JDYZF Enhances Cognitive Function in APP/PS1 Transgenic Mice
After 56 days of treatment, Morris water maze (MWM) tests were employed to examine the impacts of JDYZF on learning and spatial memory in AD mice. During the place navigation test, it was observed that the AD group displayed a more erratic swimming trajectory, a prolonged escape latency (p < 0.01), and a reduced number of platform crossings compared to the control group. Conversely, the Donepezil group demonstrated a movement trajectory primarily concentrated in the platform quadrant, with a shorter distance to locate the platform (second quadrant). Meanwhile, the JDYZF group exhibited a disrupted swimming trajectory and a decreased escape latency (p < 0.01) (Fig. 1A). These findings suggest that spatial cognitive impairment was evident in the AD mice, and JDYZF treatment significantly ameliorated cognitive deficits in this population. Throughout the place navigation experiment, latency changes over five consecutive days for each group of rats were recorded, revealing a gradual decrease in escape latency as training time increased (Fig. 1B). In the spatial probe test, the number of times and distance that each group of rats crossed the effective area within 60 seconds were recorded, and representative images are presented in Fig. 1C. Notably, the control group mice crossed the platform most frequently (p < 0.01) in comparison to the model group. Furthermore, both the Donepezil and JDYZF groups demonstrated a significant increase in platform crossings (p < 0.05), with no significant difference between the two treatment groups (Fig. 1D).
JDYZF Reduces Brain Histopathological Changes in APP/PS1 Transgenic Mice
After conducting optical microscopy of H&E-stained brain tissue sections, the impact of JDYZF on brain tissue morphology was investigated. The results presented in Fig. 2 showed that in the Control group mice, the frontal cortex and hippocampus exhibited uniform staining, with normal cell structures, clear cell membranes, nucleoli, and nuclear membranes. Neurons in the hippocampus and frontal cortex were well arranged and showed no evidence of nodules, congestion, or swelling. The cell density was normal, and the arrangement was orderly. However, in the Model group mice, several cellular abnormalities were observed, including condensed nuclei, deep cytoplasm staining, unclear cell and nuclear membranes, reduced cell density, disordered cell arrangement, and noticeable shrinkage, congestion, and swelling. Subsequent to a 56-day drug treatment, both JDYZF and Donepezil were found to significantly prevent these pathological changes. As a result of the treatment, neuronal cells in the JDYZF and Donepezil groups exhibited organized arrangements. In the JDYZF group, there were no signs of shrinkage, congestion, or swelling, and a more orderly arrangement and clearer membrane boundaries were observed. Furthermore, the incidence of condensed nuclei and deep cytoplasm staining was significantly lower in the JDYZF treatment group compared to the Model group.
Neurons contain specific structures called Nissl bodies. To further investigate neurons, Nissl staining was conducted on tissue slices, and the stained samples were then observed through optical microscopy. In the control group, neurons in the hippocampus and frontal cortex of the mouse brain exhibited organised patterns with distinct cell boundaries and intact morphology, as depicted in Fig. 3A. These neurons displayed few instances of missing cells, with the nuclei staining light blue. Additionally, a greater number of Nissl bodies were observed in the cytoplasm, maintaining normal morphology. The Nissl bodies manifested as dark blue granules or patches. In comparison, the cellular arrangement in the hippocampal CA1, CA3, DG regions, and frontal cortex of AD model mice showed a more dispersed distribution, with neurons displaying disorder and the presence of vacuoles in localized areas, indicative of neuronal loss. Conversely, mice treated with JDYZF and Donepezil presented with a more uniform cell arrangement and an increased presence of Nissl bodies within the cytoplasm. Statistical analysis of Nissl bodies in the field of view showed that compared with the model group, the JDYZF and Donepezi groups had an increase in the number of Nissl bodies in hippocampal CA1, CA3, and frontal cortex (Fig. 3B, 3C, 3E), all with statistical differences (p < 0.05); The number of Nissl bodies in the hippocampal DG area of the JDYZF group also increased (p < 0.05), while the number of Nissl bodies in the DG area of the Donepezil group changed (Fig. 3D), but it was not statistically significant (p > 0.05).
