Synergistic effects of curcumin-donepezil therapy in a Drosophila Alzheimer’s disease model

ABSTRACT

Background

Despite emerging therapeutic options, Alzheimer´s disease (AD) management remains suboptimal due to multimodal pathogenesis. We investigated curcumin-donepezil combination therapy, as curcumin demonstrates antioxidant, anti-inflammatory, and anti-amyloidogenic properties that may complement donepezil’s cholinesterase inhibition.

Research design and methods

We employed the elav-Gal4/UAS-hAPP-BACE-1 Drosophila melanogaster model alongside molecular docking simulation and ADMET prediction to evaluate curcumin-donepezil combination versus monotherapy. Fruit flies received the treatment regimen, and were tested for survival, memory performance, and biochemical markers, including BACE-1 activity and oxidative stress parameters.

Results

Combination therapy significantly improved survival rates and memory performance compared to individual treatment. The combination effectively modulated multiple AD-related pathways, demonstrating reduced BACE-1 activity and decreased oxidative stress markers. Molecular docking confirmed favorable drug interactions, and ADMET profiles supported therapeutic viability.

Conclusions

Curcumin-donepezil combination therapy shows promise as a multi-target approach for AD management. However, translation to clinical applications requires validation in higher-order models and human trials

Plain Language Summary

Alzheimer’s disease is a brain condition that affects memory and thinking. Current medicines only help with some symptoms but don’t treat all the different problems that cause the disease, so patients still get worse over time. We studied combining two treatments: donepezil (a current Alzheimer’s medicine) with curcumin (a natural compound from turmeric spice). Curcumin can fight harmful substances in the brain and reduce swelling. We wanted to see if using both together works better than using each one alone.

We used fruit flies that were specially bred to have brain problems similar to Alzheimer’s disease. We gave different groups of flies either donepezil alone, curcumin alone, or both medicines together. Then we measured how long they lived, how well they remembered things, and checked for harmful substances in their brains. We also used computer models to predict how the two medicines would work together. Flies that received both medicines together lived longer and had better memory than flies that got only one medicine. The combination treatment also reduced harmful brain substances better than single treatments. Our computer tests showed the two medicines work well together and should be safe.

Using curcumin and donepezil together might be a better way to treat Alzheimer’s disease because it attacks the disease from different angles. However, we need more studies in people to confirm safety and effectiveness.

1. Introduction

Alzheimer’s disease (AD) represents one of the most challenging neurodegenerative disorders, characterized by progressive cognitive decline and neuronal loss. The pathological hallmarks of AD include the accumulation of amyloid-beta (Aβ) plaques and neurofibrillary tangles, accompanied by neuroinflammation, synaptic dysfunction, and oxidative stress [1,2]. Central to AD pathogenesis is the amyloid precursor protein (APP), a transmembrane protein that undergoes sequential enzymatic cleavage to produce Aβ peptides. When APP is cleaved by β-secretase (BACE1) and subsequently by γ-secretase, it generates Aβ fragments that can aggregate into soluble oligomers and insoluble plaques, initiating a cascade of neurodegeneration [3,4].

Current therapeutic approaches for AD fall into two main categories, each with significant limitations. Anti-amyloid strategies, including recently FDA-approved monoclonal antibodies, directly target Aβ but have shown modest clinical benefits with concerning side effects such as brain edema and hemorrhage [5]. Many anti-amyloid trials have failed due to the complex relationship between plaque burden and cognitive symptoms, questioning whether Aβ reduction alone is sufficient for meaningful clinical improvement [6]. Neuromodulators, including cholinesterase inhibitors (donepezil, rivastigmine, galantamine) and NMDA receptor antagonists (memantine), provide temporary symptomatic relief by enhancing neurotransmission but do not address underlying pathological processes or halt disease progression [7]. The limited success of monotherapies has highlighted the multimodal nature of AD pathogenesis and sparked interest in combination therapeutic strategies. These approaches aim to simultaneously target multiple pathological pathways, including Aβ and tau pathology, neuroinflammation, oxidative stress, and neurotransmitter dysfunction [8,9]. The combination of curcumin and donepezil represents a particularly rational pairing that addresses complementary mechanisms: while donepezil enhances cholinergic neurotransmission through acetylcholinesterase inhibition, curcumin targets upstream amyloidogenic processes through direct Aβ interaction, metal chelation, and anti-inflammatory activity [10–12]. This dual approach potentially addresses both symptomatic relief and disease modification.

To investigate these mechanisms systematically, Drosophila melanogaster offers distinct advantages as a model system. The fruit fly model allows for precise genetic manipulation of human AD-related genes, rapid screening of therapeutic interventions, and detailed analysis of behavioral and molecular endpoints within a tractable timeframe. Importantly, key components of amyloid processing, including APP and BACE1, are conserved between Drosophila and humans, making it an appropriate model for studying amyloidogenic responses to combination therapy [13].

Our previous investigations have demonstrated that curcumin-donepezil combination therapy improves antioxidant, anti-inflammatory, and neurolocomotor endpoints in animal models of dementia [14,15]. Building on these findings, we hypothesized that curcumin-donepezil combination therapy would demonstrate superior efficacy compared to individual treatments by simultaneously targeting multiple AD pathological pathways. Specifically, we aimed to evaluate the survival and memory-enhancing effects of curcumin-donepezil combination versus monotherapies in a transgenic Drosophila AD model, characterize the molecular mechanisms underlying combination therapy effects on BACE-1 activity and oxidative stress pathways, and validate drug-target interactions through computational molecular docking approaches. Through this integrated methodology, we aim to provide mechanistic insights into multi-target AD treatment strategies for curcumin-donepezil combination therapy.

