NEDD8-activating enzyme (NAE) is an essential E1 enzyme of the NEDD8 conjugation (neddylation) pathway, which controls cancer cell growth and survival through activation of cullin-RING ubiquitin ligase complexes (CRL). In this study, we describe the preclinical profile of a novel, highly potent, and selective NAE inhibitor, TAS4464. TAS4464 selectively inhibited NAE relative to the other E1s UAE and SAE. TAS4464 treatment inhibited cullin neddylation and subsequently induced the accumulation of CRL substrates such as CDT1, p27, and phosphorylated IκBα in human cancer cell lines. TAS4464 showed greater inhibitory effects than those of the known NAE inhibitor MLN4924 both in enzyme assay and in cells. Cytotoxicity profiling revealed that TAS4464 is highly potent with widespread antiproliferative activity not only for cancer cell lines, but also patient-derived tumor cells. TAS4464 showed prolonged target inhibition in human tumor xenograft mouse models; weekly or twice a week TAS4464 administration led to prominent antitumor activity in multiple human tumor xenograft mouse models including both hematologic and solid tumors without marked weight loss. As a conclusion, TAS4464 is the most potent and highly selective NAE inhibitor reported to date, showing superior antitumor activity with prolonged target inhibition. It is, therefore, a promising agent for the treatment of a variety of tumors including both hematologic and solid tumors. These results support the clinical evaluation of TAS4464 in hematologic and solid tumors.
This article is featured in Highlights of This Issue, p. 1183
The ubiquitin-like modifier NEDD8 controls the stability and activity of its target proteins via a conjugation cascade (the neddylation pathway; ref. 1). NEDD8-activating enzyme (NAE; a heterodimer of APP-BP1 and UBA3), an E1 enzyme, is a key regulator of the neddylation pathway. NAE starts the neddylation pathway by the transfer NEDD8 to its E2 enzyme UBE2M also called as UBC12; similarly, UAE (an E1 enzyme) transfers Ub to UBE2C (an E2 enzyme) in the ubiquitination pathway and SAE (an E1 enzyme) transfers SUMO1 to UBE2I (an E2 enzyme) in the SUMOylation pathway (2, 3). Overexpression of NEDD8 and NAE is reported in multiple types of cancer (4–7), and elevated levels of NEDD8 transcripts are reported to correlate with the poor prognosis of patients with bortezomib-treated multiple myeloma (8). The major function of NEDD8 is to activate cullin-RING ubiquitin ligases (CRL) by covalent modulation of the cullin protein within the CRL complexes (9, 10). Activated CRLs conjugate ubiquitin to their substrate proteins, and the ubiquitinated proteins are degraded by proteasomes (11). Hyperactivation of CRL complexes is also reported in several cancer types (4, 5, 12, 13). CRL-mediated substrate degradation controls a variety of cellular processes, including proliferation, cell cycle, DNA damage responses, autophagy, senescence, and apoptosis (14–22). Therefore, NAE inhibition should disrupt the turnover of the substrate proteins of CRLs via CRL inactivation and is expected to show antitumor efficacy by disturbing the amount of the various proteins that contribute to cancer cell growth and survival.
Although several studies have sought to identify NAE inhibitors as new potent anticancer agents, only MLN4924/TAK-924 (pevonedistat) has been formally clinically developed as an NAE inhibitor (15, 23, 24). On the basis of the preclinical evidence (8, 25–27), clinical trials of MLN4924 have been conducted to examine its efficacy in solid and hematologic tumors with various dosing regimens. The results of the recently published phase I trials of MLN4924 indicate that although some complete responses were observed in the acute myeloid leukemia (AML) trial, the antitumor efficacy of MLN4924 as a single agent was marginal (28–31). Moreover, acute and severe elevation of liver functional markers were caused by 50 mg/m2 of MLN4924 when given as five consecutive doses and it was shown to be dose limiting in this dosing schedule (30). To minimize the risk of liver toxicity, MLN4924 is needed to be administered with some rest periods in the clinical trials (32). In addition, because of the limited antitumor efficacy of MLN4924 as a stand-alone agent, trials examining MLN4924 in combination with other chemotherapeutic agents are ongoing (NCT03268954).
Here, we present the preclinical profile of a novel, highly potent, and selective NAE inhibitor, TAS4464, which exerts profound antitumor activity as a single agent due to its long-acting NAE inhibition in tumors when dosed intermittently, and show TAS4464 has a wider therapeutic index compared with that of MLN4924.
