Fluorescence-Based Methods for Screening Writers and Readers of Histone Methyl Marks

SAHH Assay

In this assay, SAHH and ADA convert SAH to homocysteine and inosine. The abundance of homocysteine was quantified using ThioGlo, which reacts with thiol groups and fluoresces strongly. To obtain the K_m and k_cat values for SAH, a SAH concentration ranging from 10 nM to 50 µM in a series of experiments was mixed with ADA (final 0.5 U/mg) in 25 mM potassium phosphate (pH 8), 1 mM EDTA, 2 mM MgCl₂, 0.01% Triton X-100, ThioGlo at a final concentration of 15 µM, and SAHH at a final concentration of 1 µM in a 20-µL reaction mix. Then, 9 µL of reaction mix, in duplicate, was transferred into each well of a 384-well plate (Axygen). The conversion of SAH to homocysteine was followed by monitoring the increase in fluorescence using a ViewLux (PerkinElmer, Waltham, MA) plate reader with 340 DUG 11 filter for excitation and 540/25 filter for emission for 20 min in a 384-well plate format. Homocysteine generated in the assay was quantified using standard curves. Activity values were corrected by subtracting background caused by SAH. K_m values were calculated using a hyperbolic function (Sigma Plot software, Systat Software, Chicago, IL). Standard deviations were calculated from two or more independent experiments.

HMT Assay

The methyltransferase activities of HMTs were measured using a coupled assay. ⁴⁹ In this assay, SAHH and ADA convert the methyltransferase reaction product SAH to homocysteine and inosine. The abundance of homocysteine was quantified using ThioGlo. For the K_m and k_cat determination, each protein concentration was selected to be within the linear range. For G9a, EHMT1, SETD7, SETD8, SUV39H2, and PRMT3, the concentrations used were 50 nM, 100 nM, 500 nM, 200 nM, 100 nM, and 500 nM, respectively. The rate of product formation for these HMTs was constant, and the plot of signal versus time was linear over the assay period (Suppl. Fig. S2). No lag at the start of the reactions was observed. The peptides used in this assay varied for each enzyme. Thus, for G9a, EHMT1, SUV39H2, and SETD7, the substrate is the first 25 residues of histone 3 (H3 [1–25], ARTKQTARKSTGGKAPRKQ LATKAA). The substrate for PRMT3 and SETD8 is the first 24 residues of histone 4 (H4 [1–24], SGRGKGGKGLG KGGAKRHRKVLRD). For the determination of K_m values of SAM and peptide, peptide or SAM was titrated into the assay mixture containing enzyme, 5 µM SAHH, about 0.5 U/mL ADA, and 15 µM ThioGlo in 25 mM potassium phosphate (pH 8), 1 mM EDTA, 2 mM MgCl₂, and 0.01% Triton X-100. A saturating concentration of the other substrate (fixed concentration of SAM or peptide) was added to start a reaction. K_m values were calculated using Sigma Plot (Systat Software). The methylation reaction was followed by monitoring the increase in fluorescence using a ViewLux (PerkinElmer) plate reader with 340 DUG 11 filter for excitation and 540/25 filter for emission or Synergy 2 microplate reader (BioTek, Winooski, VT) with a 360/40-nm excitation filter and 528/20-nm emission filter for 20 min in a 384-well plate format. Homocysteine generated in the assay was quantified using standard curves. Activity values were corrected by subtracting background caused by the peptide and the protein. ThioGlo has an inhibitory effect on SETD8 activity due to a reactive cysteine residue near the active site. Therefore, this enzyme was assayed using the end point assay. Reactions mixtures (all components of the assay) were incubated for 4 min before adding ThioGlo at a final concentration of 15 µM and quantifying the amount of produced SAH. For assays performed over a longer period of time, a decrease in protein concentration may be required to maintain linearity of the reaction.

