Panobinostat, An Oral Pan-Histone Deacetylase (HDAC) Inhibitor Activates a Microrna Signature That Targets Rad51 To Attenuate Homologous DNA Repair and Sensitize AML Cells To Sapacitabine

Acute myelogenous leukemia (AML) is characterized by multiple genetic and epigenetic abnormalities including a profound dysregulation of microRNA expression. Specific expression signatures of miRNA are associated with disease behavior and clinical outcome.

Effective clinical treatment of AML has largely depended on a class of antimetabolites – the nucleoside analogs. Of these, sapacitabine is a nucleoside analog prodrug that is in development for the therapy of AML. It is converted to its active metabolite 2-C-cyano-2-deoxy-1-β(-D-arabino-pentafuranosyl) cytosine (CNDAC), which interferes with DNA synthesis by initially causing a single stranded DNA break that is converted into a double strand break in the subsequent replicative cycle. Such double strand breaks are primarily repaired by the homologous recombination repair (HR) pathway. Consequently, efficient HR may offer a potential resistance mechanism to therapy with sapacitabine.

Panobinostat and vorinostat are pan HDAC inhibitors (HDACi) that modulate gene expression, induce apoptosis, and are clinically active against specific leukemias and lymphomas. In pilot experiments, we identified that exposure of AML cell lines to a combination of panobinostat or vorinostat and sapacitabine had a greater than additive effect on cell viability. Loss of cell survival was preceded by a rapid and marked decline in the levels of Rad51, a protein that is critical in repairing CNDAC-induced damage via the HR repair pathway, both in AML cell lines and primary tumor cells that were exposed to either panobinostat or vorinostat. Therefore, we hypothesized that exposure of AML cells to HDACi would activate the expression of a set microRNAs, that would target Rad51. Loss of Rad51 protein, would in turn impede the successful completion of HR rendering the cells incapable of repairing the damage caused by CNDAC and thus sensitizing them to the nucleoside analog.

Our preliminary data using microRNA arrays indicated that exposure of AML cells to panobinostat resulted in the induction of nine microRNA that putatively target Rad51. We then ectopically expressed three of the nine Rad51- directed miRNA and found that miR-182 efficiently targeted the Rad51 protein. Using real time-PCR we found that the levels of miR-182 were significantly lower in primary AML cells when compared to hematopoietic cells from normal donors. Using chromatin immunoprecipitation (ChIP) assays we identified that HDAC1 and HDAC2 were recruited to the miR182 promoter in AML cell lines as well as in primary AML cells, which was probably linked to loss of gene expression. Correspondingly, HDAC inhibition led to the accumulation of activating chromatin modifications at the miR-182 promoter, which was followed by the induction of miR-182 levels in both AML cells lines and primary AML cells. To determine whether HDAC inhibition targets Rad51 via miR-182, we expressed antagomiRs that interfere with the binding of miR-182 to Rad51, and found that the levels of Rad51 were preserved in cells expressing anti-miR182 before being exposed to panobinostat. Finally, we identified that exposure to panobinostat suppressed ongoing HR as measured by the absence of Rad51 foci in cells exposed to this agent, whereas cells exposed to CNDAC alone exhibited robust Rad51 foci formation, indicating an active HR pathway.

In conclusion, HDAC inhibition results in the induction of miR-182 which targets Rad51 and attenuates HR to sensitize cells to sapacitabine in AML. This therapeutic strategy may be effective in circumventing potential resistance mechanisms to nucleoside analog therapy in AML.