2. Materials and methods

2.1. Chemicals and reagents

Curcumin (≥95% purity, CAS 458–37–7, Cat# C1386) and Donepezil hydrochloride (≥98% purity, CAS 120,011–70–3, Cat# D6821) were purchased from -Aldrich Sigma(St. Louis, MO, USA). Ethanol (200 proof, molecular biology grade, Cat# E7023) was obtained from Sigma-Aldrich. N-Benzoyl-DL-arginine-4-nitroanilide hydrochloride (DL-BAPNA, Cat# B4875), n-n-diethyl-para-phenylenediamine (DEPPD, Cat# D3134), thiobarbituric acid (TBA, Cat# T5500), 5,5’-dithiobis (2-nitrobenzoic acid) (DTNB, Cat# D8130), and pyrogallol (Cat# P0381) were all purchased from Sigma-Aldrich. TrizolTM reagent (Cat# 15596026) was obtained from ThermoFisher Scientific (Waltham, MA, USA). Luna® Universal qPCR Master Mix (Cat# M3003) was purchased from New England Biolabs (Ipswich, MA, USA). All other reagents were of analytical grade. Glass-distilled water was prepared using a Milli-Q water purification system (Millipore, Burlington, MA, USA).

2.2. Drosophila stocks and maintenance

2.2.1. Fly strains

Wild-type D. melanogaster (Oregon-R strain, Stock# 5) was obtained from the Bloomington Drosophila Stock Center (Indiana University, Bloomington, IN, USA). The transgenic fly lines used were UAS-APP695-N-myc, UAS-BACE1, and elav-Gal4 driver line P{GawB}elavC155.

2.2.2. Fly culture conditions

All fly stocks were maintained at 25 ± 1°C with 65 ± 5% relative humidity on a 12:12 hour light: dark cycle in a temperature-controlled incubator. Flies were cultured on standard cornmeal-yeast-agar medium containing:

• Cornmeal: 50 g/L

• Yeast extract: 25 g/L

• Agar: 10 g/L

• Sucrose: 50 g/L

• Propionic acid: 3 mL/L (as mold inhibitor)

• Tegosept (methyl paraben): 2 g/L dissolved in ethanol [15].

2.2.3. Generation of Alzheimer’s disease model flies

To generate flies expressing human amyloid precursor protein (APP) and β-secretase (BACE1) in neurons, virgin elav-Gal4 females were crossed with UAS-APP695-N-myc, UAS-BACE1 males at 25°C. F1 progeny were identified by the presence of both transgenes through PCR genotyping and phenotypic markers including crumpled wings and melanotic masses on the proboscis. Only F1 flies confirmed to carry both transgenes were used in experiments.

2.2.4. Genotype confirmation

Genomic DNA was extracted from individual flies using standard protocols. PCR amplification was performed using the following primers:

• elav-Gal4: Forward 5’-CGACGCAAACTTCGACACT-3,’

  Reverse 5’-GTTCGAATCGTATCGGCAATAG-3’

• UAS-APP: Forward 5’-GAAGTTCATCTGCACCACCG-3,’

  Reverse 5’-GGTCTTGTAGTTGCCGTCGTC-3’

• UAS-BACE1: Forward 5’-ACGGACATGAAGCCTAACATAG-3,’

  Reverse 5’-TTGGAGATGGTGACCTGGTAG-3’

PCR conditions: Initial denaturation at 95°C for 3 min, followed by 35 cycles of 95°C for 30 s, 58°C for 30 s, 72°C for 1 min, and final extension at 72°C for 10 min [15].

2.3. Experimental design and treatment groups

2.3.1. Sample size determination

Sample size was calculated using G*Power 3.1.9.7 software assuming α = 0.05, β = 0.20 (power = 80%), and effect size d = 1.2 based on our pilot studies. This yielded a minimum of n = 15 per group. n = 20 per group was used to account for potential mortality during the treatment period.

2.3.2. Randomization and group assignment

Three to five-day-old flies of both sexes were collected under light CO₂ anesthesia and allowed to recover for 2 hours. Flies were then randomly assigned to treatment groups using a random number generator (random.org). Group assignments were made by a researcher not involved in data collection to ensure blinding.

2.3.3. Treatment groups

Seven experimental groups were established (n = 20 each):

1. Control (elav-Gal4): elav-Gal4 driver flies + standard diet

2. AD Model (Aβ): elav-Gal4/UAS-APP-BACE flies + standard diet

3. Curcumin Low: AD model flies + diet containing 0.5 mg/g curcumin

4. Donepezil Low: AD model flies + diet containing 12.5 μg/mL donepezil

5. Donepezil High: AD model flies + diet containing 25 μg/mL donepezil

6. Combination Low: AD model flies +0.5 mg/g curcumin +12.5 μg/mL donepezil

7. Combination High: AD model flies +0.5 mg/g curcumin +25 μg/mL donepezil

2.3.4. Drug preparation and administration

Curcumin Preparation: Curcumin was dissolved in absolute ethanol to create a 50 mg/mL stock solution. This was then diluted in melted food medium to achieve a final concentration of 0.5 mg/g food (equivalent to approximately 0.05% w/w). The final ethanol concentration in food was < 0.1%.

Donepezil Preparation: Donepezil hydrochloride was dissolved directly in melted food medium to achieve final concentrations of 12.5 μg/mL and 25 μg/mL.

Vehicle Control: Control foods contained equivalent volumes of ethanol (≤0.1%) without active compounds.