Z Factor Determination

Z factors were determined to evaluate the suitability of each screen in a 384-well format. ⁶³ To calculate the Z factor of the assay, the HMT reactions (for each enzyme separately; see above for details) were run in eight complete rows (half-plate) of 16 × 24 of a 384-well plate; the other eight rows were used for background controls (the reaction mix containing protein but not SAM). The reaction volume for G9a, EHMT1, SUV39H2, and PRMT3 was 10 µL. For SETD7 and SETD8, the reaction volume was increased to 20 µL to obtain a better signal-to-noise ratio. To obtain a Z′ factor value for the assay, a known inhibitor (UNC-0638 in case of G9a) was added at 1 µM to the reaction mix. The Z factor for each protein was calculated as described by Zhang et al. ⁶³

Fluorescence Polarization

H3K27me3 (QLATKAARK(me3)SAPATG) peptide for EED and MLL_Win (GSARAEVHLRKS) peptide for WDR5 were synthesized, N-terminal-labeled with isothiocyanate–fluorescein, and purified by Tufts University Core Services (Boston, MA). Compound binding assays were performed in a 10-µL volume at a constant labeled peptide concentration of 40 nM and protein concentrations of 80 µM for EED and 5 µM for WDR5. For EED, the buffer contained 20 mM Tris (pH 8.0), 50 mM NaCl, 2 mM dithiothreitol (DTT), and 0.01% Triton X-100. For WDR5, the buffer was 100 mM potassium phosphate (pH 8.0), 150 mM NaCl, and 0.01% Triton X-100. Fluorescence polarization assays were performed in 384-well Axygen plates using a Synergy 2 microplate reader (BioTek). The excitation wavelength of 485 nm and the emission wavelength of 528 nm were used. To determine K_d values, the data were fit to a hyperbolic function using Sigma Plot software (Systat Software). The K_d values represent the average and standard deviation of three measurements.

Assay and Screening Writers of Histone Methyl Markers

The SAHH-coupled assay developed by Collazo et al. ⁴⁹ is a convenient method to assay a wide range of HMTs. In this assay, the HMTase activity is coupled with SAHH activity. This coupled assay relies on SAHH using SAH, the product of the HMTase reaction as a substrate and producing L-homocysteine and adenosine. The free SH group of homocysteine then reacts with ThioGlo, and the fluorescence signal will increase ( Fig. 1 ). Such an increase in fluorescence signal reflects HMTase activity, and the amount of produced SAH can be quantified. In this study, we have optimized this assay in a 384-well format as a continuous assay and determined the kinetic parameters for G9a, EHMT1, SUV39H2, SETD7, and PRMT3. SETD8, however, was not amenable to the continuous assay and was characterized using the end point assay.

OPEN IN VIEWER

Figure 1.

S-adenosylhomocysteine hydrolase (SAHH) assay using fluorescence detection. L-homocysteine, the product of SAHH function, reacts with ThioGlo, and emission at 513 nm (excitation of 384 nm) can be monitored to detect S-adenosylhomocysteine (SAH) production and therefore SAHH activity. Presence of adenosine deaminase takes away adenosine, the by-product of the SAHH reaction, by converting it to inosine to prevent possible reverse reaction

Assay Optimization

Characterization of SAHH at 28 °C from Sulfolobus solfataricus (SsSAHH) ⁴⁹ with a K_m of 1.8 ± 0.1 (µM) for SAH and k_cat of 20 ± 0.3 h⁻¹ and Aeropyrum pernix (ApSAHH) with a K_m of 1.2 ± 0.1 (µM) and k_cat of 21 ± 0.1 h⁻¹ ( Table 1 ) indicated that these two proteins have more or less similar kinetic parameters, with ApSAHH having slightly higher specific activity (18 h⁻¹ µM⁻¹) compared to that for SsSAHH (11 h⁻¹ µM⁻¹). However, we selected SsSAHH for assay optimization because all preliminary experiments were previously performed using this SAHH.

OPEN IN VIEWER

Table 1.

Characterization of SsSAHH and ApSAHH

Protein	Source	Km(µM)	kcat(h−1)	kcat/Km (h−1 µM−1)
SsSAHH	Sulfolobus solfataricus	1.8 ± 0.1	20 ± 0.3	11
ApSAHH	Aeropyrum pernix	1.2 ± 0.1	21 ± 0.1	18

Collazo et al. ⁴⁹ used an end point assay to determine kinetic parameters of HMTs by initiating the reaction with addition of a saturating concentration of SAM in a large volume and aliquoting the reaction mix and stopping the reaction at evenly spaced time intervals starting at time zero by cold isopropanol. After the reaction was completed, a solution of ThioGlo in DMSO was added to each well and allowed to react for 10 min at room temperature, and fluorescence was then measured. To determine the quantity of homocysteine generated in the assay, reduced CoA was used to generate standard curves.

