INTRODUCTION
Narcolepsy is a persistent neurological condition marked by disturbances in sleep-wake regulation or fragmented sleep patterns [1]. Most often, this condition is recognized by an array of typical symptoms such as excessive daytime sleepiness (EDS), sleep paralysis, disrupted sleep-wake cycles, and hypnagogic/hypnopompic hallucinations [2]. It has been observed that, narcolepsy impairs brain’s functioning to regulate sleep-wake cycles by targeting the secretion of hypocretin neuropeptide in hypothalamus region, due to which individual experience an overwhelming urge to sleep throughout the day and difficulty during nocturnal sleep [3,4]. As per the International Classification of Sleep Disorders–Third Edition (ICSD3), this disorder can be classified into two subtypes narcolepsy type-1 (NT1) that is marked by the cataplexy, recognised by sudden weakening of muscle tone initiated by intense emotions and orexin deficiency, and narcolepsy type-2 (NT2), which lacks cataplexy [5,6]. A cross-sectional study by Niijima and Wakai [7] compared clinical and electrophysiological characteristics of 118 untreated patients with NT1 and NT2. NT1 patients were characterized by cataplexy and low cerebrospinal fluid (CSF) orexin-A levels and exhibited more severe EDS, higher body mass index, and increased prevalence of sleep paralysis compared to NT2 patients. Moreover, polysomnography and multiple sleep latency tests revealed shorter sleep latency and more frequent sleep-onset apid eye movement (REM) periods in NT1.
Orexin, also known as hypocretin, is a neuropeptide critical for regulating wakefulness and suppressing REM sleep [8]. Hypocretin deficiency is linked to the development of cataplexy, a hallmark symptom of narcolepsy [1]. Studies using neuroimaging and CSF analysis have shown that individuals with narcolepsy, particularly those with cataplexy, have significantly lower orexin levels in their CSF compared to healthy controls [9,10]. Orexin neurons typically inhibit REM sleep-associated muscle atonia (weakness or paralysis) during wakefulness [11]. Loss of these neurons results in the inappropriate intrusion of REM sleep-related atonia into wakefulness that trigger cataplexy [4]. Among these symptoms, EDS represents the most debilitating burden for patients, which significantly impact overall health, work productivity and quality of life. Current treatments for EDS often include stimulants such as modafinil, armodafinil, methylphenidate, and amphetamines along with solriamfetol, oxybates, and pitolisant [12]. Cataplexy and REM sleep disturbances are mainly managed with oxybates, antidepressants, and pitolisant [13]. These medications provide relief at some extent, while many individuals still experience persistent EDS and side effects that limit the effectiveness of the treatments [14]. For instance, systematic review and meta-analysis by Pitre et al. [15] assessed the effectiveness and safety of three wake promoting agents solriamfetol, modafinil, and pitolisant for managing EDS in patients with obstructive sleep apnea (OSA) and narcolepsy. Their study included 14 randomized controlled trials with 3,085 patients, assessing outcomes such as Maintenance of Wakefulness Test (MWT), Epworth Sleepiness Scale (ESS), and side effects. The findings confirm that solriamfetol, modafinil, and pitolisant significantly reduce EDS compared to placebo. Solriamfetol showed the greatest improvement in ESS scores (mean difference: −3.85, 95% confidence interval [CI]: −5.24 to −2.50) and MWT (standardized mean difference: 0.9, 95% CI: 0.64 to 1.17), followed by modafinil and pitolisant. However, the study highlighted that, many patients continued to experience EDS despite ongoing treatment. Common adverse effects included headache, insomnia, and anxiety. Both modafinil (relative risk [RR]: 2.01, 95% CI: 1.14 to 3.51) and solriamfetol (RR: 2.07, 95% CI: 0.67 to 6.25) were linked to a higher likelihood of treatment discontinuation due to these side effects. Furthermore, narcolepsy is frequently accompanied by other sleep dysfunctions including OSA, restless legs syndrome, and REM sleep behaviour illness, as well as comorbidities including obesity, depression, anxiety, and metabolic disorders, which further complicate treatment strategies [16]. For instance, Li et al. [17] suggested that depressive symptoms were found to be higher in individuals suffering with narcolepsy condition. Jennum et al. [18] observed autonomic response of orexin insufficiency and concluded NT1 link with cardiovascular disease. Moreover, study by Mohammadi et al. [19] mentioned, individuals with narcolepsy have greater occurrence of obesity, type 2 diabetes mellitus, dyslipidemia and hypertension in comparison with healthy individual.
