Inhibitory activities of flavonoids derivatives on cyclooxygenase-2 (COX-2): in silico modelling and cell-based study / Siti Norhidayu Mohd Amin

COX-2 inhibitors are introduced as a better alternative to replace the traditional
NSAIDs. This is because it is associated with lower incidences of gastrotoxicity that
commonly occurs among the users of the traditional NSAIDs. Therefore, this study aims
to identify the lead compounds based on flavonoid scaffolds which can be further
developed into drug(s) as treatment for inflammation. To achieve this, several
pharmacophore hypotheses on COX-2 were developed. These pharmacophore
hypotheses were then validated and the best pharmacophore was chosen and named as

“phore1”. The “phore1” was used in virtual screening of datasets (ASINEX and in-
house datasets). The hit compounds with score of more than 45.00 were chosen for

molecular docking. Then, the compounds with binding energy better than -8.00

kcal/mol at the COX-2 binding site were chosen as the final hit compounds. For the in-
house virtual screening, additional methods were applied. The in-house dataset was

prefiltered by Lipinski’s Rule of Five and verified using COX peroxidase assay.
Simultaneously, the filtered compounds were screened using virtual screening method

mentioned before and both of the results were compared and validated. From the in-
house dataset, five compounds were successfully predicted through in silico screening

but only compound F3 was exhibited potency towards COX-2 with IC50 of 24.30 μM.
For ASINEX dataset, fifteen (15) hit compounds successfully screened using in silico
study and they are known as BAS00127074, BAS00384673, BAS00428711,
BAS00547888, BAS00643043, BAS00643060, BAS00643061, BAS00654798,
BAS00791751, BAS01121975, BAS01121978, BAS01316535, BAS01316573,
BAS02332476, and BAS02557914. It is concluded from this study that the in silico
COX-2 model has been successfully developed and is useful for the screening and
identification of new potential COX-2 inhibitors. This model would allow researchers
to screen more compounds without doing biological assay.