Abstract

   This study explores the potential of Ailanthus excelsa Roxb. for
   managing obesity by evaluating its effects on key metabolic enzymes. We
   evaluated a hydroalcoholic extract and its fractions for their ability
   to modulate important metabolic enzymes, porcine pancreatic lipase,
   HMG-CoA reductase, α-glucosidase, and α-amylase activities. Our
   methodology integrated in vitro enzymatic assays with cluster analysis
   (MCODE, ClueGO, and Cluepedia) and network pharmacology to elucidate
   interactions between metabolites and target enzymes. Protein-protein
   interaction (PPI) analysis, consisting of a network with 51 nodes and
   264 edges, was performed using the CytoNCA plugin to calculate
   topological parameters. Key targets were identified based on degree
   centrality, including ADIPOQ, PPARA, PPARG, IL6, TNF, and AKT1. Cluster
   analysis of the PPI networks, conducted using the MCODE plugin,
   highlighted a top cluster with a high score of 22.82. Network
   pharmacology has identified key targets associated with obesity,
   including HK1, HK2, PIK3CA, AKT1, MTOR, CD36, ACACB, SLC2A4, CPTIA,
   INSR, ACACA, FASN, and ADIPOQ. These targets are linked to highly
   modulated metabolic pathways. Isoquercetin shows significant binding
   affinities: − 7.11 for HMG-CoA Reductase (PDB ID: 1HW9), -9.96 for
   lipase (PDB ID: 1LPB), − 8.96 for α-amylase, and − 10.41 for
   α-Glucosidase (PDB ID: 3A47). The ethyl acetate fractions exhibit
   notable inhibition of Porcine Pancreatic lipase (IC[50]:
   56.25 ± 4.85 µg/mL) and HMG-CoA reductase (IC[50]:
   108.27 ± 3.38 µg/mL), α-glucosidase (IC[50]: 117.08 ± 3.28 µg/mL), and
   α-amylase (IC[50]: 125.93 ± 2.29 µg/mL). Moreover, metabolites showed
   stronger binding affinities to all four enzymes than reference
   molecules. By integrating in vitro assays with molecular modeling, this
   study highlights the promising potential of A. excelsa and its
   fractions in obesity management, offering valuable insights into its
   therapeutic applications.

Supplementary Information

   The online version contains supplementary material available at
   10.1038/s41598-025-14420-2.

   Keywords: Ailanthus excelsa Roxb, Cluster analysis, Molecular dynamic
   simulation, α-glucosidase

   Subject terms: Computational biology and bioinformatics, Drug discovery

Introduction

   Obesity, a global health crisis, has become one of the most pressing
   public health challenges of the 21st century^[40]1. Defined by
   excessive fat accumulation, obesity significantly increases the risk of
   metabolic diseases such as hypertension, type 2 diabetes, and
   dyslipidaemia. The prevalence of obesity has surged worldwide,
   affecting both adults and children. According to the World Health
   Organization (WHO), over 650 million adults were obese in 2016, a
   staggering increase since 1975^[41]2. This dramatic rise is primarily
   driven by urbanization, sedentary lifestyles, and unhealthy diets,
   coupled with an aging population^[42]3. Obesity’s role as a precursor
   to type 2 diabetes mellitus (T2DM) is particularly concerning. T2DM,
   which accounts for most diabetes cases, often stems from insulin
   resistance, where changes in receptor structure reduce hormone-receptor
   affinity or impair insulin secretion. In contrast, type 1 diabetes
   mellitus arises from a complete disruption in insulin secretion^[43]4.
   The International Diabetes Federation (IDF) reported that 537 million
   adults were living with diabetes in 2021, with 6.7 million deaths
   attributed to the disease. The burden of diabetes is expected to
   escalate further, with projections indicating that 643 million people
   will be affected by 2030 and 783 million by 2045, further compounding
   the global health crisis^[44]5.

   Given the strong connection between obesity and T2DM, effective
   strategies to manage both conditions are crucial^[45]6. One promising
   approach involves inhibiting enzymes responsible for carbohydrate
   digestion, such as α-amylase and α-glucosidase^[46]7. These enzymes
   hydrolyze α-1,4-glycosidic bonds in carbohydrates, such as starches,
   producing absorbable monosaccharides. Inhibiting α-amylase and
   α-glucosidase slows carbohydrate digestion, thereby reducing
   postprandial glucose levels and contributing to the management of both
   diabetes and obesity^[47]8. Furthermore, inhibiting lipase and HMG-CoA
   reductase enzymes play a key role in regulating lipid metabolism,
   offering additional therapeutic benefits in managing obesity and its
   related complications.

   In addition to these enzymatic pathways, oxidative stress is a
   significant complication associated with diabetes and obesity.
   Oxidative stress arises from an imbalance between reactive oxygen
   species (ROS) production and the body’s antioxidant defence
   mechanisms^[48]9. This imbalance leads to cellular damage and is
   implicated in the progression of chronic diseases, including diabetes
   and cardiovascular disease^[49]10. While conventional pharmaceutical
   treatments for diabetes and obesity are available, they often come with
   limitations, such as side effects and long-term safety concerns. Oral
   drugs, for instance, may be useful for individuals who are allergic to
   insulin or prefer not to use insulin injections^[50]11. However,
   prolonged use of these medications can increase the risk of heart
   attacks, acute kidney injury, and liver toxicity^[51]12. Consequently,
   there is a growing interest in developing traditional medicines that
   offer a safer and more sustainable alternative for managing these
   conditions^[52]13.

   Traditional medicines, derived from natural sources, have long been
   used to treat various ailments and are considered a valuable reservoir
   of bioactive compounds with therapeutic potential. In this context, A.
   excelsa, a versatile medicinal plant, has garnered attention for its
   potential anti-obesity, anti-diabetic, and antioxidant properties.
   Commonly known as the “tree of heaven,” A. excelsa belongs to the
   Simaroubaceae family and is native to Asia, Africa, and
   Australia^[53]14. Historically, different parts of the A. excelsa tree,
   including its bark, leaves, and roots, have been utilized in
   traditional herbal medicine for their purported medicinal
   properties^[54]15. These include anti-asthmatic, hepatoprotective, and
   hypoglycaemic effects. However, the potential impact of A. excelsa
   extract on obesity and its mechanisms of action remains largely
   unexplored^[55]16.

   This study explores the potential of A. excelsa as a therapeutic agent
   for managing obesity and related metabolic disorders. By investigating
   the Pharmacognostical properties and metabolite profiles of its
   hydroalcoholic extract, the research evaluates its anti-obesity,
   anti-diabetic, and antioxidant activities through in vitro assays. The
   integration of computational techniques, such as network pharmacology
   and molecular modeling, provides insights into the molecular
   interactions of the plant’s bioactive compounds. This study aims to
   contribute to traditional medicine by highlighting the therapeutic
   potential of A. excelsa and its relevance in developing safe,
   plant-based treatments for chronic diseases like diabetes and obesity.
   Specifically, this study addresses the limited exploration of Ailanthus
   excelsa as a potential therapeutic agent for obesity and metabolic
   disorders. Although traditional uses of A. excelsa have been
   documented, its mechanisms of action, particularly in targeting key
   metabolic pathways such as PPAR and PI3K-Akt signaling, remain
   understudied. The integration of network pharmacology, molecular
   docking, and in vitro assays offers a comprehensive approach to
   elucidate these mechanisms. By focusing on lipid modulation, enzyme
   inhibition, and antioxidant activities, this research provides new
   insights into the plant’s bioactive compounds, addressing gaps in both
   Pharmacognostical profiling and therapeutic evaluation.

Methods

Collection of plant material

   The plant material of Ailanthus excelsa was collected in June from the
   Hidakal Dam region, Belagavi. This species is not listed under the
   Convention on International Trade in Endangered Species of Wild Fauna
   and Flora (CITES) or any other endangered species lists. The collection
   was conducted in accordance with the Plant Protection Act (PPA) and the
   International Plant Protection Convention (IPPC) guidelines, ensuring
   that no harm was caused to any endangered flora. Sustainable harvesting
   practices were strictly followed throughout the process. The plant was
   identified and authenticated by a qualified taxonomist at the Indian
   Council of Medical Research (ICMR) – National Institute of Traditional
   Medicine (NITM), and assigned the authentication reference number
   RMRC-1715. A voucher specimen has been deposited in the herbarium of
   ICMR-NITM for future reference. Following collection, the plant parts
   were manually separated and air-dried under shade on a muslin cloth for
   10–15 days. After complete drying, the material was ground into a
   coarse powder for further processing.

Standardization parameters and physicochemical evaluation

   By the guidelines of the Indian Pharmacopeia and Herbal Pharmacopeia,
   physicochemical characteristics such as acid-insoluble ash, total ash,
   water-soluble ash, and extractive values were investigated. They were
   assessing the ash content aids in identifying low-quality products,
   exhausted drugs, and the presence of excessive sand or other earthy
   materials. Microscopic characteristics of powdered plant substances
   were observed in scanning electron microscopy. This method is essential
   for quality control, identification of species, and detection of
   contaminants in herbal medicines and spices.

Preparation of plant extract and fractions

   The crude drug was thoroughly washed, cut into small pieces, and
   shade-dried before being coarsely powdered. The coarse powder was then
   extracted using a hydroalcoholic solvent (70:30) through the maceration
   technique. After extraction, the solvent was evaporated under reduced
   pressure using a rotary vacuum evaporator, and the resulting extract
   was stored in an airtight container for further use. A portion of the
   extract was suspended in distilled water and subjected to successive
   liquid-liquid partitioning with n-hexane, ethyl acetate, chloroform,
   and petroleum ether to obtain different fractions^[56]17.

