Sage Journals: Discover world-class research

Abstract

Histone deacetylase (HDAC) enzymes modify the acetylation state of histones and other important proteins. Aberrant HDAC enzyme function has been implicated in many diseases, and the discovery and development of drugs targeting these enzymes is becoming increasingly important. In this article, the authors report the evaluation of homogeneous, single-addition, bioluminogenic HDAC enzyme activity assays that offer less assay interference by compounds in comparison to fluorescence-based formats. The authors assessed the key operational assay properties including sensitivity, scalability, reproducibility, signal stability, robustness (Z′), DMSO tolerance, and pharmacological response to standard inhibitors against HDAC-1, HDAC-3/NcoR2, HDAC-6, and SIRT-1 enzymes. These assays were successfully miniaturized to a 10 µL assay volume, and their suitability for high-throughput screening was tested in validation experiments using 640 drugs approved by the Food and Drug Administration and the Hypha Discovery MycoDiverse natural products library, which is a collection of 10 049 extracts and fractions from fermentations of higher fungi and contains compounds that are of low molecular weight and wide chemical diversity. Both of these screening campaigns confirmed that the bioluminogenic assay was high-throughput screening compatible and yielded acceptable performance in confirmation, counter, and compound/extract and fraction concentration-response assays.

Keywords

histone deacetylase sirtuin assay development drug discovery high-throughput screening

Introduction

The histone deacetylase (HDAC) enzymes play an important role in the transcriptional regulation of eukaryotic gene expression by modifying the acetylation state of histones and other important proteins.¹ Aberrant HDAC enzyme function has been implicated in many diseases including various forms of cancer,² asthma and allergic diseases,³ and inflammatory and CNS disorders.⁴ Consequently, the discovery and development of drugs targeting the HDAC enzymes has become an increasingly important activity within the pharmaceutical and biotechnology industry.⁵

The HDAC enzymes are composed of 18 family members classified in four classes depending on sequence identity and domain organization.⁶ The 11 so-called classical HDAC enzymes of class I, II, and IV are Zn²⁺ dependent.⁷ The remaining seven class III HDAC enzymes are referred to as sirtuin (SIRT) enzymes and require NAD⁺ as an essential cofactor.⁸ The first-generation pan HDAC enzyme targeted drug approved by the Food and Drug Administration (FDA) was vorinostat, a hydroxamic acid derivative that exhibits broad activity against HDAC class I and II enzymes for the treatment of cutaneous T-cell lymphoma.⁹ Subsequent to this, the natural product romidepsin, which targets the Zn²⁺-dependent HDAC enzymes, was approved by the FDA also for the treatment of cutaneous T-cell lymphoma.¹⁰ Both of these drugs have significant side effect profiles, and thus, a key goal for hit-finding studies, particularly in non-oncology indications, is to identify inhibitors from synthetic small-molecule libraries as well as natural products libraries that are both potent and selective for individual HDAC enzymes and thereby offer the potential to overcome the side effects of the current generation of HDAC-targeted drugs.^11–13 This requirement necessitates assays that are both cost-effective and adaptable to high-throughput screening (HTS).

The initial approach for monitoring HDAC enzyme activity required the use of radiolabeled histones as the substrate.¹⁴ However, more recently, HDAC enzyme assays utilized in drug discovery typically rely on artificial substrates, for example, a peptide containing a ε-acetylated lysine that is C-terminally coupled to 4-methyl-coumarin-7-amide. The product of the HDAC enzyme–mediated turnover of this substrate can be monitored after a development step, in a fluorescence intensity format.¹⁵ This type of fluorogenic assay has drawbacks, which if not fully controlled for in counter and confirmation studies, can lead to confounding results, which risk diverting resources from the development of real hits to the pursuit of artifacts.^16,17

In this report, we evaluate homogeneous, single-addition, bioluminogenic assays that can be used to monitor the activity of a range of HDAC class I/II and SIRT enzymes. See Figure 1 for a schematic representation of the HDAC-Glo and SIRT-Glo assays. In the case of the HDAC-Glo I/II assay, the proluminogenic substrate contains an acetylated lysine peptide sequence derived from Histone 4 conjugated to aminoluciferin. HDAC enzyme–mediated deacetylation of the lysine residue facilitates luminogenic substrate susceptibility to specific proteolytic cleavage by the enzyme in the developer reagent. The aminoluciferin product of that cleavage can then act as a substrate for luciferase, and the amount of light produced in this reaction is proportional to HDAC enzyme activity. The analogous SIRT-Glo assay makes use of a proluminogenic substrate that has been optimized for class III enzymes. We have investigated the HDAC-Glo Class I/II and SIRT-Glo assays using HDAC-1, HDAC-3/NcoR2, HDAC-6, and SIRT-1 enzymes, and the studies undertaken included determining their sensitivity, scalability, reproducibility, signal stability, robustness (Z′), DMSO tolerance, and pharmacological response to standard inhibitors. Finally, their performance in an automated HTS setting was assessed using 640 FDA-approved drugs and the Hypha Discovery MycoDiverse natural products library, which is a collection of 10 049 extracts and fractions with unusually high chemical novelty. Both of these screening campaigns confirmed that the HDAC-Glo I/II and SIRT-Glo bioluminogenic assays were HTS compatible and yielded acceptable performance in confirmation, counter, and compound/extract and fraction concentration-response assays.

