Summary paragraph
“Undruggable” proteins, such as RAS proteins, remain problematic despite efforts to discover inhibitors against them. KRAS mutants are prevalent in human cancers. Although KRAS G12C inhibitors have been developed recently, there are no effective inhibitors for KRAS G12D/V. Here, we described the development of a novel chemical knockdown strategy, termed CANDDY (Chemical knockdown with Affinity aNd Degradation DYnamics). This strategy, which is not an inhibition strategy, involves a CANDDY tag modified from a proteasome inhibitor. The tag induces direct proteasomal degradation. We constructed TUS-007 as a multispecific small molecule tethered from a KRAS interactor and CANDDY tag to target KRAS G12D/V. TUS-007 successfully suppressed tumors due to the degradation of KRAS G12D/V. We confirmed that the CANDDY tag-induced degradation was independent of target ubiquitination. The CANDDY technology could represent a simple and practical way to degrade currently “undruggable” proteins.
The majority (75%) of disease-causing proteins are “undruggable” (i.e., difficult to inhibit with small molecules). This difficulty reflects in the presence of smooth surfaces and lack of deep pockets, including proteins associated with cancer drivers and many interfaces for protein-protein interaction (PPI) (1, 2, 3). RAS family members (e.g., KRAS, HRAS, and NRAS) are most challenging to inhibit by small molecules (1, 4, 5, 6, 7). Mutations of KRAS, especially G12C, G12D, and G12V, are frequent in human cancers (7). The RAS protein activities depend on nucleotide loading in their GTP- binding pockets. The inhibition of this pocket has been attempted for nearly 40 years. However, progress has been hindered by the exceptionally high affinity between GTP and RAS proteins (1, 4, 8). While clinically effective inhibitor candidates targeting KRAS G12C have been developed recently (6, 9), there is still no effective inhibitor for KRAS G12D/V. These are important targets found in 95% of pancreatic cancers and 64% of colon cancers (7, 8). Inhibitors of the PPI between RAS and Son of Sevenless 1 (SOS1) (RAS-SOS inhibitor) have been investigated. Inhibitors that directly bind to RAS are not effective owing to their low affinity (5). Thus, the current inhibition technologies are not enough to be effective for undruggable proteins.
Pharmaceutical research on undruggable proteins, such as RAS, focuses on novel modalities instead of inhibition (10, 11, 12). Protein destabilization technologies, using matchmakers between the target and E3 ligase for target ubiquitination, are expected to be effective against undruggable proteins (13, 14, 15, 16, 17). However, matchmaker design has been hampered by the dependency on target ubiquitination (16). It is difficult to select a suitable E3 ligase for a target of interest because of limited knowledge about the mechanism underlying substrate recognition by E3 ligases. Even if the target has an established corresponding E3 ligase, there may be no available ligand for the E3 ligase. Moreover, target ubiquitination may not always induce proteolysis, as in the case of RAS, in which ubiquitination also regulates protein localization and activation (18). Despite such difficulties, an effective inhibitor (9) was used recently for a proteolysis inducer of KRAS G12C (19). However, no proteolysis inducer for KRAS G12D/V has been reported. Current protein destabilization technologies that depend on ubiquitination have limited efficacy for undruggable targets (14). Therefore, to modulate diverse undruggable targets, the difficulties in matchmaking should be eliminated (13).
Here, we report successful KRAS G12D/V degradation, mediated by a novel tool named CANDDY (Chemical knockdown with Affinity aNd Degradation DYnamics).
This innovative approach induces direct proteasomal degradation (i.e., chemical knockdown) of the target using a CANDDY tag derived from a proteasome inhibitor that lacks the site for inhibitory activity (inhibitor site). In principle, chemical knockdown occurs without ubiquitination, so it can bypass the difficulties of designing a matchmaker in current protein destabilization techniques (13). To evaluate the utility of CANDDY, we developed a CANDDY molecule called TUS-007 for KRAS G12D/V and demonstrated its application in KRAS G12D/V chemical knockdown in cell-free, in vitro, and in vivo assays and in vivo tumor suppression.
A CANDDY molecule for targeting KRAS G12D/V proteins
CANDDY molecules are bispecific molecules constructed from two modules; a target interactor, which enables specific binding to the target, and a CANDDY tag, which induces proteasomal degradation (Fig. 1a). The CANDDY tag is essential in CANDDY technology and is a derivative lacking the inhibitor site of MLN2238 (MLN) (20), a clinical proteasome inhibitor (Fig. 1b and Additional Information). To design TUS-007, we employed a RAS-SOS inhibitor (5) (shown in Fig. 1c) which directly binds to KRAS G12D, KRAS G12V, and wild-types of KRAS and HRAS as the target interactor module. Since the binding to wild-type NRAS have not been reported, it was expected to avoid the severe toxicity observed in a pan-RAS inhibitor (21). RAS-SOS inhibitor was conjugated to the CANDDY tag using an NH2 linker (Fig. 1c and Additional Information).
a, Principle of the CANDDY technology. Each CANDDY molecule includes a target interactor (burgundy hexagon) and a CANDDY tag (yellow oval). The target protein is bound by the CANDDY molecule via the target interactor and is directly degraded by proteasome via the CANDDY tag. b, The CANDDY tag was synthesized by chemically inactivating the inhibitor site of MLM2238 (MLN). c, Structures of the RAS-SOS inhibitor (ref. 3), RAS-SOS-NH2, and TUS-007. d, Evaluation of the target-binding affinity of TUS-007 to KRAS mutants by a thermal shift assay. Recombinant KRAS was treated with the vehicle, RAS-SOS inhibitor, or TUS-007 for 30 min, heated at the indicated temperature for 20 min, then analyzed by immunoblotting. e, KRAS G12D protein level was lower after incubation with TUS-007 at the indicated concentrations in the presence of 26S proteasome, indicating successful KRAS G12D degradation by TUS-007 (mean ± SEM; n = 3). *P < 0.05 and ** P < 0.01 vs. DMSO. f, KRAS G12D and G12V levels, but not the G12C level, were decreased in RAS-less MEFs after incubation with TUS-007 for 72 h. Representative blots are shown. The relative amounts of KRAS mutants normalized to GAPDH are shown in the bar graph (mean ± SEM; n = 3–4). *P < 0.05 and ** P < 0.01 vs. DMSO. g, Decreased cell viability was observed in RAS-less MEFs expressing human KRAS G12D and G12V, but not those expressing G12C, after incubation with TUS-007 for 72 h (mean ± SEM; n = 3–4). *P < 0.05 and ** P < 0.01 vs. DMSO.