Preliminary phytochemical screening

   Preliminary phytochemical screening helps identify the secondary
   metabolites present in plants carried out by previously reported
   methods^[57]18.

Quantitative analysis (Hydro alcohol, Ethyl acetate and hexane)

Determination of total phenol content (TPC)

   The total phenol content in the plant extracts was determined using the
   Folin-Ciocalteu method, with gallic acid as the standard, incorporating
   slight modifications. The reaction mixture consisted of 1 ml of the
   extract and 9 ml of distilled water. Subsequently, 1 ml of
   Folin-Ciocalteu phenol reagent was added, and the mixture was shaken
   well. After 5 min, 10 ml of a 7% sodium carbonate solution was added.
   Standard solutions of gallic acid were prepared at concentrations of
   20, 40, 60, 80, and 100 µg/ml. The mixture was incubated for 90 min at
   room temperature, and the absorbance of the test and standard solutions
   was measured against a reagent blank at 550 nm using an ultraviolet
   (UV)/visible spectrophotometer. The total phenol content was expressed
   as mg of gallic acid equivalents (GAE) per gram of extract^[58]19.

Determination of total flavonoid content (TFC)

   According to Martínez S et al.^[59]20, the total flavonoid content in
   the plant extract was determined using the aluminium chloride method.
   The reaction mixture consisted of 0.25 ml of plant extract in 1.25 ml
   of distilled water, followed by the addition of NaNO[3]. The mixture
   was placed in the dark for 6 min, after which 0.15 ml of 10% AlCl[3]
   was added and incubated for an additional 15 min. Subsequently, 0.5 ml
   of NaOH and 0.275 ml of distilled water were added to the mixture.
   Standard solutions of quercetin were prepared at concentrations of 20,
   40, 60, 80, and 100 µg/ml. The absorbance of the standard and extract
   solutions was measured against a reagent blank at 510 nm using a
   UV/Visible spectrophotometer. The total flavonoid content was
   determined from the calibration curve and expressed as milligrams of
   quercetin equivalent (QE) per gram of extract.

Determination of total alkaloid content (TAC)

   The plant extract (1 mg) was dissolved in dimethyl sulfoxide (DMSO) and
   then acidified with 1 mL of 2 N hydrochloric acid (HCl). The resulting
   solution was filtered and transferred to a separating funnel.
   Subsequently, 5 mL of bromocresol green solution and 5 mL of phosphate
   buffer were added. The mixture was subjected to vigorous shaking with
   chloroform, and the organic phase was collected in a 10-mL volumetric
   flask. The volume was then adjusted to the mark with chloroform. For
   calibration, a series of reference standard solutions of atropine (20,
   40, 60, 80, and 100 µg/mL) were prepared using the same procedure. The
   absorbance of both test and standard solutions was measured at 470 nm
   using a UV/Visible spectrophotometer, with the reagent blank as the
   reference. The total alkaloid content was expressed as mg of atropine
   equivalent (AE) per gram of extract^[60]21.

Biological evaluation

Determination of antioxidant activity and free radical scavenging activities

   The radical scavenging activity of the extracts was studied using DPPH
   free radical assay, ferrous ion chelating, nitric oxide scavenging
   activity, ABTS radical scavenging activity, and total antioxidant
   activity by phosphomolybdenum method were determined according to
   established methods^[61]22,[62]23.

Determination of Porcine pancreatic lipase Inhibition assay

   Porcine pancreatic Lipase Inhibition activity was carried out by
   Andersone et al.^[63]24 method with slight modification. A stock
   solution of A. excelsa extract and standard were prepared. Serial
   dilution of extract and standard were prepared in the range of
   (20–100 µg/ml). Collect different test tubes and label them as samples
   (hydro alcohol extract and n-hexane, petroleum ether, chloroform, and
   ethyl acetate fractions). In the sample tube, add 1 ml extract and
   0.5 ml lipase solution. Incubate for 30 min. Then, add the PNPB
   substrate, and continue incubation for 30 min. After that, absorbance
   was recorded at 410 nm. Orlistat was used as a standard and formed in
   the same manner as the sample.

α-amylase Inhibition assay

   The study aimed to evaluate the inhibitory potency of A. excelsa
   extract against the α-amylase enzyme. The α-amylase inhibition
   properties of samples were performed as described in Tian et
   al.^[64]25. α-amylase (3 units/mL) was dissolved in 0.1 M buffered
   phosphate-buffered saline, pH 6.9. Samples at various concentrations
   (0–100 µg/mL) were pre-incubated with the enzyme for 10 min at 37 °C.
   The reaction commenced upon adding substrate (0.1% starch) to the
   incubation medium. After 10 min, the reaction was halted by adding 250
   µL of dinitro salicylic (DNS) reagent (comprising 1% 3,5-dinitro
   salicylic acid, 0.2% phenol, 0.05% Na[2]SO[3], and 1% NaOH in aqueous
   solution), followed by boiling the reaction mixture for 10 min in a
   water bath. Subsequently, 250 µL of 40% potassium sodium tartrate
   solution was added. After cooling to room temperature in a cold-water
   bath, the absorbance was measured at 540 nm.

α-Glucosidase inhibition assay

   The α-glucosidase inhibitory activities of samples were evaluated using
   the method and Kifle ZD et al.^[65]26. α-glucosidase was dissolved in
   phosphate buffer (50 mM, pH 6.9) and treated with different
   concentrations of samples (0–100 µg/mL) individually for 10 min at
   37 °C. The reaction commenced with adding 50 µL of 5 mM
   p-nitrophenyl-α-D glucopyranoside in phosphate buffer. The enzyme
   reaction proceeded at 37 °C for 30 min. The reaction was terminated by
   adding Na[2]CO[3] (1 M), followed by measuring the absorbance at
   405 nm.

HMG-CoA reductase inhibition assay

   The inhibitory effect of various samples on HMG-CoA reductase was
   evaluated using a modified method from Baskaran et al.^[66]27. Each
   sample was tested at concentrations ranging from 25 to 125 µg/mL, mixed
   with a reaction solution containing nicotinamide adenine dinucleotide
   phosphate (400 µM), HMG-CoA substrate (400 µM), and potassium phosphate
   buffer (100 mM, pH 7.4) with potassium chloride (120 mM), ethylene
   diamine tetra acetic acid (1 mM), and dithiothreitol (5 mM). HMG-CoA
   reductase (2 µL) was added, and the mixture was incubated at 37 °C.
   Absorbance was measured at 340 nm after 10 min. Simvastatin
   (Sigma-Aldrich Co.) was a positive control, and distilled water was
   used as a negative control.

   Using a positive control, the percentage of inhibition was determined
   using the following formula for Porcine pancreatic Lipase, α-amylase,
   α-glucosidase, and HMG-CoA reductase inhibition assay: (OD of blank-OD
   of test/OD of Blank) * 100. The results were represented in IC[50]
   values.

   All in vitro experiments were performed in triplicate (n = 3) to ensure
   reproducibility and reliability of results. Data were expressed as
   mean ± standard deviation (SD). The IC[50] values were calculated using
   non-linear regression analysis.

Computer-aided studies

Identification of active components of Ailanthus excelsa and target retrieval

   The biologically active phytochemicals present in Ailanthus excelsa
   were identified through a comprehensive literature review and public
   repositories, including Dr. Duke’s Phytochemical and Ethnobotanical
   Database (Dr. Duke’s DB), IMPPAT database and PubChem
   ([67]https://pubchem.ncbi.nlm.nih.gov/) database. The potential
   molecular targets of these phytoconstituents were predicted using
   DIGEP-pred, SwissTargetPrediction, and SuperPred databases.
   Additionally, obesity-related targets were retrieved from multiple
   databases, including DisGeNET
   ([68]https://www.disgenet.org/browser/0/1/0/C0028754/), the Therapeutic
   Target Database ([69]https://db.idrblab.net/ttd/), and Gene Cards
   ([70]https://www.genecards.org/).

Ailanthus excelsa - obesity common target network construction

   A venny 2.0 intersect the common targets between A.excelsa-Obesity.
   After that, import the common targets into Cytoscape to build the A.
   Excelsa-Obesity shared target network. Analysis of the main targets and
   bioactive components of A. excelsa against Obesity according to the
   degree.

Protein-protein interaction

   Constructing protein-protein interaction (PPI) networks for common
   targets between the compounds and disease targets were visualized using
   the STRING database. PPI data were subsequently exported to Cytoscape
   for further analysis of centrality measures. In these networks, nodes
   represented targets, while edges depicted interactions between the
   targets. The platform assessed the mutual information score for each
   protein, where a higher score indicated stronger interactions between
   targets. Isolated targets were excluded from the study, and the PPI
   network was constructed with a confidence score threshold of 0.04.
   Subsequently, key topological properties were analyzed, including
   degree centrality (DC), betweenness centrality (BC), closeness
   centrality (CC), eigenvector centrality (EC), information centrality
   (IC), local average connectivity-based method (LAC), network centrality
   (NC), and subgraph centrality (SC) using the CytoNCA plugin in
   Cytoscape^[71]28. These metrics were employed to identify core targets
   within the network. Finally, cluster analysis was performed using the
   MCODE clustering module in Cytoscape 3.10.2, with the clustering
   coefficient set to identify three distinct clusters.

