Praeruptorin C alleviates cognitive impairment in type 2 diabetic mice through restoring PI3K/AKT/GSK3β pathway

Diabetic encephalopathy is a common consequence of diabetes mellitus that causes cognitive dysfunction and neuropsychiatric disorders. Praeruptorin C (Pra‐C) from the traditional Chinese medicinal herb Peucedanum praeruptorum Dunn. is a potential antioxidant and neuroprotective agent. This study was conducted to investigate the molecular mechanisms underlying the effect of Pra‐C on diabetic cognitive impairment. A novel object recognition test and the Morris water maze test were performed to assess the behavioral performance of mice. Electrophysiological recordings were made to monitor synaptic plasticity in the hippocampus. A protein–protein interaction network of putative Pra‐C targets was constructed, and molecular docking simulations were performed to predict the potential mechanisms of the action of Pra‐C. Protein expression levels were detected by western blotting. Pra‐C administration significantly lowered body weight and fasting blood glucose levels and alleviated learning and memory deficits in type 2 diabetic mice. Network pharmacology and molecular docking results suggested that Pra‐C affects the PI3K/AKT/GSK3β signaling pathway. Western blot analysis confirmed significant increases in phosphorylated PI3K, AKT, and GSK3β levels in vivo and in vitro upon Pra‐C administration. Pra‐C alleviated cognitive impairment in type 2 diabetic mice by activating PI3K/AKT/GSK3β pathway.

International Diabetes Federation, an estimated 463 million adults worldwide were diagnosed with DM in 2019.This number is projected to increase to 700 million by 2045 (Cho et al., 2018;Walters et al., 2021).Mounting evidence indicates that persistent hyperglycemia in DM can increase the risk of serious complications such as retinopathy, nephropathy, neuropathy, and cardiovascular disease (Yan et al., 2022).
Diabetic encephalopathy (DE) is a major complication of DM, characterized by brain infarction (Ward et al., 2018), atrophy of the cortex and hippocampus (Schmidt et al., 2004), cerebrovascular dysfunction and morphological changes (Chen et al., 2018), and impairment of synaptic plasticity (Fried et al., 2017).Cognitive impairments associated with DE, which are often difficult to reverse, can ultimately lead to dementia.Studies show that patients with DM, especially those with type 2 diabetes mellitus (T2DM), have a higher risk of developing dementia and Alzheimer's disease (Tang et al., 2022;Zheng et al., 2021).Furthermore, learning and memory deficits have been observed in various diabetic mouse models, such as obesity induced by feeding a high-fat diet (HFD), pancreatic β cell damage by streptozotocin (STZ) injection, or genetic construction (db/db mice; Rom et al., 2019;Taylor et al., 2015;Yan et al., 2016).Thus, DE has emerged as a critical area of study for preventing diabetes.However, existing medications have not proven effective at treating DE, highlighting the need for new and effective therapeutic agents and strategies.
Praeruptorin C (Pra-C; Figure 1a) is the main bioactive constituent isolated from the dried roots of Peucedanum praeruptorum Dunn (also known as Qian Hu).Qian Hu is a traditional Chinese medicinal herb used to treat respiratory diseases such as asthma and bronchitis (Su et al., 2021).Studies have shown that Pra-C can lower blood pressure (Wang et al., 2014), dilate coronary arteries (Zhu & Jiang, 2018), exert anti-inflammatory (Yu et al., 2012), and antiosteoporosis effects (Liu et al., 2017), protect against NMDA-induced apoptosis by downregulating GluN2B levels and calcium influx (Yang et al., 2013), and alleviate motor deficits and depression-like behaviors in a mouse model of Huntington's disease (Wang et al., 2017).
Given Pra-C's beneficial effects on neuronal apoptosis and neurodegenerative animal models, we hypothesized it could be an effective agent for treating DE.Therefore, we aimed to investigate the therapeutic effects of Pra-C on cognitive impairment in type 2 diabetic mice and explore its underlying mechanisms to lay the groundwork for the clinical use of Pra-C in intervening and improving DE.
All mouse experiments were conducted in a randomized manner by an investigator who was blinded to the experimental groups.To induce T2DM, mice received HFD (60% kcal fat, Ke'ao Biotech, Xi'an, China) for 4 weeks, and a single dose of STZ (60 mg/kg, dissolved in 0.1 M citrate buffer) was intraperitoneally injected after 12 h of fasting, while control mice were given regular food and the same volume of citrate buffer (0.1 M).Model mice were then continuously fed an HFD for 1 week and subjected to a blood glucose test using single-tail tip pricks.HFD-fed mice with fasting blood glucose levels >13 mM were considered to have successfully established T2DM (Zhao et al., 2022).Blood glucose levels were measured every 2 weeks.All mice were allowed to acclimate for 24 h in the laboratory environment before any room changes were made.