One promising avenue for narcolepsy treatment involves targeting the brain’s histaminergic system, which plays a critical role in regulating wakefulness. Histaminergic neurons, located in the tuberomammillary nucleus of the posterior hypothalamus, project to wake-promoting brain regions and are integral to maintaining alertness and wakefulness [20]. Histamine, a small monoamine neurotransmitter enhances wakefulness by activating postsynaptic histamine 1 (H1) receptors on wake-promoting neurons [21]. Furthermore, presynaptic histamine 3 (H3) receptors act as auto and heteroreceptors that modifies the secretion of histamine along with other neurotransmitters such as dopamine, serotonin, and norepinephrine, which are also implicated in wakefulness and cataplexy regulation [22]. It has been noted that, inhibition of H3 receptors can influence the release of these neurotransmitters, which offers a potential way out to manage EDS and cataplexy [22]. Pitolisant is an H3 receptor inverse agonist/antagonist, approved for narcolepsy treatment and has demonstrated excellent efficacy in reducing EDS and cataplexy symptoms. However, certain limitations such as drug-drug interactions and cardiovascular safety concerns draw attention toward alternative H3-targeting agents [23,24]. Phytochemicals are naturally occurring compounds derived from plants, which possess great therapeutic strength to combat several neurological conditions such as Alzheimer’s disease [25], Parkinson [26], insomnia [27], and narcolepsy [4]. Unlike synthetic drugs, phytochemicals offer a favourable safety profile and reduced risk of drug interactions, which make them attractive candidates for therapeutic development. For example, Ansari et al. [28] examined the potential of phytocompounds derived from Withania somnifera leaf extract, and identified phytochemicals hexadecanoic acid, 2-methoxy-4-vinylphenol, N,N-bis(2-hydroxyethyl)dodecanamide, 1,2 benzene dicarboxylic acid diethyl ester, and oxacycloheptadec-8-en-2-one potentially inhibit monoamine oxidase A (MAO-A) enzyme, responsible for pathophysiology of anxiety and depression. Plant-based constituents have demonstrated potential in preventing and treating various diseases, particularly chronic conditions like cardiovascular disease, diabetes, and certain cancers. These compounds, found in fruits, vegetables, and other plant-based foods which exhibit antioxidant, anti-inflammatory, and immune-modulating properties [29]. For example, phytochemicals like flavonoids and isoflavones can help improve heart health by reducing inflammation, improving blood pressure, and preventing atherosclerosis [30]. Moreover they play essential role in regulating glucose metabolism, which improve insulin sensitivity, thus reducing the risk of type 2 diabetes [31]. Study reported that, certain phytochemicals have shown potential in preventing and fighting cancer by inhibiting tumor growth, promoting apoptosis (programmed cell death), and protecting against DNA damage [32,33].
W. somnifera also widely recognized as ashwagandha (Hindi), Indian ginseng, or winter cherry. This is an evergreen shrub of the Solanaceae family that has been a cornerstone of traditional Indian medicine for centuries [34]. W. somnifera bears a solitary axillary inflorescence and produces a globose berry enclosed in a persistent calyx. The fruit is typically red when ripe and contains numerous seeds [35]. The estimated production of W. somnifera is estimated at around 1,500 tonnes per year, while annual demand reaches up to 7,000 tonnes. India stands as the leading producer and exporter, cultivating it on over 10,780 hectares and yielding approximately 8,429 tonnes, with Madhya Pradesh, Gujarat and Haryana as the primary growing regions [36,37]. Extensively used in Ayurvedic and Unani systems, this plant is primarily cultivated for its roots, which are rich in withanolides bioactive steroidal lactones known for diverse pharmacological effects [38]. Native to the drier regions of India, the Middle East, and parts of Africa and Asia, W. somnifera has been historically utilized as a Rasayana (rejuvenative agent) for promoting longevity, vitality, and resistance to stress and disease [39]. It has demonstrated therapeutic potential in neurological disorders such as epilepsy, Alzheimer’s, Parkinson’s disease, and cerebral ischemia, along with anti-inflammatory, cardioprotective, hepatoprotective, and anticancer properties [40–42]. Plant parts such as roots, leaves, flowers, fruits, and seeds are utilized in managing a wide array of ailments, which may include arthritis, infertility, skin disease, and microbial infections that makes it a valuable multipurpose herb in both traditional and modern medicine [43]. Moreover, W. somnifera exhibits notable neuroactive properties such as GABA-mimetic effects and modulation of serotonergic and cholinergic neurotransmission, which support its potential to alleviate stress, enhance cognition, and support sleep regulation [44]. Study mentions that, withanolide-rich extracts from W. somnifera have demonstrated chronobiotic activity by promoting non-REM sleep and improving spatial memory in sleep-deprived animal models [45].
This study explores the therapeutic potential of selected phytochemicals derived from W. somnifera as H3 receptor inhibitors, aiming to establish their efficacy and safety in preclinical models of narcolepsy. By leveraging the wake-promoting effects of these compounds, this research seeks to contribute to the development of novel, plant-based interventions for the management of EDS and other narcolepsy symptoms.
METHODS
Site description and climatic conditions
The investigation was carried out inside a controlled greenhouse facility at the Department of Botany, Aligarh Muslim University (AMU), situated at coordinates 27.9135° N latitude and 78.0782° E longitude in Aligarh, India. The region experiences a humid subtropical climate, with ambient temperatures typically ranging from 28°C to 38°C during the experimental period. A randomized complete block design was employed for the experimental setup. Plants were grown under natural light conditions, maintaining a daily photoperiod consisting of 16 hours of light followed by 8 hours of darkness.
Plant propagation and experimental setup
Seeds of W. somnifera were obtained from the Central Institute of Medicinal and Aromatic Plants (CIMAP), Lucknow. To ensure sterility, the seeds were initially treated with 70% ethanol (HPLC-grade), followed by immersion in a 4% sodium hypochlorite solution (both reagents sourced from Sigma-Aldrich). Subsequently, the seeds were rinsed multiple times with sterile distilled water to eliminate any chemical residues. Sterilized sandy loam soil (3 kg) was transferred into earthen pots measuring 25 cm in diameter and height of 30 cm. Five seeds were sown in each pot, and two weeks after emergence, seedlings were thinned to retain only one vigorous plant per pot. Soil moisture was maintained at approximately 80% of the field capacity by adding 300 mL of water per pot as needed.
Extraction and analysis of essential oils
Fresh leaves of W. somnifera (approximately 250 grams) were subjected to hydro-distillation using a Clevenger-type apparatus (Borosil) for 3 hours to obtain essential oil. The obtained oil was dried using anhydrous sodium sulfate (Na2SO4; Sigma-Aldrich), to eliminate any remaining moisture. The chemical composition of the essential oil was analyzed through gas chromatography–mass spectrometry (GC–MS) employing a PerkinElmer® Clarus® 680 GC system coupled with a Clarus SQ 8 mass spectrometer [46]. The separation was carried out on a TG WAX MS fused silica capillary column. Operating conditions included an injector temperature of 260°C, a column oven starting at 50°C, and a split ratio of 110. Helium served as the carrier gas at a flow rate of 1.21 mL/min. The ion source and interface temperatures were set at 220°C and 270°C, respectively, with a mass scan range of 40 to 650 m/z [47].
