Dreams to Reality
Drug Discovery World
Powerful new tools and methods for drug discovery
Dr Peter Banks, Scientific Director, BioTek Instruments, Inc.
Insights 2015

The Human Genome Project (HGP) was completed in February 2001. The sequencing was conducted by 20 universities and research centres around the globe, took more than a decade and cost almost five billion in 2015 inflation-adjusted US dollars.

At the time, the International Human Genome Sequencing Consortium described an initial gene index which contained 15,000 known genes and about 17,000 gene predictions. The pharmaceutical industry certainly took notice and augmented its target-based approaches to drug discovery with the additional genes from the index.

This period marked the height of industrial scale screening of millions of compounds in biochemical assays using purified proteins. The validation of these proteins as drug targets was typically shallow compared to the depth of the chemical libraries used for screening.

A few years on, target-based approaches had been around long enough for some of the chemical entities that were produced by the screening campaigns and refined by lead optimisation studies to make it into clinical trials. While there were some successes, new small molecule drug approvals per year showed a definite decline relative to the period before the completion of the HGP and numerous pharmaceutical companies complained of declines in R&D productivity. This loss in productivity was mostly due to a lack of compound efficacy caused by poor validation of putative drug target early on in the drug discovery process and poor ADME/Tox characteristics caused by refining binding affinity of compounds.

Five years ago, I wrote an article for this journal that described Pharma’s response to declining R&D productivity and how I thought that building relationships with academia could potentially solve issues in drug discovery through collaboration, particularly with early drug discovery efforts, like drug target validation. Over the past five years, we have seen much collaboration between Pharma and Academia. In 2013, $1.73 billion was spent by industry to fund early stage drug discovery efforts in academia and in the full four years since my original article, 93% of drug-related small molecule compounds and biologics that were associated with patentprotected intellectual property cited academic research1.

Pharma R&D has also transitioned by focusing more on proper target validation and revisiting its past with classical pharmacology methods in the form of phenotypic screening of compounds which is replacing target-based approaches to a certain degree. Unlike target-based approaches, phenotypic assays are not built from a molecular hypothesis of disease, but instead rely on relatively complex models of disease that kinetically probe living cells and tissues. One of the most popular models for phenotypic assays is the use of 3D cell culture methods that purport to mimic living human tissue and thus be a close surrogate to the human patient. BioTek Instruments has embraced these methods and has developed versatile tools to enable not only phenotypic screening, but also mechanism of action studies following on from the screen (Figure 1). Over the next five years, I believe phenotypic screening will continue to supplant target-based approaches in preclinical drug discovery.

But target-based approaches are not going away, they are becoming much more sophisticated due to the ability to rapidly sequence whole genomes at low cost. Today, next generation sequencing (NGS) platforms can sequence human genomes in about a day while shaving off costs by more than six orders of magnitude relative to the original HGP. Pharmaceutical companies are using this technology to sequence exomes of thousands of patients to search for new drug targets for specific diseases due to common gene mutations. Nowhere is this approach more applicable than in the fight for cancer where carcinogenesis typically involves multiple mutations of various genes, including proto-oncogenes, tumour suppressor genes and DNA repair genes. NGS is used for large-scale sequencing of tumour biopsies to identify non-synonymous mutations and determine which genes should be targeted for possible therapeutic intervention.

To my mind, however, functional genomics offers a more powerful approach for target-based approaches in drug discovery, especially with the new gene editing tool CRISPR/Cas9. For about a decade, RNAi has been used in target validation for gene loss of function studies, but it is limited by incomplete gene knockdown and off-target effects that can confound results. CRISPR/Cas9 differs from RNAi in this application area by producing complete loss-of-function through interacting with genomic DNA rather than partial loss-of-function of transcripts. This yields high screening sensitivity and enables the ability to target the entire genome, not just the exome.

Perhaps more importantly, CRISPR/Cas9 is a gene editing tool that allows for alteration of genomic DNA. Thus it is more comparable to zinc finger nucleases or TALENs. The principle differences between CRISPR/Cas9 and these gene editing tools is ease of use and cost. CRISPR/Cas9 is not only demonstrating high utility in preclinical drug discovery, but it has shown the potential for gene and cell replacement therapies. Recently, researchers injected CRISPR/Cas9 into the livers of mice to snip out mutated DNA associated with the disease tyrosinemia (FAH mutation in hepatocytes) and replace it with the correct sequence2. By expansion of the edited hepatocyte population, the researchers were able to reverse the disease phenotype of weight loss. Thus CRISPR/Cas9 holds potential for treating many genetic disorders. This technology will have explosive growth in use in drug discovery and general research over the next five years and beyond. BioTek Instruments is currently exploring new work flows that will enable the use of CRISPR/Cas9 and other functional genomics tools.

Advancing Translational Research for Biomedical Innovation, Battelle Technology Partnership Practice (2015).

2 Hao Yin, H, Xue, W, Chen, S, Bogorad, R, Benedetti, E, Grompe, M, Koteliansky, V, Sharp, P, Jacks, T and Anderson, D. Genome editing with Cas9 in adult mice corrects a disease mutation and phenotype. Nature Biotechnology 32, 551–553 (2014).