Background: Previous studies have described that the SEC23A gene is involved in the occurrence and development of various tumor entities. However, little is known about its expression and relevance in stomach adenocarcinoma (STAD). The aim of this study was to bioinformatically analyze the role of SEC23A in STAD, followed by patient tissue sample analyses. Materials and methods: SEC23A expression levels in STAD and normal gastric tissues were analyzed in the Cancer Genome Atlas and Gene Expression Omnibus databases; results were verified in fresh clinical STAD specimens on both gene and protein expression levels. SEC23A expression correlated with survival parameters by Kaplan–Meier and multivariate Cox regression analyses. The top genes co-expressed with SEC23A were identified by gene set enrichment analysis (GSEA) using the clusterProfiler package in R. Furthermore, the R package (immunedeconv), integrating the CIBERSORT algorithm, was used to estimate immune cell infiltration levels in STAD. Results: SEC23A gene and sec23a protein expression were both significantly upregulated in STAD, and this correlated with the pT stage. Moreover, high SEC23A expression was associated with poor disease-free and overall survival of STAD patients. Cox analyses revealed that besides age and pathologic stage, SEC23A expression is an independent risk factor for STAD. GSEA indicated that SEC23A was positively associated with ECM-related pathways. In the CIBERSORT analysis, the level of SEC23A negatively correlated with various infiltrating immune cell subsets, including follicular helper T cells, Tregs, activated NK cells and myeloid dendritic cells. Finally, the expression levels of immune checkpoint-related genes, including HAVCR2 and PDCD1LG2, were significantly increased in the high SEC23A expression group. Conclusions: We observed the significantly upregulated expression of SEC23A in STAD, an association with disease progression, patients’ prognosis and infiltrating immune cell subsets. Thus, we propose SEC23A as an independent prognostic factor with a putative role in immune response regulation in STAD.

IntroductionFor pre-clinical drug development and precision oncology research, robust cancer cell models are essential. Patient-derived models in low passages retain more genetic and phenotypic characteristics of their original tumors than conventional cancer cell lines. Subentity, individual genetics, and heterogeneity greatly influence drug sensitivity and clinical outcome.Materials and methodsHere, we report on the establishment and characterization of three patient-derived cell lines (PDCs) of different subentities of non-small cell lung cancer (NSCLC): adeno-, squamous cell, and pleomorphic carcinoma. The in-depth characterization of our PDCs included phenotype, proliferation, surface protein expression, invasion, and migration behavior as well as whole-exome and RNA sequencing. Additionally, in vitro drug sensitivity towards standard-of-care chemotherapeutic regimens was evaluated.ResultsThe pathological and molecular properties of the patients’ tumors were preserved in the PDC models HROLu22, HROLu55, and HROBML01. All cell lines expressed HLA I, while none were positive for HLA II. The epithelial cell marker CD326 and the lung tumor markers CCDC59, LYPD3, and DSG3 were also detected. The most frequently mutated genes included TP53, MXRA5, MUC16, and MUC19. Among the most overexpressed genes in tumor cells compared to normal tissue were the transcription factors HOXB9, SIM2, ZIC5, SP8, TFAP2A, FOXE1, HOXB13, and SALL4; the cancer testis antigen CT83; and the cytokine IL23A. The most downregulated genes on the RNA level encode the long non-coding RNA LANCL1-AS1, LINC00670, BANCR, and LOC100652999; the regulator of angiogenesis ANGPT4; the signaling molecules PLA2G1B and RS1; and the immune modulator SFTPD. Furthermore, neither pre-existing therapy resistances nor drug antagonistic effects could be observed.ConclusionIn summary, we successfully established three novel NSCLC PDC models from an adeno-, a squamous cell, and a pleomorphic carcinoma. Of note, NSCLC cell models of the pleomorphic subentity are very rare. The detailed characterization including molecular, morphological, and drug-sensitivity profiling makes these models valuable pre-clinical tools for drug development applications and research on precision cancer therapy. The pleomorphic model additionally enables research on a functional and cell-based level of this rare NCSLC subentity.