Abortitristoside A and desrhamnosylverbanscoside: the potential COX-2 inhibitor from the leaves of Nyctanthes arbor-tristis as anti-inflammatory agents based on the in vitro assay, molecular docking and ADMET prediction

The plant Nyctanthes arbor-tristis belongs to the genus Nyctanthes (Oleaceae family) which is used in the treatment of many diseases like diabetes, cancer, inflammation, intestinal problem and tuberculosis in many countries of Asia. In this analysis, we have investigated the anti-inflammatory activity of various extracts of leaves of Nyctanthes arbor-tristis by the in vitro method as the cyclooxygenase inhibitory potential. The reported phytochemical constituent of the plant was in silico screened to find the cyclooxygenase inhibitory potential by molecular docking method. The molecular docking study of pre-isolated phytochemicals from the leaves of plants shows that compounds possess a great inhibition potential against COX-1 and COX-2 therefore, an in vitro comparative study is required to conclude the selective and potential COX-2 inhibitor for the development of new anti-inflammatory drug without causing any gastrointestinal ulcer. Here, we deliver a new strategy to design a selective COX-2 inhibitor from the plant origin natural compound by using molecular docking technique and in vitro screening of plant extract. In the other part of the paper, abortitristoside A (1) and abortitristoside B (2) were isolated from the leaves part of the plant. The isolated compound possesses good activity against COX-2 and COX-1 (IC50 value 7.91–8.00 μM against COX-2) which is also established by structure–activity relationship analysis.


Introduction
Inflammation is a responsive phase of physical, chemical injury, or damage by the attack of a pathogen like bacteria or viruses (Nagajyothi et al. 2015;Hasan et al. 2017). The components involved in the inflammation are degradative enzyme and superoxide (Cia X et al. 2017), chemotactic factors (Recinella et al. 2020;Cooper et al. 2019), vasoactive substances, and neuropeptide substances (Mallick et al. 2017). To treat the inflammation NSAIDs block, the enzyme cyclooxygenase (COX) also known as prostaglandin synthase hence inhibiting the synthesis of prostaglandin at the site of inflammation (Ameen et al. 2017).
The discovery of two different COX isoenzymes that is COX-1 (present throughout the gastrointestinal tract) and COX-2 (present at the site of inflammation) leads to the improvement of NSAIDs. Although only COX-2 involve in inflammation, therefore a COX-2 selective molecule is required so that it inhibits only COX-2 at the site of injury but not COX-1 throughout the gastrointestinal tract to treat the inflammation without causing any ulcer throughout the gastrointestinal tract (Leone et al. 2019). Different type of stimuli triggers the activation of cytosolic phospholipase A 2 (cPLA 2 ) set amino acid (AA) free from membrane phospholipids (Smith et al. 1996). COX-1 or, and COX-2 induced and catalyzed the biosynthesis of endoperoxidase PGH 2 which is highly unstable and converted into prostaglandins (PGD, PGE and PGF), prostacyclin's (PGI) and thromboxane's (TX) by specific catalyst (Fig. 1) (Rouzer et al. 2011;Sharma et al. 2010;Smith et al. 2011;Clària et al. 2003). Biosynthesize PGE 2 increases cell proliferation which causes various prolonged inflammatory diseases like the progression of cancer, Alzheimer's disease, arthritis, cardiovascular disorder, neurological disorders and autoimmune (Hyde et al. 2009;Willerson et al. 2004;Abou-Raya et al. 2006;Coussens et al. 2002;Lucas et al. 2006). Tricyclic COX-2 selective inhibitors like Refocoxib, Celecoxib and others act by blocking the COX-2 pocket. The selectivity is achieved due heavy size of the tricyclic COX-2 inhibitor. Another important difference is the presence of leucine in COX-2 at the place of phenylalanine in COX-1, which shows greater flexibility in the upper side of the active site in COX-2. NSAIDs bind to COX-1 by reversible hydrogen bonding and inhibition by simple steric hindrance which leads to a great variation between COX-1 and COX-2 and in this way selectivity can be achieved (Choi et al. 2018;Chuang et al. 2019). Studies of fluorescence quenching suggest that the outcome of COX-2 inhibitors is time-dependent which depends upon an active process with blocking of the lower enzyme site . Due to the presence of COX-2 mostly in the perinuclear envelope and COX-1 in the perinuclear membranes and endoplasmic reticulum (ER) the accessibility of the amino acid (AA) released by different  (Shi et al. 2018;Puleri et al. 2017). It was shown by Kulmacz and Wang (1995) that COX-2 could show dominant catalysis at low hydroperoxide levels, whereas COX-1 proceed at high hydroperoxide levels (Hu et al. 2017;Zago et al. 2017). It is therefore believable that limited AAs accessibility in specific cellular conditions COX-2 activity is favored (Garg et al. 2018;Hewett et al. 2016). As another option, it has been projected that both COX-1 and COX-2 exhibit different attachments to the specific terminal prostanoid synthases. This thought was primarily recommended for the link between COX isoenzymes and PGE synthase enzymes (PGES) and then between COX isoenzymes and other terminal prostanoid synthases (Jensen et al. 2018). So far, three enzymes that catalyze the formation of PGE2 from PGH2 somewhat exactly have been recognized; namely membrane-bound PGES (mPGES)-1 (Uribe et al. 2018), mPGES-2 and cytosolic PGES (cPGES) (Krishnamachary et al. 2017).
Nyctanthes arbor-tritis (Family: Oleaceae) is well known medicinal plant generally known as Night jasmine or Harsingar native to India, wildly distributed in the sub-Himalayan region and is also used as an ornamental plant in many regions of Asia (Chetty et al. 2008). The root part of the plant is used to treat anorexia and fever, the leaf part is used to treat vermifuge, fever, diaphoretic and diabetes, the bark part is used as an expectorant while the whole plant is used to control cancer (Mathuram et al. 1991;Saxena et al. 1987). Ethnobotanical importance shows that different extracts of the different parts of the plant possess anti-arthritis, antimalarial activity (Amarite et al. 2007), anti-oxidant and anti-inflammatory activity (Rathee et al.2007;Omkar et al. 2006). The sap of greeneries exhibits digestive, diuretic and antivenom properties and is also used as a tonic (Nadkarni 1982;Kirtikar & Basu 2000;Tempany 1950). The plant showed antiviral activity against the Semliki Forest virus and Encephalomyocarditis virus, while leaves were used to reduce the enlargement of the liver spleen (Gupta et al. 2008). The plant is very famous for its pharmacological properties from ancient science. Extensive research has been done on the plant Nyctanthes arbor-tristis because of its pharmacological properties. In the present study, we have tried to find a potential COX-2 inhibitor by in vitro analysis of the different extracts of various parts and by virtual screening of phytochemicals using molecular docking techniques.

