Biomarker Quantification, Spectroscopic, and Molecular Docking Studies of the Active Compounds Isolated from the Edible Plant Sisymbrium irio L.

Phytochemical investigation of the ethanolic extract of the aerial parts of Sisymbrium irio L. led to the isolation of four unsaturated fatty acids (1–4), including a new one (4), and four indole alkaloids (5–8). The structures of the isolated compounds were characterized with the help of spectroscopic techniques such as 1D, 2D NMR, and mass spectroscopy, and by correlation with the known compounds. In terms of their notable structural diversity, a molecular docking approach with the AutoDock 4.2 program was used to analyze the interactions of the identified fatty acids with PPAR-γ and the indole alkaloids with 5-HT1A and 5-HT2A, subtypes of serotonin receptors, respectively. Compared to the antidiabetic drug rivoglitazone, compound 3 acted as a potential PPAR-γ agonist with a binding energy of −7.4 kcal mol−1. Moreover, compound 8 displayed the strongest affinity, with binding energies of −6.9 kcal/mol to 5HT1A and −8.1 kcal/mol to 5HT2A, using serotonin and the antipsychotic drug risperidone as positive controls, respectively. The results of docked conformations represent an interesting target for developing novel antidiabetic and antipsychotic drugs and warrant further evaluation of these ligands in vitro and in vivo. On the other hand, an HPTLC method was developed to quantify α-linolenic acid in the hexane fraction of the ethanol extract of S. irio. The regression equation/correlation coefficient (r2) for linolenic acid was Y = 6.49X + 2310.8/0.9971 in the linearity range of 100–1200 ng/band. The content of α-linolenic acid in S. irio aerial parts was found to be 28.67 μg/mg of dried extract.

Compound 4 was isolated as a white amorphous solid. The HRESIMS showed quasimolecular ion peaks at m/z 329.2325 [M+H] + and 351.2140 [M+Na] + , consistent with a molecular weight of 328 amu, and a molecular formula of C 18 H 32 O 5 . IR absorption at 3315, 1697, and 1455 cm −1 indicated the presence of hydroxyl group, carboxylic group, and olefinic double bonds, respectively.
Compound 4 was isolated as a white amorphous solid. The HRESIMS showed quasimolecular ion peaks at m/z 329.2325 [M+H] + and 351.2140 [M+Na] + , consistent with a molecular weight of 328 amu, and a molecular formula of C18H32O5. IR absorption at 3315, 1697, and 1455 cm −1 indicated the presence of hydroxyl group, carboxylic group, and olefinic double bonds, respectively.

