aNDR1 was screened and identified as an agonist of NDR1. The X-ray crystal structure of NDR1 (code: 6BXI) was downloaded from the RCSB PDB database (http://www.rcsb.org). FTMAP (http://ftmap.bu.edu/login.php) was used to identify potential binding sites for ligands by utilizing small-molecule probes. The contact frequency per residue with probes was generated to quantify the contribution of each residue to ligand binding. The results of FTMAP were visualized using PyMOL to generate an isopotential diagram depicting the molecular pocket (Seeliger and de Groot, 2010). The synthetic protocol of aNDR1 is shown in Supplementary Figure S1A, and its structure was verified through LC‒MS and ¹H NMR. The purity of aNDR1 was greater than 95%, and it was dissolved in DMSO and stored at −20 °C.

Human embryonic kidney cells (HEK293T), WPMY-1 cells, and human prostate cancer cell lines (PC3 and DU145) were purchased from the cell bank of the Chinese Academy of Sciences (Shanghai, China). The cells were cultured in RPMI-1640 or DMEM containing 10% fetal bovine serum (FBS) and 1% penicillin‒streptomycin in a humidified 5% CO₂ atmosphere at 37 °C.

The PC3, DU145 and WPMY-1 cells were seeded in 96-well plates at 5 × 10³ cells per well. Following 24 h of incubation, the cells were treated with different concentrations of aNDR1 for 24 h, followed by the addition of 10 µL CCK-8 to each well for 4 h. The absorbance was measured at 450 nm using a microplate reader. Cell viability (%) = OD (treated)/OD (control) × 100%.

aNDR1 was added to the cell culture medium at a concentration of 5 μM. The samples were obtained at 0, 24, 48, and 72 h and analyzed by UPLC-HRMS (Thermo Fisher Scientific, United States) to measure the stability of aNDR1. For protein binding activity, the medium with 10% FBS and 5 μM aNDR1 was centrifuged (230,000 g) for 5 h at 37 °C. The supernatant was analyzed by UPLC‒HRMS (Thermo Fisher Scientific, United States) to detect unbound aNDR1. The permeability experiment model of Caco-2 cells was established according to a previous study (Xiao et al., 2022). aNDR1 was added to the apical compartment at a final concentration of 5 μM at 37 °C. Samples were taken from the compartments 2 h before and after incubation and then analyzed by UPLC‒HRMS (Thermo Fisher Scientific, United States).

In short, pMD2.0G and psPAX were cotransfected with sh-NDR1 or control vectors into HEK293T cells using Lipofectamine 3,000 (Invitrogen, United States) according to the manufacturer’s protocol. The supernatant containing lentivirus was collected after 48 h and subjected to infection of DU145 cells to construct NDR1-knockdown cells. Similarly, PCDNA3.1 and NDR1 plasmids were transfected by using Lipofectamine 3,000 (Invitrogen, United States) to overexpress NDR1 in HEK293T cells.

E. coli BL21 was transformed with the pGEX-GST-NDR1 plasmid. Mid-logarithmic phase cells were induced with 1 mM isopropyl-β-D- thiogalactopyranoside (IPTG) overnight at 16 °C. The bacteria were collected and lysed in lysis buffer (20 mM Tris-Cl, pH 8.0, 100 mM NaCl, 0.2 mM EDTA, and 0.5% NP-40) together with lysozyme and phenylmethylsulfonyl fluoride (PMSF) via an ultrasonic cell crusher. Then, the GST-fused NDR1 proteins in the supernatant were purified on glutathione-agarose (GE Healthcare, United States) and detected by SDS‒PAGE-coupled Coomassie blue staining. According to the manufacturer’s protocol, purified NDR1 proteins were incubated with substrate peptide (KKRNRRLSVA), ATP, reaction buffer and different concentrations aNDR1 and the activity of NDR1 kinase was detected by a Kinase-Lumi™ luminescent kinase assay kit (Beyotime, China).

