DNA Barcoding, Phytochemical analysis, and Computational Insights into the Neuroprotective Potential of Benincasa hispida

Benincasa hispida, commonly referred to as winter melon or ash gourd, is a traditionally employed medicinal plant with multitude of therapeutic benefits known to be effective for many disorders, such as respiratory disease, neurological disorders, heart diseases, gastrointestinal problems, diabetes mellitus and urinary diseases. The phytopharmacological features of Benincasa hispida is listed in Table 1.

Table 1. Phytopharmacological information of Benincasa hispida

Among the broad spectrum of therapeutic benefits, various parts of the plant, particularly the fruit have been specifically indicated for managing CNS disorders, including cognitive dysfunction, schizophrenia, epilepsy, and depressive disorders. In addition to these traditional claims, plant extracts have been scientifically validated for their neuroprotective effects to some extent. For instance, different solvent-based extracts confirmed anxiolytic activity². In addition to these findings, the fruit extract of Benincasa hispida has demonstrated protective effects in a colchicine-induced Alzheimer’s disease (AD)³ and, the fruit peel extract demonstrated anticonvulsant effect. Notably, treatment with the plant extract has been shown to restore acetylcholine, dopamine, and serotonin levels while reducing amyloid beta aggregation, which are considered key hallmarks of AD pathology⁴. Despite its well-documented ethnopharmacological relevance and demonstrated neuroprotective effects in experimental models, the capacity of its phytoconstituents to interact with critical molecular targets implicated in AD remains incompletely characterised. It has to be noted that, in silico methods are increasingly employed to elucidate potential interactions between disease-associated proteins and drug-like molecules^{5, 6}. In this context, the present study aimed to comprehensively evaluate the neuroprotective potential of phytochemicals derived from Benincasa hispida against AD targets using computational approaches.

Initially, the plant material was authenticated through morphological assessment and DNA barcoding to ensure accurate identification. Phytoconstituents were characterised via literature survey and gas chromatography–mass spectrometry (GC-MS) analysis. A combination of virtual screening, molecular docking, and network pharmacology approaches was subsequently applied to assess the binding affinities of these compounds against selected AD-related targets, including acetylcholinesterase (AChE), β-secretase 1 (BACE1), gamma-aminobutyric acid receptor (GABABR), and glycogen synthase kinase-3β (GSK3β).

2. MATERIALS AND METHODS

2.1 Plant authentication

Accurate and scientific method of authentication is essential for the species identification and further evaluation of medicinal plants for their medicinal properties.

2.1.1 Morphological identification

The plant was collected from the botanical garden of the Department of Botany, University of Kerala, Trivandrum, and was identified and authenticated by the taxonomist of the same department. The leaf specimen of the plant along with the flowers was labelled and preserved on an herbarium sheet, and a corresponding herbarium voucher/reference number was assigned.

2.1.2. DNA Barcoding

After the identification based on the morphological characteristics the plant was further subjected to species-level identification using DNA barcoding. In this process, a short, conserved region of the rbcL gene was sequenced to determine related species. Fresh leaf samples the plant was collected and submitted for sequencing at the Rajiv Gandhi Centre for Biotechnology (RGCB), Trivandrum. Universal forward and reverse primers targeting the rbcL gene, RBCL-AF(ATGTCACCACAAACAGAGACTAAAGC) [12] and RBCL-724R (TCGCATGTACCTGCAGTAGC) [13], respectively were used for amplification. DNA was isolated using the NucleoSpin® Plant II Kit, and the quality of the extracts was assessed by agarose gel electrophoresis. Gels were visualized with a UV transilluminator (Genei), and images were captured using a Gel documentation system (Bio-Rad). Sanger sequencing was performed in a PCR thermal cycler (GeneAmp PCR System 9700, Applied Biosystems) employing the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, USA), based on the manufacturer’s protocol. The quality of the resulting sequences was evaluated using Sequence Scanner Software v1 (Applied Biosystems). Sequence alignment and editing were carried out in Geneious Pro v5.1 to assemble consensus sequences for each plant. The consensus sequence was compared to existing records using the Basic Local Alignment Search Tool (BLAST) [https://blast.ncbi.nlm.nih.gov/Blast.cgi] to identify related species. Later, the sequence was deposited in GenBank [https://www.ncbi.nlm.nih.gov/genbank/].

