Optimising the Physicochemical Properties of Lead Compounds

Lead optimisation is a defining point in a drug discovery project. To this point, time and resources have been dedicated to virtual and in vitro screening to identify a chemical space that fits predefined project requirements. Requirements include affinity towards the target at dose responsive concentrations in the low (M range); freedom from intellectual property; chemical tractability; reasonable solubility in water (>100 μM); absence of acute cytotoxicity; and compliance with drug-likeness filters such as Lipinski’s ‘Rule of 5’ (1). While these help focus the project from millions of compounds to a manageable handful of well-characterised molecules, there is generally still room for improvement. The ultimate objective of lead optimisation is to make slightly modified analogues – using the same underlying criteria used in early lead discovery – that will ultimately provide drug candidates with more realistic chances of completing development (see Figure 1). In lead optimisation, scientists strive to simultaneously improve potency and drug-like properties of a lead compound while minimising liabilities. Off-target activity is curtailed to avoid side effects, and the physicochemical and metabolic properties are driven toward reasonable in vivo pharmacokinetics (PK), pharmacodynamics (PD), and absorption, distribution, metabolism and excretion (ADME) values. Optimisation is carried out through empirical (trial-and-error) modification of the structure, and/or using structure-based design when sufficient information is available. Usually, improvement efforts are focused on one parameter at a time. It is important, however, that the comprehensive suitability of a compound as a drug be respected. This can be trickier for purely trialand-error approaches, where structure-activity relationships are more easily explored one parameter at a time, leading to many syntheses to investigate each individual concern.