A generic high-throughput screening assay for kinases: protein kinase a as an example

Siamese Recurrent Neural Network with a Self-Attention Mechanism for Bioactivity Prediction

Activity prediction plays an essential role in drug discovery by directing search of drug candidates in the relevant chemical space. Despite being applied successfully to image recognition and semantic similarity, the Siamese neural network has rarely been explored in drug discovery where modelling faces challenges such as insufficient data and class imbalance. Here, we present a Siamese recurrent neural network model (SiameseCHEM) based on bidirectional long short-term memory architecture with a self-attention mechanism, which can automatically learn discriminative features from the SMILES representations of small molecules. Subsequently, it is used to categorize bioactivity of small molecules via N-shot learning. Trained on random SMILES strings, it proves robust across five different datasets for the task of binary or categorical classification of bioactivity. Benchmarking against two baseline machine learning models which use the chemistry-rich ECFP fingerprints as the input, the deep learning model outperforms on three datasets and achieves comparable performance on the other two. The failure of both baseline methods on SMILES strings highlights that the deep learning model may learn task-specific chemistry features encoded in SMILES strings.

Label-Free Fluorescent Kinase and Phosphatase Enzyme Assays with Supramolecular Host-Dye Pairs

The combination of the macrocyclic hosts p-sulfonatocalix[4]arene and cucurbit[7]uril with the fluorescent dyes lucigenin and berberine affords two label-free enzyme assays for the detection of kinase and phosphatase activity by fluorescence monitoring. In contrast to established assays, no substrate labeling is required. Since phosphorylation is one of the most important regulatory mechanisms in biological signal transduction, the assays should be useful for identification of inhibitors and activators in high-throughput screening (HTS) format for drug discovery.

A Simple Theoretical Model for Fluorescence Polarization Binding Assay Development

Fluorescence polarization is a screening technology that is radioactivity free, homogeneous, and ratiometric. The signal measured with this technology is a weighted value of free and bound ligand. As a consequence, saturation curves are accessible only after calculation of the corresponding concentrations of free and bound ligand. To make this technology more accessible to assay development, the authors propose a simple mathematical model that predicts fluorescence polarization values from ligand and receptor total concentrations, depending on the corresponding dissociation constant. This model was validated using data of Bodipy-NDP-αMSH binding to MC5, obtained after either ligand saturation of a receptor preparation or, conversely, receptor saturation of a ligand solution. These experimental data were also used to calculate the actual concentration of free and bound ligand and receptor and to obtain pharmacological constants by Scatchard analysis. A general method is proposed, which facilitates the design of fluorescence polarization binding assays by relying on the representation of theoretical polarization values. This approach is illustrated by the application to 2 systems of very different affinities.

Scintillation proximity assay

Scintillation proximity assay technologies provide a rapid non-separation method to measure common biological interactions using radioactively tagged molecules. This unit identifies potential uses of the technology for the measurement of receptor-ligand binding, cAMP accumulation, GTP binding to heterotrimeric G proteins, protease activity and cellular uptake.

Development, validation and implementation of immobilized metal affinity for phosphochemicals (IMAP)-based high-throughput screening assays for low-molecular-weight compound libraries

This protocol describes assay development, validation and implementation of automated immobilized metal affinity for phosphochemicals (IMAP)-based fluorescence polarization (FP) and time-resolved fluorescence resonance energy transfer (TR-FRET) high-throughput screening (HTS) assays for identification of low-molecular-weight kinase inhibitors. Both procedures are performed in miniaturized kinase reaction volumes and involve the stepwise addition of test or control compounds, enzyme and substrate/ATP. Kinase reactions are stopped by subsequent addition of IMAP-binding buffer. Assay attributes of the IMAP FP and TR-FRET methodologies are described. HTS assays developed using these procedures should result in Z-factors and low assay variability necessary for robust HTS assays. Providing that the required reagents and equipment are available, one scientist should be able to develop a 384-well, miniaturized HTS assay in approximately 6-8 weeks. Specific automated HTS assay conditions will determine the number of assay plates processed in a screening session, but two scientists should expect to process between 100 and 150 assay plates in one 8-h screening day.

Single-label kinase and phosphatase assays for tyrosine phosphorylation using nanosecond time-resolved fluorescence detection

The collision-induced fluorescence quenching of a 2,3-diazabicyclo[2.2.2]oct-2-ene-labeled asparagine (Dbo) by hydrogen atom abstraction from the tyrosine residue in peptide substrates was introduced as a single-labeling strategy to assay the activity of tyrosine kinases and phosphatases. The assays were tested for 12 different combinations of Dbo-labeled substrates and with the enzymes p60c-Src Src kinase, EGFR kinase, YOP protein tyrosine phosphatase, as well as acid and alkaline phosphatases, thereby demonstrating a broad application potential. The steady-state fluorescence changed by a factor of up to 7 in the course of the enzymatic reaction, which allowed for a sufficient sensitivity of continuous monitoring in steady-state experiments. The fluorescence lifetimes (and intensities) were found to be rather constant for the phosphotyrosine peptides (ca. 300 ns in aerated water), while those of the unphosphorylated peptides were as short as 40 ns (at pH 7) and 7 ns (at pH 13) as a result of intramolecular quenching. Owing to the exceptionally long fluorescence lifetime of Dbo, the assays were alternatively performed by using nanosecond time-resolved fluorescence (Nano-TRF) detection, which leads to an improved discrimination of background fluorescence and an increased sensitivity. The potential for inhibitor screening was demonstrated through the inhibition of acid and alkaline phosphatases by molybdate.