| Pra-C treatment
After the establishment of the T2DM model, mice were randomly divided into two groups (12 mice per group), T2DM and Pra-C treatment, and continuously fed an HFD.Pra-C was dissolved in 0.9% saline and administered intraperitoneally (i.p., 3 mg/kg/day) at 9 a.m.once a day for 4 weeks, starting in the fifth week, while the T2DM group mice received an equal volume of 0.9% saline.The dose of Pra-C was selected based on a previous study (Su et al., 2021).Control mice were fed conventional chow and administered equal amounts of 0.9% saline.Behavioral tests were performed 3 weeks after treatment.Fasting blood glucose levels and body weight were measured before sacrifice.

| Novel object recognition test
A novel object recognition (NOR) test was performed in a soundproof polyvinyl chloride box (25 cm Â 25 cm bottom, 25 cm height) using a digital camera on the roof.On Day 1, a mouse was placed in the box and allowed to acclimate to the new environment for 10 min.On Day 2, the mice were placed in a box with two cylinders (5 cm in diameter and 5 cm in height) fixed to a symmetrical corner of the box (training trial).On Day 3, each group was randomly divided into two groups.
For NOR, one cylinder was replaced with a cone (5 cm in diameter and 5 cm in height; testing trial).For the object-place recognition test, the former cylinders were fixed in parallel corners, with one in the former location used as the training trial and the other in a different location.In both the training and testing trials, the mice were allowed to explore for 10 min freely, and the time spent exploring each object or location was recorded and scored using the DigBehv-LR4 system (Shanghai Jiliang Software Science & Technology Co., Ltd., Shanghai, China).After each experiment, the boxes and objects were wiped with 75% ethanol.The proportion of exploration time on a novel object or novel object location was defined as the "discrimination index (DI)," expressed by the ratio of (TN À TF)/(TN + TF) (TN, time spent on exploring the novel object or location; TF, time spent on exploring the familiar object or location).

| Morris water maze test
The Morris water maze (MWM) test was conducted in a circular tank (120 cm in diameter and 50 cm deep) with white walls decorated with bold marks in four directions.An escape platform (10 cm in diameter) was erected midway between the center and the wall of the tank.
Before the experiment, opaque water (24 C± 1 C, with an unharmed white coating) was filled in the tank to a height of 1.0 cm above the platform.Mice were placed in water and trained for four trials (different starting positions) per day with 10 min intertrial intervals for 4 consecutive days.During each trial, the mice were allowed to explore until they found the hidden platform and were held on it for 20 s before returning to the holding cage.Mice that failed to find the platform within 60 s were guided to it using a wooden stick.
The probe test was performed 24 h after the final training trial, and the mice were placed in the same height water-filled tank with an undersurface-hidden platform for 90 s, and their swim paths were recorded.

