Baicalein alleviates Alzheimer’s disease pathology in AD models by activating the PI3K/AKT/GSK3β pathway to inhibit neuronal apoptosis

Abstract

Baicalein, a bioactive flavonoid from Scutellaria baicalensis, exhibits anti - inflammatory, antioxidant, and neuroprotective properties. Its potential for Alzheimer’s disease (AD) therapy is emerging. This study evaluated its therapeutic efficacy and molecular mechanisms using APP/PS1 transgenic mice and SH - SY5Y - APP/PS1 cells. Morris water maze and novel object recognition tests assessed cognitive deficits in mice. Hippocampal pathology and neuronal apoptosis were examined via Nissl staining, transmission electron microscopy, and TUNEL assay. For SH - SY5Y - APP/PS1 cells, CCK − 8 optimized drug concentrations, while Annexin V - FITC/PI measured apoptosis and ELISA analyzed Aβ1–42. Network pharmacology identified PI3K/AKT as a key target. Baicalein improved cognition, reduced hippocampal neurodegeneration, decreased neuronal apoptosis, and lowered Aβ1–42 accumulation. It activated the PI3K/AKT pathway, increasing p - PI3K/p - AKT/BCL2 and decreasing p - GSK3β/Caspase − 3. These findings demonstrate that baicalein ameliorates AD by inhibiting neuronal apoptosis via the PI3K/AKT/GSK3β cascade, validating its potential as an AD therapeutic agent.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-47003-w.

Introduction

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder characterized by an insidious onset and gradual deterioration¹. Its hallmark pathological features include senile plaques (SP) formed by Aβ deposition, neurofibrillary tangles (NFTs) resulting from hyperphosphorylated Tau protein, and progressive neuronal loss². Excessive neuronal apoptosis is closely associated with various neurodegenerative disorders and neural injuries³. In AD, neuronal apoptosis constitutes a primary cause of neuronal loss and is strongly linked to the pathological alterations in Aβ and Tau⁴. Aβ oligomers can activate apoptotic pathways via death receptors or through endoplasmic reticulum and mitochondrial dysfunction, while Tau abnormalities can directly or indirectly trigger the apoptotic cascade⁵. Concurrently, pro-inflammatory cytokines released during neuroinflammation synergistically amplify apoptotic signaling⁶. Furthermore, the suppression of anti-apoptotic pathways, such as PI3K/AKT, heightens neuronal susceptibility to apoptosis. Collectively, these mechanisms drive the progression of AD.

The PI3K/AKT signaling pathway, a crucial intracellular signal transduction pathway, plays a pivotal role in regulating neuronal apoptosis^7,8. Following its activation, phosphatidylinositol-3-kinase (PI3K) generates phosphatidylinositol-3,4,5-trisphosphate (PIP3) as a second messenger, which subsequently recruits and activates protein kinase B (AKT). AKT, a serine/threonine protein kinase, modulates diverse biological processes including cell proliferation, survival, and apoptosis by phosphorylating numerous downstream substrates upon activation. Activation of the PI3K/AKT pathway induces the phosphorylation of Bad, leading to its dissociation from B-cell lymphoma-2 (Bcl-2). The liberated Bcl-2 then exerts its anti-apoptotic effect⁹. Furthermore, the PI3K/AKT pathway can directly phosphorylate and thereby inactivate Caspase-3, inhibiting Caspase-3-mediated apoptosis¹⁰. Glycogen synthase kinase-3β (GSK3β) is a multifunctional kinase whose activation can promote cytoskeletal protein degradation, inhibit cell adhesion, and accelerate apoptosis. AKT activation phosphorylates GSK3β at Ser9, inhibiting its kinase activity and consequently reducing apoptosis¹¹.

Baicalein is a primary active constituent extracted from the dried roots of Scutellaria baicalensis Georgi, a perennial herb belonging to the Lamiaceae family and a prominent medicinal plant in traditional Chinese medicine (TCM). It serves as one of the key material foundations for the pharmacological effects of the TCM herb Huangqin (Scutellariae Radix). Studies indicate that baicalein exhibits neuroprotective effects, and one mechanism involves mitigating neuroinflammation by inhibiting microglial M1 polarization¹². Furthermore, the combined application of baicalein and memantine has been demonstrated to reduce oxidative stress and effectively prevent pathology in β-amyloid-induced Alzheimer’s disease rat models¹³. Additional research suggests that baicalein’s modulation of the gut microbiota may be a critical factor in ameliorating memory and cognitive deficits in APP/PS1 mice¹⁴. However, to date, no studies have specifically elucidated the precise mechanism underlying baicalein’s inhibition of neuronal apoptosis.

This study provides the first elucidation of the neuroprotective and anti-apoptotic effects of Baicalein in both APP/PS1 Alzheimer’s disease (AD) mice and SH-SY5Y-APP/PS1 cells. Data from both in vivo and in vitro experiments demonstrate that Baicalein exerts a significant inhibitory effect on neuronal apoptosis. This effect is likely mediated by Baicalein binding to PI3K and subsequently activating the PI3K/AKT signaling pathway. Activation of this pathway leads to the suppression of pro-apoptotic factors GSK3β and Caspase-3, while promoting the expression of the anti-apoptotic factor Bcl-2.

Materials and methods

ARRIVE guidelines adherence

All in vivo procedures were conducted in full accordance with the Animal Research: Reporting of In Vivo Experiments (ARRIVE) 2.0 guidelines. Rigorous measures—balanced group sizes, randomisation, blinding and appropriate statistical analyses—were implemented to minimise bias and maximise reproducibility. No animals were excluded from the final dataset.

Experimental animal grouping and drug administration

Thirty male 6-month-old APP/PS1 double transgenic mice and ten male C57BL/6 mice were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). Animals were housed under standard laboratory conditions: room temperature of 25 °C, relative humidity of 50–60%, and a 12-hour light-dark cycle. All mice had free access to food and water for a one-week acclimatization period.

Following acclimatization, the 30 APP/PS1 transgenic mice were randomly divided into three groups (n = 10 per group): the Model group (Model), the Positive Control group (Donepezil), and the Baicalein Treatment group (Baicalein). The ten C57BL/6 mice served as the Blank Control group (Control).