2.3.5. Dose rationale

• Curcumin (0.5 mg/g): Based on a previous dose-response study showing optimal neuroprotective effects without toxicity at this concentration [16]

• Donepezil (25 μg/mL): Established effective dose demonstrating significant cognitive improvement [17]

• Donepezil (12.5 μg/mL): Sub-optimal dose selected to test synergistic potential with curcumin.

2.3.6. Treatment protocol

Flies were transferred to vials containing drug-supplemented food immediately after group assignment. Treatment duration was 7 days with continuous drug exposure. Flies were transfered to fresh treatment-containing food medium every 48 hours to maintain drug stability. Flies were maintained at 25 ± 1°C throughout the treatment period.

2.3.7. Mortality monitoring

Fly mortality was recorded daily at approximately 10:00 AM by a blinded observer. Dead flies were removed immediately to prevent contamination. Survival curves were constructed and analyzed using the Kaplan-Meier method.

2.4. Behavioral assessments

2.4.1. Negative geotaxis (climbing) assay

Timing: Performed on day 7 post-treatment initiation, 2–4 hours into the light cycle.

Procedure:

1. Surviving flies were transferred to empty plastic vials (11 cm length × 3.5 cm diameter) without anesthesia

2. Flies were allowed to acclimatize for 10 minutes at room temperature

3. Vials were tapped firmly on the bench to knock flies to the bottom

4. The number of flies climbing above a 6-cm mark within 15 seconds was recorded

5. Three trials were performed with 1-minute intervals between trials

6. Climbing index was calculated as: (number of flies above 6 cm/total flies) × 100

Controls: Parallel testing of all groups under identical conditions by a blinded observer.

2.4.2. Olfactory associative learning and memory

Training Phase (Day 7):

1. Flies were placed in a T-maze apparatus with two choice arms

2. Training period: 4 minutes alternating exposure to:

   • 3-octanol (OCT, 1:100 dilution in mineral oil) paired with 75 V electric shock (12 pulses/minute)
   • Methylcyclohexanol (MCH, 1:25 dilution) without shock

3. Shock delivery: 1.5-second pulses at 5-second intervals using a custom shock generator.

Memory Testing (10 minutes post-training):

1. Flies were placed at the T-maze choice point with both odors present (no shock)

2. Distribution between OCT and MCH arms was recorded after 2 minutes

3. Performance Index (PI) = [(flies in MCH arm – flies in OCT arm)/total flies] × 100

4. PI values range from −100 (complete attraction to shock-paired odor) to + 100 (complete avoidance)

Controls: Reciprocal training (MCH paired with shock, OCT unpaired) performed on separate cohorts to control for odor preferences [15].

2.5. Biochemical analyses

2.5.1. Sample preparation

Tissue Collection: On day 8, flies were anesthetized with CO2, and heads were separated on ice using fine forceps. Twenty heads per sample were collected for biochemical analysis.

Homogenization: Fly heads were homogenized in ice-cold potassium phosphate buffer (0.1 M, pH 7.4) using a motorized pestle (Kontes, Vineland, NJ) at 4°C. The homogenization buffer consisted of 0.1 M KH₂PO₄, 0.1 M K₂HPO₄, and a protease inhibitor cocktail (1:100, Sigma P8340). The pH was adjusted to 7.4 with NaOH.

Protein Quantification: Total protein was determined using the Bradford assay (Bio-Rad Protein Assay Kit) with bovine serum albumin as standard.

2.5.2. β-secretase (BACE-1) activity assay

Principle: Measurement of BACE-1 enzymatic activity using the synthetic substrate DL-BAPNA.

Procedure:

1. Reaction mixture (200 μL total):

   • 50 μL tissue homogenate (1:2 dilution)

   • 50 μL sodium acetate buffer (50 mM, pH 4.5)

   • 100 μL DL-BAPNA substrate (6 mM)

2. Incubation: 65 minutes at 37°C in water bath

3. Absorbance measurement: 405 nm using spectrophotometer (Thermo Fisher Evolution 201)

4. Activity calculation: ΔA405/hour normalized to protein content

Controls:

• Blank: Buffer without tissue homogenate

• Negative control: Boiled homogenate (100°C, 10 min)

• Positive control: Commercial BACE-1 enzyme (Sigma)

2.5.3. Reactive oxygen species (ROS) assay

Principle: DEPPD-based colorimetric assay measuring hydrogen peroxide equivalents.

Reagent Preparation:

• DEPPD solution: 6 mg/mL DEPPD + 4.37 μM FeSO4 in 0.1 M sodium acetate (pH 4.8)

• Prepared fresh daily and stored on ice

Procedure:

1. Reaction mixture (245 μL total):

   • 140 μL sodium acetate buffer
   • 5 μL tissue homogenate
   • 100 μL DEPPD reagent

2. Incubation: 5 minutes at 37°C

3. Absorbance measurement: 505 nm

4. Quantification using H₂O₂ standard curve (0–100 μM)

Quality Control: Standard curve R² > 0.99 required for acceptance.

2.5.4. Lipid peroxidation (TBARS) assay

Principle: Measurement of malondialdehyde (MDA) as thiobarbituric acid reactive substances.

