Fig 2: TGF‐β1 transcriptionally activates PSG9 by enhancing recruitment of Smad3 and Smad4 onto the PSG9 promoter regions containing two putative Smad‐binding elements (SBEs). A‐D, MCF‐7 cells were transfected with siNC or two independent siRNAs targeting Smad2 (siSmad2), Smad3 (siSmad3), or Smad4 (siSmad4). After 24 hours of transfection, cells were cultured in serum‐free media for 24 hours and then treated with or without 10 ng/mL of TGF‐β1 for another 24 hours. Immunoblotting analyses were carried out with the indicated antibodies (A‐C). The expression levels of PSG9 were normalized to those of vinculin. qPCR analyses were carried out to detect PSG9 mRNA levels (D). E, MCF‐7 cells were transfected with siNC, siSmad3, or siSmad4. After 24 hours of transfection, cells were transfected with pGL3‐PSG9 and cultured in serum‐free media for 24 hours, followed by treatment with or without 10 ng/mL of TGF‐β1 for another 24 hours. The PSG9 promoter activity was determined using a Dual‐Luciferase Reporter Assay System and normalized to the values of Renilla luciferase. F, MCF‐7 cells were cultured in serum‐free media for 24 hours and treated with or without 10 ng/mL of TGF‐β1 for another 24 hours. ChIP assays were performed with control IgG or a specific antibody against Smad2, Smad3, or Smad4, followed by qPCR analysis with specific primers amplifying PSG9 promoter regions (region 1 and region 2) containing two putative Smad‐binding elements (SBEs). Recruitment of Smad3 and Smad4 to the PSG9 promoter regions containing two putative SBEs was normalized to the Input. G, MCF‐7 cells were transfected with pGL3, pGL3‐PSG9, or mutant pGL3‐PSG9 and cultured in serum‐free media for 24 hours, followed by treatment with or without 10 ng/mL of TGF‐β1 for another 24 hours. The PSG9 promoter activity was determined using a Dual‐Luciferase Reporter Assay System and normalized to the values of Renilla luciferase. Note: mut1, the first SBE was mutated; mut2, the second SBE was mutated; mut 1&2, both SBEs were mutated. **P < .01; ***P < .001; NS, no significance

Fig 3: PSG9 contributes to TGF‐β1‐induced epithelial‐mesenchymal transition (EMT) and breast cancer cell migration and invasion. A, MCF10A cells stably expressing pCDH and HA‐PSG9 were cultured in serum‐free media for 24 hours, treated with or without 10 ng/mL of TGF‐β1 for another 24 hours, and photographed under a phase contrast microscopy. Scale bar: 100 μM. B, MCF10A cells stably expressing pCDH and HA‐PSG9 were cultured in serum‐free media for 24 hours, treated with or without 10 ng/mL of TGF‐β1 for another 24 hours, and then subjected to immunoblotting analysis with the indicated antibodies. C and D, LM2‐4175 and Hs578T cells stably expressing shNC and shPSG9 were cultured in serum‐free media for 24 hours and treated with or without 10 ng/mL of TGF‐β1 for another 24 hours, and then subjected to Boyden's chamber migration assays and Matrigel‐coated invasion assays. Representative images (C) and quantitative results of migrated and invaded cells from three biological replicates (D) are shown. **P < .01

Fig 4: PSG9 enhances breast cancer cell migratory and invasive potential in vitro and lung metastasisin vivo. A and B, MDA‐231 cells stably expressing pCDH and HA‐PSG9 were subjected to wound‐healing assays (A) or Boyden's chamber migration assays and Matrigel‐coated invasion assays (B). Representative images (A, left) and quantitative results of percent wound closure from three biological replicates (A, right). Representative images (B, left) and quantitative results of migrated and invaded cells from three biological replicates (B, right). C and D, LM2‐4175 and Hs578T cells stably expressing shNC and shPSG9 were subjected to wound‐healing assays (C) or Boyden's chamber migration assays and Matrigel‐coated invasion assays (D). Representative images (C, left) and quantitative results of percent wound closure from three biological replicates (C, right). Representative images (D, left) and quantitative results of migrated and invaded cells from three biological replicates (D, right) are shown. E and F, MDA‐231 cells stably expressing pCDH and HA‐PSG9 were injected into 6‐week‐old female BALB/c nude mice (n = 6) through the tail vein. After 6 weeks of injection, the lungs were harvested and stained with Bouin's solution. Representative images of lung metastasis (E) and quantitative results of lung nodules (F) are shown. G and H, LM2‐4175 cells stably expressing shNC and shPSG9 were injected into 6‐week‐old female BALB/c nude mice (n = 6) through the tail vein. After 6 weeks of injection, the lungs were harvested and stained with Bouin's solution. Representative images of lung metastasis (G) and quantitative results of lung nodules (H) are shown. *P < .05; **P < .01; ***P < .001

Fig 5: PSG9 levels are elevated in tumor tissues and plasma specimens from breast cancer patients and are associated with poor prognosis. A, Total lysates from nine pairs of matched adjacent normal breast tissues, primary breast tumors, and metastatic lymph node (LN) tissues were subjected to immunoblotting analysis with the indicated antibodies. Oncoprotein MTA1 was used as a positive control. B, Total 161 surgical specimens from breast cancer patients with clinical follow‐up information were subjected to immunohistochemical staining of PSG9 with a specific antibody against PSG9 (Novus, #NBP2‐19980). The representative images for PSG9 staining are shown. Bar, 25 μM. C, Kaplan‐Meier curves of disease‐free survival (DFS) of 161 breast cancer patients with high or low PSG9 expression based on the median of IRS value. Statistics analysis was performed by the log‐rank test. D, Detection of PSG9 levels by ELISA in plasma specimens from 20 healthy controls and 60 patients with breast cancer. E, Kaplan‐Meier curves of DFS of 161 breast cancer patients with high or low PSG9 plasma levels. *P < .05; **P < .01; ***P < .001