Materials and Methods
Recombinant E1, E2, and their Ubls (Ubiquitin-like proteins) were all purchased from commercial suppliers. NAE (UBA3/APP-BP1), GST-UBE2M, GST-UBE2C, and GST-UBE2I were purchased from Ubiquigent Ltd., and biotin-NEDD8 and UAE were purchased from Enzo Life Sciences International, Inc. Biotin-Ubiquitin, biotin-SUMO1, and SAE (SAE1/UBA2) were purchased from R&D Systems, Inc.
TAS4464 was designed and synthetized at Taiho Pharmaceutical Co., Ltd. The structure was confirmed by ESI-MS and NMR, and its purity exceeded 99% as measured by HPLC. 14C-radiolabeled TAS4464 was synthetized at Curachem Inc. MLN4924 was synthesized at the NARD institute Ltd. Bortezomib was purchased from Selleck Chemicals LLC. Ibrutinib was synthetized at Taiho Pharmaceutical Co., Ltd. Doxorubicin hydrochloride was purchased from Kyowa Hakko Kirin Co., Ltd. Pazopanib was purchased from LC Laboratories, Inc.
The E1 enzyme inhibitory activities of the NAE inhibitors were evaluated in an assay of Ub/Ubl thioester transfer activity from E1 to E2 (27). Carbonic anhydrase II (CA2) assays were performed by Eurofins Panlabs, Inc.
Cell lines and cell culture
Cell lines were purchased as follows: human T-cell acute lymphoblastic leukemia CCRF-CEM (EC85112105-F0) from Dainippon Pharmaceutical Co., Ltd.; human mantle cell lymphoma (MCL) GRANTA-519, human follicular lymphoma DOHH2, and human diffuse large B-cell lymphoma (DLBCL) OCI-LY18 and U-2940 from DSMZ; human multiple myeloma KMS-26 from the Japanese Collection of Research Bioresources Cell Bank; human chronic lymphocytic leukemia BHL-89 from DS Pharma Biomedical Co., Ltd.; and human AML THP-1, human Burkitt's lymphoma Daudi, DLBCL Pfeiffer, human colon carcinoma HCT116, human small-cell lung cancer (SCLC) NCI-H211, and human clear cell sarcoma SU-CCS-1 from the ATCC. TMD8 cells were supplied by Dr. Shuji Tohda, Tokyo Medical and Dental University (Tokyo, Japan). To establish a systemic model, TMD8 cells were engineered to stably express luciferase (TMD8-Luc) by using a lentiviral vector (Life Technologies). All cell lines used were authenticated by means of short tandem repeat–based DNA profiling (Supplementary Materials and Methods).
THP-1 was cultured in RPMI1640 medium supplemented with 0.05 mmol/L 2-mercaptoethanol and 10% FBS. CCRF-CEM was cultured in RPMI1640 supplemented with 20% FBS. BHL-89 was cultured in RPMI1640 supplemented with 15% FBS. Daudi, DOHH2, U-2932, OCI-LY18, Pfeiffer, U-2940, KMS-26, NCI-H211, SU-CCS-1, and TMD8-Luc were cultured in RPMI1640 (ATCC Modification) supplemented with 10% FBS. HCT116 was cultured in McCoy's 5A medium supplemented with 10% FBS. GRANTA-519 was cultured in DMEM supplemented with 10% FBS.
Growth inhibition assay of cell lines
Cells were treated with TAS4464 for 72 hours. Cell viability was measured by using the CellTiter-Glo 2.0 Assay (Promega Corporation). Compound sensitivity large-screen data were obtained by using the OncoPanel cancer cell line panel (OncoPanel: http://www.Eurofinspanlabs.com).
Growth inhibition assays of patient-derived samples
Human peripheral blood mononuclear cells (PBMC) from three donors were purchased from Cellular Technology Ltd. Gibco human pooled cryopreserved hepatocytes from 5 donors (mixture) were purchased from Thermo Fisher Scientific K.K.
DLBCL permanent models (patient-based, xenograft-derived cells) and primary AML patient cells were tested at Oncotest. Oncotest's patient-derived models were collected from patients in agreement with existing German and European laws. Small-cell lung cancer permanent model LU5195 (patient-based, xenograft-derived cells) was tested at Crown Biosciences Inc. Ovary and endometrial cancer patient–based, xenograft-derived cells were tested at Fukushima Medical University (Fukushima, Japan). All samples were obtained from patients with prior, free, and written informed consent. All procedures were carried out according to guidelines from the Declaration of Helsinki and performed after approved by an institutional review board. PBMC, AML, DLBCL, and SCLC cells were treated with TAS4464 for 72 hours, after which viability was measured. Patient-derived ovary and endometrial cells were maintained in low-attachment plates and treated with TAS4464 or chemotherapeutic agents for 6 days. Cell viability was then measured by using the CellTiter-Glo 3D Cell Viability Assay (Promega Corp.). Hepatocytes were cultured in Cell-able plate (Sumitomo Bakelite Co., Ltd.) for 10 days and then treated with TAS4464 for 24 hours, after which viability was measured by using CellTiter-Glo 3D Cell Viability Assay.
