Résumé
Multiple myeloma (MM) is a virtually incurable clonal plasma cell malignancy which mainly resides in the bone marrow. Like in other cancers, numerous epigenetic aberrations have been described in MM, resulting in deregulated gene expression, disease progression and drug resistance. Targeting epigenetic modifiers involved in this deregulated epigenetic landscape therefore represents an interesting therapeutic approach. G9a (EHMT2) and GLP (EHMT1) are 2 histone methyltransferases which catalyze mono- and dimethylation of histone 3 lysine 9 (H3K9). Importantly, G9a is overexpressed in several cancers, correlating with a poor prognosis. Currently, data about the expression and role of G9a/GLP in MM is lacking. The aim of this study is therefore to investigate the functional role of G9a and GLP in MM pathogenesis.
Here we report that high expression levels of both G9a and GLP are associated with a worse disease outcome in the UAMS-TT2 newly diagnosed MM patients cohort (n=345, GSE4581). Gene set enrichment analysis of high G9a/GLP expressers displays a significant enrichment of genes involved in pathways associated with MM disease progression, including the RAS pathway, NF-ĸB canonical pathway, IRF4 multiple myeloma program and mRNA splicing. Using a panel of 10 human cell lines, 3 murine cell lines and 5 primary samples we further evaluated the effect of the specific small molecule G9a/GLP inhibitors UNC0638 and BIX01294 on MM cell viability, cell cycle progression and apoptosis. We found BIX01294 and UNC0638 to significantly and potently reduce MM cell viability. Moreover, both inhibitors also induce cell cycle arrest and apoptosis. When comparing between both inhibitors, BIX01294 was found to be the most potent to induce apoptosis. Mechanistic studies for BIX01294 furthermore indicate that BIX01294 treatment results in autophagic programmed cell death as evidenced by a strong increase in the formation of LC3B puncta and an increase in LC3II and beclin-1 protein levels. In addition, we found that BIX01294 sensitizes MM cells to the clinically relevant proteasome inhibitor bortezomib and the Bcl-2 inhibitor ABT199. The in vivo anti-MM activity of therapeutic BIX01294 treatment was tested using the murine 5TGM1 model. Difference in overall survival between the vehicle and BIX01294 treated groups was assayed with a log-rank test and survival curves plotted using the Kaplan-Meier method. Therapeutic treatment of 5TGM1 inoculated mice with BIX01294 resulted in a clear delay in tumor progression, as evidenced by a clear decrease in tumor burden and a significant increase in the overall survival of BIX01294 treated mice compared to vehicle treated mice. Lastly, 5TGM1 inoculated mice were treated with a suboptimal dose of bortezomib or BIX01294, and the combination of both in a therapeutic setting. We observed a significant reduction in BM plasmacytosis in mice treated with bortezomib (p<0.05) and with the combination of both bortezomib and BIX01294 (p<0.001) compared to vehicle mice. Although there was no significant difference between BM plasmacytosis levels in mice treated with bortezomib and in the combo treated mice, we could observe a clear trend of an additional reduction in tumor burden in combo treated mice versus bortezomib treated mice.
In conclusion, our data indicates for the first time the importance of the histone methyltransferases G9a/GLP in MM pathogenesis. Furthermore, we demonstrate that specific targeting of G9a/GLP induces MM cell apoptosis, enhances MM sensitivity to ABT- 199 and bortezomib and significantly delays tumor progression in the murine 5TGM1 model. Thus, G9a/GLP targeting represents a promising strategy to improve treatment of MM.
No relevant conflicts of interest to declare.