Procedure:

1. Reaction mixture (0.7 mL total):

   • 0.05 mL tissue homogenate

   • 0.15 mL SDS (8.1% w/v)

   • 0.25 mL acetic acid-HCl solution (pH 3.4)

   • 0.25 mL thiobarbituric acid (0.8% w/v)

2. Incubation: 100°C for 60 minutes in boiling water bath

3. Cooling: Ice bath for 5 minutes

4. Centrifugation: 3000×g for 10 minutes

5. Absorbance measurement: 532 nm

6. MDA calculation using extinction coefficient (ε = 1.56 × 10⁵ M⁻¹ cm⁻¹) [15]

2.5.5. Total thiol content

Procedure:

1. Reaction mixture (270 μL total):

   • 200 μL potassium phosphate buffer (0.1 M, pH 7.4)

   • 25 μL tissue homogenate

   • 35 μL distilled water

   • 10 μL DTNB (10 mM in phosphate buffer)

2. Incubation: 30 minutes at room temperature in dark

3. Absorbance measurement: 412 nm

4. Quantification using cysteine standard curve

2.5.6. Antioxidant enzyme activities

Superoxide Dismutase (SOD):

1. Reaction mixture (7.5 mL total):

   • 50 μL sample

   • 650 μL Tris-HCl buffer (50 mM, pH 8.2)

   • 50 μL pyrogallol (60 mM, freshly prepared)

2. Absorbance monitoring: 420 nm for 3 minutes (30-second intervals) Activity calculation:

3. Units/mg protein (1 unit = 50% inhibition of pyrogallol auto-oxidation)

Catalase:

1. Reaction mixture (0.8 mL total):

   • 0.05 mL tissue homogenate

   • 0.25 mL phosphate buffer (0.01 M, pH 7.0)

   • mL H2O2 (2 M)

2. 2. Incubation: 1 minute at room temperature

3. 3. Reaction termination: 0.4 mL dichromate-acetic acid reagent

4. 4. Absorbance measurement: 620 nm

5. 5. Activity calculation: μmol H2O2 consumed/min/mg protein [15].

2.6. Gene expression analysis

2.6.1. RNA extraction

Total RNA was extracted from fly heads using TrizolTM reagent following manufacturer’s protocol:

1. Homogenization: 20 fly heads in 500 μL Trizol

2. Phase separation: Addition of 100 μL chloroform

3. RNA precipitation: Isopropanol precipitation

4. Washing: 75% ethanol washes

5. Resuspension: RNase-free water

Quality Control: RNA integrity was assessed using 1.5% agarose gel electrophoresis. RNA concentration and purity were determined using NanoDrop 2000 (Thermo Scientific). Only samples with A260/A280 ratio > 1.8 were used.

2.6.2. cDNA synthesis

First-strand cDNA was synthesized using SolisScript Reverse Transcriptase (Solis BioDyne):

• Template:1 μg total RNA

• Primers: Random hexamers (50 ng/μL; Table 1)

• Incubation: 42°C for 60 minutes, 70°C for 15 minutes

• Storage: −20°C until qPCR analysis

Table 1.

Sequence of RT-qPCR primers.

Gene	Full Name	Primer Sequence (5′–3′)	Ascension number
hBACE-1	Human Beta Secretase	Left: TGAGAAAGGAGCCTGCCAAG Right: CACCCACTTCACCCTGTCTC	NG_029372.1
hAPP	Human Amyloid Precursor Protein	Left: GCCCTTCTCGTTCCTGACAA Right: ACTGCATGTCTCTTTGGCGA	NM_000484.4
dmActin	D. melanogaster Actin	Left: TACCCCATTGAGCACGGTAT Right: GGGTCATCTTCTCACGGTTG	FBgn0286832
dmcnc/Nrf2	D. melanogaster Nuclear factor erythroid 2–related factor 2	Left: CGGCGACAAATATCATCCTT Right: CCAACTTCCTCAAGGAGCAG	FBgn0262975
dmGPDH	D. melanogaster Glyceraldehyde 3 phosphate dehydrogenase	Left: GCTCCTCAATGGTTTTTCCA Right: ATGGAGATGATTCGCTTCGT	FBgn0001128

2.6.3. Quantitative real-time PCR (qPCR)

Equipment: StepOnePlus Real-Time PCR System (Applied Biosystems)

Reaction Conditions:

• Total volume: 20 μL per reaction

• Template: 2.5 ng/μL cDNA

• Master mix: 10 μL Luna Universal qPCR Master Mix

• Primers: 0.25 μM each (forward and reverse)

Thermal Cycling:

• Initial denaturation: 95°C for 3 minutes

• Amplification: 40 cycles of:

  • Denaturation: 95°C for 15 seconds
  • Annealing: 60°C for 10 seconds
  • Extension: 72°C for 30 seconds

• Melt curve analysis: 95°C for 10 seconds, 55°C for 1 minute, 95°C for 15 seconds

Data Analysis:

• Relative gene expression calculated using 2^(-ΔΔCt) method

• Normalization: Geometric mean of GAPDH and β-Actin

• Technical replicates: Triplicate reactions per sample

• Biological replicates: n = 6 per group

Quality Control:

• Primer efficiency: 90–110% (determined by standard curve)

• Melt curve analysis: Single peak confirmation

• No-template controls: Included in each run

• Genomic DNA contamination: Checked with no-RT controls [15]

2.7. Molecular docking studies

In order to study further the interactions between the two compounds and selected protein targets implicated in Alzheimer’s diseases, a molecular docking study was undertaken.

2.7.1. Compound and protein data sets

The 3D structures of the selected compounds were downloaded in a structure data file (SDF) format from a chemical repository linked to the National Centre for Biotechnological Information (NCBI) called PubChem (PubChem (nih.gov)). Donepezil’s main therapeutic use is in the treatment of Alzheimer’s disease, where it is used to increase cortical acetylcholine. β-Secretase of Humans with PDB ID: 4LXM, Acetylcholinesterase with PDB ID: 4M0E, Butyrylcholinesterase with PDB ID: 5DYW, β-Amyloid with PDB ID: 7Q4B, and Drosophila melanogaster Acetylcholinesterase with PDB ID: 6XYU was used for the virtual screening.