Molecular Docking of Fatty Acids (1-4) with PPAR-γ
The isolated fatty acids (1-4) from S. irio were subjected to molecular docking experiments with peroxisome proliferator-activated receptor gamma (PPAR-γ or PPARG), also known as the glitazone reverse insulin resistance receptor, based on the reported antidiabetic activity for similar compounds [24][25][26][27]. The relative binding of fatty acids to the PPAR-γ substrate-binding site is described in Table 2 and Figures 3 and 4. The details of the protein-ligand interaction are presented in Supplementary Table S3. It is obvious that all isolated fatty acids have a docking energy in the range of −6.0 to −7.4 kcal/mol (Table 2), and the lowest binding energy was revealed by compound 3. To strengthen our finding, the binding potentials of (1-4) were compared with rivoglitazone as a positive control. The details of interactions between PPAR-γ and fatty acids, along with rivoglitazone, are discussed in the subsequent sections.    The analysis revealed that the PPAR-γ-rivoglitazone complex was stabilized by hydrophobic bonding ( Figure 3A,B), including two Pi-Sigma hydrophobic bonds with ILE 341 and three amide-Pi-stacked interactions with GLY 284 (two interactions) and CYS 285 (one interaction). Further, there were five Pi-Alkyl hydrophobic interactions with CYS 285 (two interactions), ARG 288 (two interactions), and LEU 330 (one interaction); and one Pi-Pi T-shaped hydrophobic interaction with HIS 449 ( Figure 3C). In addition, the protein-ligand complex was stabilized by three Pi-S bonds with MET 348 (two interactions) and MET 364 (one interaction); two Pi-donor hydrogen bonds with CYS 285 and SER 289 ; and one conventional hydrogen bond with TYR 473 . The van der Waals interactions were also present with amino acid residues like ARG 280 , ILE 281 , PHE 282 , GLN 286 , HIS 323 , ILE 326 , TYR 327 , VAL 339 , LEU 353 , LEU 453 , and LEU 469 . The PPAR-γ-rivoglitazone complex was stabilized by an estimated free energy of −8.2 kcal mol −1 , which corresponds to a dissociation constant of 1.03 × 10 6 M −1 (Table 2).
The analysis of molecular docking also suggests that compound 2 occupied the active site of PPAR-γ ( Figure 4B). The compound 2-PPAR-γ complex was stabilized by one  It is obvious that all isolated fatty acids have a docking energy in the range of −6.0 to −7.4 kcal/mol (Table 2), and the lowest binding energy was revealed by compound 3. To strengthen our finding, the binding potentials of (1-4) were compared with rivoglitazone as a positive control. The details of interactions between PPAR-γ and fatty acids, along with rivoglitazone, are discussed in the subsequent sections.
Finally, the analysis of the compound 4-PPAR-γ interaction revealed that the complex was stabilized by hydrogen bonds as well as hydrophobic (alkyl) interactions. Compound  (5)(6)(7)(8) were screened for antidepressant activity using 5-HT 1A and 5-HT 2A serotonin receptors as potential targets. The relative binding of indole compounds is described in Tables 3 and 4 and Figures 5-8, and the detailed protein-ligand interaction is presented in Supplementary Tables S4 and S5.         The results exhibited that all indole compounds (5)(6)(7)(8) were able to bind to the substrate binding site of 5-HT 1A ( Figure 5A,B) and 5-HT 2A ( Figure 7A,B) receptors, with more affinity towards the 5-HT 2A receptor, and their binding energies varied in the range between −6.4 and −6.9 kcal mol −1 and between −7.3 and −8.1 kcal mol −1 , respectively. The binding potentials of indole compounds towards 5-HT 1A and 5-HT 2A receptors were compared with serotonin and risperidone, respectively, as positive controls (Tables 3 and 4). The details of interactions between indole compounds and 5-HT 1A and 5-HT 2A , along with positive controls, are discussed in the subsequent sections.

