This study from the laboratory of Dr. Ricky Johnstone in Australia is focused on the underlying cause of the profound thrombocytopenia that occurs in about 80 percent of patients taking histone deacetylase inhibitors (HDACi). This thrombocytopenia is often the dose-limiting toxicity with all types of HDACi, including panobinostat (a pan-HDACi), romidepsin (HDAC1/2-selective), and vorinostat.
HDACs regulate chromatin structure by deacetylating histones that form the cores of nucleosomes around which DNA wraps. Deacetylation of histones is associated with more tightly wrapped chromatin and transcriptional repression. Because cancer cells are known to silence tumor suppressors, HDACi are thought to have anti-tumor activity by reactivating expression of these genes that normally prevent cell transformation. This is likely an oversimplification, because some genes are downregulated after exposure to HDACi and HDACs may also deacetylate non-histone proteins.
HDAC inhibitors are effective in many types of cancer including Hodgkin lymphoma, T-cell lymphoma, and multiple myeloma. Although their molecular mechanisms are thought to be known, it is not yet clear how their anti-tumor effects are mediated. Because the histone targets are spread throughout the entire genome, the effects of HDACi are considered to be quite broad, so specific genes whose reactivation is induced by these drugs in each different malignancy have not been be identified. Similarly, the mechanism by which these relatively nonspecific drugs cause thrombocytopenia has not been identified.
In this study, the investigators show that multiple HDACi used in the clinic can also induce thrombocytopenia in mice, and they use this murine system along with cell lines to determine a mechanism for the thrombocytopenia. They show that, unlike the thrombocytopenia that occurs with conventional chemotherapy, HDACi do not have anti-mitotic or pro-apoptotic effects on hematopoietic stem and progenitor cells, and that both thrombopoietin levels and megakaryocyte numbers are elevated. Thus, they pursued studies to determine why thrombocytopenia occurs despite normal cellularity with increased megakaryocytes in the marrow.
Culturing the cells in vitro, they visualized a defect in proplatelet formation by megakaryocytes from mice that had been treated with HDACi. Proplatelet formation, which occurs when multiple long projections extend out from the megakaryocytes, is required for release of functional platelets and requires the coordinated action of actin and tubulin cytoskeletal elements. The mediators of these complex cytoskeletal changes are the G proteins Rac, Rho, and CDC25. In order for proplatelets to form normally, a specific myosin molecule (myosin light chain 2) needs to be dephosphorylated, which occurs when RhoA is inactivated. The investigators found that in megakaryocytes that had been exposed to HDACi, myosin light chain 2 was maintained in a highly phosphorylated form, presumably due to alterations in the network of proteins that regulate RhoA activity.
Importantly, the investigators could prevent HDAC inhibitor-induced thrombocytopenia in the mice by administering a mouse-specific thrombopoietin-mimetic (AMP-4), which returned the platelet count to levels similar to those in untreated controls. AMP-4 is the murine version of the FDA approved thrombopoietin-mimetic romiplostim (AMG- 531).
In Brief
These studies begin to answer the question of why thrombocytopenia occurs in response to HDACi, but it remains to be determined how HDACi promote RhoA stimulation. These findings may also be relevant to the thrombocytopenia that can occur in patients treated with the anti-epileptic drug valproic acid, which also functions as an HDACi. Seemingly not directly related to the findings regarding RhoA activation, this study also shows that thrombopoietin mimetics may be clinically useful for preventing HDACi-induced thrombocytopenia even in the setting of elevated levels of endogenous thrombopoietin, thus allowing HDAC inhibitors to be more widely used as anti-tumor agents.
Competing Interests
Dr. Krause indicated no relevant conflicts of interest.