The Baicalein group received Baicalein at a dose of 40 mg/kg/day¹⁵. The Donepezil group received Donepezil hydrochloride at a dose of 0.65 mg/kg/day. Both the Model group and the Blank Control group received an equivalent volume of physiological saline via oral gavage. All mice had free access to food and water throughout the dosing period. Treatments were administered once daily by oral gavage for 40 consecutive days.Baicalein Source: China National Institutes for Food and Drug Control (NIFDC); Batch No.: 111,595–202,309; CAS: 491-67-8.Donepezil Source: Eisai (China) Pharmaceutical Co., Ltd.; Product Batch No.: H20050978.

Upon completion of treatment and behavioral testing, mice (31.72 ± 1.98 g body weight) were deeply anesthetized by intraperitoneal injection of 1.25% tribromoethanol (Sigma-Aldrich, T48402) at 0.2 mL per 10 g body weight. After complete loss of corneal and pedal withdrawal reflexes, animals were immediately euthanized by cervical dislocation. Whole brains were rapidly and precisely dissected on a pre-chilled surgical ice plate, and target regions were micro-dissected within 3 min. Tissues intended for pathological examination were processed accordingly, whereas the remainder were snap-frozen in liquid nitrogen and transferred to − 80 °C for long-term storage.

This study was approved by the Experimental Animal Ethics Committee of Hubei University of Chinese Medicine (No.[2024]IEC025). All experimental procedures were conducted in strict accordance with relevant ethical guidelines for animal research.

Cell culture

SH-SY5Y cells(Procell, CL-0208) were cultured in DMEM/F12 medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin. Cells were maintained under standard conditions at 37 °C in a humidified atmosphere containing 5% CO₂.Cell morphology and confluence were monitored daily using an inverted microscope. Upon reaching 80–90% confluence, cells were washed twice with phosphate-buffered saline (PBS) and subsequently detached using 0.25% trypsin-EDTA solution containing 0.1% collagenase for 1–2 min. Digestion was terminated by adding complete culture medium. The cell suspension was then centrifuged at 1000 × g for 5 min. Cells were subcultured at a 1:3 split ratio into fresh culture flasks.For cryopreservation, cells were resuspended in freezing medium containing 10% dimethyl sulfoxide (DMSO), subjected to controlled-rate freezing, and subsequently stored in liquid nitrogen for long-term preservation.

Establishment of the SH-SY5Y-APP/PS1 cell model

To generate a stable SH-SY5Y cell line expressing APP/PS1, genetic modification was performed using a lentiviral transduction system. The APP/PS1 overexpression plasmid (Miaolingbio; pLV3-CMV-PSEN1-T2AAPP-Puro) was co-transfected with the packaging plasmids psPAX2 and pMD2.G into HEK293T cells (Procell, CL-0005) using a lipofectamine-based transfection reagent. Viral supernatants were collected at 48 h and 72 h post-transfection.SH-SY5Y cells were plated in antibiotic-free medium. When cells reached approximately 50% confluence, they were incubated with the viral supernatant supplemented with 8 µg/mL Polybrene (Servicebio, G1803). Following 24 h of transduction, cells were subjected to selection with complete medium containing 1.5 µg/mL puromycin (Servicebio, G4017) for 3 days. Selection continued until complete cell death was observed in non-transduced control cells, yielding a stably transduced cell line with integrated APP/PS1 genes.All procedures were performed under sterile conditions in a biological safety cabinet. Successful protein expression was confirmed by Western blot analysis and quantitative real-time PCR (qPCR).

Morris water maze test

Spatial learning and memory in all experimental groups were assessed using a Morris water maze system (Anhui Zhenghua Biological Instrument Equipment Co., Ltd., Morris model). Following completion of the gavage administration period, mice underwent five consecutive days of place navigation training with two daily trials. During training sessions, escape latency (time to locate the hidden platform), average swimming speed, total path length, and swimming trajectories were recorded. After a 24-hour rest period, a spatial probe test was conducted on Day 7 with the platform removed, during which the number of crossings through the former platform location and swimming trajectories were recorded. All testing was performed under standardized environmental conditions.

Novel object recognition test

Prior to testing, mouse cages were transferred to the experimental room 30 min in advance for environmental and experimenter habituation. On Day 1, during the familiarization phase, two identical black cylinders (Objects A and B) were placed in the arena. Each mouse was introduced facing the lower left corner and allowed to freely explore for 5 min per session; this procedure was repeated three times with thorough cleaning of the arena using 75% ethanol between trials to eliminate olfactory cues. After a 24-hour interval, mice underwent the testing phase: they explored the arena containing Objects A and B for 5 min while exploration time, frequency, and distance toward each object were recorded via video tracking. One hour later, Object B was replaced with a novel red cube (Object C), and mice were reintroduced for 5 min with identical parameters recorded. The recognition index (RI) was calculated as: RI = [Time exploring novel object/(Time exploring novel object + Time exploring familiar object)] × 100%. All procedures were conducted under standardized conditions.

Nissl staining

Hippocampal tissues were fixed in formaldehyde solution, embedded in paraffin, and sectioned. Tissue sections underwent deparaffinization and rehydration through sequential immersion in xylene and a graded ethanol series. After rinsing with distilled water, slides were stained in toluidine blue solution for 40 min at 50–60 °C. Sections were then rinsed with distilled water, dehydrated through immersion in 75% ethanol (3 min), 95% ethanol (5 min), and absolute ethanol (3 min), followed by clearing in xylene (two changes, 5 min each). Finally, representative sections were examined and imaged under a light microscope.

Transmission electron microscopy (TEM)

Following anesthesia, mice were euthanized by cervical dislocation. Hippocampal tissues were rapidly dissected on an ice-cold surface and immediately immersed in glutaraldehyde-based primary fixative (Servicebio, Cat# G1102) at room temperature for 2 h. Specimens were subsequently transferred to 4 °C for storage until further processing. Ultrastructural analysis of hippocampal neurons, including detailed examination of nuclear morphology, mitochondrial integrity, and synaptic structures, was performed using a HITACHI HT7800 transmission electron microscope (Hitachi High-Technologies Corporation, Tokyo, Japan).