Essential oil component identification
Individual compounds in the essential oil were identified by comparing their mass spectra and retention indices with those available in the NIST-11 and Wiley spectral libraries. Additional comparisons were made with literature-reported data to ensure accurate compound characterization. The relative concentrations of identified constituents were calculated as percentages based on their peak areas in the total ion chromatogram.
Ligand retrieval and preparation
Ligands were retrieved from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/) based on their chemical identifiers [48]. The three-dimensional (3D) structures of the ligands were downloaded in SDF format. Preparation of the ligands was performed by UCSF Chimera (version 1.16) [49]. The preparation process involved adding hydrogen atoms, assigning Gasteiger charges, and minimizing the energy of the ligand structures using the AMBER ff14SB force field [50]. The prepared ligands were formatted as PDBQT files for further molecular docking analysis.
Retrieval of protein and its preparation
The structure of the target protein was retrieved from the Research Collaboratory for Structural Bioinformatics Protein Data Bank (RCSB PDB, https://www.rcsb.org/) in PDB format [51]. Protein preparation was performed utilizing UCSF Chimera (version 1.16) [49]. Protein preparation steps included water molecules removal, ligands, and non-standard residues, followed by the addition of hydrogen atoms and assignment of AMBER ff14SB force field charges. The final prepared protein structure was formatted in PDBQT format for docking purpose.
Molecular docking
Molecular docking was conducted using PyRx (version 0.8) employing the AutoDock Vina algorithm [52]. The prepared ligand and protein structures were imported into PyRx, and the binding site was defined using a grid box cantered on the active site of protein. The dimensions of the grid box were recorded as 45.90× 103.64×46.86 Å, while centred at −19.58, 45.36, 1.16. Docking was carried out using an exhaustiveness value of 8, with the highest-scoring binding poses chosen according to binding affinity (kcal/mol). The top docked conformations of the complexes were saved for subsequent analysis.
Physicochemical and drug-likeness properties
The physicochemical characteristics of the ligands were assessed using the SwissADME server (http://www.swissadme.ch/) [53]. Essential parameters such as molecular weight, logP, hydrogen bond acceptors, and acceptors were analyzed to assess drug-like characteristics. Drug-likeness was further evaluated using the Molsoft Drug likeness and Molecular Property Prediction tool (http://molsoft.com/mprop/) [54], which provided a drug likeness score in accordance of Lipinski’s Rule of Five and other relevant filters.
Blood–brain barrier
Blood-brain barrier (BBB) is highly selective and semipermeable in nature that shields from harmful substances in the bloodstream while regulating the transport of essential molecules. For drugs targeting neurological or sleep-related disorders, such as narcolepsy, insomnia, epilepsy, or depression, the ability to effectively cross the BBB is crucial for therapeutic efficacy. To assess the BBB permeability of selected drug-like compounds, an in silico prediction was performed using the MolSoft online server (https://molsoft.com/mprop/) [55]. This tool evaluates pharmacokinetic and physicochemical properties, including the likelihood of a compound crossing the BBB, which is a critical consideration for central nervous system (CNS)-targeted drug candidates.
ADMET analysis
The Deep-PK server (https://biosig.lab.uq.edu.au/deeppk/) was used to predict the absorption, distribution, metabolism, excretion, and toxicity (ADMET) properties of the ligands [56]. The server provided complete insights into pharmacokinetic parameters, including solubility, permeability, and metabolic stability.
Molecular dynamics simulation
Molecular dynamics (MD) simulations were executed using the WebGro server (https://simlab.uams.edu/) [42] to assess the stability of the protein-ligand complex. The top-ranked docked complex was used as the starting structure. The GROMACS engine with the GROMOS96 43a1 force field was employed for simulation of 50 ns. The system was solvated in a cubic box configured with TIP3P water molecules, neutralized with counterions, and subjected to energy minimization. The simulation was executed under periodic boundary conditions at 300 K and 1 atm pressure, employing a time step of 2 fs. Trajectory analysis included root mean square deviation (RMSD), root mean square fluctuation, and hydrogen bond interactions to evaluate the stability of the complex [57].
MM/GBSA binding free energy calculation
Binding free energy calculations for the protein-ligand complexes were calculated using the molecular mechanics/generalized Born surface area (MM/GBSA) approach implemented in Schrödinger Maestro 11.2 [58,59]. The top-ranked docked complex from the molecular docking study was used as input structure. The MM/GBSA calculation was performed using the OPLS4 force field and the VSGB 2.0 solvation energy model. Binding free energy (ΔGbind) was calculated according to the following equation:
where Gcomplex is free energy of the protein ligand complex, Gprotein is free energy of the unbound protein, and Gligand is free energy of the unbound ligand.
Each free energy term was calculated as:
where EMM is molecular mechanics energy (sum of bonded, electrostatic, and van der Waals interactions), Gsolv is solvation free energy (polar contribution from the generalized Born model), and GSA is non-polar solvation energy (proportional to the solvent-accessible surface area).
The calculations were performed without entropy contributions to reduce computational cost, focusing on enthalpic contributions to binding affinity. The binding free energy was reported in kcal/mol.
Visualization of interactions
Protein-ligand interactions were visualized using Schrodinger Maestro 11.2 [60]. The docked complexes were analysed to identify key interactions such as hydrogen bonds, hydrophobic interactions, and pi-pi stacking. Two-dimensional (2D) interaction diagrams and 3D representations were generated to illustrate the binding mode and critical residues involved in ligand binding.