| Field electrophysiological recording
Hippocampal slices were prepared as previously described (Yang et al., 2016).Briefly, slices (300 μm) were cut in an oxygenated solution (in mM): 250 Sucrose, 2.5 KCl, 0.5 CaCl 2 , 6 MgSO 4 , 1.2 NaH 2 PO 4 , 25 NaHCO 3 , and 10 D-glucose.After cutting, the slices were recovered at 34 C in artificial cerebrospinal fluid (ACSF) (in mM) as follows: 124 NaCl, 4.4 KCl, 2 CaCl 2 , 1 MgSO 4 , 1 NaH 2 PO 4 , 25 NaHCO 3 , and 10 D-glucose.After 10 min, the slices were placed in ACSF at room temperature for an additional 1-2 h, gassed with 95% O 2 to 5% CO 2 .A commercial MED64 recording system (Panasonic Alpha-Med Sciences, Osaka, Japan) was used to record field excitatory postsynaptic potentials (fEPSPs) as described in a previous study (Yang et al., 2015).Briefly, the single hippocampal slice was placed in a probe (MED-P515A, 8 Â 8 array; interpolar distance, 150 μm), positioned on the Schaffer collateral-CA1 pathway of the dorsal hippocampus, with oxygenated ACSF (30.0 C, 1-2 mL/min) perfusion.After a recovery period of at least 1 h for the slices in the recording chamber, biphasic constant-current pulse stimulation (0.2 ms) was applied to the stimulation channel to evoke fEPSPs in the channels closest to the stimulation site.Stable baseline responses were recorded for at least 40 min, late-phase long-term potentiation (L-LTP) was induced by strong theta burst stimulation (TBS; five bursts at 5 Hz, repeated five times at 10 s intervals, four pulses at 100 Hz for each burst; Liu et al., 2018), and the test stimulus was repeated once per minute for at least 3 h.The fEPSPs were measured and analyzed using the Mobius software (Panasonic Alpha-Med Sciences, Tokyo, Japan).

| Cell culture and treatment
Primary hippocampal cells were prepared as previously described (Kulkarni et al., 2020) with minor modifications.Briefly, the hippocampi were isolated from 18-day-old C57BL/6 mouse embryos.Neuronal tissues were enzymatically digested with 0.25% trypsin for 15 min at 37 C. Neurons were plated in poly D-lysine-coated 6-well plates at a density of 1 Â 10 6 cells per well.Cultures were incubated with DMEM containing 20% FBS at 37 C in a 5% CO 2 incubator.
After 6 h, the medium was replaced entirely with a maintenance medium (neurobasal medium supplemented with 2% B27 and 1 Â GlutaMAX).Every 3-4 days, half of the medium was replaced.
Primary hippocampal cells were used for experiments at 8 days in vitro.
To confirm the effects of Pra-C on the PI3K/AKT/GSK3β signaling pathway, the cells were pretreated with Pra-C (10 μM) for 1 h in either the absence or presence of LY294002 (10 μM) before exposure to a high glucose solution.A high-glucose solution was prepared using maintenance media with additional D-glucose at a final concentration of 100 mM.The dose of Pra-C was selected based on a previous study (Su et al., 2021).After 24 h of culture, the cells were harvested for further analysis.

| Western blot analysis
Western blotting was performed as described previously (Yang et al., 2014).After behavioral and blood glucose tests, mice were initially anesthetized with urethane (1.5 g/kg) and rapidly decapitated.
The brain was immediately placed on ice; the hippocampus was removed, respectively, collected in 1.5 mL tubes, and stored at À80 C until further analysis.After treatment, cells were harvested using an ice-cold RIPA lysis buffer.Tissues and cells were homogenized by RIPA lysis buffer and separated by centrifugation at 13,000 rpm for 20 min at 4 C, and the supernatant was collected.All proteins were quantified using a BCA protein assay kit (Thermo Fisher Scientific, Waltham, USA), and equal amounts of protein were subjected to western blot analysis.Samples were separated via SDS-PAGE gel and electrotransferred onto PVDF membranes (Merck KGaA, Darmstadt, Germany), which were blocked with primary antibodies at 4 C overnight, washed three times with Tris-phosphate buffer containing 0.05% Tween 20 for 10 min each time, and further incubated with HRP-conjugated secondary antibodies.The bands were developed using enhanced chemiluminescence detection and imaged using a ChemiScope imaging system (Clinx Science Instruments Co., Ltd., Shanghai, China).