TUNEL staining

Following euthanasia, brain tissues were rapidly excised and fixed in 4% paraformaldehyde. After sequential dehydration through a graded ethanol series, clearing in xylene, and paraffin embedding, coronal hippocampal Sect. (5 μm thickness) were obtained. Sections underwent deparaffinization in xylene followed by rehydration in graded ethanol. Subsequent processing included: (1) proteinase K digestion (20 µg/mL, 37 °C, 20 min) to expose DNA fragmentation sites, (2) endogenous peroxidase blockade with 3% H₂O₂, and (3) application of TUNEL reaction mixture (Servicebio, Cat# G1502) containing terminal deoxynucleotidyl transferase (TdT) and fluorescein-dUTP. Slides were incubated in the dark at 37 °C for 60 min to allow specific fluorescent labeling of DNA breaks. Nuclei were counterstained with DAPI (Servicebio, Cat# G1012), and sections were mounted with anti-fade medium. Apoptotic neurons exhibiting red nuclear fluorescence (TUNEL-positive signals) were visualized using a fluorescence microscope (Nikon Eclipse Ni). Quantitative analysis of neuronal apoptosis was performed by measuring fluorescence intensity in hippocampal subregions (CA3, CA1, DG) with ImageJ software (NIH).

Network pharmacology

Putative targets of baicalein were identified using the SwissADME database, while potential therapeutic targets for Alzheimer’s disease (AD) were mined from the OMIM, GeneCards(Relevance score ≥ 7), and DisGeNET databases(Confidence score ≥ 0.3). Common targets were obtained through Venn diagram analysis. Protein-protein interaction (PPI) networks of overlapping targets were constructed using the STRING database(v12.0; minimum interaction score: 0.7) and subjected to topological analysis and visualization with Cytoscape 3.10.0 software. Functional enrichment analysis of core targets in the PPI network was performed through Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG).

Molecular docking

The three-dimensional structures of PI3K and AKT were retrieved from the Protein Data Bank (PDB; https://www.rcsb.org/)¹⁶. The acquired protein PDB files were processed in AutoDockTools (version 1.5.7; https://ccsb.scripps.edu/mgltools/downloads/)¹⁷ by removing water molecules and adding hydrogen atoms. These prepared structures were designated as receptors and saved in pdbqt format. The mol2 file of Baicalein was obtained from the TCMSP database, hydrogenated in AutoDockTools, designated as the ligand, and saved as a pdbqt file. Molecular docking was performed using AutoDock Vina (version 1.2.3; https://vina.scripps.edu/)^18,19, with results visualized in PyMOL (version 2.6.0; http://www.pymol.org/pymol).

Drug affinity response target stability (DARTS) assay

SH-SY5Y-APP/PS1 cells were lysed and centrifuged to collect the supernatant. Protein concentration was determined using a BCA kit, followed by incubation with varying concentrations of Baicalein (0, 0.1, 0.3, 1, 3, 10, and 30 µM) at room temperature for 2 h. Pronase E (Sigma, P5147) was then added at a 1:20,000 ratio and incubated at room temperature for 10 min. The reaction was terminated by adding 2× SDS-PAGE loading buffer. After denaturation at 95 °C for 5 min, samples were subjected to Western blot analysis.

Cell counting kit-8 assay

Cells in the logarithmic growth phase were seeded in 96-well plates at 1 × 10⁴ cells/well and pre-cultured for 24 h at 37 °C with 5% CO₂ to facilitate adhesion. After cell attachment, the medium was replaced with fresh medium containing Baicalein (0, 0.3, 1, 3, 10, or 30 µM). Each concentration group included six replicate wells. Following 24 h of incubation, CCK-8 reagent (Beyotime, C0038) was added to each well at 10% (v/v) of the culture volume. Plates were protected from light and incubated for 2 h. Absorbance at 450 nm was measured using a microplate reader after gentle agitation to ensure uniform cell distribution. Cell viability was calculated as: (ODₜ_rₑₐₜₑd/OD_control) × 100%.

Flow cytometric analysis of apoptosis

Cell apoptosis was assessed using the Annexin V-FITC/PI dual-staining assay. Cells in logarithmic growth phase from each group were seeded in 6-well plates at 3 × 10⁵ cells/well. Following treatment, cells were harvested using EDTA-free trypsin and washed twice with ice-cold PBS. Cell pellets were resuspended in 100 µL of diluted 1× Annexin V Binding Buffer. Subsequently, 2.5 µL Annexin V-FITC reagent and 2.5 µL PI reagent (50 µg/mL) were added to the suspension. After vortex mixing, samples were incubated at 25 °C for 20 min protected from light, followed by addition of 400 µL diluted 1× Annexin Binding Buffer.

Analysis was performed on a BD FACSCanto™ flow cytometer (488 nm excitation) with 10,000 cellular events recorded per sample. Data processing was conducted using FlowJo v10.8 software. Following debris exclusion by FSC/SSC gating, total apoptosis rate was calculated as the sum of Annexin V⁺/PI⁻ (early apoptosis) and Annexin V⁺/PI⁺ (late apoptosis) populations.

RT-qPCR

Total RNA was extracted using TRNzol reagent (Tiangen Biotech), with purity and concentration measured by a microspectrophotometer (Kaiao, K5600C); cDNA was synthesized using FastKing gDNA Dispelling RT SuperMix (Tiangen, KR118). RT-qPCR was performed in 20 µL reactions containing 10 µL FastReal qPCR PreMix (SYBR Green; Tiangen, FP207), 0.6 µL each of forward/reverse primers (10 µM), 50 ng cDNA, and nuclease-free water, using a Bioer FQD-96 A system under cycling conditions: 95 °C for 2 min, followed by 40 cycles of 95 °C for 5 s and primer-specific annealing/extension at optimized temperatures for 30 s. GAPDH served as endogenous control, with relative expression calculated via the 2⁻^ΔΔCT method (primers in Table 1).

Table 1.

Primer sequences.

