Genomic and Transcriptomic Analyses of Malignant Pleural Mesothelioma (MPM) Samples Reveal Crucial Insights for Preclinical Testing

Cell lines serve as tools for understanding molecular signatures of predictive biomarkers. However, the use of highly passaged commercial cell lines has to be questioned, as these do not closely resemble the originating tumor. Here, we established patient-derived cell lines from MPM tumors and performed whole exome and mRNA sequencing to understand genomic and transcriptomic differences between MPM tumors, patient-derived cell lines, and commercial cell lines. Our results show that the genome and transcriptome of tumors correlate to a higher degree with patient-derived cell lines than with commercial cell lines. These findings are of major relevance for the scientific community in regard to using cell line models to study predictive biomarkers and to interpret preclinical results which cannot be translated in clinical practice.Cell lines are extensively used to study cancer biology. However, the use of highly passaged commercial cell lines has to be questioned, as they do not closely resemble the originating tumor. To understand the reliability of preclinical models for Malignant pleural mesothelioma (MPM) studies, we have performed whole transcriptome and whole exome analyses of fresh frozen MPM tumors and compared them to cell lines generated from these tumors, as well as commercial cell lines and a preclinical MPM mouse model. Patient-derived cell lines were generated from digested fresh tumors and whole exome sequencing was performed on DNA isolated from formalin-fixed, paraffin-embedded (FFPE) tumor samples, corresponding patient-derived cell lines, and normal tissue. RNA sequencing libraries were prepared from 10 fresh frozen tumor samples, the 10 corresponding patient-derived cell lines, and 7 commercial cell lines. Our results identified alterations in tumor suppressor genes such as FBXW7, CDKN2A, CDKN2B, and MTAP, all known to drive MPM tumorigenesis. Patient-derived cell lines correlate to a high degree with their originating tumor. Gene expressions involved in multiple pathways such as EMT, apoptosis, myogenesis, and angiogenesis are upregulated in tumor samples when compared to patient-derived cell lines; however, they are downregulated in commercial cell lines compared to patient-derived cell lines, indicating significant differences between the two model systems. Our results show that the genome and transcriptome of tumors correlate to a higher degree with patient-derived cell lines rather than commercial cell lines. These results are of major relevance for the scientific community in regard to using cell lines as an appropriate model, resembling the pathway of interest to avoid misleading results for clinical applications.