5-HT 1A Binding Interaction
The molecular docking of serotonin (positive control) showed that it occupied the active site of 5-HT 1A and interacted primarily through hydrogen bonds and hydrophobic interactions ( Figure 5A,B). It formed four hydrogen bonds with ASP 116 , VAL 117 , THR 121 , and TYR 390 ; one Pi-Pi T-shaped hydrophobic interaction with PHE 361 ; and four Pi-Alkyl hydrophobic interactions with VAL 117 (two interactions), ILE 189 , and ALN 203 ( Figure 5C). In addition, the protein-ligand complex was stabilized by van der Waals interactions with CYS 120 , ILE 124 , ILE 167 , SER 199 , PHE 362 , ALA 365 , and ASN 386 . The 5-HT 1A -serotonin complex was stabilized by −6.1 kcal mol −1 free energy, which corresponds to a dissociation constant of 6.1 × 10 4 M −1 ( Table 3).
The analysis of molecular docking also suggests that the 5HT 1A -7 complex was stabilized by hydrophobic interactions, including two Pi-Sigma hydrophobic interactions with VAL 117 , one Pi-Alkyl interaction with ILE 189 , one Pi-Sulfur bond with CYS 120 , and one Pi-Pi T-shaped bond with PHE 361 ( Figure 6C). In addition, it formed van der Waals interactions with ASP 116 , THR 121 , ILE 124 , TRY 195 , SER 199 , ALA 203 , and PHE 362 . Gibb's free energy of the complex formation was −6.5 kcal mol −1 , which corresponds to a dissociation constant of 5.85 × 10 4 M −1 (Table 3). Finally, Compound 8 was able to bind at the central binding cavity of 5-HT 1A , and the resulting complex was mainly stabilized by hydrophobic interactions ( Figure 6D). Compound 8 formed one hydrogen bond with THR 196 and six hydrophobic (Pi-Alkyl) interactions with ILE 189 , LYS 191 , ALA 365 , and PRO 369 ( Figure 6D). In addition, it formed van der Waals interactions with VAL 117 , SER 190 , TYR 195 , SER 199 , THR 200 , PHE 361 , and PHE 362 . The binding energy and dissociation constant for compound 8 and 5-HT 1A interactions were −6.9 kcal mol −1 and 1.15 × 10 5 M −1 , respectively (Table 3).
The analysis of molecular docking suggests that compound 5 occupied the active site of 5-HT 2A and the formed complex was stabilized by two hydrogen bonds with SER 159 and THR 160 , and six hydrophobic interactions with TRP 336 , PHE 340 , VAL 156 , and ILE 163 ( Figure 8A). In addition, compound 5 formed van der Waals interactions with ASP 155 , SER 242 , PHE 243 , PHE 332 , PHE 339 , and TYR 370 . Gibb's free energy of the complex formation was −7.3 kcal mol −1 , which corresponds to a dissociation constant of 2.60 × 10 5 M −1 ( Table 4).
The analysis of molecular docking suggests that compound 7 occupied the active site of 5-HT 2A , and the complex was stabilized via hydrogen bond formation and hydrophobic interactions. Compound 7 formed two hydrogen bonds with ASP 155 , four Pi-Pi T-shaped hydrophobic interactions with TRP 336 and PHE 340 , one Pi-Alkyl interaction with VAL 156 , and two Amide-Pi stacked interactions with SER 159 and THR 160 ( Figure 8C). In addition, compound 7 formed van der Waals interactions with LEU 123 , THR 160 , ILE 163 , GLY 238 , SER 242 , PHE 243 , PHE 332 , PHE 339 , and TYR 370 . Gibb's free energy of the complex formation was −7.4 kcal mol −1 , which corresponds to a dissociation constant of 2.68 × 10 5 M −1 (Table 4).
Finally, the analysis of the 5-HT 2A -8 complex revealed that it was mainly stabilized by the formation of three hydrogen bonds with ASP 155 , THR 160 , and TYR 370 ; three Pi-Pi Tshaped hydrophobic interactions with TRP 336 and PHE 340 ; and two Pi-Alkyl hydrophobic interactions with ILE 163 and VAL 156 ( Figure 8D). A salt bridge was also formed between ASP 155 and compound 8. In addition, it formed van der Waals interactions with LEU 123 , SER 159 , SER 242 , PHE 243 , PHE 332 , PHE 339 , and VAL 366 . The binding energy and dissociation constant for 5-HT 2A -8 interactions were −8.1 kcal mol −1 and 2.68 × 10 5 M −1 , respectively (Table 4).

Prediction of Physicochemical, Pharmacokinetic, Drug-Likeness, and Toxicity
The physicochemical, pharmacokinetic (ADME parameters), drug-likeness, and toxicity properties of the investigated compounds were predicted using the SwissADME online tool [28]. These properties include lipophilicity, water-solubility, topological polar surface area (TPSA), GI absorption, blood-brain barrier (BBB) permeability, P-glycoprotein pump (P-gp) efflux, Lipinski's rules, bioavailability, PAINS, and CYP 450 inhibition. In general, the results showed compliance with Lipinski's rules ( Table 5). The radar plot ( Figure S9) is a representation of the mean values of six descriptors that are significant for oral bioavailability and used for a rapid appraisal of drug-likeness, including molecular size (SIZE), polarity (POLAR), lipophilicity (LIPO), flexibility (FLEX), saturation (SATU), and solubility (INSOLU). The red lines of the investigated molecules have to fall entirely in the pink area of the radar plot to be considered drug-like [28]. Most of the investigated compounds fall inside the pink area of their radar representations, indicating potential drug-like properties.
Lipophilicity is a crucial property for BBB permeability, a major obstacle in the delivery of antidepressant drugs to the brain [29]. Compounds 5, 6, and 7 showed potential BBB permeability; however, the permeability of the polar indole derivative (8) and the polyhydroxylated fatty acids (3 and 4) was not probable. The P-gp reduces the drug permeability through the BBB by allowing the efflux of many drugs back into the blood (multidrug resistance). No active efflux was observed by the P-gp for all investigated indole derivatives ( Figure S10).