RESULTS
GC–MS profiling
The GC–MS analysis was performed to characterize the phytochemical profile of W. somnifera leaf extract cultivated under greenhouse conditions. The chromatographic spectrum (Figure 1) and the corresponding compound list (Table 1) revealed a total of 43 phytoconstituents. Among these, the predominant compounds included 2-methoxy-4-vinylphenol (1.84%), 3-methylbutyl hexanoate (3.01%), N,N-bis(2-hydroxyethyl)dodecanamide (1.02%), 1,2-benzenedicarboxylic acid diethyl ester (0.63%), tetradecanoic acid (1.09%), hexadecanoic acid methyl ester (1.09%), phytol (0.89%), oxacycloheptadec-8-en-2-one (0.70%), and 1,8,11-heptadecatriene (Z,Z) (1.63%). Alongside these, a wide array of minor constituents was also detected, collectively contributing to the extract’s rich and diverse phytochemical landscape.
Molecular docking of the phytochemicals against H3R
Table 2 summarizes the molecular docking results obtained from a virtual screening study performed using PyRx, targeting histamine H3 receptor, a G-protein coupled receptor (GPCR) implicated in pathophysiology of narcolepsy. Docking scores reflect the predicted binding affinities between each compound and the receptor, with more negative values indicating stronger theoretical binding potential. Among all screened molecules, methotrexate exhibited strongest binding affinity with score of −8.8 kcal/mol. This antifolate chemotherapeutic agent is structurally complex and possesses multiple functional groups capable of forming strong interactions including hydrogen bonds and electrostatic contacts with receptor binding pocket. Its high docking score suggests a significant interaction profile. 1,2-Benzenedicarboxylic acid exhibited a docking score of −6.9 kcal/mol, indicating a moderate to strong affinity for the H3 receptor. As an aromatic dicarboxylic acid, its binding may be driven by π-π stacking interactions and hydrogen bonding, which supports its consideration as a low-molecular-weight ligand scaffold in further structure-activity relationship studies. Octadecenoic acid, a monounsaturated fatty acid, showed a docking score of −6.6 kcal/mol, suggesting effective receptor interaction potentially mediated through hydrophobic contacts. Fatty acids are known to interact with lipid-facing regions of GPCRs, and this compound may modulate receptor conformation allosterically. Adenosine 3′,5′-cyclic monophosphate (cAMP), an endogenous secondary messenger, yielded a notable docking score of −7.3 kcal/mol. Although not traditionally associated with direct GPCR binding, this result may reflect possible allosteric modulation or interaction with intracellular receptor domains, indicating a unique mechanism of action worth further investigation. 9,12-Octadecadienoyl chloride, (Z,Z)−, a chlorinated linoleic acid derivative, scored −6.2 kcal/mol, which points to moderate binding affinity. Its interaction likely arises from both hydrophobic and potential polar contacts facilitated by its reactive acid chloride group, presenting opportunities for structural derivatization and lead optimization. These top five phytocompounds on the basis of docking scores were selected for further analysis.
Physicochemical properties of the compounds
The physiochemical characteristics of screened phytocompounds investigated by the SwissADME webserver, which provide critical insights into their drug-likeness and potential suitability for therapeutic development (Table 3). Methotrexate (molecular weight [MW]: 454.44 g/mol) exhibits a balanced profile with moderate molecular weight, high polar surface area (topological polar surface area [TPSA]: 210.54 Å2), and significant hydrogen-bonding capacity (9 acceptors, 5 donors), aligning with its known bioavailability despite its relatively large size. Its aromaticity (16 aromatic heavy atoms) and rotatable bonds (n= 10) suggest conformational flexibility, which may influence receptor binding kinetics. 1,2-Benzenedicarboxylic acid (MW: 278.34 g/mol) displays lower polarity (TPSA: 52.60 Å2) and no hydrogen-bond donors, which could enhance membrane permeability but may limit solubility. Its rigid aromatic core (6 aromatic atoms) and moderate rotatable bonds (n=10) indicate potential for stable interactions with hydrophobic receptor pockets. Octadecenoic acid (MW: 282.46 g/mol), a long-chain fatty acid, shows high lipophilicity (Fraction Csp3: 0.83) and low polar surface area (37.30 Å2), typical of lipid-like molecules. Its 15 rotatable bonds suggest high flexibility, which may aid in adapting to receptor conformations but could pose challenges for metabolic stability.
cAMP (MW: 329.21 g/mol) stands out with its high polarity (TPSA: 164.65 Å2) and hydrogen-bonding features (9 acceptors, 3 donors), consistent with its role as a hydrophilic signaling molecule. Its limited rotatable bonds (n=1) imply restricted conformational freedom, potentially favoring selective binding. 9,12-Octadecadienoyl chloride (MW: 298.89 g/mol) exhibits low polarity (TPSA: 17.07 Å2) and no hydrogen-bond donors, suggesting high lipophilicity and likely rapid membrane penetration. However, its reactive acyl chloride group may raise concerns about chemical stability and off-target effects. All compounds show drug-likeness as per Lipinski’s Rule of Five.
Validation of drug-likeness using Molsoft
The drug-likeness scores predicted by the Molsoft server provide critical insights into therapeutic potential of these phytocompounds as viable candidates for possible drug development (Figure 2). An ideal drug-likeness score typically falls above 0.5, reflecting excellent bioavailability, metabolic stability, and adherence to established guidelines like Lipinski’s Rule of Five. Scores between 0.2 and 0.5 suggest good drug potential, though minor optimizations may be needed to improve pharmacokinetics. Compounds scoring between 0 and 0.2 are considered marginal, often requiring significant modifications to address issues like poor solubility or absorption. Scores below 0 indicate poor drug-likeness, typically due to structural features that violate multiple drug-like criteria, such as excessive molecular weight, high reactivity, or unfavorable polarity. Methotrexate exhibits a modest drug-likeness score of 0.29, reflecting its relatively large molecular size and complexity, which may impact its pharmacokinetic properties despite its established clinical use. 1,2-Benzenedicarboxylic acid shows a slightly better score of 0.38, suggesting a more favorable balance of physicochemical properties, though its lack of hydrogen-bond donors could limit solubility. Octadecenoic acid has a marginal score of 0.30, indicative of its highly lipophilic nature, which may pose challenges for formulation and absorption. In contrast, cAMP achieves the highest score among the evaluated compounds (0.45), aligning with its balanced molecular features and polar yet drug-like characteristics. However, the notably poor score of 9,12-octadecadienoyl chloride (−1.18) highlights its non-drug-like properties, likely due to its reactive acyl chloride group, which raises concerns about stability and toxicity.
