Modulation of γ-aminobutyric acid (GABA) levels has been required in various disorders. GABA itself cannot be directly introduced into central nervous system (CNS) because of the blood brain barrier; inhibition of GABA aminotransferase (GABA-AT), which degrades GABA in CNS, has been the target for the modulation of GABA levels in CNS. Given that root extract of valerian (Valeriana officinalis) has been used for millennia as anti-anxiolytic and sedative, in silico approach was carried out to investigate valerian compounds exhibiting GABA-AT inhibiting activity. The 3D structure of human GABA-AT was created from pig crystal structure via homology modeling. Inhibition of GABA-AT by 18 valerian compounds was analyzed using molecular docking and molecular dynamics simulations and compared with known GABA-AT inhibitors such as vigabatrin and valproic acid. Isovaleric acid and didrovaltrate exhibited GABA-AT inhibiting activity in computational analysis, albeit less potent compared with vigabatrin. However, multiple compounds with low activity may have additive effects when the total extract of valeriana root was used in traditional usage. In addition, isovaleric acid shares similar backbone structure to GABA, suggesting that isovaleric acid might be a valuable starting structure for the development of more efficient GABA-AT inhibitors for disorders related with low level of GABA in the CNS.
Keywords: GABA aminotransferase (GABA-AT); Homology modeling; Molecular docking; Molecular dynamics; Valerian (Valeriana officinalis).