Analysis of Free Energy Calculations
The free energy (MM-GBSA) of the interaction between a protein and a ligand sheds light on the effect of solvent on the formation of a protein-ligand complex. Here, we calculated the free energy of the interaction of PPAR-γ with compounds 1-4, and serotonin receptors (5-HT 1A and 5-HT 2A ) with compounds 5-8 (Table 6). It is clear that the free energies of compounds 1 (−54.55 kcal/mol) and 3 (−57.23 kcal/mol) were the lowest, suggesting that they formed a stable complex with PPAR-γ. Likewise, the free energies of compound 8 (−57.98 kcal/mol) for 5-HT 1A and compound 8 (−57.90 kcal/mol) for 5-HT 2A were the lowest, indicating that these compounds interacted favorably with 5-HT 1A and 5-HT 2A , respectively. It is also imperative to note that van der Waals interactions (∆G vdW ), Coulombic interactions (∆G Coulomb ), and non-polar solvation energy (∆G SA ) or lipophilic interactions (∆G Sol_Lipo ) were the primary driving forces for the formation of a stable protein-ligand complex. On the other hand, polar solvation energy (∆G Solv or ∆G SolGB ) and covalent (∆G Covalent ) interactions were the main forces to destabilize a protein-ligand complex. It is worth noting that we selected only compounds 1 and 3 for PPAR-γ, and compound 8 for 5-HT 1A as well as 5-HT 2A to gain an in-depth analysis of interaction by molecular dynamics (MD) simulation.

Analysis of Moelcular Dynamics Simulation (MDS) 2.5.1. Root Mean Square Deviation (RMSD)
RSMD is a measure of deviation in the structure of a protein in the presence or absence of a ligand from its initial structure during the course of simulation, which in turn reflects on the system's stability [30]. In this study, RMSDs in Cα-atoms of PPAR-γ, 5-HT 1A , and 5-HT 2A

Root Mean Square Fluctuation (RMSF)
During MD simulation, any fluctuations in the side chain of a protein due to the binding of a ligand are measured by monitoring the RMSF. Here, the RMSF values of PPARγ 5-HT 1A and 5-HT 2A alone or in the presence of their respective ligands were determined as a function of simulation time ( Figure 10). The RMSF plot of PPAR-γ-1 and PPAR-γ-3 complexes overlapped with the RMSF plot of PPAR-γ alone, suggesting the absence of any significant changes in PPAR-γ conformation due to its interaction with compounds 1 and 3 ( Figure 10A). Similarly, the RMSF plots of 5-HT 1A -8 and 5-HT 2A -8 complexes overlapped with the RMSF plots of 5-HT 1A and 5-HT 2A alone, respectively, showing that there were no significant changes in 5-HT 1A and 5-HT 2A due to the binding of ligands and hence the formation of stable protein-ligand complexes ( Figure 10B,C). Any minor fluctuations in RMSF plots were due to the binding of ligands to proteins.

Root Mean Square Fluctuation (RMSF)
During MD simulation, any fluctuations in the side chain of a protein due to the binding of a ligand are measured by monitoring the RMSF. Here, the RMSF values of PPAR-γ 5-HT1A and 5-HT2A alone or in the presence of their respective ligands were determined as a function of simulation time ( Figure 10). The RMSF plot of PPAR-γ-1 and PPAR-γ-3 complexes overlapped with the RMSF plot of PPAR-γ alone, suggesting the absence of any significant changes in PPAR-γ conformation due to its interaction with compounds 1 and 3 ( Figure 10A). Similarly, the RMSF plots of 5-HT1A-8 and 5-HT2A-8 complexes overlapped with the RMSF plots of 5-HT1A and 5-HT2A alone, respectively, showing that there were no significant changes in 5-HT1A and 5-HT2A due to the binding of ligands and hence the formation of stable protein-ligand complexes ( Figure 10B,C). Any minor fluctuations in RMSF plots were due to the binding of ligands to proteins.

Principal Component Analysis (PCA) or Essential Dynamics (ED) Analysis
The global motion of a protein in the presence or absence of a ligand is generally monitored by PCA or ED [31]. In this study, the conformational sampling of Cα-atoms along PC1 and PC2 of PPAR-γ, 5-HT1A, and 5-HT2A was performed in the absence or presence of their respective compounds ( Figure 13). A conformational state of a protein is represented by the red and black dots. On the other hand, each red and black cluster shows the presence of distinct energetically favorable conformational spaces.

