Computer-aided design and synthesis of 5,7-dihydroxyflavone derivatives as glycogen phosphorylase inhibitors

Abstract

Diabetes is a chronic disease that affects approximately 415 million people worldwide and this number is expected to rise to 642 million by the year 2040. 90% of cases of diabetes are type 2. Although some treatments are already available for type 2 diabetes, there are considerable side-effects associated with these drugs that include a risk of hypoglycaemia. Hence, there is an immediate need for new and more effective drugs. Naturally occurring flavonoids are known inhibitors of glycogen phosphorylase, which is a validated target for controlling hyperglycemia in type 2 diabetes. By exploiting computational methods such as molecular docking with Glide and post-docking binding free energy calculations using advanced QM/MM-PBSA calculations, we have screened a large library of 1239 5,7-dihydroxyflavone analogues which have the potential to bind at the inhibitor site of glycogen phosphorylase. We compared these results to our benchmark ligand chrysin, which can inhibit glycogen phosphorylase with a Ki of 19.01 µm and identified thirteen ligands which are predicted to have better binding affinities exploiting a rigorous consensus scoring approach to reduce the chance of false positives. Favourable substituents on the predicted flavonoids included B ring hydroxyl and halogenated substituents in the ortho position and small hydrophobic groups on the meta/para positions directed towards the hydrophobic cavity of the binding site. Towards synthesis of these analogues, after many adjustments to the synthetic procedure it was identified that using methyl protection of the hydroxyls at the 5 and 7 position in parallel with the Baker-Venkataraman re-arrangement managed to achieve high purity initial compounds with a relatively good yield. The procedure is now ready to be employed to synthesize all other predicted 5,7-dihydroxyflavone derivatives. Kinetics experiments will validate their potency against glycogen phosphorylase and structure activity relationship analysis can help further guide lead optimization alongside pharmacokinetics and in vivo studies, until we can achieve a more potent and drug-like inhibitor of glycogen phosphorylase for the treatment of type 2 diabetes.