D-galactose (purity ≥99%) and donepezil hydrochloride (purity ≥98%) were purchased from Aladdin biochemical technology Inc. (Shanghai, China). Dihydroquercetin (purity ≥98%) was purchased from Xi’an Benfeng Biotechnology Co., Ltd (Xi’an China). RIPA buffer was obtained from NCM biotech Co., Ltd. (Suzhou, China). ELISA kits for IL-6, IL-1β and TNFα were supplied by Boster biological technology Co., Ltd. (Nanjing, China). The primary antibodies for TLR4, NF-κB p65, IL-1β, CASP1, NLRP3, Bax, and Bcl2 were purchased from Proteintech group, Inc. (Wuhan, China). The primary antibodies for phospho-NF-κB p65 was obtained from Affinity bioscience Co., Ltd. (Jiangsu, China). The primary antibodies for CASP3, beta-actin, and horseradish peroxidase (HRP)-conjugated affinity pure goat anti-rabbit IgG were purchased from Boster biological technology Co., Ltd. (Nanjing, China).

Male Sprague-Dawley rats (180–220 g, 8 weeks old) were obtained from Zhuhai Baishitong Biotechnology Co. Ltd. (No.SCXK(Yue)2020-0051). The animals were cultivated under specific pathogen-free conditions as follows: 25 °C ± 3 °C temperature, 55% ± 10% relative humidity, a 12-h light/dark cycle, and free access to food and water.

After 1 week acclimation, SD rats were randomly divided into the four groups, with six rats per group: (i) control group (CON, 0.9% saline i.p.+0.9% saline, i.g.), (ii) model group (MOD, D-galactose 250 mg/kg i.p.+0.9% saline, i.g.), (iii) Dihydroquercetin group (DHQ, D-galactose 250 mg/kg i.p.+dihydroquercetin 9 mg/kg, i.g.), and (iv) Donepezil group (DON, D-galactose 250 mg/kg i.p.+donepezil 1.35 mg/kg, i.g.). With the exception of the control group, all rats underwent intraperitoneal injection of D-galactose (250 mg/kg) daily to induce ageing rat models, as evidenced by literature (Aydın et al., 2016; Fasina et al., 2022; Ibrahim et al., 2019; Mathimaran et al., 2021), with minor modifications. Meanwhile, the control group subjected to equivalent volume of saline. Subsequent to each modelling procedure, rats in the various treatment groups were given the corresponding drugs orally, while equal amounts of the vehicle were administered to rats in the CON and MOD groups. The dosage of dihydroquercetin was determined on the basis of the recommended daily intake of ≤100 mg/day for adults, as established by China’s National Health Commission (National Health Commission of the People’s Republic of China, 2021). Meanwhile, 15 mg/day of donepezil was designated as the adult dosage. The body surface area (BSA) normalization method was utilised for the conversion of doses across species. Body weight was measured at 2-week intervals for each rat throughout the duration of the treatment. The experiment was conducted over a period of 12 consecutive weeks.

The experimental protocol involving the use of animals was conducted in accordance with the guidelines of the Animal Experiments Committee, and was approved by the Animal Ethics Committee of China Technology Industry Holdings (Shenzhen) Co., Ltd. (approval NO.202300167). All methodologies employed in this study conformed to the established guidelines and regulations.

As stated in the “Clinic terminology of traditional Chinese medical diagnosis and treatment-Part 2: Syndromes/patterns (GB/T 16,751.2-2021)”, in conjunction with the extant literature (Cui et al., 2014; Li D et al., 2025), a quantitative scoring table for rat physical signs was formulated, as illustrated in Table 1. Quantitative scoring was conducted for the physical signs of each group of rats. Each scoring was carried out independently by two trained researchers who were blinded to the group allocations (CON, MOD, DHQ, DON). A physical evaluation of the rats was conducted, and the results were documented. The mean score of the two operators was then taken as the final score for that physical sign.