Fig. 1.

Schematic representation of (A) the HDAC-Glo I/II and (B) SIRT-Glo assay. The histone deacetylase (HDAC) and sirtuin (SIRT) enzymes deacetylate their respective luminogenic peptide substrate, making it sensitive to a specific proteolytic cleavage event that liberates aminoluciferin. The free aminoluciferin is subsequently measured using the firefly luciferase reaction to produce a stable, persistent emission of light. All three enzymatic events occur in a coupled, near simultaneous reaction.

Materials and Methods

Reagents

Human recombinant C-ter-His-FLAG-HDAC1 (50051), HDAC-3/NcoR2 (50003), N-ter-GST-HDAC-6 (50006), and N-ter-GST-SIRT-1 (50012) were purchased from BPS Bioscience (San Diego, CA). HDAC-Glo Class I/II Kits and SIRT-Glo Kits were provided by Promega Corporation (Madison, WI). Porcine trypsin (T-0303), Tergitol NP-9 (NP9), Triton X-100 (X-100), NAD⁺ (N-1511), and the standard inhibitors valproic acid (P4543), trichostatin A (T8552), sodium butyrate (B5887), and suramin (S2671) were purchased from Sigma-Aldrich (St. Louis, MO). The standard inhibitors were dissolved to yield stock solutions in 100% v/v DMSO purchased from Roth (Karlsruhe, Germany) and stored at −20 °C. The 640 FDA-approved drugs (BML-2841-0100) were purchased from Enzo Life Sciences (Lörrach, Germany) and supplied at a concentration of 2 mg/mL in 100% v/v DMSO. The Hypha Discovery MycoDiverse natural products library was made available by Hypha Discovery as part of a collaboration and is composed of biomass extracts and extracellular fractions from fermentations of higher fungi (mushrooms and toadstools) and contains compounds that are of low molecular weight with wide chemical diversity. Each extract and fraction was dissolved in 100% v/v DMSO. Plate handling was performed using a Cell Explorer HTS platform (Perkin Elmer, Waltham, MA) equipped with an Echo 550 (Labcyte, Sunnyvale, CA) and Multidrop (Thermo, Waltham, MA) liquid-handling systems. Luminescence measurements were made using an EnVision Multilabel Reader (Perkin Elmer) or PolarStar (BMG LABTECH, Ortenberg, Germany). Assays were performed in white 96-well polystyrene, TC-treated, clear, flat-bottom, microtiter plates (3903) purchased from Corning (Lowell, MA) or white 384-well polystyrene LUMITRAC 200 (medium binding), flat-bottom microtiter plates (781075) purchased from Greiner Bio-One GmbH (Frickenhausen, Germany).

Development of HDAC-Glo I/II and SIRT-Glo kits

Luminogenic Substrates

Substrates

The two luminogenic substrates Boc-GAK(Ac)-aminoluciferin (HDAC-Glo I/II substrate) and Z-QPK(Me₂)K(Ac)-aminoluciferin (SIRT-Glo substrate) were synthesized in a similar manner to previously described peptidylaminoluciferins.¹⁸ The nonacetylated control substrates Boc- GAK-aminoluciferin (HDAC-Glo I/II control substrate) and Z-QPK(Me₂)K-aminoluciferin (SIRT-Glo control substrate) were synthesized in a similar manner as previously described¹⁸ but using orthogonal protecting groups on the lysine rather than an acetyl group, followed by deprotection as the final synthetic step.

HDAC-Glo I/II and SIRT-Glo assay buffers

Assay buffers contained 25 mM Tris buffer, pH 8.0, supplemented with 137 mM NaCl, 2.7 mM KCl, and 1 mM MgCl₂. The HDAC-Glo assay buffer also contained 1% v/v Triton X-100, and the SIRT-Glo assay buffer also contained 1% v/v Tergitol NP-9.

The HDAC-Glo I/II and SIRT-Glo assay protocols

The HDAC-Glo I/II assay reagent was prepared by (1) rehydration of lyophilized HDAC-Glo I/II substrate (with an acetylated peptide concentration of 100 µM) in 10 mL HDAC-Glo I/II assay buffer and (2) addition of 10 µL of developer reagent (containing trypsin). The SIRT-Glo assay reagent was prepared similarly but with a final acetylated peptide concentration of 200 µM. The HDAC enzymes diluted as appropriate using the HDAC-Glo I/II assay buffer were dispensed into microtiter plates, followed by an equal volume of HDAC-Glo I/II assay reagent to initiate the reaction. The microtiter plates were mixed briefly by orbital shaking (500–700 rpm) and luminescence measured at steady-state signal:background, which was achieved after 20 min using an Envision or PolarStar. The SIRT-Glo assay protocol was the same as described above except that the SIRT-Glo substrate was used instead with the inclusion of β-nicotinamide adenosine dinucleotide (NAD⁺) at 5 mM in the assay. During adaptation of the assay to the Cell Explorer HTS platform, it was shown that for each enzyme, no difference in the luminescence signal was obtained when using a low control composed of either no enzyme or an inhibitor at concentrations >>IC₅₀.

