Inhibition of NSD2 in t(4;14) myeloma induces changes in mitochondrial priming Free

Introduction: Myeloma patients with the t(4;14) subtype are classified as high-risk with lower progression free and overall survival. Currently there are no therapeutic regimens specific for this subtype, thereby representing an unmet need in myeloma treatment. Previously, we demonstrated that mitochondrial priming in t(4;14) myeloma cell lines is heterogeneous and cannot be easily categorized as seen with t(11;14) myeloma. Further, our data showed inhibition or loss of NSD2 in KMS11 isogenic cell lines resulted in increased priming and sensitivity to BCLXL and possibly BCL2 inhibition. Here we have expanded our study to include additional t(4;14) cell lines.

Methods: The occupancy of anti-apoptotic BCL2 proteins with pro-apoptotic ones is referred to as mitochondrial priming and indicates the proximity of a cell to the apoptotic threshold. This is measured through an assay called BH3 profiling. Dynamic BH3 profiling (DBP) was performed by treating KMS11, KMS34, KMS26, KMS18, and H929 for six days with 1 mM of the NSD2 inhibitor NSD2i-1 (Excenen), followed by BH3 profiling. Drug-induced changes in priming (Delta priming) were calculated by subtracting the % cytochrome c loss of the control samples from the % cytochrome c loss of the NSD2 inhibitor-treated samples. Cell death studies were performed by drug treatment for 24-48 h followed by Annexin V/propidium iodide staining and flow cytometry. Loss of H3K36 dimethylation was verified via intracellular flow.

Results: Consistent with our BH3 priming data, the change in mitochondrial priming following NSD2 inhibition was also heterogeneous. Treatment of H929 resulted in an average delta priming of 28.5% with BIM peptide (targets all BCL2 family members) and 30% with the MS1 peptide (targets MCL1). NSD2i-treated H929 also showed increased sensitivity to the MCL1 inhibitor AZD5991. We previously demonstrated H929 is highly primed on MCL1 and these data suggest this priming increases with NSD2i. Treatment of KMS34 led to a 10-12% delta priming with the BAD peptide as well as increased sensitivity to the BCL-X_L/BCL2 inhibitor AZD4320. Our data indicate KMS34 is primed on BCL-X_L, and after NSD2i this priming expands to include BCL2. Previous BH3 profiling data showed KMS26, KMS11, and KMS18 are not primed on a single BCL2 family member and targeting more than one is needed to induce cytochrome c release. In KMS26 this pattern remained largely unchanged following NSD2 inhibition. Dynamic profiling of KMS11 indicated increased priming on BCL-X_Land potentially BCL2 with delta priming of 14-21% with BAD and 14% with HRK peptides (targets BCL-XL), as well as increased sensitivity to AZD4320 and DT2216 (BCL-X_LPROTAC). These data are consistent with what we previously reported in the KMS11-NTKO and -TKO cell lines. Unexpectedly, inhibition of NSD2 in KMS18 resulted in negative delta priming. Cytochrome c loss decreased in response to the BIM, BAD, and HRK peptides following treatment (average delta of -13%, -20.3%, and -13.3% respectively). Interestingly, when we tested KMS18 NSD2-low cells (loss of translocated allele) we observed lower overall priming (BIM) and lower priming on MCL1 (MS1) and BCL2/BCLXL (BAD). Additionally, we found these cells were less sensitive to AZD5991 and AZD4320 than the control cells, suggesting loss of NDS2 in KMS18 leads to decreased priming.

Conclusions: Dynamic mitochondrial profiling post-NSD2 inhibition in t(4;14) myeloma displayed cell line specific patterns, perhaps owing to the heterogenous baseline priming in this high-risk subtype. The molecular basis for these differences is under current investigation and could be related to the differences in genes influenced by NSD2 activity in these cells. RNAseq analysis of 3 myeloma cells lines following NSD2i treatment has shown that while over 400 genes are significantly downregulated in each cell line, only 49 were common to all 3 lines, consistent with the presence of cell-specific events. Alternatively, events secondary to t(4;14) (e.g., TP53 loss or mutation, 1q gain, 1p loss, activating mutations in FGFR3) could also influence changes in priming. Together these findings suggest that functional studies may need to be performed along with genomic analysis to determine which t(4;14) patients will most likely benefit from NSD2 inhibition as well as determine the best ways to combine NSD2 inhibitors with other myeloma or novel therapies.