A substrate-independent TR-FRET histone deacetylase inhibitor assay.

Developing molecularly targeted therapeutics with minimal off-target effects is facilitated by an understanding of compound selectivity. However, for HDAC inhibitors, a clear understanding of specificity has been challenging. In particular, it has been suggested that use of nonspecific substrates and the presence of multiple HDAC activities in enzyme preparations may complicate interpretation of inhibitor experiments. To overcome these and other potential limitations of activity-based HDAC assays, the authors have developed an assay format based on measurement of the binding affinity of inhibitors rather than measurement of enzyme activity. One advantage of this format is that it does not require use of a substrate and thus ameliorates concerns about lack of specificity of existing substrates. This assay is based on an Alexa Fluor® 647-labeled HDAC inhibitor or "tracer," which binds with a high affinity to Class I and Class IIb HDACs. Unlike activity assays, which can be affected by the presence of residual untagged endogenous HDACs from the host expression system, the signal in this format is dependent on the presence of an epitope tag on the specific HDAC of interest. The authors demonstrate the utility of this method by determining the potencies of commonly used inhibitors for six human HDACs.

Development and applications of a broad-coverage, TR-FRET-based kinase binding assay platform.

The expansion of kinase assay technologies over the past decade has mirrored the growing interest in kinases as drug targets. As a result, there is no shortage of convenient, fluorescence-based methods available to assay targets that span the kinome. The authors recently reported on the development of a non-activity-based assay to characterize kinase inhibitors that depended on displacement of an Alexa Fluor 647 conjugate of staurosporine (a "tracer") from a particular kinase. Kinase inhibitors were characterized by a change in fluorescence lifetime of the tracer when it was bound to a kinase relative to when it was displaced by an inhibitor. Here, the authors report on improvements to this strategy by reconfiguring the assay in a time-resolved fluorescence resonance energy transfer (TR-FRET) format that simplifies instrumentation requirements and allows for the use of a substantially lower concentration of kinase than was required in the fluorescence-lifetime-based format. The authors use this new assay to demonstrate several aspects of the binding assay format that are advantageous relative to traditional activity-based assays. The TR-FRET binding format facilitates the assay of compounds against low-activity kinases, allows for the characterization of type II kinase inhibitors either using nonactivated kinases or by monitoring compound potency over time, and ensures that the signal being detected is specific to the kinase of interest and not a contaminating kinase.

Identification of pregnane X receptor ligands using time-resolved fluorescence resonance energy transfer and quantitative high-throughput screening.

The human pregnane X nuclear receptor (PXR) is a xenobiotic-regulated receptor that is activated by a range of diverse chemicals, including antibiotics, antifungals, glucocorticoids, and herbal extracts. PXR has been characterized as an important receptor in the metabolism of xenobiotics due to induction of cytochrome P450 isozymes and activation by a large number of prescribed medications. Developing methodologies that can efficiently detect PXR ligands will be clinically beneficial to avoid potential drug-drug interactions. To facilitate the identification of PXR ligands, a time-resolved fluorescence resonance energy transfer (TR-FRET) assay was miniaturized to a 1,536-well microtiter plate format to employ quantitative high-throughput screening (qHTS). The optimized 1,536-well TR-FRET assay showed Z'-factors of >or=0.5. Seven- to 15-point concentration-response curves (CRCs) were generated for 8,280 compounds using both terbium and fluorescein emission data, resulting in the generation of 241,664 data points. The qHTS method allowed us to retrospectively examine single concentration screening datasets to assess the sensitivity and selectivity of the PXR assay at different compound screening concentrations. Furthermore, nonspecific assay artifacts such as concentration-based quenching of the terbium signal and compound fluorescence were identified through the examination of CRCs for specific emission channels. The CRC information was also used to define chemotypes associated with PXR ligands. This study demonstrates the feasibility of profiling thousands of compounds against PXR using the TR-FRET assay in a high-throughput format.

Development of the predictor HERG fluorescence polarization assay using a membrane protein enrichment approach.

The life-threatening consequences of acquired, or drug-induced, long QT syndrome due to block of the human ether-a-go-go-related gene (hERG) channel are well appreciated and have been the cause of several drugs being removed from the market in recent years because of patient death. In the last decade, the propensity for block of the hERG channel by a diverse and expanding set of compounds has led to the requirement that all new drugs be tested for hERG channel block in a functional patch-clamp assay. Because of the need to identify potential hERG blockers early in the discovery process, radiometric hERG binding assays are preferred over patch-clamp assays for compound triage, because of relative advantages in speed and cost. Even so, these radiometric binding assays are laborious and require dedicated instrumentation and infrastructure to cope with the regulatory and safety issues associated with the use of radiation. To overcome these limitations, we developed a homogeneous, fluorescence polarization-based assay to identify and characterize the affinity of small molecules for the hERG channel and have demonstrated tight correlation with data obtained from either radioligand binding or patch-clamp assays. Key to the development of this assay was a cell line that expressed highly elevated levels of hERG protein, which was generated by coupling expression of the hERG channel to that of a selectable cell surface marker. A high-expressing clone was isolated by flow cytometry and used to generate membrane preparations that contained >50-fold the typical density of hERG channels measured by [(3)H]astemizole binding. This strategy enabled the Predictor (Invitrogen, Carlsbad, CA) hERG fluorescence polarization assay and should be useful in the development of other fluorescence polarization-based assays that use membrane proteins.