Primers	Primer sequences(5’−3’)	Primer length(bp)
PI3K	Forward: AGAGAGACTTAGAGAGCACAGGG Reverse: TATTACATCACACACCCCAGCC	223
AKT	Forward: TGTTTCTACTGTGGGCAGCA Reverse: TGGTCGCGTCAGTCCTTAAT	162
Caspase3	Forward: TAACAGGAAGGTGGCAACGG Reverse: ACCAGGAGGACCGTCAGATT	223
GSK3β	Forward: GCTGGAGTACACACCTACCG Reverse: GGAGGGATAAGGATGGTGGC	180
APP	Forward: CGTGATTCCTTACCGTTGCC Reverse: TGCCAGTGAAGATGGGTCTC	120
GAPDH	Forward: CGATGCCCCCATGTTTGTGA Reverse: GAGCCCTTCCACAATGCCAA	145

Western blot analysis

Hippocampal tissue lysates were prepared using lysis buffer (Elabscience, Cat. No. E-BC-R327), with total protein concentration quantified by BCA Protein Assay Kit (Beyotime, Cat. No. Poo12). Proteins were separated by gel electrophoresis, transferred to membranes, and blocked with 5% skim milk. Membranes were incubated overnight at 4 °C with primary antibodies against: GAPDH (Proteintech, 10494-1-AP, 1:5000), PI3K (CST, 4257, 1:1000), p-PI3K (ABclonal, AP0427, 1:1000), AKT1 (Abways, CY5561, 1:1000), p-AKT1 (CST, 4060, 1:1000), Caspase3 (Proteintech, 19677-1-AP, 1:1000), Bcl-2 (Proteintech, 26593-1-AP, 1:1000), GSK3β (Proteintech, 22104-1-AP, 1:1000), APP (Proteintech, 25524-1-AP, 1:1000), and p-GSK3β (Abways, CY6248, 1:1000). After three 10-min TBST washes, membranes were incubated with horseradish peroxidase-conjugated secondary antibodies: anti-mouse (Huabio, HA1006, 1:10000) or anti-rabbit (Huabio, HA1001, 1:20000) for 90 min at room temperature. Protein bands were visualized using ECL substrate (MeilunBio, MA0186-1) and detected with a chemiluminescence imaging system (Guangzhou Boluteng, GelView 6000 Pro II). Band intensity was quantified using ImageJ software.

Statistical analysis

Data are presented as mean ± standard error of the mean (mean ± SEM), and statistical analyses were performed using GraphPad Prism 9.0 software. Two-tailed Student’s t-test was used for comparisons between two groups; for multiple group comparisons, one-way analysis of variance (one-way ANOVA) was first performed. If significant differences were observed, Tukey’s post-hoc test was further applied for pairwise comparisons. The significance level was uniformly set at P < 0.05.

Results

Effects of baicalein on cognitive function in APP/PS1 mice

Spatial learning and memory capabilities across groups were assessed using the Morris water maze. Representative swimming trajectories from the place navigation and probe trials are presented in Fig. 1A. No statistically significant differences in swimming speed were observed among the groups (p > 0.05, Fig. 1B). Compared with the Control group, mice in the Model group exhibited significantly prolonged escape latencies during the training phase on days 2–5 (p < 0.01, Fig. 1C), indicating impaired learning ability. During the probe trial, the Model group showed significantly fewer platform crossings (p < 0.0001, Fig. 1D), significantly prolonged escape latency (p < 0.0001, Fig. 1E), and significantly reduced time spent in the target quadrant (p < 0.0001, Fig. 1F). These results demonstrate impaired spatial memory function in the Model group. In the subsequent novel object recognition test, the novel object recognition index was significantly lower in the Model group compared to the Control group (p < 0.001, Fig. 1G). Collectively, these results indicate significant cognitive dysfunction in the Model group.

Fig. 2.

Effects of baicalein on Aβ deposition and neuronal degeneration in APP/PS1 mice. (A) Representative Nissl-stained images of the hippocampal region (CA1: cornu ammonis 1; CA3: region 3; DG: dentate gyrus). (B-D) Quantitative analysis of surviving neuron density in the CA3 (B), CA1 (C), and DG (D) subregions. (E) Aβ₁₋₄₂ levels in hippocampal tissues measured by ELISA. (F) Ultrastructural images of hippocampal neurons by transmission electron microscopy (N: nucleus; M: mitochondria; red arrows indicate pyknotic nuclei and swollen mitochondria; ×5000; Scale bar = 5 μm).Data are presented as mean ± SEM (n = 3 per group).Statistical significance: ns, not significant (p > 0.05); *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, (intergroup comparisons indicated by connecting lines).

Fig. 1.

Baicalein improves learning and memory function in APP/PS1 mice. (A) Swimming paths showing spatial learning (top, Day 5 training) and memory retrieval (bottom, Day 7 probe trial). (B) Average swimming speed of mice during the water maze tests. (C) Escape latency to reach the hidden platform during the training phase. (D) Number of platform location crossings during the probe trial. (E) Latency to first reach the platform location during the probe trial. (F) Time spent in the target quadrant during the probe trial. (G) Discrimination index in the novel object recognition test. Data are presented as mean ± SEM (n = 10 mice per group). Statistical significance: ns, not significant (p > 0.05); *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, (intergroup comparisons indicated by connecting lines).

Following 40 days of treatment with Donepezil or Baicalein, compared to the Model group, both the Donepezil and Baicalein groups showed significantly reduced escape latencies during the place navigation trials on days 3–5 (p < 0.05, Fig. 1C). In the probe trial, both treatment groups exhibited significantly increased platform crossings (Donepezil: p < 0.001; Baicalein: p < 0.05; Fig. 1D), significantly reduced escape latency (p < 0.01, Fig. 1E), and significantly increased time spent in the target quadrant (Donepezil: p < 0.05; Baicalein: p < 0.01; Fig. 1F). In the subsequent novel object recognition test, both the Donepezil and Baicalein groups showed significantly elevated novel object recognition indices compared to the Model group (Donepezil: p < 0.05; Baicalein: p < 0.01; Fig. 1G). These results demonstrate that both Donepezil and Baicalein ameliorated cognitive deficits in the mice.

Baicalein ameliorates AD-like pathology in APP/PS1 mice

Nissl staining was employed to assess histopathological damage in brain tissue. Compared to the Control group, mice in the Model group exhibited a significant reduction in Nissl bodies within the hippocampal formation, accompanied by sparse and disorganized neuronal cell distribution. In contrast, both the Donepezil and Baicalein treatment groups showed increased numbers of Nissl bodies, relatively denser neuronal populations, and more organized cellular arrangements in the hippocampal region compared to the Model group (Fig. 2A). Quantitative analysis revealed that neuronal counts in the CA3, CA1, and DG regions were significantly decreased in the Model group relative to the Control group (p < 0.001). Conversely, neuronal counts in these corresponding regions were significantly increased in both the Donepezil and Baicalein groups compared to the Model group (p < 0.01) (Fig. 2B-D). These results indicate that both Donepezil and Baicalein exert protective and reparative effects on neuronal structure.

Fig. 4.