2.7.2. Preparation and protein target selection

The three-dimensional structure 3D of β-Secretase of Humans (PDB ID: 4LXM) with resolution of 2.30 Å, Acetylcholinesterase (PDB ID: 4M0E) with resolution of 2.00 Å, Butyrylcholinesterase (PDB ID: 5DYW) with resolution of 2.50 Å, Β-Amyloid (PDB ID: 7Q4B) with resolution of 2.50 Å, and Drosophila melanogaster Acetylcholinesterase with (PDB ID: 6XYU) with resolution of 2.51 Å were selected from the Research Collaboratory for Structural Biology (RCSB)’s Protein Database (PDB) (rcsb.org). The number of chains available was checked, and the chain with the preferred target was left while the unwanted chains were deleted. Water molecules, ions, and ligands were removed to make the protein nascent. Energy minimization was performed via the protein preparation and minimization tools using Chimera© software, version 1.13.1 (https://www.cgl.ucsf.edu/chimera/).

2.7.3. Molecular docking of compounds against target proteins

The 3D of the selected compounds sourced from PubChem was facilitated by the Open Babel tool integrated within Python Prescription (version 0.8) and converted into their most energetically favorable and stable conformations using the Merck Molecular Force Field (MMFF94) [18]. The molecular docking was done using AutoDock Vina in open-source Python Prescription (PyRx) 0.8 software [19,20] at neutral pH to acquire possible orientations and binding energy (BE) of compounds at the binding site of 4LXM, 4M0E, 5DYW,7Q4B, and 6XYU. A target region for the 4LXM was attuned with the aid of the grid box with dimensions (19.6024 ×19.1107 ×17.1585) Å, and the center was adjusted based on the site of binding in 4LXM, comprising ASP228, ASP32, GLN73, THR72, THR232, and GLN12. For the protein 4M0E, the attuned target region with the grid box displayed dimensions of (9.2752 ×16.3662 ×14.8352) Å on the site of binding in 4M0E, comprising TRP286, TYR72, and TYR341. The protein 5DYW was attuned at a target region with the aid of the grid box with dimensions (21.2880 ×21.2821 ×21.2281) Å, and the center was adjusted based on the site of binding in 5DYW, comprising VAL288, LEU286, TRP231, TRP82, ASP70, SER198, HIS438, GLU325. For the protein 7Q4B, it attuned the target region with the aid of the grid box displayed dimensions of (52.9612 ×26.8418 ×18.1068) Å. The center was adjusted based on the site of binding in 7Q4B, comprising GLY9, ALA42, HIS14, GLU22, and VAL12. For the protein 6XYU, A target region for the 6XYU was attuned with the aid of the grid box with dimensions (19.6804 ×24.5964 ×20.2352) Å, and the center was adjusted based on the site of binding in 6XYU, comprising TRP321, TRP83, GLY150, GLY151, HIS480, SER238, TRP271, GLU367, and ALA239. After the completion of docking, the compounds were subjected to molecular visualization to analyze their molecular interaction fingerprints through PyMOL© Molecular Graphics (version 2.4, 2016, Schrödinger LLC) and Discovery Studio (version 16.1.0.15, Dassault Systèmes) [19]. The analysis for each protein-ligand complex examined specific interactions, including hydrogen bonds, hydrophobic interactions, and electrostatic interactions, between the amino acid residues of the protein and the atoms of the ligand, as shown in Figure 5(A–E).

Figure 5.

2D visualization of interaction (Ai). Curcumin and BACE-1 (4LXM) (Aii). Donepezil and β-Secretase (4LXM) (Bi). Curcumin and acetylcholinesterase (4M0E) (Bii). Donepezil and acetylcholinesterase (4M0E) (Ci). Curcumin and butyrylcholinesterase (5DYW) (Cii). Donepezil and butyrylcholinesterase (5DYW) (Di). Curcumin and β-amyloid (7Q4B) (Dii). Donepezil and β-amyloid (7Q4B) (Ei). Curcumin and D. melanogaster acetylcholinesterase (6XYU) and (Eii). Donepezil and D. melanogaster acetylcholinesterase (6XYU).

2.7.4. ADMET prediction in silico

ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) is important at the early stage of drug discovery and in the design of pipelines to analyze the pharmacokinetics and pharmacodynamics of the proposed compounds that could serve as a drug. The SwissADME server was used to forecast the ADMET properties of the compounds with the best hits after molecular docking analysis [21]. The SMILE Strings of the compounds from PubChem (https://pubchem.ncbi.nlm.nih.gov/compound) were supplied on the server through its search bar and ran for the prediction of ADMET properties.

2.8. Statistical analysis

2.8.1 Software and Data Management

All statistical analyses were performed using GraphPad Prism version 8.2.1 (GraphPad Software, San Diego, CA). Shapiro-Wilk test was used to test for normality of the data distribution, while the ROUT method (Q = 1%) was used for outlier detection.

2.8.2. Statistical tests

The study analyzed the interactions among the different groups of samples using a two-way ANOVA and conducted a Bonferroni post-hoc comparison with Tukey’s multiple comparisons test. Results were presented as mean ± standard deviation (SD), unless otherwise specified. Statistical significance was set at p < 0.05.

