In silico studies of cannabinoids targeting COX-2 and CB2 receptors for anti-inflammatory drug design

Phytocannabinoids have been studied for their medicinal purposes. This study examines their binding affinity to cyclooxygenase-2 (COX-2) and cannabinoids receptor type 2 (CB2), which are key therapeutic targets for inflammation. Semiempirical quantum mechanical (SQM) approaches hold promise for accurately describing noncovalent interactions in protein–ligand complexes, though calculating binding free energy for large complexes remains challenging. In this research, molecular docking simulations using AutoDock4 were initially employed to sampling ligand poses. False negative errors generated by AutoDock4 for non-steroidal anti-inflammatory drugs (NSAIDs) were identified using GFN2-xTB, a tight-binding SQM method. This approach, coupled with the ALPB solvation model, was used to compute the binding free energy of fully relaxed receptor-cannabinoid complexes in implicit aqueous solvation. This study also reports the performance of selected SQM methods in modeling noncovalent interaction of benchmark datasets. In addition to the solvation effect, thermostatistical contributions were included to obtain a more accurate binding free energy (∆Gbind,solv). Results showed that non-psychoactive acid derivatives such as cannabichromenic acid (CBCA), cannabinolic acid (CBNA), and cannabielsoic acid (CBEA) exhibited strong affinities for COX-2 and CB2. To enhance their anti-inflammatory potency, a sulfonamide group was incorporated to interact with Arg499 of COX-2. This modification of the CBCA analog yielded a novel anti-inflammatory compound with a computed binding free energy of -48.41 kcal/mol for COX-2, which is lower than that of celecoxib (-32.02 kcal/mol), a known NSAID. The predicted drug-like properties of the modified cannabinoid analogs provide valuable insights for developing novel oral anti-inflammatory leads.