GO and KEGG pathway enrichment analyses of PPI network and cluster for
obesity using clue GO and Clue Pedia plugins

   Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG)
   pathway enrichment analyses were performed using the ClueGO and
   CluePedia plugins within the Cytoscape software environment. GO terms
   related to Biological Process (BP), Molecular Function (MF), and
   Cellular Component (CC) categories were selected for functional
   annotation (GO database: [72]www.geneontology.org; KEGG:
   [73]www.kegg.jp/kegg/kegg1.html). ClueGO
   ([74]https://apps.cytoscape.org/apps/cluego) was used to visualize and
   functionally group enriched GO terms and KEGG pathways, facilitating
   the interpretation of complex gene-function relationships. To enhance
   the analysis, CluePedia ([75]https://apps.cytoscape.org/apps/cluepedia)
   was integrated with ClueGO, enabling deeper exploration of gene–pathway
   interactions and identification of potential novel biomarkers. Further
   SRplot: A free online platform was used for data visualization and
   graphing ([76]https://www.bioinformatics.com.cn/en). This combined
   approach provided a systems-level understanding of the functional
   associations and biological significance of the gene
   dataset^[77]29,[78]30.35.

Network construction and analysis

   Cytoscape 3.10.2 created a network connecting the phytocompounds of A.
   excelsa with their respective targets. This network visualized the
   interactions between the active constituents and target genes. Using
   the “Network Analyzer” tool as a direct command, edge count was
   prioritized from high too low for modulation. Additionally, network
   analysis was conducted based on “Node degree distribution” and
   “Betweenness by degree,” considering parameters such as eccentricity,
   neighbourhood connectivity, in-degree distribution, and outdegree
   distribution. In this network construction, nodes represented the
   components, targets, and pathways, while edges depicted the
   interactions among these factors. This representation elucidated the
   relationship between the active ingredients and disease targets. The
   degree value of interactions was used to assess the importance of nodes
   in each network^[79]31.

Molecular docking

   To investigate the binding interaction and orientation of 31
   phytoconstituents as inhibitors in the α-amylase, α-glucosidase,
   porcine pancreatic Lipase and HMG-CoA pocket, molecular docking study
   was taken into consideration using Schrodinger’s glide extra precision
   mode (XP)^[80]32 Using Maestro, all of the identified ligand structures
   were represented^[81]33. Ligprep in Schrödinger’s suite was used for
   ligand construction, employing OPLS4 for energy minimization. Torsional
   flexibility was applied to enhance pose complementarity. Crystal
   structure coordinates of the α-amylase (PDB ID: 4W93), α-glucosidase
   (PDB ID: 3TOP), Porcine pancreatic Lipase (PDB ID: 2OXE) and HMG-CoA
   (PDB ID: 1HW9) were retrieved from a protein data bank
   ([82]https://www.rcsb.org). The proteins underwent pre-processing,
   which involved deleting ligands and water molecules from the PDB
   dataset, assigning bond orders, and generating PKa values at PH 7 ± 2.
   The protein was further refined by energy reduction using an OPLS4
   force field to achieve a stable structure for more analysis. A. A
   co-crystal ligand was used to construct a grid to identify the binding
   location inside the target receptor. The binding affinity of ligands
   with receptors was described regarding docking score, hydrogen bonds,
   and pi-pi interactions using Glide software in extra precision (XP)
   mode.

Molecular dynamics (MD) simulations

   Molecular dynamics (MD) simulations were conducted using the Desmond
   Molecular Dynamics System (version) for 100 ns. The receptor-ligand
   complexes (Isoquercetin-HMG-CoA and Isoquercetin-Lipase) were solvated
   in a cubic TIP3P water box with 10 Å side lengths. System preparation
   included neutralizing the net charge and achieving an ionic strength of
   0.15 M with salt addition to maintain electroneutrality. Simulations
   employed the Smooth Particle Mesh Ewald (SPME) algorithm for accurate
   short-range non-bonded interactions with a 9 Å cutoff and efficiently
   handled long-range electrostatics. Initial equilibration simulations of
   10 ns validated system suitability, followed by a multi-stage
   relaxation protocol in an NPT ensemble to prepare equilibrated
   structures. The protocol included restrained minimization (two rounds
   of 2000 steps/round) with 50 kcal/mol/Ų harmonic restraints on solute
   atoms, followed by four MD stages: (1) 12 ps of NVT ensemble at 10 K
   with constant volume, (2) 12 ps of NPT ensemble at 10 K, (3) 24 ps of
   NVT ensemble with temperature ramped to 300 K, and (4) 24 ps of NPT
   ensemble at 300 K with unrestrained MD. Subsequently, two phases of
   relaxation were performed: a brief 1.2 ns phase and a longer
   unrestrained 10 ns phase, with a Nose-Hoover thermostat maintaining
   310 K temperature and pressure controlled isotropically using a
   Martyna-Tobias-Klein barostat. Periodic boundary conditions and the
   multistep RESPA integrator were consistently applied throughout. The MD
   trajectories were analyzed using the Desmond software suite to assess
   system stability and molecular interactions^[83]5,[84]34.

Results and discussion

Pharmacognostical evaluation

Standardization parameters

   By evaluating the ash levels, it is easy to determine where the
   powdered material was contaminated with sand and other inorganic
   material. The quantity of inorganic material contained in the crude
   medication may be determined using the water-soluble ash (3.31 ± 0.25).
   In contrast, the amount of sand and other debris can be determined
   using the in acid-insoluble ash (2.05 ± 0.33) and sulphated ash was
   found to be 2.1 ± 0.47. The varying extraction values using various
   solvents have been identified. Alcohol has the highest extractive value
   (6.44 ± 0.42), followed by water (3.24 ± 0.53), Hexane (3.54 ± 0.12),
   ethyl acetate (2.88 ± 0.32), petroleum ether (2.32 ± 0.95) and
   chloroform (2.04 ± 0.83). The extractive values play a role in
   determining the best solvent to utilize for extracting the most active
   principle (Table [85]S1).

Powder microscopy

   Powder microscopy of A. excelsa revealed distinct cellular components,
   including sclereids, fibers, cork cells, calcium oxalate crystals,
   medullary rays, and xylem cells. They were scanning electron microscopy
   images (Fig. [86]1). Showed fibers grouped, characterized by their
   relatively large size, thin walls, and numerous pits. Sclereids were
   identified as lignified, thick-walled, rectangular cells in clusters.
   Thin-walled, colorless cork cells were also observed. The sample
   contained minute, needle-shaped acicular raphides and idioblasts of
   calcium oxalate crystals. Medullary rays varied from biseriate to
   tetraseriate, with some being uniseriate. Xylem vessels were abundant,
   exhibiting bordered pitting, and often associated with other xylem
   elements. Starch grains were primarily simple and oval, with both
   single and compound forms present.

Fig. 1.

   [87]Fig. 1
   [88]Open in a new tab

   Scanning electron micrographs of Ailanthus excelsa bark powder.

Preliminary phytochemical screening

   The preliminary phytochemical screening of A. excelsa extract revealed
   the presence of various secondary metabolites, including alkaloids,
   phenols, tannins, steroids, triterpenoids, flavonoids, and glycosides,
   as detailed in Table S2.

Total phenol, flavonoid, and tannin content estimations

   In a comparative study of sample and fractions, it was found that Ethyl
   acetate fraction exhibited the highest concentrations of polyphenols,
   flavonoids and alkaloids among the tested samples. Specifically, Ethyl
   acetate fraction revealed the highest content of total polyphenols
   (82.39 ± 2.13 mg GAE/g), flavonoids (95.24 ± 3.21 mg QE/g), and
   alkaloids (185.31 ± 1.42 mg AT/g) (Table [89]1).

Table 1.

   Results of total phenol, flavonoid, and tannin contents.
                  Hydro alcohol Ethyl acetate** N- Hexane**
   TAC (mg AT/g)  148.86 ± 3.96  185.31 ± 1.42  156.96 ± 2.95
   TPC (mg GAE/g) 67.97 ± 4.97   82.39 ± 2.13   54.97 ± 5.17
   TFC (mg QE/g)  79.43 ± 2.31   95.24 ± 3.21   65.98 ± 3.92
   [90]Open in a new tab

   Showing mean ± S.D (n = 3), ** Fractions.

In-vitro studies

In-vitro antioxidant activity

   The assessment of antioxidant activity in the hydroalcoholic extract
   and fractions of A. excelsa involves various tests. The efficacy of
   plant extracts in combating free radicals is contingent upon their
   composition and the testing conditions. Antioxidant capacities are
   subject to numerous influencing factors, necessitating the use of
   multiple assays to comprehensively assess the diverse mechanisms of
   antioxidant action. The ethyl acetate fraction and hydroalcoholic
   extract exhibited notable free radical scavenging activity. Both ethyl
   acetate fraction and standard substances demonstrated increasing
   inhibition values with concentration. Table [91]2. provides the IC[50]
   values for Ailanthus excelsa in various antioxidant tests.

Table 2.

