Multicellular Tumor Spheroids Combined with Mass Spectrometric Histone Analysis To Evaluate Epigenetic Drugs.

Multicellular tumor spheroids (MCTS) are valuable in vitro tumor models frequently used to evaluate the penetration and efficacy of therapeutics. In this study, we evaluated potential differences in epigenetic markers, i.e., histone post-translational modifications (PTMs), in the layers of the HCT116 colon carcinoma MCTS. Cells were grown in agarose-coated 96 well plates, forming reproducible 1-mm-diameter MCTS. The MCTS were fractionated into three radially concentric portions, generating samples containing cells from the core, the mid and the external layers. Using mass spectrometry (MS)-based proteomics and EpiProfile, we quantified hundreds of histone peptides in different modified forms; by combining the results of all experiments, we quantified the abundance of 258 differently modified peptides, finding significant differences in their relative abundance across layers. Among these differences, we detected higher amounts of the repressive mark H3K27me3 in the external layers, compared to the core. We then evaluated the epigenetic response of MCTS following UNC1999 treatment, a drug targeting the enzymes that catalyze H3K27me3, namely, the polycomb repressive complex 2 (PRC2) subunits enhancer of zeste 1 (EZH1) and enhancer of zeste 2 (EZH2). UNC1999 treatment resulted in significant differences in MCTS diameter under drug treatment of varying duration. Using matrix-assisted laser desorption/ionization (MALDI) imaging, we determined that the drug penetrates the entire MCTS. Proteomic analysis revealed a decrease in abundance of H3K27me3, compared to the untreated sample, as expected. Interestingly, we observed a comparable growth curve for MCTS under constant drug treatment over 13 days with those treated for only 4 days at the beginning of their growth. We thus demonstrate that MS-based proteomics can define significant differences in histone PTM patterns in submillimetric layers of three-dimensional (3D) cultures. Moreover, we show that our model is suitable for monitoring drug localization and regulation of histone PTMs after drug treatment.