Immunoblotting and antibodies
Chemiluminescense or fluorescent Western blot analysis was carried out as described previously (33), using antibodies against NEDD8 from Abcam plc and Santa Cruz Biotechnology, Inc.; NRF2, Cullin1, and α-tubulin from Abcam plc; CDT1, phosphorylated IκBα (p-IκBα), p21, p27, Cullin4, cleaved PARP, cleaved caspase-3, and cleaved caspase-8 from Cell Signaling Technology, Inc.; and GADD34, UBC12 and PTTG from Santa Cruz Biotechnology, Inc.; Cullin2 from Thermo Fisher Scientific Inc.; Cullin3 from Bethyl Laboratories, Inc. In chemiluminescence Western blot, blots were then developed by means of luminol-based enhanced chemiluminescence (Thermo Fisher Scientific Inc.) and images were captured with an LAS-3000 imaging system (Fuji Photo Film Co., Ltd.). For fluorescent Western blot analysis, IRDye antibodies were provided from LI-COR, Inc. and two-color multiplex detection were performed using near-infrared fluorescence. Images were captured on the LI-COR Biosciences Odyssey Infrared Imaging System.
In vivo efficacy studies
CCRF-CEM cells or SU-CCS-1 cells were subcutaneously implanted into 6-week-old male BALB/cAJcl-nu/nu mice (CLEA Japan, Inc.). GRANTA-519 cells were subcutaneously implanted into 6-week-old male C.B-17 SCID mice (Charles River Laboratories Japan, Inc.). These cells were allowed to grow to a volume of >100 mm3 before the mice were randomized. Six mice were assigned to each group for each experiment, and MLN4924, TAS4464, or TAS4464-vehicle [5% (w/v) glucose solution] was then administered intravenously either weekly or twice weekly. Tumor volume was calculated with the following formula: [length × (width)2]/2. Statistical significance was calculated by using Dunnett test to assess the difference in tumor volume between the control (TAS4464-vehicle treated) and TAS4464-treated groups. Welch t test was used to compare the tumor volume in the TAS-4464 treatment group and the comparative group. For all tests, P < 0.05 was considered statistically significant. For pharmacodynamic analysis, tumors were harvested at the indicated time points after administration of TAS4464 or MLN4924. The excised tumors were homogenized in lysing matrix D (MP Biomedicals) containing lysis buffer. Lysates were then centrifuged at 17,800 × g for 10 minutes and the supernatants were collected. The amounts and phosphorylation status of proteins were evaluated by means of Western blotting with the appropriate antibodies.
TMD8-Luc cells were injected (1 × 107) into tail vein of 6-week-old SCID mice. Eleven days after transplantation, TAS4464 was intravenously administered at a dosage of 100 mg/kg/day on days 1, 4, 8, and 11 of a 21-day cycle for 8 cycles. Mice were intravenously administered d-luciferin potassium salt (Promega Corp.) at a dose of 150 mg/kg/day, and then dorsal and ventral bioluminescence images were taken with an IVIS Lumina II Imaging System (Perkin Elmer). Images were analyzed with the Living Image 3.1 software (Perkin Elmer). The treatment efficacy of TAS4464 was determined by log-rank statistical analyses and plotted by using the Kaplan–Meier method.
These animal experiments were performed with the approval of the institutional animal care and use committee of Taiho Pharmaceutical Co., Ltd. and carried out according to the guidelines for animal experiments of Taiho Pharmaceutical Co., Ltd.
LU5266 SCLC patient derived cells (P5) were subcutaneously implanted into 6-week-old female NOD-SCID mice. The cells were allowed to grow to a volume of >150 mm3 before the randomization. 75 mg/kg/day of TAS4464 or TAS4464-vehicle [5% (w/v) glucose solution] was then administered intravenously either weekly or twice weekly for 3 weeks. 8 mg/kg/day of etoposide and 4 mg/kg/day of cisplatin were administered intravenously on days 1, 2, 3, and day 1. This study was in strict accordance with applicable Crown Bioscience, Inc. Guidelines and Standard Operation Procedures.