   The IC[50] values for the antioxidant properties of the Ailanthus
   excelsa extract and fractions were determined using various methods.
   Samples DPPH (µg/ml) Phosphomolybdenum (µg/ml) Ferrous Ion (µg/ml) NO
   Scavenging (µg/ml) ABTS (µg/ml)
   AE Extract 43.90 ± 3.3 85.59 ± 2.4 94.51 ± 6.6 95.86 ± 2.5 73.96 ± 2.7
   Ethyl acetate** 38.38 ± 2.92 73.30 ± 4.5 72.87 ± 2.4 76.97 ± 3.3
   59.98 ± 4.2
   Chloroform** 156.42 ± 5.8 109.27 ± 3.6 183.01 ± 4.7 121.97 ± 3.9
   107.35 ± 3.8
   Petroleum Ether** 68.58 ± 1.6 106.80 ± 5.8 147.60 ± 2.4 128.2 ± 4.4
   113.86 ± 2.1
   N-hexane** 41.64 ± 3.5 84.51 ± 1.3 82.79 ± 1.1 116.97 ± 2.8
   96.97 ± 3.02
   Std (5–25 µg/ml) 13.93 ± 0.7 (AA)

   29.35 ± 1.4

   (EDTA)

   62.85 ± 2.32

   (AA)

   64.98 ± 1.2

   (AA)
   25.84 ± 2.3 (BHA)
   [92]Open in a new tab

   Showing mean ± S.D (n = 3), ** Fractions.

Porcine pancreatic lipase Inhibition assay

   The inhibitory effects on porcaine pancreatic lipase were assessed
   using PNPB as the substrate, with different concentrations of both
   standard and A. excelsa extract and fractions being evaluated. The
   hydroalcoholic extract of A. excelsa exhibited significant inhibition
   of pancreatic lipase with an IC[50] value of 97.92 ± 2.37 µg/mL, while
   the ethyl acetate fraction showed superior inhibitory activity with an
   IC[50] value of 56.25 ± 4.85 µg/mL. The n-hexane fraction demonstrated
   moderate activity (IC[50]: 106.84 ± 3.96 µg/mL), but chloroform and
   petroleum ether fractions showed no significant inhibition (NA). The
   inhibition profile of ethyl acetate was notable, as opposed to the
   standard orlistat (IC[50]: 49.86 ± 4.37 µg/mL). This suggests that
   specific components within the ethyl acetate fraction may contribute to
   enhanced lipase inhibition. Figure [93]2a bargraph reprrsents the
   IC[50]value of Porcine Pancreatic Lipase Inhibition.

Fig. 2.

   [94]Fig. 2
   [95]Open in a new tab

   IC[50] values for enzyme inhibition. (a) Porcine pancreatic lipase
   inhibition (b) HMG-CoA reductase (c) α-amylase inhibition (d)
   α-glucosidase inhibition.

HMG-CoA reductase inhibitory activity

   HMG-CoA reductase inhibitory activity was done by standard protocol,
   findings revealed Inhibition of HMG-CoA reductase, a critical enzyme in
   cholesterol biosynthesis, was most effective with the ethyl acetate
   fraction (IC[50]: 108.27 ± 3.38 µg/mL), followed by the A. excelsa
   extract (IC[50]: 117.92 ± 4.92 µg/mL). The n-hexane fraction showed
   weaker inhibition (IC[50]: 138.27 ± 2.47 µg/mL), while chloroform and
   petroleum ether fractions demonstrated no significant activity (NA).
   Simvastatin, used as a reference, inhibition (IC[50]:
   45.92 ± 1.84 µg/mL), indicating that while A. excelsa and its fractions
   possess moderate HMG-CoA reductase inhibitory activity. Figure [96]2b
   bargraph reprrsents the IC[50] value of HMG-CoA reductase inhibitory
   activity. Nevertheless, the ethyl acetate fraction’s effectiveness
   suggests the presence of bioactive compounds with potential
   lipid-lowering effects, results are represented in Table [97]3.

Table 3.

   The IC[50] values for Porcine pancreatic Lipase, HMG coa reductase,
   α-amylase and α-glucosidase inhibitory activity.
   Samples Porcine Pancreatic lipase HMG-CoA reductase α-amylase
   α-glucosidase
   Ailanthus excelsa Extract 97.92 ± 2.37 117.92 ± 4.92 139.63 ± 4.21
   132.97 ± 2.23
   Ethyl acetate* 56.25 ± 4.85 108.27 ± 3.38 125.93 ± 2.29 117.08 ± 3.28
   Chloroform* NI NI 154.32 ± 4.94 153.08 ± 2.86
   Petroleum Ether* NI NI 165.94 ± 3.72 148.38 ± 4.32
   N-hexane* 106.84 ± 3.96 138.27 ± 2.47 134.82 ± 3.37 128.42 ± 4.63
   Orlistat 49.86 ± 4.37 – – –
   Simvastatin – 45.92 ± 1.84 – –
   Acarbose – – 28.73 ± 2.97 39.72 ± 3.52
   [98]Open in a new tab

   Showing mean ± S.D (n = 3), ** Fractions, NI: No Inhibition.

α-amylase inhibition

   The inhibition percentages of α-amylase for one extract and four
   fractions were calculated relative to acarbose. The ethyl acetate
   fraction displayed the highest inhibition of α-amylase (IC[50]:
   125.93 ± 2.29 µg/mL), followed by the n-hexane fraction (IC[50]:
   134.82 ± 3.37 µg/mL) and the A. excelsa extract (IC[50]:
   139.63 ± 4.21 µg/mL). The chloroform and petroleum ether fractions
   exhibited weaker inhibitory effects with IC[50] values of
   154.32 ± 4.94 µg/mL and 165.94 ± 3.72 µg/mL, respectively. Acarbose,
   the standard α-amylase inhibitor, demonstrated a significantly stronger
   inhibition (IC[50]: 28.73 ± 2.97 µg/mL). Figure [99]2c shows the IC[50]
   values for α-amylase inhibition. Although the plant fractions are less
   effective than acarbose, the ethyl acetate fraction’s activity is
   noteworthy, indicating potential for carbohydrate digestion modulation.
   .

α-glucosidase Inhibition

   Evaluating the percentage inhibition of α-glucosidase for one extract
   and four fractions. Similar trends were observed in the inhibition of
   α-glucosidase, with the ethyl acetate fraction showing the highest
   activity (IC[50]: 117.08 ± 3.28 µg/mL), followed by the A. excelsa
   extract (IC[50]: 132.97 ± 2.23 µg/mL). The n-hexane fraction also
   exhibited moderate inhibition (IC[50]: 128.42 ± 4.63 µg/mL), while the
   chloroform and petroleum ether fractions were less effective, with
   IC[50] values of 153.08 ± 2.86 µg/mL and 148.38 ± 4.32 µg/mL,
   respectively. Acarbose, as the standard, had an IC[50] value of
   39.72 ± 3.52 µg/mL, indicating a much stronger inhibition compared to
   the plant fractions. However, the ethyl acetate fraction’s inhibition
   of α-glucosidase is promising for controlling postprandial
   hyperglycemia. Figure [100]2d shows the IC50 values for α-glucosidase
   inhibition.

   The ethyl acetate fraction consistently showed the highest inhibition
   across all four metabolic enzymes, suggesting it contains potent
   bioactive compounds. While the inhibition of pancreatic lipase and
   HMG-CoA reductase was notable, these values are still higher compared
   to reference standards such as orlistat and simvastatin. However, the
   combination of moderate inhibition across multiple enzymes indicates
   that A. excelsa and its fractions may offer a holistic approach to
   managing obesity through the simultaneous modulation of lipid and
   carbohydrate metabolism. Further isolation and characterization of the
   active compounds from the ethyl acetate fraction could lead to the
   development of effective multi-target therapeutic agents.

Network pharmacology

Ailanthus excelsa component-target network

   The components of A. excelsa were obtained from Dr Duke, IMPPAT and
   Published articles, a total of 31 bioactive components were obtained.
   The chemical information of all 31 components given in (Table S3).
   Action targets for these 31 bioactive components were predicted using
   Digep pred, Swiss target prediction and super pred databases. After
   removing duplicates in total, 838 targets of from Digep pred, 705
   targets from Swiss target prediction and 365 targets from super pred
   databases were predicated (Fig. [101]3a). The A.s excelsa component
   target network comprised 1581 nodes and 34493 edges, 43.6 average node
   degree, 0.381 avg. local clustering coefficient, 23017 expected number
   of edges and < 1.0e-16 PPI enrichment p-value (Fig. [102]S1). With
   Apigenin (degree = 561), 3-hydroxyphenyl valeric acid (degree = 228),
   Cinnamic acid (degree = 106), Caftaric acid (degree = 84), Quinic acid
   (degree = 73), 3, 4 Dihydroxy phenylacetic acid (degree = 58),
   Canthin-6-one (degree = 58) identified as the primary bioactive
   components (Fig. S2). “The values of SUID, Average Shortest Path
   Length, Betweenness Centrality, Closeness Centrality, Clustering
   Coefficient, Degree, Layout, Eccentricity, Edge Count, In-degree,
   Component Name, Out-degree, Partner of Multi-edged Node Pairs, and
   Self-loops for 31 bioactive components are provided in Table S4.

Fig. 3.

   [103]Fig. 3
   [104]Open in a new tab

   (a) Common targets representation of digep pred, Swiss target
   prediction and super pred databases. (b) The venn diagram
   representation of common genes between AE components and Obesity
   related targets.

Selection of disease (Obesity) related targets

   To achieve a comprehensive collection of obesity-related targets, data
   were gathered from several key databases. Specifically, DisGeNET
   ([105]https://www.disgenet.org/browser/0/1/0/C0028754/), which compiles
   gene-disease associations, was utilized. The Therapeutic Target
   Database (TTD, [106]https://db.idrblab.net/ttd/) contributed 802
   targets related to obesity, offering detailed information on
   therapeutic proteins, gene targets, associated diseases, pathways, and
   relevant drugs. Gene Card ([107]https://www.genecards.org/) provided an
   extensive set of 10,019 obesity-related targets covering various
   aspects such as genomics, transcriptomics, proteomics, genetics,
   clinical data, and functional annotations. Additionally, 22 targets
   were sourced from the Pharmacogenomics Knowledgebase (PharmGKB,
   [108]https://www.pharmgkb.org/), which focuses on the interplay between
   actionable genes and clinical drug responses. Following the removal of
   duplicate entries, a consolidated set of 546 unique obesity-related
   targets was established.