The androgen receptor T877A mutant recruits LXXLL and FXXLF peptides differently than wild-type androgen receptor in a time-resolved fluorescence resonance energy transfer assay.

The interactions of the ligand binding domain (LBD) of androgen receptor (AR) and the AR T877A mutant, found in prostate cancer, with peptides from coactivator and corepressor proteins or random phage display peptides were investigated using in vitro time-resolved fluorescence resonance energy transfer (TR-FRET). Interaction of wild-type AR LBD with the random phage display peptide D11FxxLF was observed with dihydrotestosterone (DHT), testosterone, R1881, estradiol, spironolactone, progesterone, and cortisol resulting in distinct dose dependency (EC50) values for each ligand and correlating well with the reported rank order potency of these agonists. Increasing concentrations of cyproterone acetate and mifepristone resulted in more complete disruption of the DHT-mediated AR-D11FxxLF peptide interaction, while flutamide, hydroxyflutamide, and bicalutamide caused only partial disruption of the complex. The mutant AR T877A LBD exhibited increased binding affinities for all ligands tested except for bicalutamide, mifepristone, DHT, and R1881 in a competitive binding assay as compared to wild-type AR LBD. This mutation was also characterized by increased ligand potency for agonist-induced peptide recruitment. Although usually an antagonist, hydroxyflutamide was more potent in the recruitment of D11FxxLF or an SRC3-1 LXXLL motif to AR T877A LBD than AR LBD. The antagonist cyproterone acetate behaved as a full antagonist of D11FxxLF recruitment to AR LBD and AR T877A LBD but as a more potent agonist in the recruitment of SRC3-1 to AR T877A LBD. These results suggest that the AR T877A mutation affects both ligand affinity and ligand dose dependency for peptide recruitment and may explain in part the altered responses of antagonists and increased transcriptional activation reported in androgen-independent prostate cancers.

Multiparameter analysis of a screen for progesterone receptor ligands: comparing fluorescence lifetime and fluorescence polarization measurements.

Direct measurement of the fluorescence lifetime (FLT) of a fluorescent label is an emerging method for high-throughput screening. Changes in the fluorescence lifetime can be correlated to changes in the non-radiative relaxation pathway(s) for the excited state of the label. These pathways can be environmentally sensitive, such as when a labeled analyte is free in solution versus bound to a receptor. Because lifetime is an intrinsic property of a fluorophore, it is not concentration dependent, and therefore has advantages similar to those of ratiometric fluorescent techniques such as fluorescence resonance energy transfer or fluorescence polarization. We have applied the FLT measurement technique to a screen of a small compound library in order to identify compounds that bind to the progesterone receptor, and compared the results to those obtained by performing the assay in fluorescence polarization mode. Each readout modality showed excellent Z'; values, with the FLT readout performing slightly better in this respect. Interfering compounds could be rapidly identified for either assay format by comparing the results between the two formats.

Overcoming compound interference in fluorescence polarization-based kinase assays using far-red tracers.

Kinase-mediated phosphorylation of proteins is critical to the regulation of many biological processes, including cell growth, apoptosis, and differentiation. Because of the central role that kinases play in processes that can lead to disease states, the targeting of kinases with small-molecule inhibitors is a validated strategy for therapeutic intervention. Classic methods for assaying kinases include nonhomogenous enzyme-linked immunosorbent assays or scintillation-based formats using [gamma-(32)P]ATP. However, homogenous fluorescence-based assays have gained in popularity in recent years due to decreased costs in reagent usage through miniaturization, increased throughput, and avoidance of regulatory costs associated with the use of radiation. Whereas the readout signal from a nonhomogenous or radioactive assay is largely impervious to interferences from matrix components (such as library compounds), all homogenous fluorescent assay formats are subject to such interferences. Interference from intrinsically fluorescent compounds or from scattered light due to precipitated compounds can interfere with assays that depend on a fluorescence intensity (or fluorescence quenching), fluorescence resonance energy transfer, or fluorescence polarization-based readout. Because these interfering factors show a greater effect at lower wavelengths, one strategy to overcome such interferences is to develop fluorescent assays using longer wavelength (red-shifted) fluorescent probes. In this article, we describe the PanVera PolarScreen far-red fluorescence polarization assay format, which mitigates assay interference from autofluorescent compounds or scattered light through the use of a far-red tracer. The tracer shows substantially less interference from light scatter or autofluorescent library compounds than do fluorescein-based tracers, and gives rise to a larger assay window than the popular far-red fluorophore Cy5.