Virtual screening based on pharmacophore for potential il-21/il-21r inhibitors in the treatment of rheumatoid arthritis

Through a pharmacophore model, the study screened compounds based on the binding score between interleukin 21 and its receptor interleukin 21R (IL-21/IL-21R). The results yielded two models: pharmacophore model P-1 (Tyr129, Met70, Asp72) and pharmacophore model P-2 (Tyr36, Met70, Asp72). Screening was conducted across two large databases, the ZINC database (18,471 compounds) and the ChemBL database (585 compounds). The screening results for model P-1 identified 24 compounds from ZINC and 50 compounds from ChemBL, while model P-2 identified 80 compounds from ZINC and 252 compounds from ChemBL. These screened compounds were further filtered using Lipinski's Rule of Five, resulting in 5 compounds from ZINC and 2 compounds from ChemBL for model Ph-1, and 7 compounds from ZINC and 9 compounds from ChemBL for model Ph-2. Docking results of the 23 selected compounds with interleukin 21, showing interactions with 8 or more amino acids, revealed that 19 out of 23 compounds had good binding affinities (∆G < -2 kJ/mol), with CHEMBL176689 (-16.3527 kJ/mol) and CHEMBL1198616 (-11.8833 kJ/mol) exhibiting the best docking scores. However, CHEMBL176689's binding site deviated from the Met70 position. This study helps to screen compounds with the potential to be developed into small-molecule drugs that inhibit IL-21/IL-21R in rheumatoid arthritis.