TAS4464 is a highly potent and selective inhibitor of NAE
TAS4464 was discovered by means of library screening and structure-based design (Fig. 1A) and its synthetic procedure is described in Supplementary Fig. S1. The IC50 values of TAS4464 against NAE, UAE, and SAE were 0.955 nmol/L, 449 nmol/L, and 1,280 nmol/L, respectively, in the assay of Ub/Ubl thioester transfer activity from each E1 enzyme to the corresponding E2 enzyme (Fig. 1B). In comparison, the IC50 value of MLN4924 against NAE was 10.5 nmol/L, which is consistent with previously reported values (1). In addition, we assessed the inhibitory activity of TAS4464 and MLN4924 against Carbonic anhydrase II (CA2), because MLN4924 is known to inhibit CA2 as an off-target (34). The IC50 values of TAS4464 and MLN4924 against CA2 were 0.730 μmol/L and 0.0167 μmol/L, respectively. These results demonstrate that TAS4464 functionally inhibits NAE and that this inhibitory activity is more potent and more selective than that of MLN4924. The E1 selectivity of TAS4464 was further evaluated at the cellular level. The transition of Ubl to its corresponding E2 was evaluated by use of Western blotting. Western blot analysis revealed that among NAE, UAE, and SAE, TAS4464 selectively inhibits NAE in cells (Supplementary Fig. S2A). Furthermore, we hypothesized that TAS4464 creates a covalent adduct with NEDD8 on NAE due to its sulfonamide structure, which is reported to be the key element to create the covalent bond between an inhibitor and Ubl on E1 enzymes in an ATP-dependent manner (35, 36). HCT116 human colon carcinoma cells were treated with 14C-radiolabeled TAS4464, and proteins were separated by SDS-PAGE and visualized by autoradiography. The labeled TAS4464 was observed in the gel where a protein the size of NEDD8 would be anticipated, indicating that TAS4464 formed an adduct with NEDD8 (Fig. 1C). This effect was nullified by pretreatment with nonlabeled TAS4464 in a dose-dependent manner. These results suggest that TAS4464 adducts with NEDD8 as a part of its inhibitory mechanisms of neddylation in cells (Supplementary Fig. S2B).
We next evaluated the effects of TAS4464 on the levels of various cellular proteins in the human cancer cell line CCRF-CEM (acute lymphoblastic leukemia) by use of Western blotting. A 4-hour treatment with TAS4464 (dose range, 0.001–1 μmol/L) induced dose-dependent decreases in neddylated cullin and dose-dependent accumulation of the CRL substrates CDT1 (37), NRF2 (38), p-IκBα (19), and p27 (39) (Fig. 1D). Neddylated cullin was detected by NEDD8 antibody at the size of cullin (95 kDa). Therefore, TAS4464 induced inhibition of neddylation pathway and accumulation of CRL substrate proteins in cells. Although MLN4924 also induced inhibition of neddylation pathway and accumulation of CRL substrate proteins, its effects were only achieved when a much higher dose range than that for TAS4464 was used (Fig. 1D). This difference reflects the strength of the enzyme-inhibitory activities of these two inhibitors.
We next examined the time dependency of the effects of TAS4464 by treating CCRF-CEM, HCT116, and THP-1 cells for 1, 4, 8, 16, and 24 hours with 100 nmol/L TAS4464 (Fig. 1E and F; Supplementary Fig. S3A). In CCRF-CEM cells, TAS4464 induced rapid decreases in neddylated Ubc12 and cullin within the first hour, and time-dependent accumulation of the CRL substrates CDT1, NRF2, p-IκBα, and p27. In addition, we tried to evaluate the amount of the neddylated and unneddylated UBC12 and cullins by fluorescent Western blot analysis (Fig. 1F; Supplementary Fig. S2C). Dual-color detection of UBC12 or cullin and NEDD8 demonstrated that the amount of UBC12 and cullin proteins was not changed by TAS4464 treatment, whereas lower levels of NEDD8 were detected at positions corresponding to the neddylated forms of the proteins. These results demonstrate that the reduction in the levels of neddylated proteins was caused by the inhibition of neddylation. Within 24 hours of its addition to the cells, TAS4464 increased the levels of cleaved caspase-3 and cleaved PARP, which are essential molecules for apoptosis. Importantly, TAS4464 did not induce the accumulation of the non-CRL substrates such as GADD34 and PTTG. Similar expression patterns were observed with other cell lines (Supplementary Fig. S3A). These results suggest that TAS4464-mediated inhibition of NAE diminishes neddylation pathway while increasing the accumulation of CRL substrates, resulting in apoptosis in the cells.