Kinetic characterization of HDAC and SIRT enzymes

Enzyme titrations and K_m determinations

Enzyme titrations were performed following the HDAC-Glo I/II and SIRT-Glo assay protocol as described above with appropriately diluted HDAC and SIRT enzymes (between 0.1 nM and 10 µM) and steady-state read time points. The K_m values for the HDAC and SIRT enzymes for their respective substrates were determined in substrate titration experiments by following the HDAC-Glo I/II and SIRT-Glo assay protocol as described above but by spiking in additional acetylated peptide to a final concentration of 400 µM, then performing twofold serial dilutions of the HDAC-Glo I/II and SIRT-Glo Assay Reagents in acetylated peptide-free luciferin detection reagent (Promega, V859A) rehydrated with assay buffer and developer reagent. The data from these duplicate experiments were fitted using the Michaelis–Menten equation, equation 1, within GraphPad Prism (version 5.02, GraphPad Software, Inc, La Jolla, CA).

v_{i} = \frac{V_{\max} {[S]}_{0}}{(K_{m} + {[S]}_{0})}

(1)

IC₅₀ determinations for standard inhibitors

The IC₅₀ values for the standard inhibitors of each enzyme were determined in order to validate the HDAC-Glo I/II and SIRT-Glo assay formats. As the standard inhibitors were dissolved in DMSO, it was necessary to determine the tolerance of the assays to this solvent. The DMSO tolerances of the HDAC-Glo I/II and SIRT-Glo assays were investigated by titration up to 10% v/v DMSO, which was substantially above its concentration proposed in any screening campaign. For standard inhibitor concentration-response experiments, the IC₅₀ values were calculated by fitting the duplicate data to the four-parameter logistic equation, equation 2 using GraphPad Prism (version 5.02, GraphPad Software, Inc).

v = \frac{v_{0}}{(1 + [I] / {IC}_{50})}

(2)

Adaptation of the HDAC-Glo I/II and SIRT-Glo assays for use in HTS campaigns

The parameters that were determined to ensure that the assays were HTS compatible included their scalability, reproducibility, signal stability, robustness (Z′), and pharmacological response to standard inhibitors. The concentration of enzyme used in these experiments was determined from the enzyme titration experiments in which linear luminescence signal was observed and there was <20% substrate turnover, and these were as follows: for HDAC-1, 0.1 nM; HDAC-3/NcoR2, 0.15 nM; HDAC-6, 0.1 nM; and SIRT-1, 2 nM.

Prior to initiating the screening campaigns, the assays were miniaturized to a volume of 10 µL, and no change in assay performance was observed. They yielded reproducible results in terms of signal stability and robustness (Z′). As the libraries to be screened were stored in 100% v/v DMSO, it was possible to obtain a suitable screening concentration of test compounds/extracts and fractions for the 640 FDA-approved drugs and Hypha Discovery MycoDiverse natural products library with 0.25% and 1% v/v DMSO, respectively. In all screening campaigns, the low control population was based on a no-enzyme control. The quality and robustness of the assay, represented as Z′, were calculated as previously described.¹⁹

In all cases, test compounds/extracts and fractions (25 nL) were added to microtiter plates. Subsequent to this, the HDAC-Glo I/II and SIRT-Glo assay protocols described above were followed. Counter assays were used to determine whether the hits identified from the screening campaigns were inhibiting the luciferase and enzymes in the developer reagent. The counter assay was essentially identical to the HDAC-Glo I/II and SIRT-Glo assay protocols except that (1) the developer reagent was used at a 1:100-fold dilution and (2) the nonacetylated control substrates were used instead at concentrations that produced luminescence comparable to their respective deacetylase assays. All screening data were processed using Activity Base XE (IDBS, UK) and visualized using Vortex (Dotmatics, Herts, UK).

Results

HDAC and SIRT enzyme substrates

The key features of the HDAC and SIRT substrates were that they contained (1) an acetylated lysine residue that would undergo HDAC and SIRT enzyme-mediated deacetylation and (2) recognition features of each enzyme as reported in the literature.^20,21 Upon HDAC or SIRT enzyme–mediated deacetylation of the lysine residue within their respective substrate, the product becomes susceptible to protease cleavage, resulting in the formation of aminoluciferin, which subsequently acts as a substrate for the luciferase enzyme, resulting in light production. The concentrations of the luciferase and trypsin enzyme in the developer reagent were at suitably high concentrations so that they were not rate limiting (data not shown). Sequence-matched, nonacetylated, luminogenic control substrates were developed for a counter assay to assess potential compound interference directed at luciferase or enzymes in the developer reagent. The concentrations of the control substrates used in the counter assays were such that they produced similar relative luminescence to that observed in their respective deacetylase assays. Any decrease in light observed in the counter assays is expected to be due to inhibition of the enzymes in the developer reagent or luciferase, indicating that the compound is a false-positive rather than a true inhibitor of the HDAC or SIRT enzymes.