JDYZF Treatment Improves the Metabolomic Profiles of Brain Tissue in APP/PS1 Transgenic Mice with AD-related Changes
Improving the Performance of UHPLC-Q-TOF/MS Quality Control
In this study, brain tissues from euthanized mice were collected and subjected to metabolic profiling using UHPLC-QTOF/MS. Firstly, all samples were combined at equal volumes to prepare quality control samples. Subsequently, total ion current (TIC) chromatograms in positive (Fig. 4A) and negative (Fig. 4B) ion modes demonstrated well-overlapped peaks for all QC samples, indicating consistent peak intensities and retention times. Following this, through the analysis of all test samples using XCMS software, a total of 924 positive ions and 539 negative ions were detected. Notably, principal component analysis (PCA) of the metabolites detected in positive and negative ion modes revealed a tight clustering of all QC samples. Additionally, correlation analysis using the Pearson method indicated a high correlation among QC samples (Fig. 4C, D, correlation coefficient > 0.9). These comprehensive results collectively demonstrate the excellent performance of the equipment and the high reliability of the obtained results.
Identification and Statistical Analysis of Metabolites
A comprehensive targeted metabolomic analysis was conducted to gain a clearer understanding of the patterns of metabolite changes in different treatments, using the UPLC-MS platform. A total of 1463 metabolites were identified in the samples, in both positive and negative ion modes (Supplementary Table 1). The identified metabolites were categorized based on their chemical taxonomy, revealing the most abundant class to be lipids and lipid-like molecules, with 380 species, followed by organic acids and derivatives with 350 species. Additionally, 144 organoheterocyclic compounds, 112 organic oxygen compounds, 100 benzenoids, 87 nucleosides, nucleotides, and analogues, 48 phenylpropanoids and polyketides, and 41 organic nitrogen compounds were identified. The proportional representation of metabolites in each subclass is illustrated in Fig. 5A. This detailed analysis provides valuable insights into the specific metabolite classes present and their abundance within the samples.
A total of 300 species of Carboxylic Acids and Derivatives were identified within the organic acid compounds, representing 85.71% of the category. In the category of lipids and lipid-like molecules, the proportional distribution of third-level classifications of metabolites is depicted in Fig. 5B, which includes 121 species of Glycerophospholipids, 108 species of Fatty Acyls, 62 species of Prenol Lipids, 62 species of Steroids and Steroid Derivatives, 20 species of Sphingolipids, 6 species of Glycerolipids, and 1 species of Glycerolipids/Fatty Acyls. Furthermore, the sub-classification of identified lipids and lipid-like molecules is presented in Fig. 5C. Notably, diverse lipids within the Model group and Control group exhibited contrasting trends, indicating different metabolic profiles between C57BL/6 mice and APP/PS1 transgenic mice. Subsequent treatment with JDYZF and Donepezil resulted in noticeable alterations in some distinct lipids, displaying a trend closer to that of the Control group. This pattern was also evident in the analysis of the top 100 significantly different organic acid metabolites across the groups, as depicted in Fig. 5D.
Univariate Statistical Analysis
Univariate statistical analysis methods are commonly relied upon in the comparison of two sample groups, typically employing techniques such as Fold Change(FC) Analysis and t-tests/non-parametric tests. Differential analysis of all identified and unidentified metabolites in both positive and negative ion modes is conducted through univariate analysis. Metabolites with a fold change > 1.5 or FC < 0.67 and a P-value < 0.05 are visually represented using volcano plots, as depicted in Fig. 5. The volcano plots denote metabolites with FC > 1.5 and p < 0.05 in pink, those with FC < 0.67 and p < 0.05 in blue, and non-significant differences in black. Notably, the Control group exhibits a greater number of increased metabolites compared to the Model group (Fig. 6A, B). Conversely, following the administration of Donepezil, there is a decrease in the number of increased metabolites as compared to the decreased ones (Fig. 6C, D). Intriguingly, following the administration of JDYZF, the numbers of both increased and decreased metabolites appear to be more similar, suggesting a significant alteration in the metabolic profile of the brain tissue in AD mice attributable to the drug (Fig. 6E, F).
