{"paper_id":"0ac4eac9-178e-4196-8400-0a6eac276511","body_text":"Construction and validation of a bioinformatics-based screen for Cuproptosis-related genes and risk model for Alzheimer's disease | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Construction and validation of a bioinformatics-based screen for Cuproptosis-related genes and risk model for Alzheimer's disease Rui Hu, Zhen Xiao, Mingyu Qiao, Chaoyu Liu, Guiyou Wu, Yunyi Wang, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3854023/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract This study aimed to validate the correlation between core cuproptosis genes (CRGs) and Alzheimer's disease (AD) from both bioinformatics and experimental perspectives and also to develop a risk prediction model. To this end, 78 human-derived temporal back samples were analyzed in GSE109887, and then the biological functions of the resulting CRGs were explored by cluster analysis, weighted gene co-expression network analysis (WGCNA), and similar methods to identify the best machine model. Moreover, a nomogram was developed to validate the model. The mRNA and protein expression of CRGs were validated using the SH-SY5Y cell model and SD rat animal model. The RT-qPCR and western blot results showed that the mRNA and protein expression content of DLD, FDX1, GLS, and PDHB decreased, and the DBT expression content increased in AD, which supported the bioinformatic analysis results. CRGs expression alterations affected the aggregation and infiltration of certain immune cells. The study results also confirmed the accuracy and validity of AD diagnostic models and nomograms. This study validated the correlation between five CRGs and AD, indicating a significant difference between AD patients and healthy individuals. Therefore, CRGs are expected to serve as relevant biomarkers for the diagnosis and prognostic monitoring of AD. Alzheimer's disease1 Cuproptosis2 immune infiltration3 machine learning model4 Nomogram5 Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Introduction Alzheimer's disease (AD, also known as dementia) is one of the most prevalent neurodegenerative diseases that imposes a heavy economic burden on societies around the world[ 1 ]. The disease is characterized by age spots, and senile plaque (SP), neurofibrillary tangle (NFT), and neuronal loss are among its major pathological symptoms[ 2 ]. Many studies have addressed the pathogenesis of AD and reported that the primary pathogenic mechanisms in the pathogenesis of AD that have significantly influenced clinical treatments are Aβ plaque-associated neurodegeneration, neurofibrillary tangle-based neuro-progenitor degeneration, mitochondrial dysfunction, and gene mutation hypothesis[ 3 , 4 ]. Because of the correlation between these pathogenic mechanisms, some researchers have proposed that several mechanisms may be involved in the development of AD. One of the major areas of research in this field is to investigate these mechanisms to predict the risk of AD[ 5 ]. Another area of research is the study of the genome as a predictor of AD risk. The genome provides important insights for investigating the molecular mechanisms of AD, and it is a potent tool for predicting the risk of the disease. Therefore, it is necessary to develop an AD risk prediction model based on the exploration of molecular mechanisms. Cuproptosis is an essential element that plays an important role in several metabolic processes in the brain[ 6 , 7 ]. Excess cuproptosis in the brain of AD patients exacerbates oxidative damage and increases the formation of amyloid plaques and neurofibrillary tangles (NFTs)[ 8 , 9 ]. The neurotoxic effect of cuproptosis is closely related to the inherent redox properties of Cu 2+ . Cuproptosis overload promotes cuproptosis[ 10 ], and cuproptosis, as a newly discovered independent mode of cell death characterized by cuproptosis-dependent and mitochondrial respiratory regulation, has a significant impact on mitochondrial functions[ 11 , 12 ]. However, numerous studies have concluded that mitochondrial dysfunction-induced oxidative stress and abnormalities in energy metabolism account for a major pathogenic mechanism involved in the development of AD[ 13 , 14 ]. Cuproptosis ions and cuproptosis are closely related to mitochondrial function[ 15 , 16 ], and recent studies have investigated the potential correlation between cuproptosis and AD[ 17 ]. There are relatively few studies exploring the potential correlation between cuproptosis and AD from the perspective of bioinformatics to develop a prognostic risk model. Therefore, this study aims to investigate the correlation between CRGs and AD and develop a machine learning-based prognostic risk model for AD. Another objective of this study is to develop a cellular model with Okadaic acid (OA)-treated SH-SY5Y cells to validate the expression of the resulting CRGs by the Aβ 25−35 dementia model in simulated rats in order to propose a new approach to the diagnosis and prognosis of AD. Materials and methods Data collection. The gene expression datasets used in this study are all accessible via the GEO public database ( https:www.NCBI.NLM.nih.gov/GEO/ ). This includes the GSE109887 dataset based on the GPL10904 platform, which contains 46 AD samples and 32 control samples (all from human temporal regression samples)[ 18 ], and the GSE33000 dataset based on the GPL 10558 platform, which contains 310 AD samples and 157 control prefrontal cortex samples[ 19 ]. A total of 16 CRGs were obtained from the KEGG public website ( https://www.kegg.jp/pathway ). Differential expression analysis and data processing. Data analysis was performed on the above datasets in R 4.2.1. The sva R package was used to remove batch effects and the limma R package was employed for differential expression analysis[ 20 , 21 ]. All data sets were then normalized in the preprocessCore R package, and differential expression analysis was performed in the limma R package. The threshold for DEG was set at | log-fold change (FC)| > 1, P < 0.05. The screened differential genes were presented using heatmaps and volcano plots, produced using the R packages ggplot2 and pheatmap. Enrichment analysis of core cuproptosis genes. Gene Ontology (GO) and KEGG pathway analyses were performed on all DEGs as well as five cuproptosis-related hub genes in the ClusterProfiler package to identify the relevant pathways and biological functions influenced by differentially expressed genes. The significance level was considered to be P < 0.05. Evaluation of immune cell infiltration. The relative abundance of 22 types of immune cells in each sample was estimated based on gene expression data in the CIBERS0RT R package[ 22 ]. The results of the correlation between CRGs and infiltrating immune cells were presented using the \"ggplot2\" R package. Unsupervised cluster analysis of AD samples. Based on the expression profiles of CRGs, an unsupervised cluster analysis was performed in the \"ConsensusClusterPlus\" R package to classify the AD samples into different clusters using a k-means algorithm. The optimal number of clusters was evaluated by combining the maximum number of subtypes k (k = 6) (see figure), the cumulative distribution function (CDF) curve, the consistency matrix, and the consistency clustering score (> 0.9). Weighted gene co-expression network analysis (WGCNA). The WGCNA R package was employed to develop co-expression networks and modules in healthy individuals and AD patients to identify key gene modules associated with AD. The top 50% of genes with the highest variance were applied to subsequent WGCNA analyses to ensure the accuracy of qualitative results. A total of 10 co-expression modules with different colors were obtained using a dynamic cutting algorithm, and a heatmap of the topological overlap matrix (TOM) was also presented, in which gene dendrograms and module colors were generated based on gene dissimilarity. The global gene expression profiles in each module are represented by module signature genes. Then an appropriate set of genes for each module is filtered by an algorithm to intersect with the set of differential genes to show the correlation. GSVA analysis. GSVA enrichment analysis was performed in the \"GSVA\" R package to elucidate the signaling pathways that are differentially enriched between different groups of CRGs.