COX-1 and COX-2 inhibition activity of various extracts, abortitristoside A (1) and abortitristoside B (2)
All four extracts, abortitristoside A (1) and abortitristoside B (2) isolated from leaves were tested to analyze their inhibitory activity toward COX-1 and COX-2. The tests were performed using the markedly available COX-(human) inhibitor screening kit (Catalog No. 701070, 701080, Cayman Chemical, PLM Pvt. Ltd., New Delhi, India). In this method, we analyze the quantity of PGF2α by using the enzyme immunoassay.
Results were expressed in the terms of percentage inhibition of COX by 30ug/ml solution of each extract and 30 μM solution of compounds 1 and 2 in triplicate observation with a two-digit standard deviation. Further IC 50 values were evaluated for the extracts having high inhibitory potential, abortitristoside A (1) and abortitristoside B (2) based on percentage inhibition (Table 1, Fig. 3). The inhibitory effects of the extracts were estimated at a concentration of 30 µg/ml, and percentage inhibition values were the means of three observations. Among all the extracts the methanol extract shows the highest percentage inhibition against COX-2 (88.97%) and COX-1 (80.87%) but reduced selectivity toward COX-2 as compared to celecoxib. Instead of methanol extract, chloroform extract also showed a remarkable percentage inhibition against COX-2 (85.23%) and COX-1 (69.19%) but lesser than celecoxib. The IC 50 values were evaluated for chloroform, methanol extracts, abortitristoside A (1) and abortitristoside B (2) against COX-2 which shows that abortitristoside A (1) possesses the lowest IC 50 value (7.91 μM) amongst all tested sample but higher than celecoxib (5.69 μM).

Structure-activity relationship (SAR)
The phytochemicals 1 and 14 contain a propenoate active group for interaction with COX-1/2 enzyme residues (Fig. 7). The compound 1-furan-2-yl-3-pyridine-2-yl-propenone (FPP-3) synthesized by Jahng and his colleagues possesses the highest COX-2 inhibition potency (IC 50 = 1.89 uM) which is a preclinical drug for the treatment of chronic inflammatory diseases like rheumatoid arthritis without causing any gastrointestinal ulcer (Shanmugam et al. 2006;Lee et al. 2006). Further, the FPP-3 confirmed that 5,6-membered aromatic/heterocyclic rings at 1, 3 positions were the basis for its increased potency (Jahng et al. 2004), and the 3-substituted larger phenyl group shows better selectivity toward COX-2 isoform to reduce side effects (Moreau et al. 2006). The presence of (-SO 2 Me) group at five or six-membered rings attached

Plant material
The plant was collected from the Chauraas Campus, HNB Garhwal University, District Tehri (UK, India), and the identification of the plant was done by the expertise of the Department of Botany, FRI, Dehradun, Uttarakhand and herbarium is deposited (210/Dis./2021/Syst.Bot./Rev. Gen./5-6, accession number 435).