Multivariate Statistical Analysis
Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) is a supervised multivariate statistical analysis method commonly used for pattern recognition. In this study, OPLS-DA analysis was conducted to investigate the impact of drugs on the metabolites in mouse brain tissue. The analysis encompassed the blank control group (Control, C57BL/6), the model group (Model, APP/PS1 mice), the positive drug group (Donepezil, APP/PS1 mice), and Jiedu Yizhi formula group (JDYZ-H, APP/PS1 mice). The results revealed distinct metabolic profiles between the Model and Control groups in both positive (A: R2X = 0.397, R2Y = 0.996, Q2 = 0.623) and negative (B: R2X = 0.415, R2Y = 0.994, Q2 = 0.571) ion modes (Fig.s 7A, 7B). Similarly, significant differences were observed between the Donepezil and Model groups (Fig.s 7C, 7D), as well as between the JDYZ-H and Model groups (Fig.s 7E, 7F). Notably, minimal intragroup differences were observed, and the group clusters showed distinct separation.
Score plots were generated using OPLS-DA for paired analysis of the Control, Model, Donepezil, and JDYZ-H groups. In this model, R2X and R2Y represent the explained variances of the model for X and Y matrices, respectively, with Q2 indicating the predictive ability of the model. The Q2 values for all compared groups are above 0.3, demonstrating the suitability of the constructed model. Furthermore, the OPLS-DA score plots illustrate distinct separation among the different comparison groups.
Identifying different metabolites through screening
The analysis of OPLS-DA effectively eliminates irrelevant influences and selects differential metabolites. To measure the impact strength and explanatory power of the expression patterns of various metabolites on the classification and discrimination of samples in each group, VIP values are obtained from the OPLS-DA model, with metabolites having VIP > 1 being generally considered to significantly contribute to the model interpretation. The identification of significant differential metabolites involves the application of VIP > 1 and p < 0.05 criteria. To further screen differential metabolites, a combination of univariate analysis and OPLS-DA analysis was utilized, and the results were visually represented using cluster plots and Venn diagrams. The comparison between the model group and the control group revealed the identification of 43 differential metabolites (28 upregulated and 15 downregulated), as evidenced in Table 1. Likewise, in the comparison between the positive drug group and the model group, 12 differential metabolites (4 upregulated and 8 downregulated) were discovered (as shown in Table 2). Furthermore, the comparison between the JDYZ-H group and the model group resulted in the identification of 18 differential metabolites (8 upregulated and 10 downregulated), as illustrated in Table 3.
Table 1
Information on the differential metabolites of Model and Control
ID | Adduct | Name | VIP | P-value | Fold change |
M369T442_1 | [2M-H]- | Phosphoserine | 2.2216 | 0.0028 | 3.3796 |
M214T396 | [M-H]- | sn-Glycerol 3-phosphoethanolamine | 6.6553 | 0.0322 | 2.0913 |
M359T42 | [M-H]- | Ursocholanic acid | 1.8743 | 0.0077 | 2.0294 |
M293T39 | (M-H)- | 9-OxoODE | 1.3677 | 0.0364 | 2.0042 |
M241T102 | [M-H]- | His-ser | 2.4783 | 0.0307 | 1.8386 |
M277T46 | [M-H]- | Pantetheine | 3.3289 | 0.0327 | 1.7854 |
M141T395 | (M-H2O-H)- | 2-Oxoadipic acid | 2.0144 | 0.0383 | 1.6480 |
M133T412 | [M-H]- | Malate | 6.9962 | 0.0042 | 1.2389 |
