ASR490, a Small Molecule, Overrides Aberrant Expression of Notch1 in Colorectal Cancer

Hyperactivation of Notch1 plays a significant role in the pathogenesis of cancer (1). Activation is triggered by binding of ligands to the receptor, which leads to protease (TACE or Kuzbanian proteases) driven sequential cleavages of the receptor followed with cleavage by γ-secretase. The cleaved Notch receptor intracellular domain (NICD) subsequently translocates to the nucleus, which induces the transcriptional activation of Notch target genes, such as HES1 (2). Cleavage of NICD initiates a signaling cascade that has multiple interactive points with other oncogenic pathways (1, 3, 4). Moreover, HES1 activation has been shown to promote colorectal cancer cell resistance to 5-Fu by inducing epithelial-to-mesenchymal transition (EMT; ref. 5). Notch induction also activates several other oncogenic pathways and negatively affects proapoptotic pathways leading to activation of cell proliferation genes (6, 7).

In colorectal cancer, Notch1 signaling is a major pathway that governs cancer cell differentiation and proliferation (8). Its dysregulation has been frequently associated with colorectal cancer pathogenesis, which is the second leading cause of cancer death in men and women (CDC Colorectal stats 2019; fightcolorectalcancer.org 2019). Although recent advances in colorectal cancer treatment have resulted in dramatic reductions in colorectal cancer–related death (9), colorectal cancer–related morbidity in young adults and chemoresistance to existing therapies is a major challenge in curing patients with colorectal cancer (10).The colorectal cancer incidence rate in adults aged ≥50 years decreased by 32%, whereas these incidence rates increased by 22% among adults aged <50 years (11).

Although screening at an early stage can significantly improve survival, most of the patients with colorectal cancer are diagnosed at an advanced stage. Neoadjuvant therapy before surgery, which is followed by chemotherapy, is recommended for such patients. However, pharmacologic therapy often is associated with toxic and harmful side effects, and patients eventually develop chemoresistance (12–14). Changes in cell signaling patterns, such as upregulation of expression or aberrant activation of several important genes such as antiapoptotic factors (BCL-2 and BCL-XL; ref. 15), survival signaling, and EMT signaling, have been shown to be the main causative factors of chemoresistance in colorectal cancer (5).

Mutations in negative regulatory region (NRR) have been attributed to ligand-independent activation of Notch1 and resulted in aggressive malignancies (16). NRR is termed as an activation switch of Notch1 receptor (17). mAbs targeting NRR have shown promise by inhibiting Notch1 cleavage, which resulted in degradation of NICD (16, 18). However, there is no report, to our knowledge, of compounds specifically binding to NRR and affecting Notch1 signaling. Increasing incidences of chemoresistance to existing therapies in advance colorectal cancer and importance of Notch1 signaling in maintenance of oncogenic phenotypes via uncontrolled proliferation, loss of apoptosis, and advancement to metastasis in colorectal cancer make it imperative to further broaden the current treatment paradigm by developing plant-derived novel small molecules, which have low toxicity profile, can target Notch1 signaling and its aberrant activation, and thus overcome these challenges.

We identified a small molecule, ASR490, using structure–activity relationship studies focused on the Withaferin A analogs. ASR490 effectively inhibits colorectal cancer cell growth in both in vitro and in vivo models. Our results also suggest that ASR490 effectively suppressed Notch1 signaling, which resulted in inhibition of EMT in colorectal cancer. Targeting the multifaceted functions of Notch1 receptor and several interlinked signaling pathways in colorectal cancer with a plant-derived potent small molecule presents a promising approach for the treatment of colon cancer.

