A Successful Virtual Screening Application: Prediction Anticonvulsant Activity of Methyl Paraben and Propyl Paraben

A. TALEVI; E. A. CASTRO; L. E. BRUNO-BLANCH

Montevideo

Encuentro; I Latinoamerican Meeting on Medicinal Chemistry; 2007

Facultad de Química, Universidad de La República

Epilepsy is a neurologic disorder that affects about 1% of world population. Nearly 80% of epilepsy sufferers come from developing countries in the south hemisphere. Current therapeutic arsenal is not effective in the seiuzure control of up to 30-40% of the patients. Besides, even the new generation of antiepileptic drugs which was approved in the 1990s decade present important side-effects, so that new therapeutic agents are urgently needed (1,2). The objective of this study was to discover new anticonvulsant agents in a rational way, through virtual screening of large compound databases. For this purpose we have applied Linear Discriminant Analysis (LDA) to derive a Discriminant Function based on topological (2D) descriptors. Several DFs were obtained from a 877 constitutional and topological descriptors pool from Dragon software (3). The DF with lowest Wilk´s U statistical and best performance in the validation steps was applied in virtual screening of a database of chemical compounds obtained from Merck Index 13th. Two of the most used preservatives in pharmaceuticals, cosmeticals and food, methyl paraben and propyl paraben, were selected through this methodology for biological testing. They proved active in the Maximal Electroshock (MES) test, at doses of 30, 100 and 300 mg/kg. This confirms the effectiveness of the applied methodology, serves as definitive validation of the selected model. Most important, two potent, new anticonvulsants of well-known safety assessments and pharmacokinetic properties have been identified, which arises perspectives of new antiepileptic medications to be developed. Alan Talevi thanks CONICET for his type I postgrado fellowship. Eduardo A. Castro is a member of CONICET. Luis E. Bruno-Blanch thanks UNLP and ANPCyT (PICT BIB 1728/06-11985) for the funds to develop this study.