A Small-Molecule Inhibitor Targeting TRIP13 Suppresses Multiple Myeloma Progression

Multiple myeloma is characterized by clonal proliferation of malignant monoclonal plasma cells in the bone marrow (1). Genomic instability, defined by a higher rate of acquisition of genomic changes per cell division compared with normal cells, is a prominent feature of multiple myeloma cells. Approximately 86,000 new patients with multiple myeloma are diagnosed worldwide each year (2). Although the prognosis of patients with multiple myeloma has improved with the increased use of autologous stem cell transplantation and combinations of approved antimyeloma agents such as proteasome inhibitors (bortezomib and carfilzomib), immunomodulatory drugs (lenalidomide and pomalidomide), and mAbs (daratumumab and elotuzumab), 5-year overall survival rate is only 45% (3). Genetic complexity and clonal heterogeneity are the main reasons for cancer treatment failure in patients with multiple myeloma (4). Thus, the identification of a key driver gene for multiple myeloma may enable the specific targeting of these malignant cells.

Accumulating evidence has shown that dysregulated thyroid hormone receptor–interacting protein 13 (TRIP13) protein levels are operational in several tumors, including breast, liver, gastric, lung, prostate cancer, human chronic lymphocytic leukemia, and Wilms tumor (5, 6). TRIP13 is the mouse ortholog of pachytene checkpoint 2 (7). During mitosis, TRIP13 regulates the spindle assembly checkpoint via remodeling of its effector MAD2 from a “closed” (active) into an “open” (inactive) form (8). During meiosis, TRIP13 was found to regulate meiotic recombination in Saccharomyces cerevisiae, Caenorhabditis elegans, and Drosophila (9). A recent study indicated that TRIP13 enhanced nonhomologous end joining (NHEJ) repair and induced treatment resistance via binding to NHEJ proteins KU70/KU80/DNA-PKcs in head and neck cancer (10).

In our previous study, TRIP13 was identified as a chromosome instability gene that was correlated with multiple myeloma drug resistance, disease relapse, and poor outcomes in patients with multiple myeloma (11). TRIP13 was first identified by yeast two-hybrid screening as a protein fragment that was associated with thyroid hormone receptor in a hormone-independent fashion (12). Overexpressing TRIP13 in cancer cells prompted cell growth and drug resistance, while targeting TRIP13 by TRIP13 shRNA inhibited multiple myeloma cell growth, induced cell apoptosis, and reduced the tumor burden in xenograft multiple myeloma mice (11). Our previous results suggested that TRIP13 might serve as a biomarker for multiple myeloma disease development and prognosis, making it a potential target for future therapies.

To identify a TRIP13 inhibitor, detailed structural information of TRIP13 is essential. Although the reported crystal structure of the TRIP13 mutant (E253Q or E253A) provided insight into the mechanism of substrate recognition (8), further structural information of the wild-type TRIP13 protein is needed for specific inhibitor development. In this study, we determined the crystal structure of the wild-type human TRIP13 at a resolution of 2.6 Å. We then identified small-molecule inhibitors of TRIP13 based on its crystal structure via molecular docking and bioassay. A small-molecule inhibitor, designated DCZ0415, was confirmed to bind to TRIP13 by pull-down, nuclear magnetic resonance (NMR) spectroscopy, surface plasmon resonance (SPR) assays. DCZ0415 exhibited significant antimyeloma activity in vitro, in vivo, and in patient multiple myeloma cells. Importantly, DCZ0415 also synergized with melphalan and the histone deacetylase (HDAC) inhibitor panobinostat in multiple myeloma cells.

U266, HEK293T, MOPC-315, and HS-5 cells were commercially obtained from the ATCC. ARP-1, OCI-MY5, RPMI-8226, and H929 cells were provided by Dr. Fenghuang Zhan (University of Iowa, Iowa City, IA). Cell lines were certificated by short tandem repeat analysis (Shanghai Biotechnology Co., Ltd.). Mycoplasma testing was performed using MycoAlert Mycoplasma Detection Kit according to the manufacturer's recommended protocols. Multiple myeloma cells were maintained in RPMI-1640 medium (Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin–streptomycin (Gibco). Human HS-5, HEK293T and mouse MOPC-315 cells were maintained in DMEM (Gibco) supplemented with 10% FBS and 1% penicillin–streptomycin. All cells were maintained in a humidified atmosphere of 5% CO₂ at 37°C, subcultured every 3 days and passaged routinely for use until passage 20. Bone marrow samples were obtained from patients with multiple myeloma after obtaining written informed consent at the Department of Hematology Shanghai Tenth People's Hospital (Shanghai, China). The protocol for collection and usage of clinical samples was approved by the Shanghai Tenth People's Hospital Ethics Committee. Informed consent was obtained in accordance with the Declaration of Helsinki.

