A Transcriptomic Analysis of Cancer-Stromal Interactome in Lung Cancer Xenograft Models

Cancer-stromal interactions play important roles in the biology of various cancers, including lung adenocarcinoma. We aimed to comprehensively analyze the lung cancer interactome and identify the key ligand-receptor pairs involved in the aggressiveness of lung adenocarcinoma. Transcriptome data were obtained from xenografts of 11 lung cancer cell lines that represented the major driver mutations in lung adenocarcinomas. A quantitative dataset was constructed in both stroma-to-cancer and cancer-to-stroma directions using the cancer-stromal interactome analysis method. The prognostic value of each factor was evaluated using multiple datasets. Analysis of 24,250 stroma-derived mouse transcripts and 26,289 human cancer-derived transcripts identified 1150 cancer-stromal interactions, from which we selected 117 interactions based on the intensity score of ligand-stromal transcript levels. Further prognostic analysis using public databases led us to identify 21 ligand-receptor pairs, including well-known as well as less well-characterized ligand-receptor pairs. Therefore, we selected tumor necrosis factor superfamily member 12/tumor necrosis factor receptor superfamily member 12A as possible factors contributing to the aggressiveness of lung adenocarcinoma via cancer-stromal interactions; immunohistochemical analysis confirmed that these factors were expressed mainly in the stroma and cancer cells, respectively, in both xenografts and primary lung adenocarcinoma. In human clinical specimens, high tumor necrosis factor receptor superfamily member 12A expression significantly correlated with tumor size, invasive diameter, and stage. Thus, tumor necrosis factor superfamily member 12 and its receptor tumor necrosis factor receptor superfamily member 12A signaling axis may be potential candidates for therapeutic intervention for lung adenocarcinoma.