Principal Component Analysis (PCA) or Essential Dynamics (ED) Analysis
The global motion of a protein in the presence or absence of a ligand is generally monitored by PCA or ED [31]. In this study, the conformational sampling of Cα-atoms along PC1 and PC2 of PPAR-γ, 5-HT 1A , and 5-HT 2A was performed in the absence or presence of their respective compounds ( Figure 13). A conformational state of a protein is represented by the red and black dots. On the other hand, each red and black cluster shows the presence of distinct energetically favorable conformational spaces.
Furthermore, the first three eigenvalues of 5-HT1A alone, 5-HT2A alone, 5-HT1A-8 complex, and 5-HT2A-8 complex occupied 58.1%, 49.4%, 56.4%, and 46.1% conformational variances, respectively. These results indicate that there was a marginal increase in the flexibility of 5-HT1A in the presence of compound 8, while the flexibility of PPAR-γ in the presence of compounds 1 and 3, and flexibility of 5-HT2A in the presence of compound 8 were similar to those of PPARγ alone and 5-HT2A alone, respectively.

HPTLC Analysis of α-Linolenic in the Aerial Parts of S. irio
The developed HPTLC method was found to furnish a compact spot for α-linolenic acid at Rf = 0.57 ± 0.004 ( Figure S11A). The regression equation/correlation coefficient (r 2 ) for α-linolenic acid was Y = 6.49X + 2310.8/0.9971 in the linearity range of 100-1200 ng/band. The limits of detection (28.89 ng/band), quantification (87.57 ng/band), and recovery (98.16-99.26%) were found satisfactory for α-linolenic acid. The intra-/interday precisions (% RSD) for the proposed method were 1.24-1.48/1.14-1.43, which indicated a good precision for the proposed method. The amount of α-linolenic acid was estimated by comparing the peak area of the standard with that of crude extract ( Figure S11B,C). Figure S11D clearly reveals that all peaks of α-linolenic acid in the extract coincided with each other at the observed UV absorption maxima (λ max = 540). The estimated α-linolenic acid content in the hexane extract of aerial parts of S. irio was 28.67 µg/mg of dried extract.

Discussion
The chosen target for molecular docking analysis of the identified fatty acids was inspired by the activities reported in the literature for polyunsaturated fatty acids (PUFAs). PUFAs are known to reduce the risk of heart disease and heart attacks by refining blood lipids and endothelial function and by employing notable anti-inflammatory and antithrombotic effects [32]. They have a significant role in the treatment of type 2 diabetes through modulation of lipid and glucose homeostasis. They also play a vital role in Alzheimer's disease and in some cancers [33].
PPAR-γ or PPARG is the peroxisome proliferator-activated receptor gamma, also known as the glitazone reverse insulin resistance receptor. It is a type II protein-regulating gene encoded by the PPAR-γ gene [34]. Polyunsaturated fatty acids (PUFAs) are known to function as agonists of PPAR-γ, a nuclear receptor that has been getting increasing interest as a novel therapeutic target for the treatment of diabetes and related metabolic disorders [35]. Studies demonstrated that activation of PPAR-γ by PUFA ligands results in a number of biologically beneficial effects, including stimulation of lipid and glucose metabolisms, anti-inflammatory effects, and favorable cardiovascular effects [36].
The results of docking of the fatty acids (1-4) revealed moderate interaction with PPARγ active residues that formed stable complexes with relatively high free energy, compared to the standard drug rivoglitazone. Compound 3 formed the most stable complex with the highest binding affinity (−7.4 kcal mol −1 ). On the other hand, the close structural similarity of indole alkaloids, some of which are of plant origin (exogenous agonists), to the endogenous neurotransmitter serotonin might explain the potential neurological activity of these compounds, as depression is mostly triggered by an imbalance in serotonin levels [37].
Assessment of the ADME properties of the isolated compounds revealed that all investigated compounds, except the polyunsaturated fatty acids 1 and 2, showed high predicted GIT absorption and oral bioavailability. Regarding metabolism, CYP450-1A2 showed possible inhibition by compounds 5, 6, and 7, whereas CYP2D6 showed potential inhibition by compounds 3 and 4, in contrast to 8, which showed no inhibition to all CYP450 subtypes. P-gp is extensively distributed in the capillary endothelial cells of the BBB and contributes to pumping xenobiotics back into the blood [38]. Bypassing the P-gp drug-efflux mechanism is a crucial property for drugs used in neurodegenerative diseases [38,39]. In other words, compounds that are not P-gp substrates are predictably more bioavailable in the brain [28]. Since no predictable active efflux was observed by the P-gp for the investigated indole derivatives (5)(6)(7)(8), they can be delivered in appropriate concentrations to the brain and used in the treatment of neurological disorders, including depression [29], whereas 3 and 4 are probable P-gp substrates due to the presence of more rotatable bonds compared to indole derivatives (5-8) [40]. Also, no PAINS (Pan-assay interference compounds) alerts were detected for any of the tested compounds (1)(2)(3)(4)(5)(6)(7)(8).
A study conducted on the extract of Mitragyna speciosa, which contains the indole alkaloids mitragynine, paynantheine, and speciociliatine as major constituents, induced an antidepressant-like effect in mouse models; the effect was speculated to be through the interaction with the hypothalamic-pituitary-adrenal (HPA) axis in the neuroendocrine system [41].
Another plant, Passiflora incarnata L. (passion flower), containing the indole alkaloids harman, harmol, and harmine, reduced anxiety and improved memory in rats in a dosedependent manner. Cortical serotonin content was depleted, with increased levels of metabolites and increased turnover. It was found that the proposed mechanism of action of passion flower involved GABAA-receptors [42]. These facts motivated us to investigate the potential neurological activity of currently isolated indole alkaloids through docking into the 5-HT 1A and 5-HT 2A receptors. The serotonin receptor subtype 5-HT 1A has been implicated in several neurological conditions, and 5-HT 1A receptor agonism represents efficacious therapeutic potential for the treatment of major depression, anxiety, schizophrenia, and Parkinson's disease.

