This proposal aims to develop in silico models for aldehyde oxidase (AOX) binding affinities (Aim 1), and in vivo human clearance (Aim 2). Three compounds have failed recently in clinical trials as a result of the lack of knowledge of AOX metabolism prior to entering humans. One of these compounds led to kidney toxicity in the test subjects. These failures are a result of the inability of any preclinical species to predict AOX metabolism. Furthermore, it has been reported that human cytosol can also fail to accurately predict in vivo human clearance. Preliminary results support our efforts to accurately predict human clearance using in silico methods. We were able to predict human clearance with a reasonable amount of accuracy (r2 = 0.8) for 8 compounds that have been administered to humans. We also provide preliminary results for the inhibition of AOX for 8 compounds. All but one of these compounds show complex inhibition kinetics. From the inhibition results, we are able to determine that two compounds, clozapine and chlorpromazine, have high enough affinities to show drug-drug interactions.
In Aim 1, we will develop quantitative structure inhibition relationships (QSIR) using data from 100 inhibitors of AOX. These QSIR models will have multiple uses, one of which will be predicting potential drug-drug interactions.
In Aim 2, we will refine our models for AOX clearance. Again, these models will have multiple uses, and provide a strong understanding of how these enzymes work in drug metabolism. We hypothesize that the models from Aim 1 and Aim 2 can be used in drug development to help eliminate drug failures.
Our knowledge of the drug metabolizing enzyme aldehyde oxidase is very weak and this has lead to a number of failures in drug development, with 3 being reported over the past year. Failures in clinical trials put people's lives in danger during the tral, and can stop or slow the development of new drugs. This grant will provide tools that can be used to help decrease failures in clinical trials.
Showing the most recent 10 out of 17 publications