Abstract
Thirty percent of medullary thyroid carcinomas (MTCs) are related to dominant germline pathogenic variants in the RET proto-oncogene. According to their aggressiveness, these pathogenic variants are classified in three risk levels: 'moderate', 'high' and 'highest'. The present study compares the metabolomics profiles of five pathogenic variants, whether already classified or not. We have generated six stable murine fibroblast cell lines (NIH3T3) expressing the WT allele or variants of the human RET gene, with different levels of pathogenicity, including the M918V variant that is yet to be accurately classified. We carried out a targeted metabolomics study of the cell extracts with a QTRAP mass spectrometer, using the Biocrates Absolute IDQ p180 kit, which allows the quantification of 188 endogenous molecules. The data were then subjected to multivariate statistical analysis. One hundred seventy three metabolites were accurately measured. The metabolic profiles of the cells expressing the RET variants were found to be correlated with their oncogenic risk. In addition, the statistical model we constructed for predicting the oncogenic risk attributed a moderate risk to the M918V variant. Our results indicate that metabolomics may be useful for characterizing the pathogenicity of the RET gene variants and their levels of aggressiveness.