Maternal Embryonic Leucine Zipper Kinase (MELK) Drives a High-Risk Gene Network and Represents an Attractive Novel Drug Target in Multiple Myeloma

Background:

In spite of significant progress in the outcome of multiple myeloma (MM), effective treatment options are still missing for high-risk patients. A better understanding of key mechanisms in high-risk disease is therefore a necessary prerequisite for the development of novel therapeutic options for patients with poor prognosis. Here, we evaluated maternal embryonic leucine zipper kinase (MELK) for its implications in the prognosis of MM, its interactions with genes involved in high-risk MM and the impact of MELK inhibition in vitro and in vivo.

Methods:

Expression levels of genes associated with high-risk MM were analyzed in two large cohorts of publically available gene expression (GEP) datasets (GSE2658 and GSE9782; n=551 and n=264, respectively) and by quantitative PCR (qPCR) in human myeloma cell lines (HMCLs). The impact of MELK inhibition was studied using a selective inhibitor of MELK (OTSSP167). MELK and OTSSP167 target genes were analyzed by qPCR and Western Blot. MELK knockdown HMCLs were generated via lentiviral transduction of MELK specific shRNA. For the in vivo experiment, C57BL/KaLwRij mice were injected intravenously with 5x10⁵ GFP-transfected 5TGM.1 cells. Starting the day after injection, two groups of mice were treated by different doses of OTSSP167 (15mg/kg/d or 7.5mg/kg/2d) and one group received the vehicle. Treatment was administered orally.

Results:

MELK expression was significantly elevated in the GEP-defined high-risk proliferation (PR) associated molecular subgroup compared to healthy donor bone marrow plasma cells, but barely detected in any other GEP defined MM subgroup. Consequently, overall survival was significantly shorter in patients with high compared to low MELK levels treated within the total therapy 2 (P=0.0003), total therapy 3 (P=0.04) and the APEX trial protocol (P=0.002). Of note, MELK expression was further elevated in relapsed patients compared to baseline (GSE31161) suggesting a potential role in drug resistance.

In line with their proliferative character, MELK expression was detected in 8 of 8 HMCLs. Treatment with the MELK inhibitor OTSSP167 downregulated MELK protein levels and led to a dose-dependent reduction of viability in all primary MM samples and HMCLs tested (median IC50: 10.16 nM, range: 7.6 - 15.2 nM). We observed induction of apoptosis in all HMCLs investigated, verified by mitochondrial membrane depolarization, annexin V/7-AAD staining, detection of cleaved caspase 3, and cleaved PARP. This was accompanied by downregulation of gene expression levels of IRF4 and MCL-1. In addition, OTSSP167 induced a G2/M cell cycle arrest which was linked to downregulation of cyclin B1, aurora kinase A and PLK-1. Importantly, we also observed reduced clonogenic growth of MM cells treated with OTSSP167 and the anti-myeloma activity of OTSSP167 was upheld in the presence of bone marrow stromal cells.

We next sought to clarify the impact of MELK on other genes implicated in high-risk myeloma. Strikingly, treatment of HMCLs with OTSSP167 reduced MELK and PLK-1 protein levels, but also those of FOXM1, EZH2 and DEPDC1. This relationship was confirmed in HMCLs with shRNA mediated MELK knockdown. All of these genes were significantly elevated in the GEP defined PR subgroup and associated with poor outcome.

Finally, the activity of OTSSP167 was analyzed in the 5TGM.1 model of MM. In vivo, we observed a dose-dependent reduction of tumor parameters by OTSSP167. Low doses of OTSSP167 significantly reduced spleen weight and serum IgG2b levels but only tended to decrease BM infiltration, while high doses of OTSSP167 significantly decreased tumor infiltration analyzed by spleen weight (267mg vs 108mg) and bone marrow infiltration (14% vs 2%, P<0.05).

Conclusion:

The current study identified almost exclusive expression of MELK in the GEP defined PR subgroup of MM. Hence, MELK expression represents a novel poor prognostic marker. Inhibition of MELK with OTSSP167 impaired myeloma cell growth and survival in vitro and in vivo. Moreover, downregulation of MELK via shRNA or OTSSP167 treatment significantly reduced protein levels of several other genes of the GEP defined PR subgroup of myeloma, suggesting a central role for MELK in the signaling network of proliferation associated high-risk myeloma. These results place MELK in the center of a MM high-risk gene network and emphasize to initiate clinical testing of MELK inhibition in MM.