Cloxiquine, a traditional antituberculosis agent, suppresses the growth and metastasis of melanoma cells through activation of PPARγ

Abstract

Melanoma is one of the most aggressive skin cancers and 5-year survival rate is only 4.6% for metastatic melanoma patients. Current therapies, especially those involving clinical chemotherapy drugs, have achieved remarkable advances. However, their side effects, such as bone marrow suppression, limit the effectiveness of available pharmacological therapies. Therefore, exploring new antimelanoma drugs with less toxicity is critical for the treatment of melanoma. In the present study, we aimed to identify the antimelanoma drugs with ability to repress the proliferation of melanoma cells by using a high-content screening of FDA-approved drug libraries. We found that cloxiquine (CLQ), a traditional antituberculosic drug, exhibited strong inhibitory effects on the growth and metastasis of melanoma cells both in vivo and in vitro. In contrast, CLQ at the tested doses did not show any apparent toxicity in normal melanocytes and in the liver. At the metabolic level, treatment with CLQ decreased glycolysis, thus potentially inhibiting the "Warburg effect" in B16F10 cells. More importantly, combination of CLQ and 2-deoxyglucose (2-DG), a well-known glycolysis inhibitor, did not show a synergistic effect on the tumor growth and metastasis, indicating that inhibition of glycolysis is potentially involved in mediating CLQ's antimelanoma function. Bioinformatics analyses revealed that peroxisome proliferator-activated receptor-gamma (PPARγ) served as a potential CLQ target. Mechanistically, CLQ stimulated the transcription and nuclear contents of PPARγ. Furthermore, the specific PPARγ inhibitor GW9662 or PPARγ shRNA partially abolished the effects of CLQ. Collectively, our findings demonstrate that CLQ has a great potential in the treatment of melanoma through activation of PPARγ.

Fig. 1. Compound screenings in B16F10 cells identify CLQ as a potential antimelanoma drug.

a The schematic workflow of the compound screen. b Cell viability of melanoma (B16F10 and A375) cells treated with indicated functional drugs. c Chemical structure of CLQ. d, e Mouse B16F10 and Melan-A melanocyte cells, as well as the human A375 and PIG1 cells were treated with the indicated concentrations of CLQ for 24 h. The inhibitory effect of CLQ on cell viability was detected by a CCK-8 assay. f Morphological analysis. All values are presented as the mean ± SD from at least three separate experiments. **P < 0.01 vs. CTL

Fig. 2. CLQ inhibits the proliferation of melanoma cells and suppresses tumor growth in vivo.

B16F10 and A375 cells were treated with the indicated concentrations of CLQ for 24 h. a EdU incorporation assay used to measure cell proliferation. b Protein expression levels of key regulators involved in cell cycle progression. c Quantification for the immunoblots from b. All values are presented as the mean ± SD from at least three separate experiments. *P < 0.05 and **P < 0.01 vs. CTL. B16F10 cells (1 × 10⁶ cells per mouse) were injected subcutaneously into the right flank of nude mice. When tumors became palpable on day 7, mice were randomly divided into three groups and were administered either vehicle (olive, i.p., daily) or CLQ (5 mg/kg, 25 mg/kg, i.p., daily) for 8 days. n = 5 for each group. d Subcutaneous tumor volumes were measured at the indicated days. e Representative tumor images. f Tumor weights. g H&E staining and immunohistochemistry analysis of Ki-67 from tumor sections. All values are presented as the mean ± SD. *P < 0.05 and **P < 0.01 vs. CTL

Fig. 3. CLQ inhibits the migration of melanoma cells and suppresses tumor metastasis in vivo.

B16F10 and A375 cells were treated with the indicated concentrations of CLQ for 24 h. a Determination of VSMC migration by transwell chamber (top) and wound-healing (bottom) assays. b Protein expression levels of ICAM-1, VCAM-1 and MMPs. c Quantification for the immunoblots from b. All values are presented as the mean ± SD from at least three separate experiments. *P < 0.05 and **P < 0.01 vs. CTL. B16F10 cells (1.5 × 10⁵ cells per mouse) were injected intravenously. Seven days later, mice were randomly divided into three groups and were administered either vehicle (olive, i.p., daily) or CLQ (5 mg/kg, 25 mg/kg, i.p., daily) for 14 days; n = 5 for each group. d Macroscopic images, e Statistical analyses and f H&E staining of lung metastatic nodules. All values are presented as the mean ± SD. *P < 0.05 and **P < 0.01 vs. CTL

Fig. 4. CLQ suppresses the glycolysis in melanoma cells.

Mouse B16F10 cells were treated with the indicated concentrations of CLQ for 24 h. a Glucose consumption. b Relative ATP production. c Lactate production. d ECAR. e RT-qPCR and f Western blot analyses of GLUT1, HK2, PKM2 and LDHA expression levels. All values are presented as the mean ± SD from at least three separate experiments. *P < 0.05 and **P < 0.01 vs. CTL

Fig. 5. Bioinformatic analyses identify PPARγ as a potential drug target of CLQ.

a A cluster of the top 10 candidates from bioinformatics analyses using PharmMapper and DRAR-CPI software. To investigate the effect of CLQ on PPARγ expression, B16F10 cells were treated with 1.5 μM CLQ for the indicated time-points. b RT-qPCR and (c) Western blot analyses of total PPARγ expression levels. **P < 0.01 vs. 0 h. B16F10 cells were treated with 1.5 μM CLQ for 12 h. d Protein levels of cytosolic and nuclear PPARγ expression. e Immunofluorescence analysis of PPARγ protein expression in B16F10 cells. f Quantitative data of (e). All values are presented as the mean ± SD from at least three separate experiments. **P < 0.01 vs. CTL

Fig. 6. PPARγ antagonist alleviates the antimelanoma effects of CLQ in B16F10 cells.

B16F10 cells were treated with 1.5 μM CLQ with/without 10 μM GW9662 for 24 h. a EdU incorporation assay. b Protein expression levels of key regulators involved in cell cycle progression. c Wound-healing assays. d Protein expression levels of ICAM-1, VCAM-1 and MMPs. All values are presented as the mean ± SD from at least three separate experiments

Fig. 7. PPARγ mediates the inhibitory effects of CLQ on glycolysis in B16F10 cells.

B16F10 cells were treated with 1.5 μM CLQ with/without 10 μM GW9662 for 24 h. a Glucose consumption. b Relative ATP production. c Lactate production. d ECAR. e RT-qPCR and (f) Western blot analyses of GLUT1, HK2, PKM2, LDHA expression levels. All values are presented as the mean ± SD from at least three separate experiments. *P < 0.05 and **P < 0.01 vs. CTL, ^#P < 0.05 and ^##P < 0.01 vs. CLQ

Fig. 8

CLQ effectively suppresses the malignant phenotypes of melanoma through activation of PPARγ