[ 23 ]. The files \"c2.cp. Kegg.v7.4.symbols\" and \"c5.go.bp.v7.5.1.symbols\" were obtained from the MSigDB web database for further GSVA analysis. The \"limma\" R package was utilized to compare the GSVA scores between different clusters of CRGs to identify the differentially expressed pathways and biological functions. Nomogram model development and validation. A nomogram model was developed for the clinical diagnosis of AD in the \"RMS\" R package. Each predicted gene received a score, and the \"total score\" represented the sum of the above predictor scores. Calibration curves and DCA were used to estimate the predictive power of the nomogram model. Machine learning and prognostic modeling. Models such as the Random Forest Model (RF), Support Vector Machine Model (SVM), Generalized Linear Model (GLM), and Extreme Gradient Boosting (XGB) were developed using the differentially expressed genes of two different clusters of CRGs and the intersecting genes of AD differentially expressed genes. The best machine learning model was identified based on the highest ROC values of the above models; accordingly, the SVM and the first five important variables were considered the key predictive genes associated with AD. Finally, the diagnostic value of the model was validated by assessing its accuracy by using the GSE33000 dataset. Animal sources and construction of AD models. Thirty male SD rats, SPF-grade, 8 months old, weighing 220–240 g, were purchased from Guangdong Viton Lihua Laboratory Animal Technology Co. to be used as subjects. The subjects were housed, modeled, and administered in an SPF-grade laboratory at the Animal Experimentation Center of Youjiang Medical College of Nationalities. The SPF-grade laboratory was equipped with a 12 h/12 h alternating light and dark barrier system, and its ambient temperature and relative humidity were 25℃ and 60–70%, respectively. In addition, the subjects were fed until December for subsequent experiments. They also passed the Right River College of National Medicine's ethical review and examination for the use of experimental animals (Ethical Review Approval No. 2023040101), and all relevant guidelines were properly followed throughout the experiment. The subjects were randomly divided into the model, saline, and normal control groups (10 rats in each group). After weighing the subjects, they were anesthetized with isoflurane (RWD, CAT. NO.hsc-454-5) using a Reward R500 general-purpose small animal anesthesia machine. They were then placed on a sterile surgical table, and the head was fixed with a stereotaxic apparatus to monitor the respiratory rate. If the respiratory rate remained stable, a stereotaxic injection was performed after the subject's head was fixed appropriately. Subjects in the model group were treated with 1 µl of Aβ 25−35 on each side of the hippocampus (Glpbio, CAT. NO.GP10082)[ 24 , 25 ], whereas those in the saline group were injected an equal amount of normal saline into each side of the hippocampus. Those in the normal control group received no surgical treatment. In addition, 100,000 units/g/day of Penicillin was injected into each subject for 7 days after surgery to prevent postoperative infections. Morris water maze (MWM) test. The MWM test was performed using the DigBehv-MG model device, Shanghai Gizo Software Technology Co., to assess the spatial learning and memory of subjects[ 26 ]. The localization and navigation test lasted for 4 days, and each subject was trained 4 times per day. The subjects were submerged from the entry point that faces the pool wall, and the time it took them to find the platform within one minute was recorded (escape latency). If a subject failed to find the platform within one minute, the subject was instructed to stand on the platform and remain there for 10 seconds in order to achieve the training purpose; the escape latency for that subject was considered to be one minute. Two days after the localization and navigation tests finished, the platform in the pool was removed and the furthest quadrant of the original platform was taken as the entry point. Then the subjects in both groups were placed into the water while facing the wall from the entry point, and the escape latency was recorded within one minute. Moreover, the subjects were placed into the water from the entry point toward the wall of the pool, and the swimming time and the number of times they crossed the original platform within one minute were recorded. All subjects were acclimatized overnight before the experiments. RNA extraction and qRT-PCR to verify the differential expression of CRGs mRNA in AD rat model. The day after the end of the trans-spatial memory ability test, the subjects were anesthetized to remove their brain tissue; the tissue samples were weighed and then they were stored in a refrigerator at -80°C. To extract RNA from the brain tissue samples, 30 mg of weighed brain tissue samples were taken from each group. Then, the centrifuge tubes were filled with the appropriate amount of TRIzol reagent (Thermo Fisher Scientific, CAT. NO.15596026) to extract the total RNA samples based on the mass of the weighed brain tissue samples. The Third Generation Variable Speed Tissue Mill (TGrinder, China) was used to grind the brain tissue samples in the centrifuge tubes until the brain tissue was totally dissolved in the TRIzol reagent. Then they were left on ice for 5 minutes for subsequent operations. RNA was reverse-transcribed into cDNA using ToloScript All-in-one RT EasyMix for qPCR kit (TOLOBIO, CAT. NO.22107), and the results were analyzed in a Light Cycle 96 real-time fluorescence PCR instrument (Roche, Germany) using SYBR Green qPCR Mix kit (TOLOBIO, CAT. NO.22204-1). The mRNA expression of the genes related to cuproptosis was determined in a Light Cyle 96 real-time fluorescence quantitative PCR instrument (Roche, Germany). GAPDH was the endogenous control gene, and the primers used for amplification were as follows. GAPDH-F:5'-GACATGCCGCCTGGAGAAAC-3 ' GAPDH-R :5 '-AGCCCAGGATGCCCTTTAGT-3 ' DLD-F:5 '-CATTTCAATCGGCTGTCTCA-3 ' DLD-R:5 '-CAAGCATGTTCCTCCTAGTGTT-3 ' GLS-F:5 '-CTGAACGAGAAAGTGGAGACCGAA-3 ' GLS-R:5 '-TGGGCAGAAACCGCCATTAG-3 ' PDHB-F:5 '-GCATTTGAACTTCCCACAGA-3 ' PDHB-R:5 '-TTCCCTCCTTAGACAATACAGC-3 ' DBT-F:5'-ACGTGTGCTCTCTGTGGGGTTATC-3 ' DBT-R:5 '-GCTGTCAAACTGAGACACCG-3 ' FDX1-F:5 '-TGGTGAAACGCTAACGACCA-3 ' FDX1-R:5 '-CAAGCCAAAGTCCCCTCACA-3 ' Protein extraction and WB verification of differential expression of CRGs in rat model of AD. A sample of 50 mg of brain tissue was taken from each group. A third-generation variable-speed tissue grinder (TGrinder, China) was used to grind the brain tissue samples in centrifuge tubes until they were completely dissolved in the RIPA lysate. They were then lysed on ice for 30 minutes, centrifuged at 12,000 g for 30 minutes at 4°C, and the supernatant was collected. All of these steps were done under the mass of the weighed rat brain tissue samples. The protein concentration was determined by the BCA protein assay. Separation gels were prepared according to the molecular weight of the target protein. The samples were boiled in the 4×protein-sampling buffer for 5 minutes and uploaded onto the gel. The upper gel was electrophoresed at 80 V for 30 minutes, while the lower gel was electrophoresed at 120 V for 90 minutes. A current of 250 mA was applied for 80 minutes for the membrane transfer process. After the transfer was completed, the membrane was sealed with protein-free rapid closure solution (1x) for 20 minutes, washed three times with TBST (20xTBST buffer 50 ml + 950 ml purified water) for 10 min each time, and incubated overnight at 4°C with the following primary antibodies: Anti-GAPDH, 36 kDa, 1:5000( Proteintech, CAT. NO.10494-1-AP); Anti-KGA/GAC (GLS). 65kDa, 1:5000( Proteintech, CAT. NO.12855-1-AP); Anti-DLD,56kDa, 1:5000( Proteintech, CAT. NO.16431-1-AP); Anti-PDHB,34kDa,1:5000 ( Proteintech, CAT. NO.14744-1-AP); Anti-DBT,53kDa,1:3000 ( Proteintech, CAT. NO.12451-1-AP). Then the membrane was washed three times with TBST for 10 minutes each time. The secondary antibody was incubated at a dilution ratio of 1:5000 for 1 hour at room temperature, and then it was washed three times with TBST for 10 min each time. Each membrane was placed in a fully automated chemiluminescence gel imager for color development after 200 µl of the ultrasensitive luminescence solution (A: B formulated at 1:1) (Monad, CAT. NO.PW30701S) was prepared. GAPDH was selected as the endogenous control protein. Cell culture and AD model development. SH-SY5Y cells were obtained from Wuhan Prosperity Life Science Co. Ltd, China, okadaic acid (OA) for AD cell modeling was purchased from MCE[ 27 ], and the cells were purchased from MCE. SH-SY5Y cells were incubated in a cell culture containing 5% C0 2 at 37°C. The medium used for the culture of cells was a combination of Dulbecco Modified Eagle medium (DMEM/F-12) with F-12, 14% Sigma Australia Fetal Bovine Serum (FBS), and a 1% Penicillin-Streptomycin Solution mixture. Dimethyl sulfoxide was used to prepare a 4 µmol/L OA master mix, and the resulting solution was then stored at 4°C. Using the above complete medium, SH-SY5Y cells were diluted to a density of 1×10 5 mL and inoculated with 200 µL per well in 96-well plates. Each well received a full medium supplemented with OA solutions at concentrations of 0, 10, 20, 40, and 80 nmol/L after 24 hours. There were four replicates for every concentration. To compare the effects of different OA concentrations on the proliferation of SH-SY5Y cells 24 hours later, CCK-8 reagent (20 µL) was added to each well and incubated for 2 hours in exposure to 5% CO 2 at 37°C. The absorbance values at 450 nm were measured using a BIOBASE-EL10A enzyme marker (BIOBASE, Shandong, China)[ 28 ]. The experiment was conducted three times, and the optimal drug concentration of 25 nM for OA-induced AD in this cell model was found by monitoring the cell growth under a microscope and calculating the absorbance values. Then 200 µl of SH-SH5Y cells was added to each well of 96-well plates, resulting in a density of 1×10 5 mL, and the plates were incubated for 24 hours. The cell model was then incubated for 2 hours at densities 0 and 25 nM. The medium was replaced with another one containing 0 and 25 nM of OA and incubated again for 24 hours. The cells were incubated for 24 hours with 20 µl of OA per well. Then 20 µl of CCK-8 reagent was added to each well, and