BBB permeability
BBB penetration scores provide important insights into the potential of these compounds to reach therapeutic targets in the CNS, which is particularly relevant for treating brain-related or sleep disorders like narcolepsy (Figure 3). The scores indicate that methotrexate (4.25), 1,2-benzenedicarboxylic acid (4.18), octadecenoic acid (4.40), and 9,12-octadecadienoyl chloride (4.36) all exhibit high BBB penetration potential (scores ≥4), suggesting they could effectively cross the BBB and reach CNS targets. These high scores are likely attributed to their moderate molecular weights and balanced lipophilicity, properties known to facilitate passive diffusion across the BBB. In contrast, cAMP shows a significantly lower BBB score (2.01), indicating poor penetration. This is expected due to its high polarity and molecular features that hinder passive diffusion, a common challenge for hydrophilic compounds. For cAMP to be effective in CNS disorders, structural modifications or advanced delivery strategies (e.g., prodrugs or nanocarriers) would be necessary to enhance its brain uptake. For sleep-related disorders like narcolepsy, targeting the histamine H3 receptor or other CNS pathways requires compounds with BBB scores ≥4 to ensure sufficient brain exposure. While the lipid-like compounds (e.g., octadecenoic acid) show excellent BBB penetration, their drug-likeness scores (from previous analysis) suggest potential metabolic stability or solubility issues that may need optimization. Conversely, methotrexate’s strong BBB penetration and known pharmacological profile make it a promising candidate for repurposing, though its broader biological effects would require careful evaluation.
ADMET analysis
Table 4 details the ADMET profiles of five phytocompounds methotrexate, 1,2-benzenedicarboxylic acid, octadecenoic acid, cAMP, and 9,12-octadecadienoyl chloride (Z,Z), evaluated for their potential in drug design, particularly for targeting the histamine H3 receptor in narcolepsy treatment. In drug design, ADMET properties are critical as they determine how a compound behaves in the body, including how well it is absorbed, distributed, metabolized, excreted, and whether it poses toxicity risks. Absorption analysis predicts that all five compounds are likely to be absorbed in the human intestine, thus they are promising for oral drug delivery. Most compounds, except octadecenoic acid, are predicted to be bioavailable at a 20% threshold, meaning they can reach systemic circulation after oral administration. None of the compounds act as P-glycoprotein inhibitors or substrates, reducing the likelihood of being actively pumped out of cells, which is favorable for maintaining drug levels. Distribution properties show that all compounds have plasma protein binding values (4.85% to 37.35%) well below 90%, indicating a favorable therapeutic index with sufficient free drug available for action. The fraction unbound (0.33 to 1.51) and steady-state volume of distribution (0.49 to 2.74 L/kg) vary, with octadecenoic acid and 9,12-octadecadienoyl chloride showing higher unbound fractions, suggesting better availability in tissues, while methotrexate has a low distribution volume, indicating it stays more in plasma. Metabolism profiles reveal differences in cytochrome P450 (CYP) enzyme interactions. 1,2-Benzenedicarboxylic acid, octadecenoic acid, and 9,12-octadecadienoyl chloride inhibit multiple CYP enzymes (e.g., CYP1A2, CYP2C19, CYP2C9, CYP3A4), which could lead to drug-drug interactions by slowing the metabolism of other drugs. Octadecenoic acid and 9,12-octadecadienoyl chloride are also CYP2C9 substrates, meaning they may be metabolized by this enzyme. Methotrexate and cAMP show minimal CYP interactions, which is advantageous for reducing metabolic complications. None of the compounds interact with OATP1B1 or OATP1B3 transporters, simplifying their metabolic profile. Excretion data indicate varied clearance rates, with 1,2-benzenedicarboxylic acid having the highest clearance (7.68 mL/min/kg) and octadecenoic acid the lowest (−0.8 mL/min/kg, suggesting potential accumulation). Half-life predictions show methotrexate and 1,2-benzenedicarboxylic acid have short half-lives (<3 hours), requiring frequent dosing, while the others have longer half-lives (≥3 hours), supporting less frequent administration. Toxicity profiles are generally favorable, with all compounds predicted to be safe for Ames mutagenesis (no DNA damage), avian toxicity, hERG blockers (no heart rhythm issues), and various nuclear receptor-related toxicities (e.g., NR-AhR, NR-AR, NR-ER). However, 9,12-octadecadienoyl chloride is flagged as potentially toxic for Ames mutagenesis and carcinogenesis, raising concerns about its safety for long-term use.
MD simulation
MD simulations conducted by using the WebGro server to analyze four key properties solvent accessible surface area (SASA), RMSD, radius of gyration (Rg), and number of hydrogen bonds over time to assess the stability and dynamics of the protein-ligand complexes.