The proliferation of PCa cells was measured using the EdU Cell Proliferation Assay Kit (Beyotime, China). DU145 and PC3 cells were treated with aNDR1 or DMSO, stained with 5-ethynyl-20-deoxyuridine (EdU) according to the manufacturer’s instructions and photographed under a fluorescence microscope.

PC3 and DU145 cells were seeded in 6-well plates. At 95% confluence, a pipette was used to draw a straight line per plate, and PBS was used to remove cell debris. Photos were taken to record the beginning of wound gaps, and then medium containing aNDR1 or DMSO was added. The gap growth of each group was observed and photographed after 24 h. The wound healing rate was calculated by the scratch length change = (the average wound length at 0 h - the average wound length at 24 h)/the average wound length at 0 h.

PC3 and DU145 cells were seeded in 6-well plates to a certain number, and then aNDR1 was added to the medium for 24 h. Next, all adherent and suspended cells were collected, and an Annexin V-FITC apoptosis detection kit (Beyotime Biotechnology, China) was used according to the manufacturer’s protocol. Apoptosis of cultured cells was detected via flow cytometry.

DU145 cells were transfected with MYC-NDR1 and treated with or without aNDR1 for 24 h. Cells were collected, and equal volumes of cell suspensions were heated at different temperatures for 5 min. Then, the cells were lysed by three freezes‒thaw cycles in liquid nitrogen, and the supernatant was collected after centrifuging the lysates at 15,000 × g for 15 min at 4 °C. The samples were detected via Western blotting analysis.

After transfecting DU145 cells with MYC-NDR1 for 24 h, cells were lysed, and the supernatant was collected after centrifuging lysates at 15,000 × g for 15 min at 4 °C. Each sample was incubated with aNDR1 for 1 h at room temperature. Then, different concentrations of pronase were added to lysates and digested for 5 min. Lysates were added to loading buffer and then subjected to boiling and Western blotting analysis.

Cells were lysed with RIPA lysis buffer (Solarbio, China) containing protease and phosphatase inhibitors (MedChemExpress, United States). An equal amount of protein in each sample was isolated by SDS‒PAGE electrophoresis, transferred to a PVDF membrane (Millipore, United States), and then blocked with 5% fat-free milk or 3% bovine serum albumin (BSA) in Tris-buffered saline with Tween 20 (TBST) at room temperature for 1 h. The membrane was then incubated with the primary antibody, and the general Western blotting assay procedure was followed. For immunoprecipitation, the starting steps were as described above but involved the use of IP lysate (Thermo Fisher Scientific, United States). A portion of the protein solution was taken to directly detect the corresponding protein expression, while the rest of the solution was added to the primary antibody and magnetic beads (Bimake, United States), respectively, according to the manufacturer’s instructions. The beads were washed with wash buffer (Thermo Fisher Scientific, United States) and heated to 100 °C for 5 min using protein loading buffer (ABclonal, China). Immunoblotting was then performed as described above.

The NDR1 (sc-365555) antibody was purchased from Santa Cruz Technology. The phospho-NDR1/2 (Thr444/Thr442) antibody was purchased from Affinity Biosciences. All the following antibodies were purchased from Cell Signaling Technology: E-cadherin (#14472), PARP (#9532), cleaved-PARP (#5625), phospho-Thr (#9386), β-actin mouse monoclonal antibody (#3700), and secondary antibody against mouse and rabbit. Bcl-2 (A19693), Bax (A19684), N-cadherin (A19083), Vimentin (A19607), Survivin (A1551), YAP (A1002), and phospho-YAP-Ser127 (AP0489) were purchased from ABclonal. All antibodies were used following the manufacturer’s instructions.