DCZ0415 was synthesized by Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai, China. Antibodies for caspase-3, caspase-8, caspase-9, mouse CD4, CD3, and CD8 were purchased from Cell Signaling Technology; TRIP13, BAX, BCL2, CDK4, CDK6, cyclin D, p-p65, α-tubulin, TUNEL, Ki-67 and β-actin were from Abcam. Annexin V-FITC and propidium iodide (PI) detection kit was purchased from BD Pharmingen. Penicillin–streptomycin was purchased from Invitrogen. Puromycin and biotin were purchased from Sigma.

Cells were harvested, aspirated, and washed with cold PBS; they were then lysed and centrifuged at 8,000 × g at 4°C. The lysate was then incubated with 10 μL of DCZ0415-biotin (50 μmol/L) or biotin (50 μmol/L) for 2 hours in the presence of neutrAvidin agarose resins (Thermo Fisher Scientific) with rotation at 4°C. The solution was then centrifuged at 1,500 × g and the supernatant was discarded; it was then washed twice with PBS, centrifuging after each wash, resuspended in SDS, and analyzed by immunoblotting.

TRIP13 protein was prepared in 10 mmol/L sodium acetate (pH = 5.5) and then covalently immobilized on CM5 sensor chip xia amine-coupling procedure. The rest binding sites of the sensor chip were blocked by ethanolamine. The kinetic measurements of compounds were performed at 25°C with Biacore T2000 (GE Healthcare). In this step, compounds were diluted at different concentrations in PBS buffer (10 mmol/L HEPES pH = 7.4, 150 mmol/L NaCl, 3 mmol/L EDTA), and were flowed over the chip at rate of 30 mL/minute. The combining time and dissociation time was set at 120 and 150 seconds, respectively. Data analysis was finished via the state model of T2000 evaluation software (GE Healthcare).

Cell viability assay was performed as described previously (13). Briefly, cells were seeded in triplicate in 96-well plates and then treated with DCZ0415. Cell viability was measured using the Cell Counting Kit (CCK)-8 assays.

Apoptosis assay was performed as described previously (14). Briefly, cells were treated with or without DCZ0415. Then, cells were collected and stained with Annexin-V for 15 minutes and then PI for 5 minutes at room temperature. Stained cells were detected via using flow cytometry.

Wild-type TRIP13 protein was mixed with AMP-PNP at a molar ratio of 1:2 and incubated on ice for 1 hour to allow complex formation. Crystallization was achieved by sitting-drop vapor diffusion at 4°C with the well solution containing 0.1 mol/L bicine, pH 9.0, and 10% (v/v) (±)-2-Methyl-2,4-pentanediol. Crystals were gradually transferred to a harvesting solution containing the precipitant solution and 25% glycerol, prior to flash-freezing them in liquid nitrogen for storage. Native and Se-Met-SAD datasets were collected under cryogenic conditions (100 K) at the beamlines BL18U1 and BL19U1 of the Shanghai Synchrotron Radiation Facility, and were processed using the program HKL3000 (15). The single-wavelength anomalous diffraction (SAD) data phases were calculated using the CCP4i (16) suite and four selenium atoms were located and refined. The initial SAD map was significantly improved by solvent flattening. A model was built into the experimental electron density using the programs CCP4i and Coot (16) and further refined in the program Phenix (17). The native structure was determined by molecular replacement using the crystal structure of Se-Met TRIP13 as the initial model, and further refined in Coot and Phenix (17). Figures of the crystal structures were generated with the program PyMOL (Schrodinger L. The PyMOL Molecular Graphics System, Version 1.8. 2015).

The oligonucleotide sequence specific for TRIP13 silencing (sgRNA) was designed. The packaging plasmids VSVG and psPAX2 were used to produce recombinant lentivirus by transfecting HEK293T cells. Lentiviral transduction of myeloma cell lines was performed using polybrene. Stable cell lines were selected with puromycin (2.5 μg/mL). The efficiency of viral transduction was >95%. Then, PCDH constructs were used to generate human TRIP13 WT and TRIP13 MT overexpression plasmids for transfection into sgTRIP13 cells.

DNA double-strand break (DSB) repair assay was performed as described previously (18). Briefly, the efficiency of NHEJ and homologous recombination (HR) was measured using a GFP-based reporter system.

