Study On The Virtual Screening Model Of JAK3 Inhibitors

JAK3(Janus Kinase 3)is involved in the signal transduction process of γc cytokines,so it is closely related to the occurrence of inflammation and autoimmune diseases.Other subtype kinases of the JAK family extensively express in various organs and tissues,while the distribution of JAK3 is limited to the immune system.It makes JAK3 a very potential therapeutic target for the treatment of inflammation and autoimmune diseases.Therefore,JAK3 has been received extensive attention,and a large number of JAK3 inhibitors have been reported successively.Currently,JAK3 inhibitors are mainly developed through targeting ATP pockets,which are mainly divided into two categories: non-covalent inhibitors and covalent inhibitors.The main difficulty in the development of non-covalent inhibitors lies in the improvement of selectivity.The high homology of amino acids around the ATP binding pockets of JAK family makes the development of selective JAK3 inhibitors more challenging.At present,most of the reported inhibitors are not really selective inhibitors of JAK3 but pan-JAK inhibitors.On the other hand,covalent inhibitors are difficult to identify through traditional experimental methods because they involve the formation and rupture of chemical bonds.Meanwhile,the computer-aided drug design methods applied to the development of covalent inhibitors are limited,and the low computational efficiency and poor accuracy of these methods make them unsuited for the large-scale virtual screening.Thus,in this present study,we developed corresponding virtual screening models based on structural characteristics for different types of JAK3 inhibitors.The main results are as follows:(1)Based on a series of reported highly active JAK3 selective inhibitors,the interaction mechanism between JAK3 inhibitiors and proteins at the molecular level was determined by using molecular simulation methods combining three-dimensional quantitative structure-activity relationships,molecular docking,and molecular dynamics simulation.The key amino acids associated with the specific inhibition of JAK3 were all highlighted,these results would guide the subsequent virtual screening of non-covalent JAK3.(2)The aboved structural features were then integrated into the virtual screening process to find some JAK3 selective inhibitors.After the bio-activity evaluation of these virtual screening hits,some active compounds were discovered,proving that this structural feature-based virtual screening method possesses high accuracy and applicability.(3)A virtual screening strategy integrating pharmacophore and covalent docking was constructed for screening JAK3 covalent inhibitors.The combined method successfully screened the reported JAK3 clinical inhibitor(PF-06651600)from the Drug Bank database,indicating this virtual screening method not only has high efficiency,but also has strong screening accuracy,which provides certain guidances for the virtual screening of JAK3 covalent inhibitors.