XPO1 inhibition by selinexor blocks nucleo-cytoplasmic trafficking of multiple members of the NF-kb family in MPN cell lines and primary myelofibrosis cells Free

Introduction: Selinexor, a first-in-class oral selective XPO1 inhibitor, in combination with ruxolitinib in JAK inhibitor-naïve patients with myelofibrosis (MF) has shown rapid, deep, and sustained spleen and symptom responses with associated reductions in proinflammatory cytokines (Tantravahi et al., ASH 2023). In a JAK2^V617F-driven mouse model, selinexor treatment selectively suppressed JAK2^V617F+ progenitors compared to normal progenitors and decreased myeloproliferation, demonstrating disease-modifying potential (Yan et al., CCR 2019). In MF, dysregulation of the NF-κB pathway contributes to disease development and progression by activating proinflammatory cytokines that perturb the bone marrow niche and suppress normal hematopoiesis. Selinexor has been shown to inhibit NF-κB activity by preventing nuclear export of both p65 and IκB, which allows IκB to bind to p65 in the nucleus, repressing the NF-κB complex's transcriptional activity. High throughput imaging flow cytometry allows for high-content image analysis and has proven useful in determining subcellular location of proteins. Here we used imaging flow cytometry to monitor selinexor-induced changes in subcellular localization of proteins relevant to MF pathogenesis in the HEL cell line as a model of JAK2^V617F MF, as well as MF patient CD34⁺ cells.

Methods: Cells were incubated with 250 nM selinexor for 16 hours (IC₅₀ at 72 hours=237 nM for HEL cells) followed by fixation with formaldehyde, permeabilization with methanol, then incubated overnight at 20°C with fluorescently-conjugated antibodies targeting various components of the NF-κB pathway, namely Transcription factor p65 (RELA), Transcription factor RelB (RELB), NF-κB inhibitor alpha (IkBα), proto-oncogene c-Rel (REL). Signals for CD45 and DAPI defined the boundaries of the cytoplasm and nucleus. Data was collected on a 4-laser, 2-camera ImageStream Mk II imaging flow cytometer. The similarity score, a derivative of linear regression with positive values noting correlated signal origin, was calculated by IDEAS v6.2 software for each cell and used to infer nuclear/cytoplasmic localization. Analysis was limited to cells in G0/G1 to ensure an intact nuclear envelope.

The patient sample analyzed was from the bone marrow of a patient with MF secondary to polycythemia vera, at transformation to acute myeloid leukemia (70% atypical CD34+ myeloblasts, CD33-negative; fibrosis 3/3). Mutations detected in the bone marrow included JAK2^V617F, SRSF2^P95R, NRAS^G12S, TET2^E155fs, TET2^F1287S, and a newly emerged RUNX1^D198N mutation. The patient was previously treated with ruxolitinib, fedratinib, and ilginatinib/NS018, with adverse events and lack of response.

Results: Selinexor induced G0/G1 cell cycle arrest at 16 hours in HEL cells. Preliminary data showed that exposure of HEL cells to selinexor led to pronounced nuclear sequestration of several NF-κB family members. Cells with nuclear localization of RELA increased from 39.4% in DMSO-treated controls to 93.2% following selinexor treatment. IκBα translocation rose from 33.1% to 85.1%, consistent with the previously described mechanism of action for selinexor. Interestingly, RELB showed a marked nuclear increase from 48.0% to 98.4% which may be mechanistically important as emerging evidence (Navarro et al., Cell Discovery 2025) points to its potential for tempering RELA-driven pro-inflammatory signaling through competition for DNA binding or direct complex formation. IKKα/β showed a modest decrease in nuclear localization from 95.1% to 89.2%, while REL localization remained largely unchanged. Similar trends were observed in a preliminary analysis of primary CD34⁺ cells, including IKBa and RELB. Additional markers are being evaluated, and experimentation is ongoing.

Conclusions Selinexor treatment of MPN cells resulted in nuclear accumulation of both p65 (total and phosphorylated) and IkBα, consistent with selinexor's mechanism of suppressing NF-kB activity through nuclear sequestration of p65 and IkB. These events occur in G0/G1 phase cells prior to apoptosis, indicating that nuclear sequestration may impede continued proliferation and survival. We are evaluating multiple additional primary MF CD34⁺ patient samples and results will be presented. Selinexor is currently being evaluated in the SENTRY (XPORT-MF-034) phase 3 clinical trial (NCT04562389) in combination with ruxolitinib for JAK inhibitor naïve patients.