Chemical Genomics Facilitates Finding Targets

Important New Mechanisms and Protein Targets Have Already Been Uncovered

Chemical genomics, using small molecules to elucidate gene, protein, and pathway functions, in a cellular context, by modulating activity of gene products, whether known molecular targets or gene products of unknown identity, is a powerful method for target validation and complements other methods since it operates on the protein, rather than the gene itself or an mRNA encoded by it.

“By one definition, we do chemical genomics every time we run a high throughput screen of our small molecule compound library against a specific target,” states David J. Matthews, Ph.D., senior director, structural biology, Exelixis (S. San Francisco).

“During target-directed drug discovery, you often discover unexpected effects of potential therapeutics, which can either enhance or diminish a drug’s activity. They may be due to unknown responses to inhibit the cognate target, or to unanticipated cross-reactivity with other targets. Either way, it’s important to understand their causes.”

Forward chemical genomics explores biological pathways. Small molecules interact with and alter a target, phenotypically changing the target’s gene product. With the target perturbed, and the phenotype altered, that pathway is explored to see how the phenotype changed. Reverse chemical genomics helps elucidate structure-based activity.

For example, purified proteins are tested against thousands, to upwards of a million, small molecules, netting tens to thousands of compounds. After these compounds are subjected to stringent conditions, those still bound are examined to reveal how these small molecules have altered the bound target.

“Until relatively recently, the predominant approach in pharma was target-independent drug discovery to look for pharmacological effects,” states James Inglese, Ph.D., director, biomolecular screening and profiling, at the NIH’s Chemical Genomics Center (NCGC).

“Reverse chemical genomics is essentially what occurs in pharma today. Economics dictates that targets be well validated before they are screened to show solid evidence of these targets’ substantial commercial potential. These identifications provide valuable support to companies’ franchises, and knowledge of the target is helpful in developing improved second-generation medicines.”

Nationwide Screening Centers

The NCGC, the first component of the NIH Roadmap Molecular Libraries Screening Center Network, nationwide screening centers, will produce biomodulators to study gene, cell, and organism functions. Assays submitted to the NCGC will be screened against a diverse collection of more than 500,000 chemical compounds, including natural products, cellular metabolites, and biosynthetic intermediates.

The goal of this research tool is to maximize chances of binding a chemical compound to a target and achieving a desired therapeutic effect via this systematic screening. “Data and tools from chemical genomics efforts like ours will increase physiologically relevant information available about a prospective drug target, and will show that small molecules can be used to manipulate it. We hope these methods enable drug discovery in ways other validation techniques do not,” says Dr. Inglese.

Petra Ross-MacDonald, Ph.D., senior research investigator, applied genomics, Bristol-Myers Squibb (BMS; Princeton, NJ), feels that academics will integrate and interrogate this information in different, though no less valuable, ways from industry.

“An industry scientist would say, That’s a dirty compound because it hits in three diverse assays,’ while an academic would ask, Why do those compounds block that set of channels and induce apoptosis?'”

Academics will also do different assays compared with industry. “The former will probably perform more cell-based pathway screens; whereas industry favors molecular target screens. Combining this information could be powerful for discovering therapeutic compounds. No matter how good the compound is at producing the desired therapeutic effect, there are strong reasons for knowing the molecular target, and few drugs make it through development without that knowledge,” Dr. Ross-MacDonald says.

“Ideally, you want to know all of the potential targets a compound could hit in a cell at relevant concentrations,” notes Kevin Fitzgerald, group leader, high content target validation/emerging technologies at BMS.

“Chemical genomics methods have uncovered several important mechanisms and protein targets for compounds with unique therapeutic activities in animal models, which turned out to have different molecular targets from what was originally thought.

“Each time an effort like this is successful, there is not only a new therapeutic target, but sometimes a whole new therapeutic pathway,” states Fitzgerald. If the original compound shows efficacy, there is a lead compound series and the ability, with the newly discovered target, to proceed through structure activity relationship assays.

BMS is combining RNA interference (RNAi), compounds, and over-expression tools with high-content screening. Fitzgerald notes that mechanistically, RNAi and small molecules each act differently in vivo. “There is a big difference in activity and kinetics between an RNAi, which typically will result in removal, over 48 hours, of an entire protein from a cell, and a compound, which within an hour, will inactivate a single site on that protein.”

An RNAi to a typical transmembrane tyrosine kinase not only removes tyrosine kinase activity of the protein, but might disrupt several protein complexes bound to other areas of the receptor, Fitzgerald explains. In contrast, the small molecule causes loss of only kinase activity, preserving the other protein complexesan important difference when interpreting a phenotypic outcome.

Fitzgerald believes that chemical genetics, genomics, high content screening, and proteomics are making it easier to find targets and may lead to a resurgence of cell-based screening techniques. “A major reason why cell-based screens fell out of fashion was because there was no easy way to find the target for compounds derived from such a screen.

“We screen millions of compounds against an isolated, purified target hoping that either inhibiting or activating that target in the complex environment of a cell will have therapeutic activity. In the past, many of the best compounds, discovered by older’ cell based assays, hit multiple therapeutic targets.