the plates were incubated for 2 hours in exposure to 5% C0 2 at 37°C. Absorbance values at 450 nm were determined using the BIOBASE-EL10A enzyme marker. The experiment was repeated three times in four replicates to obtain the cell survival index at a drug concentration of 25 nM. RNA extraction and qRT-PCR to verify the differential expression of CRGs mRNA in SH-SY5Y cells. Total RNA samples were extracted using TRIzol reagent. To this end, TRIzol reagent was added to cell samples of the model group, which were induced with 25 nM of OA, and cell samples of the control group, which received no treatment. The subsequent operation was the same as the animal part. GAPDH was selected as the endogenous control gene, and the primers were amplified as follows. GAPDH-F:5'-CAGGAGGCATTGCTGATGAT-3 ' GAPDH-R:5 '-GAAGGCTGGGGGCTCATTT-3 ' DLD-F:5 '-GTGATTTACACACACACCCTGA-3 ' DLD-R:5 '-GTCTGTCGATTTCTGCCCAA-3 ' GLS-F:5 '-CTGAGCCCTGAAGCAGTTCG-3 ' GLS-R:5 '-AGGAGACCAGCACATCATACCC-3 ' PDHB-F:5 '-GTGTCTGGCTTGGTGCGGAG-3 ' PDHB-R:5 '-ACCTTGTATGCCCCATCATACTG-3 ' DBT-F:5'-CCATTGCATTTGCTCGTGGA-3 ' DBT-R:5 '-ACCCTGCTCAGTATCCATTGC-3 ' FDX1-F:5 '-CTTGTTCAACCTGTCACCTCATCT-3 ' FDX1-R:5 '-CAGCCACTGTTTCAGGCACTC-3 ' Protein extraction and WB verification of differential expression of CRGs in SH-SY5Y cells. A protein blotting assay was performed to assess the expression of proteins in different groups of cells, with GAPDH as the endogenous reference protein. The culture flasks were washed with pre-cooled PBS, and the cells were collected by using a cell spatula and then centrifuged in a 15-ml tube at 1200 rpm for 5 minutes. The centrifuge tube was filled with 100:1:1 RIPA tissue/cell lysate, protein phosphatase inhibitor mixture, and protease inhibitor and then lysed on ice for 30 minutes. Subsequent operations were the same as those in the animal part. Statistical results. The obtained data were processed and statistically analyzed in R -4.2.1, SPSS-25, and GraphPad Prism-9. Analysis of variance (ANOVA) or t-test was used to determine whether there were significant differences between subgroups in DEG mRNA levels. The t-test was employed to compare RNA and protein expression levels, cell viability, and other variables. In addition, the MWN was analyzed using repeated-measures ANOVA. Differences at P < 0.05 were considered statistically significant. Ethical statement . The study is reported in accordance with ARRIVE guidelines. Results Identification of differentially expressed genes . This study analyzed gene expression array data from 78 human-derived temporal regression samples from the GEO database; Fig. 1. A presents the volcano plots of the 78 DEGs. A total of five differentially expressed CRGs were identified in this study, including one upregulated and four downregulated expressed genes(Figure 1. B). Figure 1. C shows the heatmap of the five differentially expressed related genes. Biological function analysis of cuproptosis-related core genes . The obtained differentially expressed CRGs were analyzed for their biological functions. GO analysis showed that the differentially expressed CRGs were mainly enriched in the cellular catabolic/metabolic processes of amino acids, and the cellular fractions were mainly enriched in the mitochondrial matrix oxidoreductase complex. In terms of molecular functions, the oxidoreductase activity, which acts on the donor's aldehyde or carbonyl group and the acceptor's NAD or NADP, as well as the oxidoreductase activity, which acts on the donor's aldehyde or carbonyl groups, were both affected (Fig. 2. A). KEGG enrichment analyses revealed that differentially expressed CRGs were mainly involved in lipoic acid metabolism, the citric acid cycle (TCA cycle), propionic acid metabolism, pyruvate metabolism, valine, leucine and isoleucine degradation, glycolysis, and carbon metabolism (Fig. 2. B). Correlation analysis of cuproptosis-related core genes . Subsequently, the obtained CRGs were correlated to explore whether they play a synergistic role in the progression of AD. The results indicated that some of the genes, such as PDHB and DLD, exhibit synergistic effects on A. However, DLAT and the remaining four genes showed significant antagonistic effects. In addition, further investigation of the correlation patterns of these CRGs revealed that both DLAT and DLD were significantly correlated with other modulators (Figure 2. C, D). Immune infiltration cell analysis of cuproptosis-related core genes . Immune infiltration analysis was performed based on the CIBERS0RT algorithm, and the results showed differences between AD patients and healthy individuals in the proportion of 22 infiltrating immune cell types. The heatmap shows the distribution of the 22 infiltrating immune cell types (Figure 3. A). The results indicated that AD patients exhibited higher levels of infiltration in monocytes and neutrophils compared to controls (Figure 3. B). The correlation between the five differentially expressed CRGs and infiltrating immune cells demonstrated that DLD was positively correlated with dendritic-activated cells, M2 macrophages, and mast-activated monocytes, whereas PDHB was basically the same as DLD in the expression of immune correlations. Moreover, GLS showed a positive correlation with dendritic-activated cells (Figure 3. C). These results suggest that CRGs may be a key factor in regulating the molecular and immune infiltration status in AD patients. Characterization of epithelial cell aggregation in AD patients . To elucidate the expression patterns associated with cuproptosis in AD, 78 samples were grouped based on the expression profiles of the five CRGs using a consistent clustering algorithm. The number of clusters was most stable when the k-value was set to 2 (k=2) and the CDF curves fluctuated within the smallest range of the common index (0.2 to 0.6) (Fig. 4. A). When the k-value ranged between 2 and 9, the area under the CDF curve exhibited a difference between the two CDF curves (k and k-I) (Fig. 4. B). Clustering the k-values ranging from 2 to 9 showed a concordance score of >0.9 across subtypes only when the k-value was equal to 2 (Fig. 4.c). Characterization of differentiation and immune infiltration of regulatory factors in two clusters of cuproptosis . Combined with the heatmap of the shared matrix, 78 AD patients were finally grouped into two clusters, including Cluster 1 (n=46) and Cluster 2 (n=32). To compare the molecular features of the clusters, the expression differences of the five CRGs between Cluster 1 and Cluster 2 were first thoroughly assessed (Fig. 5. A). Different expression profiles of CRGs were observed between the two cuproptosis intoxication patterns, with Cluster 1 exhibiting high expression levels of DBT and Cluster 2 characterized by increased expression of DLD, FDX1, GLS, and PDHB (Figure 5. B). The immune infiltration analysis revealed a difference between the clusters in the immune infiltration profile (Figure 5. C); the percentage of macrophage M1 was higher in Cluster 1, whereas the percentage of dendritic activated cells was higher in Cluster 2 (Figure 5.D). Development of a co-expression networks . The WGCNA algorithm was used to develop co-expression networks and modules in healthy individuals and AD patients to identify key gene modules associated with AD. After calculating the variance of each gene expression in GSE 109887, the top 25% of genes with the highest variance were selected for further analysis. Co-expressed gene modules were identified when the value of soft power was set to 11 and the scale-free R2 was equal to 0.5 (Figure 6. A). A total of 10 gene modules with different colors were obtained using the dynamic cutting algorithm, and then a heatmap of the topological overlap matrix (TOM) was drawn (Fig. 6. B-D). These genes were then used in each of the ten color modules in turn to examine the degree of similarity and proximity of module-clinical feature (control and AD) co-expression; the turquoise module demonstrated the strongest correlation with AD (Fig. 6. E). In addition, a positive association was observed between the turquoise module and module-associated genes (Fig. 6. F). A similar approach was also employed to further analyze the differentially expressed genes between the two cuproptosis clusters. β=19 and R2=0.8 were screened as the most suitable soft threshold parameters for developing the scale-free network (Fig. 7. A). The heatmaps demonstrated the TOM of all module-related genes (Fig. 7. B-D). An analysis of the relationship between modules and clinical features (Clusters 1 and 2) revealed a strong correlation between turquoise modules and AD clusters (Fig. 7. E). The correlation analysis also showed a significant relationship between the turquoise module genes and selected modules (Fig. 7. F). Identification of specific genes and GSEA . A total of 134 related genes were identified by analyzing the crossover between the modular genes associated with AD (Fig. 8. A). GSVA analysis was used to further investigate the functional differences between the two clusters in terms of cluster-specific DEGs. Cluster 1 was found to be enhanced in non-small cell lung cancer, glycosaminoglycan biosynthesis of keratan sulfate, basal cell carcinoma, Kegel's chronic granulocytic leukemia, primary immunodeficiency, drug metabolism of cytochrome P450, isobiotic metabolism by cytochrome P450, and Notch signaling activity (Fig. 