Development of High-Throughput Assays to Study Methylases, Demethylases and Deacetylases Targeting Histone H3K4, H3K27 and H3K36 Residues Mathieu Arcand, Mireille Caron, Julie Blouin, Claire Normand, Anne Labonté, Hendrick Plante, Lucille Beaudet & Jaime Padrós PerkinElmer, 1744 William St., Montreal, QC H3J 1R4, Canada 3 µM 120 min JMJD3 1 nM H3K27me3 50 nM 2OG 1 µM 45 min H3K36me3 LSD1 H3K4ac 2 nM HDAC1 300 nM H3K4me1 1 nM 2OG + NAD 200 nM H3K27ac 2 µM 30 min 500 µM 30 min N/A 3 nM 60 min N/A 45 min EZH2 150 ng/well H3 (21-44) 500 nM SAM 3 µM 180 min JMJD3 5 nM H3K27me3 200 nM 2OG 0.5 µM 120 min JMJD2A 1 nM H3K36me3 250 nM 2OG 5 µM 30 min Enzyme Titration & Time-Course 1,000,000 750,000 500,000 250,000 0 0 30 60 90 120 150,000 EX-527 IC50 = 0.68 µM Suramin IC50 = 0.26 µM 75,000 0 - -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 30 60 90 LANCE Signal (665nm) LANCE Ultra LANCE Signal (665 nm) 0 120 60,000 Nicotinamide IC50 = 180 µM 40,000 EX-527 IC50 = 0.67 µM 20,000 Suramin IC50 = 0.20 µM 0 - 1,500,000 [LSD1] (nM) 1,250,000 8 4 2 1 0 1,000,000 750,000 500,000 250,000 0 40 60 80 100 120 200,000 100,000 0 [LSD1] (nM) 5 4 3 2 1 0 150,000 100,000 50,000 0 40 60 80 Time (min) -5 -4 -3 -2 -1 100 120 175,000 Tranylcypromine IC50 = 74.3 µM 150,000 125,000 100,000 75,000 50,000 25,000 0 - -7 -6 -5 -4 -3 -2 -1 Log [Tranylcypromine] (M) LSD1 Assay Buffer: 50 mM Tris-HCl pH 9.0, 50 mM NaCl, 1 mM DTT and 0.01% Tween-20. SAHA EC50 = 399 nM 40,000 20,000 0 120 60,000 150 100 75 50 25 0 50,000 40,000 30,000 20,000 10,000 0 30 60 90 120 150 180 [EZH2] (ng/well) 50,000 200 150 100 75 0 40,000 30,000 20,000 10,000 0 60 90 120 150 400,000 300,000 200,000 100,000 0 80 100 120 [JMJD3] (nM) 10 30,000 5 1 2 0.5 0 20,000 10,000 0 40 60 80 Time (min) 120 0 - -8 -7 100,000 2 1 0.5 0.25 0.125 0 80,000 60,000 40,000 20,000 0 20 40 60 80 -6 -5 -4 Log [2,4-PDCA] (M) 100 50,000 2,4-PDCA IC50 = 1.5 µM 40,000 30,000 20,000 10,000 0 120 - -8 -7 -6 -5 -4 Log [2,4-PDCA] (M) Assay Robustness AlphaLISA LSD1 assay 20,000 Sinefungin IC50 = 420 M 10,000 0 - -6 -5 -4 -3 -2 -1 30,000 25,000 No inhibitor 600,000 500,000 400,000 Z' = 0.80 S/B = 643 300,000 200,000 100,000 3 mM Tranylcypromine 0 0 10 20 30 40 50 AlphaLISA LANCE Ultra S/B Z' S/B Z' SIRT1 384 0.77 15 0.75 LSD1 643 0.80 20 0.88 HDAC1 2.9 0.69 3.6 0.80 EZH2 52 0.71 4.6 0.80 JMJD3 69 0.70 7.4 0.75 JMJD2A 144 0.85 7.3 0.74 Well # 15,000 10,000 Enzyme 700,000 20,000 Sinefungin IC50 = 780 M 5,000 0 -6 -5 -4 -3 -2 -1 Z’-factor values and signal to background ratios (S/B) were determined for each AlphaLISA and LANCE Ultra optimized epigenetic assay by analyzing 48 assay wells for both total and inhibited signals. Calculated Z’-factor values were ≥ 0.69 and remained stable after overnight incubation (not shown). 