M223T209 | (M + CH3COO)- | D-Quinovose | 1.1874 | 0.0330 | 1.2386 |
M89T236 | [M-H]- | Dl-lactate | 6.6952 | 0.0483 | 1.2192 |
M119T351_1 | [M-H]- | Tartronate | 1.1051 | 0.0143 | 1.1989 |
M181T347 | [M-H]- | 3,4-dihydroxyhydrocinnamic acid | 1.0818 | 0.0274 | 1.1429 |
M141T348_2 | [M-H]- | Cis,cis-muconic acid | 5.9623 | 0.0124 | 1.1082 |
M269T37 | [M-H]- | Heptadecanoic acid | 1.1345 | 0.0114 | 0.8512 |
M811T192 | [M-H]- | 1-stearoyl-2-arachidonoyl-sn-glycero-3-phosphoserine | 1.3607 | 0.0262 | 0.8202 |
M188T394 | [M-H]- | N-acetyl-l-glutamate | 1.8446 | 0.0243 | 0.8104 |
M181T281 | [M-H]- | 1-methyluric acid | 1.0743 | 0.0249 | 0.7874 |
M253T51 | [M-H]- | Cis-9-palmitoleic acid | 7.1654 | 0.0049 | 0.7255 |
M767T75_2 | [M-H]- | Pg 36:5 | 2.0803 | 0.0116 | 0.6835 |
M367T46 | [M-H]- | Lignoceric acid | 1.0132 | 0.0041 | 0.5666 |
M865T61 | [M + Hac-H]- | Pc(16:1e/17-hdohe) | 1.6816 | 0.0479 | 0.5416 |
M329T172 | M-H | 4,2'-dihydroxy-3,4',6'-trimethoxychalcone | 1.4495 | 0.0129 | 0.4740 |
M699T147 | [M-H]- | (2-aminoethoxy)[2-[hexadec-9-enoyloxy]-3-[octadeca-1.11-dien-1-yloxy] | 1.6639 | 0.0000 | 0.4265 |
M617T173 | [M-H]- | 7-ethyl-10-(4-n-aminopentanoic acid)-1-piperidino)carbonyloxycamptothecin | 2.8112 | 0.0073 | 0.4173 |
M295T523 | [M + Cl]- | Propanoic acid, 3-[[[2-[(aminoiminomethyl)amino]-4-thiazolyl]methyl]thio]- | 5.0798 | 0.0204 | 0.3987 |
M191T502 | [M-H]- | Citrate | 17.2546 | 0.0346 | 0.3551 |
M136T171 | [M + H]+ | Trans-zeatin | 3.5697 | 0.0038 | 0.3800 |
M268T152 | [M + H]+ | Adenosine | 20.5946 | 0.0214 | 0.2380 |
M427T468 | [M + H]+ | L-cysteine-glutathione disulfide | 2.7070 | 0.0261 | 0.2626 |
M136T149_2 | [M + H]+ | Adenine | 8.5663 | 0.0284 | 0.4626 |
M473T36_2 | [M + H]+ | Maslinic acid | 2.0007 | 0.0230 | 0.7796 |
M220T233_1 | [M + H]+ | Pantothenic acid | 2.5093 | 0.0341 | 1.3257 |
M385T34 | [M + HC20H42O4NP]+ | 1-(1z-octadecenyl)-2-sn-glycero-3-phosphoethanolamine | 1.5058 | 0.0111 | 1.4827 |
M401T305 | [M + Na]+ | 5-[5-(acetyloxymethyl)-1,2,4a-trimethyl-7-oxo-3,4,8,8a-tetrahydro-2h-naphthalen-1-yl]-3-methylpentanoic acid | 1.0506 | 0.0126 | 1.6805 |
M146T215_2 | [M]+ | Acetylcholine | 1.8489 | 0.0081 | 1.8557 |
M356T53 | [M + H]+ | Arachidoyl ethanolamide | 1.7890 | 0.0393 | 1.8685 |
M338T35 | [M + H]+ | Erucamide | 2.2709 | 0.0375 | 1.9740 |
M524T182_2 | [M + H]+ | 1-Stearoyl-sn-glycerol 3-phosphocholine(LPC(18:0)) | 11.0870 | 0.0151 | 2.0630 |
M134T211_2 | [M + H]+ | Thiazolidine-4-carboxylic acid | 2.1821 | 0.0055 | 2.7946 |
M763T181 | [M + H]+ | 2-oleoyl-1-palmitoyl-sn-glycero-3-phosphoserine | 1.1181 | 0.0466 | 2.8496 |
M551T35_2 | [M + H-H2O]+ | 1,2-dihexadecanoyl-sn-glycerol | 1.1146 | 0.0472 | 3.4154 |
M704T171 | [M + H]+ | Palmitoyl sphingomyelin | 4.3488 | 0.0307 | 5.3082 |
M814T166 | [M + H]+ | N-nervonoyl-d-erythro-sphingosylphosphorylcholine | 1.7646 | 0.0347 | 12.8660 |
Table 2
Information on the differential metabolites of Donepezil and Model
ID | Adduct | Name | VIP | P-value | Fold change |
M337T33 | [M-H]- | 2-(2-hydroxybut-3-en-2-yl)-3a,6,6,9a-tetramethyl-2,4,5,5a,7,8,9,9b-octahydro-1h-benzo | 1.4785 | 0.0292 | 1.6913 |
M103T195 | [M-H]- | Dl-a-hydroxybutyric acid | 1.8175 | 0.0270 | 1.6551 |
M133T412 | [M-H]- | Malate | 3.6375 | 0.0126 | 0.8387 |
M163T209 | [M-H]- | 1,5-anhydro-d-sorbitol | 1.4189 | 0.0073 | 0.6994 |
M223T209 | (M + CH3COO)- | D-Quinovose | 1.4184 | 0.0043 | 0.6904 |
M147T261 | [M-H]- | 3-hydroxyglutaric acid | 1.4591 | 0.0363 | 0.6481 |