Nude mice (6 weeks old) were purchased from Shanghai SLAC Laboratory Animal Co., Ltd. (Shanghai, China). Human H929 cells (1 × 10⁶) in 100 μL of serum-free culture medium were subcutaneously injected into the upper flank region of the nude mice. After the tumor growth of mice, mice were randomly assigned to two groups: the control group (DMSO, Tween-80 and saline) and 50 mg/kg DCZ0415-treated group (dissolved in DMSO, Tween-80 and saline solution). BALB/c mice (6 weeks old) were injected subcutaneously in the right flank with or without 5 × 10⁶ MOPC-315 cells in a volume of 0.1 mL. After tumor growth of mice, mice were randomly assigned to two groups: the control group (DMSO, Tween-80, and saline), 25 mg/kg DCZ0415-treated group (dissolved in DMSO, Tween-80, and saline solution). Mice were then administered with or without 25 mg/kg DCZ0415 via intraperitoneal injection every day for 15 days. All mice were euthanized at the end of the experiment and tumors were photographed. Tumor volumes were measured using a Vernier caliper and calculated using the formula: Tumor volume (mm³) = 1/2 × (relatively shorter diameter)² × (relatively longer diameter). All animal studies were approved by the Institutional Review Board of Shanghai Tenth People's Hospital (ID: SYXK 2011-0111).

Statistical analyses were performed using Prism software (GraphPad). Data are expressed as means ± SD. Data were considered statistically significant at P < 0.05. The Student t test was used to compare two groups. The log-rank test was used for survival curves. The combination index (CI) values were calculated by median dose-effect analysis using commercially available software (CalcuSyn; Biosoft). All tests of statistical significance were two sided.

We set out to study the crystal structure of wild-type human TRIP13 (wt_hTRIP13) to aid the discovery and rational design of TRIP13 inhibitors for multiple myeloma therapeutics. Wild-type human TRIP13 protein exists as a mixture of monomer and oligomers in solution (Supplementary Fig. S1A). To obtain homogeneous sample for crystallization, we further purified wt_hTRIP13 monomer using mono-Q ion exchange chromatography and gel filtration chromatography. We used AMP-PNP (a nonhydrolyzable analogue of ATP) to cocrystallize with wt_hTRIP13. AMP-PNP has a binding affinity of 49 μmol/L to wt_hTRIP13 (Supplementary Fig. S1B). We determined the crystal structure of wt_hTRIP13 by single-wavelength anomalous dispersion at a resolution of 2.6 Å (Supplementary Table S1). There is one copy of wt_hTRIP13 per asymmetric unit and the crystal packing belongs to the space group P6₅. Similar to recently reported mutant structure of human TRIP13 (TRIP13^E253Q; ref. 8), wt_hTRIP13 assembled into a helical filament instead of the classic hexamer ring of the AAA⁺ ATPase family, probably due to the crystal packing (Supplementary Fig. S1C).

The wt_hTRIP13 monomer comprises three domains: An N-terminal domain that is involved in substrate recognition, and the large and small AAA domains that form the catalytic site for ATP hydrolysis (Fig. 1A and B). Positive electron densities could be observed inside the ATP-binding cleft of TRIP13 in the Fo-Fc map of the wt_hTRIP13 structure, but we failed to fit the AMP-PNP molecule into this density. Superposition of the wt_hTRIP13 structure with the structure of TRIP13^E253Q mutant in complex with ATP (PDB:5VQA) revealed that the base and phosphate groups of ATP happens to occupy the positive electron densities (Fig. 1C). The key residues that participate in coordinating and hydrolyzing ATP, such as K185/T186 of the Walker A motif, N300 of the Sensor 1 motif, and R386 of the Sensor 2 motif, adopt almost identical conformations between the two structures (Fig. 1C). However, as for the residue E253 of the Walker B motif, its side chain sticks outward from the ATP-binding cleft and seems not to participate in ATP-binding as the residue Q253 in the TRIP13^E253Q mutant (Fig. 1C). Residue E253 has a higher b factor compared with the other parts of the structure, indicating that its conformation might be dynamic and prone to alternate.

On the basis of our wt_hTRIP13 structure, structure-based molecular docking was applied to virtually screen our in-house compound library with 8,000 compounds using Smina, a fork of AutoDock Vina, with default parameters (19, 20). Of the compounds identified, 76 were selected for biological testing (Supplementary Table S2). Viability test of multiple myeloma cells revealed some active compounds, of which, DCZ0415 was the most promising inhibitor based on biology screening (Fig. 2A).

