Loss of PPM1A expression enhances invasion and the epithelial-to-mesenchymal transition in bladder cancer by activating the TGF-β/Smad signaling pathway

Abstract

The transforming growth factor-β (TGF-β) signaling pathway is believed to contribute to carcinoma development by increasing cell invasiveness and metastasis and inducing the epithelial-to-mesenchymal transition (EMT). Protein phosphatase PPM1A has been reported to dephosphorylate TGF-β-activated Smad2/3, thus inhibiting the TGF-β signaling pathway. In this study, we investigated the role of PPM1A in bladder cancer. PPM1A protein expression was analyzed in 145 bladder cancer specimens. The loss of PPM1A expression was predictive of poor survival and high muscle-invasiveness. PPM1A was more commonly deficient among muscle-invasive relapse samples compared to primary tumors in twenty paired bladder cancer tissues. Functional studies indicated that blockade of PPM1A through lentivirus-mediated RNA interference significantly promoted urinary bladder cancer (BCa) cell motility, the EMT in vitro and metastasis in vivo, and these effects were dependent on the TGF-β/Smad signaling pathway. The increase in p-Smad2/3 induced by TGF-β1 correlated with the degree of PPM1A depletion in BCa cells, which resulted in an altered expression profile of TGF-β-inducible genes. The correlations between PPM1A and biomarkers related to the TGF-β signaling pathway and tumor invasion were also detected in BCa samples. These results demonstrate that loss of PPM1A is associated with the development of tumor invasion in bladder cancer patients.

Figure 1. PPM1A expression correlated with prognosis and muscle invasion in patients with BCa

(A) Representative IHC staining of the PPM1A protein expression in bladder cancer tissues of different histological differentiation grades is shown (× 400). (B) Matched pairs of primary tumors and recurrent BCa samples were analyzed for PPM1A protein expression. The expression of PPM1A in recurrent muscle-invasive BCa was significantly lower than that of primary non-muscle-invasive BCa (× 200). (C) Correlation between the expression of PPM1A and tumor muscle invasion in 145 BCa samples. (D) Kaplan-Meier curves with log-rank analyses for BCa patients with negative PPM1A-expressing tumors (n = 46) versus positive PPM1A-expressing tumors (n = 99).

Figure 2. Downregulation of PPM1A significantly promoted cellular invasion, which was dependent on TGF-β1, in vitro

(A) Western blotting analysis of PPM1A expression in five bladder cancer cell lines. Establishment and selection of notably and stably PPM1A-silenced T24 and 5637 cells, Western blotting. (B) and real-time RT-PCR. (C) were performed to detect the protein and mRNA expression. (D) The effect of PPM1A knockdown on the migration rates of BCa cells treated with TGF-β1 (200 pM) or vehicle was compared in wound healing assays in T24 and 5637 cells. (E) The invasive capacity of BCa cells treated with TGF-β1 (200 pM) or vehicle was compared using Matrigel invasion assays in T24 and 5637 cells.

Figure 3. PPM1A suppressed tumor cell invasion and metastasis in BCa cells in vivo

(A and B) The effect of PPM1A on tumor growth was evaluated in a nude mouse xenograft model, T24 control and T24-PPM1A-RNAi (loss expression of PPM1A) cells were xenografted and the weights of tumors formed in the xenograft model. (C) The effect of PPM1A on tumor growth was evaluated in a nude mouse xenograft model, T24 vector, T24-PPM1A (high expression of PPM1A) cells were xenografted, and the weights of tumors formed in the xenograft model. (D and E) At postmortem examination, the pseudo-capsules of the xenograft tumors formed from the T24-PPM1A-RNAi (low-expression PPM1A cell) cells were absent, whereas the xenograft tumors formed from T24 control PPM1A-RNAi cells demonstrated a smooth surface with well-formed pseudo-capsules. (F) Lung metastasis assay in vivo. The metastatic lesions in the mice's lungs were captured in T24-PPM1A-RNAi, T24-PPM1A and control T24 cell lines.

Figure 4. PPM1A terminated TGF-β signaling in BCa cells by dephosphorylating TGF-β-activated Smad2/3

PPM1A was silenced in T24 and 5637 cells, which were then treated with vehicle, TGF-β1 and/or SB431542 as indicated. (A and B) At the indicated time intervals, Western blotting was performed to analyze the levels of phosphorylated Smad2 and Smad3. GAPDH served as a loading control. (C) Cells were co-treated with or without TGF-β1 (200 pM) and SB431542 (5 μM) for 24 h. Western blotting was performed to assess the expression of phosphorylated Smad2 and Smad3 and GAPDH. (D) Cells were either untreated or induced with TGF-β1 (200 pM) for 24 h. Total RNA was extracted, and the expression of TGF-β1, CTGF and PAI-1 was assessed by real-time RT-PCR.

Figure 5. PPM1A inhibited BCa cells invasion and EMT, which was dependent on TGF-β/Smad signaling

PPM1A was silenced in T24 and 5637 cells, treated with vehicle, TGF-β1 and/or SB431542, as indicated. (A and B) The number of cells that had invaded through the membrane with 1% gelatin was counted. (C) Western blotting analysis of the expression of MMP2 and MMP9. GAPDH was used as a loading control. (D) mRNA was extracted, and MMP2 and MMP9 expression was analyzed by real-time RT-PCR. (E) Western blotting analysis of the expression of E-cadherin, CK19 and vimentin. GAPDH was used as a loading control. F. mRNA was extracted, and the expression of vimentin, CK19 and E-cadherin was analyzed by real-time RT-PCR.

Figure 6. Immunohistochemical staining for PPM1A, p-Smad2/3, MMP2 and E-cadherin in 145 BCa samples

(A) Representative images of hematoxylin and eosin staining and IHC for PPM1A, p-Smad2/3 and MMP2 from consecutive sections are shown (× 200). (B) The correlation between PPM1A and p-Smad2/3 was analyzed in tissues from 145 cases of BCa. (C) Representative IHC images for PPM1A and E-cadherin from consecutive sections are shown (× 200).