Intersection of AE components -obesity common-target network

   The intersection of A. excelsa components and obesity-related targets
   identified 255 common targets (Fig. [109]3b). The resulting network,
   which consists of 255 nodes and 4,619 edges includes 28 bioactive
   components of A. excelsa (Fig. [110]4). The main targets and bioactive
   components were analyzed based on their degree values, which represent
   the number of connections each node has within the network. The results
   highlighted that Apigenin (degree = 67), with the highest number of
   connections, plays a central role, indicating its significant
   interaction with obesity-related targets. 3-Hydroxyphenyl valeric acid
   (degree = 32) and Quinic acid (degree = 18) also showed considerable
   connectivity, suggesting their importance. Other components such as
   Caftaric acid (degree = 16), Canthin-6-one (degree = 14),
   3,4-Dihydroxyphenylacetic acid (degree = 12), Oleanolic acid
   (degree = 12), Selinidin (degree = 12), and Malic acid (degree = 9)
   were identified as critical bioactive compounds due to their moderate
   degree values, reflecting their roles in the network. In contrast,
   Salicylic acid, Protocatechuic acid, and Glaucarubine were found to
   have no common targets with the obesity-related disease targets, as
   indicated by their degree values of zero.

Fig. 4.

   [111]Fig. 4
   [112]Open in a new tab

   Ailanthus excelsa Roxb components targets and obesity related common
   target network.

Protein-Protein interaction of AE- obesity targets

   The 255 common targets were imported into String database, and then
   unconnected nodes like NUCB2 and IRAK1BP1 target was discarded. The PPI
   network included 225 nodes and 4619 edges. After that PPI network was
   exported to Cytoscape 3.10.1 for further analysis. The topological
   parameters of the A. excelsa -Obesity PPI network was calculated. The
   significant targets of the A. Excelsa-Obesity PPI network were obtained
   based on DC, BC, CC, EC, IC, LAC, NC, and SC. Finally, important
   targets were obtained, namely, Interleukin-6 (IL6), Tumor necrosis
   factor (TNF), RAC-alpha serine/threonine-protein kinase (AKT1),
   Vascular endothelial growth factor A (VEGFA), Serum albumin (ALB),
   Glyceraldehyde-3-phosphate dehydrogenase (GAPDH), Catenin
   beta-1(CTNNB1), Epidermal growth factor receptor (EGFR), Interleukin-1
   beta (IL-1B), Peroxisome proliferator-activated receptor gamma (PPARG),
   Signal transducer and activator of transcription 3 (STAT3), Estrogen
   receptor (ESR1), Hypoxia-inducible factor 1-alpha (HIF1A),
   NAD-dependent protein deacetylase sirtuin-1 (SIRT1),
   Serine/threonine-protein kinase mTOR (MTOR), Toll-like receptor 4
   (TLR4), Peroxisome proliferator-activated receptor alpha (PPARA),
   Adiponectin (ADIPOQ), Receptor tyrosine-protein kinase erbB-2 (ERBB2)
   and Peroxisome proliferator-activated receptor gamma coactivator
   1-alpha (PPARGC1A) (Fig. [113]5).

Fig. 5.

   [114]Fig. 5
   [115]Open in a new tab

   (a) Ailanthus excelsa Roxb-Obesity PPI network, (b) PPI network by the
   screening criteria of DC ≥ 38, BC ≥ 1248.6538, CC ≥ 0.2555332,
   EC ≥ 0.04289, IC ≥ 9.871477, LAC ≥ 11.105263, NC ≥ 18.84516, and
   SC ≥ 1.0275286E21. (c)Ailanthus excelsa Roxb-Obesity PPI network by the
   screening criteria of DC ≥ 79, BC ≥ 1061.192, CC ≥ 0.266806,
   EC ≥ 0.11179, IC ≥ 11.329657, LAC ≥ 29.544, NC ≥ 54.41227, and
   SC ≥ 6.97917E21. The size and color shading of the nodes were ordered
   by degree. The thickness and color shading of the edges were set
   according to the degree of interaction between the nodes.

MCODE cluster analysis of PPI network

   In order to further analyse the 255 common targets of A. excelsa and
   Obesity, find the key sub-networks, and speculate the main targets, the
   AE-Obesity PPI network was further computationally analyzed by MCODE.
   The PPI network was divided into 7 clusters (Figs. [116]6a–g). The
   score of clusters 1 (score = 22.82) was far higher than the other
   clusters. The degree of the targets in cluster 1 was higher, such as
   IL6 (degree = 98), TNF (degree = 93), AKT1 (degree = 88), VEGFA
   (degree = 88), IL1B (degree = 81), PTGS2 (degree = 78), PPARG
   (degree = 78).

Fig. 6.

   [117]Fig. 6
   [118]Open in a new tab

   Cluster analysis of Ailanthus excelsa Roxb-Obesity PPI Network (a)
   Cluster 1, Composed of 30 nodes and 231 edges (Score = 22.828) (b)
   Cluster 2, composed of 39 nodes and 281 edges (Score = 14.789). (c)
   Cluster 3, composed of 28 nodes and 105 edges (Score = 7.778). (d)
   Cluster 4, composed of 7 nodes and 14 edges (Score = 4.667). (e)
   Cluster 5, composed of 6 nodes and 9 edges (Score = 3.600). (f) Cluster
   6, composed of 6 nodes and 8 edges (Score = 3.200). (G) Cluster 7,
   composed of 3 nodes and 3 edges (Score = 3.00).

GO and KEGG pathway enrichment analyses of PPI network

   Enrichment analysis of the A. excelsa -Obesity PPI network was
   obtained. In GO enrichment analysis, 255 AE-Obesity common targets were
   highly enriched in 2027 biological processes (BP), 101 cellular
   fractions (CC), and 187 molecular functions (MF) by setting p-values
   less than 0.05. The top 20 significant results for each were taken as
   shown in the Fig. [119]7. Biological processes (BP) show a strong
   correlation with responses to oxygen-containing compounds, organic
   substances, chemicals, endogenous stimuli, organic cyclic compounds,
   hormones, and lipids. These responses, both general and cell-specific,
   are crucial for managing obesity. Additionally, regulatory mechanisms
   such as the regulation of biological quality, multicellular organismal
   processes, and cell communication play significant roles. Obesity
   influences these responses through external stimuli like diet and
   environment. The positive regulation of biological processes suggests
   mechanisms that enhance biological activity, which may be crucial for
   understanding the body’s adaptation to fat accumulation. This indicates
   a potential role for A. excelsa in obesity management.

Fig. 7.

   [120]Fig. 7
   [121]Open in a new tab

   Gene ontology analysis of PPI network (BP: Biological process, CC:
   Cellular component, MF: Molecular function).

   Response to oxygen-containing compound (GO:1901700) score the lowest
   false discovery rate of 6.75E-84 via the modulation of 145 genes i.e.
   CFTR, OTC, DRD4, CPB2, SIRT1, HMOX1, TIMP1, CTSH, PON1, SERPINE1, GAA,
   RPS6KB1, CCL2, CD38, NFKB1, IL2, CYP27B1, NR3C1, IGFBP2, GHSR, AHR,
   CD68, KCNA5, SLC6A4, IL1B, PIK3CA, BCHE, STAT3, DNMT3A, GCKR, CPT1A,
   GAL, SOD1, SLC6A3,REN, SRD5A1,EGFR, HTR2C, CA3,PPARG, ACE,
   CCR5,HMGB2,IL15,NOS3,MMP3,ACHE, BCR, INSR, ADRB2,DRD5,CXCL8,OGG1,GLUL,
   PRKACA, PPARD, GH1,HNF4A, EDNRA,
   HSD11B2,ROCK2,NQO1,EZH2,SLC2A4,HSPA5,PRKCD, CD86,MAPT, HRH3,ESR2,PDE4D,
   PRKAA1,FYN, C5AR1,ANO1,RET, HP, ARG1,PDE3A, RXRG, CES1,NTRK3,MTOR,
   DRD2,PARP1,PTGS2,BGLAP, CNR1,SCD, PTPN1,MMP9,ABL1,TLR4,CNR2,RXRB, AR,
   ABCA1,F7,WT1,FOXO1,AKR1C3,JAK2,DRD3,TFRC, G6PD,
   DRD1,CYP19A1,NFE2L2,BCL2,CRHR1,ROCK1,PPARA, IL6,ADIPOQ, TNF,
   ESR1,GLRA1,CD36,SLC2A1,PIK3CG, RXRA, AMY2A, IDO1,TYK2,JAK3,CASP1,HTR2A,
   HIF1A, TSHR, NR1H4,DIO2,AKT1,TRPV1,H6PD, CCL5,NR1H3,NOS1,CYP1B1,ACACA,
   CTNNB1,PRKCB, IGF1R, ENPP1,CALM1,and JAK1against 1547 background gene
   at a strength of 0.86.

