Designing small molecules with increased potential for CNS bioavailability

(Laura Chico, Northwestern University)

Importance of addressing bioavailability in CNS drug discovery – the drug needs to be available at the target body tissue(s).

  • Efficacy can be limited if the drug is unable to reach the target in sufficient amounts during the appropriate time window. Can be caused by lack of BBB penetration, metabolism of drug, efflux out of the brain.
  • The molecular properties of the drug influence its absorption, distribution, and interactions with transporters/metabolizing enzymes.
  • Keep in mind that bioavailability does *not* guarantee drug efficacy.

Example: First-generation antihistamines caused undesirable sedative side effects since the drugs were not substrates for P-glycoprotein efflux. Second-generation antihistamines ameliorated this problem with chemical modifications that made the compounds targets for efflux, thus limiting CNS bioavailability.

Genetic polymorphisms of metabolizing enzymes may also impact bioavailability, safety/efficacy of drug.

  • Example: Cytochrome P450 2D6 (CYP2D6) is a major enzyme isoform involved in CNS drug metabolism. There are two phenotypes – “ultra-rapid” metabolizers and “poor” metabolizers. Codeine is metabolized by CYP2D6 to morphine, so the “poor” metabolizer phenotype is associated with decreased efficacy while the “ultra-rapid” phenotype is associated with increased toxicity risk.

Physical properties affect how drugs interact with the body (ADME).

  • Factors: size, solubility, lipophilicity
  • Factors are inter-related – modifying one property can have significant impact on other properties (see FIG. 2)

Lipinski “Rule of 5” – Properties analyses bias efforts towards bioavailable compounds.

  • Poor absorption/permeation is more likely if:
    • molecular weight > 500
    • hydrogen bond donor atoms > 5
    • partition coefficient (LogP) > 5 (measure of lipophilicity)
    • N + O atoms > 10
  • Keep in mind that the Rule of 5 is not a simplistic “hard” filter; it is also not CNS specific.

For CNS drug discovery, the most relevant molecular properties are lipophilicity (LogP), molecular weight, and polar surface area (N and O atoms); these properties affect transport across the BBB.

  • These values can be calculated experimentally or in silico (see resource list for products/companies). In silico may be better for early stage development to estimate values.
  • Prior to purchasing or screening a chemical library, look at the property landscape for CNS relevance.

CNS drugs have a more restricted molecular properties space.

  • PSA discriminates CNS suitable compounds better than LogP.
  • CNS penetrant compounds that are Pgp substrates have higher LogP and MW values than those that are not Pgp substrates.
  • For increased probability of CNS penetrant compounds with low efflux (more restrictive than Lipinski’s Rule):
    • LogP < 4
    • MW < 400
    • PSA < 80

Using properties guidelines can help prioritize CNS drug discovery efforts.

  • Simple filters can narrow down the search to fewer high priority candidate compounds.
  • These guidelines are useful throughout the discovery process, from initial library screening/selection to hit-to-lead refinement.
  • But remember these are guidelines, not rules – move beyond, if there’s a good rationale.

Keep in mind that bioavailability does not guarantee efficacy – other ADME issues are important.

  • Most property analyses focus on one outcome/endpoint, but CNS bioavailability is complex (e.g. need to consider CYP2D6 metabolism in addition to BBB penetrance).
  • The future direction of the field will emphasize properties analysis on multiple outcomes and overlap results. By superimposing analyses, “hotspots” possessing multiple desirable properties can be identified while undesirable combinations can be avoided.
  • These multidimensional properties analyses will help refine “CNS space”
Basics of high throughput screening (HTS): Bridging chemistry and biology >