Here we modified the reaction so that we can monitor the production of SAH (through detecting homocysteine) continuously. A reaction mix was prepared with all components of the assay, including protein and SAM but not peptide. Freshly prepared ThioGlo was added to this reaction mix along with SAHH and ADA. The reaction was started by adding the respective peptide substrate for each protein, and the increase in SAH production was monitored by recording the emission at 528 nm (excitation at 360 nm) using Synergy2, Synergy4, or ViewLux plate readers. It was crucial to prepare the ThioGlo solution freshly each time to obtain reproducible data. The ThioGlo should be stored in 100% DMSO and then added to buffer assays, which are slightly alkaline (pH 8). To ensure that the SAHH reaction is not a rate-limiting step and SAH conversion to homocysteine is properly monitored, we performed the assay for G9a, the most active HMT in our set, at saturating concentrations of SAM and H3 (1–25) peptide and varying the SsSAHH concentration ( Fig. 2A ). The fluorescence signal increased by increasing the SAHH up to 4 µM and plateaued completely at a 5-µM concentration. To ensure that conversion by SAHH is not a rate-limiting step, we also tested the activity of the HMTs using 5 and 10 µM SAHH. Doubling the concentration of SAHH made no significant difference (Suppl. Fig. S3). Therefore, we selected the 5-µM concentrations of SAHH for all following experiments in our HMT set. Another component of the coupled assay to consider was ADA. This enzyme was originally included in this assay to take away adenosine, the by-product of the SAHH reaction ( Fig. 1 ), to reduce the possibility of reverse reaction and SAH production upon SAHH-catalyzed condensation of adenosine and homocysteine. ⁴⁹ However, it was later reported that ADA may not be required in the plant caffeic acid O-methyltransferases assay. ⁵³ We also performed the SAHH-coupled assay for G9a at 5 µM SsSAHH and varying the concentration of ADA ( Fig. 2B ). No significant effect of ADA concentration on activity of SAHH was observed. However, due to a small increase in the level of activity at low concentrations of ADA ( Fig. 2B ), we selected to use 5 µM SAHH and 0.5 U/mL ADA in all HMT assays. The concentration of HMTs in the assay must be kept low enough so that all produced SAH is simultaneously converted to homocysteine and subsequently reacts with ThioGlo. The SAHH reaction should not be a rate-limiting step at any point in the continuous assay. Therefore, we performed all the assays at 5 µM SAHH and around 50, 100, 100, 500, and 500 nM of G9a, EHMT1, SUV39H2, SETD7, and PRMT3, respectively.

OPEN IN VIEWER

Figure 2.

Effect of S-adenosylhomocysteine hydrolase (SAHH) and adenosine deaminase (ADA) on histone methyltransferase (HMT) assay readout. G9a activity was assayed at 0.5 U/mL ADA and different concentrations of SAHH (A) and at 5 µM SAHH and different concentrations of ADA (B). The buffer was 25 mM potassium phosphate (pH 8), 1 mM EDTA, 2 mM MgCl₂, and 0.01% Triton X-100. Reaction mix contained saturating concentrations of S-adenosyl-L-methionine (SAM; 10 µM) and peptide (10 µM) substrate (H3 [1–25]). Experiments were performed at 28 °C

Determining the Kinetic Parameters for HMTs

We employed this continuous assay to characterize G9a, EMHT1, SUV39H2, SETD7, PRMT3, and SETD8. However, due to interaction of ThioGlo with a free cysteine thiol in SETD8 and a clearly related inhibitory effect, we excluded SETD8 from this set and determined its kinetic parameters using the end point assay. ⁴⁹ Kinetic parameters for all these HMTs are summarized in Table 2 . G9a, EHMT1, and SUV39H2 methylate lysine 9 (H3K9), and SETD7 methylates lysine 4 (H3K4) of histone H3. We performed all kinetic assays for these four HMTs using a 25-mer peptide corresponding to the first 25 residues of histone H3 (H3 1–25) as the substrate. SETD8, on the other hand, methylates lysine 20 of histone H4, and PRMT3, being an arginine methyltransferase, methylates arginine 3 of histone H4 (H4R3). A peptide corresponding to the first 24 residues of histone H4 was used as a substrate for these two proteins. To facilitate purification of all these HMTs, we used shorter active constructs of these proteins ( Table 2 ). For the three H3K9 HMTs, the K_m values for the peptide substrate (H3 1–25) are more or less in the same range (1.7–3.4 µM). However, the size of the peptide substrate significantly can affect the K_m of the protein for the peptide. For example, with G9a, the K_m for H3 (1–11) was determined to be 50 ± 7 µM, almost 30 times higher than that for H3 (1–25). Among these HMTs, G9a showed the highest specific activity (450 h⁻¹ µM⁻¹). Reproducibility of determining the kinetic parameters for these proteins confirmed the suitability of this assay for characterizing most HMTs.