| Molecular docking
The crystal structure of PI3K was downloaded from RCSB (http:// www.pdb.org) in the PDB format (ID:4OVV).The 3D structure of Pra-C was optimized using the MMFF94 force field and Discovery Studio software 2019.All hydrogen atoms and Gasteiger charges were added to the modeled receptor using AutoDock Tools (Morris et al., 2009), and molecular docking was performed using AutoDock Vina (Trott & Olson, 2010).The grid box was set as center_x: 74.2297, center_y: 45.6328, and center_z: 86.6748 with dimensions size_x:40, size_y:40, and size_z:40 within a druggable pocket based on PrankWeb server prediction (Jendele et al., 2019).The poses were ranked by their docking scores, and the pose with the lowest score in terms of binding free energy (kcal/mol) was selected as the proteinligand complex to be used in molecular dynamics (MD) simulations.
Interactions between the protein and ligand were analyzed using PyMol (https://pymol.org).

| MD simulation
MD simulations were performed using the complex chosen in the previous step as the initial conformation using the standard dynamics cascade protocol in Discovery Studio 2019.CHARMM36 force field was used, and the protein-ligand complex was solvated in an orthorhombic cell shape.The simulation was carried out with the natural pressure and nearly 300 K temperature.An equilibration simulation was conducted for 20 ps, followed by a production simulation for 200 ps.The results were analyzed using root mean squared deviation (RMSD) and root mean squared fluctuation (RMSF) time profiles.

| Data analysis
Statistical analyses were performed using GraphPad Prism software version 8.0.All data are presented as the mean (SEM).The statistical significance of the differences between groups was analyzed using one-way analysis of variance, followed by least significant difference and S-N-K(s) t-tests.

| Effects of Pra-C on body weight and fasting blood glucose levels
Obesity and high saturated fat intake are risk factors for T2DM that can harm the brain and cognitive function (Mueller et al., 2012).Thus, successfully controlling blood glucose levels can improve T2DMassociated learning and memory impairments.In this study, no animals in the experimental group died and there was no significant difference in fur color between the experimental and control groups.The effects of Pra-C on body weight and fasting blood glucose levels were evaluated in HFD/low-dose STZ-induced T2DM mice to determine its potential as an intervention for these impairments.The animal model induction and Pra-C administration are shown in Figure 1b.T2DM mice showed significantly higher body weights than mice on a normal diet, whereas body weight decreased after treatment with Pra-C (3 mg/kg/day intraperitoneally) for four consecutive weeks (Figure 1c).Additionally, T2DM mice exhibited hyperglycemia with significantly elevated fasting blood glucose levels compared to normal controls (Figure 1d).However, blood glucose levels were attenuated in T2DM mice treated with Pra-C.

| Pra-C relieves cognitive impairment in T2DM mice
We performed NOR and MWM tests to investigate whether Pra-C could mitigate memory and learning impairments.In the NOR test (Figure 2a), T2DM mice spent less time exploring novel objects than control mice, whereas Pra-C-treated mice explored objects longer than untreated T2DM mice (Figure 2b).Moreover, the novel location recognition test showed that the DI of T2DM mice was significantly lower than that of controls.In contrast, Pra-C-treated mice had higher DIs than untreated T2DM mice (Figure 2c).
In the MWM test, control mice found the platform faster over 4 days, whereas T2DM mice took significantly lower.After Pra-C treatment, the escape latency of T2DM mice was significantly reduced (Figure 2d).On the MWM testing day, T2DM mice spent less time in the target quadrant than the controls, whereas the Pra-C-treated group dwelled longer than untreated T2DM mice (Figures 2e,f).These behavioral results indicate that Pra-C enhances the learning and memory abilities of T2DM mice, demonstrating its ameliorative effects on cognitive impairment in diabetes.

| Pra-C reverses the impairment of L-LTP facilitation in T2DM mice
LTP is a cellular model of information storage in the brain linked to learning and memory (Bliss & Collingridge, 2013).Diabetic animals exhibit impaired LTP facilitation at Schaffer collateral-CA1 synapses in the hippocampus (Stranahan et al., 2008;Winocur et al., 2005).Here, we investigated L-LTP, a temporal protein synthesis-dependent phase of LTP in the hippocampal CA1 region, using MED64 recordings (Figure 3a).Hippocampal slices from all groups showed elevated fEPSP slopes to the TBS in the Schaffer collateral axons compared with the baseline recording slopes (Figure 3b).In control slices, L-LTP persisted above baseline for at least 3 h.However, the fEPSP slopes in T2DM mice decreased significantly 2 h postinduction.Pra- C-treated slices showed lower slopes than controls at completion, yet still significantly higher than those from T2DM mice (Figure 3c).These results indicate that Pra-C mitigates L-LTP impairment in T2DM mice.

