The PDI-ATF6-axis as a novel treatment target in CSF3R driven chronic neutrophilic leukemia (CNL) Free

Introduction CNL is a rare myeloproliferative neoplasm (MPN) with a 5-year-survival-rate of 28 percent, and no standard of care for treatment. The majority of CNL patients harbors the membrane proximal T618I mutation in the colony-stimulating factor 3 receptor (CSF3R), which leads to constitutive activation of the JAK1/STAT3-pathway. However, JAK1/2-inhibitor Ruxolitinib achieves only a limited response in CNL patients, suggesting the involvement of additional signaling pathways. Our previous data established that CSF3R^T618I is retained in the endoplasmic reticulum (ER), where it interacts with Calnexin, induces unfolded protein response (UPR) and ER-Phagy. In this study we analyzed the role of ER-stress sensor ATF6 for CSF3Rmutations mediated cell proliferation and signaling, and explored potential signaling pathways crucial for Ruxolitinib-persistent CSF3R^T618I expressing cells.

Methods PDI inhibitor 16F16 treatments in Ba/F3, 32D and HEK293 cells were conducted to target ATF6 and analyze effects on membrane proximal-, truncation-, and compound-mutated CSF3R, and mutant Calreticulin (CALR)del52 mediated oncogenic signaling pathways, using MTT-assays, Western Blots (WB) and co-Immunoprecipitation (IP). RNA-seq was performed with Ba/F3 cells and peripheral blood samples from CNL-patients from the German Study Group for Myeloproliferative Neoplasms (GSG-MPN) (n=28). NIH-3T3 and 32D cells expressing CSF3R^T618I, CSF3R^WT-KDEL and CSF3R compound mutations were used to assess the significance of altered protein structure on signaling and UPR, using confocal microscopy, MTT-assays and WB. A PamGene assay was utilized to analyze differential kinase activity in Ruxolitinib-persistent CSF3R^T618I expressing Ba/F3 cells.

Results RNA-seq revealed upregulation of ER-stress-sensor ATF6 and ATF6 target genes in CSF3R^T618I, CSF3R^W791* and CSF3R^T618I+Y748* expressing Ba/F3 cells, and in CNL-patients harboring CSF3R^T618I or CSF3R-compound-mutations. Compound-mutated CSF3R proteins displayed a specific pattern of UPR induction compared to CSF3R^T618I, suggesting an important role of the cytoplasmic tail of CSF3R in ER-stress-sensor activation. Mechanistically, PDI inhibitor 16F16 led to decreased ATF6 activation in 32D and Ba/F3 cells, and impaired dimerization of CSF3R^T618I and CSF3R^W783*, but not CSF3R^WT expressing, flag-/HA-tagged CSF3R constructs containing, HEK293 cells. Targeting ATF6 activation and PDI activity with 16F16 led to a decrease in cell proliferation and reduced STAT3, Src kinase and AKT activation in membrane proximal-, truncation-, and compound-mutated CSF3R, as well as Ruxolitinib-persistent CSF3R^T618I, but not CSF3R^WTexpressing 32D and Ba/F3 cells. PamGene assay analysis revealed enhanced activity of PKC-family-kinases in Ruxolitinib-persistent CSF3R^T618I expressing Ba/F3 cells compared to non-persisters. Co-IP showed interaction of CSF3R^T618I and PKC-family-kinases and WB-results corroborated a JAK-independent activation mechanism. Interestingly, 32D cells expressing thrombopoietin receptor (MPL) and CALRdel52 also showed ATF6 protein activation and UPR induction. Treatment with 16F16 led to significantly decreased cell proliferation compared to CALR^WT expressing 32D cells.

ConclusionsCSF3R^T618I strongly induces UPR through the PDI-ATF6-axis. Activation of this pathway requires specific regions of the cytoplasmic domain of CSF3R, acting in concert with protein folding alterations induced by T618I. Our results suggest a shift to enhanced PKC-family-kinase mediated signaling as a possible mechanism for CSF3R^T618I expressing cells to evade Ruxolitinib mediated JAK1/2 inhibition, potentially explaining limited treatment efficacy in CNL patients. However, 16F16 treatment significantly decreases cell proliferation mediated by various CSF3R mutants, including Ruxolitinib-persistent CSF3R^T618I expressing cells, potentially through a multilayered mechanism, comprising impaired dimerization and enhanced degradation of CSF3R mutants due to blockage of the PDI-ATF6-axis. CALRdel52 mediated proliferation, as seen in MPNs such as essential thrombocythemia, can also be diminished by 16F16. These results implicate the PDI-ATF6-axis as a novel and effective treatment target for CSF3R mutations mediated CNL and other MPNs.