ASR490 (Pyridine-2-carboxylic acid {17-[1-(5-hydroxymethyl-4-methyl-6-oxo-3,6-dihydro-2H-pyran-2-yl)-ethyl]-10,13-dimethyl-1-oxo-, 4,6,7,8,9,10,11,12,13,14,15,16,17-tetradecahydro-20-oxa-cyclopropa[5,6]cyclopenta[a]phenanthren-4-yl}ester) was synthesized starting from Withaferin A (4β,5β,6β,22R)-4,27-Dihydroxy-5,6:22,26-diepoxyergosta-2,24-diene-1,26-dione) according to a synthetic strategy recently developed in our laboratory (Supplementary Material Section S1 and Supplementary Fig. S1A) with modifications in earlier reported protocols (19, 20). Briefly, to a stirred solution of Withaferin A (0.470 g, 1.0 mmol) and triethylamine (0.278 mL, 2.0 mmol) in CH₂Cl₂ (10.0 mL) at 0°C under nitrogen atmosphere was added pyridine-2-carbonyl chloride hydrochloride (0.195 g, 1.10 mmol), and the resulting reaction mixture was stirred overnight at room temperature. After completion of the reaction as indicated by thin layer chromatography (TLC), the reaction mixture was quenched with saturated NaHCO₃ solution (5 mL). The organic layer was separated, followed with extraction of aqueous layer with CH₂Cl₂ (2 × 10 mL). The combined organic layers were washed with brine (20 mL), dried over anhydrous Na₂SO₄, and concentrated under reduced pressure to afford the crude ASR490 which was purified by silica gel column chromatography with eluent ethyl acetate: hexane (2:8) to afford the pure ASR490 (0.471 g, 82%) as a white solid. The compound was characterized by NMR and MS, and its purity (≥98%) was determined by HPLC. ¹H NMR (600 MHz, CDCl₃): δ 8.78 (d, J = 4.2 Hz, 1H), 8.15 (d, J = 7.8 Hz, 1H), 7.86 (td, J = 7.8, 1.2 Hz, 1H), 7.52−7.48 (m, 1H), 6.94 (dd, J = 6.0, 4.2 Hz, 1H), 6.22 (d, J = 10.2 Hz, 1H), 5.30−5.20 (m, 2H), 4.50−4.46 (m, 1H), 4.45−4.35 (m, 2H), 4.14 (q, J = 7.2 Hz, 1H), 3.78 (d, J = 6.0 Hz, 1H), 3.25 (1H, brs), 2.60−2.50 (m, 1H), 2.20−2.15 (m, 1H), 2.05 (2s, 6H), 2.00−1.98 (m, 1H), 1.90−1.85 (m, 1H), 1.70−1.65 (m, OH, 3H), 1.55−1.48 (m, 2H), 1.40 (s, 3H), 1.30−1.25 (m, 3H), 1.05−1.00 (m, 3H), 0.72 (s, 3H). ESI-MS m/z 576 (M+H)⁺.

HCT116, SW-620, TCCSUP, UMUC3, HT1376, 5637, T24, and RT4 cells were purchased from the ATCC. HCT116, T24, and RT4 were maintained in McCoy's medium, TCCSUP, UMUC3, and HT1376 in EMEM, SW620 in DMEM, and 5637 in RPMI medium, respectively, and supplemented with 10% FBS and penicillin (100 units/mL) and streptomycin (100 units/mL; Millipore Sigma) in the presence of 5% CO₂ at 37°C. pCMV6-NOTCH1, vector pCMV6-Entry [NOTCH1 (NM_017617) Human ORF Clone; Origene], and NOTCH1 Human siRNA Oligo Duplex were obtained from Origene Technologies Inc. Lipofectamine 2000 reagent was used following the manufacturer's (Cat# 11668019; ThermoFisher Scientific) instruction, transfection with overexpression vectors was performed with 500 ng plasmid concentration, whereas the siRNAs were used in 25 nmol/L concentration. Cells were allowed to be transfected for 48 hours and later harvested or treated for further analysis. Neomycin (1 μg/mL) selection media were used to cultivate Notch1-overexpressing HCT116 clones (C1, C2, C3, C4, and C5).

The growth-inhibitory effect of ASR490 (reconstituted in 10 mmol/L DMSO) was determined by the MTT (3-[4, 5-Dimethylthiazol-2-yl]-2, 5-diphenyltetrazolium bromide) assay. Six biological replicates were used for MTT assays, and it was repeated twice for each experiment. Colon and bladder cancer cell lines were treated with varying concentrations of ASR490 (0–1.6 μmol/L). The anchorage-independent growth assay was performed and repeated in triplicate as described previously (21).