2.8.3. Blinding and quality control.

• Behavioral assessments: Performed by investigators blinded to treatment groups

• Biochemical analyses: Sample coding ensured blinded analysis

• Data analysis: Group identities revealed only after the completion of statistical analysis completion

• Independent verification: Key analyses verified by a second investigator

2.9. Ethical considerations

All experimental procedures were approved by the Institutional Animal Care and Use Committee of the Federal University of Technology, Akure (Approval# 2023–0156). Drosophila maintenance and experimental procedures followed standard ethical guidelines for invertebrate research.

3. Results

3.1. Effect of curcumin-donepezil combination therapy on survival and neurolocomotor function in hAPP and BACE-1 expressing Drosophila melanogaster

The survival rate, locomotor Performance, and memory index of hAPP and BACE-1 expressing flies is shown in Figure 1. “In Aβ-expressing flies, both survival rate and memory performance were significantly lower compared to controls. Treatment with either curcumin or donepezil alone, or their combination, rescued these impaired phenotypes. However, hAPP and BACE-1 expression did not significantly impact locomotor performance.

Figure 1.

Effect of curcumin-donepezil combination therapy on (A). Survival rate (B). Locomotor performance (C). Memory index in hAPP and BACE-1 expressing Drosophila melanogaster. Values represent mean ±SD. * values are significantly different compared to control; # values are significantly different compared to transgenic Aβ flies; Ψ values are significantly different compared to curcumin alone; Φ values are significantly different compared to donepezil alone. Abbreviations: Aβ flies – amyloid-beta flies, cur – curcumin, donep – donepezil.

3.2. Effect of curcumin-donepezil combination therapy on BACE-1 and hAPP activity markers in BACE-1 and hAPP expressing Drosophila melanogaster

We confirmed elevated BACE-1 and hAPP levels in Aβ-expressing BACE-1 showed increased enzymatic activity (Figure 2(A)) and higher relative expression levels (Figure 2(B)), while hAPP demonstrated increased relative expression (Figure 2(C)). Treatment with curcumin or donepezil, either individually or in combination, significantly reduced both BACE-1 and hAPP expression at all tested doses.

Figure 2.

Effect of curcumin-donepezil combination therapy on (A). BACE-1 activity (B). Relative mRNA levels of BACE-1 (C). Relative mRNA levels of hAPP in hAPP and BACE-1 expressing Drosophila melanogaster. Values represent mean ±SD. * values are significantly different compared to control; # values are significantly different compared to transgenic Aβ flies; Ψ values are significantly different compared to curcumin alone; Φ values are significantly different compared to donepezil alone. Abbreviations: Aβ flies – amyloid-beta flies, cur – curcumin, donep – donepezil.

3.3. Effect of curcumin-donepezil combination therapy on oxidative stress and antioxidant activity markers in hAPP and BACE-1 expressing Drosophila melanogaster

Oxidative stress markers were significantly altered in Aβ-expressing flies. ROS and TBARS levels were markedly elevated in Aβ flies compared to controls (Figure 3(A,B)), indicating increased oxidative damage. Treatment with curcumin or donepezil, significantly attenuated these elevated oxidative stress markers. The combination therapy further significantly attenuated the elevated oxidative stress markers.

Figure 3.

Effect of curcumin-donepezil combination therapy on oxidative stress and antioxidant activity markers in hAPP and BACE-1 expressing Drosophila melanogaster. (A). ROS levels (B). TBARS levels. (C). Total thiol (D). SOD activity (E). Catalase activity (F). Relative expression of cnc/Nrf2. Values represent mean ±SD. * values are significantly different compared to control; # values are significantly different compared to transgenic Aβ flies; Ψ values are significantly different compared to curcumin alone; Φ values are significantly different compared to donepezil alone. Abbreviations: Aβ flies – amyloid-beta flies, cur – curcumin, donep – donepezil.

The antioxidant defense system was also substantially impacted in Aβ flies. Total thiol content was significantly decreased (Figure 3(C)), while SOD activity was elevated (Figure 3(D)), likely representing a compensatory response to oxidative stress. Catalase activity was notably reduced in Aβ flies (Figure 3E), indicating compromised antioxidant capacity. Both curcumin and donepezil treatments effectively restored these parameters toward control levels. Specifically, they normalized the elevated SOD activity, enhanced catalase function, and restored thiol content.

Gene expression analysis (Figure 3(F)) demonstrated significant alterations in cnc/Nrf2 mRNA levels in Aβ flies compared to controls. Treatment interventions modulated these expression patterns, with the 25 μg/ml combination therapy showing distinct modulatory effects. This suggests that the therapeutic benefits of curcumin and donepezil may be mediated through both direct antioxidant actions and transcriptional regulation of key oxidative stress response genes

3.4. Molecular docking simulation and ADMET profiling of curcumin and donepezil interaction with selected protein targets implicated AD