   Cellular components (CCs) are closely correlated with membrane rafts,
   receptor complexes, plasma membrane components, vesicles, cytoplasm,
   cell periphery, and extracellular space. Membrane rafts and receptor
   complexes in the plasma membrane are critical for signaling pathways
   that regulate appetite and energy homeostasis. The integral and
   intrinsic components of the plasma membrane, along with vesicles, are
   essential for nutrient uptake, signal transduction, and lipid/glucose
   transport. Cytoplasm and cell periphery drive metabolic reactions,
   while extracellular signaling molecules influence cellular behavior.
   The endomembrane system, including mitochondria, is vital for protein
   and lipid processing in energy metabolism. The cell surface mediates
   interactions with the extracellular environment, impacting nutrient
   absorption and communication. Neural components such as the
   somatodendritic compartment regulate appetite and energy expenditure,
   underscoring the complex interplay in obesity regulation. Membrane raft
   (GO: 0045121) has the lowest false discovery rate (8.20E-17) via
   modulation of 34 genes (HMOX1, SLC6A2, GHSR, KCNA5, SLC6A4, NPC1,
   SLC6A3, EGFR, NOS3, INSR, S1PR1, PRKACA, SLC2A4, TRPM8, MAPT, FYN,
   ANO1, RET, DPP4, PTGS2, CNR1, ABCA1, JAK2, TNF, CD36, SLC2A1, HTR2A,
   LCK, NOS1, HK1, CTNNB1, ITGAM, IGF1R, ABCG2) against 320 background
   genes, with a strength of 0.91.

   MF was associated with Protein binding, Binding, Identical protein
   binding, signaling receptor activity, nuclear receptor activity, Ion
   binding and signaling receptor binding etc., the obtained molecular
   functions that are significant in the context of obesity, describing
   how different proteins and molecules interact and participate in
   cellular processes. Protein binding and identical protein binding refer
   to the interactions between proteins, which can affect signaling
   pathways and metabolic processes. Signaling receptor activity and
   signaling receptor binding are crucial as they relate to how cells
   receive and respond to external signals, which can influence appetite,
   energy expenditure, and fat storage. Nuclear receptor activity,
   including nuclear steroid receptor activity, involves receptors that
   regulate gene expression in response to hormones, impacting metabolism
   and fat distribution. Ion binding, such as cation and transition metal
   ion binding, affects enzyme function and cellular signaling. Catalytic
   activity, enzyme binding, and protein tyrosine kinase activity are
   involved in biochemical reactions and signal transduction pathways that
   regulate metabolism. Organic cyclic compound binding and small molecule
   binding relate to how the body processes various compounds, including
   drugs and hormones. Phosphotransferase activity, including kinase and
   protein kinase activities, involves phosphorylation processes critical
   for regulating metabolic pathways. Transferase activity, transferring
   phosphorus-containing groups, is essential for modifying proteins and
   metabolites, impacting energy metabolism. Understanding these molecular
   functions helps in deciphering the complex biological processes that
   contribute to obesity.

   In addition, a total of 202 pathways were enriched by KEGG enrichment
   analysis, and the obesity related important pathways and its
   connectiveness with targets, gene ratio, and its gene count were shown
   in the Fig. [122]8. The results indicated that A. excelsa may regulate
   Obesity through Adipocytokine signaling pathway, PPAR signaling
   pathway, Insulin signaling pathway as well as through Fatty acid
   biosynthesis, Fat digestion and absorption and Glycolysis /
   Gluconeogenesis. A. excelsa also regulates metabolism pathways such as
   the Galactose metabolism, Fructose and mannose metabolism, Cholesterol
   metabolism, Fatty acid metabolism, Starch and sucrose metabolism and
   metabolic pathways.

Fig. 8.

   [123]Fig. 8
   [124]Open in a new tab

   Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment
   analysis of Ailanthus excelsa.

GO and KEGG pathway enrichment analyses of cluster 1

   GO and KEGG enrichment analysis was performed on cluster 1
   (score = 22.828), which had the highest score in the cluster analysis,
   to further obtain the functional information and pathway information of
   cluster 1. GO enrichment analysis was performed on cluster 1 by using
   ClueGO and CluePedia plugins. 30 targets present in cluster 1 were
   highly enriched in 227 biological processes by setting p-values less
   than 0.05. The significant results for each were taken as shown in the
   Fig. [125]9. GO:0001516 (prostaglandin biosynthetic process) involves
   genes like IL1B, PTGS2, and SIRT1, which are pivotal in inflammation
   and metabolic regulation. GO:0001935 (endothelial cell proliferation)
   and GO:0001937 (negative regulation of endothelial cell proliferation)
   highlight the role of genes such as CCL2, PPARG, and TNF in modulating
   vascular health and inflammation. GO:0002237 (response to molecule of
   bacterial origin) involves a range of immune response genes (e.g.,
   IL1B, TNF, TLR4) crucial for managing systemic inflammation. GO:0002367
   (cytokine production involved in immune response) and GO:0002526 (acute
   inflammatory response) emphasize the importance of cytokines like IL1B
   and TNF in inflammation. GO:0010874 and GO:0010875 (regulation of
   cholesterol efflux) show the role of APOA1, PPARG, and SIRT1 in lipid
   metabolism. GO:0010883 and GO:0010888 (regulation of lipid storage) and
   GO:0032368 (regulation of lipid transport) further highlight the
   involvement of PPARG and SIRT1 in managing lipid levels. Collectively,
   these genes and their associated processes are critical in controlling
   inflammation, lipid metabolism, and vascular health, all of which are
   essential for effective obesity management.

Fig. 9.

   [126]Fig. 9
   [127]Open in a new tab

   Cluster 1 Biological process (a) ClueGO analysis of biological process
   (b) CluePedia analysis of biological process.

   Furthermore, a total of 30 pathways were enriched by KEGG enrichment
   analysis, and the pathways were shown in the Fig. [128]10. In treating
   obesity, various KEGG pathways play crucial roles in modulating
   metabolic and inflammatory responses. The PPAR signaling pathway
   (KEGG:03320), involving genes such as APOA1, MMP1, PPARA, and PPARG,
   regulates lipid metabolism and enhances fat oxidation. The NF-kappa B
   signaling pathway (KEGG:04064), with genes like CXCL8, IL1B, PTGS2,
   TLR4, and TNF, controls inflammation, which is often elevated in
   obesity. Similarly, the HIF-1 signaling pathway (KEGG:04066) and the
   Toll-like receptor signaling pathway (KEGG:04620) involve genes such as
   HMOX1, IL6, TLR4, and TNF, which impact metabolic adaptations and
   inflammatory responses. The Adipocytokine signaling pathway
   (KEGG:04920) and Insulin resistance pathway (KEGG:04931), featuring
   PPARA, SLC2A1, and TNF, improve insulin sensitivity and glucose
   homeostasis, critical in obesity management. Additionally, pathways
   like the AGE-RAGE signaling pathway (KEGG:04933) and lipid and
   atherosclerosis (KEGG:05417) address oxidative stress and
   cardiovascular risks associated with obesity. By targeting these
   pathways, it is possible to balance energy metabolism, reduce
   inflammation, and improve insulin sensitivity, offering a comprehensive
   approach to obesity treatment. The bar graph and pie chart of
   biological process of cluster 1 given in supplementary Fig. S3, S4. [GO
   and KEGG pathway enrichment analyses of the cluster 2 and 3 given in
   supplementary file fig. S5-S8].

Fig. 10.

   [129]Fig. 10
   [130]Open in a new tab

   Cluster 1 KEGG pathway analysis (a) ClueGO analysis (b) CluePedia
   analysis (c) Pie chart representation (d) Bar graph of KEGG pathways.

Network construction and analysis

   The network analysis revealed potential relationships between compounds
   and targets, implying the pharmacological mechanisms of the compounds.
   Nodes with the highest degree of connections represent hubs in the
   network, indicating potential drugs or targets. The core
   component-target (C-T) network of A. excelsa for obesity was analyzed
   based on degree values. Apigenin emerged as a highly modulated
   component, with a degree of 40, eccentricity of 5, neighbourhood
   connectivity of 3.22, radiality of 0.9652, stress of 35,312, and a
   topological coefficient of 0.1392. Alongside Apigenin,
   3-hydroxyphenylvaleric acid (degree = 10), selinidin (degree = 9), and
   quinic acid (degree = 7) also demonstrated significant modulation
   within the C-T network. These findings suggest that a single compound
   can target multiple receptors, contributing to the network’s
   interconnectedness.

   These compounds targeted key modulators of obesity, such as HK1, HK2,
   and ADIPOQ, as well as other targets involved in obesity-related
   complications, including PPARA and INSR. Furthermore, several targets,
   such as FASN, PIK3CA, MTOR, and AKT1, were synergistically regulated by
   different compounds, emphasizing the multitargeting nature of obesity
   treatment. The top 15 significant targets identified were HK1, HK2,
   PIK3CA, AKT1, AMY2A, GAA, MTOR, CD36, ACACB, SLC2A4, CPT1A, INSR,
   ACACA, FASN, and ADIPOQ.

   Pathway analysis, crucial in systems pharmacology, bridges
   receptor-ligand interactions with pharmacodynamics outcomes. To focus
   on disease relevance, obesity-associated pathways were used to
   construct the target-pathway (T-P) network. Using integrated KEGG
   pathway analysis, 238 interactions between 127 targets and 19
   disease-associated pathways were mapped into the T-P network using
   Cytoscape 3.10.1 (P < 0.01). The T-P network identified metabolic
   pathways as the most modulated pathway (degree = 51, average shortest
   path length = 2.16, betweenness centrality = 0.45344, closeness
   centrality = 0.4629, eccentricity = 5, neighbourhood
   connectivity = 3.01961, radiality = 0.97725, stress = 100,380, and
   topological coefficient = 0.0594), followed by insulin resistance
   (degree = 26), AMPK signaling (degree = 24), JAK-STAT signaling
   (degree = 20), and insulin signaling pathways (degree = 19). The roles
   of these high-degree pathways in obesity are well-established.