M196T322 | [M + H-C6H7NO3]+ | Cyclopiazonic acid | 1.0166 | 0.0006 | 0.2183 |
M242T257 | [M + H-C7H14]+ | 4-penten-1-one,3,3,4-trimethyl-1-[1-[(tetrahydro-2h-pyran-4-yl)methyl]-1h-indol-3-yl] | 1.3332 | 0.0216 | 0.6724 |
M220T233_1 | [M + H]+ | Pantothenic acid | 3.3468 | 0.0291 | 0.7134 |
M152T251_2 | [M + H-C5H8O3]+ | Deoxyguanosine | 1.4571 | 0.0196 | 0.8642 |
M323T36_2 | [M + H-H2O]+ | 15s-hydroperoxy-11z,13e-eicosadienoic acid | 1.4159 | 0.0470 | 1.1369 |
M348T399 | [M + H]+ | Adenosine 2'-monophosphate | 1.2807 | 0.0343 | 1.3478 |
Table 3
Information on the differential metabolites of JDYZF and Model
ID | Adduct | Name | VIP | P-value | Fold change |
M103T195 | [M-H]- | Dl-a-hydroxybutyric acid | 2.2249 | 0.0016 | 1.7752 |
M737T57 | [M-H]- | Pe(16:1e/14,15-epete) | 1.1088 | 0.0067 | 1.6834 |
M743T37 | [M-H]- | Pg 34:3 | 1.2858 | 0.0001 | 1.3427 |
M129T200 | [M-H]- | Glutaconic acid | 1.7609 | 0.0308 | 1.2970 |
M181T281 | [M-H]- | 1-methyluric acid | 1.0040 | 0.0337 | 1.1408 |
M313T34 | [M-H-C14H10O]- | 1h-imidazo[4,5-c]pyridine-6-carboxylic acid, 1-[[4-(dimethylamino)-3-methylphenyl]methyl]-5-(2,2- diphenylacetyl)-4,5,6,7-tetrahydro | 1.2676 | 0.0475 | 0.8478 |
M796T72 | [M-H]- | Pg 38:5 | 4.2926 | 0.0468 | 0.8086 |
M223T209 | (M + CH3COO)- | D-Quinovose | 1.4213 | 0.0187 | 0.7835 |
M774T39 | [M-H-NH3]- | 1,2-distearoyl-sn-glycero-3-phospho-l-serine | 3.7723 | 0.0121 | 0.7791 |
M163T209 | [M-H]- | 1,5-anhydro-d-sorbitol | 1.6200 | 0.0101 | 0.7733 |
M721T85 | [M-H]- | 1,2-dipalmitoyl-sn-glycero-3-phospho-(1'-rac-glycerol) | 3.2170 | 0.0244 | 0.7344 |
M181T192 | [M-H]- | Hydroxyphenyllactic acid | 1.1827 | 0.0168 | 0.6323 |
M508T187 | [M + H]+ | 1-(1z-octadecenyl)-sn-glycero-3-phosphocholine | 1.0707 | 0.0128 | 0.2628 |
M464T165 | [M + H-NH3]+ | Bis(2,2,6,6-tetramethyl-4-piperidyl) sebacate | 1.0839 | 0.0214 | 0.3533 |
M233T473 | [M + H]+ | N-desbutylbupivacaine | 1.3383 | 0.0346 | 0.9129 |
M231T370 | [M + 2H]2+ | Lys-Trp-Lys | 1.9246 | 0.0354 | 1.1787 |
M167T260 | [M + H]+ | Phenyllactic acid | 1.2265 | 0.0253 | 1.3388 |
M137T318_2 | [M + H-C7H12]+ | Siduron | 2.2526 | 0.0017 | 2.0217 |
(Note: In mass spectrometry analysis, the metabolite's identification number (ID) is represented as a unique identifier. Information related to the compound's adduct ion is denoted by the term "adduct." The metabolite's name is referred to as "Name". VIP is a measure reflecting the significance of a variable, with higher values indicating greater importance. FC represents the ratio of abundance levels between two conditions. The p-value is an indicator of the significance level of the analysis, with smaller values suggesting a higher significance of the observed difference) |
The identification of common substances across different groups is facilitated by visualizing the differential metabolites among the comparison groups (Fig. 8A). Notably, three substances—Malate, Pantothenic acid, and D-Quinovose—show significant changes among the Control, Model, and Donepezil groups. Likewise, among the Control, Model, and JDYZ-H groups, two substances—1-methyluric acid and D-Quinovose—demonstrate significant alterations. A visual representation of the number of upregulated and downregulated metabolites in each comparison group is provided in Fig. 8B.
