Abstract Lung cancer is the leading cause of cancer mortality due in part to an inability to diagnose and intercept the disease at its earliest and potentially most curable stage. In order to improve lung cancer prevention and early detection, a detailed understanding of both the genomic alterations driving carcinogenesis and the mechanisms by which altered cells escape immune surveillance is required. Toward this goal, our group is profiling RNA and DNA from bronchial premalignant lesions (PMLs), precursors of lung squamous cell carcinoma (LUSCC), to identify alterations associated with their development and progression. Results from the analysis of 197 human endobronchial biopsies representing all histologic grades profiled by RNA-Seq, identified a class of high-grade PMLs with high activity of pathways associated with epithelial proliferation and differentiation. Among this class of PMLs, progression-associated pathways included immunoproteasome, antigen processing, immune inhibition, and macrophage activation. Further elucidation of the epithelial and immune changes that characterize progression of PMLs and testing of interception strategies based on these human findings require a preclinical model with similar histology and molecular alterations. The N-nitrosotris-(2-choroethyl)urea (NTCU) mouse model of LUSCC is a promising preclinical model that develops histologically comparable PMLs to those found in humans; however, the molecular alterations associated with these histologic changes have not yet been characterized. In order to establish molecular concordance between NTCU-treated mouse and human PMLs, we isolated RNA from 40 fresh-frozen whole lung sections and laser capture microdissected (LCM) tissue from SWR/J and A/J mice treated with NTCU. We profiled by RNA-Seq tissue that represented a range of histologic grades including normal (n=9), metaplasia/mild dysplasia (n=8), moderate/severe dysplasia (n=19), and carcinoma in situ/SCC (n=4). RNA-Seq data was aligned to mm9 and gene-level counts were summarized using Ensembl annotation. After removing samples of low quality, linear modeling adjusting for sample quality, sample type, and mouse strain was used to identify differentially expressed genes associated with histologic grade. Immune cell abundance was estimated using CIBERSORT and EnrichR was used to identify pathways associated with histologic grade. Concordance between the mouse data and human data was established using GSEA and GSVA. We identified 596 genes differentially expressed with increasing histologic severity across all mouse samples (FDR<0.05). The genes are involved in cancer signaling including protein processing, DNA damage, and p53 signaling as well as immune regulation including interferon signaling and lymphocyte and myeloid cell regulation (FDR<0.05). Upon further investigations of the immune-associated changes, we observed increases in T-cell populations in moderate dysplasia samples that declined in more severe lesions. Comparing the mouse strains, the SWR/J strain had more variability in immune cell content and types of immune cells, and the A/J strain had a more prominent macrophage polarization with increasing histologic severity. We also identified significant concordance of gene expression differences associated with histology between mouse and human (FDR<0.05). Concordant enrichment included alterations in DNA damage, cell cycle, and mitochondrial respiration in addition to immune modulation associated with progression of PMLs in human lesions. The results suggest that molecular changes and immune modulation in the NTCU model are similar to those observed in human PMLs. This work indicates that the NTCU mouse model may be a relevant preclinical model for investigating targeted agents to halt or reverse PML progression to reverse or delay lung cancer development in the damaged epithelium. Citation Format: Sarah Mazzilli, Sherry Zhang, Gang Liu, Marc Lenburg, Avrum Spira, Jennifer Beane. The carcinogen-induced NTCU model: A preclinical mouse model for lung cancer interception [abstract]. In: Proceedings of the AACR Special Conference: Advances in Modeling Cancer in Mice: Technology, Biology, and Beyond; 2017 Sep 24-27; Orlando, Florida. Philadelphia (PA): AACR; Cancer Res 2018;78(10 Suppl):Abstract nr B18.
Read full abstract