Kinetic characterization of HDAC and SIRT enzymes

Enzyme titrations

The initial kinetic characterization of the HDAC and SIRT enzymes involved titration experiments that allowed an assessment of the sensitivities of the HDAC-Glo and SIRT-Glo assay reagents. These studies also allowed the determination of appropriate enzyme concentrations that should be used in their subsequent kinetic characterization while providing a suitable assay window for their use in HTS campaigns. Each enzyme was titrated at concentrations dictated by their stock concentrations, and the profiles for HDAC-6 and SIRT-1 are shown in Figures 2A and 2B , respectively. The profiles for HDAC-1 and HDAC-3/NcoR2 were similar (data not shown). The concentration of enzyme that provided a linear increase in signal varied for each enzyme, and it is noteworthy that the optimal concentrations required were relatively low and as follows: for HDAC-1, 0.1 nM; for HDAC-3/NcoR2, 0.15 nM; for HDAC-6, 0.1 nM; and for SIRT-1, 2 nM. The SIRT-1 assay yielded a signal:background ˜80:1 during after 20 min (see Supplemental Fig. S1). The HDAC-1, HDAC-3/NcoR2, and HDAC-6 assays also yielded a similar signal:background (data not shown).

Fig. 2.

Titration of (A) HDAC-6 and (B) SIRT-1 enzymes using the HDAC-Glo I/II and SIRT-Glo assays, respectively. Reactions were performed at room temperature with 50 µM substrate. Reactions were initiated by the addition of enzyme to the assay reagent, which converted the product of the enzyme-mediated deacetylation of the substrate into oxyluciferin and light (mediated by the luciferase enzyme). (C) Determination of the K_m for the substrate for HDAC-1 enzyme (45 ± 3 µM). The duplicate data from these experiments were fitted using the Michaelis–Menten equation (equation 1) within GraphPad Prism (version 5.02, GraphPad Software, Inc.). See the Materials and Methods section for full experimental details.

Substrate K_m and IC₅₀ determinations for standard inhibitors

The K_m for the HDAC-Glo and SIRT-Glo substrates for each enzyme were also determined in this study. The K_m determinations were made by titrating the substrates with an appropriate fixed concentration of each enzyme when <20% substrate turnover took place and fitting the data using the Michaelis–Menten equation, equation 1 within GraphPad Prism (version 5.02, GraphPad Software, Inc.). The K_m value determined for HDAC-1 was 45 ± 3 µM (see Fig. 2C ). The K_m values determined for the other enzymes were as follows: for HDAC-3/NcoR2, 192 ± 15 µM; for HDAC-6, 31 ± 3 µM; and for SIRT-1, 183 ± 16 µM (data not shown). All K_m values were comparable to values reported in the literature.²¹

The tolerance of the HDAC-Glo I/II and SIRT-Glo assays toward DMSO up to 10% v/v was investigated. The HDAC-6 enzyme was able to tolerate up to 1% v/v DMSO with negligible loss in enzyme activity (see Supplemental Fig. S2), which was acceptable as 0.25% and 1% v/v DMSO was used when screening the 640 FDA-approved drugs and Hypha Discovery MycoDiverse natural products library, respectively. The HDAC-1, HDAC-3/NcoR2, and SIRT-1 assays also yielded a similar sensitivity toward DMSO (data not shown). Four compounds that have been reported to inhibit HDAC and SIRT enzymes were also investigated (valproic acid, trichostatin A, sodium butyrate, and suramin; see Fig. 3A ). Standard inhibitor concentration-response assays were performed with HDAC-1, HDAC-3/NcoR2, HDAC-6, and SIRT-1, and the data from these experiments were fitted using the four-parameter logistic equation (equation 2) using GraphPad Prism (version 5.02, GraphPad Software, Inc.). All concentration-response data sets were of a high quality, and those for HDAC-1 and trichostatin A, HDAC-3 and sodium butyrate, HDAC-6 and trichostatin A, and SIRT-1 and suramin are shown in Figures 3B to 3E , respectively. The results obtained in the bioluminogenic HDAC and SIRT assays for the standard inhibitors were comparable to those reported in the literature (see Table 1 ).

Fig. 3.

(A) Structures of the standard inhibitors valproic acid, trichostatin A, sodium butyrate, and suramin used in the present study. (B) Compound concentration-response curve for HDAC-1 and trichostatin A, IC₅₀ = 1.26 ± 0.1 nM (Hill slope = 1.01 ± 0.12). (C) Compound concentration-response curve for HDAC-3 and sodium butyrate, IC₅₀ = 135 ± 5 µM (Hill slope = 0.77 ± 0.06). (D) Compound concentration-response curve for HDAC-6 and trichostatin A, IC₅₀ = 3.6 ± 0.8 nM (Hill slope = 1.45 ± 0.11). (E) Compound concentration-response curve for SIRT-1 and suramin, IC₅₀ = 1.5 ± 0.2 µM (Hill slope = 0.64 ± 0.06). The duplicate data from these experiments were fitted using the four-parameter logistic equation (equation 2) using GraphPad Prism (version 5.02, GraphPad Software, Inc.). See the Materials and Methods section for full experimental details.

Table 1.

