Skin Sensitisation Mechanisms Provide Potency Predictions

« Mechanism-based Qsar modelling used
Researchers have developed quantitative structure-activity relationship (Qsar) models to help predict skin sensitisation potency, by first grouping chemicals based on sensitisation mechanisms.


The main way to determine a chemical’s skin sensitisation potential is to use a rodent test called the local lymph node assay (LLNA), which indicates potency. The test looks for stimulation of T cells in the lymph nodes following chemical exposure.

To date, most in silico prediction studies have been qualitative, simply indicating whether or not a chemical is a sensitiser. However, potency information is vital for risk assessment purposes and there is a “pressing need” for non-animal methods to give a quantitative indication of it, writes a team from the UK, US and Estonia.
The scientists used a computer programme and “additional expert knowledge” to allocate 204 skin-sensitising chemicals from large published datasets to mechanistic categories, before developing Qsar models.

The key to a succcessful model is to keep it simple by limiting the number of descriptors used, said co-author Steve Enoch from Liverpool John Moores University. Research suggests that there should be no more than five chemicals for each descriptor, he added.

The Qsar team stresses the importance of assigning chemicals to mechanistic categories, before attempting to predict potency using in silico approaches. “The chemistry of how the molecule is interacting with the protein [in the skin] is very important. You can’t mix that chemistry,” said Dr Enoch.
The scientists acknowledge that they have had to make compromises and that their model is not perfect. For example, the training set Qsars used to build the model were not fully independent of the test chemicals.

The Qsar model would not be used in isolation but as part of a battery of tests, said Dr Enoch. The idea is to build weight of evidence by using the model to make a prediction, and then combining the information with results obtained using other non-animal test methods, he said. »

The study is published in Chemical Research in Toxicology.

Mechanism-based QSAR modeling of skin sensitization
John C Dearden, Mark Hewitt, David W. Roberts, Steven Enoch, Philip Rowe, Katarzyna Przybylak, Daniel Vaughan-Williams, Megan Smith, Girinath G. Pillai, and Alan R. Katritzky
Chem. Res. Toxicol., Just Accepted Manuscript
DOI: 10.1021/acs.chemrestox.5b00197
Publication Date (Web): September 18, 2015

Many chemicals can induce skin sensitization, and there is a pressing need for non-animal methods to give a quantitative indication of potency. Using two large published data-sets of skin sensitizers, we have allocated each sensitizing chemical to one of ten mechanistic categories, and then developed good QSAR models for the seven categories with a sufficient number of chemicals to allow modeling. Both internal and external validation checks showed that each model had good predictivity.

Sources :
Abstract :

Cet article n’engage que son auteur/ This article is the sole responsibility of the author