The protein–ligand binding was first studied by cellular thermal shift assay (CETSA) by following previously described protocol (22). Briefly, we treated the cells (3 × 10⁶) with ASR490, incubated at different temperatures (38°C–60°C) to denature and precipitate proteins, performed cell lysis, and centrifuged at 13,000 × g for 10 minutes to collect the soluble fraction. Equal amount of cell lysate was used for ELISA with NRR antibody (Cat: NBP2–62557; Novus Biologicals). GloMelt Thermal Shift Protein Stability assay was performed as per the kit instructions (GloMelt Thermal Shift Protein Stability Kit; Biotium). Briefly, a qPCR reaction was set up with the purified NRR protein (Origene technologies: TP606288), the GloMelt fluorescent dye, and ASR490 (10 μg per reaction). A protein melt run profile was generated, and Tm (melting temperature) was calculated using DNA melt curve software. To analyze protein melting, the Tm was considered at the lowest −dF/dT value (at the lowest point on the curve).

For molecular docking studies, the structure of NRR domain was downloaded with resolution 2 Å from the RCSB database (PDB ID: 3ETO; refs. 23). All bound crystal water molecules and ligands were removed prior to building missing residues through homology modeling using Modeller 9v15. Simultaneously, we built and optimized the structure of ASR490 using Marvin sketch workspace (arXiv.org). We relaxed the NRR structure using Chiron (24) and Gaia (25) for subsequent docking studies with ASR490 compound. To evaluate the extent of interaction between ASR490 and the NRR domain of Notch1 receptor, we performed molecular docking using MedusaDock (26). H-bond interactions between Notch1-NRR domain and ASR490 compound are represented as blue dotted lines. Notch1-NRR domain is shown as carton with α-helices in cyan, β-sheets in magenta, and loops in deep salmon color. ASR490 is shown in green licorice representation, and water molecules mediating the interaction between NRR and ASR490 are shown in red spheres.

Note that 0.3 × 10⁶ cells were seeded in a 6-well plate and were cultured until 70% to 80% confluence was achieved. The cells were then treated with ASR490 for 24 hours. To quantify apoptosis, flow cytometry analysis of the Annexin V–FITC against propidium iodide (PI) assay was performed following a previously described protocol (27). The apoptosis detection kit was purchased from BD Pharminogen. All experiments were repeated in triplicate to achieve statistical relevance.

The invasive capability of pCMV/HCT116 and Notch1/HCT116 was evaluated in Boyden chambers, as described in earlier studies (28). HCT 116, SW-620, pCMV/HCT116, and Notch1/HCT116 cells were analyzed for migration capability with protocols already described in an earlier study (28). All experiments were performed in triplicate to achieve statistical relevance.

Mammalian protein extraction reagent (Cat:78503; Thermo Scientific) was used to extract total protein from pCMV/HCT116, C4, and C5 cells as well as bladder cancer cells were prepared according to the manufacturer's instructions. Western blotting was performed using specific antibodies against Notch1 (Cat:3447S), BCL-2 (Cat:2872S), E-cadherin (Cat:3195S), N-cadherin (Cat:13116S), β-catenin (Cat:8480S), NF-κB (p65), Bax (Cat:5023S), cleaved PARP (Cat:9541L; Cell Signaling Technology), β-actin (Santa Cruz Biotechnologies), and HES1 (Genescript; Lot QC1851). Actin presented in the images represents the loading control for one or more markers from same cell lysates. Chemiluminescence was used to detect the positive bands on the membrane.

Six- to 8-week-old BALB/c athymic nude mice (nu/nu) (Jackson Laboratory) were subcutaneously injected with pCMV/HCT116 and C4 (1 × 10⁶ cells). The monitoring and measurements were performed as described previously (21).

The tumor samples from the pCMV/HCT116 and Notch1/HCT116 xenografts were subjected to IHC analysis as we previously described (28). Primary antibodies against Ki67, Notch1, HES1, and p65 were used in this study.