To further validate whether DCZ0415 targeted TRIP13, a series of assays were performed. We employed an affinity pull-down target verification system, in which DCZ0415 was conjugated with a biotin. Compared with the unconjugated biotin, the addition of DCZ0415-biotin to the cell lysate brought down endogenous TRIP13 (Fig. 2B). And a 2% input of total cell lysate was tested by immunoblotting analyses (right; Fig. 2B). In this study, we measured the interaction between DCZ0415 and TRIP13 using NMR. We observed that the CPMG spectra of 200 μmol/L DCZ0415 with the addition of 5, 8, 10 μmol/L TRIP13 (Fig. 2C) and the STD spectrum of 200 μmol/L DCZ0415 with the addition of 5 μmol/L TRIP13 both showed the interaction (Fig. 2D). In our study, we measured the interaction between TRIP13 and DCZ0415 using SPR assay. The binding affinity (K_d) was calculated from SPR data. The measurements displayed a strong affinity with a K_d value of 2.42 ± 1.26 μmol/L (Fig. 2E and F).

To evaluate the inhibitory effect of DCZ0415, multiple myeloma cells were treated with DCZ0415 for 72 hours and cell viability was assessed. Data showed that DCZ0415 induced a significant dose-dependent decrease in viability (Fig. 3A). Using CCK-8 assay, the IC₅₀ value of DCZ0415 was 1.0–10 μmol/L calculated by CalcuSyn in multiple myeloma cell lines. To further investigate the antimyeloma activity of DCZ0415, two representative cell lines were treated with DCZ0415 (0–40 μmol/L) for 24, 48, and 72 hours. We observed that DCZ0415 decreased cell viability in a time- and dose-dependent manner (Supplementary Fig. S2A). To examine the effect of DCZ0415 on the colony formation of multiple myeloma cells, soft-agar clonogenic assays were performed. Multiple myeloma cells treated with DCZ0415 showed a significant decrease in colony formation, indicating that this compound inhibits cell proliferation (Fig. 3B; Supplementary Fig. S2B). EdU assays were employed to examine whether DCZ0415 affects DNA synthesis. Compared with that in the control group, the percentage of EdU-positive cells was significantly decreased with DCZ0415 treatment, indicating that DCZ0415 exerts cytotoxic effects by inhibiting DNA synthesis in multiple myeloma cells (Supplementary Fig. S2C).

Bone marrow stromal cells (BMSC) mediated the paracrine growth of multiple myeloma cells and protect against the cytotoxicity of antimyeloma agents via cytokine secretion (21). To determine whether DCZ0415 could overcome the protective effects of the BM microenvironment, multiple myeloma cells were cultured with or without HS-5 BMSCs in the presence or absence of DCZ0415 (Fig. 3C; Supplementary Fig. S2D). As a positive control, cells were treated with melphalan with or without BMSCs for the same period of time. Results revealed that DCZ0415 treatment significantly inhibited multiple myeloma cell viability in the presence and absence of BMSCs, but melphalan-induced cytotoxicity could be abrogated by BMSCs (Supplementary Fig. S2E).

Cytokines IL6 and insulin growth factor-1 (IGF1), which are secreted by multiple myeloma cells and BMSCs, have been reported to promote multiple myeloma cell proliferation, migration, and drug resistance (22). We therefore examined the effect of DCZ0415 on IL6- or IGF1–induced multiple myeloma cell growth. Multiple myeloma cells were cultured either alone or with IL6 or IGF1. Significant inhibition of IL6- or IGF1-induced multiple myeloma cell growth was observed with DCZ0415 treatment (Fig. 3D; Supplementary Fig. S2F), suggesting that this compound not only directly targets multiple myeloma cells, but can also overcome the cytoprotective effects of the host BM microenvironment.