Effects of baicalein on neuronal apoptosis in APP/PS1 mice. (A) TUNEL staining of apoptotic neurons (Scale bar: 50 μm; N = 3). (B-D) Quantification of apoptosis-positive signals (relative fluorescence intensity) in hippocampal subregions: (B) CA3 (C) CA1 (D) Dentate gyrus (DG).Data are presented as mean ± SEM; Statistical significance: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (intergroup comparisons indicated by connecting lines).

Transmission electron microscopy (TEM) was used to examine neuronal ultrastructure. Hippocampal neurons in the Control group displayed intact cellular and nuclear structures, with aggregated chromatin within the nucleus and abundant mitochondria exhibiting intact morphology. Hippocampal neurons in the Model group exhibited significant pathology, including shrunken (pyknotic) cell bodies, collapsed nuclear membranes (often fragmented or absent), condensed or dispersed chromatin, markedly reduced numbers of mitochondria, and widespread mitochondrial swelling or rupture. Neurons from both the Donepezil and Baicalein groups also showed some degree of damage; however, nuclear morphology was largely preserved with intact structure, and only some mitochondria exhibited swelling (Fig. 2F).

As Aβ oligomerization initiates synaptic toxicity and tau hyperphosphorylation, constituting the core pathogenic mechanism of AD²⁰, we quantified Aβ deposition in hippocampal tissue across groups using a specific ELISA kit (Elabscience, E-EL-M3010). Hippocampal Aβ levels were significantly elevated in the Model group compared to the Control group (p < 0.0001). In contrast, both Donepezil and Baicalein treatment groups exhibited significantly reduced Aβ levels compared to the Model group (p < 0.0001)(Fig. 2E).

Collectively, these results demonstrate that 8-month-old APP/PS1 mice develop significant neuropathological damage characteristic of AD. Treatment with either Donepezil or Baicalein ameliorated these AD-like pathologies to varying extents.

Prediction of the mechanism underlying baicalein’s improvement of cognitive function in APP/PS1 mice

To explore the specific mechanism of action of Baicalein in treating AD, we conducted a network pharmacology study. AD-related targets were sourced from three databases: DisGeNET (https://www.disgenet.org/), GeneCards (http://www.genecards.org/), and OMIM (https://omim.org/). From the GeneCards database, 13,567 targets were initially retrieved. These were filtered using the median function (MEDIAN) based on the “relevance score” parameter, setting the threshold at relevance score > 11.40120888, resulting in 1,699 targets. From the OMIM database, 43 AD targets were obtained. From the DisGeNET database, 3,397 AD targets were acquired and filtered using the median function (MEDIAN) with a disease association score threshold of score > 0.05, yielding 973 AD targets (see Supplementary Table S1). After merging targets from all three databases and removing duplicates, a final set of 2,210 potential disease targets was obtained (see Supplementary Table S2). Potential targets of Baicalein were predicted using the SwissTargetPrediction database (http://www.swisstargetprediction.ch), yielding 100 potential targets (see Supplementary Table S3). The intersection between disease targets and drug targets was identified using Venny, resulting in 54 overlapping targets. These overlapping targets were subjected to GO (Gene Ontology) and KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway enrichment analysis using the DAVID database (david.ncifcrf.gov). GO analysis revealed that Baicalein is involved in biological processes such as cell proliferation, programmed cell death, and apoptosis. Its actions localize to cellular components including the plasma membrane, neuronal cell body, and cytoplasm. Molecular functions include protein kinase activity and protein kinase binding (see Supplementary Table S4) (Fig. 3A). KEGG²¹ analysis identified a strong association with the “PI3K-Akt signaling pathway” (Fig. 3B). The enrichment analysis results suggest that Baicalein likely exerts its effects by regulating neuronal proliferation and apoptosis via the PI3K-Akt signaling pathway.

Fig. 6.

Establishment and validation of the SH-SY5Y-APP/PS1 cell model. (A) APP mRNA expression after APP/PS1 plasmid transfection (N = 3). (B-C) APP protein expression post-transfection (N = 3): (B) Representative western blots, (C) Quantitative analysis of band intensity.(D) Intracellular Aβ₁₋₄₂ levels detected by ELISA (N = 6). (E-F) Apoptosis analysis by flow cytometry (N = 3): (E) Flow cytometry scatter plots, (F) Statistical analysis of apoptotic cell ratio.Data are expressed as mean ± SEM; Statistical significance: ***p < 0.001, ****p < 0.0001.

Fig. 3.

Mechanistic prediction of baicalein-mediated neuroprotection in APP/PS1 mice. (A) Gene Ontology (GO) enrichment analysis(Top 5 significantly enriched terms per category: Biological Process [BP], Cellular Component [CC], Molecular Function [MF]). (B) KEGG pathway enrichment analysis(Top 10 key signaling pathways). (C) Molecular docking models of baicalein with PI3K (left) and AKT (right).Binding energy: PI3K = −8.9 kcal/mol; AKT = −8.3 kcal/mol. (D) Drug Affinity Responsive Target Stability (DARTS) assay coupled with Western blot validation of direct binding between baicalein and PI3K/AKT.Lanes: 1. Untreated control; 2–7. Baicalein (0.1, 0.3, 1, 3, 10, 30 µM) + Pronase E (1:20,000 w/v, room temperature for 10 min).Enrichment significance: p < 0.05 with FDR correction. (E) Proposed mechanism of baicalein-mediated neuroprotection.

Subsequently, we validated the interaction capability of Baicalein with PI3K and AKT. We first performed molecular docking experiments. The 3D structures of PI3K (PDB ID: 6dgt) and AKT (PDB ID: 7uxf) were retrieved from the PDB database. The mol2 file for Baicalein was sourced from the TCMSP database. The molecular docking results yielded binding energies of −8.9 kcal/mol for Baicalein-PI3K and − 8.3 kcal/mol for Baicalein-AKT (Fig. 3C-D). According to established criteria²²: a binding affinity < −4.25 kcal/mol indicates binding activity between the ligand and target; < −5.0 kcal/mol indicates good binding activity; and < −7.0 kcal/mol indicates strong docking activity. These results clearly demonstrate that Baicalein exhibits strong interaction effects with both PI3K and AKT. Next, we conducted Drug Affinity Responsive Target Stability (DARTS) assays to further validate the interactions between Baicalein and PI3K/AKT. The results showed that Baicalein protected PI3K from proteolysis by pronase, and this protective effect increased with higher concentrations of Baicalein. However, Baicalein did not protect AKT from proteolysis by pronase. This indicates that Baicalein possesses binding capability specifically with PI3K (Fig. 3E).