The experimental data are presented as the mean ± standard deviation (SD or SEM). Unpaired Student t test was used to determine the significance of the differences between different test groups. The significant differences were established at P < 0.05. Prism 6 software purchased and licensed from GraphPad Software Inc. was used to perform the statistical analyses

To examine the therapeutic potential of ASR490 (Fig. 1A) in colorectal cancer, we assessed the cell viability of ASR490-treated HCT116 and SW620 using the MTT assay. Cell viability was significantly reduced with 24-hour (HCT 116, IC₅₀: 750 nmol/L; P = 0.007 and SW-620, IC₅₀: 1.2 μmol/L; P = 0.0008) and 48-hour (HCT 116, IC₅₀: 600 nmol/L; P = 0.005 and SW-620, IC₅₀: 850 nmol/L; P = 0.007) treatment (Fig. 1B and C). To determine the molecular mechanism by which ASR490 inhibits the growth of colorectal cancer cells (HCT116 and SW620), we treated the cell lines with ASR490 and performed immunoblot analysis. Significant downregulation in the expression of NICD and its downstream effector HES1 protein were observed in HCT116 and SW620 cells (Fig. 1D; Supplementary Fig. S1B). Interestingly, an apparent decline in Notch1 and HES1 mRNA expression was observed (Supplementary Fig. S1C), whereas no change in Notch2 and Notch3 expressions was seen in ASR490-treated HCT116 cells (Supplementary Fig. S1D). To confirm that ASR490 inhibits the colorectal cancer cell growth through Notch1, we silenced Notch1 expression by siRNA in HCT116 cells (Supplementary Fig. S2A), then treated with vehicle or ASR490. As seen in Supplementary Fig. S2B, ASR490 failed to inhibit the growth of Notch1-silenced HCT116 cells as compared with scrambled transfected HCT116 cells. Bladder cancer cells have low basal level of Notch1, and we treated these cells with ASR490, which failed to inhibit their growth (Supplementary Fig. S2C and S2D). These two experiments suggest that Notch1 may be a target for ASR490.

Next, to analyze whether ASR490 binds directly to Notch1, we performed molecular docking studies. The CASTp predictions confirmed binding sites of ASR490 in NRR of Notch1 (Fig. 1E). The catalytic pocket in NRR is lined by the residues: Lys-1462, Cys-1464, Asp-1479, Cys-1480, Leu-1482, Asn-1483, Ala-1708, Gly-1711, Leu-1713, Asn-1714, Ile-1715, Tyr-1717, Lys-1718, Ile-1719, and Glu-1720. The estimated binding energy between the NRR domain and ASR490 was -52.55 kcal/mol which signifies strong interaction between ASR490 and NRR domain. The residue-wise interaction analysis estimated three hydrogen-bond interactions between ASR490 and NRR residues Asn-1483, Glu-1673, and Gly-1664 mediated by water molecules (Fig. 1E). To further confirm the binding at protein level, we performed protein thermal shift differential scanning fluorimetry assay with purified NRR protein (Fig. 1F). The results suggest an increased stabilization of NRR protein in presence of ASR490 than with vehicle (DMSO). Further, traditional CETSA was performed, and NRR-specific antibody was used in ELISA as detection method. The absorbance profile (495 nm for TMB substrate) of ASR490-treated HCT116 cells confirmed that ASR490 binds directly to NRR (Fig. 1G). To further confirm that the inhibition of Notch1 activation alters the expression of key genes that regulate cancer cell survival signaling in colorectal cancer cells, we analyzed the effect of ASR490 treatment on prosurvival genes. As shown in Fig. 1H, ASR490 treatment significantly inhibited p65 and BCL-2 expression in colorectal cancer cells.

To examine whether inhibition of Notch1 signaling facilitates induction of proapoptotic signaling, we performed apoptotic assays in ASR490-treated colorectal cancer cells. Induction of apoptosis in ASR490-treated HCT116 (19.9%, P = 0.01; 24 hours) as well as SW-620 (9.57%, P = 0.011; 24 hours) cells in FACS analysis showed significant apoptotic cell death (Fig. 2A and B). A time-dependent upregulation of the proapoptotic markers Bax and cleaved PARP expression was observed in ASR490-treated HCT116 and SW620 cells (Fig. 2C and D). More importantly, inhibition of the migratory (25.18%, P = 0.05; HCT116 and 32.36%, P = 0.032; SW620) capability of colorectal cancer cells was observed in response to ASR490 treatment for 24 hours (Fig. 2E and F). In addition, the time-dependent increase in the E-cadherin (an epithelial marker) and a significant decrease in mesenchymal markers N-cadherin and β-catenin expression were observed in colorectal cancer cells treated with ASR490 for both 12 and 24 hours (Fig. 2G and H).