We next examined whether DCZ0415 functioned by inducing apoptotic cell death (23). Multiple myeloma cells were exposed to various concentrations of DCZ0415, leading to significant increases in the proportion of early- (Annexin-V⁺/PI⁻) and late-stage (Annexin-V⁺/PI⁺) apoptotic cells compared with that in control cells exposed to DMSO (Fig. 3E; Supplementary Fig. S3A). To determine whether DCZ0415 could induce apoptosis in patient multiple myeloma cells, purified CD138⁺ cells were examined from six newly diagnosed patients and four refractory/relapsed patients with multiple myeloma who were refractory to bortezomib. Patient cells showed a dose-dependent relationship between DCZ0415 treatment and apoptotic cell death. This confirmed that DCZ0415 can trigger cytotoxicity in bortezomib-resistant primary myeloma cells. In contrast, DCZ0415 does not significantly induce normal peripheral blood mononuclear cells' (PBMC) apoptosis (Fig. 3F). We tested the sensitivity of cells of PBMCs and patients to TRIP13 expression and the results showed that cells with high expression of TRIP13 were more sensitive to DCZ0415 than those with low expression of TRIP13 (Fig. 3F). Because BCL2 family proteins affect apoptosis via the regulation of cytochrome C release, which then mediates caspase activation (24), the effects of DCZ0415 on the expression of caspase enzymes, antiapoptotic BCL2, and proapoptotic protein BAX were evaluated. Caspase-8 and -9 activities increased in a dose-dependent manner in multiple myeloma cells treated with DCZ0415 for 48 hours. Furthermore, DCZ0415 treatment decreased the expression of BCL2 in a dose-dependent manner but increased the expression of BAX (Supplementary Fig. S3B). Also reduction of TRIP13 induced similar increase in BAX and decrease in BCL2 (Supplementary Fig. S3C). To determine the dependence of DCZ0415-induced apoptosis on the caspase pathway, we assessed the ability of the pan-caspase inhibitor Z-VAD-FMK to protect against cell apoptosis. As shown in Fig. 3G, Z-VAD-FMK partially blocked DCZ0415-induced cell apoptosis as determined by Annexin V–PI staining. These data demonstrate that DCZ0415 triggers caspase-dependent apoptosis in multiple myeloma cells. In addition to apoptosis, effects on cell-cycle progression might be important for the action of anticancer drugs. We therefore evaluated the effects of DCZ0415 on cell-cycle progression by flow cytometry analysis. Treatment induced a significant accumulation in G₀–G₁ multiple myeloma cells (Fig. 3H; Supplementary Fig. S3D). As essential components of the cell-cycle machinery, cyclins function to bind and activate specific cyclin-dependent kinase (CDK) partners. Thus, protein levels of CDK4, CDK6, and cyclin D were evaluated. In agreement with flow cytometry data, we observed a marked dose-dependent decrease in cyclin D, CDK4, and CDK6 after DCZ0415 administration (Supplementary Fig. S3E).

To determine whether the antimyeloma activity of DCZ0415 is dependent on TRIP13, TRIP13-sgRNA and point mutation plasmids were designed. Treatment of sgTRIP13 cells with DCZ0415 led to the loss of sensitivity compared with that of sgControl-transfected wild-type cells (Fig. 4A). sgTRIP13 and sgControl cells were separately treated with melphalan or left untreated. The results showed that sgTRIP13 and sgControl cells were both sensitive to melphalan (Supplementary Fig. S4A and S4B). This finding suggested that cells with deleted TRIP13 were specifically resistant to DCZ0415. We evaluated the effects of DCZ0415 on cell-cycle progression in sgControl and sgTRIP13 cells by using flow cytometry analysis. The sgControl cells were blocked at G₀–G₁ stage under DCZ0415 treatment. However, compared with in sgControl cells, sgTRIP13 cells showed no significant changes in the cell cycle under DCZ0415 treatment (Supplementary Fig. S4C). These results suggested that the antimyeloma activity of DCZ0415 depends on TRIP13. On the basis of the structure of TRIP13 and DCZ0415, we speculated that valine (V140), serine (S187), and arginine (R386) of TRIP13 are essential for DCZ0415 binding. We thus mutated TRIP13 V140/S187/R386 (TRIP13 WT) to TRIP13 alanine (A) 140/187/386 (TRIP13 MT); we then downregulated endogenous TRIP13 expression using the sgTRIP13 target intron sequence and overexpressed TRIP13 WT or TRIP13 MT. Pull-down assay shown that DCZ0145 bound to TRIP13 WT cells, but not TRIP13 MT cells (Fig. 4B). Cell viability of was compared among the groups (TRIP13 WT and TRIP13 MT) with DCZ0415 treatment. We found that DCZ0415 decreased cell viability in the TRIP13 WT group, whereas TRIP13 MT cells were resistant to DCZ0415, as compared with TRIP13 WT cells. The protein expression of TRIP13 was also examined among these groups (Fig. 4C). This suggests that DCZ0415–TRIP13 binding is essential for the antimyeloma activity of DCZ0415. Our findings suggested that TRIP13 is essential for the antimyeloma activity of DCZ0415.

