Drug Discovery Stage definitions

Whilst the Drug Discovery process is continuous and can vary depending on target it is often useful to split the process into stages with key milestones and targets clearly defined and met before a project moves from one stage to the next.

Target Identification

Demonstration that the potential molecular target is present in human or in parasite/infective agent. At this point it is usual to identify an assay to evaluate modulation of the target and determine the feasibility of screening.

The Centre for Therapeutic Target Validation platform brings together information on the relationships between potential drug targets and diseases. The core concept is to identify evidence of an association between a target and disease from various data types.

A target can be a protein, protein complex or RNA molecule, but we integrate evidence through the gene that codes for the target. In the same way, we describe diseases through a structure of relationships called the Experimental Factor Ontology (EFO) that allows us to bring together evidence across different but related diseases.The platform supports workflows starting from either a target or disease and presents the evidence for target – disease associations in a number of ways through association and evidence pages.

DisGeNET is a discovery platform integrating information on gene-disease associations (GDAs) from several public data sources and the literature doi.

The current version contains (DisGeNET v3.0) contains 429111 associations, between 17181 genes and 14619 diseases, disorders and clinical or abnormal human phenotypes.

Target Validation

Proof that modulation of the identified target in a model system has the desired impact on biological activity and can be linked to therapeutic utility. This might be achieved by identification of genetic mutations in the human population for example CCR5 mutations and HIV, gene knock-out studies in mice, or siRNA experiments, for example siRNA and neuropathic pain. There may also be known small molecules that exert the desired effect but perhaps with a sub-optimal profile. It may be possible to identify potential liabilities at this point or key off-targets issues. Screening assay in place together with functional screen. If you are using literature data for target validation your mantra should be "Trust but verify".

This is an absolutely critical step, almost everything else can be fixed.

Hit Identification

Identification of the starting point for the drug discovery program, this might be a natural ligand or a molecular from the literature, however it is more likely to be derived from a screening campaign. A wide variety of screening technologies are available from high-throughput screening, fragment-based screening to virtual screening.

“The single most important factor determining the likelihood of success of a project is the quality of the starting lead”, Anon

Whilst most screens will identify many potential hits the critical step is to validate the hit, the degree of validation may depend on the screening technology (e.g. fragment hits may be too weak to show functional activity) but the list below gives some ideas.

• Activity confirmed with a fresh sample in which the structure has been confirmed by spectroscopy and the purity determined.
• No obvious undesirable chemical motifs or known toxicophores
• Many screening assays simply measure binding, it may be desirable to evaluate functional activity.
• Selectivity against key undesired targets
• Activity in human target, and activity in species used for in vivo models and species likely to be used in safety studies.
• Evidence of structure activity, activity of near neighbours
• Synthetic tractability/ patent position

Lead Identification

Demonstration that one or more of the small molecule hits has the potential to provide a series that could be optimised to afford a candidate. The following list provides some thoughts

• Sub micro molar activity in primary binding or functional assay
• Demonstrable selectivity (>10-fold) against key undesired targets
• Activity in cell-based assays, with SAR tracking activity in primary assay
• Defined structure-activity from analogues, perhaps linked to X-ray structure or other biophysical information
• Representative examples evaluated in in vitro ADME assays (CYP inhibition, turnover in microsomes, Cell penetration, plasma protein binding)
• Representative examples evaluated in in vitro hERG assay
• Lead compound(s) ADME profiles in vivo determined, include species to be used for in vivo efficacy model
• Tool compounds identified and profiles suitable for establishing in vivo assays
• Evidence for activity in the in vivo models even if using sub-optimal dosing regimes. This could be i.v. infusion, implanted mini pumps, s.c. dosing or multiple oral doses.
• Synthetic tractability/ patent position confirmed and chemistry strategy in place to address any short-comings

All key in vitro assays in place and in vivo efficacy models identified, strategy for predicting clinical dose in place (Remember an in vivo efficacy model in which you block the effect of an applied agonist can obviously be manipulated by changing the dose of agonist).

