The Integrated World of Small Molecule Oncology Drug Discovery

In cancer drug discovery, CROs are increasingly a key player in delivering compounds to the clinic

Integrated drug discovery programmes are large, multi-layered work plans that can go all the way from identifying a molecular target suitable for therapeutic intervention to delivery of a compound that is fully ready to be tested in patients. Typical clients are small to medium biotechnology companies that may be a spin off from an academic group. These companies have built the scientific case and sufficient understanding of the target to attract funding for drug development, but they may not have the capabilities or the drug discovery expertise to take their initial findings all the way into the clinic. Contract research organizations, like Charles River, are there to help carry out the process (or parts of the process) on their behalf.

For a deeper dive into integrated drug discovery and how developers can collaborate with CROs to advance their products, we turned to Elizabeth Anderson, PhD, Science Director, Oncology for Charles River Early Discovery.

Eureka: When a client approaches a CRO do they already have a target in hand?

Dr. Anderson: The drug discovery process is broken down into discrete “work packages” and the contents and complexity of these packages depends on the stage of the project. In most cases, the client has already identified the target and what it does before approaching a CRO but sometimes, they come with compounds (rather than targets) that have been pulled out by phenotypic screens. These phenotypic screens identify compounds by monitoring changes in cellular functions (e.g., inhibition of cell proliferation) rather than by altering the biochemical activity of a known target such as an enzyme. An important part of the subsequent drug discovery programme is identification of the target causing these phenotypic changes (known as target deconvolution). In other cases, a CRO might be asked to carry out further target validation, which requires teams of specialists in cellular and molecular biology. Their job is to show that the target is expressed where we think it should be expressed and that if you take it away (or add it back in) you get the expected cellular changes.

Eureka: How hard is it to identify these targets in a disease such as cancer?

Dr. Anderson: Recent advances in understanding cancer biology and the explosion of novel, targeted therapeutic agents for oncology mean that most of the obvious targets have already been worked on. But with the right expertise, CROs are able to work with the less obvious and more difficult targets and are often commissioned to work on those that are particularly problematic or even “undruggable”. They can generate the reagents and understanding of the target’s biology to facilitate the drug discovery process.

Eureka: How significant a role does automation and AI play in finding these targets?

Dr. Anderson: The drug discovery process often starts with a high throughput screen (HTS) which is where robotic automation has been introduced. Sadly, you won’t find R2-D2 or C-3PO roaming the labs, but you will find robotic arms that can dispense hundreds of thousands of compounds into assay plates. The robots are used to help process thousands of assays per week. They don’t work in isolation; a huge team of humans is also required ranging from compound management to structural biologists, who synthesise and purify the target (or the active bit of it) for use in the assays, to chemists who work in Computer Aided Drug Design (CADD) and decide which compounds to include in the HTS, to instrumentation specialists and biologists who devise and optimise the assays. The goal of an HTS is to identify several hits – compounds that bind to the target and have the desired effect. At this point, the compounds may not be very potent, but they should have properties that make them suitable for further development with the potential to be patentable. AI methods such as machine learning and deep learning are being increasingly used by medicinal chemists as they can design and optimise huge numbers of complex molecules and the routes by which they are synthesised.

Eureka: What happens once a promising “hit” has been identified?

Dr. Anderson: Once hits have been identified, scientists work on them to improve their properties to the point where they become “lead” compounds. At this stage, a large project team is put together that encompasses chemistry, biology, structural biology, CADD (and sometimes artificial intelligence or AI) together with specialists in drug metabolism and pharmacology (DMPK). This team instigates a full screening cascade and the design-make-test cycle. The screening cascade or funnel is a sequence of in vitro assays that tests new compounds as they are synthesised by the medicinal chemists and provides information on potency, selectivity, cellular activity and in vitro pharmacokinetic properties. CADD and AI may be used to design new compounds based on the data provided by the cascade to gain an understanding of the relationship between the structure of the compound and its activity against the target. Members of the structural biology group may also be involved at this point to determine the structure of the target in the presence and absence of potential drugs to guide compound design.

Eureka: How long does this hit-to-lead process usually take?

Dr. Anderson: About nine months. By then you should have identified a small number of compounds with different chemical backbones suitable for lead optimisation.

Eureka: Can you describe what lead optimisation looks like?

Dr. Anderson: Lead optimisation (LO) or, as I like to call it, advanced tinkering, is exactly what you think it is. This process takes at least 12 months during which the lead compounds are further optimised using and extending all the knowledge about the relationship between compound structure, activity and PK properties that were gained in the hit-to-lead process. The number of compounds being synthesised and tested increases massively as all the possibilities for improving potency, selectivity, and PK properties are explored. The cascade is also expanded to include regular in vivo PK screening, PK/PD studies, and in vivo anti-tumour efficacy studies. Once some promising candidates have been identified, pharmaceutics will be involved to address formulation and we will start investigating how to scale up the chemistry to produce large quantities of compound (or drug substance) suitable for formulation into the final drug product. Exploratory toxicology will also be initiated. At the end of LO, we should have identified or nominated a candidate molecule that is potent and safe enough to be given to humans. Nomination kicks off the pre-clinical toxicology studies that are carried out in the Safety Assessment arm of CRL and which are needed before regulatory authorities such as the FDA can give the go ahead for first in human clinical trials. Then it’s off to the races!

Dr. Anderson has extensive knowledge of solid and haematological tumor types as well as novel methods of cell and tissue explant culture, biomarker development, validation and quality control of immunohistochemistry, development of PDX animal models for the study of normal and malignant human tissue. She has more than 60 peer-reviewed publications and is currently a Deputy Editor for Breast Cancer Research, a specialist oncology journal with a 2y impact factor of 6.142.

Charles River has a wide number of sites that collaborate on oncology drug discovery including the Early Discovery sites at Chesterford Park, Harlow and Leiden that carry out the in vitro work, as well as in vivo sites in Morrisville, NC and Freiburg, Germany, and a number of strategic partners.