NSD2 inhibition suppresses t(4;14) gene expression and oncogenic signaling to impair proliferation in multiple myeloma

Abstract

Abstract Introduction: The t(4;14) translocation is present in ~12% of Multiple Myeloma (MM) patients and uniformly results in high levels of NSD2 expression. NSD2 catalyzes mono- and di-methylation of histone H3 lysine 36 (H3K36me2), which is dramatically increased in t(4;14) MM. How NSD2 expression in t(4;14) and the subsequent accumulation of H3K36me2 promotes malignant transformation of plasma cells in MM is poorly understood. Methods: H929, KMS11, KMS18, KMS26, LP1, and XG-7 MM t(4;14) cell lines were treated with the NSD2 catalytic inhibitor KTX-1031 (K36; hereafter NSD2i; 1 uM) or the substantially inactive enantiomer (KTX-1030) as a control. Growth assays used flow cytometry-based counting over 7 days with or without HS-5 stromal cells, or HS-5 conditioned media. Changes in H3K36me2 and H3K27me3 were assessed by immunoblot using total H3 as a loading control and histone mass spectrometry performed at Northwestern proteomics core. Gene expression changes were determined by RNAseq (Kapa). Changes in chromatin were assessed using Cleavage Under Targets and Tagmentation (CUT&Tag; Epicypher) for H3K36me2, H3K27me3, H3K27ac. Recombinant CRISPR/Cas9 was used to generate multiple single cell clones of KMS11 and KMS18 with the NSD2 translocated allele disrupted. Results NSD2 inhibition significantly depleted H3K36me2 by immunoblot analysis in all cell lines tested. Histone mass spec of H929, KMS11, KMS18 indicated H3K36me2 at 57, 57, and 24% of H3.1 histones, respectively. NSD2 inhibition uniformly reduced H3K36me2 to 4-5% of histones in all three cell lines (p<10e-6). In 5 of 6 t(4;14) cell lines tested, NSD2 inhibition significantly reduced growth (p<0.05). Interestingly, H929, which did not show growth inhibition in mono-culture, was significantly inhibited in stromal cell co-culture, or in conditioned media (p<0.01). This was due to a growth advantage conferred by co-culture and conditioned media that was ablated by NSD2 inhibition. Indeed, KMS11 and KMS26 cell lines also showed growth advantages in co-culture that were, in part, NSD2-dependent. RNAseq of H929, KMS11, KMS18 indicated 663, 554, and 888 genes dysregulated by NSD2 inhibition, respectively (FDR<0.01; fold-change>1.5). Most genes were downregulated upon NSD2 inhibition, consistent with reports of H3K36me2 functioning to activate transcription. Surprisingly, the overlap of NSD2-regulated genes was only 54 genes. Similarly, Gene Set Enrichment Analysis, indicated only a few commonly regulated pathways, and these were all downregulated. Nonetheless, NSD2 inhibition significantly downregulated the same genes impacted by NSD2 genetic ablation (p<0.0001). Furthermore, the gene expression program of t(4;14) MM samples from Arkansas (Zhan et al., Blood 2006) and from the MMRF CoMMpass study were potently downregulated by NSD2 inhibition in all cell lines tested (p<0.008). Chromatin analysis by CUT&Tag of NSD2 inhibited cells indicated a dramatic depletion of H3K36me2, consistent with the mass spec and immunoblot data. Interestingly, H3K36me2 was enriched at enhancers marked by H3K27ac. This was further supported by the mass spec data indicating that K36me2 was present on 40-70% of H3.1 peptides with K27ac. NSD2 inhibition resulted in decommissioning of 192 to 1,319 enhancers in the three cell lines analyzed, and these NSD2-dependent enhancers were proximal to genes downregulated by NSD2 inhibition (p <0.001). Conclusions Here, we employed pharmacologic and genetic approaches to understand how NSD2 regulates MM biology, and found NSD2 inhibition reverses the t(4;14) gene expression program. This resulted in significant growth inhibition that was sometimes dependent upon exogenous signals. Mechanistically, NSD2 inhibition reduced H3K36me2 from 57-24% to ~5% of all histones, and reprogrammed the chromatin state, disrupting NSD2-dependent enhancers that were tightly correlated with gene expression changes. We propose a model where NSD2-mediated H3K36me2 accumulation in t(4;14) facilitates a permissive chromatin environment and potentiates common oncogenic signaling pathways in MM. These results have implications for ongoing clinical trials testing NSD2 inhibition in patients (NCT05651932).