Activity-Guided Characterization of COX-2 Inhibitory Compounds in L. Extracts.

Inflammation is the body's response to infection or tissue injury in order to restore and maintain homeostasis. Prostaglandin E2 (PGE-2) derived from arachidonic acid (AA), via up-regulation of cyclooxygenase-2 (COX-2), is a key mediator of inflammation and can also be induced by several other factors including stress, chromosomal aberration, or environmental factors. Targeting prostaglandin production by inhibiting COX-2 is hence relevant for the successful resolution of inflammation. L. is a traditional medicinal plant whose extracts have demonstrated COX-2 inhibitory properties. However, the compounds responsible for the activity remained unknown. For the preparation of extracts with effective anti-inflammatory properties, characterization of these substances is vital. In this work, we aimed to address this issue by characterizing the substances responsible for the COX-2 inhibitory activity in the extracts and generating prediction models to quantify the COX-2 inhibitory activity without biological testing. For this purpose, an extract was separated into fractions by means of centrifugal partition chromatography (CPC). The inhibitory potential of the fractions and extracts against the COX-2 enzyme was determined using a fluorometric COX-2 inhibition assay. The characterizations of compounds in the fractions with the highest COX-2 inhibitory activity were conducted by high resolution mass spectrometry (HPLC-MS/MS). It was found that these fractions contain alpha-linolenic acid, linoleic acid and oleic acid, identified and reported for the first time in leaf extracts. After analyzing their contents in different extracts, it could be demonstrated that these fatty acids are responsible for up to 41% of the COX-2 inhibition observed with extract. Additional quantification of secondary metabolites in the extract fractions revealed that substances from the group of steroidal saponins and triterpenoid saponins also contribute to the COX-2 inhibitory activity. Based on the content of compounds contributing to COX-2 inhibition, two mathematical models were successfully developed, both of which had a root mean square error (RMSE) = 1.6% COX-2 inhibitory activity, demonstrating a high correspondence between predicted versus observed values. The results of the predictive models further suggested that the compounds contribute to COX-2 inhibition in the order linoleic acid > alpha linolenic acid > steroidal saponins > triterpenoid saponins. The characterization of substances contributing to COX-2 inhibition in this study enables a more targeted development of extraction processes to obtain extracts with superior anti-inflammatory properties.