Cytotoxicity profile of TAS4464
The effects of concentration (0.001 to 10 μmol/L) and exposure time (1 to 72 hours) on the growth-inhibitory effect of TAS4464 were assessed at 72 hours after initiation of exposure to cells (Fig. 2A; Supplementary Fig. S3B and S3C). The net growth (%) of CCRF-CEM cells under the various exposure conditions is shown in Fig. 2A, and the net growth (%) of HCT116 and THP-1 cells is shown in Supplementary Fig. S3B and S3C. In each cell line examined, the net growth (%) decreased as the concentration and exposure time of TAS4464 increased, indicating that TAS4464 exerted antiproliferative and cytotoxic activity in a concentration- and time-dependent manner. In all cell lines tested, the GI50 values were almost identical for the treatment schedules of 24 and 72 hours of exposure, indicating that exposure to TAS4464 for 24 hours is sufficient to exert maximal antitumor efficacy. In addition, the cell-killing activity of TAS4464 [net growth (%) of < 0] was seen after only 1 hour of exposure. These results suggest that TAS4464 would exert strong antiproliferative activity and cell-killing activity even following short-term exposure.
We compared the cellular potency of TAS4464 with that of MLN4924 in several cell lines; the 72-hour GI50 values obtained are summarized in Fig. 2B. TAS4464 displayed 3- to 64-fold higher potency than MLN4924 against all tested cell lines. Because TAS4464 showed much higher anti-neddylation activity than MLN4924 (Fig. 1C), we considered these stronger antiproliferative effects of TAS4464 to be the result of greater NAE inhibitory activity in cells.
To determine tumor-type sensitivity in preclinical models, we profiled TAS4464 against 240 human tumor cell lines of various tissue origin and molecular background (OncoPanel). TAS4464 showed widespread antiproliferative activity against a panel of cancer cell lines (Supplementary Table S1); Fig. 2C shows a waterfall plot of the IC50 values of TAS4464 by tumor type. TAS4464 was highly active against most hematologic malignancy cell lines, including leukemia-, lymphoma-, and myeloma-derived cell lines. A number of solid tumor-derived cells were also sensitive to TAS4464. These data suggest that TAS4464 has a potential to treat not only hematologic malignancies but also solid tumors. To determine the molecular marker of the TAS4464 sensitivity in solid tumors, we investigated the correlation with gene expression profile and TAS4464 sensitivity (Supplementary Fig. S4 and Supplementary Materials and Methods). There was no single biomarker that predicted the sensitivity; however, we identified that gene signature consisting from neddylated substrates (40) can be used to classify the sensitive/insensitive cells.
Cytotoxicity of TAS4464 in patient-derived cells
The growth-inhibitory effects of TAS4464 in cancer patient-derived cells were also evaluated. TAS4464 showed cytotoxicity in AML patient-derived cells (Fig. 3A) and DLBCL patient-derived cells (Fig. 3B). There is a high unmet medical need in the treatment of patients with platinum-refractory or relapsed SCLC. Therefore, to evaluate the antiproliferative activity of TAS4464, we chose cells obtained from a patient who had been treated with a platinum agent. TAS4464 showed antiproliferative activity at a lower dose range than of cisplatin in SCLC patient-derived cells (Fig. 3C). While the IC50 values of TAS4464 in patient-derived AML, DLBCL, and SCLC were 1.6–460 nmol/L, 0.7–4,223 nmol/L, and 0.2 nmol/L, respectively, the IC50 value of TAS4464 in human PBMCs was only 5.24 ± 3.63 μmol/L (Fig. 3D). From these results, TAS4464 would be expected to show cytotoxicity in cancer cells selectively at a dose range with no PBMC sensitivity in a 72-hour cytotoxicity assay.
In addition to the 2D cytotoxicity assay for solid tumor patient-derived cells, we also tested 3D colony formation and viability after exposure to TAS4464 or a chemotherapeutic agent. Endometrial and ovarian cancer patient-derived cells were maintained in low-attachment plates and treated with TAS4464, carboplatin, or paclitaxel for 6 days and then viability was measured. In this assay, TAS4464 showed cytotoxicity even in chemotherapy-resistant cells (Table 1). Even in a 3D culture condition, TAS4464 showed minimal cytotoxicity in human normal cells. The cultured human primary hepatocytes were treated with 300 nmol/L or 1,000 nmol/L of TAS4464 for 24 hours and relative viability was 82% and 83%, respectively.
|.||.||Treatment .||IC50 (μmol/L) .|
|Type .||Sample .||history .||TAS4464 .||Paclitaxel .||Carboplatin .|
|#4||PTX+CBDCA, DXR, CDDP||4.11||0.01||8.49|
|#5||PTX+CBDCA, DXR, CDDP||>10||0.4||>10|
|#2||CPT-11, CDDP, PTX, CBDCA||0.92||>10||>10|
|#3||PTX, CBDCA, CPT-11, CDDP, GEM||0.44||0.01||3.51|
|.||.||Treatment .||IC50 (μmol/L) .|
|Type .||Sample .||history .||TAS4464 .||Paclitaxel .||Carboplatin .|
|#4||PTX+CBDCA, DXR, CDDP||4.11||0.01||8.49|
|#5||PTX+CBDCA, DXR, CDDP||>10||0.4||>10|
|#2||CPT-11, CDDP, PTX, CBDCA||0.92||>10||>10|
|#3||PTX, CBDCA, CPT-11, CDDP, GEM||0.44||0.01||3.51|
Abbreviations: CBDCA, carboplatin; CDDP, cisplatin; CPT-11, camptothecin-11; DXR, doxorubicin; GEM, gemcitabine; PTX, paclitaxel.
