A feasibility study on the detection of clear cell renal cell carcinoma by mass spectrometry-based metabolomics

MALENA MANZI; MARÍA E. MONGE; MARÍA E. KNOTT; LYDIA I. PURICELLI ; MARIO O. SALAZAR

Congreso; 2do Simposio internacional: Frontiers in Bioscience; 2016

Renal Cell Carcinoma (RCC) is among the 10 most common cancers in both men and women and accounts for nearly 300,000 cases worldwide. 1Clear cell Renal Cell Carcinoma (ccRCC), characterized by high lipid and glycogen reserve, is the most common (75%) subtype of RCC. 1Current research has shown that several metabolic alterations are associated with ccRCC tumor progression and metastasis. 2 However, these alterations are still not well understood, and more studies are needed to find robust biomarkers for diagnosis, prognosis and follow upcare. Two parallel metabolomic studies were conducted to investigate the metabolic profiles of ccRCC. An untargeted lipidomic profiling pilot study of age-matched serum samples from patients with advanced ccRCC (stage IV, n=5) and healthy individuals (n=5) was performed using ultraperformance liquid chromatography coupled to quadrupole-time-of-flight mass spectrometry. Metabolites were extracted from the serum samples with isopropanol. High resolution mass spectra were acquired in negative ionization mode across the range of m/z 50?1200. Metabolic features (retention time, m/z pairs) were obtained via Progenesis QI software, and analyzed using a cross-validated orthogonal projection to latent structures-discriminant analysis (OPLS-DA) model. This supervised model classified samples as advanced ccRCC patients or healthy individuals with high accuracy (R2Y=0.9985 and Q2=0.9956). Discriminant metabolic features of this study suggest alterations of glycerophospholipids and cholesterol metabolism in agreement with previous studies. In parallel, an in vitro metabolomics study was conducted on conditioned media of ccRCC cell lines, 786-O (VHL-/-) and Caki-1 (VHL+/+), and the non-tumoral cell line HEK-293, to investigate altered metabolic networks associated with the tumor biology in these ccRCC models. Sample processing, data acquisition and data processing were performed as described above. Principal component analysis showed 3 separated clusters associated to the studied cell lines. Supervised statistical models are being performed to find discriminant metabolic features of thesein vitromodels. REFERENCES1.Hakimi, A. A.; Pham, C. G.; Hsieh, J. J.Nat.Genet.2013, 45, 849-850.2.Lin, L.; Huang, Z. Z.; Gao, Y. et al.J. Proteome Res.2011, 10, 1396-1405.