission range of 9 and found that the HDAC inhibition potential of 9 was weaker than first observed. Finally, Compound 9 was found to be very soluble in water (418 μM) and membrane permeable (>48 % flux). This study highlights the need for rigorous validation of results. In our case, two orthogonal testing methods were not sufficient to catch all the confounding factors involved in measurement of HDAC inhibition, and a third approach was required to identify the actual inhibition of 9 against HDAC9 and 11.