All wild-type male BALB/c nude mice were purchased from the Experimental Animal Center of Xiamen University (Xiamen, China). DU145 cells (1 × 10⁶) were injected subcutaneously into the backs of nude mice (6–8 weeks). When the tumor volume grew to 50–100 mm³, the mice were randomly divided into two groups: the control group and the aNDR1 treatment group. These two groups were treated with the same concentrations of placebo or aNDR1 (5 mg/kg) every other day. The body weight of the mice and the length and width of the tumor were measured every other day, and the tumor volume was calculated using Equation V= (length x width²)/2, where length was always the longest size of the tumor. At the end of the experiment, the mice were sacrificed, and the solid tumors were resected, weighed and sacrificed. The tumor tissues and important organs were analyzed by H&E staining. For the metastasis model, DU145 cells (1 × 10⁶) were intravenously injected into the tail veins of nude mice, and the mice were randomly divided into two groups: the control group and the aNDR1 treatment group. Similarly, these two groups were treated with the same concentrations of placebo or aNDR1 (5 mg/kg) every other day. After 14 days, the mice were sacrificed, and the number of lung metastasis colonies was counted.

GraphPad Prism 8 software was used to perform statistical analysis. All experiments were repeated at least three times except for the animal experiments. The significance of the differences between the two groups was assessed using Student’s t-test. Univariate and repeated measure analysis of variance (ANOVA) was used to compare the significance. ^* p < 0.05; ^** p < 0.01; ^*** p < 0.001; NS, not significant.

Considering that NDR1 belongs to the NDR kinase family, the activation loop is very important for kinase activity. The activation loop usually begins with the Asp-Phe-Gly (DFG) motif and ends with the Ala-Pro-Glu (APE) motif. In the fully active state, the DFG motif adopts the so-called DFG-in state, wherein the Asp of DFG points inside the active site cleft (Xie et al., 2020). Many small-molecule drugs function by activating or inhibiting activation loops, especially the DFG motif. To identify a specific small molecule agonist of NDR1, we initially conducted druggability analysis of the NDR1-full and NDR1-protein kinase domain using FTMAP (Ngan et al., 2012). The analysis revealed that the activation loop region (from Asp230 to Glu292) was the most promising druggable pocket because of the high probe binding frequency in the NDR1-full (Figures 1A–C) and NDR1-protein kinase domain (Figures 1D–F). Subsequently, we focused on ChEMBL, a database of bioactive molecules with drug-like properties (Gaulton et al., 2012), to screen for small-molecule agonists of NDR1, and it provided 31 potential compounds that may activate NDR1 (Figure 1G). Then, we followed the principles of “Rule of 5" (Ro5) (Brown and Wobst, 2021). As a result, we obtained 19 potential small-molecule compounds (Supplementary Table S1).

To further evaluate the binding affinity and the most promising compound, we performed molecular docking (Pinzi and Rastelli, 2019) using computer software. Surprisingly, we found that the Compound C19H14ClFN6O3S (CHEMBL1822054, PubChem CID: 44124863) could form H-bonds with Asp230 Gln130, with a high docking score of −6.098 kcal/mol, which stabilized the DFG-in conformation and facilitated ATP binding, possibly promoting kinase activity (Figure 1H). We named this small-molecule compound aNDR1 (Figure 1I). Additionally, we synthesized aNDR1 (Supplementary Figure S1A) and verified its structure through ¹H NMR (Supplementary Figure S1B) and LC‒MS (Supplementary Figure S1C).

First, we evaluated the effects of aNDR1 in the CRPC cell lines PC3 and DU145. In cell viability assays, after 24 h of treatment with different concentrations of aNDR1, significant growth inhibition was observed in PCa cells. The IC₅₀ values for PC3 and DU145 cells were determined to be 1.178 μM and 1.763 μM, respectively (Figure 2A). Subsequently, we assessed the cytotoxicity of aNDR1 in PCa cells using an LDH release assay. The results showed that aNDR1 had no obvious cytotoxicity (Figure 2B). To verify whether the effect of aNDR1 is tumor specific, we also performed similar experiments on the normal prostate cell line WPMY-1. The results showed that the IC₅₀ of aNDR1 on WPMY-1 was greater than 500 μM (Figure 2C). Additionally, aNDR1 did not display noticeable cytotoxicity in WPMY-1 cells (Figure 2D). These findings suggest that aNDR1 selectively targets CRPC cells and its efficacy is not solely dependent on its own pharmacochemical properties.

