Physiochemical, Genotoxicity, and Blood-Brain Barrier Passage Properties of Chemicals
The traditional model used by medicinal chemists, pharmacologists, and toxicologists to understand how xenobiotics (compounds not naturally produced by the body) enter and leave the body considers the absorption, metabolism, distribution, excretion, and toxicity (ADMET) of the compound. Although remarkable progress has been made in recent years regarding the development of computational tools for predicting toxicological endpoints,6 accurately anticipating toxicological hazard remains a significant challenge.7 Currently, the most effective means to proactively reduce toxicological risk of environmental or industrial chemicals is to design molecules that are not readily absorbed by a biological system.8,° An additional factor is to redesign structural motifs to avoid metabolism of compounds into more toxic intermediates.1041 A review by Mangiatordi et al. describes in silico procedures for predicting metabolic processes.12 In drug research, medicinal chemists have developed models to predict which physicochemical properties make a molecule more “drug-like”: that is, likely to be absorbed into the body, to be distributed within the body to certain targets, and to produce an effect.13 Those same principles can provide guidance to design molecules that are less likely to interact with biological targets. these guiding principles are primarily concerned with the likelihood that a compound will be absorbed if it is administered orally, but they overlap significantly with the rules of dermal and respiratory absorption as well. generally speaking, the rules that govern the likelihood of a compound crossing the cellular membrane are consistent across routes of exposure. Lipinski’s rules are a commonly used set of five properties that are used to predict whether a compound is likely to be readily absorbed through oral ingestion. The rules are as follows: chemicals with (1) more than five hydrogen bond donors, (2) more than 10 hydrogen bond acceptors, (3) a molecular weight >500 Da, or (4) a logP value (sometimes called log Kow) >5 are unlikely to be well absorbed, unless the compound is (5) a substrate for a biological transporter, in which case it can be an exception to the previous rules.14 A comparable analysis by Veber and colleagues similarly found that compounds with >10 rotatable bonds, >12 total hydrogen bond acceptors and donors, or a polar surface area >140 A were unlikely to be orally bioavailable in rats.15 while numerous properties are instrumental in the absorption of exogenous compounds, lipophilicity, charge, similarity to endogenous substances, blood-to-gas partition, molecular weight, and polar surface area appear to be of the greatest value when designing molecules.8,° That is, molecules that are large, hydrophilic, charged at neutral pH, and that possess a large polar surface area are not readily absorbed in the gastrointestinal tract. If the molecule of interest is a substrate for a one of the body’s many biological transport proteins, it may be actively transported into cells, and possibly distributed to the rest of the body. More complex and sophisticated models have been developed for predicting intestinal absorption based on in vitro data collected in the Madin-Darby Canine Kidney Epithelial cells (MDCK) or Caco-2 cell lines, which are used as models for intestinal absorption.16’17 MDCK and Caco-2 permeability predictions are not included in many computational platforms, but are notable when they appear, as they provide some direct suggestions about bioavailability. Given the importance of genetic toxicity in the evaluation of commercial compounds, computational tools that predict whether a chemical has mutagenic potential are common.11’18 The Ames test is an in vitro test which principally uses the bacterium Salmonella typhimurium to determine the potential mutagenicity of chemicals.19-21 A positive Ames result should prompt reconsideration of structural features (such as reactive nucleophiles) likely to elicit mutagenic activity, as well as additional chemical biotransformations that can change mutagenic potential. Currently, assessing the potential mutagenicity of metabolites is limited to specific computational platforms and in most cases, potential metabolite structures have to be assessed separately.
Knowing if a chemical can cross the blood-brain barrier (BBB) is a significant aspect of drug design. The highly selective permeability of the BBB can make it challenging to deliver some drugs to the brain as intended, but it can also allow other compounds through that are not meant to affect the central nervous system (CNS), resulting in undesirable effects.22-24 Predictive models for BBB permeability use lipophilicity, polar surface area, and whether the compound is a substrate for specific transporters to identify compounds that are likely to cross the BBB and potentially interact with neurological pathways. However, computational predictions of BBB permeability must be interpreted cautiously, because the BBB is a complex membrane that is difficult to model,25,26 and because poor penetration alone may not be sufficient to prevent a chemical with chronic systemic exposure from achieving significant levels in the brain. The key knowledge required for a chemist wishing to use such predictive toxicology tools is to understand which molecular interactions and characteristics that are most benign or worrisome for a physiological system— concepts which are central to the design of medicinal compounds.