Confirmation of the mechanism: baicalein ameliorates cognitive impairment in APP/PS1 mice by activating the PI3K/AKT pathway

Building upon our earlier findings, we hypothesized that Baicalein ameliorates cognitive impairment in APP/PS1 mice by inhibiting neuronal apoptosis. We assessed neuronal apoptosis levels in the hippocampal regions across groups using TUNEL assay. The results, as shown in Fig. 4A, revealed that compared to the Control group, APP/PS1 model mice exhibited a significant increase in TUNEL-positive signal intensity (indicating apoptosis) in the hippocampal CA1, CA3, and DG subfields (p < 0.001). Following Baicalein treatment, the TUNEL fluorescence intensity was significantly reduced in each of these subregions (p < 0.01) (Fig. 4B-D). These findings demonstrate that Baicalein effectively inhibits neuronal apoptosis in APP/PS1 mice.

Fig. 8.

Baicalein suppresses apoptosis by activating the PI3K/AKT pathway in SH-SY5Y-APP/PS1 cells. (A) Representative western blots of PI3K/AKT key pathway proteins. (B-F) Quantitative analysis of band intensity: (B) p-PI3K/PI3K ratio analysis, (C) p-AKT/AKT ratio analysis, (D) p-GSK-3β/GSK-3β ratio analysis, (E) BCL-2 expression quantification, (F) Cleaved caspase-3 expression quantification.Data expressed as mean ± SEM, (N = 3); *P < 0.05, *P < 0.01 vs. Model group (statistical comparisons indicated).

Based on the previous predictions regarding the mechanism of action, we examined the PI3K/AKT pathway and its key downstream apoptosis-related targets.

Western blot (WB) results showed that compared to the Control group, the Model group exhibited significantly reduced relative expression levels of p-PI3K/PI3K, p-AKT/AKT, and BCL2 (p < 0.01), while the relative expression levels of p-GSK3β/GSK3β and Caspase-3 were significantly increased (p < 0.01). Compared to the Model group, treatment with either Donepezil or Baicalein resulted in significantly increased relative expression levels of p-PI3K/PI3K, p-AKT/AKT, and BCL2 (p < 0.05), and significantly decreased relative expression levels of p-GSK3β/GSK3β and Caspase-3 (p < 0.05) (Fig. 5A-F).

Fig. 5.

Baicalein inhibits apoptosis by activating the PI3K/AKT pathway in APP/PS1 mice. (A) Representative western blots. (B-F) Quantitative analysis of band intensity: (B) p-PI3K/PI3K ratio, (C) p-AKT/AKT ratio, (D) p-GSK3β/GSK3β ratio, (E) BCL2 expression, (F) Cleaved caspase-3 expression. (G-J) RT-qPCR analysis of mRNA levels: (G) PI3K, (H) AKT, (I) GSK3β, (J) Caspase-3.Data expressed as mean ± SEM, N = 3; Statistical significance: ns, not significant (p > 0.05); *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, (intergroup comparisons indicated by connecting lines).

Simultaneously, RT-qPCR results revealed no statistically significant differences in the mRNA expression levels of PI3K, AKT, and GSK3β across groups (Fig. 5G-I). However, compared to the Control group, the Model group showed a significant increase in Caspase-3 mRNA expression level (p < 0.0001). Compared to the Model group, treatment with Donepezil or Baicalein significantly decreased Caspase-3 mRNA expression levels (p < 0.001)(Fig. 5 J).

Analysis of the WB results indicates that Baicalein inhibits neuronal loss by activating the PI3K/AKT pathway and thereby regulating the downstream apoptosis-related molecules GSK3β, Caspase-3, and BCL2. Furthermore, the RT-qPCR results, showing no statistically significant differences in PI3K, AKT, and GSK3β mRNA expression levels, demonstrate that Baicalein exerts its effects primarily through the phosphorylation of PI3K, AKT, and GSK3β proteins, rather than via transcriptional regulation.

Construction and validation of the SH-SY5Y-APP/PS1 cell line model

First, we verified the transfection efficiency of the APP/PS1 overexpression plasmids. RT-qPCR analysis confirmed APP mRNA overexpression in the transfected SH-SY5Y cells compared to controls (Fig. 6 A, p < 0.001). Western blot (WB) analysis further demonstrated overexpression of APP protein in the transfected SH-SY5Y cells (Fig. 6B, p < 0.001). Additionally, using an Aβ1–42 ELISA kit (Elabscience, E-EL-H0543), we assessed whether Aβ1–42 accumulation occurred post-transfection. The results revealed robust secretion of Aβ1–42 in the transfected SH-SY5Y cells (Fig. 6 C). These findings collectively demonstrate the successful establishment of our AD cell model.

Next, we evaluated whether this model exhibited increased susceptibility to apoptosis. Annexin V-FITC/PI staining confirmed this characteristic, showing that SH-SY5Y cells transfected with APP/PS1 exhibited a markedly higher apoptotic rate compared to the Vector control group (Fig. 6D-E, p < 0.0001).

Baicalein inhibits apoptosis in SH-SY5Y-APP/PS1 cells

We confirmed that Baicalein significantly inhibits apoptosis in SH-SY5Y-APP/PS1 cells in vitro. The optimal concentration of Baicalein was determined using the CCK-8 assay. Results showed that Baicalein concentrations up to 3 µM had no effect on the viability of wild-type SH-SY5Y cells, whereas concentrations of 10 µM and 30 µM significantly inhibited their viability (Fig. 7 A). Subsequently, SH-SY5Y-APP/PS1 cells were treated for 24 h with 3 µM Baicalein and/or 10 µM LY294002 (a PI3K inhibitor, MCE, Cat# 154447-36-6).

Fig. 7.

Baicalein inhibits apoptosis in SH-SY5Y-APP/PS1 cells. (A) Cytotoxicity screening of baicalein by CCK-8 assay. 3 µM was identified as the maximum non-toxic concentration (N = 6). (B) Cell viability assessed by CCK-8 assay after baicalein treatment (N = 6). (C) Apoptosis detection by Annexin V-FITC/PI dual staining. Representative flow cytometry plots. (D) Quantitative analysis of apoptotic cells (N = 3). Data expressed as mean ± SEM; Statistical significance: ns, not significant (p > 0.05); *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, (intergroup comparisons indicated by connecting lines).