OPEN IN VIEWER

Table 2.

Kinetic Parameters of HMTs

				Km (µM)
Protein	Construct	Number of Amino Acids	Peptide Substrate	Peptide	SAM	kcat (h−1)	kcat/Km (h−1 µM−1)
G9a	913–1193	281	H3 (1–25)	1.7 ± 0.1	1.5 ± 0.1	765 ± 8	450
EHMT1	951–1235	253	H3 (1–25)	3.4 ± 0.1	1.2 ± 0.2	465 ± 30	136
SUV39H2	52–350	299	H3 (1–25)	3 ± 0.1	8 ± 0.1	660 ± 5	220
SETD7	1–366	366	H3 (1–25)	160 ± 40	0.6 ± 0.2	25 ± 2	0.15
SETD8	186–352	167	H4 (1–24)	74 ± 14	38 ± 6	960 ± 40	12
PRMT3	211–531	230	H4 (1–24)	13 ± 3	25 ± 5	50 ± 4	4

HMT, histone methyltransferase; SAM, S-adenosyl-L-methionine.

Among H3K9 methyltransferases studied here, G9a and EHMT1 mono- and dimethylate lysine 9. SUV39H2, however, also trimethylates H3K9. ⁶² Our data indicate that the K_m of the peptide increases for higher methylation states (un- < mono- < dimethylated H3 [1–25]; Table 3 ). With k_cat values slightly increasing for higher methylation states, SUV39H2 shows much higher specific activity with unmethylated peptide.

OPEN IN VIEWER

Table 3.

Substrate Specificity of SUV39H2

		Km (µM)
Protein	Peptide Substrate	Peptide	SAM	kcat (h−1)	kcat/Km (h−1 µM−1)
SUV39H2	H3 (1–25)	3 ± 0.1	8 ± 0.1	660 ± 5	220
SUV39H2	H3 (1–25) K9Me1	12 ± 2	ND	860 ± 55	71
SUV39H2	H3 (1–25) K9Me2	60 ± 8	ND	940 ± 110	15

SAM, S-adenosyl-L-methionine; ND, not determined.

HTS Using Continuous Assay

HMTs methylate histones and affect chromatin condensation and gene expression. HMTs have been implicated in many diseases, including various types of cancer.^1,34,64 Identifying potent inhibitors of HMTs with in vivo activity could be very useful to further investigate such effects and may provide the basis for drug discovery. Two sets of potential inhibitors can be identified for HMTs: (1) peptide mimetic compounds that bind to the peptide binding site and therefore compete with peptide such as the inhibitors reported for G9a^50–52,56 and (2) compounds that may bind to SAM binding site and compete with SAM. To identify peptide competitive inhibitors, one needs to screen HMTs against libraries of compounds at low concentrations of peptide substrates while keeping the concentration of SAM close to saturation. Typically, screening at a peptide concentration close to the K_m of peptide and a SAM concentration about five times the K_m of SAM should provide an opportunity for compounds to compete with the peptide for binding to the peptide binding site and make hit detection easier. We assessed such screening in a 384-well format for all five HMTs in this study, and Z factors of 0.64, 0.76, 0.77, 0.84, and 0.84 were obtained for SETD7, G9a, EHMT1, SUV39H2, and PRMT3, respectively, indicating that this continuous assay is very suitable for HTS (Suppl. Table S1 and Fig. 3A ). We employed this method to screen SUV39H2 and PRMT3 against a library of thousands of diverse compounds, and we identified hits with low micromolar IC₅₀ values (e.g., PDB ID: 3SMQ). A Z factor of 0.84 was also determined for screening SETD8 at a K_m of peptide using the end point SAHH-coupled assay.

OPEN IN VIEWER

Figure 3.

