LAUSR

Among the driver mutations in the MPNs, the mechanism by which CALR mutations activate JAK/STAT signaling is unique in that the novel C-terminus of CALR mutant proteins binds directly to MPL leading to constitutive activation [1, 2]. CALR mutants have also been shown to activate the MAPK pathway [3]. Whether CALR mutants further activate PI3K/AKT signaling is controversial, with several studies reporting a lack of (or modest) activation [2, 4], but another demonstrating potent activation [5]. The former studies relied upon transduction of Ba/F3 TpoR cells with CALR mutant alleles whereas the latter probed the pathway in cells lines with megakaryocytic potential. Moreover, the extent to which PI3K/AKT signaling is a target in CALR mutant MPNs has been debated, with at least one study reporting that inhibitors of this pathway do not synergize with ruxolitinib [2]. To address the discrepancy, we assayed for activation of JAK/STAT and AKT in multiple settings. First, we overexpressed the CALR type 1 (del52) and type 2 (ins5) mutants in primary murine c-kit bone marrow progenitors and cultured the cells for 48 h. We observed enhanced activation of both STAT5 and AKT (Fig. 1a). Next we transplanted bone marrow transduced with CALRdel52 mutant or empty vector to irradiated recipient mice and assessed AKT activation. Both intracellular flow and western blots of spleen cells revealed increase in p-AKT (Figs. 1b, c). Finally, we assayed for total and phosphoAKT in primary CALR mutant MPN samples. We observed robust increase in AKT phosphorylation in patient derived CD34 cells (Fig. 1d and Supplementary Table 1). This was accompanied by an increase in cyclinD3 consistent with the increased proliferation of the CALR mutant cells. These cells also expressed MPL, consistent with the manner in which CALR mutants activate signaling (Fig. 1d). We previously reported that inhibition of AKT activity with the selective AKT inhibitor MK-2206 resulted in reduced growth of MPLW515L expressing cells both in vitro and in vivo, suggesting that this pathway is a therapeutic target in ET and PMF [6]. To determine whether CALR mutant expressing cells are similarly dependent on PI3K/AKT signaling, we treated CD34+ cells from PMF patients or healthy individuals with MK-2206. We found that CALR mutant cells were susceptible to AKT inhibition and more sensitive than healthy progenitor cells in colony assays (Figs. 1e, f). Of note, the differential effect was much more significant for megakaryocyte colonies (Fig. 1f) than for myeloid colonies (Fig. 1e), consistent with a reliance of megakaryocytes on MPL signaling. Finally, while the level of p-AKT was higher in bone marrow cells from mice transplanted with CALRdel52 expressing cells compared to those with empty vector, the degree of inhibition was similar (Fig. 1g). MK-2206 was well tolerated in healthy C57Bl/6 mice with no evidence of myelosuppression or impaired hematopoiesis [6]. To determine the extent to which the drug suppressed thrombocytosis and megakaryopoiesis induced by mutant CALR, we transplanted CALRdel52 mutant expressing hematopoietic progenitor cells into irradiated