A High Throughput Luminescent Assay for Glycogen Synthase Kinase-3β Inhibitors

A high throughput luminescent assay based on the Kinase-Glo() system (Promega, Madison, WI) has been developed for screening against glycogen synthase kinase-3beta (GSK-3beta). Careful optimization of assay parameters allowed us to develop a robust, reproducible, and sensitive assay. Its usefulness has been demonstrated in a high throughput screening run when screening 55,000 compounds. This campaign yielded five chemical classes of hits, including several highly potent GSK-3beta inhibitors.

Assay principle for modulators of protein-protein interactions and its application to non-ATP-competitive ligands targeting protein kinase A

Targeting sites that modulate protein-protein interactions represents an ongoing challenge for drug discovery. We have devised an assay principle, named ligand-regulated competition (LiReC), in an effort to find non-ATP competitive small-molecule regulators for type Ialpha cAMP-dependent Protein kinase (PKA-Ialpha), a protein complex that is implicated in disease. Our assay based on the LiReC principle utilizes a competitive fluorescent peptide probe to assess the integrity of the PKA-Ialpha complex upon introduction of an allosteric ligand. The developed fluorescence polarization method screens for small molecules that specifically protect (antagonists) or conversely activate (agonists) this protein complex. In high-throughput format, various cyclic nucleotide-derived agonists and antagonists are successfully detected with high precision. Furthermore, assay performance (Z'-factors above 0.7) far exceeds the minimum requirement for small-molecule screening. To identify compounds that operate through novel modes of action, our method shields the ATP-binding site and purposely excludes ATP-competitive ligands. These proof-of-principle experiments highlight the potential of the LiReC technique and suggest its application to other protein complexes, thereby providing a novel approach to identify and characterize modulators (small molecules, proteins, peptides, or nucleic acids) of protein-protein systems.

Assay development and data analysis of receptor-ligand binding based on scintillation proximity assay

In this paper, we described the optimization of a generic binding assay to measure ligand-receptor interactions for peroxisome proliferator-activated receptors (PPARs). The assay is based on scintillation proximity assay, in which a protein is coated on scintillant-incorporated beads, and a radiolabeled ligand stimulates the beads to emit a signal by binding to the immobilized protein. An intrinsic binding affinity of unlabeled ligands is determined by competitive displacement of the radioligand. The protein coating and ligand binding are achieved in one step by simply mixing ligands, protein and beads in sequence. No additional steps of pre-coating and washing of beads are required. Protein is captured on beads effectively by electrostatic interactions, thus no affinity labeling of protein is required. In data analysis, ligands are grouped into two classes based on their binding affinities. For tight binding ligands, an equation is derived to accurately determine the binding affinity. Otherwise a general equation applies. This quantitative and high throughput assay provides a tool to screen a large library of molecules in search of potent ligands.

ADP-specific sensors enable universal assay of protein kinase activity

Two molecular sensors that specifically recognize ADP in a background of over 100-fold molar excess of ATP are described. These sensors are nucleic-acid based and comprise a general method for monitoring protein kinase activity. The ADP-aptamer scintillation proximity assay is configured in a single-step, homogeneous format while the allosteric ribozyme (RiboReporter) sensor generates a fluorescent signal upon ADP-dependent ribozyme self-cleavage. Both systems perform well when configured for high-throughput screening and have been used to rediscover a known protein kinase inhibitor in a high-throughput screening format.

High-throughput screening with quantitation of ATP consumption: a universal non-radioisotope, homogeneous assay for protein kinase

A number of assays have been developed for high-throughput screening (HTS) of potentially bioactive compounds. To screen millions of chemical compounds efficiently, the best detection technology prior to initiating HTS must be chosen. Ideally, a non-radioisotope (non-RI), homogeneous method, equivalent to the most reliable assay for a particular target, should be selected as an HTS method. Protein kinases are among the most important classes for drug discovery because they participate in various signaling pathways. Several HTS technologies are available for kinase activity: SPA (Amersham, Piscataway, NJ, U.S.A.), HTRF (CIS-US, Inc., Bedford, MA, U.S.A.), IMAP (Molecular Devices, Sunnyvale, CA, U.S.A.), and Z'-LYTE (Invitrogen, Carlsbad, CA, U.S.A.). The amount of phosphorylated product is detected by different methods in these assays. Recently, Kinase-Glo Luminescent Kinase Assay, a non-RI, homogeneous, adenosine triphosphate (ATP) quantitative kit useful for kinase activity detection, has become available from Promega (Madison, WI, U.S.A.). ATP is a universal substrate for kinases. Thus, the Kinase-Glo assay shows promise for becoming the primary method of determining kinase activity in HTS. We have developed a Kinase-Glo system for cyclin-dependent kinase 4 (Cdk4), and compare its results with those of the filtration method, the most reliable assay for in vitro Cdk4 activity. In addition, the reliability and sensitivity of the Kinase-Glo are discussed.