We've updated our Privacy Policy to make it clearer how we use your personal data.

We use cookies to provide you with a better experience. You can read our Cookie Policy here.

High-Throughput Screening Methods for Drug Discovery

High-Throughput Screening Methods for Drug Discovery

High-Throughput Screening Methods for Drug Discovery

High-Throughput Screening Methods for Drug Discovery

Read time:

Want a FREE PDF version of This Article?

Complete the form below and we will email you a PDF version of "High-Throughput Screening Methods for Drug Discovery"

First Name*
Last Name*
Email Address*
Company Type*
Job Function*
Would you like to receive further email communication from Technology Networks?

Technology Networks Ltd. needs the contact information you provide to us to contact you about our products and services. You may unsubscribe from these communications at any time. For information on how to unsubscribe, as well as our privacy practices and commitment to protecting your privacy, check out our Privacy Policy

Drug development is a long and expensive process. A study reported in the Journal of the American Medical Association found that between 2009 and 2018, the average investment needed for a drug to make it all the way through the development pipeline so that it may be considered for regulatory approval was US $1.3 billion. One reason for the huge expenditure is high drug failure rates. Research published in Nature Biotechnology showed that between 2003 to 2011, 90% of drugs that went on to clinical trials failed to gain US Food and Drug Administration (FDA) approval. Over the years, the pharmaceutical industry has been experiencing a decline in research and development productivity which in turn worsens unmet clinical needs. In this article, we will discuss how the use of different high-throughput screening (HTS) methods can help to accelerate the drug discovery process.

The drug discovery process

Drug discovery typically starts with a hypothesis of a biological mechanism implicated in a disease, which if targeted, may be helpful in treating the disease. Following this, assays are developed and vast libraries of compounds are screened. Screening can either be target-based or phenotypic-based. This is followed by an iterative process whereby identified “hits” are refined to produce “leads”. These compounds are then further optimized before the most promising lead is chosen to be taken forward for preclinical testing and clinical trials.


Due to the sheer number of possible compounds available for screening, pharmaceutical companies have been adopting HTS methods that can screen their chemical libraries rapidly and inexpensively to identify the most promising compounds with activity against a specific biological target. HTS usually involves the use of automation (e.g., liquid handling robots), miniaturization such as 1536-well plate formats and large scale data analysis such as incorporation of artificial intelligence.

“HTS has been a staple of the drug discovery industry for decades. It allows automated profiling of drug activities in very specifically tailored biochemical assays. However, so far, the readouts for these purposely designed assays have been limited to a very small number of targets,” says Dr. Chaoyang Ye who developed a cost-effective RNA sequencing protocol for HTS.

In the following sections, the use of newly developed HTS platforms to support drug discovery will be highlighted.

Fluorescence read-out through imaging

Fluorescence-based assays are a convenient way to visualize biological responses to specific compounds in a high-throughput setting. Based on the fluorescence intensity, the degree of biological activity can also be determined. Fluorescence assays can be designed such that signals are emitted or quenched upon production or elimination of a target molecule. Recently, Cecilia Eydoux, a research engineer at the AFMB laboratory, and colleagues created a fluorescence-based HTS assay for the severe acute respiratory syndrome coronavirus (SARS-CoV) RNA synthesis complex. This complex is made up of protein non-structural-protein (nsp)-7, nsp8 and nsp12 and due to it being highly homologous to SARS-CoV-2, it has the potential to be exploited as a means to design inhibitors of protein synthesis. Making use of known protein interactions, the research team created an assay when upon the association of nsp-7/8 to nsp-12 (a RNA dependent RNA polymerase), fluorescence dye would intercalate the double-stranded RNA to produce fluorescence signals.

The researchers validated their assay by screening against 1520 FDA-approved compounds from the Prestwick Chemical Library®. This compound library was selected for its high chemical and pharmacological diversity and known bioavailability and safety in humans. The screening was performed with the help of an automated liquid handling robot and fluorescence read-out with a plate reader. Based on a cut-off of 30% inhibition, the calculated hit rate was 3% with drug candidates belonging to the anthracyclins and tetracyclins families.