| PPI network of putative Pra-C targets revealed potential involvement of PI3K/AKT pathway in Pra-C activity
A network pharmacology approach was used to explore the potential mechanism of Pra-C action.Due to the limited number of studies on Pra-C and its relatively high Pra-C content in the bioactive Qian Hu extracts, Qian Hu was used as a search term instead of Pra-C to identify relevant studies in this part.These studies provide a preliminary indication of the putative targets of Pra-C.Based on a database search, 1587 genes were associated with Qian Hu function, 1429 with cognitive impairment-related, and 1359 with DM.From these, 86 targets were identified as putative targets of Qian Hu against diacognitive impairment (Figure 4a).These were then submitted to STRING to construct a PPI network (Figure 4b) at a confidence score of 0.95, comprising 86 nodes and 1585 edges.GO and KEGG pathway enrichment analyses were performed to explore the functions of the identified targets further.Figure 4c  functions.A total of 170 signaling pathways were identified as significant through KEGG enrichment analysis (p < 0.05), and the top 20 significantly enriched pathways are shown in Figure 4d.The PI3K/ AKT signaling pathway was identified as the most significant (hsa04151, Gene Ratio = 28/79), suggesting its involvement in the Qian Hu/Pra-C action mechanism.These results support the hypothesis that the PI3K/AKT pathway underlies the therapeutic activity of Pra-C.

| Molecular docking and MD simulations reveal interactions between Pra-C and PI3K
PI3K is a lipid kinase with catalytic (p110) and inhibitory (p85) subunits.PI3K phosphorylates phosphatidylinositol 4,5-bisphosphate (PIP2) into phosphatidylinositol 3,4,5-trisphosphate (PIP3; Zhang et al., 2019).PIP3 also recruits AKT to the plasma membrane.The PI3K/AKT signaling pathway regulates glucose homeostasis, cell proliferation, and protein synthesis mediated by glucose-related hormones in the digestive and central nervous systems (Ahima et al., 1996).MD simulations were thus performed to evaluate a potential interaction between Pra-C and PI3K.For this purpose, a druggable pocket (blue mesh in Figure 5a) with a higher score than that of the ATP-binding site (red mesh in Figure 5a) was identified for docking using the PrankWeb server.Pra-C nested into a hydrophobic pocket formed by residues TRP 424, PRO447, PRO449, THR 679, and PRO1011 with a docking score of À8.1 kcal/mol.Moreover, Pra-C also interacted with the key TYR467 residue of p85 via hydrogen bonding (Figure 5c).MD simulations were performed to validate the molecular docking results.As shown in Figure 5d, the protein-ligand complex remained stable after 100 ps, with a mean RMSD of 1.78 Å. RMSF values (Figure 5e) further indicated the reliability of our docking-based prediction.These results indicate that PI3K may be a potential target of Pra-C, requiring experimental validation.

