MB-88: QUANTITATIVE PROTEOMIC PROFILING IN MEDULLOBLASTOMA RECAPITULATES SUBGROUPING BASED ON TRANSCRIPTIONAL PROFILES

There is an important need to identify clinically reliable protein markers, which can be used for medulloblastoma subgroups assignment. Using our established proteomics-based target discovery program, we propose to identify and validate protein-based markers that are main pro-oncogenic signaling components in different molecular and clinical subgroups in medulloblastoma. Here, we report description and pilot results of a novel approach based on relative protein quantification by liquid chromatography tandem mass spectrometry (LC-MS/MS) and tandem mass tag (TMT) technology in a set of medulloblastoma samples of all 4 known molecular subgroups in clinically standard risk patients. Summary scores that characterized variability of concentration ratio of the measured proteins were computed as between- to within-groups sum of squares ratios. Several proteins with highest score were then selected for cluster analysis. Two cluster algorithms were used - either based on Euclidean distance or correlation measure. Results were presented as heatmaps with cluster dendrograms for proteins and patients. Interestingly, our proteomic based clustering shows, for the first time to our knowledge, recapitulation of subgrouping in medulloblastoma as based on transcriptional profiles. Proteins identified as linked to different medulloblastoma molecular subgroups were mainly involved in regulation of mitochondrial energy metabolism and actin cytoskeleton dynamics. The study was supported by the grant AZV MZCR NV15-30657A and the project MEYS - NPS I - LO1413