To assess the proliferative attribute of Notch1 in colorectal cancer, we first generated Notch1 expressing stable HCT116 cell lines, i.e., clones C1, C2, C3, C4, and C5 (Fig. 3A). C4 and C5 were used for further studies as they expressed higher Notch1 compared with other clones. pCMV/HCT116 and Notch1/HCT116 clones C4 and C5 were assessed for cellular growth. Notch1 transfectants showed a significantly higher growth compared with the control (pCMV/HCT116) cells (C4: 42%, P = 0.0226; and C5: 25.8%, P = 0.0236; Fig. 3B). ASR490 treatment significantly inhibited cell growth in both C4 (IC₅₀: 800 nmol/L; P = 0.0016) and C5 (IC₅₀: 1.1 μmol/L; P = 0.0028) transfectants, showing the ability of ASR490 to override Notch1-mediated overgrowth of colorectal cancer cells (Fig. 3B). To understand the effect of Notch1 in increasing the tumorigenic capability of colorectal cancer cells, we performed a colony-forming assay. The colony-forming ability in Notch1 transfectants increased significantly (C4 −42%; P = 0.0238 and C5 −32.4%; P = 0.0238) compared with vector-transfected HCT116 cells. However, ASR490 treatment significantly reduced the colony-forming ability of pCMV/HCT116 (22.3%; P = 0.0169), C4 (38%; P = 0.0406), and C5 (26.66%; P = 0.0127) cells (Fig. 3C). Immunoblot and densitometry analysis of ASR490-treated transformants (C4 and C5) demonstrated inhibition of Notch1 and HES1 expression (Fig. 3D and E). In addition, p65 and BCL-2 expression (survival markers) was downregulated (Fig. 4A and B). Next, we observed increase in expression of proapoptotic genes such as cleaved-PARP and Bax (Fig. 4C) along with induction of apoptosis in ASR490-treated pCMV/HCT116 (25.8%, P = 0.0165) as well as C4 (13.4%, P = 0.0102) and C5 (13.2%, P = 0.0112) cells during FACS analysis (Fig. 4D).

Next, we determined whether Notch1 overexpression influences EMT signaling in Notch1/HCT116 cells. The invasive capability in C4 and C5 cells increased by 65.5% (P = 0.0316) and 63.5% (P = 0.0253), respectively (Fig. 5A). Similarly, we observed a 29% (P = 0.036) and 28.68% (P = 0.0474) increase in the migratory capability of C4 and C5 colorectal cancer cells, respectively (Fig. 5B) compared with pCMV/HCT116 cells. To analyze whether ASR490 treatment can inhibit the enhanced migratory and invasive capability of C4 and C5 cells, both transfectants were treated with the respective IC₅₀ doses of ASR490 for 24 hours. A significant decline in the migratory potential of pCMV/HCT116 (25.11%; P = 0.0073) and Notch1 transfectants (C4 −49.3%; P = 0.0031 and C5 −44%; P = 0.0130) and the invasive capacity of pCMV/HCT116 (40%; P = 0.0047) and both C4 (60.75%; P = 0.0305) and C5 cells (65.5%; P = 0.0301) was observed (Fig. 5A and B). Next we analyzed ASR490-treated pCMV/HCT116, C4, and C5 cells for expression of genes that regulate EMT. EMT markers such as N-cadherin and MMP-9 were significantly downregulated, whereas the epithelial marker E-cadherin expression upregulated, which are hallmarks of EMT (Fig. 5C and D). Notch1 plays an active role in the EMT process, and the results collectively indicate that ASR490 can overcome Notch1-induced EMT signaling in colorectal cancer cells.

