QSAR and ADME

https://doi.org/10.1016/j.bmc.2003.11.037Get rights and content

Abstract

The prediction from structure of ADME (absorption, distribution, metabolism, elimination) of drug candidates is an important goal to achieve since it can considerably reduce the cost of drug development. Using our database of 10,700 QSAR, we are now reaching the point where we can make many useful comparisons that illustrate how ADME is a practical way to describe the way organic compounds react with living systems. We also show that Caco-2 cells are useful to model absorption, but the most generally useful parameter is the octanol/water partition coefficient. It should be noted, however, that in our opinion, an in silico prediction of ADME is still a long way in the future.

The prediction from structure of ADME (absorption, distribution, metabolism, elimination) of drug candidates is an important goal to achieve since it can considerably reduce the cost of drug development. Using our database of 10,700 QSAR, we are now reaching the point where we can make many useful comparisons that illustrate how ADME is a practical way to describe the way organic compounds react with living systems. We also show that Caco-2 cells are useful to model absorption, but the most generally useful parameter is the octanol/water partition coefficient. It should be noted, however, that in our opinion, an in silico prediction of ADME is still a long way in the future.

Introduction

The latest buzzword in drug development is ADME (absorption, distribution, metabolism, elimination). Except for metabolism, it is possible to make some useful predictions from structure that can guide early drug development. To obtain some understanding how enormously complex this problem is, one needs a large database of QSAR from as many diverse studies as possible. At the present time our group has some 18,800 QSAR, of which 10,700 are from biological systems. These range in complexity from DNA to enzymes, organelles, membranes, organs and whole animals including humans. About 8800 of the equations are from physical organic chemistry for comparative, mechanistic purposes. In this report, we are considering reports measuring the permeability of drugs and other organic compounds into Caco-2 cells, which are models for the intestinal lining. This is not a simple problem because in essence, one wants to know about the levels reached in many parts of whole animals. The subject has been recently reviewed by Waterbeemd and Gifford1 and Bergström et al.2 to suggest an algorithm for general use. In this report we are interested in the use of parameters that we have been developing for the past 40 years to characterize the properties that determine the interaction of organic chemicals with living systems or their parts. The most generally useful parameter is the octanol/water partition coefficient. More than half (5522) of the QSAR in our bio QSAR database contains such a term, often accompanied by parameters modeling electronic and/or size effects. Terms for polarizability3 (CMR or NVE) occur in 2576 equations and steric parameters (B1, B5, Es)4 occur in 1974 examples. We have used ClogP and CMR programs from the BioByte Corporation.5

Section snippets

Results

First we consider work done with Caco-2 cells to assess drug absorption via the intestines. At present, we have 29 QSAR using data reported on such cells. Of these, 26 contain Clog P terms. Two are based on Pi, the hydrophobic parameter derived for substituents. Of these 26, 9 are simple QSAR with single linear logP terms as the only parameter. Ten contain a single logP term plus an additional negative term: MgVol or CMR. Thus we can see a close relationship between penetration of Caco-2 cell

Discussion

There has been a recent tendency to refer to ADMET where T stands for toxicology. It is wishful thinking to believe that toxicology can be treated adequately via comparative QSAR. One must force the fact that at the present time, no one knows how to deal with metabolism. Cytochrome P-450 enzymes play the major role in metabolism. We have 96 QSAR for the activity of this varied class of enzymes. We plan to attempt to relate these QSAR to ADME but it is too complex for the current report. While

References (18)

  • C Hansch et al.

    Pharm. Sci.

    (1987)
  • A Gerritsen et al.

    Ecotoxicol. Environ. Saf.

    (1998)
  • H Ellgehausen et al.

    Ecotoxicol. Environ. Saf.

    (1980)
  • H van de Waterbeemd et al.

    Nat. Rev.

    (2003)
  • C.R.S Bergstrom et al.

    J. Med. Chem.

    (2003)
  • C Hansch et al.

    J. Chem. Inf. Comp. Sci.

    (2003)
  • C Hansch et al.

    Exploring QSAR. Fundamentals and Application in Chemistry and Biology

    (1995)
  • BioByte Corporation, 201 West 4th Street, Claremont, CA 91711...
  • A Kulkarni et al.

    Chem. Inf. Comput. Sci.

    (2002)
There are more references available in the full text version of this article.

Cited by (164)

  • Computational approaches for drug-metabolizing enzymes: Concepts and challenges

    2022, Biochemistry of Drug Metabolizing Enzymes: Trends and Challenges
  • Opportunities and challenges related to saturation of toxicokinetic processes: Implications for risk assessment

    2021, Regulatory Toxicology and Pharmacology
    Citation Excerpt :

    Many in vitro methods are available to estimate ADME parameters or investigate if ADME processes may be nonlinear over the range of possible administered doses used in animal studies (OECD et al., 2021). Parameters that are commonly measured with in vitro methods are fractional plasma protein binding (Rotroff et al., 2010; Wetmore et al., 2012), oral and dermal absorption rates (Hansch et al., 2004; Lehman et al., 2011; Potts and Guy, 1992; Zhao et al., 2003), tissue/plasma partitioning (Smith and Waters, 2018), and metabolism by whole cells (such as hepatocytes), key proteins in cellular membranes (such as microsomes), or individual enzyme (such as CYPs) (Franzosa et al., 2021; Obach et al., 1997; Rotroff et al., 2010; Wetmore et al., 2012). Some of these parameters can also be predicted in silico based on chemical structure and physiochemical properties (Dawson et al., 2021; Ingle et al., 2016; Obach et al., 2008; Poulin and Krishnan, 1996a, 1996b; Poulin and Theil, 2000; Sarigiannis et al., 2017; Schmitt, 2008).

View all citing articles on Scopus
View full text