Some anticancer agents sensitize cancer cells by inducing DNA DSBs and DNA damage responses (25). In the response of mammalian cells to DNA DSBs, phosphorylation of histone H2AX (γH2AX) at sites proximal to the DNA breaks has been reported (26). In this study, γH2AX levels were evaluated in multiple myeloma cells following DCZ0415 treatment by immunofluorescence analysis. The results revealed that γH2AX levels were higher in multiple myeloma cells treated with DCZ0415 than baseline γH2AX levels (untreated control), which reflected ongoing DNA damage (Fig. 5A). Ataxia telangiectasia mutated (ATM) protein kinase is a key mediator of this DNA damage response, which induces cell-cycle arrest and facilitates DNA repair by activating downstream targets such as the cell-cycle checkpoint kinase (CHK2; refs. 27, 28). The protein levels of phosphorylated (p)-ATM and phosphorylated (p)-CHK2 were found to be increased in a dose-dependent manner in DCZ0415-treated multiple myeloma cells compared with the untreated control cells (Supplementary Fig. S5A). A DNA damage response accompanied by efficient and appropriate repair of DSBs is essential for the preservation of genomic integrity. However, in cancer, the repair of anticancer agent-induced DSBs by the NHEJ or HR repair pathways promotes treatment resistance and subsequent relapse in patients (26). TRIP13 also enhanced NHEJ repair and induced treatment resistance via binding to NHEJ proteins KU70/KU80/DNA-PKcs in head and neck cancer (10). Thus, we evaluated whether DCZ0415 impaired DNA repair via the NHEJ repair pathways using GFP-based reporter assay, which was an excellent tool to measure the efficiency of NHEJ repair. The results indicated that DCZ0415 suppressed the NHEJ repair pathway (Fig. 5B), which was consistent with TRIP13 promoting DSB-induced NHEJ repair and we confirmed that TRIP13 interacted with the NHEJ key regulator KU70/KU80 (Supplementary Fig. S5B). Besides, our results have shown that 10 μmol/L DCZ0415 had a little effect on HR repair pathway, but 20 μmol/L DCZ0415 could significantly inhibit the HR repair (Supplementary Fig. S5C). With DCZ0415 treatment, the protein level of γH2AX increased in TRIP13 WT cells. However, the protein level of γH2AX in TRIP13 MT cells, which did not bind DCZ0415, showed no significant change under DCZ0415 treatment (Fig. 5C). We studied NHEJ repair in TRIP13 MT and WT cells using a GFP-based reporter assay. Following DCZ0415 treatment, greater levels of NHEJ repair were detected in TRIP13 MT cells compared with TRIP13 WT cells (Fig. 5D). This suggested that TRIP13 MT cells were resistant to DCZ0415, associated with diminished DNA damage and greater NHEJ repair compared with TRIP13 WT cells. Besides, to address the impact of DCZ0415 on mitotic spindles, α-tubulin/DAPI staining was performed. As shown in Supplementary Fig. S5D, DCZ0415 induced spindle multipolarity, suggesting that DCZ0415 acted as a spindle poison, which supported the involvement of TRIP13 in regulation of the mitotic spindle.

NF-κB is aberrantly activated in MM and promotes cell survival and malignancy by upregulating antiapoptotic genes (29). To investigate the possible contribution of the NF-κB signaling pathway to pathogenesis, we investigated whether iκBα and NF-κB p65 phosphorylation were decreased by DCZ0415 treatment. Compared with the untreated control cells, the protein levels of phosphorylated (p)-iκBα and phosphorylated (p)-NF-κB p65 were decreased in multiple myeloma cells treated with DCZ0415 (Fig. 5E), and DCZ0415 showed a strong inhibitory effect on NF-κB–promoter luciferase activity (Fig. 5F). As shown in Fig. 5G, DCZ0415-induced cell apoptosis could be partially rescued by TNFα. These results suggested that cell death induced by DCZ0415 may be mediated via the inhibition of the NF-κB signaling pathway. Importantly, TRIP13 increased NF-κB–promoter luciferase activity (Supplementary Fig. S5E).

To further evaluate its preclinical efficacy, we investigated the effects of DCZ0415 on multiple myeloma cell growth when used in combination with other antimyeloma agents. The cytotoxicity of combined DCZ0415 and melphalan (30) was examined in multiple myeloma cells. Melphalan-induced growth inhibition was enhanced with increasing concentrations of DCZ0415. Calculation of the CI values using CalcuSyn software revealed synergistic effects of DCZ0415 on melphalan against multiple myeloma cells (Fig. 6A and B). Then we tested DCZ0415 in combination with HDAC inhibitor panobinostat (31) in multiple myeloma cells. Isobologram analysis revealed synergy between DCZ0415 and panobinostat with a CI less than 1 (Fig. 6C and D). This provided evidence for the beneficial effects of combination therapy, which could effectively reduce the required concentration of other antimyeloma agents, thereby reducing the potential side effects.

To investigate the therapeutic potential of DCZ0415 in vivo, an multiple myeloma mouse xenograft model was employed. The administration of DCZ0415 significantly reduced the growth of multiple myeloma cells-induced tumors in immunodeficient mice compared with control mice (Fig. 7A). There were no significant differences in the body weight of nude mice in each group, suggesting that DCZ0415 was well tolerated (Fig. 7B). The toxicity of DCZ0415 was also examined by hematoxylin and eosin (H&E) staining of major organs and no significant histologic changes were observed in the liver and kidney of the mice (Supplementary Fig. S6A), indicating that the side effects of DCZ0415 were minimal. Significantly, treatment with DCZ0415 resulted in a significant prolongation in overall survival compared with vehicle-treated animals (Fig. 7C).