Earlier we reported that colorectal cancer xenografts with overexpression of AKT (Notch1 and AKT signaling are interlinked) are significantly aggressive compared with control-transfected colorectal cancer xenografts (21). To determine antitumor potential of ASR490, pCMV/HCT116 and Notch1/HCT116 (clone 4; C4) cells were subcutaneously injected into nu/nu mice. The MTD was checked for ASr490 in nu/nu mice, and ASR490 was found to be safe till 500 mg/kg dose. Nocth1/HCT116 tumors showed rapid and aggressive growth compared with pCMV/HCT116 tumors (Fig. 6A). On the other hand, significant tumor growth inhibition was noted in both the ASR490-treated (5 mg/kg of mouse body weight for 4 weeks) pCMV/HCT116 and Notch1/HCT116 (C4) xenografts.

We then assessed the survival (p65, Notch1, HES1) and proliferation (Ki67) in both pCMV/HCT116 and Notch1/HCT116 (C4) tumors. Notch1 and HES1 expression was higher in Notch1/HCT116 tumors compared with pCMV/HCT116 tumors (Fig. 6B). ASR490 treatment resulted in significant reduction of Ki67 expression (cellular proliferation) in pCMV/HCT116, as well as Notch1-overexpressing Notch1/HCT116 (C4) tumors (Fig. 6B). In addition, ASR490 treatment significantly reduced the expression of the prosurvival marker and p65 in all tumors (Fig. 6B). Next, to assess the effect of ASR490 treatment on Notch1 signaling in xenografts, we analyzed Notch1 and HES1 protein expression in pCMV/HCT116 and Notch1/HCT116 (C4) tissues. Consistent with the IHC results, an inhibition in Notch1 and HES1 protein levels were observed (Fig. 6C).

In the current study, we have demonstrated that aberrant Notch1 overexpression causes colorectal cancer cells to grow rapidly and demonstrates aggressive migratory behavior, and our newly identified small molecule, ASR490, overrides aberrant overexpression of Notch1 in in vitro and in vivo colorectal cancer models to achieve antiproliferative and antitumorigenic effects.

Modern treatment concepts in colorectal cancer are multimodal and use interdisciplinary approaches, including the use of adjuvant, neoadjuvant chemotherapy, radiotherapy, and immunotherapy, which are followed based on the colorectal cancer stage and localization (29). However, it is well understood that the need for optimization of adjuvant therapies (30) and increasing instances of resistance in neoadjuvant therapies (31) warrant identification of new therapeutic targets and support the search for new compounds with low toxicity profiles and better bioavailability.

The recent progress in determination of the crystal structure of the NRR has improved the understanding of mechanisms that are responsible the self-inhibitory effects of HD domain on the processing and activation of NOTCH receptors (32). Recently, mAbs have been used to target NRR region in order to stabilize the region and prevent ligand-independent activation and wild-type Notch1 activation and thus decrease in NICD expression (16, 18). In our results, molecular docking studies suggest ASR490 binds to NRR region of Notch1. Further, the CETSA and GloMelt protein thermal shift assay were performed with NRR-specific antibody of Notch1 on the purified NRR protein, and those results further confirmed that ASR binds to NRR of Notch1 and downregulated Notch1 expression. Similarly, Notch3 antibodies against NRR domain have shown to inhibit expressions of NICD and HES1, whereas the anti-LBD antibody failed to achieve that (18). It is possible that ASR490 may elicit the similar response as anti-NRR antibodies, although the exact mechanism needs to be elucidated in detail. In addition to downregulation of NICD expression, we have also seen inhibition of Notch1 gene expression. However, the exact mechanism by which ASR490 inhibits Notch1 gene expression is yet to be elucidated.

