Kinome Expression Profiling in Multiple Myeloma Identifies New Therapeutic Targets

Multiple Myeloma (MM) account for approximately 10% of hematological malignancies and is the second most common hematological disorder. Active research on MM allowed great improvement in treatment, including proteasomes inhibitors or immunomodulatory agents, that enhanced significantly the patients median survival from 3-4 years in the 90's to 7-8 years nowadays. However there is a vital need for additional therapies since until today, MM is a condition that cannot be cured, and all patients finally relapse. Although numbers of kinases inhibitors are currently in preclinical and clinical development in cancers, kinases have only be poorly studied in MM. Thus only one study (Tiedeman RE Blood 2010) examined the kinome in MM by using a screening of siRNA in Human Myeloma Cell lines (HMCL).

Here, in contrast to Tiedeman et al., we evaluated the potential of kinase involvement directly on patient outcome. We first performed kinome in silico analysis in three independent cohorts of patients, and identified 36 kinases significantly and identically involved in MM patient's survival. We built a Kinome Index (KI) from the 36 kinases expression, and showed that KI: 1) is related to OS and EFS prognosis in 3 independent cohorts, 2) is linked to proliferation and poor prognosis subgroups of MM molecular classification, 3) is associated with relapse. Among the 36 targets, 7 had available inhibitors and were not already consistently studied in MM (PBK, SRPK1, CDC7-DBF4, MELK, CHK1, PLK4, MPS1/TTK). We then tested the therapeutic interest of these 7 kinase inhibitors using 4 HMCLs representing two IL-6 independent (AMO1, OPM2) and two IL-6 dependent (XG1, XG21) cell lines. All tested inhibitors significantly reduced viability of the cells, and had IC50 from the nanomolar (for MELKi, CHK1i and PLK4i) to the micromolar range (for PBKi, SRPK1i, CDC7-DBF4i and MPS1/TTKi). Annexin, PARP and cell cycle analysis following treatment with selected concentration of the inhibitors in AMO1, showed significant increase in apoptosis, associated for most of them with a perturbation of the cell cycle. Treatment of primary myeloma cells from patients with MELKi, CHK1i and PLK4i induced MM cell death while they didn't impacted significantly the cellular microenvironment. Similarly experiments using murine MM cells, from 5TMM models, confirmed the potential of these three inhibitors.

Next, IC20 of the different inhibitors were tested in co-treatment with melphalan, lenaledomide or velcade in XG1, XG21, AMO1 and OPM2 to challenge the capabilities of the selected kinases inhibitors to synergize with major routinely used therapeutics. We didn't observed any synergy for the co-treatments with bortezomib. However, CHK1i, MELKi, PBKi, DBF4i synergized with melphalan to induce MM cell death. Furthermore, CHK1i, MELKi, PLK4i, PBKi, SRPK1i demonstrated also a synergistic effect when combined with lenalidomide, highlighting their potential therapeutic interest in MM. Analysis of apoptosis and cell cycle following co-treatments with melphalan and lenaledomide using AMO1 cell line demonstrated an increase in PARP cleavage and annexin V staining in association with significant cell cycle deregulation. One of the outstanding co-treatment effects was the reversion of the natural AMO1's resistance to lenaledomide with CHK1, MELK and PBK inhibitors used at their IC20. We then tested their therapeutic interest to overcome melphalan drug resistance using wild-type (XG7-WT) and melphalan-resistant (XG7-Mres) cell lines. PLK4 and DBF4 inhibitors demonstrated higher activity on XG7-Mres cells compared to their WT control. In addition CHK1, PBK and DBF4 inhibitors (IC20) were able to re-sensitize XG7-Mres to melphalan at a similar level than XG7-WT. Altogether we identified 36 new potential kinases deregulated in MM in association with a poor prognostic value. Seven of them represent potential new MM therapeutic targets alone or in combination with conventional therapies in high-risk MM patients.