Targeting GRP75 with a Chlorpromazine Derivative Inhibits Endometrial Cancer Progression Through GRP75-IP3R-Ca2+-AMPK Axis

Abstract

Tumors often overexpress glucose-regulated proteins, and agents that interfere with the production or activity of these proteins may represent novel cancer treatments. The chlorpromazine derivative JX57 exhibits promising effects against endometrial cancer with minimal extrapyramidal side effects; however, its mechanisms of action are currently unknown. Here, glucose-regulated protein 75 kD (GRP75) is identified as a direct target of JX57 using activity-based protein profiling and loss-of-function experiments. The findings show that GRP75 is necessary for the biological activity of JX57, as JX57 exhibits moderate anticancer properties in GRP75-deficient cancer cells, both in vitro and in vivo. High GRP75 expression is correlated with poor differentiation and poor survival in patients with endometrial cancer, whereas the knockdown of GRP75 can significantly suppress tumor growth. Mechanistically, the direct binding of JX57 to GRP75 impairs the structure of the mitochondria-associated endoplasmic reticulum membrane and disrupts the endoplasmic reticulum-mitochondrial calcium homeostasis, resulting in a mitochondrial energy crisis and AMP-activated protein kinase activation. Taken together, these findings highlight GRP75 as a potential prognostic biomarker and direct therapeutic target in endometrial cancer and suggest that the chlorpromazine derivative JX57 can potentially be a new therapeutic option for endometrial cancer.

Keywords: GRP75; MAM; chlorpromazine; endometrial cancer.

Figure 1

CPZ derivatives JX57 and JX66 exhibit anticancer properties in an endometrial cancer organoid (ECO) model. A) Schematic diagram showing the drug repurposing screening and structure modification strategy used to obtain the preferred compounds JX57 and JX66. B) Histological images showing the organization structure in primary tumor and ECO models. Scale bar: 100 µm. C) IC₅₀ anti‐proliferation regression fitting curves for CPZ, JX57, and JX66 (concentration units = µm) generated using the ECO model; n = 3. Data are represented as mean ± SD.

Figure 2

Combining activity‐based protein profiling (ABPP) and transcriptome results identified GRP75 as the direct target of JX57. A) Schematic diagram of the ABPP technology used for the identification of JX57 and JX66 target proteins. B) Protein—protein interaction analysis of candidate targets. Redder colors indicate more central, yellower colors indicate exterior. C) Ingenuity pathway analysis suggests the enrichment pathways that exhibited the most substantial up (positive)‐ and down (negative)‐regulation in EC cells treated with JX57. D,E) Microscale thermophoresis results showing the binding of the GRP75 protein to JX57 and JX66 (concentration units: mol/L); n = 3. F) Molecular docking patterns of JX57 and JX66 with GRP75. G) Root mean square deviation of the protein backbone in complex with JX57 and JX66. H) Root mean square fluctuation (RMSF) values of the protein backbone throughout the simulations, where the ordinate is the root mean square fluctuation RMSF (nm) and the abscissa is the residue number. I) Radius of the protein complex gyration profile in the presence of the selected ligands after 50 000 ps molecular dynamics simulations. Data show the mean ± SD.

Figure 3

High GRP75 expression contributes to tumorigenesis and correlates with a poor EC prognosis. A) Western blot analysis showing successful knockdown of GRP75 in ISK and KLE cells. B,C) CCK‐8 assays showing the effects of the GRP75 knockdown on EC cell proliferation; n = 6. D) Knockdown of GRP75 in ECO cells. E) Representative images of ECO cells; scale bar = 100 µm. F) CCK‐8 assays showing the effects of the GRP75 knockdown on ECO cell proliferation; n = 6. G) Representative images of GRP75 staining in the EC tumor and paracancerous tissues; scale bar = 50 µm. H) GRP75 expression in 110 tumor tissues compared with levels in 15 adjacent non‐cancerous tissues. I) High GRP75 expression is correlated with poorer EC differentiation. J) Kaplan–Meier survival analysis of 110 patients with EC based on tumor GRP75 expression. K) GRP75 expression was analyzed with Kaplan–Meier Plotter in patients with breast cancer, lung adenocarcinoma, and liver hepatocellular carcinoma from TCGA database. Data show the mean ± SD. B,C) p values were calculated using one‐way ANOVA; F,H) p values were calculated using two‐tailed unpaired t‐tests; ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, and ^**** p < 0.0001.