Apparatus and Chemicals
IR spectrum was acquired using a JASCO 320-A spectrometer (JASCO International Co., Ltd., Easton, MD, USA). Normal and reversed-phase silica gels (Merck, Darmstadt, Germany) were used for column chromatography (CC) and thin-layer chromatography (TLC). The compounds were visualized on TLC by spraying with 15% H 2 SO 4 /ethanol, followed by heating.
NMR spectroscopy was performed using deuterated solvents in an UltraShield Plus 500 (Bruker, Billerica, MA, USA) spectrometer operating at 500 MHz for 1 H and 125 MHz for 13 C at the College of Pharmacy, Prince Sattam Bin Abdulaziz University. The twodimensional NMR analyses (COSY, HSQC, and HMBC) were conducted using the standard Bruker pulse program. Chemical shift values are reported in δ (ppm) relative to an internal standard (TMS), and coupling constants (J) are reported in Hertz (Hz).
HRMS was performed using a Thermo Scientific UPLC RS Ultimate 3000 Q Exactive Hybrid Quadrupole-Orbitrap Mass Spectrometer (Mundelein, IL, USA) combined with high-performance quadrupole precursor selection with high resolution, accurate-mass (HR/AM) Orbitrap™ detection. The instrument was located at Prince Sattam Bin Abdulaziz University, College of Pharmacy. The detection was performed in negative and positive modes, and the experiment run time was 1 min using nitrogen as the supplementary gas with a scan range from 160-1500 m/z.

Plant Material
The aerial parts of S. irio L. were collected from a farm near Riyadh city in the Najd region of Saudi Arabia in March 2019 and kindly identified by a taxonomist at the Pharmacognosy Department, College of Pharmacy, King Saud University. A voucher specimen (no. 14380) has been deposited in the herbarium of the Pharmacognosy Department, College of Pharmacy.

Extraction and Isolation of Compounds
The shade-dried aerial parts of the plant (1 kg) were coarsely powdered and extracted with 80% ethanol. The ethanolic extract was concentrated under reduced pressure using a rotary evaporator (R-210, BUCHI) to give 28 g of brownish-black mass. The obtained extract was fractionated using different polarity solvents, starting with n-hexane (n-Hex.), followed by dichloromethane (CH 2 Cl 2 ), and finally n-butanol (n-BuOH) to obtain the corresponding fractions.
Part of the CH 2 Cl 2 fraction (6 g) was purified using a chromatotron (1 mm, mobile phase, 4% MeOH/CHCl 3 ) to obtain several subfractions ; further purification of sub-fraction 17 resulted in compound 6 in pure form.

