Analysis of TP53 mutation status in human cancer cell lines: a reassessment

Abstract

Tumor-derived cell lines play an important role in the investigation of tumor biology and genetics. Across a wide array of studies, they have been tools of choice for the discovery of important genes involved in cancer and for the analysis of the cellular pathways that are impaired by diverse oncogenic events. They are also invaluable for screening novel anticancer drugs. The TP53 protein is a major component of multiple pathways that regulate cellular response to various types of stress. Therefore, TP53 status affects the phenotype of tumor cell lines profoundly and must be carefully ascertained for any experimental project. In the present review, we use the 2014 release of the UMD TP53 database to show that TP53 status is still controversial for numerous cell lines, including some widely used lines from the NCI-60 panel. Our analysis clearly confirms that, despite numerous warnings, the misidentification of cell lines is still present as a silent and neglected issue, and that extreme care must be taken when determining the status of p53, because errors may lead to disastrous experimental interpretations. A novel compendium gathering the TP53 status of 2,500 cell lines has been made available (http://p53.fr). A stand-alone application can be used to browse the database and extract pertinent information on cell lines and associated TP53 mutations. It will be updated regularly to minimize any scientific issues associated with the use of misidentified cell lines (http://p53.fr).

Keywords: TP53; cancer cell line; cross-contamination; misidentification; recommendation.

Figure 1

Mutation spectrum in tumors and cell lines. Mutational events for the TP53 gene in cell lines and tumors have been compared with the pattern of mutational events obtained from exome sequencing of cell lines and tumors. Data were obtained from the last release of the TP53 mutation database (p53.free.fr), the TCGA portal (http://www.cbioportal.org/public-portal/), and the COSMIC database (http://cancer.sanger.ac.uk/cancergenome/projects/cosmic/).

Figure 2

Overview of the various features found in the TP53 Mutations in Cell Lines Compendium. After selecting the cancer type search (red stamp 1), the search panel is activated (red stamp 2) and several filtering criteria are proposed, such as: cancer type (green stamp 1), TP53 status (wild type vs. mutant, green stamp 2), mutant frequency (green stamp 3), number of mutations per cell lines (green stamp 4), and mutation type (missense, frameshift, or nonsense, green stamp 5). After selection, the list of all cell lines matching the various search criteria is displayed (red stamp 3). Clicking on a cell line (green stamp 6) displays the TP53 mutant with description of the mutations in two panels (red stamp 4 and 5). The first panel displays specific information related to the mutant including its frequency in the database and various structural information such as potential posttranslational modification or domain localization. It also includes pathogenicity prediction using multiple complementary tools such as Sift, Mutassessor, or PolyPhen. The second panel summarizes the consequences of the mutation for the eight transcripts and 12 protein isoforms expressed by the TP53 gene. Each panel is described in detail in Supp. Figure S1 with several examples of various searches.