Figure 4 shows the SASA of the H3 receptor in complex with the five compounds over 50 ns. SASA represents surface area of the protein, accessible to the solvent (water), providing insight into the protein’s exposure and conformational changes. Initially, all complexes exhibit SASA values around 225 nm2, but they quickly fluctuate and stabilize around 210–220 nm2 after 1,000 ps (1 ns). Methotrexate (red) and 1,2-benzenedicarboxylic acid (purple) show slightly higher SASA values (around 220 nm2) throughout the simulation, indicating greater solvent exposure, possibly due to looser binding or conformational shifts that expose more of the protein surface. In contrast, cAMP (cyan) and octadecenoic acid (blue) maintain lower SASA values (around 210 nm2), suggesting a more compact or stable interaction that shields the protein from the solvent. 9,12-Octadecadienoyl chloride (green) fluctuates between these extremes, indicating moderate stability. The overall trend of decreasing SASA after the initial phase suggests that the complexes undergo minor conformational adjustments early on, stabilizing thereafter, which is a positive sign for consistent binding behaviour over the simulation period.
The number of hydrogen bonds formed between the H3 receptor and each ligand over 50 ns is depicted in Figure 5. Hydrogen bonds are critical for stabilizing protein-ligand interactions, contributing to binding affinity. Initially, the number of hydrogen bonds spikes to around 330 for all complexes, likely due to the system adjusting during equilibration. However, this drops rapidly within the first 500 ps to around 200–250 bonds and fluctuates ranging 180 to 260 for the remainder of simulation. Methotrexate (red) and cAMP (cyan) consistently maintain a higher number of hydrogen bonds (around 220–250), indicating stronger and more stable interactions with the receptor, which aligns with their favorable docking scores (−8.8 and −7.3, respectively). In contrast, 9,12-octadecadienoyl chloride (green) and octadecenoic acid (blue) show fewer hydrogen bonds (around 180–200), suggesting weaker interactions, consistent with their moderate docking scores (−5.2 and −5.6). 1,2-Benzenedicarboxylic acid (purple) fluctuates around 200–220 bonds, reflecting intermediate stability. The overall decrease and subsequent stabilization of hydrogen bonds suggest that after initial rearrangements, the complexes reach a dynamic equilibrium, with methotrexate and cAMP forming more persistent interactions, making them promising candidates for further study.
Figure 6 displays the RMSD of the H3 receptor backbone in each complex over 50 ns. RMSD measures the structural deviation of the protein from its initial conformation, serving as an indicator of stability—lower and more stable RMSD values suggest a more rigid, stable complex. All complexes start with an RMSD of around 0.2 nm, but this increases rapidly within the first 10 ns, peaking at around 1.5–2.0 nm. Octadecenoic acid (blue) reaches the highest peak (around 2.0 nm) around 15 ns, indicating significant structural adjustment, before stabilizing at 1.0–1.5 nm. Methotrexate (red) and cAMP (cyan) show lower RMSD values (stabilizing around 0.5–1.0 nm after 20 ns), suggesting greater structural stability, which correlates with their higher docking scores and more hydrogen bonds. 9,12-Octadecadienoyl chloride (green) and 1,2-benzenedicarboxylic acid (purple) fluctuate between 0.8–1.5 nm, indicating moderate stability. The initial rise in RMSD reflects the protein adapting to the ligand, while the stabilization after 20 ns suggests that the complexes reach a stable conformation, with methotrexate and cAMP exhibiting the least deviation, making them more structurally reliable candidates for narcolepsy drug design.
Rg of the H3 receptor in each complex over 50 ns is depicted in Figure 7. Rg measures the compactness of the protein structure lower and more stable Rg values specify a more compact, stable protein. All complexes start with an Rg of around 2.6 nm, increasing to 2.8–3.1 nm within the first 1,000 ps (1 ns) as the protein adjusts to the ligand. 9,12-Octadecadienoyl chloride (green) and octadecenoic acid (blue) show the highest Rg values, peaking at 3.1 nm and stabilizing around 2.9–3.0 nm, suggesting a less compact structure and potentially weaker binding, consistent with their lower docking scores. Methotrexate (red) and cAMP (cyan) maintain lower Rg values (around 2.7–2.8 nm after 20 ns), indicating a more compact and stable conformation, aligning with their better RMSD profiles and higher docking scores. 1,2-Benzenedicarboxylic acid (purple) fluctuates around 2.8–2.9 nm, showing intermediate compactness. The overall trend of Rg stabilizing after the initial increase suggests that the protein-ligand complexes achieve a consistent structural state, with methotrexate and cAMP demonstrating greater compactness, supporting their potential as more stable candidates for targeting the H3 receptor in narcolepsy treatment.
MMGBSA free energy calculation
MM/GBSA binding free energy scores (ΔG_bind) for five compounds—methotrexate, 1,2-benzenedicarboxylic acid, octadecenoic acid, cAMP, and 9,12-octadecadienoyl chloride, (Z,Z)− —interacting with the histamine H3 receptor, was calculated using Schrödinger Maestro (Figure 8). The MM/GBSA ΔG_bind scores, reported in kcal/mol, estimate the binding affinity of each ligand to the receptor, with more negative values indicating stronger binding. The scores range from −25.54 to −52.96 kcal/mol, reflecting varying binding strengths. 9,12-Octadecadienoyl chloride, (Z,Z)− stands out with the most favorable ΔG_bind score of −52.96 kcal/mol, suggesting the strongest binding affinity among the compounds, which could imply a highly stable interaction with the H3 receptor. Octadecenoic acid follows with a ΔG_bind of −36.67 kcal/mol, and 1,2-benzenedicarboxylic acid scores −34.09 kcal/mol, both indicating good binding potential, though less than 9,12-octadecadienoyl chloride. Methotrexate and cAMP have ΔG_bind scores of −25.96 and −25.54 kcal/mol, respectively, suggesting relatively weaker binding compared to the others. However, these MM/GBSA results must be interpreted in context with other data, such as the docking scores, ADMET profiles, and MD simulation results. For instance, while 9,12-octadecadienoyl chloride shows the best MM/GBSA score, its docking score (−5.2) was moderate, and its ADMET profile indicated toxicity concerns (Ames mutagenesis and carcinogenesis), which could limit its suitability as a drug candidate. In contrast, methotrexate and cAMP, despite having less favorable MM/GBSA scores, demonstrated better docking scores (−8.8 and −7.3), more stable MD trajectories (lower RMSD, Rg, and higher hydrogen bonds), and safer ADMET profiles, making them potentially more viable for further development. The MM/GBSA scores highlight the binding potential of these compounds, but their overall suitability for narcolepsy treatment via H3 receptor targeting requires balancing these energies with pharmacokinetic and safety considerations.