Molecular Docking
Interaction of the active constituents of S. irio with serotonin receptors (5-HT 1A and 5-HT 2A ) and PPAR-γ was studied by performing molecular docking using AutoDock 4.2 [43,44]. The two-dimensional structures of ligands (active constituents) were drawn in ChemDraw Ultra 7.0 and converted to three-dimensional structures using OpenBabel. In ligands, Gasteiger partial charges were added, non-hydrogen atoms were merged, and rotatable bonds were defined using AutoDock Tools (ADT). The energies of all the ligands were minimized using the Universal Forcefield (UFF). The three-dimensional coordinates of different drug targets were obtained from the Protein Data Bank (www.rcsb.org, accessed on 21 August 2022). The X-ray crystal structure (PDB ID: 5U5L) of PPAR-γ in complex with rivoglitazone was resolved to 2.55 Å [10]. Similarly, the X-ray crystal structures of both serotonin receptors, namely 5-HT 1A (PDB ID: 7E2Y) and 5-HT 2A (PDB ID: 6A93), bound with serotonin and risperidone, respectively, were resolved to 3.00 Å [45,46].
Prior to molecular docking, the target proteins were cleaned by removing any heteroatoms, including non-essential water molecules, and adding hydrogen atoms. Also, Kollman-united atom type charges and solvation parameters were added with the help of ADT. For PPAR-γ, grid boxes were defined as 35Å × 35Å × 35Å centered at −5 Å, 33 Å, and 131 Å coordinates. Similarly, the dimensions of grid boxes of 5-HT 1A and 5-HT 2A were set at 28Å × 28Å × 28Å placed at 101 Å, 115 Å, and 108 Å; and 35Å × 28Å × 29Å centered at 16 Å, −0.2Å, and 60Å, respectively. Molecular docking was performed using the Lamarck Genetic Algorithm (LGA) along with the Solis and Wets search methods. The position, torsion, and orientation of ligands were set randomly, and all rotatable torsions were released. For each docking run, a maximum of 2.5 × 10 6 energy calculations were computed. The population size, translational step, quaternions, and torsion steps were set at 150, 0.2, 5, and 5, respectively. For each docking experiment, the lowest-energy docked structure was selected from 10 runs. Discovery Studio Visualizer was used to prepare and analyze the results and prepare figures. The dissociation constant (K d ) was evaluated from binding free energies (∆G) using the following equation.
where R and T were the universal gas constant (=1.987 cal/mol/K) and temperature (=298K), respectively.

Prediction of Physicochemical, Pharmacokinetic, Drug-Likeness, and Toxicity
The physicochemical, pharmacokinetic, drug-likeness, and toxicity properties of the investigated compounds were predicted using the SwissADME web tool hosted by the Swiss Institute of Bioinformatics (http://www.sib.swiss, accessed on 24 August 2022) [28].

Molecular Dynamics Simulation (MDS)
The MDS of PPARγ, 5-HT 1A , and 5-HT 2A along with their respective ligands (compounds 1, 3, 7, and 8) was performed using Desmond-2018 (Schrodinger, LLC, NY, USA), as described earlier [47,48]. The MDS was performed in an orthorhombic box by placing the initial protein-ligand docked pose at the center of the box, keeping a distance of at least 10 Å from the box boundaries. The simulation box was solvated with TIP3P water molecules, and Na + or Cl − ions were added to neutralize the system. Salt (150 mM NaCl) was added to the system to mimic the physiological condition. The system was iterated with 1000 steps with a convergence criterion of 1 kcal/mol/Å using an OPLS3e force field in order to minimize its energy. A 100 ns production run was initiated using the OPLS3e force field under NPT conditions of 298 K temperature and 1 bar pressure. A Nose-Hoover chain thermostat and Martyna-Tobias-Klein barostat were employed to maintain the NPT conditions of the system, respectively [49,50]. A time step of 2 fs was kept in all MDS, and at every 10 ps, energies and structures were saved in the trajectory. The trajectories were analyzed for root mean square deviation (RMSD), root mean square fluctuation (RMSF), radius of gyration (Rg), and solvent-accessible surface area (SASA).