8. B). Therefore, it was concluded that it may be involved in various immune responses as well as metabolism. Development and evaluation of machine learning models . To identify specific genes with high diagnostic value, four validated machine learning models were developed based on the intersection of 134 cluster-specific DEG and AD differentially expressed genes. These four models were Random Forest Model (RF), Support Vector Machine Model (SVM), Generalized Linear Model (GLM), and Extreme Gradient Boosting (XGB). The \"DALEX\" software package was employed to plot the residual distributions of each model in the test dataset. The SVM and RF models presented relatively low residuals (Fig. 9. A, B). The top 15 significant feature variables of each model were then ranked based on the root mean square error (RMSE) (Fig. 9. C). The discriminative performance of the four machine learning algorithms in the test dataset was evaluated by calculating the subject operating characteristic (ROC) curves based on a 5-fold cross-validation. The SVM showed the highest area under the ROC curve (AUC); the AUC of other models was as follows: GLM=0.585; SVM=0.803; RF=0.761; and XGB=0.624 (Figure 9.D). Combined with these results, the SVM was shown to distinguish well between patients belonging to different clusters. Finally, the top five SVM variables (CACNG3, TNS1, TCEAL2, FAM65C NPTN, and THY1) were selected as predictor genes for further analysis. Nomogram construction and validation of predictive modeling accuracy . A nomogram was developed to estimate the risk of cuproptosis in 78 AD patients to further generalize the RF model to clinical applications (Fig. 10. A). Correction curves and decision curve analysis (DCA) were then employed to evaluate the predictive efficiency of the nomogram model. The calibration curve showed that there was a small error difference between the actual AD aggregation risk and the predicted risk (Fig. 10. B). Since the nomogram’s calibration curve confirmed its high accuracy, it can be used as a basis for clinical decisions (Fig. 10. C). The gene prediction model developed in this study was validated on two external tissue datasets, and the ROC curves showed that it performed well in the GSE 33000 dataset, with an AUC of 0.837 (Fig. 10.D). These results suggest the outstanding diagnostic power of the prediction model. Characterization of CRGs expression in animal models of AD . After the MWM confirmed the model group's modeling, hippocampal tissue was collected for qRT-PCR and Western blot experiments to verify CRG expression. No animals perished during the MWM after modeling ten rats in each group, and the analysis revealed that five modeling attempts were successful for the model group (n=5). The MWM was used to test the effect of Aβ 25-35 on cognitive function in SD rats. Figure 11(A) shows the distance traveled by each group of rats in the MWM. The 4-day localization cruise experiment showed that the mean swimming speed of rats in the AD group (treated with Aβ 25-35 ) reduced and the avoidance latency increased (Figure 11. B-C, *p<0.05, **p<0.01, *p<0.001). The spatial exploration experiment also indicated that the rats in the model group traversed the platform less frequently than those in the control and saline groups (Fig. 11.D). These results suggest that the memory function was impaired in the AD group. The qRT-PCR showed that the relative expression of DLD, FDX1, GLS, and PDHB mRNA in the model group was lower than that in the control and saline groups, whereas the expression level of DBT mRNA in the model group was higher than that of the other two groups (Fig. 11. E). Furthermore, the relative expression level of DLD, GLS, and PDHB proteins showed a reduction and that of DBT showed an increase in the model group (Figure 11. F). These results indicated that the expression of DLD, FDX1, GLS, and PDHB was downregulated and the expression of DBT was upregulated in the AD model group, which supports the bioinformatic analysis results. Characterization of CRGs expression in AD cell models . The modeling drug concentration of 25 nM was determined based on LC50 after OA action on SH-SY5Y cells at various concentrations was observed (Fig. 12. A). Then a drug concentration of 25 nM was added to the cells for 24 hours, and cell viability was determined to investigate the effects of this concentration on the cell proliferation rate (Figure 12. B). The AD model was then created using this drug concentration. According to the qRT-PCR results, the expression of DLD, FDX1, GLS, and PDHB was lower and the expression of DBT was higher in the cells of the drug group than that of the normal group (Fig. 12. C). Western blot results also showed that the DLD, GLS, and PDHB content decreased and the DBT content increased in the drug group (Figure 12.D). These findings support the results of the bioinformatics analysis and the animal model. Discussion The pathogenesis and pathologic characteristics of AD are complex. Apart from the primary pathological characteristics such as neurofibrillary tangles, senile plaques, and neuronal loss, AD patients exhibit signs of mitochondrial dysfunction, including changed mitochondrial morphology, decreased activity of cytochrome oxidase (respiratory chain complex IV), and decreased glucose metabolism[29]. Metal ions can cause neurodegeneration in AD patients by upsetting the mitochondrial functional cycle, which has been suggested to be closely associated with mitochondrial respiration by cuproptosis ions and cuproptosis, a type of cell death with cuproptosis dependence[30,31]. Since cuproptosis ions induce Aβ aggregation and the generation of reactive oxygen species (ROS), they may play a role in the progression of AD[32]. For example, Aβ aggregation-mediated oxidative stress leads to mitochondrial dysfunction and accelerates the progression of AD[33,34]. Ying Zhang et al. suggested that cuproptosis-activated oxidative stress may affect memory function in AD patients[35]. The findings of Alexander Pilozzi et al. also verified the effects of cuproptosis redox on AD amyloid pathology[36]. There is a close relationship between AD and cuproptosis in mitochondrial function and the oxidative respiratory chain. This is consistent with the enrichment analysis results in this study. For example, GO and KEGG analyses showed that CRGs, i.e., FDX1, DLD, and PDHB, are closely related to the oxidative respiratory chain and play a role in mitochondrial functions. This is consistent with the findings of Starkov et al. and Yang Weiguang et al[37,38]. It can be hence assumed that there are similarities between the mechanisms of AD and cuproptosis. It has been shown that immune cells in the human brain, such as microglia and astrocytes, are involved in AD pathogenesis[39]. Several studies have investigated the effects of cuproptosis ions and cuproptosis on the infiltration of immune cells[40,41]. The correlation analysis of CRGs of infiltrating immune cells in this study showed that almost all genes were strongly associated with dendritic cell activation and monocytes. Other studies have also proven the significantly higher infiltration of primitive B cells and neutrophils in the hippocampus of AD patients[42,43]. This is consistent with the results of the immune infiltration analysis for cuproptosis core genes. Subsequent experiments revealed the high similarity between DLD and PDHB in the infiltration of immune cells, supporting the strong positive correlation between these two variables, as shown by the correlation analysis. The study results confirmed the acceptable diagnostic power of the AD risk prediction model developed based on five cuproptosis genes (including DLD, FDX1, GLS, PDHB, and DBT). FDX1 gene promotes lipoylation of pyruvate dehydrogenase (PDH) and α-ketoglutarate dehydrogenase (α-KDH), which in turn affect the TCA cycle. It has been further investigated in studies on neurodegenerative disorders and cuproptosis mortality[44,45]. On the other hand, the pyruvate dehydrogenase E1 subunit beta (PDHB) gene and the dihydrolipoic acid dehydrogenase (DLD) gene are involved in encoding the pyruvate dehydrogenase complex (PDC), which is essential for mitochondrial respiration[46] and is also closely related to the regulation of FDX1 and AD. The branched-chain alpha-ketoacid dehydrogenase complex (BCKD) is an endomitochondrial enzyme complex that influences the catabolism of the branched-chain amino acids isoleucine, leucine, and valine. It has been closely linked to the pathogenesis of AD. The dihydrolipoamide branched-chain transacylase E2 (DBT) gene encoding the transacylase (E2) subunit is also involved in the formation of this complex[47]. In contrast, the glutaminase (GLS) gene, an ammonia metabolism gene, has been reported to be associated with a wide range of glutamate signaling disorders, including AD[48]. An algorithmic analysis revealed a significant difference between the AD group and the normal group in the expression level of DLD, FDX1, GLS, and PDHB. The results showed that the DBT expression increased and the DLD expression decreased in the AD group. Considering the results of validation experiments performed on SH-SY5Y cells of SD rats, there is a need for further studies to investigate the exact mechanism of action. GSEA analysis showed all of these key genes to be associated with neurodegenerative diseases. The difference between