The association analysis of the significant differential metabolites among the groups revealed three substances of significance among the Control, Model, and treatment groups. Malate increased in the Model group in comparison to the Control group (p < 0.01), but the quantity of Malate decreased in the Positive Drug group compared to the Model group (p < 0.05, Fig. 9A). 1-methyluric acid exhibited decreased expression in the Model group, whereas it showed increased expression in the Control and Treatment groups (Fig. 9B). Furthermore, the metabolite D-Quinovose showed an elevation in the Model group but a significant decrease in the Control, Positive Drug, and Treatment groups (Fig. 9C).
Analysis of Differential Metabolite Bioinformatics
Upon the identification of significantly different metabolites meeting both OPLS-DA VIP > 1 and P-value < 0.05 criteria, subsequent bioinformatics analyses were conducted. These analyses comprised cluster analysis, correlation analysis, pathway analysis, and other related content for a comprehensive examination. In order to provide a more comprehensive and intuitive representation of the relationships between samples and the differential expression patterns of metabolites across different samples, a distance matrix calculation was performed using all samples and differential metabolite expression levels. Subsequently, hierarchical clustering analysis was employed, involving the calculation of distances between samples or metabolites, followed by iterative merging of the two closest clusters until all clusters were combined into one cluster. The hierarchical clustering analysis of significant differential metabolites (VIP > 1, p < 0.05) is illustrated in Fig. 10. In Fig. 10A, 13 clusters were formed among the differential metabolites between the Control group and Model group, indicating similar expression patterns within the same group. Additionally, Fig. 10B illustrates 3 clusters formed among the differential metabolites between the Donepezil group and Model group, suggesting similar functionalities. Lastly, Fig. 10C displays 6 clusters among the differential metabolites between the JDYZF group and Model group, which collectively indicate the involvement of grouped metabolites in the same metabolic processes.
Correlation analysis is used to quantify the metabolic proximity between significantly different metabolites (VIP > 1, p < 0.05). It provides insights into the mutual regulatory relationships among metabolites during biological state changes. Metabolites with expression correlations often participate together in specific biological processes, indicating functional relevance. Furthermore, different metabolites can exhibit synergistic or mutually exclusive relationships, as illustrated by the positive and negative correlations. A positive correlation (r-value between 0 and 1) indicates that the trend of change in one class of metabolites is the same, whereas a negative correlation (r-value between − 1 and 0) suggests an opposite trend of change. The strength of the correlation is determined by the absolute value of r-values closer to 1 indicate a stronger correlation, while values closer to 0 indicate a weaker correlation. Positive correlations among metabolites may imply that they originate from the same biosynthetic pathway, while negative correlations may indicate potential decomposition for the synthesis of other metabolites, illustrating synthetic transformation relationships. The Pearson correlation coefficient is employed for this analysis as it measures the linear correlation between two quantitative variables. Specifically, the Pearson product-moment correlation coefficient assesses the degree of linear correlation between two variables.