Cyclooxygenase inhibition assay
The repressive action of various extracts against COX-2 was calculated by the fair marketable quality COX-(human) inhibitor screening kit (Catalog No. 701070, 701080, Cayman Chemical, PLM Pvt. Ltd., New Delhi, India) based on the producer's directions. The prepared reaction buffer contains 160 uL, 0.1 M Tris-HCl (pH 8.0), 5 mM EDTA and 2 mM phenol. COX-1/2 (10 uL) was added in 10uL, 30uM solution of each extract in the presence of 10 uL Heme solution. All these test solutions were relaxed for 10 min at 37 °C. Further, 10 uL arachidonic acid and 30 uL stannous chloride were added to the test solution (sample) and blank solution (control) to start the reaction to produce the PGF2α by the reduction of PGH2 and relaxed exactly for 5 min at 37 °C. The amount of produced PGF2α was calculated for sample and control by immunoassay method using the UV microplate reader at the wavelength 410 nm to calculate the percentage inhibition. Results were expressed in the terms of percentage inhibition of COX by 30ug/ml solution of each extract and 30 μM solution of compounds 1 and 2 in triplicate observation with a two-digit standard deviation. Further IC 50 values were evaluated for the extracts having high inhibitory potential, abortitristoside A (1) and abortitristoside B (2) based on percentage inhibition (Table 1, Fig. 3). The inhibitory effects of the extracts were estimated at a concentration of 30 µg/ml, and percentage inhibition values were the means of three observations.

Protein preparation
The groundwork of protein was done employing UCSF Chimera 1.11.2 in the dock prep module, where the deletion of inherent ligand attached to the crystal structure, hydrogen addition, water, solvents, replacing partial residues employing Dunbrack rotamer library, renovation of Bromo-UMP to UMP, renovation of selenomethionine to methionine, alteration of methylselenyl-dCMP to CPM. AMBER.ff14SB force field was used to dispense the charges on protein molecules

Molecular docking, docking parameters and post-docking modeling
The docking of all reference compounds and the selected phytochemicals of Nyctanthes arbor-tristis was performed using PyRx against COX-1 (PDB ID -3KK6) and COX-2 (PDB ID -3LN1). This software used generic algorithms for the docking. The defined parameters in PyRx for docking with Grid center (X = 26.0916, Y = 157.3507, Z = 10.1225 for COX-1 and X = 48.1391, Y = 33.8036, 25.3268 for COX-2), Dimension (X = 82.5649, Y = 89.4395, Z = 118.3433 Å for COX-1 and (X = 133.5613, 89.4395, 118.3433 Å for COX-2), population size (n = 200), generations (g = 10) and the number of solutions for each compound (s = 1). Based on the above set parameters, the compounds were screened. The post-dock modeling of docked poses was executed by Discovery studio visualizer v3.5. The top four compounds were chosen by considering the lowest energy of docking/binding. The binding energy of a ligand to the protein is given by: where E VDW means Vander Waal energy, E H-BOND means hydrogen bonding energy and E ELECTROSTATIC means for electro statistic energy (Forli et al. 2016). Further, K i values were calculated using the formula K i = exp(∆G/RT), where ∆G is the docking energy, R is 1.98719, and T is 298.15 (Storici et al. 2004).

Conclusion
As in the case of inflammation COX-2 enzyme is responsible while COX-1 involves maintaining normal physiology. Most of the known anti-inflammatory drugs inhibit both the isoenzyme and cause a peptic ulcer or gastrointestinal ulcer therefore a selective drug molecule is required. The in vitro analysis of all four fractions of the leave of Nyctanthes arbor-tristis shows that the methanol fraction has the greatest potential to inhibit the COX-2 as compared to other fractions. Further in vitro and in vivo analysis of methanol fraction in the different models also may provide a good therapeutic option to treat the inflammation. The molecular docking study of pre-isolated phytochemicals from the leave of the plant also shows that compounds 6 (− 9.5 kcal/mol), 21 (− 9.2 kcal/mol), 2 (− 8.8 kcal/mol) and 1 (− 8.3 kcal/ mol) has the highest binding affinity/docking score for 3KK6 (COX-1 enzyme) while ligands 14(− 10.3 kcal/mol), 1(− 10.2 kcal/mol), 2(− 10.2 kcal/mol) and 6(− 9.8 kcal/ mol) have highest binding affinity for 3LN1 (COX-2 E BINDING = E VDW + E H−BOND + E ELECTROSTATIC enzyme) but compound 1, 2 and 14 also shows docking selectivity for COX-2 express the novelty for finding good NSAIDs. Therefore, the methanol fraction was subjected to isolation and two compounds (abortristoside A and abortristoside B) were isolated and tested against COX-1 and COX-2. The in vitro screening against COX-1 and COX-2 reflexes that compound 1 and compound 2 have selectivity for COX-2 over COX-1 with IC 50 near to celecoxib which is also established by structure-activity relationship analysis. Further in vitro analysis of 1 and 2 in the different models also may provide a good therapeutic option to treat the inflammation therefore, in vivo analysis of abortristoside A (1) and abortristoside B (2) in the different models may provide a good therapeutic option to treat the inflammation without causing any gastrointestinal ulcer. Desrhamnosylverbanscoside a verbascoside reported phytochemical of the Nyctanthes arbor-tristis shows the highest docking score which isolation and screening against COX-1/2 may give the potential anti-inflammatory agent. Further, we provide a new strategy to design potential selective COX-2 inhibitors from the plant origin natural compound by using molecular docking techniques and in vitro screening of plant extract and isolated compounds.