NDR1 is known as an autophosphorylated kinase (Cook et al., 2014), which is fundamental for activity regulation. To eliminate the influence of autophosphorylation, we purified the NDR1 protein in E. coli BL21 and confirmed it by Coomassie blue staining (Figure 2E). The kinase activity of NDR1 was assessed using a chemiluminescence kinase activity assay kit. The results indicated that NDR1 activity increased in a dose-dependent manner (Figure 2F). The Thr pan-phosphorylation of NDR1 was gradually elevated in response to the increased concentration of aNDR1, and Ser pan-phosphorylation had no obvious change (Figure 2G). These findings confirm that aNDR1 promotes NDR1 kinase activity.

Then, we investigated the physicochemical properties of aNDR1, including its chemical structure stability and plasma protein binding capacity as well as its cell membrane permeability by UPLC‒HRMS. The results revealed that aNDR1 maintained its structural integrity, with 84.93%, 83.80%, and 80.33% of the compound remaining intact after 1, 2, and 3 days of incubation in cell culture medium, respectively (Supplementary Figure S2A). In addition, 60.15% of aNDR1 was found to bind with proteins in the culture medium (Supplementary Figure S2B). Furthermore, aNDR1 exhibited efficient permeability across Caco-2 cells, with an apparent permeability coefficient (Papp) of 8.31 × 10⁻⁶ cm s^-1 (Supplementary Figure S2C). These data suggest that aNDR1 possesses favorable drug-like physicochemical properties.

We further investigated the effects of aNDR1 on prostate cancer cells in vitro. PC3 and DU145 cells were exposed to varying concentrations of aNDR1 for 24 h, followed by labeling with EdU and Hoechst. As shown in Figure 3A, the proportion of EdU-positive cells decreased with increasing drug concentrations, indicating that aNDR1 could effectively inhibit cell proliferation. A wound healing assay was used to assess cell migration, and the results showed that aNDR1 significantly inhibited the migration of prostate cancer cells in a dose-dependent manner (Figure 3B). Epithelial mesenchymal transformation (EMT) plays a critical role in tumor metastasis. Our previous study found that NDR1 could inhibit the metastasis of prostate cancer cells by inhibiting EMT progression (Yue et al., 2018). Therefore, we detected protein markers of EMT in DU145 cells by Western blotting. The results showed that the epithelial marker E-cadherin was obviously upregulated, and mesenchymal markers such as N-cadherin and vimentin were downregulated (Figure 3D), suggesting the negative regulation of aNDR1 toward NDR1-mediated EMT processing. NDR1 has been reported to promote apoptosis as a member of the Hippo pathway, which also participates in CRPC (Vichalkovski et al., 2008b). Therefore, we explored the effect of aNDR1 on apoptosis. After treatment with aNDR1 for 24 h, the apoptosis of PC3 and DU145 cells was measured by flow cytometry using an Annexin V-FITC apoptosis detection kit. We found that the proportion of apoptotic cells was significantly increased in a dose-dependent manner by aNDR1, and most of them were late apoptotic cells (Figure 3C). To further confirm the cell apoptosis induced by aNDR1, the expression levels of apoptosis-related proteins were measured. The expression of the antiapoptotic proteins Survivin, Bcl-2 and PARP was downregulated, and the proapoptotic proteins Bax and cleaved-PARP were upregulated (Figure 3D). In combination, these results indicated that aNDR1 had good inhibitory effects on tumor-related cellular processes, such as proliferation, migration, EMT and antiapoptosis, based on prostate cancer cell lines in vitro.