   Additionally, other signaling pathways, such as the Adipocytokine
   signaling pathway and PPAR signaling pathway, also contribute to the
   development of obesity. An example of the synergism of compounds acting
   on metabolism pathways includes Fatty acid metabolism, Galactose
   metabolism, Fructose and mannose metabolism, Cholesterol metabolism,
   and Carbohydrate digestion and absorption (Fig. [131]11). In addition
   to disease-associated pathways, the compounds also influence
   non-disease-associated pathways, including the AGE-RAGE signaling
   pathway in diabetic complications (hsa04933), PI3K-Akt signaling
   pathway (hsa04151), Glucagon signaling pathway (hsa04922), and
   Non-alcoholic fatty liver disease (hsa04932). Since obesity can
   potentially develop into diabetes mellitus, it can be inferred that
   components present in A. excelsa could also prevent the development of
   diabetes mellitus. Detailed pathway information, including KEGG IDs,
   observed gene counts, background gene counts, false discovery rates,
   and matching proteins, is provided in Table S5.

Fig. 11.

   [132]Fig. 11
   [133]Open in a new tab

   Network analysis of Phytoconstituents present in Ailanthus excelsa and
   obesity related targets, pathways.

   One of the most significant pathways for A. excelsa in treating
   obesity, according to the degree value, is the metabolic pathways,
   indicating that A. excelsa may facilitate enzymatic reactions within
   cells that convert substrates into various end products. These pathways
   support energy production, biosynthesis of essential molecules, and
   waste elimination. Enzymes catalyse each step, ensuring efficient and
   regulated transformation of molecules, supporting vital cellular
   functions, and maintaining homeostasis in the body. By considering
   targets and pathways, A. excelsa could exert its therapeutic effects
   through influencing multiple pathways and acting on multiple targets in
   each pathway.

Probable molecular mechanism modulated by Ailanthus excelsa components

   Genes involved in the treatment of obesity impact various metabolic
   pathways and processes. Activation genes include SIRT1, which enhances
   insulin sensitivity through protein deacetylation; PPARGC1A, which
   boosts mitochondrial function and oxidative metabolism; CPT1A, which
   facilitates fatty acid oxidation; FABP4, which aids fatty acid
   transport; and ADIPOQ, which improves insulin sensitivity and fatty
   acid metabolism. PPARG regulates adipocyte differentiation and lipid
   metabolism, while ACLY is crucial for fatty acid and cholesterol
   biosynthesis. FOXO1 and STAT3 play roles in gluconeogenesis and energy
   balance, respectively, and PRKACA activates protein kinase A,
   influencing glucose metabolism and lipolysis. In contrast, Inhibition
   genes such as FASN and ACACA promote lipid accumulation when
   overactive, while SCD modulates fat accumulation through fatty acid
   conversion. MTOR and NFKB1 are associated with metabolic disorders and
   inflammation linked to obesity, and PTGS2 contributes to
   obesity-related inflammation. Association genes like PIK3CA and INSR
   are involved in insulin signaling and glucose uptake, with
   dysregulation linked to obesity. CD36 influences fatty acid storage,
   and IL6 and HMGCS1 are associated with inflammation and metabolic
   regulation. RXRA and JAK2 are involved in lipid metabolism and cytokine
   signaling, respectively, while ADIPOQ is crucial for insulin
   sensitivity and obesity risk (Fig. [134]12). These genes collectively
   influence energy metabolism, fat storage, inflammation, and insulin
   sensitivity, highlighting their roles in obesity treatment ([36, 37]).

Fig. 12.

   [135]Fig. 12
   [136]Open in a new tab

   Probable molecular mechanism of Ailanthus excelsa Roxb components and
   its targeted proteins in AMPK signalling pathway (KEGG ID: map04152).

Molecular docking

Molecular docking studies on α-amylase and α-glucosidase

   Among the selected phytocompounds from the literature, compound
   Isoquercitrin demonstrating highest binding affinity at − 8.96 and
   glide energy − 55.83 kcal/mol and displayed similar bond interactions
   as opposed to standard acarbose with ASP197, GLU233, ILE235, ASP300
   (Fig. [137]14a). and Acarbose displays binding affinity at -10.36 and
   glide energy − 54.0 kcal/mol and displays hydrogen bond interaction
   with ASP197, GLU233, ILE235, ASP300, ARG195, LYS200, GLU240, TYR151
   residues in α-amylase protein (PDB ID:4W93 / Fig. [138]13a). Similarly,
   in α-glucosidase binding pocket (PDB ID: 3A47). Isoquercitrin
   demonstrating highest binding affinity at -10.41 and glide energy
   − 59.44 kcal/mol and displayed similar bond interactions as opposed to
   standard acarbose with PRO309, GLU304, ASP408 (Fig. [139]13b) and
   Vitexin demonstrating second highest binding affinity at − 10.21 and
   glide energy − 52.75 kcal/mol with SER308, PHE157, LYS155, HIE239 and
   Acarbose displays binding affinity at -10.2 and glide energy
   − 62.38 kcal/mol and displays hydrogen bond interaction with PRO309,
   ASN241, GLU304, ASP408, PHE157 residues (Fig. [140]14b).

Fig. 14.

   [141]Fig. 14
   [142]Open in a new tab

   Standard Acarbose in (a) α-amylase protein with ASP197, GLU233, ILE235,
   ASP300, ARG195, LYS200, GLU240, TYR151 (PDB ID:4W93) (b) α-glucosidase
   binding pocket with PRO309, ASN241, GLU304, ASP408, PHE157 amino acid
   residues (PDB ID:3A47).

Fig. 13.

   [143]Fig. 13
   [144]Open in a new tab

   Isoquercitrin in (a) α-amylase protein with ASP197, GLU233, ILE235,
   ASP300 residues (PDB ID: 4W93) (b) α-glucosidase binding pocket with
   SER308, PHE157, LYS155, HIE239 (PDB ID: 3A47).

Molecular docking studies on lipase and HMG-CoA

   Similarly, in Lipase (PDB ID: ILPB) binding pocket compound
   Isoquercitrin demonstrating highest binding affinity at -9.96 and glide
   energy − 52.84 kcal/mol and displayed bond interactions with ARG44,
   LEU41, ARG337, ASP331, ALA332, GLN29 and standard Orlistat displayed
   − 2.8 and glide energy − 43.12 kcal/mol and interactions with ARG44,
   LEU41 and ARG337 Fig. [145]15.

Fig. 15.

   [146]Fig. 15
   [147]Open in a new tab

   2D Orientation of (a) Isoquercitrin with ARG44, LEU41, ARG337, ASP331,
   ALA332, GLN29 residues (b) Standard orlistat with ARG44, LEU41 and
   ARG337 residues in lipase binding pocket (PDB ID: 1LPB).

   Figure. 16. shows the binding interactions of Isoquercitrin and
   simvastatin within the HMG-CoA binding pocket (PDB ID: 1HW9).
   Isoquercitrin exhibits a binding affinity of − 7.11 and a glide energy
   of − 37.60 kcal/mol. Its interactions include key residues such as
   ALA751, LYS735, GLU559, and GLY560. In contrast, the standard drug
   simvastatin demonstrates a lower binding affinity of -2.13 and a glide
   energy of -26.82 kcal/mol. Simvastatin interacts primarily with GLU559
   and ASN755, forming a salt bridge with LYS735 (Fig. [148]16).

Fig. 16.

   [149]Fig. 16
   [150]Open in a new tab

   2D Orientation of (a) Isoquercitrin with ALA751, LYS735, GLU559, GLY560
   residues, (b) Standard simvastatin with GLU559, ASN755 and salt bridge
   with LYS735 residues in HMG-CoA binding pocket (PDB ID: 1HW9).

Molecular dynamics simulation study

   A molecular dynamics simulation was conducted to investigate the
   stability and conformational changes of specific Isoquercitrin within
   the binding pocket of lipase (PDB ID: 1LPB) and HMG-CoA protein (PDB
   ID: 1HW9). The Desmond program was utilized to analyze interactions
   during the 100ns simulation, with key properties such as Root Mean
   Square Deviation (RMSD), Root Mean Square Fluctuation (RMSF), and
   hydrogen bond contacts being assessed. The RMSD values reflect
   alterations in both protein and ligand conformations relative to the
   initial reference structure (Frame 0) chosen for the simulation.

   Compound Isoquercitrin was analysed by Protein-ligand RMSD plot.
   Protein structure RMSD to reference C-alpha RMSD and Backbone RMSD
   value showed a maximum deviation of 2.169 Ǻ and 2.186 Ǻ (frame 125)
   respectively. Ligand stability was assessed by ligand RMSD to Ligand
   aligned on protein and ligand aligned on ligand indicating a maximum
   deviation of 2.481 Ǻ and 1.478 Ǻ (Frame 147) in Lipase protein. The
   Protein RMSF plot of Isoquercitrin showed ASP_249 showed maximum
   fluctuation of 2.142 Ǻ and 1.943 (frame 337) for C-alpha and protein
   backbone RMSF. The Ligand RMSF plot showed high fluctuation with RMSF
   values of 2.220 Ǻ and 1.389 Ǻ (frame 23) aligned on protein and aligned
   on ligand respectively (Fig. [151]17).

Fig. 17.

   [152]Fig. 17
   [153]Open in a new tab

   (a) Protein Ligand RMSD Plot of Isoquercitrin, (b) Protein RMSF Plot,
   (c) Protein Ligand Contacts timeline, (d) Ligand RMSF plot over period
   of 100 ns within lipase protein.

