Abstract
Background and Aims
A gain-of-function mutation in Janus kinase 2 (JAK2V617F) is at the basis of the majority of chronic myeloproliferative neoplasms (MPN). The dual JAK1/JAK2 inhibitor ruxolitinib (ruxo) determined rapid and sustained responses in splenomegaly and symptomatic improvement in patients with myelofibrosis (MF), supporting the central role of dysregulated JAK2 signaling. Enhanced activation of other downstream pathways including the PI3K/mTOR pathway has been documented as well. We previously reported (Bogani et al, PlosOne 2013;8:54828) that targeting mTOR by the allosteric inhibitor RAD001 resulted in inhibition of JAK2VF mutated cells and produced clinical benefits in a phase I/II trial (Guglielmelli et al, Blood 2011;118:2069). In this study we evaluated the effects of BKM120, a specific PI3K inhibitor, alone and in combination with ruxolitinib, in in-vitro and in-vivo MPN models.
Methods
To evaluate cell proliferation, colony formation, apoptosis, cell cycle and protein phosphorylation status we used mouse BaF3 and BaF3-EPOR cells expressing wild type (WT) or VF mutated JAK2, the human VF-mutated HEL and SET2 cell lines, and primary MPN CD34+ cells from patients with MF or polycythemia vera (PV). Effect of drug combination was analyzed according to Chou and Talalay calculating the combination index (CI); a CI <1 indicates synergistic activity. For in vivo studies we used two mouse models: (1) SCID mice receiving iv BaF3-EPOR VF-luciferase (luc) cells (gift of T. Radimerski) were randomized on day 6 to different treatment groups based on baseline luminescence. (2) C57Bl6/J JAK2 VF Knock-in mice were generated by insertion of the reversed JAK2V617F exon 13 sequence; mating with Vav-Cre transgenic mice activates the VF allele producing a MPN phenotype in progenies with VF heterozygous expression (Hasan et al, Blood 2013;Epub). Mice were treated for 15 days, then blood, spleen and bone marrow cells were analyzed.
Results
We found that BKM120 preferential inhibited BAF3 VF and BaF3-EpoR VF cells (IC50: 364±200nM and 1100±207nM, respectively) compared to their respective WT counterpart (5300±800nM and 3122±1000nM: p<.05). HEL and SET2 cells resulted also sensitive to BKM120 (2000±500nM and 1000±300nM). Interestingly we found that BKM120 significantly increased G2/M phase and decreased S phase of cell cycle (p<.01) and induced apoptosis (IC50, SET2=10µM, BaF3-EPOR VF=1.8 µM). Western blot analysis showed marked reduction of phospho-mTOR and its target phospho-4EBP1 as well as downregulation of phospho-STAT5 at 6 and 24h of treatment. BKM120 impaired colony formation from MF and PV CD34+ cells at doses 2 to 8-fold lower than healthy controls (p<.01). BKM120 strongly inhibited EEC colony growth from PV pts (IC50, 9±4nM). Co-treatment of BKM120+ruxo resulted in synergistic inhibition of proliferation of SET2 (median CI=0.45) and BaF3-EPOR VF (median CI=0.8) cells. Triple combinations including BKM120/ruxo plus either RAD001 (Torc1 inhibitor) or PP242 (Torc1/2 inhibitor) resulted highly synergistic (median CI=0.27 and 0.52) to indicate the importance of complete mTOR inhibition. BKM120 at 45mpk and 60mpk increased mean lifespan of BaF3 VF luc mouse model from 21d in control mice to 27.2d and 28d in BMK120 treated mice. In KI mice, co-treatment with 60mpk BKM120 + 60mpk ruxo resulted in improvement of splenomegaly (median spleen weight: 1.4, 0.82, 0.8 and 0.6 g respectively for controls, 60mpk BKM120, 60mpk ruxo and 60mpk BKM120+60mpk ruxo) and reduction of leukocytosis and reticulocyte count. The level of phosho-STAT5 and -4EBP1 in the spleen was significantly reduced in mice receiving BKM120+ruxo as compared to single drug treatment. We finally analyzed the effects of BKM120+/-ruxo on the in-vitro clonogenic growth of BM cells from VF and WT KI mice mixed in a 1:1 ratio. The proportion of VF-positive colonies resulted reduced in a dose dependent manner by 19%, 33% and 44% (p<.03) compared to controls with 50nM, 100nM and 300nM BKM120 respectively. A 25% and 39% of VF-positive colonies reduction was achieved with 50nM and 100nM ruxolitinib. The combined treatment with 100nM BKM120 + 50nM ruxo resulted in a 50% decrease of the number of mutated colonies (p<.02).
Conclusions
Inhibition of PI3K by BKM120 alone and combined with JAK2 inhibitor ruxolitinib resulted in enhanced activity in preclinical models of MPN, providing a rationale for the ongoing combination clinical trial.
Disclosures:
Vannucchi: Novartis: Membership on an entity’s Board of Directors or advisory committees.
Prospect of JAK2 inhibitor therapy in myeloproliferative neoplasms