The CCK-8 assay revealed that compared to the Vector group, cell viability was significantly decreased in the APP/PS1 group (p < 0.0001). Treatment with Baicalein significantly increased cell viability compared to the APP/PS1 group (p < 0.001). Notably, co-treatment with Baicalein and LY294002 resulted in a significant decrease in cell viability compared to Baicalein treatment alone (p < 0.0001), and the viability level was comparable to that observed with LY294002 treatment alone (p > 0.05) (Fig. 7B). These findings indicate that Baicalein effectively enhances the survival rate of SH-SY5Y-APP/PS1 cells, and this therapeutic effect is completely inhibited by LY294002.

Apoptosis inhibition by Baicalein was further validated using Annexin V-FITC/PI dual staining. Results demonstrated a significant increase in apoptosis in the APP/PS1 group compared to the Vector group (p < 0.0001). Baicalein treatment significantly inhibited apoptosis in SH-SY5Y-APP/PS1 cells (p < 0.01). However, this therapeutic effect was abolished by LY294002 co-treatment (Baicalein vs. Baicalein + LY294002, p < 0.0001; LY294002 vs. Baicalein + LY294002, p > 0.05) (Fig. 7C-D).

This study provides in vitro evidence that Baicalein inhibits apoptosis in SH-SY5Y-APP/PS1 cells, suggesting that its mechanism of action likely involves activation of the PI3K/AKT pathway.

Confirmation of the mechanism: baicalein inhibits neuronal apoptosis by activating the PI3K/AKT pathway

We next investigated the effect of Baicalein on the PI3K/AKT pathway in the SH-SY5Y-APP/PS1 cell model. Western blot (WB) results revealed that compared to the Vector group, SH-SY5Y cells transfected with APP/PS1 exhibited significantly reduced relative expression levels of p-PI3K/PI3K, p-AKT/AKT, and BCL2 (p < 0.0001), while the expression levels of p-GSK3β/GSK3β and Caspase-3 were significantly increased (p < 0.01) (Fig. 8A-F).

Compared to the APP/PS1 group, treatment with Baicalein significantly increased the relative expression levels of p-PI3K/PI3K, p-AKT/AKT, and BCL2 (p < 0.01) and significantly decreased the expression levels of p-GSK3β/GSK3β and Caspase-3 (p < 0.05). Notably, compared to the Baicalein group alone, co-treatment with Baicalein and LY294002 resulted in significantly reduced relative expression levels of p-PI3K/PI3K, p-AKT/AKT, and BCL2 (p < 0.0001) and significantly increased expression levels of p-GSK3β/GSK3β and Caspase-3 (p < 0.05). Furthermore, the expression levels in the co-treatment group remained comparable to those in the LY294002-alone group (p > 0.05) (Fig. 8A-F). These findings demonstrate that Baicalein activates the PI3K/AKT pathway and regulates the expression levels of apoptosis-related proteins, including p-GSK3β, BCL2, and Caspase-3, thereby exerting anti-apoptotic therapeutic effects.

Discussion

Alzheimer’s disease (AD), as a devastating neurodegenerative disorder, has progressive and irreversible neuronal loss in specific brain regions (such as the hippocampus and entorhinal cortex) as one of its core pathological hallmarks²³. This loss directly contributes to cognitive decline and dementia. Although the mechanisms underlying neuronal loss are complex and multifactorial, programmed cell death (PCD), particularly apoptosis, has been established as a key driver²⁴. The traditional view held that neuronal death in AD was primarily necrotic. However, recent research indicates that apoptosis represents a major form of neuronal loss in AD^25,26. Post-mortem studies have revealed significantly higher levels of activated caspase-3 (an executioner caspase in apoptosis) in hippocampal and cortical neurons of AD patients compared to age-matched controls. Furthermore, these levels show a strong correlation with disease severity and decreased neuronal density^27,28.

Recent research has profoundly elucidated the complex interplay between neuronal apoptosis and key AD pathological factors (such as β-amyloid and phosphorylated tau protein) as well as critical signaling pathways (like PI3K/AKT), providing new perspectives for understanding AD pathogenesis and developing targeted therapeutic strategies. The two core pathological factors of AD—Aβ and tau protein—have been confirmed as potent inducers of neuronal apoptosis^29,30.Aβ oligomers (oAβ), rather than fibrillar plaques themselves, strongly activate both the extrinsic and intrinsic apoptotic pathways. This occurs through direct binding to death receptors (such as Fas and p75NTR) on the neuronal surface^31,32 or indirectly by inducing endoplasmic reticulum stress³³ and mitochondrial dysfunction³⁴. Notably, recent studies have found that oAβ specifically activates caspase-8 (an initiator of the extrinsic apoptotic pathway), subsequently triggering the downstream caspase cascade, including caspase-3^35–37.The hyperphosphorylation, aggregation, and aberrant intracellular localization of tau protein not only impair neuronal function but also act as apoptotic signaling molecules themselves³⁸. Phosphorylation of tau at Ser-262 and Ser-356 represents an early molecular event in the formation of soluble tau assemblies (STA). STA induces neuronal apoptosis by activating the caspase-8/caspase-3 pathway, and antibodies targeting these phosphorylation sites can specifically recognize early tau pathology³⁹.Aβ acts upstream of tau in AD pathogenesis, triggering the conversion of tau from its normal state to a toxic form. Concurrently, toxic tau enhances Aβ toxicity through feedback loops⁴⁰.Furthermore, sustained neuroinflammation within the AD brain (microglial/astrocytic activation) releases large quantities of pro-inflammatory cytokines (such as TNF-α and IL-1β). These cytokines can activate both the neuronal death receptor pathway and the mitochondrial apoptotic pathway, synergizing with Aβ/tau to promote apoptosis^41–43.