   Compound Isoquercitrin was analysed by Protein-ligand RMSD plot.
   Protein structure RMSD to reference C-alpha RMSD and Backbone RMSD
   value showed a maximum deviation of 2.671 Ǻ and 2.687 Ǻ (frame 228)
   respectively. Ligand stability was assessed by ligand RMSD to Ligand
   aligned on protein and ligand aligned on ligand indicating a maximum
   deviation of 2.456 Ǻ and 2.246 Ǻ (Frame 53) in HMG-CoA protein. The
   Protein RMSF plot of Isoquercitrin showed LYS735 showed maximum
   fluctuation of 2.606 Ǻ and 2.633 (frame 294) for C-alpha and protein
   backbone RMSF. The Ligand RMSF plot showed high fluctuation with RMSF
   values of 4.653 Ǻ and 0.670 Ǻ (frame 14) aligned on protein and aligned
   on ligand respectively (Fig. [154]18).

Fig. 18.

   [155]Fig. 18
   [156]Open in a new tab

   (a) Protein Ligand RMSD Plot of Isoquercitrin, (b) Protein RMSF Plot,
   (c) Protein Ligand Contacts timeline, (d) Ligand RMSF plot over period
   of 100 ns within HMG-CoA protein.

Intermolecular contacts in molecular dynamics simulation

   The stacked bar chart in Fig. [157]19. Presents a stacked bar chart
   depicting intermolecular contacts between inhibitors and the protein
   observed throughout the molecular dynamic’s simulation. This
   visualization categorizes interactions into distinct types, including
   hydrogen bonds, hydrophobic interactions, ionic interactions, and water
   bridges. Each interaction type is further subdivided to provide
   detailed insights into the nature of the interactions occurring during
   the simulation. The simulation interaction diagram explores these
   interactions, and Isoquercitrin is identified as the best-scoring
   inhibitor, with H-bonds and water bridges with SER35, LEU36, ARG38,
   LEU41, LYS239, ASP247, ILE248. Hydrophobic interactions are observed
   with ASP332 and the 2D interaction of Isoquercitrin has a polar
   interaction with GLN29, water bridge with LEU41, ASP389, GLU13, and
   charged (negative) bond with ASP331, ASP247, ASP31 amino acids in
   lipase pocket in lipase protein and Similarly, Isoquercitrin is
   identified as the best-scoring inhibitor, with H-bonds and water
   bridges with GLU559, GLY560, ARG568, LEU721, ASN755, GLY849 and the 2D
   interaction of Isoquercitrin has charged (negative) bond with GLU559
   and water bridge with GLY849 amino acids in HMG-CoA protein.

Fig. 19.

   [158]Fig. 19
   [159]Open in a new tab

   (a) Protein ligand contacts of compound Isoquercitrin with the
   respective amino acids of the lipase protein (b) Ligand protein
   contacts of Isoquercitrin with the respective amino acids of the
   HMG-CoA protein.

Discussion

   The present study was designed to investigate the interaction of
   bioactives from A. Excelsa with proteins involved in obesity and
   diabetes, and further propose the molecular mechanism of A. Excelsa in
   the management of metabolic dysfunction like obesity and diabetes via
   integrating various computational chemistry, system biology tools like
   molecular dynamic simulations, gene enrichment analysis, gene ontology,
   and cluster analysis with experimental in vitro results.

   Initially, the phytochemical analysis of the three fractions, which
   include hydro alcoholic, ethyl acetate, and N-hexane fractions, was
   subjected to estimate the total phenol, flavonoid, and tannin content,
   among which ethyl acetate was found to possess a high content of
   flavonoids and alkaloids. Further the phytoconstituent, from all the
   extract were retrieved from Dr. Dukes, IMPPAT and literature search, we
   identified macromolecule targets involved in pathogenesis of obesity
   and diabetes. Hence alteration in the molecular function of these
   identified protein targets can be possible by binding of the bioactive
   molecules. Hence it becomes the rationale behind the usage of
   phytochemicals of A. Excelsa in treatment of metabolic dysfunction like
   obesity and diabetes.

   Adiponectin (ADIPOQ) produced by the adipocytes, is one of the highly
   modulated protein by the bioactives of A. Excelsa and is associated
   with obesity and insulin resistance^[160]37. In obesity adiponectin
   levels are decreased^[161]38, the cholesterol accumulation in adipose
   tissue dysregulates adipocyte functions, owing to reduced adiponectin
   secretion^[162]39. Also, the dysregulation of the mevalonate pathway
   which occurs via the HMG-CoA reductase enzyme causes to inflammation
   and oxidative stress, which leads to suppressing of adiponectin
   levels^[163]40,[164]41. Thus, the molecules which inhibit the HMG-CoA
   reductase enzyme have shown to increase adiponectin levels previous
   studies^[165]42. Hence in present study, the three extract were
   screened for their inhibitory potential against the, HMG-CoA enzyme
   inhibition. Among them, Ethyl acetate extract was found to be the
   better inhibitor of the HMG-CoA. Also, the molecular docking studies
   revealed the binding of phytoconstituent in the binding pocket of
   HMG-CoA enzyme. Isoquercetin showed highest binding interaction with
   the binding Glide energy of -37 kcal/mol with three hydrogen bond
   interaction with ALA 751, GLU 559, and GLY 560 as compared to standard
   simvastatin. The MD simulation results revealed there was fluctuation
   seen in the protein ligand RMSD of Isoquercetin-HMG-CoA reductase
   complex, which was stabilized after the 85ns. However, the, RMSF was
   less than 3Å with the amino acid residues in contact with the ligand,
   owing to its better binding throughout the simulation.

   Similarly, there exist a close association of adiponectin and lipase
   activity, as adiponectin levels are inversely related to lipase
   activity, increase in the lipase activity cause enhanced fat absorption
   and storage leading to adiposity which further cause in reduction of
   adiponectin levels^[166]43,[167]44. Thus, reduced adiponectin level in
   obesity cause impaired lipid metabolism and insulin sensitivity. Hence
   targeting the pancreatic lipase enzyme can help increase the
   adiponectin levels leading to better lipid metabolism^[168]45,[169]46.
   Molecular docking at the lipase target, displayed Isoquercetin as
   better binding bioactive with the binding glide energy of
   -52.84 kcal/mol with five hydrogen bond interactions with LEU 41, ARG
   337, ASP 331, ALA 332, and GLN 29, as compared to standard orlistat.
   The RMSD analysis of isoquercetin-lipase enzyme displayed better
   stability with deviation of less than 3Å. Also, the RMSF was seen to be
   fluctuating between 1Å and 2Å indicating stable complex throughout the
   simulation.

   Postprandial hyperglycemia is strongly associated with the progression
   of obesity and metabolic disorders, α-amylase and α-glucosidase are the
   two main enzymes involved in the breakdown polysaccharides into
   glucose, which cause surge in blood glucose levels in diabetic
   condition which leads to fat accumulation and lipogenesis, which causes
   the suppression of adiponectin levels, which ultimately leads to
   increased fat storage and expansion of adipose tissue^[170]47. Herein
   the molecular docking results displayed Isoquercetin to be the highest
   docked bioactive against the α-amylase enzyme with the glide binding
   energy of -55.83 kcal/mol with hydrogen bond interactions with ASP 197,
   GLU 233, ILE 235, and ASP 300. Isoquercetin showed glide binding energy
   of -59.44 kcal/mol and hydrogen bonding with PRO 309, GLU 304, and ASP
   408 residues as compared to standard acarbose.

   Moreover, it is important to recognize that a single compound has the
   potential to interact with multiple proteins via modulating multiple
   pathways^[171]48. The current study further emphasizes the evaluation
   of other lead compounds derived from A. Excelsa to explore their
   potential for modulating multiple pathways involved in the pathogenesis
   of obesity and validating the other protein targets interacting with
   the bioactives.

Conclusion

   This study represents the first comprehensive evaluation of the
   antioxidant and enzyme inhibitory activities of Ailanthus excelsa,
   incorporating cluster analysis to further elucidate its effects. Our
   findings provide compelling evidence of the plant’s therapeutic
   potential for managing obesity through the modulation of key metabolic
   enzymes. Network pharmacology analysis identified critical targets
   associated with obesity, including ADIPOQ, PPARA, PPARG, IL6, TNF, and
   AKT1, which are involved in modulating metabolic pathways. Isoquercetin
   demonstrated strong binding affinities to HMG-CoA reductase, lipase,
   α-amylase, and α-glucosidase, supporting its role as a bioactive
   metabolite in the plant’s anti-obesity effects. The ethyl acetate
   fractions exhibited potent inhibitory activities, with IC[50] values of
   56.25 ± 4.85 µg/mL for lipase, 108.27 ± 3.38 µg/mL for HMG-CoA
   reductase, 117.08 ± 3.28 µg/mL for α-glucosidase, and
   125.93 ± 2.29 µg/mL for α-amylase. Future studies should focus on
   isolating and characterizing the individual bioactive compounds within
   the plant to better understand their structural and functional
   attributes. Furthermore, the promising results from this study pave the
   way for exploring the antidiabetic and anti-obesity properties of A.
   excelsa in greater detail. Continued investigation into its bioactive
   components and their interactions with metabolic pathways will provide
   valuable insights for the development of novel therapeutic strategies
   for obesity and related metabolic disorders.

Supplementary Information

   Below is the link to the electronic supplementary material.
   [172]Supplementary Material 1^ (3.4MB, docx)

Acknowledgements