Identification ofNTN4, TRA1, andSTC2as Prognostic Markers in Breast Cancer in a Screen for Signal Sequence Encoding Proteins

Abstract

In a previous screen using a signal-trap library, we identified a number of secreted proteins up-regulated in primary tumor cells isolated from invasive breast cancers. The purpose of this study was to assess the expression of these genes in human invasive breast tumors and to determine the significance of their expression for prognosis in breast cancer. A tissue microarray containing 245 invasive breast tumors from women treated with curative surgery followed by anthracycline-based chemotherapy and hormone therapy for the estrogen receptor (ER)-positive tumors was screened by in situ hybridization with probes against thrombospondin 3 (TSP3), insulin-like growth factor binding protein 7 (IGFBP7), tumor rejection antigen 1 (TRA1), stanniocalcin 2 (STC2), and netrin 4 (NTN4). Correlations between categorical variables were done using the chi(2) test and Fisher's exact test. Cumulative survival probabilities were calculated using the Kaplan-Meier method and multivariate survival analysis was done with Cox hazard model. A series of breast cancers were also stained with NTN4 antibodies. All five genes examined were expressed in invasive breast tumor cells. NTN4 protein expression was also confirmed by immunohistochemistry. Together, these data validate the design and screening of the signal-trap library. Univariate survival analysis revealed that expressions of TRA1, STC2, and NTN4 are correlated with longer disease-free survival and that TRA1 and NTN4 are associated with longer overall survival. Multivariate analysis showed that NTN4 is an independent prognostic factor of overall survival. This article describes the identification of three secreted proteins, NTN4, TRA1, and STC2, as potential novel prognostic markers in breast cancer.