Pancreatic RECK inactivation promotes cancer formation, epithelial-mesenchymal transition, and metastasis
Tomonori Masuda, … , Makoto Noda, Hiroshi Seno
Tomonori Masuda, … , Makoto Noda, Hiroshi Seno
Published September 15, 2023
Citation Information: J Clin Invest. 2023;133(18):e161847. https://doi.org/10.1172/JCI161847.
View: Text | PDF
Research Article Gastroenterology Oncology

Pancreatic RECK inactivation promotes cancer formation, epithelial-mesenchymal transition, and metastasis

Abstract

RECK is downregulated in various human cancers; however, how RECK inactivation affects carcinogenesis remains unclear. We addressed this issue in a pancreatic ductal adenocarcinoma (PDAC) mouse model and found that pancreatic Reck deletion dramatically augmented the spontaneous development of PDAC with a mesenchymal phenotype, which was accompanied by increased liver metastases and decreased survival. Lineage tracing revealed that pancreatic Reck deletion induced epithelial-mesenchymal transition (EMT) in PDAC cells, giving rise to inflammatory cancer-associated fibroblast–like cells in mice. Splenic transplantation of Reck-null PDAC cells resulted in numerous liver metastases with a mesenchymal phenotype, whereas reexpression of RECK markedly reduced metastases and changed the PDAC tumor phenotype into an epithelial one. Consistently, low RECK expression correlated with low E-cadherin expression, poor differentiation, metastasis, and poor prognosis in human PDAC. RECK reexpression in the PDAC cells was found to downregulate MMP2 and MMP3, with a concomitant increase in E-cadherin and decrease in EMT-promoting transcription factors. An MMP inhibitor recapitulated the effects of RECK on the expression of E-cadherin and EMT-promoting transcription factors and invasive activity. These results establish the authenticity of RECK as a pancreatic tumor suppressor, provide insights into its underlying mechanisms, and support the idea that RECK could be an important therapeutic effector against human PDAC.

Authors

Tomonori Masuda, Akihisa Fukuda, Go Yamakawa, Mayuki Omatsu, Mio Namikawa, Makoto Sono, Yuichi Fukunaga, Munemasa Nagao, Osamu Araki, Takaaki Yoshikawa, Satoshi Ogawa, Kenji Masuo, Norihiro Goto, Yukiko Hiramatsu, Yu Muta, Motoyuki Tsuda, Takahisa Maruno, Yuki Nakanishi, Toshihiko Masui, Etsuro Hatano, Tomoko Matsuzaki, Makoto Noda, Hiroshi Seno

×

Figure 7

Involvement of MMPs in RECK-induced upregulation of E-cadherin and suppression of EMT in PDAC cells.

Options: View larger image (or click on image) Download as PowerPoint
Involvement of MMPs in RECK-induced upregulation of E-cadherin and suppr...
(A) Detection of MMP2, MMP3, and MMP9 in culture supernatant of Reck-null PDAC cells transfected with control or RECK expression vector by immunoblot assay. Note that we adjusted the sample volume based on cell number counted at the time of harvest for direct comparison. Data are shown as the mean ± SEM of the densitometry measurements of the immunoblot band. P = 0.027 (MMP2), P = 0.031 (MMP3), P = 0.155 (MMP9); 2-tailed Student’s t test. (B) Detection of 72 kDa gelatinase (MMP2) in culture supernatant by gelatin zymography. Culture supernatants as used in A were subjected to gelatin zymography. (C) Effects of an MMP inhibitor (GM6001) on E-cadherin expression in Reck-null PDAC cells. Cells treated as shown in the top diagram were subjected to immunoblot assay using antibodies against β-actin (loading control, top) and E-cadherin (bottom). Data are shown as the mean ± SEM of the densitometry measurements of the immunoblot band. P = 0.024; 2-tailed Student’s t test. (D) Effects of GM6001 on EMT marker genes. Total RNA from the cells as used in C was subjected to RT-qPCR to estimate the levels of indicated mRNAs. Data are shown as the mean ± SEM. P = 0.01 (Zeb1); P = 0.033 (Snai2); P = 0.00009 (Twist1), 2-tailed Student’s t test. (E) Effects of GM6001 on PDAC cell invasion. Matrigel invasion assay was performed using Reck-null PDAC cell lines (n = 6) treated with either DMSO or GM6001. Top: Images of cells that migrated through the Matrigel insert membrane after staining (blue). Scale bar: 50 μm. Bottom: Data are shown as the mean ± SEM of the optical density (OD) at 570 nm. P = 0.00000014; 2-tailed Student’s t test. (F) Models consistent with our findings.