The anti-inflammatory activity of baicalein—a key bioactive constituent of the ethnomedicinal plant Scutellaria baicalensis—scientifically validates its traditional application in ‘qingre jiedu’ (heat-clearing and detoxifying) treatments, achieved through inhibition of inflammatory signaling pathways^44,45.In the present study, we initially observed that APP/PS1 mice treated with baicalein intervention exhibited improved learning and memory capabilities, as assessed by the Morris water maze and novel object recognition tests. Their performance in these tasks reached levels comparable to those of the donepezil group.Pathological examinations revealed the neuroprotective effects of baicalein. Transmission electron microscopy (TEM) observations showed that hippocampal neurons in the AD model group displayed pyknosis, nuclear collapse, and fragmented or even dissolved nuclear membranes. In contrast, neurons in the baicalein and donepezil groups, while exhibiting varying degrees of injury, maintained relatively intact cellular structures and nuclei. Nissl staining demonstrated that compared to the Control group, the model group mice had sparse Nissl bodies and loosely arranged neurons in the hippocampal CA1 region. Conversely, both the baicalein and donepezil groups exhibited significantly increased density of Nissl bodies and neurons. Furthermore, an Aβ1–42 immunoassay kit detected that baicalein reduced Aβ accumulation in the hippocampus of APP/PS1 mice. TUNEL staining confirmed that baicalein inhibited apoptosis of hippocampal neurons.Concurrently, to investigate the neuroprotective effects of baicalein in vitro, we established an SH-SY5Y-APP/PS1 cell line via plasmid transfection. Model characterization confirmed that these cells displayed Aβ pathology characteristic of AD. Furthermore, continued culture for one week resulted in a progressive reduction in cell viability and increased apoptosis. Baicalein intervention significantly enhanced the viability of SH-SY5Y-APP/PS1 cells and suppressed apoptosis.Collectively, this compelling in vivo and in vitro evidence demonstrates that baicalein possesses neuroprotective properties, effectively inhibiting neuronal apoptosis, thereby exerting therapeutic effects against AD.

To further investigate the specific mechanism underlying baicalein’s neuroprotective effects, we conducted network pharmacology analysis. Notably, GO and KEGG analyses revealed significant enrichment in biological processes related to programmed cell death and apoptosis, as well as in the PI3K/AKT pathway. Studies indicate that activation of the PI3K/AKT pathway confers neuroprotection, potentially through the inhibition of neuronal apoptosis^11,46,47.Subsequently, we performed molecular docking of baicalein with PI3K and AKT. The results indicated strong binding potential between baicalein and both PI3K and AKT. However, in the subsequent Drug Affinity Responsive Target Stability (DARTS) combined with Western blotting (DARTS-WB) experiment, we observed a significant interaction only between baicalein and PI3K. This finding identifies PI3K protein as the direct target of baicalein. We therefore hypothesize that baicalein inhibits neuronal apoptosis by activating the PI3K/AKT pathway.The PI3K/AKT signaling pathway is a central regulator of cell survival, proliferation, and metabolism, playing a crucial role in maintaining neuronal homeostasis and conferring resistance to apoptosis^48,49. Activated AKT exerts potent anti-apoptotic effects by phosphorylating multiple downstream substrates. However, recent research confirms that PI3K/AKT signaling activity is significantly suppressed in the brain tissue of AD patients and across various AD models, representing a key factor contributing to increased neuronal susceptibility to apoptosis⁵⁰.Activated GSK3β promotes the apoptotic process and is a major contributor to the dysregulation of neuronal apoptosis^51–53. Activated AKT phosphorylates GSK3β at Ser9, leading to its inactivation and thereby reducing neuronal apoptosis. Bcl-2 is a critical apoptosis regulatory protein; its active state exerts anti-apoptotic effects, inhibiting neuronal apoptosis, other forms of neuronal programmed cell death, and axonal damage⁵⁴. The PI3K/AKT pathway can exert its anti-apoptotic effect by activating Bcl-2^55–57. Caspase-3, the executioner of apoptosis, is inhibited by the PI3K/AKT pathway, consequently reducing neuronal apoptosis^58,59.We validated the impact of baicalein on the PI3K/AKT pathway and downstream apoptotic factors both in vivo and in vitro. In the APP/PS1 mouse model, Western blot (WB) results demonstrated that baicalein upregulated the levels of p-PI3K, p-AKT, and BCL2, while suppressing the levels of p-GSK3β and Caspase-3. Concurrently, RT-qPCR results showed no statistically significant changes in PI3K, AKT, or GSK3β mRNA expression levels. This indicates that baicalein acts primarily through the phosphorylation of PI3K, AKT, and GSK3β proteins, rather than via transcriptional regulation.In the SH-SY5Y-APP/PS1 cell model, we similarly found that baicalein upregulated p-PI3K, p-AKT, and BCL2 levels while suppressing p-GSK3β and Caspase-3 levels. Importantly, this therapeutic effect was effectively blocked by the PI3K inhibitor LY294002. These results collectively demonstrate that baicalein activates the PI3K/AKT pathway by directly targeting the PI3K protein, thereby exerting its anti-neuronal apoptosis effect.

Conclusions

Our study confirms that baicalein, an active constituent derived from the traditional Chinese medicine Scutellaria baicalensis, ameliorates cognitive deficits, alleviates AD-like pathological alterations, and inhibits neuronal apoptosis in APP/PS1 mice. The research identified that baicalein interacts with PI3K, thereby activating the PI3K/AKT signaling pathway. Both in vivo and in vitro experiments consistently demonstrated that baicalein rescues from neuronal apoptosis through activation of the PI3K/AKT pathway. These findings elucidate the molecular mechanism underlying baicalein’s action in AD, indicating its potential as a promising therapeutic candidate for AD treatment.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

AD

Alzheimer’s disease

ELISA

Enzyme-linked immunosorbent assay

FITC

Fluorescein isothiocyanate isomer I

PI3K

Phosphatidylinositol-3-kinase

AKT

Activates protein kinase B

GSK3β

Glycogen synthase kinase-3β

Bcl-2

B-cell lymphoma-2

RI

The recognition index

DARTS

Drug affinity response target stability

TEM

Transmission electron microscopy

APP

Amyloid precursor protein

Aβ

Amyloid-beta

WB

Western blotting

RT-qPCR

Reverse transcription quantitative polymerase chain reaction

Author contributions

Wei Cheng: Conceptualization, Writing – review & editing, Writing – original draft, Data Curation, Validation, Software. Shuo Yang: Project administration, Visualization, Formal analysis, Methodology. Yi Lu: Writing–review & editing, Validation.Ying Wang: Data curation, Investigation.Ai-hua Tan: Project administration, Funding acquisition, Writing - review & editing.

Funding

This study was supported by grants 82405523 from the National Natural Science Foundation of China, grants 202520204 from the Open Fund of Hubei Key Laboratory of Germplasm Improvement and Utilization of Dabie Shan Dao-di Herbs.

Data availability

Data is provided within the manuscript and supplementary information files.

Declarations

Competing interests

The authors declare no competing interests.