Summary of the Results from the Evaluation of Standard Inhibitors (Valproic Acid, Trichostatin A, Sodium Butyrate, and Suramin) against HDAC-1, HDAC-3/NcoR2, HDAC-6, and SIRT-1

	Valproic Acid		Sodium Butyrate		Trichostatin A		Suramin
Enzyme	Literature IC₅₀	Bioluminogenic Activity Assay IC₅₀	Literature IC₅₀	Bioluminogenic Activity Assay IC₅₀	Literature IC₅₀	Bioluminogenic Activity Assay IC₅₀	Literature IC₅₀	Bioluminogenic Activity Assay IC₅₀
HDAC-1	700²⁸ 1600³⁰ 400³³	660 ± 9	300²⁸ 350³¹	630 ± 9	0.003²⁹ 0.002³⁰ 0.003³¹ 0.0015³²	0.0013 ±0.0003	NA	>1000
HDAC-3	1000²⁸ 3070³⁰	>1000	NA	135 ± 5	0.011²⁹ 0.004³⁰ 0.0006³²	0.0015 ± 0.0001	NA	>1000
HDAC-6	>20 000²⁸ >10³⁰	>1000	NA	>1000	0.006²⁹ 0.003³⁰	0.0036 ± 0.0008	NA	>1000
SIRT-1	NA	>1000	NA	>1000	NA	>1000	0.3³⁴	1.5 ± 0.2

All IC₅₀ values are in µM. The data obtained using the HDAC-Glo I/II and SIRT-Glo assay have been compared with values reported in the literature. See the Discussion section for full details. NA = not available.

Automation of the HDAC-Glo I/II and SIRT-Glo assays and execution of the HTS campaigns against the 640 FDA-approved drugs and Hypha Discovery MycoDiverse natural products library

After the preliminary enzyme characterization experiments as described above were completed, the key parameters necessary for their successful adaptation to HTS were determined. These included their sensitivity, scalability, reproducibility, and robustness (Z′). The concentrations of each enzyme for the screening and compound/extract and fraction concentration-response experiments were fixed as follows: for HDAC-1, 0.1 nM; HDAC-3/NcoR2, 0.15 nM; HDAC-6, 0.1 nM; and SIRT-1, 2 nM, and reactions were allowed to proceed that yielded <20% substrate turnover. The assay volume was also reduced from 30 µL to 10 µL with minimal effort while retaining assay robustness and pharmacology with respect toward the standard inhibitors. The read time per microtiter plate was approximately 4 min, and this level of throughput would enable >50 000 assays to be performed per day on an automated screening platform. The overall quality and robustness of the assays were determined in 384-well microtiter plates containing 352 test compounds (located in columns 1–22), the high control (DMSO alone, located in column 23), and the low control (no enzyme, located in column 24). Each assay well contained either 0.25% or 1% v/v DMSO depending on the library screened.

The output from the 640 FDA-approved drugs screening campaign

HDAC-1, HDAC-3/NcoR2, HDAC-6, and SIRT-1 enzymes were screened in duplicate against the 640 FDA-approved drugs. The Z′ values were >0.7 for each of the 16 microtiter plates (see Supplemental Fig. S3). The duplicate data sets for the HDAC-6 and SIRT-1 screening campaigns are shown in Figures 4A and 4B , respectively. The duplicate data sets for HDAC-1 and HDAC-3/NcoR2 screening campaigns were similar (data not shown). The correlation coefficients for each duplicate data set were as follows: for HDAC-1, r² = 1.03; HDAC-3/NcoR2, r² = 0.84; HDAC-6, r² = 0.99; and SIRT-1, r² = 0.87. Hits were classified as those compounds that yielded >50% inhibition at their screening concentration, from which the hit rates were determined as follows: for HDAC-1, 15 compounds, 2.3% hit rate; HDAC-3/NcoR2, 16 compounds, 2.5% hit rate; HDAC-6, 22 compounds, 3.4% hit rate; and SIRT-1, 20 compounds; 3.1% hit rate. The hits in each assay were evaluated in their respective confirmation and counter assays, which yielded the following statistics: the confirmation rate for each assay was >80% and the false-positive hit rate (actives in the counter assay) was <20%. A selection of the hits was subsequently evaluated in compound concentration-response assays.

Fig. 4.

Correlation plots for the screening of 640 Food and Drug Administration–approved drugs against (A) HDAC-6 enzyme (r² = 0.99) and (B) SIRT-1 enzyme (r² = 0.87). See the Materials and Methods section for full experimental details.

The output from the Hypha Discovery MycoDiverse natural products library screening campaign

HDAC-6 and SIRT-1 enzymes were screened against the Hypha Discovery MycoDiverse natural products library using the same strategy as for the 640 FDA-approved drugs screening campaign but in singlicate. The Z′ values were >0.65 and >0.85 for the 34 microtiter plates screened against HDAC-6 (see Supplemental Fig. S4) and SIRT-1 (see Supplemental Fig. S5) respectively. The hit rates (extracts and fractions producing >50% inhibition at their screening concentration) for each enzyme were as follows: for HDAC-6, 72 extracts and fractions, 0.71% hit rate, and for SIRT-1, 59 extracts and fractions, 0.59% hit rate. As in the case for the screening of the 640 FDA-approved drugs, evaluation of the hits from each assay in their respective confirmation and counter assays yielded an acceptable false-positive hit rate.