Your Guide to Luminex Multiplex Immunoassays

Multiplexing allows multiple biological target analytes to be simultaneously examined and quantified in a single sample. It’s a technique widely used by scientific researchers of all disciplines and provides many benefits including maximizing the use of limited samples while minimizing experimental variability. Download this guide to learn more about how to accurately detect natural proteins, precision and reproducibility and best practices and techniques. 

View Guide

Mass spectrometry

A limitation of chemiluminescence and fluorescence readout is that it is an indirect measurement of biological activity. Dr. Andreas Luippold et al. recently developed a HTS method using mass spectrometry. The authors argue that mass spectrometry is a label-free technique and provides evidence of direct substrate to product conversion which is more physiologically relevant for drug discovery. This method also has reduced risk of compound interference which can lead to false positives and negatives, and it does not require tailored signal mediators to amplify signals, such as in fluorescence readouts. However, so far, it has been challenging to integrate mass spectrometry with liquid handling robots and the buffer used in a mass spectrometer may also be incompatible with preserving protein enzymatic activity.

Luippold and co-workers configured matrix-assisted laser desorption/ionization (MALDI)-time of flight (TOF) equipment and fitted it with a liquid-handling robot to make it compatible with 1536-well plates (or 4 x 384-well plates). This enabled robust and reproducible transfer of samples from assay plates and matrix solution onto MALDI target plates, allowing them to screen a library of ~4900 inhibitors against protein tyrosine phosphatase 1B – a known therapeutic target for the treatment of diabetes and obesity. The authors also performed a head-to-head comparison of their method against a biochemical assay and automated spotting technique. To generate a million data points (i.e., screening for a million compounds/molecular fragments), the MALDI-TOF HTS method would take just 5.1 days instead of 11 days for biochemical assay and 7 days for automated spotting technique.

True High-Throughput Analysis of Transporter Activity Screens

Two key ADME (absorption, distribution, metabolism and excretion) properties used to predict the in vivo performance of a drug candidate are its ability to permeate a cell membrane and its potential to act as a substrate for transmembrane transporter proteins. Download this app note to discover a system that is sensitive, reproducible and accurate. 

View App Note

Cell systems

Diseases affecting the central nervous system are on the rise but high failure rates among CNS drug candidates remain, due to their inability to pass through the blood–brain barrier (BBB). The success rate for CNS drugs is only 7% and it takes roughly 10.5 years to develop therapeutics in this area, which is significantly longer than other diseases. Dr. Elisa Moya, researcher at the University of Artois, and colleagues recently developed a miniaturized in vitro BBB model to study the impact of drug toxicity and BBB permeation rates for early-stage drug discovery before embarking expensive preclinical studies.

Moya highlights some of the advantages of HTS in this context: “In the BBB in vitro field, at the preclinical study phase, the development of HTS is vital for a number of reasons. For example, it enables researchers to evaluate a large number of drugs/compounds rapidly and is better adapted to the high yields needed in early-stage drug discoveries for CNS compound discovery.” She also notes that their method reduces material and plastic waste and allows researchers to screen out weak candidates early on, thus reducing the amount of further testing required using in vivo models.

“Based on a patented and well-established human in vitro BBB model created by the University of Artois, we developed a miniaturized and automated replicate. The replicate uses a 96-well system format combined with automated technology,” notes Moya.

The team created human BBB in vitro models using endothelial cells derived from CD34+ hematopoietic stem cells isolated from human umbilical cord blood and seeded different numbers of cells into 96-well plate with well inserts. Pericytes, a main cell type of the BBB, were also seeded at the bottom of the well plates. The authors then assessed BBB integrity by analyzing the transport of the fluorescence paracellular marker, sodium fluorescein, past the endothelial cell layer. When Moya and co-workers compared the BBB integrity which are manually and robotically created, they were comparable in different well plate formats, suggesting that a tightly packed network was formed. Confocal imaging also confirmed the presence of adherens junction proteins such as cadherin and tight junction proteins like claudin-5 in BBB seeded by robots.

