LSD1 Impairs the Epithelial-Mesenchymal Transition (EMT) and Osteoclastogenesis Potency in Multiple Myeloma and Synergistically Induces Cytotoxicity with HDAC Inhibitors

Lysine-specific demethylase 1 (LSD1) is a FAD-dependent histone demethylase, which selectively removes mono- and di-methyl groups from histone 3 lysine 4 or 9 residues (H3K4, H3K9) leading to either repression or activation of transcriptome. Previous studies have shown that lenalidomide and pomalidomide cause cell cycle arrest in Multiple Myeloma (MM) by modifying the chromatin structure of the p21WAF-1 promoter through LSD1 demethylation. LSD1 forms a co-repression complex with HDAC1 and HDAC2, mSin3a, and MMSET. However, the functional role of LSD1 in MM and its contribution in aggressive traits of the disease is largely unknown.

First, we evaluated the expression of LSD1 in different datasets of MM patients (GSE2113, GSE16122) and observed significant overexpression in patients with symptomatic MM and Plasma Cell Leukemia (PCL) (p<.001). The expression of LSD1 in a panel 45 HMCLs was also pronounced. We confirmed the expression and both its nuclear and cytoplasmic localization by immunoblotting analysis in 4 different HMCLs and primary bone marrow plasma cells from newly diagnosed, relapsed MM and PCL patients (N=8).

We further evaluated the LSD1-mediated effect on proliferation and survival by performing loss- and gain of function studies. LSD1 knockdown in LP1 and MM1S cells resulted in modest cytotoxicity. After a combination silencing of JARID1 members and LSD1 we were able to observe a further significant decrease in survival of MM cells lacking JARID1C and LSD1, indicating that the overlapping demethylation of H3K4 is of high importance for the cell survival.

We examined the post-translational histone modifications by immunobloting after LSD1 knockdown and as expected, we observed significant increase of K4me2/3 and K9me2 marks, but more interestingly, alteration of acetylation status of K9. Therefore, we performed cytotoxicity and proliferation experiments in MM after knockdown of LSD1 in combination with HDAC inhibitors (SAHA, LBH589) and we observed that LSD1 depletion enhances the cytotoxicity effect of HDACs inhibitors. LSD1 depletion resulted in significant reduction of mRNA levels by using real-time PCR and protein expression by immunoblotting of HDAC1 and HDAC2.

Furthermore, based on findings of higher expression of LSD1 in more aggressive types of MM, we sought to investigate the impact of LSD1 in epithelial-mesenchymal transition (EMT). LSD1 depletion in MM1S and LP1 cells inhibited significantly the migratory ability estimated by transwell migration assay, invasion and wound healing assays. More importantly, MM cells lacking LSD1 expressed significant lower levels of E-cadherin, N-cadherin and Vimentin evaluated by immunoblotting and immunocytochemistry. We confirmed the suppression of EMT-involved gene expression by performing a PCR-microarray assay.

Finally, given the presence of osteolytic lesions as a hallmark of disease, and consequent impact on outcome, we evaluated the impact of LSD1 on osteoblast differentiation and osteoclastogenesis. LSD1 depletion/ and pharmacological inhibition (S2101) resulted in significant inhibition of osteoclastogenesis and RANKL-induced resorption evaluated by double TRAP/ALP staining, survival of OCs, and mRNA expression level of osteoblast markers (APL, BSP, OC). In contrast, LSD1 overexpression confirmed the upregulation of Wnt/b-catenin pathway suggesting a possible underlying mechanism for the osteoclastogenesis potency in MM patients with high expression of LSD1.

Taken together, our findings demonstrate a promising epigenetic approach in myeloma therapeutics by targeting the deregulated LSD1-methylome in MM patients earlier than aggressive disease phase.