PAK inhibitor FRAX486 decreases the metastatic potential of triple-negative breast cancer cells by blocking autophagy

Abstract

Background: Triple-negative breast cancer (TNBC) is a unique breast cancer subtype with a high risk of metastasis and recurrence and a poor prognosis. Epithelial-mesenchymal transition (EMT) endows epithelial cells with the ability to move to distant sites, which is essential for the metastasis of TNBC to organs, including the lung. Autophagy, an intracellular degradation process that involves formation of double-layered lipid autophagosomes that transport cytosolic cargoes into lysosomes via autophagosome-lysosome fusion, is involved in various diseases, including cancer and neurodegenerative, metabolic, cardiovascular, and infectious diseases. The relationship between autophagy and cancer has become relatively clear. However, research on pharmacological drugs that block cancer EMT by targeting autophagy is still in the initial stages. Therefore, the re-evaluation of old drugs for their potential in blocking both autophagy and EMT was conducted.

Methods: More than 2000 small molecule chemicals were screened for dual autophagy/EMT inhibitors, and FRAX486 was identified as the best candidate inhibitor of autophagy/EMT. The functions of FRAX486 in TNBC metastasis were detected by CCK-8, migration and wound healing assays. The effects of FRAX486 on autophagy and its target PAK2 were determined by immunoblotting, immunofluorescence, immunoprecipitation analysis and transmission electron microscopy. The findings were validated in mouse models.

Results: Here, we report that FRAX486, a potent P21-activated kinase 2 (PAK2) inhibitor, facilitates TNBC suppression both in vitro and in vivo by blocking autophagy. Mechanistically, FRAX486 inhibits autophagy in TNBC cells by targeting PAK2, leading to the ubiquitination and proteasomal degradation of STX17, which mediates autophagosome-lysosome fusion. The inhibition of autophagy by FRAX486 causes upregulation of the epithelial marker protein E-cadherin and thus suppresses the migration and metastasis of TNBC cells.

Conclusions: The effects of FRAX486 on TNBC metastasis suppression were verified to be dependent on PAK2 and autophagy inhibition. Together, our results provide a molecular basis for the application of FRAX486 as a potential treatment for inhibiting the metastasis of TNBC.

Fig. 1. FRAX486 inhibits metastasis and the autophagic degradation of E-cadherin in TNBC cells.

a CCK-8 assay was measured in MDA-MB-231 and E0771 cells treated with the appointed concentrations of FRAX486 for 72 h. b FRAX486 inhibited migration of TBNC cells determined by migration assay. Transwell chambers images of cell migration (left), number of migration cells (right). Scale bar: 500 μm. c FRAX486 inhibited migration of TBNC cells determined by wound healing assay. Wound healing images (left), wound healing rate (right). Scale bar: 500 μm. d Nude mice (n = 5 per group) were injected with 2 × 10⁶ WT, ATG7 knockout, PAK2 knockdown MDA-MB-231 cells via tail vein, and treated with vehicle and 20 mg/kg FRAX486 by gavage for 5 days each week. After 7 weeks’ treatment, lung tissues were collected, and metastatic nodules were quantified. e Lung metastasis of MDA-MB-231 cells in Nude mice (n = 5 per group). f The number of metastatic nodules in the lung. g Pathologic analysis of lung tissues by HE staining. Scale bar: 500 μm. h Tumour area ratio of lung metastatic. i Immunoblots analysis of E-cadherin and Vimentin in TNBC cells treated with the indicated concentrations of FRAX486 for 24 h. j The effect of MG132 or CQ on E-cadherin degradation in TNBC MDA-MB-231 and E0771 cells. k Effect of ATG7 knockout and FRAX486 on E-cadherin protein levels in TNBC MDA-MB-231 and E0771 cells. The data shown as mean ± SD and represent 3 independent experiments, and *p < 0.05; **p < 0.01 compared with DMSO treatment only.

Fig. 2. FRAX486 suppresses TNBC migration and metastasis via autophagy inhibition.

a, b Wound healing assay detected the migration in WT and ATG7 knockout TNBC MDA-MB-231 and E0771 cells treated with FRAX486. c, d Migration assay detected the migration in WT and ATG7 knockout TNBC cells treated with or without FRAX486. e Lung metastasis of ATG7 knockout MDA-MB-231 cells in Nude mice (n = 5 per group). f The number of metastatic nodules in the lung. g pathologic analysis of lung tissues by HE staining. h Tumour area ratio of lung metastatic. Scale bar: 500 μm. The data shown as mean ± SD and represent 3 independent experiments, and *p < 0.05; **p < 0.01.