Cluster 1 and Cluster 2 in the marker pathway activity also revealed a significant increase in Notch signaling pathway activity. Arunima Kapoor et al. and Perna et al. also reported a relationship between this signaling pathway and AD [49,50]. The accuracy of the AD diagnostic model developed based on CRGs was revalidated using external datasets, and then a nomogram was developed to assess the risk of AD using the CRGs score. The calibration curve and DCA confirmed the validity and the improved application of the graph. Limitations Similar to any other research project, this study faced some limitations. Firstly, there were not enough relevant AD samples or datasets for further analysis. Second, this study demonstrated the correlation between cuproptosis and AD, but the specific mechanism of action of differentially expressed genes in AD was not validated. Meanwhile the single method of modeling animal models needs to be improved. Conclusions This study investigated the complex relationship between cuproptosis and AD by identifying five core cuproptosis genes (DLD, FDX1, GLS, PDHB, and DBT) associated with AD. The in vivo and in vitro experiments also validated the expression level of these five core cuproptosis genes in AD models. In addition, a practical prognostic risk model was developed to assess the risk of cuproptosis and the pathological consequences of AD. The correlation between the five core cuproptosis genes and AD was validated by multiple data analyses and in vivo and ex vivo experiments. The study results revealed the downregulation of DLD, FDX1, GLS, and PDHB and the upregulation of DBT in AD patients, compared with healthy individuals. This differential expression is expected to provide new ideas for the diagnosis and prognostic monitoring of AD. Declarations Data availability Te data used to support the fndings of this study is available. Acknowledgments We thank the GEO and KEGG databases for providing the platform and contributors for uploading meaningful datasets. Author contributions R.H, Z.X, M.Y.Q, C.Y.L, G.Y.W, Y.Y.W, Z.S.H, and M.Y.D participated in this study. R.H conducted the validation experiments and collected the experimental data, and worked with Z .X on the statistical analysis and paper writing. M.Y.Q, C.Y.L, G.Y.W and Y.Y.W assisted in the animal feeding and sampling process. Z.S.H and M.Y.D participated in the conceptualization of the whole article and revised the manuscript, and participated in the technical support. All authors approved the manuscript. Funding Project of Enhancement of Basic Research Ability of Young and Middle-aged Teachers in Guangxi Universities (No.2023KY0554); Baise City Science and Technology Program Project (No.20211807, No.20224139). Competing interests Te authors declare no competing interests. References Gónzalez de San Román E, Manuel I, Giralt MT, Ferrer I, Rodríguez-Puertas R. Imaging mass spectrometry (IMS) of cortical lipids from preclinical to severe stages of Alzheimer's disease. Biochim Biophys Acta Biomembr, 2017;1859(9 Pt B):1604-1614. Kent SA, Spires-Jones TL, Durrant CS. The physiological roles of tau and Aβ: implications for Alzheimer's disease pathology and therapeutics. Acta Neuropathol, 2020;140(4):417-447. Rostagno AA. Pathogenesis of Alzheimer's Disease . Int J Mol Sci. 2022;24(1):107. Di Fede G, Catania M, Maderna E, Ghidoni R, Benussi L, Tonoli E, Giaccone G, Moda F, Paterlini A, Campagnani I, Sorrentino S, Colombo L, Kubis A, Bistaffa E, Ghetti B, Tagliavini F. Molecular subtypes of Alzheimer's disease. Sci Rep. 2018;8(1):3269. Scheltens P, De Strooper B, Kivipelto M, Holstege H, Chételat G, Teunissen CE, Cummings J, van der Flier WM. Alzheimer's disease. Lancet. 2021;397(10284):1577-1590. Kawamata H, Manfredi G. Import, maturation, and function of SOD1 and its copper chaperone CCS in the mitochondrial intermembrane space. Antioxid Redox Signal. 2010;13(9):1375-1384. Scheiber I, Dringen R, Mercer JF. Copper: effects of deficiency and overload. Met Ions Life Sci. 2013;13:359-387. Gromadzka G, Tarnacka B, Flaga A, Adamczyk A. Copper Dyshomeostasis in Neurodegenerative Diseases-Therapeutic Implications. Int J Mol Sci. 2020;21(23):9259. Sayre LM, Perry G, Harris PL, Liu Y, Schubert KA, Smith MA. In situ oxidative catalysis by neurofibrillary tangles and senile plaques in Alzheimer's disease: a central role for bound transition metals. J Neurochem. 2000;74(1):270-279. Chen LL, Fan YG, Zhao LX, Zhang Q, Wang ZY. The metal ion hypothesis of Alzheimer's disease and the anti-neuroinflammatory effect of metal chelators. Bioorg Chem. 2023;131:106301. Tsvetkov P, Coy S, Petrova B, Dreishpoon M, Verma A, AbdusamAD M, Rossen J, Joesch-Cohen L, Humeidi R, Spangler RD, Eaton JK, Frenkel E, Kocak M, Corsello SM, Lutsenko S, Kanarek N, Santagata S, Golub TR. Copper induces cell death by targeting lipoylated TCA cycle proteins . Science. 2022 Apr 22;376(6591):eabq4855]. Xie J, Yang Y, Gao Y, He J. Cuproptosis: mechanisms and links with cancers . Mol Cancer, 2023;22(1):46. Swerdlow RH. Mitochondria and Mitochondrial Cascades in Alzheimer's Disease. J Alzheimers Dis. 2018;62(3):1403-1416. Macdonald R, Barnes K, Hastings C, Mortiboys H. Mitochondrial abnormalities in Parkinson's disease and Alzheimer's disease: can mitochondria be targeted therapeutically?. Biochem Soc Trans. 2018;46(4):891-909. Tang D, Chen X, Kroemer G. Cuproptosis: a copper-triggered modality of mitochondrial cell death. Cell Res. 2022;32(5):417-418. Garza NM, Swaminathan AB, Maremanda KP, Zulkifli M, Gohil VM. Mitochondrial copper in human genetic disorders. Trends Endocrinol Metab . 2023;34(1):21-33. Wang Q, Sun J, Chen T, Song S, Hou Y, Feng L, Fan C, Li M. Ferroptosis, Pyroptosis, and Cuproptosis in Alzheimer's Disease. ACS Chem Neurosci. 2023;14(19):3564-3587. Lardenoije R, Roubroeks JAY, Pishva E, Leber M, Wagner H, Iatrou A, Smith AR, Smith RG, Eijssen LMT, Kleineidam L, Kawalia A, Hoffmann P, Luck T, Riedel-Heller S, Jessen F, Maier W, Wagner M, Hurlemann R, Kenis G, Ali M, Del Sol A, Mastroeni D, Delvaux E, Coleman PD, Mill J, Rutten BPF, Lunnon K, Ramirez A, van den Hove DLA.Alzheimer's disease-associated (hydroxy)methylomic changes in the brain and blood. Clin Epigenetics . 2019;11(1):164. Narayanan M, Huynh JL, Wang K, Yang X, Yoo S, McElwee J, Zhang B, Zhang C, Lamb JR, Xie T, Suver C, Molony C, Melquist S, Johnson AD, Fan G, Stone DJ, Schadt EE, Casaccia P, Emilsson V, Zhu J. Common dysregulation network in the human prefrontal cortex underlies two neurodegenerative diseases. Mol Syst Biol. 2014;10(7):743. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, Smyth GK. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43(7):e47. Varma S. Blind estimation and correction of microarray batch effect. PLoS One. 2020;15(4):e0231446. Newman AM, Liu CL, Green MR, Gentles AJ, Feng W, Xu Y, Hoang CD, Diehn M, Alizadeh AA. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods. 2015;12(5):453-457. Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data [J]. BMC Bioinformatics, 2013, 14(7): 1471-2105. Stepanichev MYu, Lazareva NA, Onufriev MV, Mitrokhina OS, Moiseeva YuV, Gulyaeva NV. Effects of doses of fragment (25-35) of beta-amyloid peptide on behavior in rats. Neurosci Behav Physiol. 1998;28(5):564-566. Li Z, Tong Q, Xu H, et al. Therapeutic Effects of TianDiJingWan on the Aβ 25-35-Induced Alzheimer's Disease Model Rats. Evid Based Complement Alternat Med. 2015;2015:307350. Othman MZ, Hassan Z, Che Has AT. Morris water maze: a versatile and pertinent tool for assessing spatial learning and memory. Exp Anim. 2022;71(3):264-280. Amonruttanapun P, Chongthammakun S, Chamniansawat S. The effects of okadaic acid-treated SH-SY5Y cells on microglia activation and phagocytosis. Cell Biol Int. 2022 Feb;46(2):234-242. Yang X, Zhong Y, Wang D, Lu Z. A simple colorimetric method for viable bacteria detection based on cell counting Kit-8. Anal Methods. 2021;13(43):5211-5215. Zhu X, Smith MA, Perry G, Aliev G. Mitochondrial failures in Alzheimer's disease. Am J Alzheimers Dis Other Demen . 2004;19(6):345-352. Xie J, Yang Y, Gao Y, He J. Cuproptosis: mechanisms and links with cancers. Mol Cancer. 2023;22(1):46. Mezzaroba L, Alfieri DF, Colado Simão AN, Vissoci Reiche EM. The role of zinc, copper, manganese and iron in neurodegenerative diseases. Neurotoxicology. 2019;74:230-241. Hureau C, Faller P. Abeta-mediated ROS production by Cu ions: structural insights, mechanisms and relevance to Alzheimer's disease. Biochimie. 2009;91(10):1212-1217. Butterfield DA, Boyd-Kimball D. Redox proteomics and amyloid β-peptide: insights into Alzheimer disease. J Neurochem. 