12 Inhibitor Titration 100,000 2,4-PDCA IC50 = 6.9 µM 80,000 Signal increase assays were developed for three demethylases (LSD1, JMJD2A and JMJD3) and the SIRT1 deacetylase taking advantage of antibody specificity for unmodified (H3K4) or dimethylated (H3K27me2 and H3K36me2) residues. 60,000 40,000 20,000 0 100 120 - -8 -7 -6 -5 -4 -3 -2 50,000 2,4-PDCA IC50 = 9.2 µM 40,000 The AlphaLISA and LANCE Ultra HDAC1 signal decrease assays showed a robust Z’-factor value (0.69 for AlphaLISA and 0.80 for LANCE Ultra), despite of S/B ratios <4. IC50 values for known inhibitors and rank order of potency were as expected from the literature with either detection technology. 30,000 20,000 IC50 and Z’-factor values remained stable after overnight incubation, allowing both online and offline HTS plate reading. 10,000 0 Summary AlphaLISA Acceptor beads and LANCE Ultra europium-labeled antimark antibodies were used for the successful optimization of robust and sensitive epigenetic assays using histone H3-derived peptides as substrates. Log [2,4-PDCA] (M) 40,000 20 100 [JMJD2A] (nM) 11 30,000 Time (min) 0 80 50,000 Log [Sinefungin] (M) 5 2 1 0.5 0.25 0 60 60 100,000 40,000 - -7 [JMJD3] (nM) 40 40 120,000 0 Inhibitor Titration 180 500,000 20 20 150,000 JMJD2A Assay Buffer: 50 mM HEPES pH 7.5, 0.01% Tween-20 and 0.1 % BSA. Note: 5 µM Fe(II) and 100 µM ascorbate were added to enzymatic reactions along with the biotinylated Histone H3K36me3 peptide substrate. H3K27me3 Demethylation by JMJD3 0 0 2,4-PDCA IC50 = 1.7 µM Time (min) Log [Sinefungin] (M) 60,000 30 0 Log [Inhibitors] (M) [EZH2] (ng/well) 0 200,000 - -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 Enzyme Titration & Time-Course 300,000 -6 100 70,000 9 400,000 - -7 80 60,000 EZH2 Assay Buffer: 50 mM Tris-HCl pH 9.0, 50 mM NaCl, 1 mM DTT, 0.01% Tween-20 and 0.01% BSA. Tranylcypromine IC50 = 31.9 µM 500,000 60 Trichostatin A EC50 = 4.5 nM H3 (21-44) Methylation by EZH2 Log [Tranylcypromine] (M) 200,000 20 0 40 400,000 200,000 Time (min) 80,000 Time (min) 600,000 Time (min) 0 20,000 0 Inhibitor Titration AlphaLISA Signal (counts) Enzyme Titration & Time-Course 20 40,000 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 H3K4me1 Demethylation by LSD1 0 60,000 20 600,000 Log [Inhibitors] (M) 0 0.25 0.5 1 2 3 Log [Inhibitors] (M) LANCE Signal (665 nm) Detection reagents, consumables and instrument. The anti-mark AlphaLISA Acceptor beads, Alpha Streptavidin Donor beads, AlphaLISA 5X Epigenetics Buffer 1 kit, LANCE Ultra europium-labeled anti-mark antibodies, ULight-Streptavidin, 10X LANCE Detection Buffer, white opaque 384-well OptiPlate™ microtiter plates, TopSeal™-A film and EnVision® Multilabel Plate Reader were all from PerkinElmer. 50,000 6 AlphaLISA Signal (counts) Reagents and inhibitors. S-(5’-adenosyl)-L-methionine chloride (SAM), αketoglutaric acid potassium salt (2OG), (+) sodium L-ascorbate, ammonium iron(II) sulfate hexahydrate (Fe(II)), nicotinamide adenine dinucleotide (NAD+), trans-2-phenylcyclopropylamine (tranylcypromine), trichostatin A, sinefungin, 2,4-pyridinedicarboxylic acid (2,4-PDCA) and nicotinamide were from SigmaAldrich. Ethylenediaminetetraacetic acid (EDTA) was obtained from Invitrogen, suberoylanilide hydroxamic acid (SAHA) from Cayman Chemical, and suramin from EMD Chemicals. 