To compare the target binding of TUS-007 with that of the RAS-SOS inhibitor, we performed a thermal shift assay. Unexpectedly, KRAS G12D/V proteins incubated with TUS-007 were more resistant against heat treatment than those incubated with RAS-SOS inhibitor (Fig. 1d). Alternatively, in a fluorescence-based thermal shift assay, the Tm value of KRAS G12D incubated with TUS-007 was higher than that of KRAS G12D incubated with RAS-SOS inhibitor (Supplementary Fig. 1). These results suggest the higher affinity of TUS-007 to KRAS compared to RAS-SOS inhibitor. In addition, we confirmed that TUS-007 had no inhibitory activity against the catalytic β-subunits of proteasome (Supplementary Fig. 2). This demonstrated that the CANDDY tag, modified from a proteasome inhibitor, hardly inhibit proteasome activity.
A CANDDY molecule degraded KRAS G12D/V proteins in cell-free assay
We attempted a chemical knockdown of KRAS G12D in the presence of 26S proteasomes and in the absence of E3 ligase in a cell-free assay. Successful degradation was performed (Fig. 1e) with 50% degradation concentration (DC50) at 4 µM (Supplementary Fig. 3a). The chemical knockdown mediated by TUS-007 was counteracted by the presence of MLN (Supplementary Fig. 3b). Additionally, RAS- SOS-NH2 as a degradation-incompetent control failed to induce the chemical knockdown (Supplementary Fig. 3b), demonstrating CANDDY tag is essential to induce the chemical knockdown. Although RAS-SOS inhibitor showed 80% inhibition of RAS-SOS PPI at 1 mM (5), DC80 of TUS-007 for KRAS G12D was estimated as 16 µM (Supplementary Fig. 3a). It implied that the conjugation with CANDDY tag drastically improved the usefulness of RAS-SOS inhibitor. We also confirmed chemical knockdown of KRAS G12V by TUS-007 in a cell free assay (Supplementary Fig. 3c).
Thus, the chemical knockdown of KRAS G12D/V by TUS-007 depended only on proteasomes and not on ubiquitination. Importantly, the conjugation of CANDDY tag enabled the direct proteasomal induction and the simple cell free assay for chemical knockdown could be applied, because of the lack of ubiquitination process.
TUS-007 induced targets-selective chemical knockdown in vitro
Using an in vitro assay, we then evaluated the chemical knockdown and investigated the selectivity of TUS-007. We utilized RAS-less mouse embryonic fibroblasts (MEFs) expressing human KRAS G12D, G12V, or G12C. These MEFs do not proliferate in the absence of RAS (22). Immunoblotting analysis revealed that TUS- 007 induced the chemical knockdown of KRAS G12D/V, but not G12C, in RAS-less MEFs (Fig. 1f). TUS-007 reduced the viability of RAS-less MEFs expressing KRAS G12D/V but did not affect the viability of the cells expressing KRAS G12C (Fig. 1g), confirming the selectivity of TUS-007 for KRAS G12D/V. Moreover, the results from RAS-less MEFs expressing KRAS G12C showed that TUS-007 did not cause non- specific cytotoxic effects even for high concentrations such as 100 µM (Fig. 1f & g).
Additionally, we assessed the selectivity of TUS-007 for the human RAS family, including wild-type KRAS, HRAS, and NRAS. TUS-007 did not affect the NRAS protein levels or the viability of RAS-less MEFs expressing NRAS but attenuated KRAS and HRAS levels and reduced the viability of RAS-less MEFs expressing KRAS and HRAS (Supplementary Fig. 4a and b). This result is consistent with the previously reported data on the selectivity of the RAS-SOS inhibitor (5). Therefore, considering that TUS-007 is not a pan-RAS degrader, it is expected not to induce intolerable toxicity, as has been previously observed with a pan-RAS inhibitor in vivo (21). Furthermore, the viability of RAS-less MEFs expressing KRAS G12C or NRAS maintained even at 100 µM TUS-007. These results suggested that TUS-007 did not have remarkable off-target effect resulting in toxicity in cell even at high concentrations.
TUS-007 exerted anti-tumor activity against KRAS G12V-driven colon cancer
Approximately 64% of human colon cancers reportedly express KRAS mutants. KRAS G12V is found in nearly half of the patients and is correlated with poor prognosis (8). Therefore, we investigated whether TUS-007 can suppress the growth of human KRAS G12V-driven colon cancer cells. We conducted an experiment with the SW620-Luc KRAS G12V homozygous human colon cancer cell line (23). TUS-007 induced chemical knockdown of KRAS (Fig. 2a), accompanied by an increase in the annexin V-positive fraction in SW620-Luc cells (Fig. 2b). However, RAS-SOS-NH2 (a synthetic intermediate without CANDDY tag, Fig. 1c) and cetuximab did not induce apoptosis (Fig. 2b and Supplementary Fig. 5). In contrast, treatment with TUS-007 did not result in significant changes in the annexin V-positive fraction in HT29-Luc RAS- independent colon cancer cells (Fig. 2c and Supplementary Fig. 6). Alternatively, the apoptosis induction in SW620-Luc cells was confirmed by caspase 3/7 activation (Fig. 2d). Importantly, TUS-007 induced the apoptosis in SW620-Luc cells at the same concentration at which the chemical knockdown of KRAS was significant (Fig. 2a and b). These findings indicated that TUS-007 selectively induces apoptosis in SW620-Luc cells by chemical knockdown of KRAS G12V in vitro. Additionally, the results in HT29-luc cells also indicated no remarkable off-target effect resulting in toxicity even at high concentrations.
a, Immunoblotting analysis of KRAS G12V chemical knockdown in SW620-Luc cells treated with TUS-007 for 48 h at the indicated doses. The KRAS band intensity was normalized to that of GAPDH (mean ± SEM; n = 3). *P < 0.05 and ** P < 0.01 vs. DMSO. b & c, TUS-007 induced apoptosis in colon cancer SW620-Luc cells (b) but not in HT29-Luc cells (c), a KRAS-independent, BRAF-dependent cell line (Ctx + Vem: 100 mg/ml cetuximab and 40 µM vemurafenib). Cells were treated with each compound for 48 h in terms of the number of annexin V-positive/PI-negative apoptotic cells detected by flow cytometry (mean ± SEM; n = 2–5). **P < 0.01, and ***P < 0.001 vs. DMSO. d, Relative activity of caspase 3/7 in SW620-Luc cells treated with TUS-007 or 5-fluorouracil (5-FU) for 48 h (mean ± SEM; n = 3). *P < 0.05 and ** P < 0.01 vs. DMSO. e, Tumor volume in mice with SW620-Luc cells transplanted subcutaneously and treated with i.p. administration every 3 days (mean ± SEM; n = 6–8). *P < 0.05 vs. vehicle alone. f, Immunoblotting of KRAS G12V chemical knockdown in tumors from the same mice used in e (mean ± SEM; n = 6). **P < 0.01 vs. vehicle alone.