Strong and sustained neddylation inhibition of TAS4464 leads to profound antitumor efficacy in CCRF-CEM xenograft model
Following these promising in vitro findings, the in vivo potential of TAS4464 was explored by examining its antitumor efficacy in tumor xenograft models. TAS4464 was intravenously administered to mice at dosages of 6.3, 12.5, 25, 50, and 100 mg/kg/day once a week for 3 weeks in a CCRF-CEM human acute lymphoblastic leukemia xenograft mouse model. The mean tumor volume of each treatment group was significantly lower than that of the vehicle control group (P < 0.05, Dunnett test). The T/C ratio (tumor volume of TA4464-treated mice versus vehicle-treated mice) was 35, 20, 10, 1, and 0% for the 6.3, 12.5, 25, 50, and 100 mg/kg dose, respectively. TAS4464 at dosages of 25, 50, or 100 mg/kg/day induced complete tumor regression. We simultaneously evaluated MLN4924 in the same xenograft mouse model by administering 120 mg/kg/day of MLN4924 intravenously to the mice twice a week because doubled dose of MLN4924 was a lethal dose in this model and found that its antitumor efficacy was limited (Fig. 4A and B). Body weight changes during the treatment period were monitored and no marked decreases were observed in any of the test groups (Supplementary Fig. S5A and S5B). These results indicate that TAS4464 had a wide therapeutic window in the CCRF-CEM model and causes tumor shrinkage at less than the maximum tolerant dose.
To study in vivo target inhibition, the effects of TAS4464 and MLN4924 on the levels of various proteins in tumor tissue were evaluated in the same model. TAS4464 decreased neddylation of cullin1 and led to accumulation of CRL substrates CDT1, NRF2, and p-IκBα (Fig. 4C). Furthermore, TAS4464 increased levels of the apoptosis-related factors cleaved caspase-3 and cleaved PARP within 24 hours of administration. Although MLN4924 also decreased neddylation of cullin1 and induced accumulation of CRL substrates at 4 hours of administration, marked increases in the levels of cleaved caspase-3 and cleaved PARP were not observed at 24 hours of administration. We also quantitatively evaluated the inhibition of cullin1-neddylation after administration of TAS4464 and MLN4924 by evaluating the NEDD8-cullin1 signal in Western blots. Stronger and more durable reduction of the NEDD8-cullin1 signal was observed as the dose of TAS4464 was increased, and importantly, the NEDD8-cullin1 signal reduction was more durable after administration of TAS4464 than that of MLN4924 (Fig. 4D). Given that TAS4464 induced accumulation of CRL substrates and increased levels of apoptosis-related factors by strong, durable inhibition of neddylation in the tumor tissue, our findings indicate that TAS4464 exerts its potent antitumor effects via inactivation of the NEDD8 conjugation pathway.
TAS4464 is active in multiple preclinical models
Because diverse cell lines were sensitive to TAS4464 treatment in vitro, the efficacy of the compound was further tested in vivo. TAS4464 treatment of three tumor subcutaneous xenograft models—the GRANTA-519 (MCL) xenograft model, the SU-CCS-1 (human clear cell sarcoma) xenograft model, and patient derived SCLC xenograft model—is shown in Fig. 5. In the GRANTA-519 xenograft model, 100 mg/kg/day of TAS4464 was intravenously administered on a weekly or twice weekly dosing schedule. At the evaluation on day 22, the antitumor activity of TAS4464 is statistically significant (P < 0.05, Dunnett test) and exceeded that of bortezomib and ibrutinib (Fig. 5A) when dosed both on a weekly or twice weekly dosing schedule (P < 0.05, Welch t test). The pharmacodynamic data revealed the massive apoptosis induction in this model (Supplementary Fig. S6A). In the SU-CCS-1 xenograft model, TAS4464 was intravenously administered to the study animals on a weekly dosing schedule. The antitumor activity of TAS4464 is statistically significant (P < 0.05, Dunnett test) and TAS4464 markedly induced tumor regression and retained its antitumor efficacy throughout the treatment period (Fig. 5B). In addition, the pharmacodynamic action of TAS4464 was examined and target inhibition in the tumor was observed and cleaved PARP, an apoptotic marker, was already induced at 4 hours after administration (Supplementary Fig. S6B). Hematoxylin–eosin staining revealed that tumors treated with TAS4464 had large necrotic areas on day 3 (Supplementary Fig. S6C) consistent with the rapid induction of apoptosis observed in Supplementary Fig. S6B. In contrast to these findings with TAS4464, the antitumor efficacies of doxorubicin hydrochloride, pazopanib, and MLN4924 were limited (in each group, P < 0.05, Welch t test). In the LU5266, patient-derived SCLC xenograft model, TAS4464 was administered to the study animals on weekly or twice a week dosing schedules for 3 weeks. For this model, the antitumor activity of TAS4464 is statistically significant (P < 0.05, Dunnett test) and a majority of the mice achieved complete tumor regression in TAS4464-treated groups and the observed tumor growth inhibition was maintained beyond 7 weeks (Fig. 5C). The TAS4464 treatment groups were more efficacious than standard-of-care (cisplatin and etoposide) treatment (P < 0.05, Welch t test). Analysis of the mean body weight changes in the studies with the GRANTA-519, SU-CCS-1, and LU5266 models showed that none of the TAS4464 groups experienced weight loss exceeding 10% of their initial body weights during the study period (Supplementary Fig. S5C–S5E).