100,000 80,000 SIRT1 Assay Buffer: 50 mM Tris-HCl pH 8.0, 150 mM NaCl, 1 mM DTT, 0.01% Tween-20 and 0.01% BSA. LANCE Signal (665 nm) Enzymes and biotinylated peptide substrates. Recombinant enzymes LSD1, SIRT1, JMJD2A, JMJD3 and the EZH2/EED/SUZ12/RbAp48/AEBP2 protein complex were from BPS Bioscience. HDAC1 was obtained from Cayman Chemical. Biotinylated peptides were from AnaSpec. 2.5 1 0.5 0.25 0.1 0 - -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 Time (min) 100,000 Time (min) AlphaLISA Materials 150,000 0 LANCE Ultra 3 [SIRT1] (nM) 120 [HDAC1] (nM) Log [Inhibitors] (M) 200,000 100 Enzyme Titration & Time-Course 300,000 225,000 80 80,000 8 375,000 Nicotinamide IC50 = 260 µM 60 HDAC1 Assay Buffer: 50 mM Tris-HCl pH 8.0, 0.1 mM EDTA, 1 mM DTT, 0.01% Tween-20 and 0.01% BSA. AlphaLISA [SIRT1] (nM) 5 2.5 1 0.5 0.25 0 1,250,000 40 0 Time (min) Inhibitor Titration 1,500,000 20 0 H3K4ac Deacetylation by SIRT1 Time (min) In AlphaLISA and LANCE Ultra epigenetics proximity assays, biotinylated histone H3-derived peptide substrates are incubated in enzymatic reactions in the presence of the required cofactors (see Methods). Detection of reaction products occurs via mark-specific antibodies coupled to Acceptor beads (AlphaLISA) or labeled with europium chelate (LANCE Ultra). The biotin moiety of the histone H3-derived peptide substrates is captured by streptavidin coupled to a Donor bead (AlphaLISA) or labeled with the ULight™ acceptor dye (LANCE Ultra). In both assay technologies, irradiation of the captured reaction products triggers an energy transfer leading to light emission proportional to the enzyme activity. Anti-H3K4 (unmodified) Anti-H3K4 (unmodified) Anti-H3K27ac AntiH3K27me2-1 AntiH3K27me2-1 AntiH3K36me2 0 LANCE Signal (665 nm) 0.5 nM 100 nM 0 200,000 Time (min) AlphaLISA Signal (counts) 0.5 nM 200,000 [JMJD2A] (nM) 1 0.5 0.25 0.125 0 AlphaLISA Signal (counts) SAM Inhibitor Titration LANCE Signal (665 nm) 100 nM AlphaLISA Signal (counts) H3 (21-44) 400,000 800,000 LANCE Signal (665 nm) 150 ng/well 600,000 SAHA EC50 = 161 nM AlphaLISA EZH2 400,000 800,000 Trichostatin A EC50 = 3.2 nM LANCE Ultra 30 min 1,000,000 H3K36me3 Demethylation by JMJD2A Enzyme Titration & Time-Course AlphaLISA Signal (counts) N/A AlphaLISA Signal (counts) 3 nM LANCE Signal (665 nm) H3K27ac AlphaLISA Signal (counts) 1 nM 600,000 LANCE Signal (665 nm) HDAC1 800,000 AlphaLISA Signal (counts) 60 min 0 0.125 0.25 0.5 1 2 Inhibitor Titration LANCE Signal (665nm) N/A AlphaLISA Signal (counts) 80 nM [HDAC1] (nM) 1,000,000 LANCE Signal (665 nm) H3K4me1 1,200,000 AlphaLISA Signal (counts) 30 min AlphaLISA 800 µM LANCE