| Pra-C reactivates PI3K/AKT/GSK3β pathway in T2DM mice
The hippocampus is a vital brain region involved in learning and memory formation and is particularly affected by hyperglycemia-mediated dysregulation of metabolism and osmolarity (Heng et al., 2011).Thus, we validated the effects of Pra-C on key proteins of the PI3K/AKT signaling pathway in the hippocampi of HFD/low-dose STZ-induced T2DM mice.As illustrated in Figure 6, T2DM mice showed significantly lower phosphorylated PI3K, AKT, and GSK3β levels in the hippocampus than those in the control group, with unchanged total protein levels.Moreover, Pra-C treatment significantly restored PI3K, AKT, and GSK3β phosphorylation levels.These results suggest that Pra-C alleviates cognition impairment in T2DM mice by reactivating the PI3K/AKT/GSK3β signaling pathway.
3.7 | Effects of Pra-C on the PI3K/AKT/GSK3β pathway in hippocampal neurons induced by high glucose levels The effects of Pra-C on the high glucose-induced inhibition of PI3K/ AKT signaling in primary hippocampal neurons were investigated.
Expressions of phosphorylated PI3K, AKT, and GSK3β were significantly downregulated in hippocampal neurons exposed to D-glucose at 100 mM.However, 10 μM Pra-C treatment reversed decreases in the expression of these proteins (Figure 7a The trimeric α-amino-3-hydroxyl-5-methyl-4-isoxazolepropionate-type glutamate receptors are responsible for fast excitatory synaptic transmission in the central nervous system, and they are composed of GluA1-GluA4 subunits that combine in various ways (Diering & Huganir, 2018;Qu et al., 2021).Serine 845 and 831 cophosphorylation of GluA1 is crucial for synaptic plasticity and regulates learning and memory (Sathler et al., 2021).Therefore, we analyzed the expression of phosphorylated and total GluA1 proteins in different treatment groups.Our findings, shown in Figure 7c,d

| DISCUSSION
This is the first study to report the effects of Pra-C on diabetic cognitive impairment.Our findings indicate that Pra-C treatment reduces body weight and blood glucose levels and alleviates cognitive impairment in a T2DM mouse model.These positive effects may be attributed to restoring PI3K/AKT/GSK3β signaling in the hippocampus and alleviating L-LTP impairment in T2DM mice.
Epidemiological evidence supports a connection between diabetes, neurodegenerative diseases, and cognitive impairment (Zhang et al., 2017).Although various strategies targeting DE have been considered, identifying an effective treatment remains challenging.Traditional Chinese medicinal herbs have long been used in China and other Asian countries for chronic diseases such as diabetes.They are increasingly accepted by Western countries due to their therapeutic effects (Cyranoski, 2018).The richness and diversity of natural herbal resources provide ample opportunities to develop drugs for various diseases, with natural products like quercetin, berberine, and curcumin having been reported to alleviate cognitive deficits in diabetic animals.
Advances in isolation and purification techniques have identified more medicinal compounds than ever.
In our previous study, we found that Pra-C exerts neuroprotective effects by downregulating GluN2B-containing Nmethyl-D-aspartate receptors (Yang et al., 2013).Additionally, Pra-C also improved behavioral performance in a Huntington's disease mouse model by regulating neurotransmitter balance (Wang et al., 2017).