Furthermore, we performed a pharmacodynamic study whereby the harvested tumors were analyzed for antiproliferation and apoptosis markers. DCZ0415-treated mice exhibited decreased tumor Ki-67 and p-p65 levels compared with control mice. However, we observed an increase in cleaved caspase-3, TUNEL (apoptosis markers), and γH2AX following DCZ0415 treatment of multiple myeloma cells compared with control cells (Fig. 7D).

To examine DCZ0415 effect on immunocompetent mice, we injected MOPC-315 cells subcutaneously into BALB/c mouse, and treated with 25 mg/kg DCZ0415 or not. With DCZ0415 treatment, the growth of multiple myeloma cells induced tumors significantly was reduced compared with control mice (Fig. 7E). There were no significant changes in the body weight of BALB/c mice in each group (Fig. 7F). IHC studies for CD4, CD3, and CD8 expression from mice postsacrifice show that treatment with DCZ0415 significantly increased immune effect cells infiltration, as evidenced by much greater CD4, CD3, and CD8 staining surrounding MOPC-315 cells remains (Fig. 7G). There was no significant histologic change in the liver and kidneys of mice, as detected by H&E staining, indicating that the side effects of DCZ0415 were minimal (Supplementary Fig. S6B).

These findings suggest that DCZ0415 yields potent antimyeloma responses in mice, prolonging their survival times. These data indicate that TRIP13 inhibitors developed using this approach may provide a roadmap for candidate therapeutic agents in multiple myeloma.

TRIP13 is thought to function as an oncogene based on meta-analysis of gene expression datasets from various cell lines, and is significantly amplified in patients with high-risk multiple myeloma (32). We previously showed that knockdown of TRIP13 inhibited the growth of multiple myeloma both in vitro and in vivo (11). We found that TRIP13 was overexpressed in human multiple myeloma cell lines, further supporting a role for TRIP13 as an oncogene in multiple myeloma. Therefore, we suggested that TRIP13 might represent an antimyeloma target and inhibition of TRIP13 could be a promising strategy in multiple myeloma therapy.

Protein crystal structures can offer invaluable insight into the molecular mechanisms of action of specific proteins. The mechanisms of substrate recognition and remodeling of TRIP13 were revealed by the crystal structure of human TRIP13^E253Q (8). However, it remained necessary to analyze the full-length, wild-type, human TRIP13. In this study, the crystal structure of the wild-type TRIP13 protein was resolved, which not only contributed toward our understanding of the molecular mechanisms of TRIP13 activity, but also helped to identify new inhibitors against TRIP13 activity.

DCZ0415 was identified via structure-based molecular docking and cellular screening from in-house compound library. Our proof-of-concept experiments, which include pull-down, NMR, and SPR assays, provide evidence that DCZ0415 binds to TRIP13. Our studies employed multiple myeloma cell lines, patient multiple myeloma cells and xenograft models, along with biochemical and genetic models, to demonstrate the antimyeloma activity of the TRIP13 inhibitor DCZ0415. DCZ0415 displayed highly potent antimyeloma activity against a large panel of multiple myeloma cell lines. Furthermore, DCZ0415 results in a significant reduction of the proliferation rate in multiple myeloma cells as evidenced by colony formation and EdU expression assays. The DCZ0415-induced reduction in proliferation was associated with inducing apoptosis, arresting the cell cycle.

We previously reported that TRIP13 was an oncogene in multiple myeloma; however, the detailed mechanism by which TRIP13 promotes cell growth was not fully explained. In this study, we found that DCZ0415 (a TRIP13 inhibitor) induced cell death via inhibition of the NF-κB signaling pathway. Our results suggest that TRIP13 acts as an oncogene by activating the NF-κB signaling pathway in multiple myeloma. Using a cellular assay to mimic multiple myeloma in its microenvironment, we observed that DCZ0415 inhibited growth inhibition of multiple myeloma cells even in the presence of BMSCs and the cytokines IL6 and IGF1. This suggested that besides its cytotoxicity on multiple myeloma cells, DCZ0415 also targets the BM microenvironment and overcomes the proliferative effects of BMSCs (Fig. 3C). In BM microenvironment, multiple myeloma cell adhesion to BMSCs triggers the NF-κB–dependent transcription and secretion of cytokines such as IL6 in BMSCs, which further stimulate multiple myeloma cell growth, survival, and drug resistance (22, 33, 34). Moreover, activation of NF-κB by cell adhesion and cytokines augments the binding of multiple myeloma cells to BMSCs, which in turn induces IL6 transcription and secretion in BMSCs. Conversely, the inhibition of NF-κB activity abrogates this response (22, 33, 34). Our data demonstrate that DCZ0415 inhibits NF-κB activity (Fig. 5E and F) and TRIP13 activated NF-κB pathway (Supplementary Fig. S5E). Thus, DCZ0415 might prevent multiple myeloma progression by targeting TRIP13 and inhibiting NF-κB–dependent transcription and secretion of cytokines in BMSCs and the expression of many cell adhesion molecules on both multiple myeloma cells and BMSCs, thereby disrupting the tumor-BM microenvironment interactions that contribute to multiple myeloma progression (33). Our results suggested that treatment of cells with DCZ0415 was correlated with inhibition of NF-κB activity, cell-cycle blockade, and apoptosis, although further in-depth study is required. In the future, we will identify the mechanism underlying DCZ0415 treatment by using genetic tools.