Ultra AlphaLISA NAD+ 2 nM SIRT1 LANCE Ultra 200 nM LSD1 JMJD2A 5 H3K4ac AlphaLISA LANCE Ultra 1 nM Anti-H3K4 (unmodified) Anti-H3K4 (unmodified) Anti-H3K27ac AntiH3K27me2-1 AntiH3K27me2-1 AntiH3K36me2 10 H3K27ac Deacetylation by HDAC1 Enzyme Titration & Time-Course LANCE Ultra AlphaLISA SIRT1 AlphaLISA Signal (counts) Assay Principles Enzyme AlphaLISA 2 OPTIMIZED ASSAY CONDITIONS Reaction Detection Substrate Cofactor time reagent LANCE Ultra Several assay methods have been developed for quantifying the activity of histone deacetylases (HDACs and sirtuins), histone methyltransferases (HMTs), and histone demethylases (HDMs). These include radioactive assays, enzymelinked immunoassays (ELISA), mass spectrometry, and enzyme-coupled detection of fluorescent peptides or reaction co-products (e.g. S-adenosylhomocysteine, formaldehyde, hydrogen peroxide). These assays suffer from various drawbacks such as low throughput, lack of sensitivity, generation of hazardous waste, requirement for expensive equipment, or artifacts associated with the use of nonphysiological fluorescent moieties or enzyme-coupled assays (generation of false positives/negatives). In this study, we describe the development and optimization of homogeneous antibody-based assays for measuring the catalytic activity of a series of epigenetic lysine-modifying enzymes acting on histone H3 Lys4 (SIRT1 deacetylase and LSD1 demethylase), Lys27 (HDAC1 deacetylase, EZH2 methyltransferase and JMJD3 demethylase) and Lys36 (JMJD2A demethylase). Two different non-radioactive, no-wash technologies were used for detection of the enzymatic reaction products: amplified luminescent proximity homogeneous (AlphaLISA®) assay and time-resolved Förster energy transfer (LANCE® Ultra) assay. Results demonstrated that all assays were sensitive, rapid and robust (Z’ factors ≥ 0.69), requiring only nanomolar concentrations of enzyme and peptide. Furthermore, profiling of known inhibitors for each epigenetic enzyme showed the expected potency with either technology. These assays will therefore be ideal for the identification of selective small molecule inhibitors. The approach described here is broadly suitable for measuring the catalytic activity of other histonemodifying enzymes by combining the appropriate biotinylated histone-derived peptides and mark-selective antibodies. 7 Methods LANCE Signal (665 nm) 4 Abstract AlphaLISA Signal (counts) 1 - -8 -7 -6 -5 -4 -3 -2 Log [2,4-PDCA] (M) JMJD3 Assay Buffer: 50 mM HEPES pH 7.5, 0.01% Tween-20 and 0.01 % BSA. Note: 5 µM Fe(II) and 100 µM ascorbate were added to enzymatic reactions along with the biotinylated histone H3K27me3 peptide substrate. A comprehensive description of these assays and their optimization is available on our website at www.perkinelmer.com/epigenetics. PerkinElmer, Inc., 940 Winter Street, Waltham, MA USA (800) 762-4000 or (+1) 203 925-4602 www.perkinelmer.com