2.2. Phytochemical analysis

After authentication, the plant’s secondary metabolites were identified through literature review and GC-MS analysis.

2.2.1. Literature analysis

A comprehensive literature search was conducted to compile previously reported phytochemicals from Benincasa hispida. The search strategy used the keywords “plant scientific name + secondary metabolites/phytochemicals/phytocompounds” in the PubMed database. Results were then filtered to include only studies reporting phytochemicals present specifically in the fruit.

2.2.2. GC-MS analysis

The collected samples were thoroughly washed with distilled water and shade-dried at room temperature. The dried material was crushed, and 25 g of the sample was measured and extracted with 250 ml of water using the Soxhlet extraction method) for 24 hours. The resulting extract was filtered through Whatman No. 1 filter paper and concentrated under reduced pressure using a rotary evaporator. Subsequently, the extracts were freeze-dried in a lyophilizer and stored in sterile pre-weighed screw-cap bottles at 4?°C. The dried extract was subjected to GC-MS analysis using a Shimadzu GC-MS system (Model QP2020) equipped with an SH-Rxi-5Sil MS column. The oven temperature was maintained at 60?°C for 8 minutes, and 1.0?µl of each sample was injected for analysis. High-purity helium gas (99.99%) served as the carrier and eluent, with a flow rate of 1 ml/min. The injector temperature was set at 250?°C, and the split ratio was maintained at 10 throughout the run. Ionization was performed at 70 eV, and mass spectra were recorded over a mass range of 10–20 m/z for approximately 8 minutes. As individual compounds eluted from the column, they were introduced into the electron ionization detector, where they were fragmented by electron bombardment into charged ions. The resulting mass-to-charge (m/z) ratios, unique to each compound, were used to generate molecular fingerprints. Identification of compounds was accomplished by comparing the acquired mass spectra with reference data from the National Institute of Standards and Technology (NIST) 17 Library.

2.3. Drug Likeness

The set of phytochemicals identified through both the literature survey and GC-MS analysis was subjected to ’Lipinski’s Rule of Five filtering’ to screen for compounds exhibiting drug-like properties, in terms of oral bioavailability. This rule determines the physicochemical properties of small molecules and assesses their pharmacological potential as orally active drugs in humans⁷. According to Lipinski’s Rule of Five, an orally active drug candidate should meet the following criteria: no more than 5 hydrogen bond donors (HBD), no more than 10 hydrogen bond acceptors (HBA), a molecular weight (MW) of 500 daltons or less, and an octanol–water partition coefficient (AlogP) not exceeding 5.

2.4. Molecular Docking

To evaluate the binding affinity of the identified compounds toward molecular targets associated with CNS dysfunction, four proteins with critical roles in neurological disorders, particularly Alzheimer’s disease were selected for the study. The 3D structures of the selected molecular targets, AChE (PDB Id:6O4W), BACE1 (PDB Id:3UQU), GABABR (PDB Id:4MS3), and GSK3β (PDB Id:4J1R) were retrieved from the Protein Data Bank (PDB), a digital repository containing atomic coordinates of biological macromolecules (https://www.rcsb.org/). Non-mutated X-ray crystallographic structures were downloaded in PDB format. The control drugs, Donepezil (PubChem ID: 3152), LY2811376 (PubChem ID: 44251605), endogenous ligand, GABA (PubChem ID: 119), and drug, Cromolyn (PubChem ID: 2882) were taken as control for AChE, BACE1, GABABR, and GSK3β respectively. The 3D structures of all identified phytochemicals along with the reference drugs were downloaded from the PubChem database in SDF format. PubChem is a publicly accessible resource containing structural data for small molecules https://pubchem.ncbi.nlm.nih.gov/

Prior to docking, all target proteins and ligands were prepared following the recommended protocols of the Discovery Studio (DS) drug docking suite, client version⁸. The ligand’s protonation states and tautomers were optimized using the ‘prepare ligand’ module. The bound ligand and water molecules of the selected targets were removed manually and the binding sites were defined based on PDB site records. Virtual screening of phytochemicals against the targets was performed using the LibDock site-feature docking algorithm in DS. LibDock is a rigid docking method that leverages the physicochemical properties of ligands to identify and dock them into protein binding sites by detecting and matching polar and apolar hot spots within the receptor cavity.