Next, to assess the effectiveness of TUS-007 in vivo, we transplanted SW620- Luc cells subcutaneously in immunodeficient mice. TUS-007 or cetuximab was administered to the xenograft mice by intraperitoneal (i.p.) injection. TUS-007 significantly attenuated tumor progression (Fig. 2e; Supplementary Fig. 7a and b) and induced KRAS G12V chemical knockdown in tumors (Fig. 2f). The body weights of the mice were not affected by TUS-007 treatment (Supplementary Fig. 7c). These results suggested that TUS-007 is effective in vivo against KRAS G12V-driven tumors.
TUS-007 exerted anti-tumor activity against KRAS G12D-driven pancreatic cancer
Approximately 95% of pancreatic cancers harbor KRAS mutations, and the G12D mutation is particularly prevalent and strongly correlated with poor prognosis (8). Therefore, we examined the effects of TUS-007 in human pancreatic cancer cell lines. As expected, TUS-007 degraded the KRAS protein (Fig. 3a) and induced apoptosis (Fig. 3b) in SW1990 KRAS G12D-driven pancreatic cancer cells (24). In addition, apoptosis was induced with caspase 3/7 activation in SW1990 cells (Fig. 3c), which was counteracted by the proteasome inhibitor MLN but not by the NAE E3 ligase inhibitor (NAEi) (Fig. 3d). Furethermore, the apoptosis induction in SW1990 cells was detected around concentrations of chemical knockdown of KRAS (Fig. 3a and b). Therefore, it suggested that TUS-007 induced apoptosis owing to target degradation independent of ubiquitination.
a, KRAS G12D chemical knockdown induced by TUS-007 in SW1990 cells. SW1990 cells were treated with TUS-007 for 72 h and analyzed by immunoblotting. The KRAS G12D band intensity was normalized to that of GAPDH (mean ± SEM; n = 4). *P < 0.05 vs. DMSO. b, TUS-007 treatment for 72 h increased the proportion of annexin V-positive apoptotic SW1990 cells compared with those treated with DMSO or RAS- SOS-NH2 (mean ± SEM; n = 3). *P < 0.05 vs. DMSO or RAS-SOS-NH2. c, TUS-007 increased the caspase 3/7 activity in SW1990 cells after treatment for 96 h (mean ± SEM; n = 3). *P < 0.05 and **P < 0.01 vs. DMSO. d, Caspase 3/7 activation by TUS-007 was counteracted by proteasome inhibition by MLN2238 (MLN) but not by E3 ligase inhibition by NAE inhibitor in SW1990 cells treated with the agents for 96 h (mean ± SEM; n = 3). **P < 0.01 vs. DMSO, ##P < 0.01 vs. TUS-007 (100 µM) only, and †no statistically significant difference compared with TUS-007 (100 µM) only. e, Tumor volume in mice with SW1990 cells transplanted subcutaneously and treated with p.o. administered TUS-007 or vehicle (mean ± SEM; n = 6–9). The agents were administered every 3 days. **P < 0.01 vs. vehicle alone. f, Body weight changes in mice with SW1990 cells transplanted subcutaneously (mean ± SEM; n = 6–9). P.o. treatment with TUS-007 did not affect the body weight.
In mice with SW1990 cells implanted subcutaneously, the oral (per os; p.o.) administration of TUS-007 significantly suppressed the tumor growth (Fig. 3e and Supplementary Fig. 8a) and reduced tumor weight (Supplementary Fig. 8b). The i.p. injection of TUS-007 also supressed subcutaneously implanted tumor growth in mice (Supplementary Fig. 8c). Both p.o. (Fig. 3f) and i.p. (Supplementary Fig. 8d) administrations showed no change in body weight. Here, we examined the chemical knockdown of KRAS, HRAS and NRAS with TUS-007 in pancreas from the identical mice used in Fig. 3e & f. In accordant with the results in RAS-less MEFs, TUS-007 induced the degradation of KRAS and HRAS but not NRAS (Supplementary Fig 10). Thus, there was no sign of toxicity such as body weight losses in contrast to a pan-RAS inhibitor (21).
Moreover, extracellular signal-regulate kinase (ERK) and AKT phosphorylation were decreased by both p.o. and i.p. administrations (Supplementary Fig. 9a & b), indicating that both RAS-mitogen-activation protein (RAS-MAP) and RAS- phosphoinositide 3-kinase (RAS-PI3K) signaling were inhibited. It is difficult for inhibitors to inhibit both RAS-MAP and RAS-PI3K signaling, even using a clinically effective KRAS G12C inhibitor (6). Thus, the chemical knockdown of KRAS by CANDDY might offer a clinical benefit. Taken together, these results indicated that TUS-007 exerted in vivo antitumor activity via the chemical knockdown of KRAS G12D/V and suppression of KRAS signaling.
Here, when an i.p dose of 80 mg/kg TUS-007 was administered to healthy mice, a concentration (45 ng/mg, equivalent to 53 µM) of TUS-007 was observed in the pancreas (Supplementary Table 1.). This TUS-007 concentration in vivo agreed with the concentration at which chemical knockdown and apoptotic indicators were observed in SW1990 cells in vitro (Fig. 3a, b and c). Therefore, these results suggest that there is no contradiction between our in vivo and in vitro results.