In addition to these subcutaneous model studies, we investigated the effects of TAS4464 on survival in a luciferase gene-expressing TMD8 (TMD8-Luc) human DLBCL systemically implanted mouse model. TAS4464 was administered at a dose of 100 mg/kg/day on days 1, 4, 8, and 11 of a 21-day cycle for 8 cycles. Kaplan–Meier survival curves for the treatment and control groups are shown in Fig. 5D and the survival time of each animal, median survival time in days, and increase in life span were >159 days and 174%. The survival time of mice administered TAS4464 was significantly longer than that of the control mice, which were administered vehicle alone (log-rank test, P < 0.05). Mean relative photon value, which was used as an index of antitumor effect, was significantly smaller in the mice administered TAS4464 than in the control mice, which received vehicle alone (Fig. 5E). These data demonstrate that TAS4464 prolongs survival and has antitumor effects even in a human DLBCL systemically implanted mouse model.
The neddylation pathway is commonly deregulated in cancer, and the neddylation-specific E1 enzyme NAE is considered a therapeutic target in cancer. Here we report the detailed biologic activity of the novel clinical drug candidate TAS4464, which is a highly potent and selective small-molecule NAE inhibitor.
TAS4464 inhibits neddylation at the enzyme and cellular levels with high potency and E1 selectivity. It results in accumulation of CRL substrate proteins but has no effect on non-CRL substrate proteins. TAS4464-mediated cytotoxicity against cancer cell lines appears to be induced by partial protein-homeostasis perturbation, caused by neddylation inhibition and subsequent CRL substrate protein accumulation. TAS4464 also has less of an off-target effect than MLN4924, since strong carbonic anhydrase II (CA2) inhibition was observed with MLN4924 but not with TAS4464. Administration of MLN4924 can cause massive red blood cell transition due to the high expression of CA2 in red blood cells, an effect that is not observed with administration of TAS4464 (Supplementary Fig. S6D, Supplementary Materials and Methods). This circumvention of CA2 inhibition should therefore lower the risk of electrolyte abnormalities linked to TAS4464 treatment compared with MLN4924 treatment, those are potentially caused by CA2 inhibition (29, 30).
The IC50 values obtained from the large cell panel analysis revealed that most hematologic malignancy-derived cells were broadly sensitive to TAS4464. Many solid tumor-derived cells were also sensitive to TAS4464. In addition, TAS4464 exhibits potent cytotoxicity against patient-derived cancer cells including cells heavily pretreated with chemotherapeutic agents and resistant to such agents. However, several cell lines were relatively insensitive to TAS4464, especially some of solid tumor-derived cells. Because NAE inhibition disrupts the amount of CRL substrate proteins, and many proteins have been recognized as CRL substrates, multiple mechanisms of action for NAE inhibition have been proposed, including NF-κB pathway inhibition by p-IκBα accumulation, Noxa induction through transient Myc accumulation, and cell-cycle–related protein accumulation (26, 41–44). The essential CRL substrate proteins that determine the sensitivity of TAS4464 may depend on the tumor type or genetic background of the cancer cells. Therefore, it seems difficult to find a single predictive biomarker of the efficacy of NAE inhibitors among the many types of cancer. Indeed, we have tried to identify potential biomarkers to predict sensitivity by using 240 human tumor cell panel, we have also not been able to delignate individual gene(s) that was(were) associated with the pattern of sensitivity. However, we found that set of 35 genes could classify sensitive/insensitive cell lines (Supplementary Fig. S3). These genes were reported as neddylated substrates and most of them were overexpressed in sensitive cells. Therefore, it is expected that cancer cells with activated neddylation pathway are sensitive to NAE inhibition. Further studies would be required to investigate the predictive power of the signature model.