Visualization of interaction
These 2D interaction diagrams, generated using Schrödinger Maestro’s Ligand Interaction Diagram tool, illustrate the specific molecular interactions between each ligand and the receptor’s amino acid residues, highlighting hydrogen bonds, hydrophobic interactions, pi-stacking, and other forces that contribute to binding affinity (Figure 9).
9,12-Octadecadienoyl chloride (Z,Z)−, a long-chain fatty acid derivative, forms multiple interactions with the receptor. It engages in hydrogen bonds (blue dashed lines) with TYR 91 and TYR 94, involving the ligand’s chlorine group and the hydroxyl groups of the tyrosine residues, which likely contribute to its strong MM/GBSA binding free energy (−52.96 kcal/mol). Hydrophobic interactions (green dashed lines) are extensive, involving ALA 190, VAL 95, TYR 91, TYR 94, CYS 188, PHE 193, and TRP 110, reflecting the ligand’s non-polar nature and its ability to fit snugly into a hydrophobic pocket of the receptor. Pi-stacking (red dashed lines) occurs with TYR 115, enhancing the binding stability through aromatic interactions. Additional residues like ARG 27, HYS 187, GLY 89, GLU 395, ASP 114, and PHE 398 contribute to the interaction network, with some (e.g., GLU 395, ASP 114) being negatively charged (orange circles), potentially interacting with the ligand’s polar regions. The combination of hydrogen bonds, hydrophobic interactions, and pi-stacking explains the ligand’s strong binding affinity, though its toxicity concerns (Ames mutagenesis and carcinogenesis) from the ADMET profile may limit its drug candidacy.
Octadecenoic acid forms hydrogen bonds with TYR 189 and CYS 188 via its carboxylic acid group, stabilizing the ligand in the binding pocket and contributing to its favorable MM/GBSA score (−36.67 kcal/mol). Hydrophobic interactions dominate, involving ALA 190, VAL 95, TYR 94, TYR 91, PHE 193, TRP 110, and LEU 111, consistent with the ligand’s long hydrocarbon chain, which likely nestles into a non-polar region of the receptor. Pi-stacking is observed with TYR 115, adding to the binding strength through aromatic interactions. Residues like ARG 27, HYS 187, GLY 89, GLU 395, ASP 114, and PHE 398 are also involved, with negatively charged GLU 395 and ASP 114 potentially interacting with the ligand’s polar head. The extensive hydrophobic interactions and hydrogen bonds align with octadecenoic acid’s good binding affinity, though its moderate docking score (−5.6) and higher RMSD in MD simulations suggest less structural stability compared to other ligands like Methotrexate.
cAMP forms multiple hydrogen bonds, primarily through its phosphate and hydroxyl groups, with ARG 132, ASN 70, HIS 416, and LYS 359, and through its adenine ring with ARG 349 and SER 309. These interactions, shown as blue dashed lines, are critical for its binding, contributing to its MM/GBSA score (−25.54 kcal/mol). Hydrophobic interactions involve VAL 356 and CYS 115, though they are less extensive due to the ligand’s polar nature. A water-mediated hydrogen bond (light blue curved line) with GLN 146 further stabilizes the complex. Pi-stacking occurs with HIS 417, enhancing the binding through aromatic interactions with the adenine ring. The presence of negatively charged ASP 393 (orange circle) near the phosphate group suggests electrostatic interactions, while polar residues like SER 309 (light blue circle) support the hydrogen bonding network. The extensive hydrogen bonding and pi-stacking align with cAMP’s stable MD simulation profile (low RMSD, high hydrogen bonds) and good docking score (−7.3), making it a promising candidate despite a less negative MM/GBSA score.
1,2-Benzenedicarboxylic acid forms hydrogen bonds with CYS 188 and TYR 94 via its carboxylic acid groups, contributing to its MM/GBSA score (−34.09 kcal/mol). Hydrophobic interactions are significant, involving ALA 190, VAL 95, TYR 91, PHE 193, TRP 110, LEU 101, and TYR 94, reflecting the ligand’s aromatic and non-polar regions fitting into the receptor’s hydrophobic pocket. Pi-stacking with TYR 94 and TYR 189 further stabilizes the complex through aromatic interactions. Residues like GLU 395 (negatively charged) and GLY 89 (polar) are also involved, potentially interacting with the ligand’s polar groups. The combination of hydrogen bonds, hydrophobic interactions, and pi-stacking supports its good binding affinity, though its moderate docking score (−5.9) and MD simulation results (intermediate RMSD and Rg) suggest it may not be as stable as methotrexate or cAMP.
Methotrexate forms an extensive network of hydrogen bonds with ARG 132, ARG 349, SER 309, LYS 359, HIS 416, and ASN 70, primarily through its pteridine ring, glutamic acid tail, and amide groups, which significantly contribute to its binding stability and MM/GBSA score (−25.96 kcal/mol). A water-mediated hydrogen bond with GLN 146 further enhances the interaction. Hydrophobic interactions are present with VAL 356 and CYS 115, though less dominant due to Methotrexate’s polar nature. Pi-stacking with HIS 417, involving the pteridine ring, adds to the binding strength. Negatively charged ASP 393 and polar SER 309 residues support the electrostatic and hydrogen bonding interactions. The extensive hydrogen bonding network aligns with Methotrexate’s excellent docking score (−8.8), stable MD simulation profile (low RMSD, Rg, and high hydrogen bonds), and favorable ADMET profile, making it a strong candidate for narcolepsy drug design despite a less negative MM/GBSA score compared to 9,12-octadecadienoyl chloride.
