Elsevier

Drug Discovery Today

Volume 8, Issue 20, 15 October 2003, Pages 927-933
Drug Discovery Today

Review
In silico prediction of blood–brain barrier permeation

https://doi.org/10.1016/S1359-6446(03)02827-7Get rights and content

Abstract

This review examines the progress that is being made towards the in silico prediction of brain permeation. Following a brief introduction to the blood–brain barrier, the datasets currently available for in silico modeling are discussed. Recent developments in in silico models of brain permeation are summarized in the context of the current state of the art in prediction accuracy. An analysis of recent models is presented, focusing on what such models reveal about the molecular properties that determine brain permeation. The review concludes by presenting the current key issues in this area of research, noting in particular, the paucity of brain permeation data available for modeling. Finally, possible future directions are suggested.

Section snippets

Introduction to the blood-brain barrier

As its name suggests, the blood–brain barrier separates the brain and central nervous system (CNS) from the bloodstream. Clearly, for the great majority of drugs aimed at CNS targets, this barrier must be crossed for a therapeutic effect to be exerted, the only exceptions being compounds delivered by invasive or intranasal routes. Conversely, for non-CNS targets, passage across the BBB could lead to undesirable side effects and so should be minimized. One of the distinguishing features of the

Datasets available for in silico modeling

Datasets of sufficient size and quality are required to build predictive models. For the prediction of BBB permeation, there are various types of dataset that are available.

Types of in silico model available and current leaders in prediction accuracy

Based on either CNS+/CNS or logBB datasets, various kinds of in silico prediction methods have been developed over the last decade or so (reviewed in Refs 9., 10., 11.). Here, some recent developments are highlighted.

What do logBB QSAR models tell us?

An examination of the descriptors (and their associated coefficients) that feature in the logBB QSAR models (summarized in Table 2), provides an insight into the molecular properties that determine brain permeation, a brief discussion of which follows.

Examples

To illustrate the aforementioned discussions with some brief examples, Table 3 collates three compounds that span the range of brain permeation, together with several computed physicochemical properties.

Trifluoroperazine is a good brain permeator (logBB = 1.44) and an examination of the calculated properties shows that the MW, PSA and logD are within the guidelines as explained earlier. Also, the Norinder ‘rule’ of ClogP – (N + O) gives a value of 2.19, which predicts a logBB >0. The pKa value

Conclusions and future directions

Over recent years, there has been a great deal of work seeking to generate predictive models for brain permeation. The few datasets that are available have been studied using a variety of molecular descriptors and statistical methods. State-of-the-art approaches seem able to achieve >80% correct classifications based on CNS+/CNS data and predictions on small logBB test sets that approach experimental error (0.3–0.4 log units). Helpful insights into the molecular determinants of passive

Acknowledgements

We would like to thank Colin Bright (Argenta Discovery, http://www.argentadiscovery.com), who provided helpful advice on BBMEC assays. The comments of the anonymous referees were also much appreciated and contributed significantly to the development of this review.

References (33)

  • N.J. Abbott

    Delivery of imaging agents into brain

    Adv. Drug Deliv. Rev.

    (1999)
  • W.M. Pardridge

    Drug and gene targeting to the brain with molecular Trojan horses

    Nat. Rev. Drug Discov.

    (2002)
  • J.A. Gratton

    Molecular factors influencing drug transfer across the blood-brain barrier

    J. Pharm. Pharmacol.

    (1997)
  • F. Atkinson

    Lipophilicity and other parameters affecting brain permeation

    Curr. Med. Chem.

    (2002)
  • D.E. Clark

    Prediction of intestinal absorption and blood-brain barrier permeation by computational methods

    Comb. Chem. High Throughput Screen.

    (2001)
  • H. van de Waterbeemd

    Estimation of blood-brain barrier crossing of drugs using molecular size and shape, and H-bonding descriptors

    J. Drug Target

    (1998)
  • Cited by (433)

    • Novel xanthone derivatives as potent sirtuin 2 inhibitors

      2024, Bioorganic and Medicinal Chemistry Letters
    • Halide counterions in FDA-approved pharmaceutical salts

      2023, Journal of Drug Delivery Science and Technology
    View all citing articles on Scopus
    View full text