Figure 4

GRP75 is required for the anti‐EC activity of JX57. A) Western blotting analysis of the GRP75 expression in ISK and KLE cells treated with JX57. B) Quantification of GRP75 in JX57‐treated ISK tumor‐bearing mice determined using ELISA; n = 8. C) ATPase activity of the recombinant GRP75 treated with gradient concentrations of JX57; n = 3. D–G) Cell viability assay; n = 6 (D, E), colony formation assay; n = 3 (F), and flow cytometry analysis; n = 3 (G) to assess the effects of JX57 (10 µmol L⁻¹) and DMSO on proliferation and apoptosis in GRP75‐deficient ISK and KLE cells. H) Cell viability assays showing the effects of JX57 on proliferation in GRP75‐deficient ECO cells; scale bar = 100 µm; n = 3. I) Harvested tumors at the growth assay endpoint show HSPA9 (GRP75) knockdown inhibits the antitumor effects of JX57; n = 6. Data show the mean ± SD. C) P values were calculated using one‐way ANOVA; D–I) pvalues were calculated using two‐way ANOVA; ns, not significant; ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, and ^**** p < 0.0001.

Figure 5

JX57 destabilizes the MAM structure via GRP75. A) Representative confocal images of ISK cells and HSPA9 (GPR75) knockdown cells treated with JX57 (10 µmol L⁻¹) and stained with Mito‐Tracker (red) and ER‐Tracker (green); scale bar = 20 µm. Manders’ co‐localization coefficient (MCC) quantification of the overlap between Mito‐Tracker and ER‐Tracker; n = 4. B) Representative TEM images of the MAM in ISK cells that were treated with DMSO and JX57 (10 µmol L⁻¹). Red lines with arrows indicate the contact distance for the Mito–ER associations; scale bar = 500 nm. Quantification of the mean distance between the Mito and ER associations; n = 3. C) Co‐immunoprecipitation analysis of GRP75 and IP3R in the ISK cells and KLE cells treated with JX57 (10 µmol L⁻¹). D) Representative confocal images of Rhod‐2 (red)‐ and Mito‐Tracker (green)‐loaded ISK cells and HSPA9 knockdown cells treated with JX57 (10 µmol L⁻¹); scale bar = 20 µm. Quantification of the [Ca²⁺]_m levels (MCC of Rhod‐2 signal overlapping with Mito‐Tracker signal); n = 4. E,F) mRNA expression of ATF5, detected by quantitative real‐time polymerase chain reaction; n = 6. G) Correlations between HSPA9 and TCA were cycle scores analyzed using the Spearman correlation. The abscissa represents the distribution of gene expression, and the ordinate represents the distribution of the pathway score. H,I) JX57 (10 µmol L⁻¹) reduces α‐KGDH levels as mediated by GRP75 in the ISK and KLE cells; n = 3. The α‐KGDH level in the shNC group was set at 100%. J,K) JX57 (10 µmol L⁻¹) reduces ATP levels as mediated by GRP75 in ISK and KLE cells; n = 3. The ATP levels in the shNC were set at 100%. Data show the mean ± SD. B,E,F) p values were calculated using two‐tailed unpaired t‐tests; A,D,H–K) p values were calculated using two‐way ANOVA; ns, not significant; ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, and ^**** p < 0.0001.

Figure 6

JX57 acts on GRP75 to activate AMP‐activated protein kinase (AMPK) signaling in EC cells to exert anticancer effects. A,B) Extracellular acidification rate (ECAR) was measured using the Seahorse XFe96 analyzer in ISK and KLE cells; n = 6. C) Expression of p‐AMPK, AMPK, and BIM determined by Western blotting in ISK and KLE cells treated with JX57. D) Western blotting of HSPA9‐deficient ISK and KLE cells treated with JX57. E) ISK and F) KLE cells were treated with JX57 (10 µmol L⁻¹) for 48 h in the presence or absence of compound C (0.5 µmol L⁻¹), and a CCK‐8 assay was used to measure cell viability; n = 6. G) ISK and H) KLE cells were treated with compound C for 48 h (0.5 µmol L⁻¹), and a CCK‐8 assay was used to measure cell viability; n = 6. I) Harvested tumors at the growth assay endpoint show the antitumor effects of JX57 (5 mg kg⁻¹) when compound C (0.2 mg kg⁻¹) was added; n = 6. Data for the glycolysis experiment are shown as mean ± SEM and other data are shown as mean ± SD. E–I) p values were calculated using one‐way ANOVA;^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, and ^**** p < 0.0001.