Discussion

The validation of HDAC and SIRT enzymes as targets for therapeutic intervention by small molecules has now been established by the FDA approval of a number of vorinostat and romidepsin.^9,10 These drugs have significant side-effect profiles, and there is a need to identify novel and selective HDAC and SIRT enzyme inhibitors, in particular in non-oncology indications. One method that is used to identify chemical starting points for drug discovery makes use of HTS against small-molecule libraries containing >500 000 compounds, and these are often screened in standard microtiter plates in 384- and 1536-well format. Although fluorescence-based assay formats have been used in drug discovery for determining in vitro HDAC activity,^20,21 they have drawbacks, and it needs to be shown that any hits that are identified from such screening campaigns are not artifacts arising from interference mechanisms due to the intrinsic fluorescence properties of compounds.²² The potential impact of interfering compounds arising from fluorescence intensity–based assays was brought into focus by the controversy involving compounds identified as activators of SIRT-1 using fluorescence-based assays.²³ Subsequent independent side-by-side comparisons using TAMRA-p53 and native-p53 peptide substrates, high-performance liquid chromatography separation, and ultraviolet detection of the substrates and deacetylated products have shown deacetylase activity only with the TAMRA-p53 peptide substrate. Further biophysical studies using nuclear magnetic resonance, surface plasmon resonance, and isothermal calorimetry showed that one of the compounds interacted directly with TAMRA peptides in a dose-dependent manner; however, the native peptides did not.^24,25 This episode demonstrates the need for additional assays to validate the hits identified from fluorescence–based assays.

In this article, we report the evaluation of homogeneous, single-addition, bioluminogenic HDAC enzyme assays with respect to their sensitivity, scalability, reproducibility, signal stability, robustness (Z′), DMSO tolerance, and pharmacological relevance toward standard inhibitors. Is it noteworthy that these assays met all criteria required for them to be compatible for screening purposes. The assays required relatively low concentrations of enzyme (nM range), thus allowing the identification of potent inhibitors and were successfully miniaturized to 10 µL assay volumes and yielded IC₅₀ values for the standard inhibitors that were comparable to those reported in the literature (see Table 1 ).

The screening of targets against known drugs or compounds that might have failed to demonstrate efficacy against the primary disease in clinical trials offer the potential for drug repositioning with reduced cost and time scale for evaluation against the alternative diseases.²⁶ The 640 FDA-approved drugs were used for this purpose and screened against HDAC-1, HDAC-3/NcoR2, HDAC-6, and SIRT-1 enzymes. These activities made use of 16 microtiter plates in total, and they all yielded a Z′ >0.7 (see Supplemental Fig. S3). In each case, hits were classified as those compounds that yielded >50% inhibition of enzyme activity at their screening concentration. The correlation coefficient (r²) for each duplicate data set was >0.87, indicating good assay reproducibility (see Fig. 4A and 4B ). The subsequent evaluation of the hits yielded a >80% confirmation rate and <20% false-positive hit rate (actives in the counter assay). A selection of the hits were subsequently evaluated in compound concentration-response assays, and examples of known drugs that were identified as HDAC-1 enzyme inhibitors included tranlistat (IC₅₀ = 18 ± 2 µM), cinanserin (IC₅₀ = 12 ± 1 µM), and leflunomide (IC₅₀ = 16 ± 2 µM); data not shown, and these contain a phenylamide core that is also found in mocetinostat, vorinostat, entinostat, and givinostat, which are drugs undergoing clinical trials as HDAC inhibitors²⁷ and provide evidence of structure-activity relationships emerging from the screening campaign.

In addition to synthetic small-molecule libraries, natural products are also valuable sources for identifying starting points in drug discovery, and this is exemplified by the FDA approval of the natural product romidepsin as an HDAC inhibitor.¹⁰ On this basis, the Hypha Discovery MycoDiverse natural products library was screened in singlicate against HDAC-6 and SIRT-1 enzymes. These activities made use of 34 microtiter plates for each enzyme. For the HDAC-6 screening campaign, all microtiter plates yielded a Z′ >0.65 (see Supplemental Fig. S4). For the SIRT-1 screening campaign, all microtiter plates yielded a Z′ >0.85 (see Supplemental Fig. S5). Although 10 049 extracts and fractions were evaluated, many of these samples contain several compounds; hence, the total number of individual compounds that were screened against each enzyme would have been significantly more than 10 049. The hit rates (extracts and fractions yielding >50% inhibition at the concentration they were screened) for each enzyme were 0.71% for HDAC-6 (72 extracts and fractions) and 0.59% for SIRT-1 (59 extracts and fractions), and these are lower than for the output from the 640 FDA-approved drugs screening campaign. The postscreening cascade and results in this case were similar to those described above for the FDA-approved drugs screening campaign as the confirmation rate for each assay was >80% and the false-positive hit rate (actives in the counter assay) was <20%. The quality of the data for the hits from Hypha Discovery’s MycoDiverse natural products screening campaign was sufficiently high to initiate the purification and characterization of the individual active natural product compounds. The unusually low molecular weight and uniqueness of the bioactive compounds discovered to date from this library indicate that such novel, potent, and selective HDAC and SIRT inhibitors for further development could be found during this subsequent work.