Hyperglycemia-induced DE impairs cognitive abilities and increases the risk of neurodegenerative diseases in diabetic patients and animals (Fried et al., 2017;Rom et al., 2019;Zhang et al., 2017).cognitive impairment by repairing impaired LTP (Choi et al., 2023;Park et al., 2021;Wang et al., 2023).LTP has two main temporal phases: the early phase (E-LTP) and the late phase (L-LTP), with the latter dependent on new protein synthesis (Huang, 1998).In this study, Pra-C treatment restored L-LTP impairment in the hippocampal CA1 region of T2DM mice.PI3K/AKT signaling plays a crucial role in regulating fundamental cellular processes under normal and pathological conditions such as obesity and T2DM, where high glucose levels can decrease the activated PI3K and AKT in various tissues, including the skeletal muscle, adipose tissue, liver, pancreas, and brain (Huang, Liu, et al., 2018).We constructed a PPI network to explore potential mechanisms of action of Pra-C on diabetic cognitive impairment.We identified PI3K/AKT signaling as the target pathway, which we further validated through molecular docking simulations to evaluate the interactions between Pra-C and PI3K.Using the PrankWeb server (Jendele et al., 2019), a druggable pocket with a score greater than the ATP-binding site was identified for docking.The result showed that Pra-C was well nested into the hydrophobic pocket with a docking score of À8.1 kcal/mol.In addition, an MD simulation was performed, and the resulting RMSD and RMSF values indicated that the molecular docking-based predictions were reliable.Crystal structures have shown that the distance between the γ phosphate of ATP and the lipid substrate PIP2 in PI3K was over 6 Å, which is too high for the phosphoryl transfer (Zhang et al., 2019).Since the selected docking pocket was close to the substrate-binding site, we hypothesized that Pra-C binding in this pocket might bring PIP2 and ATP closer to each other, activating PI3K.
Our in vivo findings confirmed that Pra-C restored the phosphorylation of PI3K and AKT in the hippocampus of T2DM mice.In vitro, findings on primary hippocampal neurons revealed similar results.
Dysfunctioning the PI3K/AKT signaling pathway leads to abnormal GSK3β activity and promotes tau hyperphosphorylation and axonopathy in T2DM-induced DE (Wang et al., 2018) gated to horseradish peroxidase (HRP) were purchased from Santa Cruz Biotechnology (Dallas, USA).DMEM-high glucose, neurobasal medium, fetal bovine serum (FBS), B27, and GlutaMAX supplements were provided by Life Technologies Corporation (Grand Island, USA).All the chemicals and reagents used were of standard biochemical quality.F I G U R E 1 Effects of praeruptorin C (Pra-C; i.p., 3 mg/kg/ day) on body weight and fasting blood glucose levels in high-fat diet (HFD)/low dose streptozotocin (STZ)-induced type 2 diabetes mellitus (T2DM) mice.(a) The chemical structure of Pra-C.(b) Overview of the experimental design.(c) Changing of body weight detected weekly after 4 weeks of HFD feeding.(d) Levels of 4 h fasting blood glucose every 2 weeks after establishment of T2DM model.Values were expressed as mean ± SEM (n = 12 in each group).#p < 0.05, ##p < 0.01 versus T2DM group.NOR, novel object recognition; MWM, Morris water maze.