The MVW test is a research instrument that assesses the spatial orientation capabilities of rats, thereby serving as a fundamental indicator in studies pertaining to learning and memory. The entire MWM test was conducted over 5 days. During the course of the MWM test, rigorous blinding procedures were consistently implemented throughout the data acquisition process, with the tracking data were conducted by an independent researcher who was unaware of the group allocations. Prior to the MWM test, all rats were subjected to an adaptive training regime. Subsequently, all rats were subjected to a directional navigation test spanning across days 1 through 4. Each rat was trained four times daily, entering the water randomly from one of four entry points (illustrated Figure 2A), with a maximum time allowance of 60 s to swim. If a rat successfully reaching a platform within 60 s, escape latency, defined as the time taken from entry to climbing onto the platform, was documented. If the rat was unable to locate the platform, the escape latency was documented as 60 s, after which the rat was navigated towards the platform and remained there for 10 s. The final day of the trial was designated for a spatial probe test. On the fifth day, the platform was removed, and the rats were released into the water from the quadrant opposite the original platform location. The observations encompassed metrics such as the number of crossings over the original platform area, the duration spent in that quadrant, and the trajectories of their swimming.

Following the completion of behavioural testing, all experimental rats were anaesthetised using the inhaled anaesthetic isoflurane. Blood samples were extracted and subjected to a centrifugation process at 3,500 rpm for a duration of 10 min. The resultant serum was subsequently separated and stored at −80 °C for subsequent analysis. The heart, liver, spleen, and thymus were collected and weighed, after which the viscera indexes were calculated. The brain tissue was collected and subjected to a rinsing process with pre-cooled 0.9% saline solution. For pathomorphological analysis, a portion of rat brain tissue samples were fixed within 4% paraformaldehyde for subsequent analysis. Moreover, the remaining part of the rat brain tissue samples were separated into cortex and hippocampus tissues and preserved in liquid nitrogen for subsequent analysis.

The brain tissue was preserved by fixation in 4% paraformaldehyde, then embedded, sectioned, and finally stained with hematoxylin and eosin. The pathological changes to the brain tissue were observed via SQS-40P slide scanner system (Teksqray, Shenzhen, China).

Nissl staining was utilised to quantify neuronal loss. The brain slices were subjected to Nissl staining for 10 min, followed by a thorough rinsing process involving two rinses. Thereafter, the slices were dehydrated using a gradient ethanol series, subsequently transparentised with xylene, and sealed with neutral gum. For quantitative analysis, Nissl-positive neurons in the CA1, CA3, and DG regions of the hippocampus were counted under light microscopy. Neurons that exhibited a discernible nucleus and clearly distinguishable cytoplasmic Nissl bodies were defined as positive neurons. Within each hippocampal subregion on each section, two fields of view were randomly selected. The mean count per field was used for statistical analysis.

The levels of inflammatory factors including TNF-α, IL-1β and IL-6 were detected in the hippocampus of rats according to the instructions of the ELISA kit.

In the present investigation, the total protein content of the hippocampus was extracted using RIPA lysis buffer and subsequently quantitated using a BCA protein assay kit. Samples of equivalent protein quantity underwent separation by SDS-PAGE gel, followed by electrophoresis and subsequent transferring to a PVDF membrane. Subsequent to blocking of the nonspecific binding sites on the membranes with 5% skim milk, the primary antibodies, including TLR4 (1:6000), NF-κB p65 (1:6000), phospho-NF-κB p65 (1:1000), IL-1β (1:1000), CASP1 (1:6000), NLRP3 (1:1000), BAX (1:6000), BCL2 (1:3000), and CASP3 (1:6000), were added and incubated at 4 °C overnight. Thereafter, the membranes were subjected to three washes with TBST, and then incubated with HRP-conjugated secondary antibodies at room temperature for 1 h. Finally, the protein bands were detected via and ECL method, and subsequent analysis via ImageJ software.

The data were expressed as the mean value ± standard deviation (S.D.). Repeated-measures analysis of variance (RM-ANOVA) was utilised to evaluate the escape latency data from day 1 to day 4 in the directional navigation test. Other differences between the groups were assessed with one-way analysis of variance (ANOVA). The probability value of p < 0.05 was utilised to ascertain the statistical significance of the results. Statistical significance was set at ***p < 0.001, **p < 0.01, *p < 0.05 vs. CON group; ###p < 0.001, ##p < 0.01, #p < 0.05 vs. MOD group.