Footnotes

Acknowledgements

We thank Hypha Discovery for providing the MycoDiverse natural products library for use in this study; Björn Windshügel for helpful discussions; Katarina Bohm, Tom Livelli, Eric Muhr, and Terry Riss for support; and Paul Steckelberg and Lisa Cherradi for technical assistance.

Supplementary material for this article is available on the Journal of Biomolecular Screening Web site at .

References

Gallinari

Di Marco

Jones

Pallaoro

Steinkühler

HDACs, Histone Deacetylation and Gene Transcription: From Molecular Biology to Cancer Therapeutics. Cell Res. 2007, 17, 195–211.

Kim

H. J.

Bae

S. C.

Histone Deacetylase Inhibitors: Molecular Mechanisms of Action and Clinical Trials as Anti-cancer Drugs. Am. J. Transl. Res. 2011, 3, 166–179.

Bhavsar

Ahmad

Adcock

I. M.

The Role of Histone Deacetylases in Asthma and Allergic Diseases. J. Allergy Clin. Immunol. 2008, 121, 580–584.

Chavan

A. V.

Somani

R. R.

HDAC Inhibitors—New Generation of Target Specific Treatment. Mini. Rev. Med. Chem. 2010, 10, 1263–1276.

Wang

Dymock

B. W.

New Patented Histone Deacetylase Inhibitors. Expert Opin. Ther. Pat. 2009, 19, 1727–1757.

de Ruijter

A. J.

van Gennip

A. H.

Caron

H. N.

Kemp

van Kuilenburg

A. B.

Histone Deacetylases (HDACs): Characterization of the Classical HDAC Family. Biochem. J. 2003, 15, 737–749.

Ficner

Novel Structural Insights into Class I and II Histone Deacetylases. Curr. Top. Med. Chem. 2009, 9, 235–240.

Haigis

M. C.

Sinclair

D. A.

Mammalian Sirtuins: Biological Insights and Disease Relevance. Annu. Rev. Pathol. 2010, 5, 253–295.

Duvic

Vorinostat: A New Oral Histone Deacetylase Inhibitor Approved for Cutaneous T-Cell Lymphoma. Expert Opin. Investig. Drugs. 2007, 16, 1111–1120.

10.

Lemoine

Younes

Histone Deacetylase Inhibitors in the Treatment of Lymphoma. Discov. Med. 2010, 10, 462–470.

11.

Schemies

Uciechowska

Sippl

Jung

NAD⁺-Dependent Histone Deacetylases (Sirtuins) as Novel Therapeutic Targets. Med. Res. Rev. 2010, 30, 861–889.

12.

Aljada

Dong

Mousa

S. A.

Sirtuin-Targeting Drugs: Mechanisms of Action and Potential Therapeutic Applications. Curr. Opin. Investig. Drugs. 2010, 11, 1158–1168.

13.

Sauve

A. A.

Pharmaceutical Strategies for Activating Sirtuins. Curr. Pharm. Des. 2009, 15, 45–56.

14.

Kölle

Brosch

Lechner

Lusser

Loidl

Biochemical Methods for Analysis of Histone Deacetylases. Methods. 1998, 15, 323–331.

15.

Wegener

Hildmann

Riester

Schober

Meyer-Almes

F. J.

Deubzer

H. E.

Oehme

Witt

Lang

Jaensch

Makarov

Lange

Busse

Schwienhorst

Identification of Novel Small-Molecule Histone Deacetylase Inhibitors by Medium-Throughput Screening Using a Fluorigenic Assay. Biochem. J. 2008, 413, 143–150.

16.

Gribbon

Sewing

Fluorescence Readouts in HTS: No Gain without Pain?

Drug Discov. Today. 2003, 8, 1035–1043.

17.

Gul

Gribbon

Exemplification of the Challenges Associated with Utilising Fluorescence Intensity Based Assays in Discovery. Expert Opin. Drug Discov. 2010, 5, 681–690.

18.

O’Brien

M. A.

Daily

W. J.

Hesselberth

P. E.

Moravec

R. A.

Scurria

M. A.

Klaubert

D. H.

Bulleit

R. F.

Wood

K. V.

Homogeneous, Bioluminescent Protease Assays: Caspase-3 as a Model. J. Biomol. Screen. 2005, 10, 137–148.

19.

Zhang

J. H.

Chung

T. D.

Oldenburg

K. R.

A Simple Statistical Parameter for Use in Evaluation and Validation of High Throughput Screening Assays. J. Biomol. Screen. 1999, 4, 67–73.

20.

Hauser

A. T.

Jung

Assays for Histone Deacetylases. Curr. Top. Med. Chem. 2009, 9, 227–234.

21.

Wegener

Wirsching

Riester

Schwienhorst

A Fluorogenic Histone Deacetylase Assay Well Suited for High-Throughput Activity Screening. Chem. Biol. 2003, 10, 61–68.

22.

Jadhav

Ferreira

R. S.

Klumpp

Mott

B. T.

Austin

C. P.

Inglese

Thomas

C. J.

Maloney

D. J.

Shoichet

B. K.

Simeonov

Quantitative Analyses of Aggregation, Autofluorescence, and Reactivity Artifacts in a Screen for Inhibitors of a Thiol Protease. J. Med. Chem. 2010, 53, 37–51.