TUS-007 exerted anti-tumor activity even in orthotopic xenograft model mice
Finally, SW1990-Luc cells were transplanted directly into the pancreas of mice as an orthotopic xenograft model. The mice were subsequently treated p.o. with TUS-007. Remarkably, in vivo imaging revealed reductions in tumor growth (Fig. 4a and b; Supplementary Fig. 11a) and tumor weight (Supplementary Fig. 11b). Results from immunohistochemical analysis (Fig. 4c left panels and left bar graph) and immunoblotting (Supplementary Fig. 11c) confirmed the chemical knockdown of KRAS G12D in the tumor tissues of mice treated with TUS-007. Furthermore, higher concentrations of positive cells were observed in the terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay in the tumor tissues of TUS-007- treated mice, implying the induction of apoptosis in the TUS-007-treated tumors (Fig. 4c right panels and right bar graph). The body weights of the orthotopic xenograft model mice were not affected by TUS-007 (Supplementary Fig. 11d). Overall, these results demonstrated that oral treatment with TUS-007 induces KRAS G12D chemical knockdown and suppresses pancreatic tumor growth without the sign of weight loss and significant toxicity in the orthotopic xenograft model.
a, Images showing the luciferase activity and total flux in orthotopic tumors transplanted SW1990-Luc cells in vivo. b, Tumor growth (bioluminescence) in identical mice treated with vehicle or TUS-007 administered p.o. every 3 days. (mean ± SEM; n = 6). *P < 0.05 vs. vehicle alone. c, Representative histochemical staining of KRAS and TUNEL in orthotopic tumor sections from the same mice used in a. The scale bar is 60 µm. The bar graphs showed quantitative analyses of the KRAS G12D-positive area (left) and TUNEL-positive area (right) (mean ± SEM; n = 6). **P < 0.01 vs. vehicle alone.
CANDDY applied to another undruggable target
To evaluate the versatility of CANDDY approach, we designed a CANDDY molecule targeting MDM2 (MDM2-CANDDY), a negative regulator of P53, using a P53-MDM2 PPI inhibitor (25) as an interactor module of a CANDDY molecule (Supplementary Fig. 12a and Additional Information). MDM2-CANDDY successfully induced the chemical knockdown of MDM2 in HCT-116 human colon cancer cells (Supplementary Fig. 12b). Although the reactivation of P53 by inhibition of MDM2 has been expected as a novel therapeutic approach (26), P53-MDM2 PPI is still well- recognized “undruggable” target. This result suggested the potential versatility of CANDDY to target other undruggable proteins.
No effective drug to inhibit or degrade KRAS G12D/V has been reported yet. TUS-007 successfully induced chemical knockdown of KRAS G12D/V, resulting in tumor suppression in vivo. Generally, it has been difficult to find effective inhibitors for undruggable targets. Nevertheless, just conjugating CANDDY tag to RAS-SOS inhibitor produced TUS-007, an effective agent even for oral administrations in vivo. Our results demonstrated that TUS-007 induced chemical knockdown of KRAS G12D/V via a proteasomal process, which was independent of ubiquitination.
Importantly, we confirmed that chemical knockdown of KRAS induced apoptosis (Fig. 2a and b; Fig. 3a and b). Also, the consistency between our in vivo and in vitro results was confirmed by the pancreatic concentration of TUS-007, which agreed with the effective concentratin in vitro (Fig. 3a, b and c; Supplementary Table1).
Additionally, a previous study showed that the moderate decrease of KRAS protein inhibited tumor growth via more strong inhibition of downstream signal (12), similarly as that in Supplementary Fig. 9a & b. In fact, the chemical knockdown of KRAS by TUS-007 was moderate (Supplementary Fig. 11c) but suppressed the tumor growth in vivo (Fig. 4a and b). The future study of TUS-007 may overcome the current clinical challenge related to the treatment refractory nature of KRAS G12D/V-positive neoplasms.
The CANDDY tag, which was modified from a proteasome inhibitor, can eliminate the difficulties of matchmaking in current protein degradation (13), considering that proteasomes, unlike E3 ligases, have no substrate selectivity. Moreover, CANDDY tag was also effective in degrading MDM2 (Supplementary Fig. 12). These results suggest that CANDDY technology could be applied to other undruggable targets without efficient inhibitors. Hence, CANDDY technology could provide a simple and practical approach to induce the chemical knockdown of currently undruggable targets.
Data availability
Source data are provided for all experiments. Other data that support the findings of this study are available from the corresponding author, upon reasonable request.
Author contributions
S. Im., L. H., S. It., and E. M. S. designed the study. S. Im., L. H., S. It., M. I., M. T., M. S. and T. Y. performed the in vitro and in vivo experiments. Y. I. performed chemical analyses. E. M. S. supervised the study. All authors discussed the study and approved the submitted manuscript.
Competing interests
The authors declare no competing interests.
Additional Information is available in the online version of the paper.
Supplementary Fig. 1. Estimation of Tm value of KRAS G12D incubated with TUS-007. KRAS G12D was mixed with DMSO, RAS-SOS inhibitor (4 µM) or TUS-007 (4 µM) and incubated under heating from 25 °C to 99 °C. The denature of KRAS G12D was monitored by the fluorescence. The typical curve of each group was shown in upper panels. The means of Tm values were shown in lower table (mean ± SEM; n = 2). * P < 0.05 vs. DMSO, # P < 0.05 vs. RAS-SOS inhibitor.
Supplementary Fig. 2. Effects of TUS-007 on proteasome activity levels. The levels of chymotrypsin-like (b5), trypsin-like (b2), and caspase-like (b1) proteasome activities was monitored by Suc-LLVY-AMC, Bz-VGR-AMC, and Z-LLE- AMC, respectively. AMC fluorescence was monitored by a plate reader with excitation and emission filters of 360 and 460 nm, respectively (DMSO, 30 min = 1) (mean ± SEM; n = 2-3).
Supplementary Fig. 3. TUS-007 induced degradation of KRAS G12D/V in cell free assay. a, Evaluation of the correlation between relative amount of KRAS G12D and concentration of TUS-007 were approximated with Rodbard. The DC50 was estimated about 4 µM. b, A proteasome inhibitor MLN2238 repressed KRAS G12D chemical knockdown by 26S proteasome. RAS-SOS inhibitor and RAS-SOS NH2 did not induce KRAS G12D chemical knockdown. KRAS G12D incubated with 26S proteasome and agents as indicated for 3 h. c, KRAS G12V protein level was lower after incubation with TUS-007 at the indicated concentrations in the presence of 26S proteasome for 1 h, indicating successful KRAS G12V degradation by TUS-007 (mean ± SEM; n = 3). * P < 0.05 vs. DMSO.
Supplementary Fig. 4. Selective chemical knockdown of RAS variants in RAS-less MEFs expressing different types of human RAS. a, Relative viability of RAS-less MEFs expressing WT human RAS family members after incubation with TUS-007 or DMSO. (mean ± SEM; n = 3-5). *P < 0.05 and **P < 0.01 vs. DMSO. b, Degradation of WT human RAS family members in RAS-less MEFs treated with TUS-007 or DMSO for 72 h (mean ± SEM; n = 4-5). *P < 0.05 and **P < 0.01 vs. DMSO.