TAS4464 demonstrated strong antiproliferative activity and cell-killing activity in several cell lines, and these effects increased in a dose- and time-dependent manner, but plateaued after 24 hours in all of the cell lines we tested. Interestingly, whereas the key contributor to the antiproliferative activity in CRL substrate proteins might have been different, the time-dependency pattern was similar in each cell line. Further analysis of the mechanism of action of TAS4464 in each cancer type is needed. Nevertheless, our results demonstrate that TAS4464 does not need continuous exposure via daily administration or long-time infusion to achieve antitumor efficacy in in vivo models. In fact, TAS4464 demonstrated superior antitumor activity in multiple subcutaneous xenograft models including patient-derived xenograft model compared with conventional therapies with weekly or twice a week dosing schedules. Moreover, TAS4464 treatment significantly prolonged survival and reduced tumor growth in the TMD8-Luc systemic model. Taken together, these results suggest that TAS4464 will be effective for the treatment of cancer when administered on intermittent dosing schedules.
In the CCRF-CEM model, the antitumor activity of TAS4464 and MLN4924 was highly associated with the inhibition of cullin neddylation. Although 25 mg/kg administration of TAS4464 or 120 mg/kg administration of MLN4924 similarly inhibited cullin neddylation after 4 hours of administration, the target inhibition was still observed after 24 hours of administration in only TAS4464 administration. This durable target inhibition caused by TAS4464 administration leads to higher antitumor activity compared with that of MLN4924. Although, we were not able to evaluate durability of the target inhibition when TAS4464 was administered at 100 mg/kg because this administration caused disappearance of the tumor, 100 mg/kg administration of TAS4464 seemed further durable target inhibition. We observed a rapid decrease in the plasma concentration of TAS4464 after intravenous injection (Supplementary Fig. S6E); therefore, durable target inhibition might not correlate with the plasma concentration. Although it is unclear why this pharmacokinetic/pharmacodynamic discrepancy occurred and further study is needed, TAS4464 has a potential to exert strong antitumor activity even in the clinical settings based on its durable NAE inhibition in the tumor tissue.
Remarkably, TAS4464 has a wide therapeutic window in mouse model. TAS4464 demonstrated antitumor efficacy over a dosage range of 6.3 to 100 mg/kg once a week for 3 weeks in a CCRF-CEM subcutaneous xenograft model without marked decreases of body weight. Because elevation of liver function test parameters by MLN4924 occurred after single administration in clinical study and consequent administration of MLN4924 caused severe hepatic toxicity, intermittent dosing schedule may be needed to reduce the risk of hepatotoxicity of NAE inhibition. In this regard, potent antitumor activity with high therapeutic window achieved by intermittent dosing schedule of TAS4464 has a potential to improve the drawback of clinically developed NAE inhibitor.
In conclusion, here we present the detailed mechanisms of action and therapeutic potential of the novel, highly potent, selective NAE inhibitor TAS4464. We report its preclinical pharmacologic activities and pharmacokinetic, pharmacodynamic, and antitumor activities in tumor xenograft models, which support the continued clinical development of this compound. TAS4464 is currently being investigated in phase I dose escalation study in patients with advanced solid tumors and hematologic malignancies.
Disclosure of Potential Conflicts of Interest
All authors are employed at Taiho Pharmaceutical Co., Ltd.
Conception and design: C. Yoshimura, S. Ohkubo
Development of methodology: H. Ochiiwa, S. Tsuji
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): H. Muraoka, H. Ochiiwa, S. Tsuji, A. Hashimoto, H. Kazuno, F. Nakagawa, Y. Komiya, T. Takenaka
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): H. Muraoka, H. Ochiiwa, H. Kazuno, F. Nakagawa, S. Suzuki, M. Kumazaki, N. Fujita
Writing, review, and/or revision of the manuscript: C. Yoshimura, H. Muraoka, H. Kazuno, Y. Komiya, S. Suzuki, S. Ohkubo
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): C. Yoshimura, H. Muraoka, F. Nakagawa, Y. Komiya, T. Mizutani
Study supervision: C. Yoshimura, S. Suzuki
The authors thank Drs. Teruhiro Utsugi, Kazuhiko Yonekura, and Kenichi Matsuo for their insightful discussions, and Hidenori Fujita, Yayoi Fujikoka, Keiji Ishida, and Hiroko Hitotsumachi for their technical assistance. This study was funded by Taiho Pharmaceutical Co., Ltd., Japan.
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.