2019;151(4):459-487. Jansen IE, van der Lee SJ, Gomez-Fonseca D, de Rojas I, Dalmasso MC, Grenier-Boley B, Zettergren A, Mishra A, Ali M, Andrade V, Bellenguez C, Kleineidam L, Küçükali F, Sung YJ, Tesí N, Vromen EM, Wightman DP, Alcolea D, Alegret M, Alvarez I, Amouyel P, Athanasiu L, Bahrami S, Bailly H, Belbin O, Bergh S, Bertram L, Biessels GJ, Blennow K, Blesa R, Boada M, Boland A, Buerger K, Carracedo Á, Cervera-Carles L, Chene G, Claassen JAHR, Debette S, Deleuze JF, de Deyn PP, Diehl-Schmid J, Djurovic S, Dols-Icardo O, Dufouil C, Duron E, Düzel E; EADB consortium; Fladby T, Fortea J, Frölich L, García-González P, Garcia-Martinez M, Giegling I, Goldhardt O, Gobom J, Grimmer T, Haapasalo A, Hampel H, Hanon O, Hausner L, Heilmann-Heimbach S, Helisalmi S, Heneka MT, Hernández I, Herukka SK, Holstege H, Jarholm J, Kern S, Knapskog AB, Koivisto AM, Kornhuber J, Kuulasmaa T, Lage C, Laske C, Leinonen V, Lewczuk P, Lleó A, de Munain AL, Lopez-Garcia S, Maier W, Marquié M, Mol MO, Montrreal L, Moreno F, Moreno-Grau S, Nicolas G, Nöthen MM, Orellana A, Pålhaugen L, Papma JM, Pasquier F, Perneczky R, Peters O, Pijnenburg YAL, Popp J, Posthuma D, Pozueta A, Priller J, Puerta R, Quintela I, Ramakers I, Rodriguez-Rodriguez E, Rujescu D, Saltvedt I, Sanchez-Juan P, Scheltens P, Scherbaum N, Schmid M, Schneider A, Selbæk G, Selnes P, Shadrin A, Skoog I, Soininen H, Tárraga L, Teipel S; GR@ACE study group; Tijms B, Tsolaki M, Van Broeckhoven C, Van Dongen J, van Swieten JC, Vandenberghe R, Vidal JS, Visser PJ, Vogelgsang J, Waern M, Wagner M, Wiltfang J, Wittens MMJ, Zetterberg H, Zulaica M, van Duijn CM, Bjerke M, Engelborghs S, Jessen F, Teunissen CE, Pastor P, Hiltunen M, Ingelsson M, Andreassen OA, Clarimón J, Sleegers K, Ruiz A, Ramirez A, Cruchaga C, Lambert JC, van der Flier W. Genome-wide meta-analysis for Alzheimer's disease cerebrospinal fluid biomarkers. Acta Neuropathol. 2022;144(5):821-842. Zhang Y, Zhou Q, Lu L, Su Y, Shi W, Zhang H, Liu R, Pu Y, Yin L. Copper Induces Cognitive Impairment in Mice via Modulation of Cuproptosis and CREB Signaling. Nutrients. 2023;15(4):972. Pilozzi A, Yu Z, Carreras I, Cormier K, Hartley D, Rogers J, Dedeoglu A, Huang X. A Preliminary Study of Cu Exposure Effects upon Alzheimer's Amyloid Pathology. Biomolecules . 2020;10(3):408. Starkov AA, Fiskum G, Chinopoulos C, Lorenzo BJ, Browne SE, Patel MS, Beal MF. Mitochondrial alpha-ketoglutarate dehydrogenase complex generates reactive oxygen species. J Neurosci. 2004;24(36):7779-7788. Yang W, Guo Q, Wu H, Tong L, Xiao J, Wang Y, Liu R, Xu L, Yan H, Sun Z. Comprehensive analysis of the cuproptosis-related gene DLD across cancers: A potential prognostic and immunotherapeutic target. Front Pharmacol. 2023;14:1111462. Rajesh Y, Kanneganti TD. Innate Immune Cell Death in Neuroinflammation and Alzheimer's Disease. Cells. 2022;11(12):1885. Suliman IH, Kim K, Chen W, et al. Metal-Based Nanoparticles for Cancer Metalloimmunotherapy. Pharmaceutics. 2023;15(7):2003. Wang L, Cao Y, Guo W, Xu J. High expression of cuproptosis-related gene FDX1 in relation to good prognosis and immune cells infiltration in colon adenocarcinoma (COAD). J Cancer Res Clin Oncol. 2023;149(1):15-24. Huang LT, Zhang CP, Wang YB, Wang JH. Association of Peripheral Blood Cell Profile With Alzheimer's Disease: A Meta-Analysis. Front Aging Neurosci. 2022;14:888946. Song L, Yang YT, Guo Q; ZIB Consortium, Zhao XM. Cellular transcriptional alterations of peripheral blood in Alzheimer's disease. BMC Med. 2022;20(1):266. Tsvetkov P, Coy S, Petrova B, Dreishpoon M, Verma A, Abdusamad M, Rossen J, Joesch-Cohen L, Humeidi R, Spangler RD, Eaton JK, Frenkel E, Kocak M, Corsello SM, Lutsenko S, Kanarek N, Santagata S, Golub TR. Copper induces cell death by targeting lipoylated TCA cycle proteins [published correction appears in Science. 2022 Apr 22;376(6591):eabq4855]. Science. 2022;375(6586):1254-1261. Zhang Y, Zhou Q, Lu L, Su Y, Shi W, Zhang H, Liu R, Pu Y, Yin L. Copper Induces Cognitive Impairment in Mice via Modulation of Cuproptosis and CREB Signaling. Nutrients. 2023;15(4):972. Patel MS, Nemeria NS, Furey W, Jordan F. The pyruvate dehydrogenase complexes: structure-based function and regulation. J Biol Chem. 2014;289(24):16615-16623. Hull J, Usmari Moraes M, Brookes E, Love S, Conway ME. Distribution of the branched-chain α-ketoacid dehydrogenase complex E1α subunit and glutamate dehydrogenase in the human brain and their role in neuro-metabolism. Neurochem Int. 2018;112:49-58. Griffin JWD, Liu Y, Bradshaw PC, Wang K. In Silico Preliminary Association of Ammonia Metabolism Genes GLS, CPS1, and GLUL with Risk of Alzheimer's Disease, Major Depressive Disorder, and Type 2 Diabetes. J Mol Neurosci. 2018;64(3):385-396. Kapoor A, Nation DA. Role of Notch signaling in neurovascular aging and Alzheimer's disease. Semin Cell Dev Biol. 2021;116:90-97. Perna A, Marathe S, Dreos R, Falquet L, Akarsu Egger H, Auber LA. Revealing NOTCH-dependencies in synaptic targets associated with Alzheimer's disease. Mol Cell Neurosci. 2021;115:103657. Additional Declarations No competing interests reported. Supplementary Files Westernblotanimal.pdf Westernblotcell.pdf Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {\"props\":{\"pageProps\":{\"initialData\":{\"identity\":\"rs-3854023\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":true,\"archivedVersions\":[],\"articleType\":\"Article\",\"associatedPublications\":[],\"authors\":[{\"id\":268158272,\"identity\":\"c25230c6-0de2-4b8e-9c12-078cabc0ac54\",\"order_by\":0,\"name\":\"Rui Hu\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Youjiang Medical College of Nationalities\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Rui\",\"middleName\":\"\",\"lastName\":\"Hu\",\"suffix\":\"\"},{\"id\":268158273,\"identity\":\"c3b3e251-50cf-4097-959d-95c14ad2ae74\",\"order_by\":1,\"name\":\"Zhen Xiao\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Youjiang Medical College of Nationalities\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Zhen\",\"middleName\":\"\",\"lastName\":\"Xiao\",\"suffix\":\"\"},{\"id\":268158274,\"identity\":\"e91e5d78-fcce-4d4d-a6fc-44a66aca574d\",\"order_by\":2,\"name\":\"Mingyu Qiao\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Youjiang Medical College of Nationalities\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Mingyu\",\"middleName\":\"\",\"lastName\":\"Qiao\",\"suffix\":\"\"},{\"id\":268158275,\"identity\":\"e9819ae1-6e5b-466e-aa8e-26a2aa1c8a12\",\"order_by\":3,\"name\":\"Chaoyu Liu\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Affiliated Hospital of Youjiang Medical College of Nationalities\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Chaoyu\",\"middleName\":\"\",\"lastName\":\"Liu\",\"suffix\":\"\"},{\"id\":268158276,\"identity\":\"afdf4725-d1a9-44a0-baa6-aae3e7ced71f\",\"order_by\":4,\"name\":\"Guiyou Wu\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Guangxi University of Traditional Chinese Medicine\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Guiyou\",\"middleName\":\"\",\"lastName\":\"Wu\",\"suffix\":\"\"},{\"id\":268158277,\"identity\":\"2336af34-afbf-4c6b-b606-659d80c93ea7\",\"order_by\":5,\"name\":\"Yunyi Wang\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Youjiang Medical College of Nationalities\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Yunyi\",\"middleName\":\"\",\"lastName\":\"Wang\",\"suffix\":\"\"},{\"id\":268158278,\"identity\":\"f226e788-03f2-46cd-acdd-a20739cc6a5e\",\"order_by\":6,\"name\":\"Zhongshi Huang\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBUlEQVRIiWNgGAWjYDACCTB5gMcATFdARXmI13KGBC0MYC2MbURokZ/dfOzhlz93ZMwlkp89/DrvcLTB7QbGB2/bGOTNcWhhnHMs3Vi27RmP5Yw0c2PZbYdzN9w5wGw4t43BcGcDdi3MEjlm0pINh3kMbiQAGSAtNxLYpHnbGBIMDmDXwgbSIvEHpCX9m7TkHLAW9t/4tPAAtUh+YANpATI+NkBsYcanRUIiLU2asQ2o5cybMmmGY+m5M28kNkvOOSdhuAGHFvkZycckf/w5bG9wPH2b5I8a69y+G8kHP7wps5HHZQs4CHgQjGYgxdjAAIsvXIDxB4JRh1flKBgFo2AUjEwAAI4fYFhkxSERAAAAAElFTkSuQmCC\",\"orcid\":\"\",\"institution\":\"Youjiang Medical College of Nationalities\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Zhongshi\",\"middleName\":\"\",\"lastName\":\"Huang\",\"suffix\":\"\"},{\"id\":268158279,\"identity\":\"7b6b9c21-eaa6-4312-b446-b6f4257f4510\",\"order_by\":7,\"name\":\"Mingyou Dong\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Affiliated Hospital of Youjiang Medical College of Nationalities\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Mingyou\",\"middleName\":\"\",\"lastName\":\"Dong\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2024-01-11 15:59:10\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-3854023/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-3854023/v1\",\"draftVersion\":[],\"editorialEvents\":[],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":50046884,\"identity\":\"89d75345-73fa-4ed8-be20-3284a344c52d\",\"added_by\":\"auto\",\"created_at\":\"2024-01-23 16:05:58\",\"extension\":\"png\",\"order_by\":1,\"title\":\"Figure 1\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":280198,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eIdentification of CRGs dysregulated in AD. (A) Volcano plot of 78 DEG expressions, with significantly up-regulated genes in red and down-regulated in green. (B) Box plot showing the expression of 13 CRGs between AD and non-AD controls.(C) Heatmap presenting the expression of 5 CRGs. * p \\u0026lt; 0.05, ** p \\u0026lt; 0.01, *** p \\u0026lt; 0.001.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage1.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/4b7daf0e0a5f2c97ffe4148d.png\"},{\"id\":50043630,\"identity\":\"6d5c5a71-7321-49e1-b306-1f2d24c5f8c3\",\"added_by\":\"auto\",\"created_at\":\"2024-01-23 15:49:58\",\"extension\":\"png\",\"order_by\":2,\"title\":\"Figure 2\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":341382,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eGO and KEGG pathway enrichment analysis of CRGs in AD. (A) Results of GO enrichment analysis of CRGs. (B) KEGG pathway demonstration. (C) Gene relationship network diagram of 5 differentially expressed CRGs. (D) Correlation analysis of the 5 differentially expressed CRGs. Green and red colors indicate negative and positive correlations, respectively. Correlation coefficients are labeled with the area of pie charts.