Selective Targeting of CML Progenitor/Stem Cells by the Class 1 Histone Deacetylase (HDAC) Inhibitor MS275 in Combination with Imatinib.

Abstract 2791

BCR-ABL tyrosine kinase inhibitors (TKI) effectively inhibit CML stem/progenitor cell proliferation and induce remission in CML patients, but do not completely eliminate primitive leukemia stem cells, which persist as potential sources of relapse. There is considerable interest in identifying additional therapeutic strategies to selectively induce apoptosis in CML stem and progenitor cells. We have shown that the pan-HDAC inhibitor Pabinostat, that targets Class 1, Class 2 and Class 4 HDACs, can eliminate CML stem cells in vitro and in vivo when combined with Imatinib (Cancer Cell 17:427, 2010). However pan-HDAC inhibitors can have significant toxicity to normal stem cells, which limits their clinical utility. Here we used siRNA knockdown to determine the role of individual HDACs in proliferation and apoptosis of CML compared with normal CD34+ cells. CD34+ cells were transfected with siRNAs to 11 individual HDAC enzymes and control non-specific siRNA, using an Amaxa nucleofector system. 50–80% knockdown of target gene expression was achieved. Knockdown of individual Class 1, 2 and 4 HDACs resulted in increased apoptosis of CML CD34+ cells, which was increased in combination with IM. On the other hand, normal CD34+ cells demonstrated increased apoptosis in response to knockdown of Class 2 and 4 HDACs, but not Class 1 HDACs. These results suggested that inhibition of Class I HDACs may allow selective targeting of CML stem/progenitor cells while sparing normal cells, and led us to evaluate the effects of MS275, an orally available HDAC Class I specific inhibitor currently being evaluated in clinical trials in solid tumors, against CML and normal stem/progenitor cells. MS275 (1μM) exposure increased acetylated Histone H3 and histone H4 levels in CML CD34+ cells, as analyzed by Western blots and flow cytometry. Treatment of CML CD34+CD38- cells and CD34+CD38+ cells with MS-275 resulted in significantly increased apoptosis compared with their normal counterparts (17.93±1.1% for CML vs. 7.9±2.3% for normal CD34+CD38- cells, p=0.01; 20.1±4.6% for CML vs. 9±1.9% for normal CD34+CD38+ cells, p=0.05). Combination of MS275 with IM resulted in significant enhancement of apoptosis of CML but not normal cells (72±11.1% for CML vs. 6.3±1.87% for normal CD34+CD38- cells, p=0.02; 41.3±9.8% for CML vs. 8.45±1.75% for normal CD34+CD38+ cells, p=0.04). Quiescent CML stem cells are especially resistant to TKI-induced apoptosis. The combination of MS275 and IM resulted in significantly increased apoptosis of non-dividing CML but not normal CD34+38- cells (72.1±14.3% for CML vs. 10.6±2.0% for normal cells, n=3). Cell cycle analysis by Ki67 and DAPI labeling demonstrated significant increase in G0, and decrease in S/G2/M phase, in CML but not normal CD34+ cells treated with MS275, alone or combination with IM. Consistent with these findings, we observed downregulation of the anti-apoptotic protein Mcl1 and up-regulation of the cell cycle inhibitor p21 in MS-275 treated CML CD34+ cells. MS275 treatment also significantly reduced colony forming cell growth from CML but nor normal CD34+CD38+ and CD34+CD38- cells in methylcellulose progenitor culture. To further evaluate the effects of MS275 on normal stem cells, normal CD34+ cells were treated with MS275, IM or the combination of MS275 and IM for 72 hours in vitro, followed by injection into sub-lethally irradiated (300 cGy) NSG mice and evaluation of human cells engraftment after 12 weeks. These studies showed that treatment with MS275 did not significantly affect engraftment of normal stem cells in NSG mice. In conclusion, the selective Class1 HDAC inhibitor MS-275 significantly increases apoptosis in CML primitive and committed progenitors, including non-proliferating cells, which is further significantly enhanced in combination with IM. In contrast to pan-HDAC inhibitors, MS275 treatment does not affect survival or growth of normal stem cells. These results indicate that use of inhibitors specific to Class 1 HDAC enzymes could avoid toxicity to normal HSC associated with pan-HDAC inhibitors, allowing more selective and effective targeting of CML LSC.