Free Energy Calculations
The free energy of protein-ligand complex formation was computed by the MM-GBSA (Molecular Mechanics-Generalized Born Surface Area) approach using the Prime-2018 module (Schrodinger, LLC, New York, NY, USA), as described previously [51]. Briefly, the Molecular Mechanics (MM) approach was first used to locally optimize the docked complexes, and then their energies were minimized employing an OPLS3e force field along with a generalized Born surface area (GBSA) continuum solvent. The following relations were utilized to calculate the binding free energies of protein-ligand complexes: where E complex , E protein , and E ligand are the minimized energies of the protein-ligand complex, the protein alone, and the ligand alone, respectively; where G solv_GB(complex) , G solv_GB(protein) , and G solv_GB(ligand) are the free energies of solvation of the protein-ligand complex, the protein alone, and the ligand alone, respectively; and where G SA(complex) , G SA(protein) , and G SA(ligand) are the surface area energies of the proteinligand complex, the protein alone, and the ligand alone, respectively.
In the Prime-MM/GBSA method, the free energy is calculated as follows:

Principal Component Analysis (PCA) or Essential Dynamics (ED)
The collective motions of proteins along with their respective ligands were measured by employing a PCA or essential dynamics (ED) approach using the Bio3D package [52,53]. In this approach, first the protein's translational and rotational motions are disregarded, followed by the calculation covariance matrix and its eigenvectors by superimposing the protein's atomic coordinates onto a reference structure. Secondly, the symmetric matrix is diagonalized by an orthogonal transformation matrix, giving a diagonalized matrix of eigenvalues. The covariance matrix (C) is calculated using the following relation: where, N, x i/j , and <x i/j > represent the number of Cα-atoms, the Cartesian coordinates of the i th /j th Cα-atom, and time average of all the conformations, respectively.

Standardization of S. irio Extract by a Validated HPTLC Method
The standardization of S. irio extract was carried out by a validated high performance thin layer chromatography (HPTLC) method using α-linolenic acid as the marker compound. Chromatography was performed on a glass-backed silica gel 60 F 254 HPTLC plate (20 × 10 cm). Different combinations of solvents were tested to develop the HPTLC method, and a mixture of acetone, n-hexane, and acetic acid in the proportion of 25:75:0.1 v/v/v was selected as the most suitable mobile phase. Application of α-linolenic acid and the extracts on chromatographic plates (band wise) was carried out by an automatic TLC sampler-4 (ATS-4) while the development of the plate took place in ADC-2 (Automatic Development Chamber-2). Post development, the plate was derivatized with vanillin sulfuric acid reagent and heated to give compact bands of the chosen marker compound. It was scanned and quantified densitometrically at λ max = 540 nm. The developed method was validated for precision, recovery, robustness, limits of detection (LOD), and limits of quantification (LOQ) in accordance with ICH guidelines.

Conclusions
Chromatographic investigation of the aerial parts of the edible plant S. irio resulted in the isolation of eight compounds, of which four (1-4) are unsaturated fatty acids and the other four (5)(6)(7)(8) are identified as indole alkaloids. The structure of compound 4 was established as the fatty acid 8,11,12-trihydroxy-9Z,15Z-octadecadienoic acid, which is reported here for the first time from a natural source. Different spectroscopic techniques such as 1D, 2D NMR, and MS were employed to confirm the identity of the isolated compounds. Further, in silico molecular docking studies of compounds 1-4 were performed against PPAR-γ, which confirmed the agonist activity of compound 3 with a binding energy of −7.4 kcal mol −1 compared to the antidiabetic drug rivoglitazone. Similarly, molecular docking studies of compounds 5-8 were performed against serotonin receptor subtypes, namely 5-HT 1A and 5-HT 2A . Compound 8 exhibited notable docking scores, suggesting the strongest affinity among the tested indoles, with binding energies of −6.9 kcal/mol to 5HT 1A and −8.1 kcal/mol to 5HT 2A , respectively, against serotonin and risperidone as positive controls. The stability of target protein and compound complexes was tested by performing molecular dynamics simulations and analyzing parameters such as RMSD, RMSF, Rg, and SASA, which confirmed the formation of stable protein complexes. Further, principal component analysis (PCA) was used to collectively monitor the motion of amino acid residues of target proteins (PPAR-γ, 5-HT 1A , and 5-HT 2A ) in the presence of their respective compounds. In addition, an HPTLC method was developed for the quantification of the biomarker compound 2, which guarantees its application in quality control of commercialized herbal drugs and formulations containing α-linolenic acid. This study's outcome may serve as a scaffold to construct novel derivatives with higher potency and desirable drug-like properties. However, further validations through in vitro and in vivo studies are required.