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage2.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/e961216b4f9aa6cf2b76d696.png\"},{\"id\":50045520,\"identity\":\"31902734-a3ee-4766-b40d-1d272a20bf58\",\"added_by\":\"auto\",\"created_at\":\"2024-01-23 15:57:58\",\"extension\":\"png\",\"order_by\":3,\"title\":\"Figure 3\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":284399,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eImmune infiltration analysis of CRGs. (A) Relative abundance of 22 infiltrating immune cells between AD and non-AD controls. (B) Shows the difference in immune infiltration between AD and non-AD controls. (C) Correlation analysis of five differentially expressed CRGs with infiltrating immune cells. * p \\u0026lt; 0.05, ** p \\u0026lt; 0.01, *** p \\u0026lt; 0.001.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage3.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/e45f29b194866fd3e9a24a8c.png\"},{\"id\":50043638,\"identity\":\"eee7c5da-0d85-4988-92da-4cc313ee2644\",\"added_by\":\"auto\",\"created_at\":\"2024-01-23 15:49:58\",\"extension\":\"png\",\"order_by\":4,\"title\":\"Figure 4\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":261159,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eIdentification of copper death-related molecular clusters in AD. (A) Consistent clustering matrix when k = 2. (B) CDF δ-area curve. (C) Consistency clustering score.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage4.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/012f4bffcc7348059ede5196.png\"},{\"id\":50043633,\"identity\":\"4dcc687e-0984-4334-8b43-6996ec6f53ab\",\"added_by\":\"auto\",\"created_at\":\"2024-01-23 15:49:58\",\"extension\":\"png\",\"order_by\":5,\"title\":\"Figure 5\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":343009,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eIdentification of molecular and immunological features between the two clusters. (A) Clinical characterization of the five CRGs between the two clusters is presented in the heatmap. (B) Box plots showing the expression differences of individual CRGs between the two clusters. (C) Relative abundance of 22 infiltrating immune cells between the two copper death clusters. (D) Shows the difference in immune infiltration between two copper-dead clusters. * p \\u0026lt; 0.05, ** p \\u0026lt; 0.01, *** p \\u0026lt; 0.0011.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage5.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/100d2d7b83b6c7e4ffd34cb0.png\"},{\"id\":50043631,\"identity\":\"d30128d3-5278-4b99-940f-ac1c822567a2\",\"added_by\":\"auto\",\"created_at\":\"2024-01-23 15:49:58\",\"extension\":\"png\",\"order_by\":6,\"title\":\"Figure 6\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":463552,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eCo-expression network of differentially expressed genes in AD. (A) Selection of soft threshold power. (B) Cluster tree dendrogram of co-expression modules. Different colors represent different co-expression modules. (C) Representative heatmap of correlations between multiple modules. (D) Representative clustering of genes characterized by modules. (E) Correlation analysis of module feature genes with clinical status. Each row represents a module; each column represents clinical status. (F) Scatterplot between module membership relationships in turquoise modules and gene significance for AD.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage6.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/df98c300f5dea59c7916abb2.png\"},{\"id\":50043636,\"identity\":\"1d637ad9-a283-4a9a-ae1d-4d1cd8804610\",\"added_by\":\"auto\",\"created_at\":\"2024-01-23 15:49:58\",\"extension\":\"png\",\"order_by\":7,\"title\":\"Figure 7\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":385954,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eCo-expression network of differentially expressed genes between two copper death clusters. (A) Selection of soft threshold power. (B) Cluster tree dendrogram of co-expression modules. Different colors represent different co-expression modules. (C) Clustering of genes representing module features. (D) Representative heatmap of correlations between multiple modules. (E) Correlation analysis of module feature genes with clinical status. Each row represents a module; each column represents clinical status. (F) Scatterplot between module membership relationships in turquoise modules and gene significance in cluster 1.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage7.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/64e94c498db3174d6687a90b.png\"},{\"id\":50045524,\"identity\":\"fe4e128a-3e5a-4c8c-b646-ad4fe643e640\",\"added_by\":\"auto\",\"created_at\":\"2024-01-23 15:57:58\",\"extension\":\"png\",\"order_by\":8,\"title\":\"Figure 8\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":312429,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eIdentification of DEGs specific between two copper death clusters and their biological characterization. (A) Cross-over between module-associated genes dying in copper and module-associated genes in the GSE109887 dataset. (B) Differences in marker pathway activity between Cluster 1 and Cluster 2 samples sorted by t-value of the GSVA method.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage8.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/4522b222fedee6231dfecf92.png\"},{\"id\":50045525,\"identity\":\"b6bac40a-ad65-4280-8f9c-61c32528e536\",\"added_by\":\"auto\",\"created_at\":\"2024-01-23 15:57:58\",\"extension\":\"png\",\"order_by\":9,\"title\":\"Figure 9\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":302241,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eConstruction of RF, SVM, GLM and XGB machine models. (A) Cumulative residual distribution for each machine learning model. (B) Box plots showing the residuals for each machine learning model. The red dots indicate the root mean square (RMSE) of the residuals. (C) Important features in RF, SVM, GLM, and XGB machine models. (D) ROC analysis of the four machine learning models based on 5-fold cross-validation in the test cohort.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage9.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/db007dc7ebf9c04c788f8dee.png\"},{\"id\":50047853,\"identity\":\"812d82e6-fc9c-462c-9b9d-1adc09290c46\",\"added_by\":\"auto\",\"created_at\":\"2024-01-23 16:14:07\",\"extension\":\"png\",\"order_by\":10,\"title\":\"Figure 10\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":219290,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003evalidation of SVM models. (A) Nomogram of 5-gene-based SVM models constructed for predicting risk of AD clusters . (B, C) Calibration curves used to assess the predictive efficiency of the column-line graph model. (B) DCA.(C) Predictive efficiency of the column-line graph model used to assess the predictive efficiency of the column-line graph model. (D) ROC analysis of a 5-gene-based SVM model based on 5-fold cross-validation in GSE33000.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage10.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/285bbf357bd716244c57e235.png\"},{\"id\":50043640,\"identity\":\"c11f19e7-6998-4fde-b1b9-6878d39ff4a7\",\"added_by\":\"auto\",\"created_at\":\"2024-01-23 15:49:58\",\"extension\":\"png\",\"order_by\":11,\"title\":\"Figure 11\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":301340,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003evalidation of the animal model. mwm: water maze, N = 5 × 3 . (A) Walking distance during spatial training. (B) Swimming speed during search for the MWM platform. (C) Escape latency in the MWM experiment. (D) Number of platforms traversed in the exploration experiment. (E) mRNA levels of DLD, FDX1, GLS, PDHB, and DBT in the hippocampus of rats in each group. (F) Protein expression levels of DLD, GLS, PDHB, and DBT in the hippocampus of rats in each group (1 is the control group, 2 is the NS group, and 3 is the AD group). * p \\u0026lt; 0.05, ** p \\u0026lt; 0.01, *** p \\u0026lt; 0.001, **** p \\u0026lt; 0.0001.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage11.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/c847111a2b26455602f3901e.png\"},{\"id\":50043642,\"identity\":\"ec7d36f8-4ed4-4054-9e1b-d3acdf54e2e0\",\"added_by\":\"auto\",\"created_at\":\"2024-01-23 15:49:58\",\"extension\":\"png\",\"order_by\":12,\"title\":\"Figure 12\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":217727,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003evalidation of cell models. (A) Calculated IC50 results at 0, 10, 20, 40, and 80 nM okadaic acid concentrations. (B) Cell viability of 25 nM okadaic acid drug concentration acting on cells for 24h. (C) mRNA levels of DLD, FDX1, GLS, PDHB, DBT in control and model groups. (D) Protein expression levels of DLD, GLS, PDHB, DBT in SH-SY5Y cells of AD model group (1 is control group, 2 is AD group). * p \\u0026lt; 0.05, ** p \\u0026lt; 0.01, *** p \\u0026lt; 0.001, **** p \\u0026lt; 0.0001.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage12.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/886ea2da9c71fbce4a65aa3d.png\"},{\"id\":54361364,\"identity\":\"9d5d5897-0968-41b4-b061-cf1f9d372f22\",\"added_by\":\"auto\",\"created_at\":\"2024-04-09 11:15:23\",\"extension\":\"pdf\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":2094316,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"manuscript.