Fig. 3. FRAX486 inhibits autophagy in TNBC cells.

a, b Immunoblot analysis of LC3 and p62 protein levels in WT or ATG knockout TNBC cells treated with the indicated concentrations of FRAX486 for 24 h. c Immunoblot analysis of LC3 and p62 protein levels in WT TNBC cells treated with FRAX486 and/or CQ for 24 h. d, e Formation of exogenous LC3 puncta in TBNC cells treated with 2 μM FRAX486 for 24 h. Scale bar: 20 μm. f, g Immunofluorescence analysis of endogenous LC3 puncta in WT and ATG7 knockout TBNC cells treated with 2 μM FRAX486 for 24 h. Scale bar: 20 μm. h, i Transmission electron microscopy detected the autophagosomes in WT and ATG7 knockout TBNC cells. Scale bar: 1 μm. Red arrow: mitochondria; black arrow: autolysosome; yellow arrow: endoplasmic reticulum (ER); green arrow: nucleus. The data shown as mean ± SD and represent 3 independent experiments, and *p < 0.05; **p < 0.01.

Fig. 4. FRAX486 inhibits autophagosome–lysosome fusion via STX17 degradation.

a, b Fluorescence microscopy detected the autophagy flux in cells with stable expression of mCherry-GFP-LC3 treated with 2 μM FRAX486. Scale bar: 20 μm. c, d Immunofluorescence analysis of co-localisation of GFP-LC3 and LAMP1 in cells treated with 2 μM FRAX486. Scale bar: 20 μm. e Schematic diagram of the role of the STX17-SNAP29-VAMP8 complex in autophagosome–lysosome fusion. f Immunoblot determined the protein levels of STX17, SNAP29 and VAMP8 in MDA-MB-231 cells treated with the indicated concentrations of FRAX486 for 24 h. g The effect of MG132 on STX17 degradation in MDA-MB-231 cells. h MDA-MB-231 cells with GFP or GFP-STX17 transfection were submitted to FRAX486 treatment or not. Immunoprecipitation (IP) with anti-GFP was used to detect STX17 poly-ubiquitin. i MDA-MB-231 cells with HA -STX17 transfection were submitted to FRAX486 treatment or not. Immunoprecipitation (IP) with anti-HA was used to detect STX17 phosphorylation. The data shown as mean ± SD and represent 3 independent experiments, and *p < 0.05; **p < 0.01.

Fig. 5. FRAX486 targets PAK2 to induce STX17 degradation.

a MDA-MB-231 cells with ctr shRNA or PAK2 shRNA transfection were subjected to the indicated concentrations of FRAX486. Immunoblot analysis of LC3 and p62 protein levels. b, c Immunofluorescence analysis of endogenous LC3 puncta in ctr shRNA and PAK2 shRNA MDA-MB-231 cells treated with 2 μM FRAX486 or not. Scale bar: 20 μm. d, e Immunofluorescence tested co-localisation of LC3 and LAMP1 in cells treated as in (b). Scale bar: 20 μm. f MDA-MB-231 cells with GFP or GFP-STX17 transfection were used to investigate the interaction of STX17 with PAK1 and PAK2. g Immunoblot analysis of GFP-STX17 protein levels in cells treated as in (a). h MDA-MB-231 cells with Ctr shRNA or PAK2 shRNA transfection were submitted to 2 μM FRAX486 treatment or not. Immunoprecipitation (IP) with anti-GFP was used to detect STX17 poly-ubiquitin. The data shown as mean ± SD and represent 3 independent experiments, and ∗p < 0.05; ∗∗p < 0.01.

Fig. 6. FRAX486 targets PAK2 to inhibit TNBC migration and metastasis.

a Wound healing assay detected the wound healing rate in MDA-MB-231 cells with control shRNA or PAK2 shRNA transfection were subjected to the 2 μmol/L FRAX486 or not. Scale bar: 500 μm. b, c Migration assay analysis of number of migration cells in cells treated as in (a). Scale bar: 500 μm. d Lung metastasis of PAK2 knockout MDA-MB-231 cells in Nude mice (n = 5 per group). e The number of metastatic nodules in the lung. f Pathologic analysis of lung tissues by HE staining. g Tumour area ratio of lung metastatic. Scale bar: 500 μm. The data shown as mean ± SD and represent 3 independent experiments, and *p < 0.05; **p < 0.01.