Next, the team screened a set of drugs such as bupropion whose values for the free/unbound fraction ratio of the brain/plasma are known in humans in vivo. The results show that the free/unbound fraction ratio of the respective drugs tested using the miniaturized BBB model was similar to the in vivo data in humans, with strong correlation (R2 = 0.8808). When Moya et al. tested the BBB integrity of their miniaturized models, they also found that neurotoxic drugs like rotenone greatly disrupted the BBB while BBB integrity was unaffected by non-neurotoxic compounds like acetaminophen, further strengthening the physiological relevance of their model.

“We’ve shown that this in vitro BBB model is capable of achieving predictive results and is suitable for the HTS of drug compounds,” Moya adds. While the utility of their model has been demonstrated, it is important to note some key limitations, including the inclusion of only two cell types and the lack of a dynamic flow environment.

Automation of High-Throughput Flow Cytometry

Integration of your flow cytometer into an automated workcell will increase throughput, enable 24/7 operation and provide consistent results. Download this whitepaper to explore what can be accomplished through automated high-throughput flow cytometry, considerations for automating flow cytometry and how to optimize high-throughput workflow and data analysis. 

View Whitepaper

RNA sequencing

RNA sequencing is a powerful method to understand the effects of drugs. This approach uses the RNA transcriptome as a proxy, but the techniques currently available are typically low throughput, costly and manpower intensive. Dr. Chaoyang Ye and colleagues developed a cost-effective RNA sequencing protocol for HTS named DRUGseq by forgoing lengthy RNA purification steps and employing a multiplexing strategy. After drug treatment, cells are lysed, and mRNAs are directly reversed transcribed (RT). The team also introduced specific barcodes into RT primers which enabled them to pool cDNAs from individual wells and reduce labor needed in multi-well processing.

To further decrease the time and costs involved, Ye and colleagues also tested the effects of reducing the read depth on the number of genes detected and the accuracy of capturing differentially expressed genes after drug treatment. DRUGseq libraries were sequenced at 2 and 13 million reads/well compared to the conventional 42 million reads per sample, and even at shallow read depth, gene expressions were consistent across wells.

The team next applied DRUGseq to screen a library of 433 compounds with known targets. The number of differentially expressed genes detected by DRUGseq correlated well with data from the Connectivity Map, a large dataset of cell perturbations due to pharmacological treatments, despite using different technology platforms.

Finally, the authors used DRUGseq to compare the effect of CRISPR knockout and compound inhibition of a well-validated compound, cycloheximide, against RPL6. They found that while CRISPR treatment reduced the level of RPL6 mRNA, cycloheximide did not. However, the transcriptomic profiles of both treatments were found to be more similar in mechanism and were clustered together when compared to DMSO treatment and non-targeting control samples. The authors concluded by highlighting the advantages of DRUGseq over competing platforms such as PLATE-seq which have lower throughput and computational bias.

“This work can enable transcription profiling of thousands of genes across hundreds to thousands of wells cost-effectively and it is compatible with existing compound screening automations in industrial settings. The challenge is that it requires investment in screening automation upfront for it to be truly cost effective. Transcript capture efficiency could also be improved by testing various reaction conditions, which will enhance detection sensitivity and lower sequencing cost per well,” says Ye.

He continues, “We hope that more labs will adopt and improve this technology in the future. As a continuation of the original study, a recent manuscript has now been submitted to bioRxiv by my colleagues, which more rigorously tested out the platform and has made the analysis workflow available to the public.”


A continual decline in drug discovery success rate would lead to unmet clinical needs and more costly drugs as pharmaceutical companies seek to recover their research and development investments. HTS is a strategy to screen vast compound libraries with speed and cost effectively. Traditional methods such as fluorescence readouts and cell culture models can be integrated with automated imaging systems, liquid handling robots and miniaturization, but often depend on indirect measurement of biological activities whereas mass spectrometers can be integrated into HTS workflows to facilitate direct measurement of molecular fragments and activity. Finally, with rapid progress in sequencing technology, RNA sequencing with HTS capacity is expected to become a powerful tool to enable novel compound mechanistic studies, compound repurposing and identification of genetic transcriptional networks with CRISPR gene editing.


Meet the Author
Andy Tay, PhD
Andy Tay, PhD