Supplementary Fig. 5. FACS plots for Annexin V- PI staining of SW620-Luc cells. SW620-Luc cells were treated with the indicated agents for 48 h, followed by staining with Annexin V-FITC and PI. The typical plots are shown.
Supplementary Fig. 6. FACS plots for Annexin V- PI staining of HT-29-Luc cells. HT-29-Luc cells were treated with the indicated agents for 48 h, followed by staining with Annexin V-FITC and PI. The typical plots are shown.
Supplementary Fig. 7. Effects of TUS-007 on the growth of colon cancer subcutaneous xenografts and toxicity. a, Tumors of SW620-Luc cells at day 21 from the same mice shown in Fig. 2e (n = 6–8). b, Comparison of SW620-Luc tumor weight at 21 days after injection (mean ± SEM; n = 6–8). *P < 0.05 vs. vehicle alone. c, Body weight changes in mice with SW620-Luc cells transplanted subcutaneously (mean ± SEM; n = 6–8). Treatment with TUS-007 did not affect the body weight.
Supplementary Fig. 8. Effects of TUS-007 on the growth of pancreatic cancer subcutaneous xenografts and toxicity. a, Tumors of SW1990 cells from the same mice shown in Fig. 3e at 21 days after p.o. administration. b, Comparison of SW1990 tumor weight in a in this figure at 21 days after p.o. administration (mean ± SEM; n = 6-9). *P < 0.05 vs. vehicle alone. c, Tumor volume in mice with SW1990 cells transplanted subcutaneously and treated with i.p. administered TUS-007 or vehicle (mean ± SEM; n = 6-9). The agents were administered every three days. *P < 0.05 vs. vehicle. d, Body weight changes in mice with SW1990 cells transplanted subcutaneously (mean ± SEM; n = 6-9). i.p. treatment with TUS-007 did not affect the body weight.
Supplementary Fig. 9. Effect of TUS-007 on RAS signaling in of pancreatic cancer subcutaneous xenografts. Immunoblotting of KRAS and downstream signaling molecules in tumors from the same mice used in Fig 3a (a: p.o treatment) and Supplementary Fig.8c (b: i.p. treatment). The quantitative analysis of the immunoblotting was shown as bar graph, where the KRAS values were normalized to GAPDH, the p-ERK values were normalized to the total ERK, and the p-AKT values were normalized to the total AKT (mean ± SEM; n = 6). *P < 0.05 vs. vehicle.
Supplementary Fig. 10. Effects of oral treatment with TUS-007 on wild type RAS in pancreas. The immunoblotting analyses of wt RAS proteins in pancreas from the same mice shown in Fig. 3e. TUS-007 degraded KRAS and HRAS but not NRAS (mean ± SEM; n = 6-9). *P < 0.05 vs. vehicle alone.
Supplementary Fig. 11. Effect of TUS-007 on growth of orthotopic pancreatic cancer xenografts expressing mutant KRAS. a, Tumor growth, as assessed by luciferase signal, in individual mice orthotopically transplanted SW1990-Luc cells treated with TUS-007 (red) or vehicle alone (blue) (mean ± SEM; n = 6). b, Pancreases from orthotopic xenograft model mice at day 21 after treatment with TUS-007 (lower-left panel) or the vehicle alone (upper-left panel) and comparison of their weights (right graph) (mean ± SEM; n = 6). *P < 0.05 vs. vehicle alone. c, Immunoblotting of KRAS G12D in tumor lysates from the same mice used in Fig. 4a. The bar graph shows quantification of KRAS G12D normalized to GAPDH (mean ± SEM; n = 6). **P < 0.01 vs. vehicle alone. d, Body weight changes in mice subjected to orthotopical transplantation of SW1990-Luc cells (mean ± SEM; n =6). Treatment with TUS-007 did not affect the body weight.
Supplementary Fig. 12. CANDDY induced degradation of MDM2, a common undruggable target. a, The structure of MDM2-CANDDY using P53-MDM2 PPI inhibitor, with IC50 value between 6-25 μM. b, The human colon cancer cells HCT-116 were incubated for 48 h with MDM2-CANDDY. MDM2-CANDDY degraded MDM2 in the dose dependent manner (mean ± SEM. n = 3). ** P < 0.01 vs. DMSO.
Methods
Cell lines
SW1990 and HCT-116 cells were purchased from the American Type Culture Collection (Manassas, VA, USA). SW620-Luc and HT29-Luc cells were purchased from the National Institute of Biomedical Innovation, Health, and Nutrition (Osaka, Japan). RAS-less MEF cell lines (wild-type (wt) KRAS, KRAS G12D, KRAS G12V, KRAS G12C, NRAS and HRAS) were obtained from the National Institutes of Health (NCI RAS Initiative at the Frederick National Laboratory for Cancer Research, Frederick, MD, USA).
Animal studies
BALB/cA-nu/nu and BALB/cA (female, 7-9 weeks old) were purchased from CLEA Japan (Tokyo, Japan). The animals were maintained in conventional housing conditions, with a daily 12-h light/dark cycle, and given food and water ad libitum. The animals were cared for in accordance with the NIH Guide for the Care and Use of Laboratory Animals. All animal experiments were approved by the Committee on Animal Experimentation of the Tokyo University of Science.
Reagents
MLN2238 (A10600) was purchased from AdooQ BioScience (Irvine, CA, USA). NAEi was purchased from AdipoGen (San Diego, CA, USA). Human 26S Proteasome Protein (E-365) and Human 20S Proteasome Protein (E-360) were purchased from R&D Systems (Minneapolis, MN, USA). Human KRAS (G12D) and the corresponding His tag (12259-H07E1) were purchased from Sino Biological (Beijing, China). Human KRAS (G12V), 2-186, and the corresponding His tag (R06-32BH) were obtained from SignalChem Lifesciences (Richmond, British Columbia, Canada).