“These compounds were effective primarily because they hit multiple targets, counteracting redundant networks built into living organisms. Methods used today for screening against purified individual targets would likely never find those compounds,” Fitzgerald states.

“The limiting factor may be how physiologically relevant an assay can be when it’s crammed into a well,” he adds.

Identifying New Targets

David Evans, Ph.D., head drug discovery, Translational Genomics Research Institute (TGen; Phoenix, AZ), uses both siRNA and small molecule screening approaches to identify new targets which, when silenced by siRNA, improve the response to existing therapeutics, or prove to be novel targets for therapeutic development and intervention.

“Global RNAi phenotypic profiling is a powerful approach for finding context-dependent gene targets, which represent vulnerable points for therapeutic intervention,” notes Dr. Evans.

One specific cellular context his group has modeled in vitro is therapeutic response. When a cell is exposed to a drug or chemical agent, cell response is determined by the status of key genes functionally involved in regulating that response. By systematically knocking down individual genes in parallel, they can empirically identify genes causally involved in controlling or regulating the phenotype following chemical exposure (the chemotype). “We call this functional chemogenomics.”

“Synergies exist using chemical screening with RNAi screening, for examining the effects of gene silencing in the entire druggable genome in a single experiment,” Dr. Evans says.

siRNA knockdowns identify gene targets essential for a particular phenotype (e.g. cell viability). This observation is complemented by using compounds with known target specificity. This approach, combined with pathway analysis, allows rapid identification and verification of an individual target’s or pathway’s importance in a diseased cell.

A second approach reverses the process by examining synergies between silencing a gene using RNAi with combined exposure to a known compound. “These experiments allow us to identify novel targets and pathways that synergize with the therapeutic treatment to increase phenotypic response, and these targets become valuable starting points for a drug discovery initiative,” Dr. Evans says.

Other high throughput methods determine disease relevance of these targets to correlate the importance of the gene with disease stage. For example, tissue microarray analysis helps determine the importance of the gene product across a variety of tissue samples, where the disease stage has been classified by a pathologist.


Quantitative real-time polymerase chain reaction (QRTPCR) validates that the gene of interest is being knocked down, while microarray platforms can monitor gene expression of other genes. Integrating data from these platforms is key to understanding the importance of genes in contributing to the disease, and whether or not inhibiting the gene products by small molecules will yield a valid therapeutic.

TGen will use its high throughput RNAi platform to discover drug response genes that can be targeted to improve existing therapies and find novel points of vulnerability for developing new drugs.

The organization is looking to collaborate and partner with pharmaceutical and biotech companies to use this approach for target identification/validation and to discover functionally relevant genes that can help stratify patients in clinical trials for particular drug regimens.

Target- and Chemistry-based Technologies

Acadia Pharmaceuticals’ (San Diego) massively parallel drug discovery platform combines target- and chemistry-based technologies. “We focus on chemistry early on to speed up drug discovery,” states Uli Hacksell, Ph.D., CEO.

Gene families the company explores include: G-protein coupled receptors (GPCRs), nuclear receptors (NRs), and tyrosine-kinase-linked receptors, which represent the largest gene families targeted by known drugs.

GPCRs, the most abundant and therapeutically relevant receptor class (about 2% of the human genome codes for GPCRs), are important in many physiological processes, and about half of all prescription drugs bind to them. Acadia has identified novel chemistries for more than 125 distinct GPCR and NR targets.

The company uses its Receptor Selection and Amplification Technology (R-SAT) cell-based assay platform, in which genes are transferred to cultured cells, to study interactions between small molecules and drug targets.

Functional activity of potential drug targets are evaluated via signal transduction pathways that lead to cell growth. Drug effects are detected as changes in color or fluorescence. R-SAT integrates high throughput pharmacology, profiling, and screening, and the company uses the platform in two ways.

First, it profiles its collection of reference drugs, primarily consisting of currently and formerly marketed central nervous system drugs, over a range of assay targets to help identify those targets responsible for clinical effects of the drugs.

In the second approach, the platform screens large numbers of targets against a proprietary library of more than 300,000 small molecules to identify the most attractive target-specific chemistries as starting points for drug programs.

Acadia used the first approach to discover ACP-104, a novel antipsychotic drug with cognitive benefits, now in Phase II clinical trials to treat schizophrenia, as well as to clinically validate targets for the rest of its drug programs.

It used the second approach to discover ACP-103, a selective 5-HT 2A receptor inverse agonist, being developed as an adjunctive therapy for schizophrenia and treatment-induced dysfunction in Parkinson’s disease, and to discover the other drug candidates in the company’s pipeline.

Stepping back and reflecting on his company’s strengths, Dr. Hacksell remarks, “despite advances in genomics tools to validate targets, ultimately, among the biggest challenges in drug discovery are gleaning important pharmacological information from in vitro and in vivo experiments.”


Mary Addonizio is a freelance writer and program manager.