To confirm that the antitumor activity induced by aNDR1 specifically targets NDR1, we generated an NDR1-knockdown cell line named DU145-shNDR1, along with a control group DU145-shMOCK. The infection efficiency was assessed by Western blotting (Figure 4A). Then, we evaluated the effects of aNDR1 on DU145-shMOCK and DU145-shNDR1#2 cells, including EdU staining assays, wound healing assays and Annexin V-FITC apoptosis detection assays. As expected, knockdown of NDR1 in prostate cancer cells effectively prevented the inhibitory effects of aNDR1 on cell proliferation and migration, as well as the promotion of apoptosis (Figures 4B–D). These results demonstrate that the effects of aNDR1 on the phenotypes of cancer cells are dependent on the presence of NDR1, indicating the specificity of aNDR1 in targeting NDR1. These results further confirm that the observed anticancer effects of aNDR1 are mediated through its specific interaction with NDR1, highlighting its potential as a targeted therapeutic agent for prostate cancer.

NDR1 is a well-known member of the Hippo pathway, which can directly increase the phosphorylation of YAP and promote its degradation (Hergovich, 2013). Meanwhile, NDR1 was upregulated by MST1 as well as its phosphorylation of Thr-444 (Vichalkovski et al., 2008a). To explore whether aNDR1 directly targets NDR1, we examined the expression levels of NDR1 in DU145 cells treated with aNDR1. The results showed that the protein level of NDR1 was significantly upregulated (Figure 5A), which was related to NDR1 stabilization (Figure 5B). In fact, an enhancement of the NDR1 half-life by aNDR1 was observed upon the inhibition of protein synthesis with cycloheximide (CHX). Additionally, we found that YAP was downregulated and p-YAP was increased in the presence of aNDR1 (Figure 5A), and this effect was abolished by NDR1 small interfering RNA (siRNA) (Figure 5C).

Considering the favorable binding affinity between aNDR1 and NDR1 based on molecular docking prediction, we used the cellular thermal shift assay (CETSA) and drug affinity responsive target stability (DARTS) assay to confirm direct interaction, which have been widely applied in identifying target proteins of small molecule drugs (Park and Marqusee, 2005; Chang et al., 2012). The results showed that aNDR1 increased NDR1 thermal stability in a dose-dependent manner (Figure 5D). Meanwhile, aNDR1 partially protected NDR1 from protease-induced degradation (Figure 5E).

All these results suggested that NDR1 was the target protein of aNDR1, providing a promising lead compound for the treatment of prostate cancer.

To evaluate the antitumor effects of aNDR1 in vivo, a prostate cancer cell xenograft model was established by subcutaneously injecting DU145 cells into athymic nude mice. Tumor growth was monitored in nude mice that were intraperitoneally injected with aNDR1 (5 mg/kg) or DMSO every other day. As shown in Figures 6A–C, the volume and weight of the tumors were observably reduced in the aNDR1-treated group compared to those in the DMSO group. Importantly, no significant difference in body weight loss was observed between aNDR1-and DMSO-treated mice, indicating that aNDR1 had few toxic side effects on mice at our therapeutic concentration (Figure 6D). We also found that NDR1 expression levels in tumor tissues in the aNDR1-treated group were higher than those in the DMSO-treated group (Figure 6E). Consistently, the density of prostate cancer cells in tumor H&E staining was much lower than that in the DMSO group when treated with aNDR1 (Figure 6F). Furthermore, no obvious toxic pathologic damage in the heart, liver, spleen, lungs, or kidneys was observed by H&E staining (Figure 6F).

To investigate the effect of aNDR1 on tumor metastasis, we performed a tail vein injection of DU145 cells to induce lung metastasis. Visual observation (Figures 6G, H) and H&E staining of the lung sections (Figure 6I) revealed that aNDR1 effectively inhibited the metastasis of prostate cancer cells compared to that in the control group.

Overall, these results demonstrate that aNDR1 exhibits potent antitumor effects in vivo, including tumor growth and metastasis inhibition at the therapeutic concentration used, showing the potential of aNDR1 as a promising therapeutic agent for prostate cancer treatment.