Measurement of 20S proteasome activity
The chymotrypsin-like (b5), trypsin-like (β2), and caspase-like (b1) activities of the 20S proteasome were measured using a 20S Proteasome Activity Kit GOLD (StressMarq Bioscience Inc., Victoria, British Columbia, Canada). Purified 20S proteasome (0.1 µg) was incubated in the presence of CANDDY molecules (1, 4, 10, 20, 40, 80, and 160 µM) or MLN2238 at the indicated concentrations for 30 min at room temperature. After incubation, 100 mM of fluorogenic peptide substrates, Suc-LLVY-AMC (β5 substrate), Bz-VGR-AMC (β2 substrate), or Z-LLE-AMC (β1 substrate), was added, followed by incubation for 1 h at room temperature. The reaction mixture was then transferred to a 96-well plate, and the fluorecence from hydrolyzed AMC groups was measured using a Synergy HT multi-channel microplate reader (BioTek Instruments, Inc., Winooski, VT, USA) with a 360 nm excitation filter and 460 nm emission filter.
Evaluation of affinity of TUS-007 for KRAS in a thermal shift assay
The KRAS G12D/V (100 nM) was preincubated with TUS-007, a RAS-SOS inhibitor, or 10% DMSO in 75 mM phosphate buffer (pH 7.5) for 30 min at 37 °C (G12D) or for 20 min at 25 °C (G12V). Aliquots of the reaction solution were sampled into separate microtubes and heated for 30 min at 70, 80, or 90 °C for G12D, or at 40, 50, or 60 °C for G12V. After centrifugation at 18,000 ×g for 20 min, the supernatants were analyzed by SDS-PAGE, followed by immunoblotting using an anti-KRAS antibody (1:1000, WH0003845MI, Sigma-Aldrich, St. Louis, MO, USA). Immunoreactive bands were detected using an iBright CL1000 chemiluminator (Thermo Fisher Scientific, Waltham, MA, USA) with an enhanced chemiluminescent substrate for detection of horseradish peroxidase (HRP; Merck, Darmstadt, Germany).
Estiamtion of Tm values using fluorescence based thermal shift assay
To estimate the Tm values, we performed a fluorescence based thermal shift assay using ProteoStat Thermal Shift Stability assay (ENZO Life Sciences, Farmingdale, NY, USA). 1 µg of KRAS G12D was mixed with DMSO (2.5 %), RAS-SOS inhibitor (4 µM) or TUS-007 (4 µM) in 1x assay buffer containing a fluorescent indicator of protein destabilization, ProteoStat TS Detection Reagent, and incubated under heating from 25 °C to 99 °C in StepOne Plus Real-time PCR system (Thermo Fisher Scientific). The kinetics of fluorescent signals were analyzed to calculate Tm values in Protein Thermal Shift Software 1.3 (Thermo Fisher Scientific).
Evaluation of KRAS degradation induced by TUS-007 in a cell-free system
KRAS G12D (final concentration: 5 ng/µL) and 26S proteasome (final concentration: 8 nM) were incubated with TUS-007 (2, 3, 5, 10, 20, 40 µM), RAS-SOS inhibitor (40 µM), RAS-SOS-NH2 (40 µM) or DMSO in 20 mM Tris-HCl buffer (pH 7.5) containing 20% glycerol for 3 h at 37 °C. To assess the effect of MLN on chemical knockdown, KRAS G12D and 26S proteasome were incubated with DMSO or TUS-007 40 µM in the presence or absence of MLN (1 µM) in the same condition. KRAS G12V (final concentration: 5 ng/µL) and 26S proteasome (final concentration: 8 nM) were incubated with TUS-007 at the indicated concentration or DMSO in 75 mM phosphate buffer, containing 1% LABRASOL (Gattefossé, Saint-Priest, France), for 1 h at 37 °C. After centrifugation at 14,000 ×g for 10 min, the supernatants were analyzed by SDS-PAGE, followed by immunoblotting using an anti-KRAS antibody (1:1000, WH0003845MI for G12D, Sigma-Aldrich; and 1:2000, F234 for G12V, Santa Cruz Biotechnology, Inc., Dallas, TX, USA). The immunoreactive bands were detected, as described above. The DC50 values were calculated by a Rodbard approximation using the band intensity obtained from four and three independent experiments for G12D. Immunoblotting was performed as mentioned above.
Cell proliferation assay
To examine the effect of TUS-007 on cell proliferation, the WST-8 assay (Cell Count Reagent SF; Nacalai Tesque, Kyoto, Japan) was performed. RAS-less MEF cells, expressing human RAS, were plated in 96-well plates at a density of 3 × 103 cells per well, cultured overnight, then treated with the indicated concentrations of TUS-007 in 4% FBS containing medium (KRAS G12D), 10% FBS containing medium (KRAS G12V) or serum-free medium (KRAS G12C, wt KRAS, HRAS, NRAS) for 72 h. After incubation, cell viability was measured spectrophotometrically using the WST-8 reagent. The culture medium was removed and WST-8 reagent was mixed with the growth medium at a ratio of 1:10 (with a final volume of 110 µL) and added to each well. The cells were incubated in an atmosphere of 5% CO2 at 37 °C for 1 to 2 h, and the absorbance at 450 nm was measured using the aforementioned Synergy HT multi-channel microplate reader (BioTek Instruments, Inc.).
Evaluation of TUS-007 specificity in cells
RAS-less MEFs were used to evaluate the selective degradation by TUS-007. RAS-less MEFs expressing one of the human WT RAS members (KRAS WT, HRAS WT, or NRAS WT) or KRAS G12C were plated into 6-well plates at a density of 3 × 104 cells per well, cultured overnight, and treated with the indicated concentrations of TUS-007 in serum-free medium for 72 h. RAS-less MEFs expressing KRAS G12D or G12V were plated into 96-well plates at a density of 3 × 103 cells per well, cultured overnight, and treated with the indicated concentrations of TUS-007 in 4% FBS or 10% FBS containing medium, respectively, for 72 h. The cells were lysed using RIPA buffer (Nacalai Tesque., Kyoto, Japan). The lysates were centrifuged at 18,000 ×g for 10 min. The supernatants were analyzed by western blotting using mouse monoclonal antibodies specific to KRAS (1:1000, WH0003845MI; Sigma-Aldrich) and glyceraldehyde 3-phosphate dehydrogenase (GAPDH, 1:10000, sc-32233; Santa Cruz Biotechnology, Inc.) and rabbit polyclonal antibodies specific for HRAS (1:1000, 18295-1-AP; Proteintech, Rosemont, IL, USA) and NRAS (1:1000, 10724-1-AP, Proteintech). Immunoreactive bands were detected using iBright CL1000 chemiluminator or LAS 3000 (Fujifilm, Tokyo, Japan) with an enhanced chemiluminescent substrate for the detection of horseradish peroxidase (HRP; Merck, Darmstadt, Germany). The intensity of the RAS band was quantified using ImageJ software and normalized to the intensity of the GAPDH band.