Screening G9a in a 384-well format. G9a was screened at peptide concentration close to the K_m for peptide (2 µM) and saturation concentration (10 µM) of S-adenosyl-L-methionine (SAM; A) and SAM concentration close to K_m of SAM (2 µM) and saturation concentration (10 µM) of H3 (1–25) peptide (B) to determine Z factors. G9a was also screened in the presence (185 samples) and absence (185 samples) of UNC0638 (a potent G9a inhibitor) at conditions similar to (A), and a Z′ factor of 0.71 was obtained

On the other hand, all these HMTs use SAM as a cofactor, and identifying compounds that can compete with SAM, yet are selective enough among HMTs, may seem more difficult. Through a systematic analysis of SAM binding site across structures of 14 human HMTs or close homologues, it was shown that SAM binding pockets are likely diverse enough to be chemically tractable, and identifying selective SAM competitive inhibitors is possible. ⁹ To identify SAM competitive inhibitors, we also assessed screening HMTs at the K_m of SAM and about five times the K_m of peptides. Generally, at this condition, less variation was observed and better Z factors were obtained (Suppl. Table S1 and Fig. 3B ). Alternatively, if one is not looking for a specific type of inhibitor, screening HMTs at SAM and peptide concentrations close to respected K_m values will allow detection of any type of inhibitors. Further kinetic analysis could determine the mechanism of inhibition. However, by lowering both SAM and peptide concentrations, the signal and reproducibility may decrease, resulting in a high number of false positives.

To see how reproducibly a potent inhibitor can be identified through such screening, G9a was screened in the presence and absence of UNC0638. ⁵⁰ Screening was reproducible in a 384-well format with a Z′ factor of 0.71 ( Fig. 3C ). Our results show that this assay can reliably be used to screen HMTs against libraries of a large number of compounds. However, in such screening, compounds that inhibit SAHH activity or quenching the fluorescence signal may also be identified as HMT inhibitors (screening hits). Therefore, the screening process should include a counterscreen to eliminate false positives. The SAHH activity assay (using SAH as a substrate for SAHH) is a suitable counterscreen that includes all assay components. HMT screening hits that also reduce the signal in this counterscreen are false positives. Although a continuous SAHH-coupled assay is perfectly suitable for characterization of HMTs and HTS, confirmation of a high number of screening hits by other high-throughput orthogonal methods, such as recently developed chemiluminescence, ⁵⁷ microfluidic capillary electrophoresis, ⁵⁹ or radioactivity-based assays, may be more convenient before proceeding with further characterization using low-throughput methods (Isothermal titration calorimetry; ITC, etc.) or follow-up chemistry. The SAHH-coupled assay is an easy to optimize assay for screening other HMTs.

Screening Histone Methyl Marks Using Peptide Displacement

Enzymatic activity of some methyltransferases such as MLL and EZH2 depends on their interaction with other proteins. MLL catalyzes histone H3 lysine 4 methylation, and its enzymatic activity depends on interaction with a three-subunit complex composed of WDR5, RbBP5, and Ash2L. ⁶⁵ PRC2 complex includes EED and SUZ12 components, which interact with EZH2. This complex methylates histone H3K27 and mediates gene silencing. PRC2 components are overexpressed in metastatic prostate cancer. ⁶⁶ As these components seem to be essential for methyltransferase activity of these complexes, we developed an assay to screen the reader components of these methyltransferase complexes.

WDR5 binds to MLL through recognition of a specific motif (WIN motif) of MLL, which includes an arginine residue (3765), which binds to the same arginine binding pocket on WDR5 that has been shown to bind histone H3. ⁴⁰ Any compound that interrupts interaction of MLL and WDR5 may potentially inhibit MLL activity. We used a fluorescein-labeled peptide (F-peptide) (GSARAEVHLRKS) corresponding to MLL WIN motif to create a fluorescence-polarized signal ( Fig. 4A ). F-peptide (40 nM) was almost saturated at 5 µM of WDR5, with a K_d value of 1.85 ± 0.35 µM, and created a strong enough signal to assess binding of possible ligands. To test if displacement of the F-peptide can be detected and monitored upon ligand binding, we used unlabeled WIN peptide to displace the labeled WIN peptide ( Fig. 4B ). F-labeled peptide was displaced by unlabeled peptide with a K_displacement (K_dis) value of 7 ± 3.4 µM. Because interaction of WDR5 with WIN peptide was strong enough to generate a reliable signal-to-noise ratio at low WDR5 concentration (5 µM), we proceeded with testing the possibility of screening WDR5 in a 384-well format. Fluorescence polarization (FP) signal was measured in the presence of F-peptide only, in 176 wells, and by addition of 25 µM of unlabeled peptide to 176 other sample wells at 5 µM of WDR5 ( Fig. 4C ). Displacement of F-peptide by unlabeled WIN peptide was clear and detectable with a Z′ factor of 0.6. We applied this method to screen WDR5 against the above-mentioned diverse library of compounds and identified hits with K_dis of 1 to 10 µM. Binding of these hits was also confirmed by ITC (e.g., PDB ID: 3SMR).