2.5. Network pharmacology

The target-ligand interaction network was generated to understand the interactions of the top- scored phytochemicals of Benincasa hispida against multiple targets of AD. The network was constructed based on the plant-phytochemical-target interaction using Cytoscape (https://cytoscape.org/).

3. RESULTS

3.1. Plant Authentication

3.1.1. Based on Morphological Features

The leaf specimen of Benincasa hispida identified based on the morphological features was preserved in an herbarium sheet with voucher number ‘KUBH11054’ (Fig. 1).

Fig 1. Herbarium specimen of leaf of Benincasa hispida

3.1.2. DNA Barcoding

The annotated genomic sequence of Benincasa hispida, with a length of 1120 base pairs, has been submitted to GenBank with the accession number ‘ON032715.1’. The identified sequence is provided below in FASTA format.

>ON032715.1 Benincasa hispida ribulose-1,5-bisphosphate carboxylase/oxygenase large subunit (rbcL) gene, partial cds; chloroplast

TTATTATACTCCTGAATATGAAACCAAAGATACTGATATCTTGGCAGCATTCCGAGTAACTCCTCAACCGGGAGTTCCACCTGAGGAAGCAGGGGCCGCTGTAGCTGCTGAATCTTCTACTGGTACATGGACAACTGTGTGGACCGATGGGCTTACCAGTCTTGATCGTTACAAAGGACGATGCTATGGAATCGAGCCTGTTCCTGGAGAAGAAAATCAATATATTGCTTATGTAGCTTATCCCCTAGACCTTTTTGAAGAAGGTTCTGTTACTAACATGTTTACTTCCATTGTCGGTAATGTATTTGGATTCAAGGCTCTACGTGCTCTACGTCTGGAGGATTTGCGAATCCCTACTGCTTATATTAAAACTTTCCAAGGCCCGCCTCATGGTATCCAGGTTGAAAGAGATAAATTGAACAAGTATGGTCGCCCTCTATTGGGATGTACTATTAAACCAAAATTGGGATTATCCGCTAAGAATTATGGTAGAGCAGTTTATGAATGTCTACGCGGTGGACTTGATTTTACCAAAGATGATGAAAACGTGAATTCCCAACCATTTATGCGTTGGAGAGACCGTTTCCTATTTTGTGCGGAAGCTATTTATAAATCACAGGCTGAAACAGGTGAAATCAAGGGACATTACTTGAATGCTACTGCAGGTACATGCGAAGAAATGATCAAAAGGGCTGTATTTGCCCGAGAATTGGGAGTTCCTATCGTAATGCATGACTACTTAACAGGTGGATTCACTGCANATACTAGCTTGGCTCATTATTGCCGAGATAATGGTCTACTTCTTCACATTCACCGTGCAATGCATGCCGTTATTGATAGACAGAAGAATCATGGTATGCACTTCCGTGTACTAGCTAAAGCGTTACGTATGTCTGGTGGAGACCATATTCACGCTGGTACCGTAGTAGGTAAACTTGAAGGGGAAAGAGAAATCACTTTAGGCTTTGTTGATTTACTACGTGATGATTTTATTGAAAAAGACCGAAGCCGCGGTATTTATTTCACTCAAGATTGGGTCTCTTTACCAGGTGTTCTACCAGTGGCTTCCGGTGGTATTCACGTTTGGCATATGCCTGCTCTAACCGAGATTTTTGGAGAT

3.2. Phytochemical analysis

3.2.1. Literature survey

From the analysis of PubMed data, 78 bioactive compounds were identified for different solvent-based extracts for Benincasa hispida. The compounds with their references are listed in Supplementary Table. S1.

