Funding
Taiwan Ministry of Science and Technology
National Health Research Institutes
Taiwan Ministry of Health and Welfare
National Cheng Kung University Hospital
ARTICLE ABSTRACT
Purpose: Accumulating evidence indicates that factors secreted by cancer epithelial cells shape the tumor microenvironment to promote cancer invasion and metastasis. Recent studies also shed light on alterations of Rab small GTPase–mediated exocytosis in tumorigenesis. However, the mechanisms for Rab-mediated exocytosis in tumor microenvironment remain elusive. We aimed to investigate the interplay between Rab37-mediated exocytosis and tumor microenvironment, focusing on endothelial cell motility and angiogenesis.Experimental Design: We performed fluorescence IHC for Rab37, thrombospondin-1 (TSP1, an antiangiogenesis factor), and angiogenesis marker CD31 in 183 surgically resected esophageal squamous cell carcinoma (ESCC) patient samples. Cell migration, invasion, angiogenesis, and tumor metastasis were measured.Results: ESCC patients with low expression of Rab37 or TSP1 significantly correlated with high CD31 expression and were associated with worse progression-free survival. The multivariate Cox regression analysis showed that concordant low expression of both Rab37 and TSP1 was an independent prognostic factor of ESCC patients. Rab37-mediated exocytosis of TSP1 led to the inhibition of neovasculature in vitro and in vivo. Secreted TSP1 from cancer cells with Rab37 exocytic function inhibited the p-FAK/p-paxillin/p-ERK migration signaling in both cancer epithelial cells and their surrounding endothelial cells. Dysfunction of Rab37 or loss of TSP1 abrogated the suppressive effects on angiogenesis and metastasis.Conclusions: Our findings suggest that Rab37-mediated TSP1 secretion in cancer cells suppresses metastasis and angiogenesis via a cross-talk with endothelial cells and reveal a novel component of the vesicular exocytic machinery in tumor microenvironment and tumor progression. Dysregulation of Rab37/TSP1 axis has clinical implications for prognosis prediction. Clin Cancer Res; 23(9); 2335–45. ©2016 AACR.
