Dysregulated IL-6/GP130/JAK Signaling in Calreticulin Mutated Myeloproliferative Neoplasms (MPN)

Introduction. Frameshift mutations in exon 9 of calreticulin (CALR) gene represent the second most frequent alteration in MPN. More than 30 types of CALR mutations are described; the most frequent are a 52bp deletion (Type1, T1) and a 5bp insertion (T2). All mutations cause a recurrent frameshift that originates a novel C-terminal peptide lacking the KDEL Endoplasmic Reticulum-retention sequence. Recent data described the interaction of mutated CALR (CALRm) with the thrombopoietin receptor MPL and the resulting activation of JAK/STATs as well MAPK and PI3K pathways (Araki M et al & Chachoua I et al, Blood 2016; Elf S et al, Blood 2018). However, full characterization of the molecular mechanisms linking CALRm with MPN remains to be fully accomplished.

Methods. By using CRISPR/Cas9 technology we generated CALR knock out (KO) and T1 variants from the GM-CSF cytokine dependent UT7 cell line; primary CD34+ cells from CALRm pts were also analyzed.

Results. Introduction of CALR T1 mutation or generation of KO status did not appreciably modified the proliferation rate, cell cycle and apoptosis of UT7 cells under standard conditions. However, UT7 T1 and KO cells were similarly able to grow in the absence of GM-CSF, otherwise necessary for WT cells; in these cytokine-depleted conditions, apoptosis was significantly increased in WT cells compared both T1 and KO (4- and 8-fold higher; p<0.01) suggesting acquisition of cytokine independency. Exploring the mechanisms leading to cytokine independence we found that signal transducer and activation of transcription 3 (STAT3) was the main constitutively phosphorylated STAT in T1 and KO cells. Among a panel of STAT3 activation candidates assessed by RT-PCR, we found that mRNA levels of interleukin 6 (IL6), which induces STAT3 phosphorylation and activation through the glycoprotein 130/Janus kinase (GP130/JAK) pathway, were significantly increased in T1 and KO compared to parental cells (2- and 3.5-fold higher; p<0.01). The levels of IL6 released in cell culture medium were also significantly increased (1.5- and 2-fold higher; p<0.05). Blocking GP130 with SC144 or IL6 receptor (IL6R) with Tocilizumab (TCZ), in cytokine-depleted conditions, led to cell growth arrest in T1 and KO with no effects on WT cells. SC144 treatment induced a marked dephosphorylation of STAT3 in T1 and KO cells. Also, the membrane expression of GP130 and IL6R resulted higher in KO and T1 compared to WT cells, whilst intracellular content was similar. Co-immunoprecipitation showed that wild type calreticulin, unlike the mutated T1, formed complexes with GP130 and IL6R suggesting that CALRm lost the capacity to negatively regulate their membrane exposure resulting in downstream increased IL6 synthesis and release.

The role of IL6 pathway in CALRm patients was evaluated by assessing megakaryocyte (Mk) colony generation in plasma clot cultures of CD34+ cells in the absence/presence of IL6, as part of cytokine cocktail optimally supporting CFU-Mk growth. We found that CFU-Mk generation by CALRm CD34+ cells was largely independent of exogenously added IL6, unlike control CFU-MK whose number was halved in the absence of IL6 (p<0.05). Moreover, inhibition of GP130 and IL6R by SC144 and TCZ, respectively, significantly reduced CFU-Mk by CALRm CD34+ cells (2.9- and 3.2-fold; p<0.05), with no effect in control cultures. As anticipated, JAK inhibitors ruxolitinib and momelotinib reduced CFU-Mk formation by CALRm CD34+ cells (2- and 3.2-fold, respectively; p<0.01), particularly evident in exogenous IL6-free conditions (5.8- and 4.8-fold, respectively; p<0.01). The combination of either ruxolitinib or momelotinib with TCZ on CFU-Mk formation was also assessed. Both combinations proved synergistic activity in reducing CFU-Mk from CALRm CD34+ cells (4.3- and 5-fold vs control cultures, respectively; p<0.01; combination compared to single agent, 2.8- fold vs ruxolitinib, 4.5-fold vs momelotinib and 4-fold vs TCZ, p<0.05 for all).

Conclusions. Overall, our data suggest a novel model of action of mutant CALR resulting from the loss of its capability to complex GP130 and IL6R and regulate their cell membrane expression. Augmented expression of IL6 receptor complex on CALRm CD34+ cells on turn results in deregulated IL6-dependent signaling and augmented megakaryocyte generation. Blockade of IL-6/GP130/JAK signaling pathway might represent a new therapeutic target in pts with CALR mutation.