Measurement of KRAS protein levels in cells by western blotting
SW620-Luc and SW1990 cells were treated with 1% DMSO or TUS-007 in serum-free medium for 48 h (SW620-Luc cells) or 72 h (SW1990 cells). The cells were lysed using a RIPA buffer (Nacalai Tesque., Kyoto, Japan). The precipitates were separated from the soluble fraction by centrifugation at 18,000 ×g for 20 min. The supernatants were analyzed by western blotting with mouse monoclonal antibodies specific for KRAS (1:1000, WH0003845MI; Sigma-Aldrich) and GAPDH (1:20000, sc-32233; Santa Cruz Biotechnology, Inc.). The immunoreactive bands were detected by LAS-3000, as described above. The intensity of the KRAS band was quantified using ImageJ software and normalized to the intensity of the GAPDH band.
Evaluation of caspase activity
To measure caspase activity, SW1990 cells were plated onto 96-well white plates and 96-well clear plates, at a concentration of 8 × 103 cells/well, in a medium containing 10% FBS. The plates were incubated overnight without CO2 equilibration at 37 °C. The medium was replaced with medium containing 2% FBS, DMSO, 100 µM TUS-007, 3 µM MLN2238, or 2.5 µM NAEi and incubated for 96 h. SW620-Luc cells were plated onto 96-well white plates and 96-well clear plates at 2 × 103 cells/well in medium containing 10% FBS. After overnight incubation, the medium was replaced with FBS-free medium containing 1% L-glutamine, DMSO, and 25, 100, or 500 µM TUS-007 for 48 h. The caspase 3/7 Glo assay (Promega, Madison, WI, USA) was performed on the white plates according to the manufacturer’s protocol. The chemiluminescence intensity was measured with a Synergy HT plate reader (BioTek). The cell viability was measured in the clear plates using the WST-8 assay, as described above. Caspase 3/7 activity was obtained as the chemical luminescence intensity of the caspase 3/7 Glo assay normalized to the cell viability.
Analysis of apoptosis by flow cytometry
SW1990, SW620-Luc, and HT-29 cells were seeded in 24-well plates, followed by the addition of compounds at the indicated concentrations. After incubation for 48 h (SW620-Luc, HT-29) or 72 h (SW1990), the cells were harvested using trypsin, washed with phosphate-buffered saline (PBS), and pelleted by centrifugation at 110 ×g for 5 min. The cells were resuspended in 85 µL of binding buffer (Annexin V-FITC Kit, Medical & Biological Laboratories Co., Ltd., Nagoya, Japan). Subsequently, 10 µL of annexin V-FITC and 5 µL of propidium iodide (PI) were added to each sample, followed by incubation in the dark at room temperature for 15 min. After incubation, 400 µL of binding buffer was added to each sample on ice. The cells were then analyzed using a FACSCanto II fluorescence-activated cell sorter (BD Biosciences, Santa Clara, CA, USA) to detect the fluorescein isothiocyanate (FITC) signal with excitation and emission filter wavelengths of 488 nm and 580 nm, respectively, to detect the PI signal. Data were analyzed using FlowJo software (FlowJo, Eugene, OR, USA).
Subcutaneous xenograft model
SW620-Luc (3 × 107 cells/mL) and SW1990 (5 × 107 cells/mL) cells were suspended in PBS, and 100 µL of the single-cell suspension was transplanted subcutaneously into the right flanks of BALB/cA-nu/nu mice using a 26 G syringe. Tumor volumes were calculated as the tumor length × width2 × 0.5. When the tumor volumes reached approximately 100 mm3, the mice were randomized into treatment and control groups of 6 to 8 animals per group. For i.p. treatment, TUS-007 was dissolved in DMSO and then diluted to a final concentration of 10% DMSO in 20% polyethylene glycol (PEG) 400/Tween 80 (1:1 ratio). The TUS-007 solution was administered intraperitoneally into the tumor-bearing mice at a dose of 80 mg/kg every three days. The control mice were treated with the PEG/Tween vehicle alone. For p.o. treatment, TUS-007 was suspended in 0.5% (w/v) carboxymethyl cellulose (CMC) and administered p.o. at a dose of 80 mg/kg every three days. CMC (0.5% w/v) was administered to control mice. Cetuximab was administered to an additional group of SW620-Luc-implanted mice at a dose of 1 mg/mouse every three days.
Orthotopic xenograft model
The pGL4.51 (Luc2/CMV/Neo) vector (Promega) was transfected into SW1990 cells. Successfully transfected cells were selected using G418. SW1990-Luc cells were mixed with growth factor-reduced Matrigel (BD Biosciences) on ice, at a ratio of 1:1 (v/v), to obtain a final cell density of 1.5 × 104 cells/µL. Ten microliters of the cell suspension was injected directly into the pancreas of BALB/cA- nu/nu mice using a 27 G syringe. Three days after cell transplantation, TUS-007 treatment was initiated in six mice. TUS-007 was suspended in 0.5% (w/v) CMC and administered orally at a dose of 160 mg/kg every 3 days. CMC (0.5% w/v) was administered to control mice (n = 6). Tumor luciferase activity was monitored based on the bioluminescence intensity. Fifty microliters of D-luciferin (30 mg/mL in saline) was injected subcutaneously in mice (Promega). Twenty minutes after the injection, bioluminescence images were obtained using an IVIS Lumina LT (Perkin Elmer, Waltham, MA, USA) with an exposure time of 15 s while the mice were under isoflurane anesthesia (Wako, Tokyo, Japan).