A molecular docking investigation was conducted to examine the interaction between dihydroquercetin and hub targets in apoptosis and TLR4/NF-κB/NLRP3/CASP1 signaling pathways, including BAX, BCL2, CASP3, TLR4, IKKA, IKKB, NFKB1, NLRP3, ASC, CASP1, and IL-1β. The chemical structure of dihydroquercetin was obtained from the TCMSP database (https://old.tcmsp-e.com/tcmsp.php), and saved as a ligand in mol2 format. The 3D crystal structures of aforementioned targets, retrieved from the Protein Data Bank database (http://www.rcsb.org/pdb), underwent a series of preprocessing, optimization, and minimization procedures, ultimately being stored as receptors. Schrödinger Maestro software suite (version 11.1, Schrödinger, L.L.C.) was utilised for the molecular docking analysis. A comprehensive evaluation of the docking score, binding energy and the creation of hydrogen bonds in the ligand-receptor complex was conducted to determine the definitive stable conformation.

As demonstrated by visual observations, rats in the model group exhibited pronounced aging phenotypes, including typical symptoms such as dry and sparse hair, reduced hair density, impaired skin elasticity, mental dullness, prolonged immobility, sluggish response, and decreased locomotor activity. As the modelling period progressed, there was a marked exacerbation of the symptoms related to the ageing process. Subsequent to week 8, the quantitative physical sign scores of the model group were found to be significantly higher than those of the control group. Conversely, rats administered dihydroquercetin exhibited a substantial alleviation of age-related symptoms (Figure 1B). It is evident that donepezil was less efficacious than dihydroquercetin in ameliorating ageing phenotypes. Throughout the experiment, no animals perished as a consequence of treatment with D-galactose, dihydroquercetin and donepezil.

As demonstrated in Figure 1C, the body weight of the rats in each group exhibited a consistent increase over time. However, rats in the donepezil group exhibited a slower rate of weight gain than the model and control groups. No statistically significant disparities were observed in body weight among the control, model, and dihydroquercetin cohorts.

As illustrated in Figure 1D, the induction of D-galactose resulted in a significant decrease in heart, liver, spleen, and thymus indexes, in comparison to the control group. Supplementation with dihydroquercetin elicited a substantial elevation.

To further investigate the effects of dihydroquercetin on the spatial learning and memory function, the MWM tests were performed. As demonstrated in Figure 2A, in the initial phase of the directional navigation test, no statistically significant variations in escape latency were observed among the entire cohort of rats. As the training days progressed, a gradual decline in the escape latency of the rats was observed. Notably, the rats in the model group exhibited an increased escape latency in comparison with the control group. This finding suggests that the spatial learning ability of rats treated with D-galactose was reduced. In contrast, comparatively reduced escape latency was observed in rats in dihydroquercetin group as well as in donepezil group (Figure 2C). In the ensuing spatial probe trial conducted on the fifth day, the rats within the model group exhibited a general decline in performance, as evidenced by a decrease in the number of platform crossings, a reduction in the time spent in the target quadrant, and a decrease in the total distance travelled in that quadrant, when compared to the control group. Furthermore, the aforementioned behaviour exhibited a marked enhancement following treatment with dihydroquercetin, with a substantial escalation in the number of platform crossings (6.58-folds), the time spent in the target quadrant (2.25-folds), and the total distance travelled in the target quadrant (2.14-folds), respectively (Figures 2B,D–F).

The examination of histopathological changes was conducted using HE staining, with a focus on the hippocampus, a region susceptible to damage from early cognitive impairment. The HE staining revealed that the neurons in the hippocampal region of the control group exhibited a well-arranged structure, with normal-appearing nuclei and abundant cytoplasm. By contrast, the model group demonstrated disordered cell arrangement and a variable size of neural cells, accompanied by a marked reduction in the number of neurons. Supplementation with dihydroquercetin has been demonstrated to alleviate the neuronal histological damage in the hippocampus caused by D-galactose, and to reduce neuronal loss (Figure 3).

Nissl staining (Figure 4A) served to further validate these observations. The neurons in the control group exhibited normal characteristics with regular, densely stained Nissl bodies. Conversely, D-galactose-induced neuronal atrophy in the model rat manifested as a decline in Nissl bodies and disruption of nuclear integrity. Furthermore, a substantial decrease in the number of Nissl-positive neurons was detected in the hippocampus' CA1, CA3, and DG regions in model rats, in comparison to the control group. It is worthy of note that, supplementation with dihydroquercetin treatment induced a substantial increase in the number of neurons. The quantification of these results is demonstrated in Figures 4B–D.