Donepezil exhibited high binding affinities with the proteins BACE-1 (4LXM), AChE (4M0E), BChE (5DYW), and β-amyloid (7Q4B) as revealed by its high docking score of −8.7, 7.9, −9.9, and −5.5 kcal/mol, respectively (Table 2; Figure 4). Curcumin produced competing binding affinities to the protein targets in question as revealed by the docking score of −7.3, −7.2, −9.0, and −5.0, respectively (Table 2; Figure 4). In Figure 5, the molecular interactions behind the docking scores observed are summarized. The two compounds were noted to have interactions with Tyr72, Phe108, Lys107, Gln73, Asp228, Gly230, and Tyr71 of BACE-1. In AChE, both compounds were observed to establish interactions with Tyr124, Val294, Tyr341, Asp74, and Ser293. In BChE, interactions were observed with Met437, Thr120, Trp82, Tyr440, and His438. In β-amyloid, Phe19, Phe20, and Ile31 establish interactions with the two compounds. And finally, Tyr71, Tyr374, Asp375, Tyr370, and Tyr73 in DmAChE were found to interact with both compounds. On the other hand, in silico ADMET predictions revealed the potential pharmacological properties and therapeutic applications of both compounds (Table 3). Curcumin exhibits a lower molecular weight when compared to Donepezil (Control drug) and a notable number of heteroatoms that are relatively lower than the control drug. Donepezil showed the highest molecular weight. The lower molecular weight may have a positive influence in terms of distribution, and the variance of the heteroatoms could impact their potential interaction within the biological system [22].

Table 2.

Docking score of selected compounds against human β-Secretase/BACE-1 (4LXM), acetylcholinesterase (4M0E), butyrylcholinesterase (5DYW), β-amyloid (7Q4B), and Drosophila melanogaster acetylcholinesterase (6XYU).

			Binding Energy (Kcal/mol)
S/No.	Compound Name	PubChem ID	BACE-1	AChe	BChe	β-amyloid	dmAChE
1.	Donepezil	3152	−8.7	−7.9	−9.9	−5.5	−9.7
2.	Curcumin	969516	−7.3	−7.2	−9	−5	−10.4

Figure 4.

3D visualization of the interaction between curcumin and (Ai). BACE-1 (4LXM) (Aii). acetylcholinesterase (4M0E) (Aiii). butyrylcholinesterase (5DYW) (Aiv). β-amyloid (7Q4B) and (Av). D. melanogaster acetylcholinesterase (6XYU), as well as the interaction between donepezil and (Bi). 4LXM (Bii). 4M0E (Biii). 5DYW (Biv). 7Q4B (Bv). 6XYU. Key: red – alpha helix, yellow – beta sheet, green – loops, blue – respective ligands.

Figure 5.

(Continued).

Table 3.

ADMET prediction of curcumin and donepezil.

ADMET Descriptors	Curcumin	Donepezil
MW	368.38	379.49
#Heavy atoms	27	28
#Aromatic heavy atoms	12	12
Fraction Csp3	0.14	0.46
#Rotatable bonds	8	6
#H-bond acceptors	6	4
#H-bond donors	2	0
MR	102.8	115.31
TPSA	93.06	38.77
MLOGP	1.47	3.06
Silicos-IT Log P	4.04	4.91
Consensus Log P	3.03	4
ESOL Log S	−3.94	−4.81
Silicos-IT class	Moderately soluble	Poorly soluble
GI absorption	High	High
BBB permeant	No	Yes
Pgp substrate	No	Yes
CYP1A2 inhibitor	No	No
CYP2C19 inhibitor	No	No
CYP2C9 inhibitor	Yes	No
CYP2D6 inhibitor	No	Yes
CYP3A4 inhibitor	Yes	Yes
log Kp (cm/s)	−6.28	−5.58
Lipinski #violations	0	0
Ghose #violations	0	0
Veber #violations	0	0
Egan #violations	0	0
Muegge #violations	0	0
Bioavailability Score	0.55	0.55
PAINS #alerts	0	0
Brenk #alerts	2	0
Leadlikeness #violations	2	2
Synthetic Accessibility	2.97	3.36

4. Discussion

Our findings demonstrate that combining curcumin with donepezil produces measurable therapeutic effects in an Aβ-expressing Drosophila model, with the combination treatment showing improvements in survival, memory performance, and molecular markers compared to vehicle controls. While both compounds individually reduced BACE-1 activity and modulated oxidative stress markers, the combination therapy appeared to provide additional benefits, particularly in restoring antioxidant enzyme balance, where SOD activity was normalized while catalase function was enhanced.

Our findings demonstrate that combining curcumin, a bioactive compound from turmeric, with donepezil, an established acetylcholinesterase inhibitor, produces enhanced therapeutic effects in an Aβ-expressing Drosophila model. The combination rescued AD-related phenotypes, including neuro-behavioral, oxidative stress, and antioxidant responses, alongside BACE-1 and hAPP expression, beyond what was achieved with either compound alone, suggesting potential synergistic effects. These results align with emerging evidence supporting combination approaches in AD treatment, particularly those incorporating functional foods with conventional therapeutics.

Aβ-expressing flies exhibited significantly reduced survival rates and impaired memory function compared to controls, while maintaining normal locomotor activity, which aligns with known AD manifestations of cognitive decline preceding motor dysfunction, as depicted in the seminal paper by Braak & Braak that used brain autopsies to stage the neuropathological changes in AD [23]. Nonetheless, the onset of sensory and motor dysfunction in AD seems to vary across people [24,25]. Importantly, the preservation of locomotor performance across all groups suggests that the recorded memory improvement in the treated groups is not confounded by general motor deficits. The enhanced survival rate in the curcumin and donepezil-treated flies suggests that the therapeutic benefits extend beyond cognitive improvement to overall health and longevity. Our assessment of the amyloidogenic pathway showed marked elevation in BACE-1 enzymatic activity and increased BACE-1 and hAPP mRNA expression in the Aβ flies compared to control, suggesting a feed-forward mechanism where Aβ accumulation may promote further amyloidogenic processing [26,27]. Both curcumin and donepezil treatments effectively reduced BACE-1 activity and expression, alongside hAPP mRNA levels, indicating that both compounds may also influence APP transcription, representing an upstream intervention point in the amyloidogenic cascade [28,29].