3.2.2. GC-MS analysis

GC?MS analysis of the water extract of Benincasa hispida fruit revealed the presence of 26 compounds. GC?MS chromatograms of these compounds are depicted in Fig. 2 and the list of identified compounds with their name, retention time (RT), percentage of area, are presented in Table 2.

Fig 2. GC-MS Chromatogram of water extract of Benincasa hispida. The compounds detected were represented as peaks. The X-axis represents retention Time (RT) and Y-axis represents the relative abundance of the compounds.

Table 2. Phytochemical Compounds Identified in Benincasa hispida Extract via GC-MS

3.3 Drug likeness

The drug-likeness probability of 79 compounds, obtained after removing duplicates from the initial list of 104 identified small molecules (78 from the literature and 26 from GC-MS analysis) was evaluated with respect to oral bioavailability. The molecular descriptors used for Lipinski’s filtering criteria for these compounds are summarized in Table 3.

Table 3. Phytochemicals of Benincasa Hispida that passed Lipinski filtering

The result showed that, out of 79 screened compounds, 74 passed the filtering criteria to be act as orally bioavailable drug. All these compounds which passed the criteria was further taken for evaluating their binding affinity towards selected molecular targets, AChE, BACE1, GABABR, and GSK3β.

3.4. Molecular Docking

Based on the docking results, 15, 36, and 22 compounds demonstrated higher binding affinity towards AChE, BACE1, and GABABR, respectively, compared to their corresponding reference molecules. In contrast, none of the compounds showed a higher docking score against GSK3β than the reference drug, cromolyn. Several compounds were found to interact with multiple targets, exhibiting greater binding affinity than the respective control molecules. The docking scores of the selected phytochemicals are represented in supplementary Table 2. It was observed that the compound, ‘9-Octadecenoic acid’ (PubChem ID: 21160048) demonstrated highest docking score against three targets, AChE (133.87), BACE1 (159.36), and GABABR (154.74) than the reference drugs. For GSK3β, the compound naringenin (PubChem ID: 439246) showed highest docking score (88.57). The interactions of top scored phytochemicals which showed highest affinity towards the selected targets are depicted in Fig. 3 and the binding mode of the ligands in the binding cleft of target proteins are shown in Fig. 4

Fig 3. 2D representation of the molecular interactions between top scored phytochemicals of Benincasa hispida against selected targets of Alzheimer’s disease. A: AChE-21160048; B: AChE-Donepezil (Control); C: BACE1-21160048; D: BACE1-LY2811376 (Control); E: GABABR-21160048; F: GABABR-GABA (Control); G: GSK3β-439246; H: GSK3β-romolyn (Control). Different colours are used to represent the types of interactions as detailed below.

Fig 4. Binding mode of top-scoring phytochemicals of Benincasa hispida within the selected targets of Alzheimer’s disease. A: AChE-21160048; B: AChE-Donepezil (Control); C: BACE1-21160048; D: BACE1-LY2811376 (Control); E: GABABR-21160048; F: GABABR-GABA (Control); G: GSK3β-439246; H: GSK3β-Cromolyn (Control). Targets are shown in ribbon and ligands are shown in ball and stick representation

The binding mode of the compound within the respective targets demonstrates that it adopts an orientation similar to that of the reference molecule, suggesting a potential regulatory effect.

3.5. Network pharmacology analysis

The target-ligand interaction network was generated to understand the interactions of the phytochemicals of Benincasa hispida against selected targets of AD. From the docking results, the top-scoring phytocompounds against each target than the reference drugs were included for the network analysis to demonstrate the interaction with multiple targets (Fig 5).

Fig 5. Target- phytochemical interaction network showing docking scores of phytochemicals of Benincasa hispida against selected molecular targets of Alzheimer’s disease. The network diagram illustrates the binding affinities of selected phytochemicals that demonstrated higher docking score than the reference drugs, against three key molecular targets AChE (604W), GABABR (4MS3), and BACE1 (3UQU). Nodes represends targets and phytochemicals, where phytochemicals are represented with their PubChem IDs. Edges represent compound–target interactions, with labels indicating docking scores. Higher scores suggest stronger binding affinity.