23.

Howitz

K. T.

Bitterman

K. J.

Cohen

H. Y.

Lamming

D. W.

Lavu

Wood

J. G.

Zipkin

R. E.

Chung

Kisielewski

Zhang

L. L.

Scherer

Sinclair

D. A.

Small Molecule Activators of Sirtuins Extend Saccharomyces cerevisiae Lifespan. Nature. 2003, 425, 191–196.

24.

Pacholec

Bleasdale

J. E.

Chrunyk

Cunningham

Flynn

Garofalo

R. S.

Griffith

Griffor

Loulakis

Pabst

Qiu

Stockman

Thanabal

Varghese

Ward

Withka

Ahn

SRT1720, SRT2183, SRT1460, and Resveratrol Are Not Direct Activators of SIRT1. J. Biol. Chem. 2010, 285, 8340–8351.

25.

Beher

Cumine

Kim

K. W.

S. C.

Atangan

Wang

Resveratrol Is Not a Direct Activator of SIRT1 Enzyme Activity. Chem. Biol. Drug Des. 2009, 74, 619–624.

26.

Ashburn

T. T.

Thor

K. B.

Drug Repositioning: Identifying and Developing New Uses for Existing Drugs. Nat. Rev. Drug Discov. 2004, 3, 673–683.

27.

Mack

G. S.

To Selectivity and Beyond. Nat. Biotech. 2010, 28, 1259–1266.

28.

Gurvich

Tsygankova

O. M.

Meinkoth

J. L.

Klein

P. S.

Histone Deacetylase Is a Target of Valproic Acid-Mediated Cellular Differentiation. Cancer Res. 2004, 64, 1079–1086.

29.

Olsen

C. A.

Ghadiri

M. R.

Discovery of Potent and Selective Histone Deacetylase Inhibitors via Focused Combinatorial Libraries of Cyclic α₃β-tetrapeptides. J. Med. Chem. 2009, 52, 7836–7846.

30.

Khan

Jeffers

Kumar

Hackett

Boldog

Khramtsov

Qian

Mills

Berghs

S. C.

Carey

Finn

P. W.

Collins

L. S.

Tumber

Ritchie

J. W.

Jensen

P. B.

Lichenstein

H. S.

Sehested

Determination of the Class and Isoform Selectivity of Small-Molecule Histone Deacetylase Inhibitors. Biochem. J. 2008, 409, 581–589.

31.

Buggy

J. J.

Sideris

M. L.

Mak

Lorimer

D. D.

McIntosh

Clark

J. M.

Cloning and Characterization of a Novel Human Histone Deacetylase, HDAC8. Biochem. J. 2000, 350, 199–205.

32.

Vannini

Volpari

Filocamo

Casavola

E. C.

Brunetti

Renzoni

Chakravarty

Paolini

De Francesco

Gallinari

Steinkühler

Di Marco

Crystal Structure of a Eukaryotic Zinc-Dependent Histone Deacetylase, Human HDAC8, Complexed with a Hydroxamic Acid Inhibitor. Proc. Natl. Acad. Sci. U. S. A. 2004, 101, 15064–15069.

33.

Phiel

C. J.

Zhang

Huang

E. Y.

Guenther

M. G.

Lazar

M. A.

Klein

P. S.

Histone Deacetylase Is a Direct Target of Valproic Acid, a Potent Anticonvulsant, Mood Stabilizer, and Teratogen. J. Biol. Chem. 2001, 276, 36734–36741.

34.

Trapp

Meier

Hongwiset

Kassack

M. U.

Sippl

Jung

Structure-Activity Studies on Suramin Analogues as Inhibitors of NAD⁺-Dependent Histone Deacetylases (Sirtuins). Chem. Med. Chem. 2007, 2, 1419–1431.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.04 MB

0.09 MB

0.04 MB

0.05 MB

0.04 MB

A Bioluminogenic HDAC Activity Assay

Abstract

Keywords

Introduction

Materials and Methods

Reagents

Development of HDAC-Glo I/II and SIRT-Glo kits

Luminogenic Substrates

Substrates

HDAC-Glo I/II and SIRT-Glo assay buffers

The HDAC-Glo I/II and SIRT-Glo assay protocols

Kinetic characterization of HDAC and SIRT enzymes

Enzyme titrations and Km determinations

IC50 determinations for standard inhibitors

Adaptation of the HDAC-Glo I/II and SIRT-Glo assays for use in HTS campaigns

Results

HDAC and SIRT enzyme substrates

Kinetic characterization of HDAC and SIRT enzymes

Enzyme titrations

Substrate Km and IC50 determinations for standard inhibitors

Automation of the HDAC-Glo I/II and SIRT-Glo assays and execution of the HTS campaigns against the 640 FDA-approved drugs and Hypha Discovery MycoDiverse natural products library

The output from the 640 FDA-approved drugs screening campaign

The output from the Hypha Discovery MycoDiverse natural products library screening campaign

Discussion

Footnotes

Acknowledgements

References

Supplementary Material

Enzyme titrations and K_m determinations

IC₅₀ determinations for standard inhibitors

Substrate K_m and IC₅₀ determinations for standard inhibitors