OPEN IN VIEWER

Figure 4.

Screening WDR5 for ligand binding by monitoring peptide displacement. Binding of a fluorescein-labeled WIN peptide (GSARAEVHLRKS; 40 nM) to an increasing amount of WDR5 generated a fluorescence-polarized (FP) signal (A). Fluorescein-labeled peptide was displaced by increasing the amount of unlabeled WIN peptide (B), resulting in a decrease in FP signal. FP signal was also measured in the presence (25 µM; 176 samples) and absence (176 samples) of unlabeled peptide in a 384-well format using 5 µM WDR5 (C), and a Z′ factor of 0.6 was determined. Experiments were performed in triplicate

EED is the reader component of the EZH2 complex and binds to H3K27me3 and may be responsible for spreading this mark along chromatin. ⁶⁷ Here we used a similar approach to assess ligand binding to EED. Binding of a fluorescein-labeled H3K27me3 peptide (QLATKAARKme3SAPATG; 40 nM) to an increasing amount of EED generated an FP signal ( Fig. 5A ). Fluorescein-labeled peptide was then displaced by an increasing amount of unlabeled H3K27me3 peptide ( Fig. 5B ), resulting in an FP signal decrease. Unlabeled H3K27me3 peptide displaced the labeled peptide with K_dis of 150 ± 25 µM. Although saturation was not reached up to 100 µM EED, using a 384-well format, we were able to screen EED using peptide displacement with a Z′ factor of 0.5 ( Fig. 5C ).

OPEN IN VIEWER

Figure 5.

Screening EED for ligand binding by monitoring peptide displacement. Binding of a fluorescein-labeled H3K27me3 peptide (QLATKAARKme3SAPATG; 40 nM) to an increasing amount of EED generated a fluorescence-polarized (FP) signal (A). Fluorescein-labeled peptide was displaced by an increasing amount of unlabeled H3K27me3 peptide (B), resulting in a decrease in FP signal. FP signal was measured in the presence (1 mM; 185 samples) and absence (185 samples) of unlabeled peptide in a 384-well format using 80 µM EED (C), and a Z′ factor of 0.5 was determined. Experiments were performed in triplicate

The kinetic parameters previously reported for the proteins characterized in this study are listed in the Supplementary Table S2.^43,68–74 These kinetic parameters were determined by variety of methods (radioactivity, fluorescence, mass spectrometry–based enzyme assays, etc.), each with certain assay limitations. Although the kinetic values reported here ( Table 2 ) are reasonably close to those reported before (Suppl. Table S2), in most cases, using different protein constructs (size and source, etc.), substrates (peptides with different length and sequence or histones as substrate), tags used to express and purify these proteins, and different assay conditions (pH, salt, temperature, etc.) often makes direct comparison of kinetic values quite difficult. For example, we observed significant differences in kinetic parameters for G9a when a different length of histone peptide was used.

In conclusion, the SAHH-coupled assay in a continuous mode is a reliable method for characterizing and screening G9a, EHMT1, SUV39H2, SETD7, and PRMT3 and probably many other HMTs. However, the amenability of other HMTs to this assay needs to be assessed before using this method for their characterization or screening as proteins such as SETD8 may react with ThioGlo and lose activity. In this report, we provided the kinetic parameters and optimized conditions for screening these HMTs. These kinetic values may vary for peptides with different amino acid sequences, and using nucleosomes as a substrate may produce significantly different values depending on the mechanism of substrate recognition. However, using peptides as substrates for HTS is often more convenient. Optimized screening conditions for WDR5 and EED also present a reliable method to set up screening conditions for other readers of histone methyl marks and peptide binding proteins. Overall, these assays are reliable screening methods and relatively easy to optimize for other writers and readers of histone methyl marks.