The release of neuroinflammatory cytokines is a salient feature in the development of cognitive impairment. As illustrated in Figure 5, D-galactose intervention led to a substantial increase in the levels of IL-6, IL-1β, and TNF-α in the model rat brain. Supplementation with dihydroquercetin induced a significant reduction the secretion of IL-6, IL-1β, and TNF-α in the brain, with a decrease of 34.37%, 29.44%, and 24.70%, respectively.

To further investigate the protective effect of dihydroquercetin against neuronal apoptosis caused by D-galactose, the expressions of important markers, including anti-apoptotic marker (BCL2) and pro-apoptotic markers (BAX, CASP3), were determined by Western blot analysis. As illustrated in Figures 6A–C, D-galactose administration resulted in a substantial augmentation in BAX and CASP3, concomitant with a conspicuous diminution in BCL2 within the hippocampus. In contrast, these changes were evidently alleviated by dihydroquercetin treatment. Specifically, the suppression of BAX, CASP3 expression and the enhancement of BCL2 expression by dihydroquercetin led to a reduction of the BAX/BCL2 ratio. The results indicate that dihydroquercetin exerts an anti-apoptotic effect on neurons by reducing the BAX/BCL2 ratio and inhibiting CASP3 activity in the rat hippocampus.

In order to provide further confirmation of the underlying mechanism of dihydroquercetin in pyroptosis induced by D-galactose, the expressions of pyroptosis-related proteins were determined by Western blot analysis. As demonstrated in Figures 7A–F, the expressions of TLR4, NF-кB, p-NF-кB, IL-1β, NLRP3, and CASP1, were significantly increased in the hippocampus of model rats induced by D-galactose (vs. CON, p < 0.05). In contrast, supplementation with dihydroquercetin effectively reduce the expressions of these factors (vs. MOD, p < 0.05). Consequently, the aforementioned results verified that dihydroquercetin attenuated D-galactose-induced pyroptosis by suppressing the TLR4/NF-κB/NLRP3/CASP1 signaling pathway.

Critically, as a first-line medication for Alzheimer’s disease in clinical practice, a substantial body of evidence supports that donepezil not only exhibits potent acetylcholinesterase inhibitory activity, but also exerts significant anti-inflammatory and anti-pyroptotic effects (Kim et al., 2021; Youn et al., 2024). The superior efficacy observed in the present study for DHQ in comparison with donepezil suggests that the multi-targeted action of DHQ on both apoptosis and pyroptosis results in a more potent neuroprotective effect.

As demonstrated in Figure 8, the docking scores, binding energy and molecular docking models between dihydroquercetin and hub targets, including TLR4 (PDBID: 2Z63), IKKα (5EBZ), IKKβ (4KIK), NFKB1 (9BOR), NLRP3 (7ALV), ASC (6KI0), CASP1 (6PZP), IL1B (5R89), BAX (8SPF), BCL2 (4AQ3), and CASP3 (3DEH), were illustrated. Dihydroquercetin exhibited a comparatively low binding score and low binding energy during docking with all the test targets, suggesting potential binding affinity. In particular, the docking scores of IKKα-DHQ, IKKβ-DHQ, and ASC-DHQ, exhibited values lower than −7.00, suggesting enhanced stability during the binding process. Observations demonstrated that the benzene ring structures and phenolic hydroxyl groups of the A and B rings of dihydroquercetin, as well as the carbonyl structure of the C ring, are capable of forming a variety of intermolecular forces with the target amino acid residues, encompassing H-bonds, Pi-Pi stacking, and Pi-cation interactions. For instance, the amino acid residues LEU 280 and ILE 285 of TLR4 on chain A were observed to form hydrogen bond interactions with the phenolic hydroxyl group of A and B rings on dihydroquercetin, while the amino acid residues THR 284 hydrogen bonded with the carbonyl of C ring on dihydroquercetin. The amino acid residues TYR 632 of NRLP3 form Pi-Pi stacking interactions with the benzene ring (A ring) on dihydroquercetin. In addition, dihydroquercetin’s benzene ring (B ring) formed Pi-cation with the amino acid residues ARG 179 of CASP1.