Measurement of RAS levels and downstream signaling molecule activation in xenograft tumors
KRAS expression in xenograft tumors and wt RAS protein expression in pancreas were evaluated by immunohistochemistry and western blot analysis. Tumor tissue and pancreas were harvested from the mice harboring subcutaneous xenografts on day 21 of the TUS-007 administration. Tumor tissues were harvested from the orthotopic xenograft mice on day 24 after the TUS-007 administration for consecutive 3 days starting on day 21. A portion of each sample was fixed in 4% paraformaldehyde (Nacalai Tesque., Kyoto, Japan), embedded in paraffin, then cut into 5 µm-thick sections. The sections were deparaffinized and incubated in 0.3% hydrogen peroxide in methanol for 30 min, and antigen retrieval was performed using a 10 mM citrate buffer (pH 6.0) at 121 °C for 20 min. Tumor tissues were subsequently blocked with 1% bovine serum albumin (BSA) in PBS for 30 min and stained with primary antibodies specific for human KRAS (1:200, 12063-1-AP; Proteintech), diluted in 1% BSA, followed by HRP-labeled secondary antibodies (1:600, ab6721; Abcam, Cambridge, UK) in 1% BSA for 30 min at room temperature. The sections were counterstained with hematoxylin and visualized with diaminobenzidine (1.02924.0001; Merck). Digital images were obtained at 40× magnification using a model BZ-9000 microscope (Keyence, Osaka, Japan). The area of KRAS staining was quantified using ImageJ software, and the perimeter-to-area ratio (KRAS-positive area/section) was calculated for a total of 30 images of 6 tumor sections from each group. The remaining part of each tumor sample was homogenized in a cell-lysis buffer (50 mM triethanolamine, 50 mM KCl, 5 mM MgCl2, 0.25 M sucrose, 1 mM phenylmethylsulfonyl fluoride, proteinase inhibitors (Nacalai Tesque., Kyoto, Japan), 1 mM dithiothreitol, and ribonuclease inhibitor (0.2 unit/µl, TaKaRa Bio., Shiga, Japan) for western blot analysis, as described above. Mouse monoclonal antibodies specific for KRAS (1:1000, WH0003845MI; Sigma-Aldrich), HRAS (1:1000, 18295-1-AP; Proteintech, Rosemont, IL, USA), NRAS (1:1000, 10724-1-AP,
Proteintech) and GAPDH (1:20000, sc-32233; Santa Cruz Biotechnology, Inc.) and rabbit antibodies specific for p-Erk1/2 (phospho-p44/42 MAPK) (1:1000, #4370; Cell Signaling Technology, Beverly, MA, USA), Erk1/2 (p44/42 MAPK) (1:1000, #4695; Cell Signaling Technology), p-Akt (1:2000, #4060; Cell Signaling Technology), and Akt (1:2000; #4691; Cell Signaling Technology) were used for the immunohistochemical staining. The immunoreactive bands were detected by LAS-3000 as described above. Quantitative analysis of the western blots was performed using ImageJ software, and the results were normalized to the GAPDH band intensity.
Measurement of apoptosis in xenograft tumors
DNA fragmentation in the tumor tissues was evaluated by a TdT-mediated dUTP nick-end labeling (TUNEL) assay. Paraffin-embedded tissue sections (5 µm thick) were deparaffinized. The TUNEL reaction was performed using the DeadEnd Fluorometric TUNEL System (Promega), according to the manufacturer’s instructions. Fluorescence images were acquired using a BZ-9000 fluorescence microscope (Keyence). TUNEL-positive cells were identified and counted using ImageJ software. In total, 30 images of 6 tumor sections were analyzed from each group.
Determination of TUS-007 in pancreas
BALB/c mice were treated i.p. with TUS-007, and pancreas tissue was collected at suitable time points. Approximately 30-60 mg of tissue was homogenized in 0.6 mL of tissue-lysis buffer (50 mM Tris-HCl pH 8.0, 20 mM EDTA, 10 mM NaCl, 1% SDS) by Micro Smash MS-100 using stainless (5.5 φ) and zirconia (1.0 φ) beads (TOMY SEIKO CO., LTD., Tokyo, Japan), and 1 μL propyl p- hydroxybenzoate (4 mg/mL in DMSO) was added to each sample as an internal standard. Lysates were centrifuged at 18,000 ×g for 10 min and 2.25 mL of methanol/chloroform (ratio of 2:1) was added to the supernatants and the mixture was allowed to stand for 30 min at room temperature. Chloroform and distilled water (0.75 mL each) were added and samples were centrifuged at 1,500 ×g for 15 min. The layer of chloroform was collected. The supernatants were dried under reduced pressure and redissolved in the mobile phase. TUS-007 was determined by HPLC (Alliance e2695, Waters Corporation., MA, USA) using COSMOSIL® C18-AR-II column (150×4.6 mm, Nacalai Tesque., Kyoto, Japan) in the isocratic elution mode with 0.1% trifluoroacetic acid (TFA) in 50:50 (v/v) acetonitrile/water mobile phase at a flow rate of 1.0 mL/min. All reagents for the assay were purchased from Nacalai Tesque. (Kyoto, Japan).
Degradation of MDM2 in HCT-116 cells
HCT-116 human colon-cancer cells were treated with 1% DMSO or TUS-007 in serum-free medium for 24 h. The cells were lysed using a RIPA buffer (Nacalai Tesque). The precipitates were separated from the soluble fraction by centrifugation at 18,000 ×g for 20 min. The supernatants were analyzed by western blotting using a rabbit polyclonal antibody specific for MDM2 (1:1000, sc-965, Santa Cruz Biotechnology, Inc.) and GAPDH (1:20000, sc-32233; Santa Cruz Biotechnology, Inc.). The immunoreactive bands were detected by LAS-3000, as described above. The intensity of the MDM2 band was quantified using ImageJ software and normalized to the intensity of the GAPDH band.
Statistical analyses
Statistical significance was determined with JMP® Pro 14 (SAS Institute Inc., Cary, NC, USA). An unpaired t-test was used to compare pairs of groups under the assumption of normality. A one-way ANOVA with Dunnett’s post hoc analysis was used to compare sets of three or more groups. P-values < 0.05 were considered statistically significant.
Acknowledgments
We thank Masaaki Ozawa, Mai Hasegawa, Takuya Yogi, Mizuki Kanazawa, and Sachi Tsuji for help with the in vitro studies, Yuri Sato for help with the in vivo studies. We thank the NCI RAS initiative for providing the RAS-less MEFs used in this work, the Uchiro Laboratory for performing the high-resolution mass spectrometry analysis at the Research Equipment Center (Tokyo University of Science (TUS), Japan), the Gunji Laboratory (TUS) for the use of their high-performance liquid chromatography instruments, the Iwakura and Abe Laboratories in the animal facility of the Research Institute for Biomedical Sciences (TUS) for maintaining the mice, Fumito Ishizuka and Naoki Takeda (NARD Institute, Ltd.) for contributions to the chemical synthesis, and Hiroyuki Ohashi for the helpful suggestions. This work was financially supported by the Program for Creating STart-ups from Advanced Research and Technology (START Program: 962238), and the joint research with FuturedMe Inc.
References