Redox homeostasis was dysregulated in the Aβ flies as ROS and TBARS levels were elevated compared with control, in line with the known ability of Aβ to generate ROS via both mitochondrial dysfunction and interaction with metal ions [30,31]. Our findings also support the detrimental feedback loop in which oxidative stress enhances Aβ production by BACE-1 upregulation [32]. We further recorded a compensatory increase in SOD activity that was counteracted by reduced total thiol levels, reduced catalase activity, and critically diminished cnc/Nrf expression, indicating a compromised antioxidant defense in the Aβ flies. Indeed, Aβ and tau accumulation reduces Nrf2 levels, thereby inhibiting antioxidant response and the autophagic turnover of Aβ and tau [33]. Both curcumin and donepezil treatments effectively modulated these oxidative stress markers. Particularly noteworthy was the restoration of the antioxidant enzyme balance, where the combination therapy normalized SOD activity while enhancing catalase function. This coordinated modulation of antioxidant enzymes suggests a more effective restoration of redox homeostasis compared to either treatment alone. Ultimately, AD therapies must factor in mechanisms to target ROS pathways and enhance antioxidant defense systems [34].

Molecular docking analyses revealed potential binding interactions between curcumin, donepezil, and key therapeutic targets in AD. The structural visualization demonstrates that curcumin can interact with AChE, BChE, and BACE-1, suggesting multiple mechanisms of action. The binding poses indicate that curcumin forms stable interactions with the catalytic sites of the enzymes. Similarly, donepezil expectedly bound to AChE and BChE, while also exhibiting potential interactions with BACE-1, which may explain its observed effects on BACE-1 activity in our experimental results. The visualization of these molecular interactions provides mechanistic insight into how these compounds might work together. The interactions that result between the two compounds and the target are hydrogen linkages, Pi-interactions, and van der Waals forces. These interactions give the protein-ligand complex stability and contribute to the binding energy of a compound toward a protein target. The ability of both compounds to interact with multiple targets suggests that their synergistic effects may arise from complementary binding interactions. This is particularly relevant given that our experimental data showed enhanced effects of the combination therapy on both cholinergic and amyloidogenic pathways, probably culminating in the improvement of cognitive outcomes.

The observed modulation of the amyloidogenic pathway by curcumin and donepezil suggests potential upstream intervention points, but whether this represents direct transcriptional effects or secondary consequences of cellular protection remains unclear and warrants further mechanistic investigation. The docking results suggest that donepezil’s superior binding affinity to BACE-1 compared to curcumin may partially explain its effectiveness in reducing BACE-1 activity, though curcumin’s multi-target interactions across different pathways could compensate for lower individual binding affinities. However, the correlation between computational binding scores and actual biological efficacy requires experimental validation, as factors such as bioavailability, cellular uptake, and metabolic stability significantly influence therapeutic outcomes beyond binding affinity predictions [35]. Importantly, although Drosophila is valuable for initial screening, the fly model is simplified and lacks the complex neuroanatomical architecture and cellular diversity of the mammalian brain. The fly model does not recapitulate the full spectrum of AD pathology, and our study notably lacks evaluation of tau pathology, which represents a significant gap given tau’s critical role in AD progression alongside amyloid accumulation. This could also be coupled with possible differences in pharmacokinetic profiles of curcumin and donepezil in Drosophila, compared to mammals. These limitations should be considered for potential translation to clinical applications.

5. Conclusion

In conclusion, our study demonstrates that combining curcumin with donepezil provides enhanced therapeutic benefits in an Aβ-expressing Drosophila model of AD. The combination therapy showed superior efficacy in improving survival and memory performance compared to either compound alone, while effectively modulating multiple pathological features of AD. The superior efficacy of the combination therapy, coupled with curcumin’s established safety profile, suggests this approach may represent a promising therapeutic strategy for AD. Future studies should investigate the translation of these findings to mammalian models and explore optimal dosing strategies for clinical applications

Funding Statement

This research is funded by the International Brain Research Organization (IBRO), [2020 Return Home Fellowship].

Article highlights

• Curcumin-donepezil combination therapy reduces BACE-1 activity and oxidative stress markers in Drosophila AD model.

• Molecular docking confirms favorable drug interactions with enhanced therapeutic potential.

• ADMET prediction supports the pharmacological viability of the combination approach.

• Multi-target strategy addresses both cholinergic deficits and amyloid pathology in Alzheimer’s disease.

Author´s contribution

Ogunsuyi Opeyemi Babatunde: Conceptualization, Funding, Methodology, Investigation, Data collection, Writing- Reviewing, Oluokun Odunayo Oluwakemi: Data collection, Data analysis, Writing- Original draft preparation. Ademiluyi Oluwaseun Adedayo: Conceptualization, Data curation, visualization, Methodology Writing- Reviewing and approving the final draft. Oboh Ganiyu: Conceptualization, Data curation, visualization, Methodology Writing- Reviewing and approving the final draft. Akeeb Rasheedat: Data collection, Data analysis, Writing- Original draft preparation. Haruna Isiyaku Umar: Data collection, Data analysis, Writing- Original draft preparation. Olawande Chinedu Olagoke: Methodology, Visualization, Writing- Original draft preparation, Writing- Reviewing and Editing. All authors agree to be accountable for all aspects of work, ensuring integrity and accuracy.

Disclosure statement

The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Ethical statement

No ethical approval applies to this research.

Availability of data and material

The data that support the findings of this study are available from the corresponding author, upon reasonable request.