Model antioxidant compounds

Chemometric modelling of flavone antioxidants has led to an improved quantitative structure-activity relationship (QSAR) technique for these molecules that reveals important information about how their ring systems contribute to activity.

Free radicals constitute an important part of various metabolic processes as well as being involved in the immune response to invading bacteria. Factors such as smoking, environmental pollution and radiation exposure increase free radical level, which leads to oxidative stress. Oxidized cholesterol, oxidation of polyunsaturated fat and excessive mental stress also add to the formation of toxic radicals. There are many aspects of ill-health that can be associated with excessive concentrations of free radicals in body tissues. For instance, free radicals are implicated in oxidative stress related to fatal diseases including Parkinson's, Alzheimer's disease, rheumatoid arthritis and cardiovascular disorders.

Our bodies have various mechanisms to reduce free radical damage and to repair damage that occurs. Antioxidants are key to these defence mechanisms and work by breaking the free radical chain reaction or through metal ion chelation, which inhibits metal-catalysed free radical reactions. Among the most interesting antioxidants are the flavones, a type of flavonoid with the 2-phenylchromen-4-one backbone found naturally in cereals and herbs.

The presence of hydroxyl groups in the flavonoids makes them particularly interesting as free radical scavengers as they react with neighbouring free radicals through the HAT (hydrogen atom transfer) mechanism.

Understanding the mode of action of flavones and what aspects of their chemical structure contribute to their beneficial health effects is important in validating theories about antioxidants in general and potentially in the development of particular supplements for specific health issues. As such, quantitative structure-activity relationship (QSARs) are an invaluable tool based on sound statistical principles for solving such problems. Several attempts have been made to build general QSAR models for the exploration of antioxidants. It is particularly powerful because it can rapidly screen libraries of designed compounds and help in rational drug design.

Indrani Mitra, and Kunal Roy of Jadavpur University and Achintya Saha of the University of Calcutta, India, explain that with this background they have developed QSAR models with various kinds of isoflavones, isoflavanes and biphenyl ketones for their ability to scavenge 1,1-diphenyl-2-picrylhydrazyl (DPPH) free radicals. They hope to unearth the requirements of this specific class of compounds for their antioxidant activities at the molecular and atomic levels. Their models should make it possible to screen flavone derivatives for improved antioxidant activity.

The researchers explain that their initial QSAR model was built on the Fujita-Ban method but it did not cope with the training set molecules so they turned to alternative chemometric tools including genetic function approximation and genetic partial least squares. These latter techniques used additional descriptors including topological, structural, spatial and quantum chemical ones. On the basis of the QSAR developed, the team concludes that, "The statistically significant models thus developed suggest that hydroxy and methoxy substituents at certain specified positions of the A and B rings of the flavone moiety chiefly influence the antioxidant activity of these molecules."