Computational Prediction of Blood-brain Barrier Permeation
Introduction
At the turn of the century, there were an estimated 35 million people aged 65 or over in the USA. This figure is expected to increase to 71.5 million by 2030 as the “baby boom” generation joins this segment of the population [1]. Similar trends are anticipated elsewhere in the developed world. This demographic shift is having profound implications for healthcare, given that disorders of the central nervous system (CNS) increase markedly in frequency after the age of 65. Consequently, CNS drugs currently represent the fastest growing segment of the pharmaceutical market and are predicted to account for 20% of blockbuster sales by 2007. There is thus a clear incentive and challenge for the pharmaceutical industry to discover and develop novel therapeutics to meet this burgeoning medical need.
Section snippets
The blood-brain barrier
For a drug to exert a therapeutic effect at a CNS target, it must be able to cross from the systemic circulation into the CNS. There are two interfaces at which this may occur: the blood-brain barrier (BBB) and the blood-cerebrospinal fluid barrier. Given that the surface area of the former (some 20 m2 [2]) is approximately 1000 times greater than that of the latter, the BBB represents the primary interface for solute exchange between the CNS and the systemic circulation [3]. There are two
Experimental blood-brain barrier permeation data
In this section, a brief overview will be given of the main types of experimental blood-brain permeation data that are available for predictive modeling. A more detailed review of the in vitro and in vivo methods used to generate such data has been published elsewhere [13].
Computational estimation of blood-brain barrier permeation
The computational models for BBB permeation that have been developed in recent years can be grouped into three classes. First, there are simple “rules of thumb” that have been derived by examining the molecular properties of compounds that do and do not cross the BBB. Second are classification models that typically predict whether or not a compound is a BBB permeator. The final class comprises models predicting continuous values of BBB permeation based on either logBB or logPS data.
Molecular determinants of blood-brain barrier permeation
From the modeling studies reported in this paper and others conducted over the years, a picture is gradually emerging of the molecular determinants of BBB permeation. Here, a brief overview will be given – more details can be found in [32].
Current issues and future directions
Probably the largest obstacle in the path to better predictions of BBB permeation is the paucity of in vivo brain permeation data, especially logPS data. For robust and reliable models, data will be required for a large number of drug-like compounds that are diverse in structure and property space and that span a wide range of experimental response. It is unlikely that such a compound collection will come into being by chance – what is required is the bespoke generation of data for modeling. In
Conclusions
The prediction of BBB permeation has developed over the last decade into a fascinating area. Over that time, there has been gradual progress in the accuracy of prediction and growing insight into the molecular determinants of BBB permeation that should aid in drug design. Early signs suggest that the current generation of models is being applied to medicinal chemistry projects with some success. In the immediate future, the acceptance of such models by medicinal chemists and other users will be
References (61)
- et al.
J. Pharm. Sci.
(2003) - et al.
Adv. Drug Deliv. Rev.
(2003) - et al.
Ann. Allerg. Asthma Immunol.
(2004) - et al.
Adv. Drug Deliv. Rev.
(2002) - et al.
Eur. J. Med. Chem.
(2001) Drug Discov. Today
(2004)Drug Discov. Today
(2004)Drug Discov. Today
(2004)- et al.
Drug Discov. Today
(2003) - et al.
Vasc. Pharmacol.
(2002)