Notch1 activation is associated with early development of cancer (33), and activation of its downstream events such as overexpression of HES1 has been linked with colorectal cancer progression (34) and metastasis (35). Silencing Notch1 activity through lentiviral-encoding Notch-1-siRNA and Notch1 inhibitors such as DAPT (N-[N-(3,5-difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl ester) has demonstrated capability to induce apoptosis in colorectal cancer cells (36), proving that Notch1 can be an effective target for colorectal cancer management. However, the current landscape of inhibitors, particularly gamma secretase inhibitors (GSI) such as LY-411,575 or DAPT (37), can have unintended biological implications because of broad substrate profile of gamma secretase (38, 39). Natural compounds such as Butein (40) and more recently compounds isolated from Nerium indicum (41) have been reported as inhibitors of Notch1. Keeping in mind the low toxicity profiles of compounds derived from natural sources and high bioavailability (42), results from our study showing the detrimental effect on Notch1 signaling by ASR490 derived from a natural compound are encouraging.

Notch1-mediated survival has been shown to be a primary driver of cell proliferation and tumor recurrence in vivo (43). Moreover, its aberrant activation has been found to be responsible for uncontrolled cellular growth in several cancer types (44). Inhibition of its expression and downstream signaling has resulted in induction of apoptosis and thus growth arrest in HT29 cells (36). A tripeptide of GSIs category inhibited the proliferation of MDA-MB231 cells (45), whereas natural compounds such as genistein (46) induce apoptosis in cancer cells by downregulating survival signaling, particularly NF-κB expression. Similar alteration in survival as well as apoptotic signaling was seen in ASR490-treated colorectal cancer cells in our study. The Notch1-overexpressing transfectants mimicking aberrant overexpression conditions also showed downregulation of proapoptotic and prosurvival markers, indicating that uncontrolled growth of colorectal cancer cells in the case of Notch1 activation can be managed by ASR490 treatment.

Notch1 signaling is also recognized as a major regulator of EMT in several cancer types (47–49) including colon cancer (50). Activation of Notch1 signaling accelerates EMT by positively regulating Snail, a slug family protein, and repressing E-cadherin function. This in turn affects the progression of tumors in cancer cells (51). In addition, elevated HES1 expression has been correlated with several neoplastic conditions (15–18). Its interaction with multiple signaling pathways has been attributed to its contribution toward promotion of cell metastasis by evading tumor cell differentiation. Alleviation of Notch1-induced EMT may well be a direct result of the inhibition of Notch1/HES1/NFκB-p65 signaling. The ability of ASR490 to overcome Notch1 signaling and inhibit tumorigenic capacity was shown in our preclinical models of colorectal cancer. Our studies suggest that ASR490 is safe up to a dose of 500 mg/kg, which is 100 times more than the dose used in our efficacy studies indicating a high therapeutic index.

In summary, our results suggest that ASR490, a potent small molecule, overcomes Notch1-mediated prosurvival signaling and EMT, which resulted in growth inhibition in preclinical models of colorectal cancer. Additional studies may require optimizing the therapeutic efficiency of ASR490 that might lead its translation to clinical settings.

S.R. Ramisetti reports grants from NIH during the conduct of the study. A.K. Sharma reports grants from NIH (R01 CA185972) during the conduct of the study and has a patent for 62/976880 pending (provisional patent filed). C. Damodaran reports grants from NIH (R01 grant [NIH/NCI-1R01CA185972]) during the conduct of the study and has a patent for 62/976880 pending (provisional patent filed). No potential conflicts of interest were disclosed by the other authors.

A. Tyagi: Data curation, software, formal analysis, investigation, methodology, writing-original draft, writing-review and editing. B. Chandrasekaran: Methodology, writing-review and editing. V. Kolluru: Data curation, software, methodology, writing-review and editing. B.V. Baby: Methodology, in vivo experiments. C.A. Sripathi: Methodology. M.K. Ankem: Resources, supervision, validation. S.R. Ramisetti: Resources, formal analysis, writing-review and editing. V.R. Chirasani: Software, formal analysis. N.V. Dokholyan: Software, formal analysis. A.K. Sharma: Resources, formal analysis, methodology, writing-review and editing. C. Damodaran: Conceptualization, resources, data curation, supervision, funding acquisition, writing-review and editing.

We acknowledge support from the NIH/NCI-1R01CA185972 to C. Damodaran, 1R35 GM134864 to N.V. Dokholyan, and the Passan Foundation. The current project was also supported by the National Center for Advancing Translational Sciences, NIH, through grant UL1 TR002014. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

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.