In summary, the interaction profiles reveal that 9,12-octadecadienoyl chloride and octadecenoic acid rely heavily on hydrophobic interactions and pi-stacking, contributing to their strong MM/GBSA scores, but their stability and safety profiles are less favorable. Methotrexate and cAMP form extensive hydrogen bonding networks, supported by pi-stacking, which aligns with their stability in MD simulations and safer ADMET profiles, making them more promising candidates for targeting the H3 receptor in narcolepsy treatment. 1,2-Benzenedicarboxylic acid shows a balanced interaction profile but lacks the stability of methotrexate and cAMP. These insights guide the selection of lead compounds for further experimental validation.
DISCUSSION
This study represents a significant step toward identifying novel plant-based therapeutic agents for the management of narcolepsy, which is a debilitating neurological disorder characterized by EDS, cataplexy, and disrupted sleep-wake cycles. By leveraging the rich phytochemical profile of W. somnifera (ashwagandha), a cornerstone of traditional Ayurvedic medicine, this research explored the potential of its bioactive compounds as histamine H3 receptor (H3R) inhibitors to address the unmet needs in narcolepsy treatment. Current pharmacological interventions modafinil, solriamfetol, and pitolisant are effective to some extent but limited by persistent EDS and some adverse effects. The GC–MS analysis of W. somnifera leaf extract revealed a diverse array of 43 phytoconstituents, with prominent compounds including 2-methoxy-4-vinylphenol, 3-methylbutyl hexanoate, hexadecanoic acid, and 1,2-benzenedicarboxylic acid diethyl ester. These compounds were subjected to an in silico pipeline to evaluate their potential as H3R inhibitors, a target critical for promoting wakefulness and mitigating cataplexy. Molecular docking studies using PyRx and AutoDock Vina identified five top-ranking compounds methotrexate, 1,2-benzenedicarboxylic acid, octadecenoic acid, cAMP, and 9,12-octadecadienoyl chloride with docking scores ranging from −8.8 to −6.2 kcal/mol, indicating strong to moderate binding affinities to H3R. Methotrexate exhibited the highest docking score (−8.8 kcal/mol), attributed to its complex structure facilitating robust hydrogen bonding and electrostatic interactions, while 9,12-octadecadienoyl chloride showed the most favorable MM/GBSA binding free energy (−52.96 kcal/mol), suggesting exceptional binding stability. Physicochemical and drug-likeness analyses via SwissADME and Molsoft confirmed that all five compounds adhere to Lipinski’s Rule of Five, with methotrexate and cAMP demonstrating balanced polarity, bioavailability, and drug-likeness scores (0.29 and 0.45, respectively). BBB permeability predictions highlighted the ability of methotrexate, 1,2-benzenedicarboxylic acid, octadecenoic acid, and 9,12-octadecadienoyl chloride to effectively cross the BBB (scores >4), a crucial property for targeting CNS disorders like narcolepsy. In contrast, cAMP’s lower BBB score (2.01) suggests the need for structural modifications or advanced delivery systems to enhance CNS uptake. ADMET profiling further validated the pharmacokinetic suitability of these compounds, with methotrexate and cAMP showing minimal cytochrome P450 interactions and no significant toxicity risks, unlike 9,12-octadecadienoyl chloride, which was flagged for potential Ames mutagenesis and carcinogenesis. To evaluate specificity, docking scores and interaction profiles were compared between orexin-related receptors and other neuroreceptors, including histamine H3 and dopamine D2 receptors. The identified phytochemicals showed stronger binding affinities and more stable interactions with orexin receptor targets, suggesting a degree of target specificity and reduced likelihood of significant off-target effects.
MD simulations over 50 ns provided comprehensions into the stability of protein ligand complexes. Methotrexate and cAMP exhibited lower root mean square deviation (RMSD, 0.5–1.0 nm), radius of gyration (Rg, 2.7–2.8 nm), and higher hydrogen bond counts (220–250), indicating superior structural stability and consistent receptor interactions. Visualization of binding modes using Schrodinger Maestro revealed that methotrexate and cAMP form extensive hydrogen bonding networks with key H3R residues (e.g., ARG 132, SER 309, HIS 416) and pi-stacking interactions (e.g., with HIS 417), enhancing their binding specificity. In contrast, 9,12-octadecadienoyl chloride and octadecenoic acid relied heavily on hydrophobic interactions, which, while contributing to strong binding energies, were associated with less stable MD trajectories. The integration of docking, MD simulations, MM/GBSA calculations, and ADMET analyses underscores methotrexate, 1,2-benzenedicarboxylic acid, and cAMP as the most promising candidates for further development. Their favorable safety profiles, stable receptor interactions, and drug-like properties position them as potential lead compounds for narcolepsy treatment. However, challenges remain, particularly for cAMP, which requires optimization to improve BBB penetration. Additionally, while 9,12-octadecadienoyl chloride demonstrated exceptional binding affinity, its toxicity concerns necessitate cautious evaluation and potential structural modifications to mitigate risks. Aforementioned, identified compounds exert chronobiotic effects by acting as histamine H3 receptor inhibitors, a mechanism crucial for promoting wakefulness and mitigating cataplexy in narcoleptic individuals.
This study highlights the therapeutic potential of W. somnifera derived phytochemicals as H3R inhibitors, offering a safer, plant-based alternative to synthetic drugs for managing narcolepsy symptoms. The findings lay ground for future preclinical studies to validate these compounds in animal models of narcolepsy, focusing on their efficacy in reducing EDS and cataplexy, as well as their long-term safety. Furthermore, the multidisciplinary approach employed here combining phytochemistry, computational modeling, and pharmacological profiling serves as a robust framework for drug discovery in other neurological disorders.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