Some anticancer agents induce apoptosis by enhancing DNA damage; however, this is amended by DNA repair, which increases cell survival and induces drug resistance (35, 36). Thus, DNA repair inhibitors have received increased attention in recent years. For example, it was reported that SCR7 was a putative inhibitor of NHEJ, blocking end-joining by interfering with ligase IV binding to DNA, thereby leading to accumulation of DSBs within the cells and culminating in cytotoxicity (36). DCZ0415 not only enhanced DNA damage, but also impeded DNA repair. As an inhibitor of TRIP13, the impediment of DNA repair by DCZ0415 is understandable, which is consistent with previous data that TRIP13 is essential for DSB repair (10). For example, it was revealed that instead of having a checkpoint role, TRIP13 was required for one of the two major classes of recombination in meiosis that was required for repairing DNA breaks. TRIP13 was also shown to be involved in DNA repair induced by programmed DSBs in meiotic recombination (37). Our study indicated that DCZ0415 impaired NHEJ, as shown in Fig. 5B. NHEJ is the major DSB repair pathway in mammalian cells and is activated by DSB ends being recognized by the KU (KU70 and KU80) heterodimer (38). It was recently reported that TRIP13 promoted NHEJ repair in head and neck cancer via binding of KU70 and KU80 (10). Consistent with this, we observed endogenous binding between TRIP13 and KU70/KU80 in multiple myeloma cells, which also indicated that TRIP13 played a key role in NHEJ repair.

Resistance mechanisms present the largest hurdle to the cure of multiple myeloma and it can arise initially or emerge during the course of treatment (39). Thus, to examine the synergistic effects of DCZ0415 and other antimyeloma agents, we tested the effects of DCZ0415 in combination with melphalan and panobinostat. Promisingly, DCZ0415 was able to enhance the cytotoxic effects of both melphalan and panobinostat in multiple myeloma cells. This suggests that DCZ0415 may have promising antimyeloma activity both alone and combined with other antimyeloma agents. In summary, our studies presented crystal structure of the wild-type human TRIP13 and identified an inhibitor of TRIP13. The inhibitor, named as DCZ0415, effectively induced antimyeloma activity in multiple myeloma, both in vitro and in vivo by impairing DSB repair and inhibiting NF-κB pathway. Our present data suggest that TRIP13 could be an important target for refractory/relapsed multiple myeloma.

No potential conflicts of interest were disclosed.

Conception and design: J. Huang, W. Zhu, J. Shi

Development of methodology: Y. Wang, H. Xue, L. Hu, Z. Xu

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Wang, B. Li, H. Xue, L. Hu, X. Sun, S. Chang, L. Gao, Y. Xie, W. Xiao, D. Yu, G. Chen, N. Zhang, Z. Mao

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y. Wang, B. Li, H. Xue, Z. Xu, S. Chang, L. Gao, Y. Tao, H. Xu

Writing, review, and/or revision of the manuscript: Y. Wang, J. Huang, B. Li, H. Xue, G. Tricot, Z. Xu, F. Zhan, W. Zhu, J. Shi

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y. Kong, X. Sun, F. Lian, X. Wu

Study supervision: J. Huang, W. Zhu, J. Shi

This work was supported by grants from the National Natural Science Foundation of China (grant no. 81870158, 81570190, 31570766, U1632130, 81602515, 81529001, and 81670194), National Key R&D Program of China (grant no. 2016YFA0501803, 2017YFA0504504, and 2016YFA0502301), and Chinese Pharmaceutical Association - Yiling Biopharmaceutical Innovation Project (CPAYLJ201908). We thank D Yao, L Wu, W Qin, and R Zhang from the beamlines BL18U1 and BL19U1 at National Center for Protein Science Shanghai (NCPSS) and Shanghai Synchrotron Radiation Facility (SSRF) for help with crystal data collection.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.