Lead Optimisation

Optimisation of one or more of the compounds from the lead series identified into a clinical development candidate, focus on:-

• Potency sufficient for activity at a clinically relevant exposure
• Selectivity in vitro optimised, in vivo selectivity evaluated. There may be occasions where efficacy requires 100% occupancy at tough but the off-target liability is seen at 10% occupancy at Cmax.
• Efficacy demonstrated in relevant in vivo models, ideally more than one, using a dosing regime that might be used in the clinic.
• Suitable in vivo ADME profile for chosen route of administration, if appropriate evaluate different vehicles.
• Distribution investigated
• PK/PD assessment, for some targets you might need to maintain 100% occupancy at trough levels in plasma.
• Routes of metabolism identified
• Viable synthetic route capable of producing multigram amounts of drug in suitable physical form. Solubility in dosing vehicle determined. Also route to labelled compound.
• Identify biomarkers suitable for use in the clinic, for CNS targets a PET ligand may be useful

Some studies may require labelled compound.

Candidate Selection

All compounds are unique and this list should only regarded as a guideline. See also Preclinical Checklist

• Activity in two in vivo preclinical efficacy models, including plasma level efficacy profile. In most cases this will need to be < 10 nM in vitro and < 10 mg/kg p.o. in vivo. Remember an in vivo efficacy model in which you block the effect of an applied agonist can obviously be manipulated by changing the dose of agonist.
• A list of all completed and ongoing preclinical ADME and human biopharmaceutical studies. A one page summary is very useful.
• Ensure the assay method that was used in quantitation of the compound in plasma and/or urine samples from laboratory animals is valid for humans. Is the method validated as per the FDA criteria?
• Complete pharmacokinetic parameters following oral and intravenous administration in two species, ideally species to be used in proposed safety studies.
• Ensure dosing/formulation/route of administration will allow adequate coverage in proposed safety studies.
• For intravenous administration check for haemolysis using the proposed vehicle.
• Check the distribution of the compound in laboratory animals, including brain and binding to melanin containing tissues. Is there a long-lived component in plasma and tissues? This will probably require labelled compound.
• Identify the routes of excretion and mass balance in laboratory animals following oral and intravenous administration of radiolabeled compound.
• Identify the pathways of metabolism of the compound in laboratory animals. Does metabolism lead to the formation of pharmacologically active metabolites?
• Define the metabolism of the compound in vitro using liver sub-cellular fractions from laboratory animals and human, including metabolite characterization and kinetics of metabolite formation.
• Identify the enzymes responsible for the metabolism of the compound in human, are they different in toxicological species?
• Complete inhibition profile of the compound against CYP enzymes and Pgp. Is the compound a pre-incubation time-dependent inhibitor of CYP enzymes? Ideally no activity below 10 uM
• The propensity of the compound to generate reactive intermediates. What is the degree of irreversible binding to liver microsomal proteins in vitro and plasma/liver proteins in vivo?
• Does the compound cause CYP induction? Ideally no activity below 10 uM
• Check the plasma protein binding and blood/plasma ratios of the compound in laboratory animals.

• Model plasma profile versus efficacy and off-target activity/toxicity. Remember for some therapeutic targets you may need to maintain 100% occupancy at trough levels in plasma.

• AMES Test (up to maximum solubility), HERG Binding, if any binding seen below 10 uM, look for QT prolongation in vivo .

• Selectivity screen, PanLabs or similar.
• Define the pharmaceutical properties of the molecule. Melting point (if possible Differential Scanning Calorimetry or Thermal Analysis), stability, solubility, purity, pKa etc. for both free base and salt if appropriate. Any evidence for polymorphs? Define formulations.
• Stability at high temperature (30-60oC) and high humidity.
• Ensure freedom to operate, including all intermediates in synthetic route.
• Ensure a chemically tractable synthesis capable of producing enough material at least for safety studies, preferably a potential manufacturing route.
• Competitive position, both for ligands for the same molecular target but also compounds addressing the same therapeutic target.

Updated 2 August 2018