New Insights into the Epigenetic Activities of Natural Compounds

Background: With their varied pharmacophores, natural products are interesting tools to open the drug discovery pipeline. Several plant secondary metabolites are components of the human diet and have reported epigenetic activities. In this study, we screened a small natural compound library for epigenetic activities. Methods: Seventy-one different natural products plus 17 controls collected from all collaborating laboratories were screened. Localized DNA methylation (DNAm) was studied on a stretch of the retinoic acid receptor gene RARβ. All genomic 5-methylated cytosine (5mC) bases were then detected by high performance liquid chromatography tandem mass spectrometry (HPLC-MS/MS). DNA methyl transferase 1 (DNMT1) enzymatic activity was measured for selected compounds. Level of histone H3 trimethylation at lysine 9 and 27 (me3H3K9 and me3H3K27) was measured by Western blot analysis. Global histone deacetylase inhibition (HDAC) was assayed first using a bioluminescent resonance energy transfer-based (BRET) assay and then with enzymatic fluorescence based-assays for most HDAC class 1. HDAC6 inhibition was measured by Western blot analysis. Sirtuin (Sirt2) inhibition was assessed first with a thermal shift assay and then using the enzymatic SIRTainty™ Class III HDAC assay for Sirt1 and Sirt2. Results: Diosmetin, (S)-equol, umbelliferone, papaverin and L-carnitine were identified as novel DNA demethylating agents. Emodin, rhein, aloin and D-glucuronic acid were identified as novel histone H3 demethylating compounds. Previously undescribed Sirt activation by apigenin, biochanin B, robinin, pinocembrin, aureusidine, brucine and boldine was also detected. Conclusions: High-throughput alpha screens are used for initial studies of diverse compound libraries; however, this approach has significant disadvantages for the study of DNAm. Indeed, finding unmethylated RARβ alleles in one cell line does not indicate the activity of the compound at the level of the entire genome over a given time-frame and a given dose. Measurement of DNMT1 activity is not useful since most natural compounds are not direct enzymatic inhibitors. When studying histone methylation, Western blot analysis is laborious but remains a cheap and effective assay under circumstances in which several histone methylases (KDMs) or demethylases may be responsible for modulation of histone methylation. Reversible epigenetic modifications of the genome remain feasible targets for nutrition-based preventive strategies. However, more accurate HDAC inhibition assays are still required for the evaluation of flavanols, which have fluorogenic properties that disturb classical fluorescence-based assays.