2. 2 |
Experimental animals and establishment of type 2 diabetic model Male C57BL/6 mice (5-7 weeks old, weighing 18-22 g) were obtained from the Experimental Animal Center of Air Force Medical University (AFMU; Xi'an, China).Mice were fed under standard laboratory conditions (12-h light/12-h dark, temperature 22 C-26 C, air humidity 55%-60%) with water and chow ad libitum.The experimental procedures were approved by the Institutional Animal Care and Use Committee of AFMU (No. 20211026) and complied with the Protection of Animals and Principles of Laboratory Animal Care (NIH pub-

F
I G U R E 2 Praeruptorin C (Pra-C) ameliorated memory and learning impairments in type 2 diabetes mellitus (T2DM) mice.(a) Schematic of the novel object recognition test, mice of each group were randomly average assigned to different tests on Day 3. (b,c) Summary of discrimination index for novel object and location.(d) Latency of mice to located platform position during learning phase of water maze test.(e) Percentage of time spent in the target quadrant of water maze during probe test.(f) Representative swimming paths of mice in water maze during probe test.Values were expressed as mean ± SEM (n = 12 in each group).*p < 0.05, **p < 0.01 versus control; #p < 0.05, ##p < 0.01 versus T2DM.
Praeruptorin C (Pra-C) rescued late-phase long-term potentiation in hippocampal CA1 of type 2 diabetes mellitus (T2DM) mice.(a) Representative average traces of last 5 min of baseline (black) and last 5 min of whole recording (red).(b) Field excitatory postsynaptic potential (fEPSP) slope recorded from hippocampal CA1 region.Arrow indicates the time point of TBS training.(c) Average fEPSP slope of last 30 min during whole recording.Values were expressed as mean ± SEM (n = 8 slices/4 mice for each group).**p < 0.01 versus control; ##p < 0.01 versus T2DM.
shows the top 15 enrichment results for biological processes, cellular components, and molecular F I G U R E 4 Network pharmacology analyses.(a) The Venn diagram showed the overlapping targets between Qian Hu, cognitive impairment and diabetes mellitus.(b) The protein-protein interaction network constructed by 86 cross-targets with each node representing a different protein.(c) Gene Ontology enrichment analysis included biological process, cellular component, and molecular function.(d) Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis.
,b), in line with in vivo results.Additionally, LY294002, a specific PI3K inhibitor, eliminated the effects of Pra-C on high glucose-induced decreases in phosphorylated PI3K, AKT, and GSK3β, confirming the ability of Pra-C to reactivate the PI3K/AKT/GSK3β pathway.
, demonstrate that Pra-C treatment significantly upregulated GluA1S831 and GluA1S845 phosphorylation compared with the high glucose group, while the total GluA1 protein level remained largely unchanged.The LY294002 treatment mitigated these positive effects.Altogether, our results suggest that Pra-C may activate the PI3K/AKT/GSK3β signaling pathway to facilitate synaptic plasticity.
In this study, T2DM mice exhibited cognitive impairment, evidenced by their poor ability to discriminate novel objects, impaired spatial memory in the novel location recognition test, increased time spent finding the platform in MWM training, and decreased time in the target quadrant during hidden platform testing.Pra-C treatment reduced body weight, fasting blood glucose levels, and improved cognitive abilities in T2DM mice.LTP is the most widely used paradigm to study molecular and cellular events underlying synaptic transmission and is considered a F I G U R E 5 Molecular docking and molecular dynamics (MD) simulation of praeruptorin C (Pra-C) with PI3K (PDB ID: 4OVV).(a) A druggable pocket (blue mesh, score: 8.97) and the ATP-binding site (red mesh, score: 8.16) of PI3K based on PrankWeb server prediction.(b) Surface representation of the PI3K-binding pocket with Pra-C.(c) H-bond (blue dotted line) and hydrophobic interactions (gray-dotted line) between Pra-C (green sticks) and PI3K.(d) Root mean squared deviation (RMSD) and (e) root mean squared fluctuation (RMSF) line plots of the PI3K-Pra-C complex during 200 ps stimulation.putativemechanism for learning and memory(Bliss & Collingridge, 1993).Impaired LTP at the hippocampal synapses is a potential neurobiological mechanism underlying cognitive impairment, as learning and memory deficits are commonly observed in animals with impaired LTP(Rong et al., 2023;Zhu et al., 2022).Several natural products and derived compounds have been reported to alleviate F I G U R E 6 Praeruptorin C (Pra-C) restored PI3K/AKT/GSK3β signaling pathway in the hippocampus.The phosphorylated and total levels of (a) PI3K, (b) AKT, and (c) GSK3β proteins, with β-actin (ACTB) as internal control.Values were expressed as mean ± SEM (n = 6 for each group).*p < 0.05, **p < 0.01 versus control; #p < 0.05 versus type 2 diabetes mellitus (T2DM).

F
I G U R E 7 Effects of praeruptorin C (Pra-C) on HG-induced primary hippocampal neurons.(a,b) Western blot analysis showed phosphorylation levels of PI3K, AKT, GSK3β, and the correspondent total forms under different treatment.(c,d) The phosphorylated GluA1S831, GluA1S845, and total GluA1 levels by western blot analysis.Data were represented as mean ± SEM (n = 4 for each group).**p < 0.01 versus control; ##p < 0.01 versus high glucose group.
. Here, downregulation of the PI3K/AKT/GSK3β pathway activity in the hippocampus was observed in T2DM mice.Restoration of phosphorylation levels in this pathway upon Pra-C treatment may be partially responsible for alleviating cognitive impairment.Additionally, serine 845 and 831 co-phosphorylation in GluA1 regulates synaptic plasticity, learning, and memory.Downregulation of GluA1S831 and GluA1S845 phosphorylation was also observed in high glucose-induced hippocampal neurons and was significantly restored by Pra-C treatment.Despite the demonstrated potential of Pra-C in improving cognitive function here, our study has also suffered from several limitations, including the lack of a positive control group.In the future, we intend to conduct a more extensive study using galantamine or memantine hydrochloride as positive controls to confirm the effect of Pra-C on cognitive impairment in mice with type 2 diabetes.The doses used in our study were within safety limits.However, the lack of complete toxicity data nevertheless presents another limitation of this study.In summary, our findings demonstrated that Pra-C exhibited therapeutic effects by improving cognitive function by reactivating PI3K/ AKT/GSK3β signaling.Further studies are required to elucidate the mechanisms underlying the regulation of glucose metabolism and neuronal function upon Pra-C treatment in mice.