The correlations between the differential metabolites in the Control and Model groups are depicted in Fig. 11A. The highest correlation was observed between 1,2-dihexadecanoyl-sn-glycerol and Palmitoyl sphingomyelin (r = 0.97). Additionally, 2-oleoyl-1-palmitoyl-sn-glycero-3-phosphoserine exhibited strong correlations with 1,2-dihexadecanoyl-sn-glycerol, Palmitoyl sphingomyelin, Erucamide, and N-nervonoyl-d-erythro-sphingosylphosphorylcholine, with correlation coefficients of 0.95, 0.95, 0.94, and 0.94, respectively. Furthermore, LPC (18:0) was found to be correlated with Arachidoyl ethanolamide (r = 0.93). Notably, 1,2-dihexadecanoyl-sn-glycerol exhibited a significant negative correlation with Maslinic acid (r =-0.92). These correlations provide valuable insights into the relationships between the metabolites in the Control and Model groups. Figure 11B illustrates the correlations between the differential metabolites in the Donepezil and Model groups. D-Quinovose exhibited the strongest correlation with 1,5-anhydro-d-sorbitol and Cyclopiazonic acid, with r-values of 0.95 and 0.83. Additionally, the correlation between 1,5-anhydro-d-sorbitol and Pantothenic acid was also found to be significant (r = 0.81). The correlations between the metabolites in the JDYZF group and the Donepezil group can be observed in Fig. 11C. Notably, deoxyguanosine demonstrates a strong correlation (r = 0.99) with His-Lys, while methylmalonic acid exhibits a high correlation (r = 0.98) with Propionic acid. Additionally, hypoxanthine is moderately correlated (r = 0.86) with Malate, and Ng,ng-dimethyl-l-arginine displays a moderate correlation (r = 0.83) with Prostaglandin F2alpha. These findings underscore the interrelated nature of the differential metabolites in both groups.The correlations between the differential metabolites in the JDYZF group and the Model group are detailed in Fig. 11D. The most substantial correlation is observed between 1,5-anhydro-d-sorbitol and palmitoyl sphingomyelin, with an r-value of 0.95. Subsequently, a strong correlation of 0.84 is found between 1,2-dipalmitoyl-sn-glycero-3-phospho and Pg 38:5. Moreover, a correlation with an r-value of 0.81 is identified between 1-(1z-octadecenyl)-sn-glycero-3-phosphocholine and 1,2-dipalmitoyl-sn-glycero-3-phospho-l-serine. Further, a correlation of 0.78 is noted between 1h-imidazo[4,5-c]pyridine-6-carboxylic acid and Pg 38:5, while 1-methyluric acid is correlated with Pg 34:3 with an r-value of 0.77.
The essay paragraph has been revised to improve logical accuracy, detail, and academic style: The MetaboAnalyst website was used to import differential metabolites from various comparison groups and match them with the KEGG database to obtain pathway information. Subsequently, enrichment analysis was conducted on the annotated results, identifying pathways with a significant enrichment of differential metabolites (p < 0.05). As depicted in Fig. 12A, the Model group and Control group exhibited 15 significantly annotated and enriched pathways. Notably, the Glycerophospholipid metabolism pathway, Citrate cycle pathway, cAMP signaling pathway, and Glucagon signaling pathway were found to be the most significant (p < 0.01). For the Donepezil group and Model group (Fig. 12B), the significant differential metabolites were primarily annotated and enriched in the Renal cell carcinoma pathway, Proximal tubule bicarbonate reclamation, Citrate cycle (TCA cycle) pathway, and Glucagon signaling pathway. In the JDYZ-H group and Model group (Fig. 12C), the significant differential metabolites were predominantly annotated and enriched in Glycerophospholipid metabolism, Caffeine metabolism, Choline metabolism in cancer, Cholinergic synapse, and Sphingolipid signaling pathway. Notably, some metabolic pathways overlap across these comparison groups, for instance, Glycerophospholipid metabolism and Caffeine metabolism pathways are present in the Control group, Model group, and treatment groups.