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/e3c9c147-de8e-4014-84ef-278c3814882a.pdf\"},{\"id\":50045521,\"identity\":\"7185659a-c4a4-4f61-8edb-cf3902af3473\",\"added_by\":\"auto\",\"created_at\":\"2024-01-23 15:57:58\",\"extension\":\"pdf\",\"order_by\":1,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":1315167,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Westernblotanimal.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/c1f3899d247d7cb0d7666e9d.pdf\"},{\"id\":50043641,\"identity\":\"4f7950c8-a34a-481f-9e85-cbacd9e19ad1\",\"added_by\":\"auto\",\"created_at\":\"2024-01-23 15:49:58\",\"extension\":\"pdf\",\"order_by\":2,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":1250328,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Westernblotcell.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-3854023/v1/4db4dd3fe732f227cb0de452.pdf\"}],\"financialInterests\":\"No competing interests reported.\",\"formattedTitle\":\"Construction and validation of a bioinformatics-based screen for Cuproptosis-related genes and risk model for Alzheimer's disease\",\"fulltext\":[{\"header\":\"Introduction\",\"content\":\"\\u003cp\\u003eAlzheimer's disease (AD, also known as dementia) is one of the most prevalent neurodegenerative diseases that imposes a heavy economic burden on societies around the world[\\u003cspan citationid=\\\"CR1\\\" class=\\\"CitationRef\\\"\\u003e1\\u003c/span\\u003e]. The disease is characterized by age spots, and senile plaque (SP), neurofibrillary tangle (NFT), and neuronal loss are among its major pathological symptoms[\\u003cspan citationid=\\\"CR2\\\" class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e]. Many studies have addressed the pathogenesis of AD and reported that the primary pathogenic mechanisms in the pathogenesis of AD that have significantly influenced clinical treatments are Aβ plaque-associated neurodegeneration, neurofibrillary tangle-based neuro-progenitor degeneration, mitochondrial dysfunction, and gene mutation hypothesis[\\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR4\\\" class=\\\"CitationRef\\\"\\u003e4\\u003c/span\\u003e]. Because of the correlation between these pathogenic mechanisms, some researchers have proposed that several mechanisms may be involved in the development of AD. One of the major areas of research in this field is to investigate these mechanisms to predict the risk of AD[\\u003cspan citationid=\\\"CR5\\\" class=\\\"CitationRef\\\"\\u003e5\\u003c/span\\u003e]. Another area of research is the study of the genome as a predictor of AD risk. The genome provides important insights for investigating the molecular mechanisms of AD, and it is a potent tool for predicting the risk of the disease. Therefore, it is necessary to develop an AD risk prediction model based on the exploration of molecular mechanisms.\\u003c/p\\u003e \\u003cp\\u003eCuproptosis is an essential element that plays an important role in several metabolic processes in the brain[\\u003cspan citationid=\\\"CR6\\\" class=\\\"CitationRef\\\"\\u003e6\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e]. Excess cuproptosis in the brain of AD patients exacerbates oxidative damage and increases the formation of amyloid plaques and neurofibrillary tangles (NFTs)[\\u003cspan citationid=\\\"CR8\\\" class=\\\"CitationRef\\\"\\u003e8\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR9\\\" class=\\\"CitationRef\\\"\\u003e9\\u003c/span\\u003e]. The neurotoxic effect of cuproptosis is closely related to the inherent redox properties of Cu\\u003csup\\u003e2+\\u003c/sup\\u003e. Cuproptosis overload promotes cuproptosis[\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e], and cuproptosis, as a newly discovered independent mode of cell death characterized by cuproptosis-dependent and mitochondrial respiratory regulation, has a significant impact on mitochondrial functions[\\u003cspan citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR12\\\" class=\\\"CitationRef\\\"\\u003e12\\u003c/span\\u003e]. However, numerous studies have concluded that mitochondrial dysfunction-induced oxidative stress and abnormalities in energy metabolism account for a major pathogenic mechanism involved in the development of AD[\\u003cspan citationid=\\\"CR13\\\" class=\\\"CitationRef\\\"\\u003e13\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR14\\\" class=\\\"CitationRef\\\"\\u003e14\\u003c/span\\u003e]. Cuproptosis ions and cuproptosis are closely related to mitochondrial function[\\u003cspan citationid=\\\"CR15\\\" class=\\\"CitationRef\\\"\\u003e15\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR16\\\" class=\\\"CitationRef\\\"\\u003e16\\u003c/span\\u003e], and recent studies have investigated the potential correlation between cuproptosis and AD[\\u003cspan citationid=\\\"CR17\\\" class=\\\"CitationRef\\\"\\u003e17\\u003c/span\\u003e].\\u003c/p\\u003e \\u003cp\\u003eThere are relatively few studies exploring the potential correlation between cuproptosis and AD from the perspective of bioinformatics to develop a prognostic risk model. Therefore, this study aims to investigate the correlation between CRGs and AD and develop a machine learning-based prognostic risk model for AD. Another objective of this study is to develop a cellular model with Okadaic acid (OA)-treated SH-SY5Y cells to validate the expression of the resulting CRGs by the Aβ\\u003csub\\u003e25\\u0026minus;35\\u003c/sub\\u003e dementia model in simulated rats in order to propose a new approach to the diagnosis and prognosis of AD.\\u003c/p\\u003e\"},{\"header\":\"Materials and methods\",\"content\":\"\\u003cp\\u003e \\u003cb\\u003eData collection.\\u003c/b\\u003e The gene expression datasets used in this study are all accessible via the GEO public database (\\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps:www.NCBI.NLM.nih.gov/GEO/\\u003c/span\\u003e\\u003cspan address=\\\"https:www.NCBI.NLM.nih.gov/GEO/\\\" targettype=\\\"URL\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e). This includes the GSE109887 dataset based on the GPL10904 platform, which contains 46 AD samples and 32 control samples (all from human temporal regression samples)[\\u003cspan citationid=\\\"CR18\\\" class=\\\"CitationRef\\\"\\u003e18\\u003c/span\\u003e], and the GSE33000 dataset based on the GPL 10558 platform, which contains 310 AD samples and 157 control prefrontal cortex samples[\\u003cspan citationid=\\\"CR19\\\" class=\\\"CitationRef\\\"\\u003e19\\u003c/span\\u003e]. A total of 16 CRGs were obtained from the KEGG public website (\\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://www.kegg.jp/pathway\\u003c/span\\u003e\\u003cspan address=\\\"https://www.kegg.jp/pathway\\\" targettype=\\\"URL\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e).\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eDifferential expression analysis and data processing.\\u003c/b\\u003e Data analysis was performed on the above datasets in R 4.2.1. The sva R package was used to remove batch effects and the limma R package was employed for differential expression analysis[\\u003cspan citationid=\\\"CR20\\\" class=\\\"CitationRef\\\"\\u003e20\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR21\\\" class=\\\"CitationRef\\\"\\u003e21\\u003c/span\\u003e]. All data sets were then normalized in the preprocessCore R package, and differential expression analysis was performed in the limma R package. The threshold for DEG was set at | log-fold change (FC)| \\u0026gt; 1, P\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05. The screened differential genes were presented using heatmaps and volcano plots, produced using the R packages ggplot2 and pheatmap.\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eEnrichment analysis of core cuproptosis genes.\\u003c/b\\u003e Gene Ontology (GO) and KEGG pathway analyses were performed on all DEGs as well as five cuproptosis-related hub genes in the ClusterProfiler package to identify the relevant pathways and biological functions influenced by differentially expressed genes. The significance level was considered to be P\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05.\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eEvaluation of immune cell infiltration.\\u003c/b\\u003e The relative abundance of 22 types of immune cells in each sample was estimated based on gene expression data in the CIBERS0RT R package[\\u003cspan citationid=\\\"CR22\\\" class=\\\"CitationRef\\\"\\u003e22\\u003c/span\\u003e]. The results of the correlation between CRGs and infiltrating immune cells were presented using the \\\"ggplot2\\\" R package.\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eUnsupervised cluster analysis of AD samples.\\u003c/b\\u003e Based on the expression profiles of CRGs, an unsupervised cluster analysis was performed in the \\\"ConsensusClusterPlus\\\" R package to classify the AD samples into different clusters using a k-means algorithm. The optimal number of clusters was evaluated by combining the maximum number of subtypes k (k\\u0026thinsp;=\\u0026thinsp;6) (see figure), the cumulative distribution function (CDF) curve, the consistency matrix, and the consistency clustering score (\\u0026gt;\\u0026thinsp;0.9).\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eWeighted gene co-expression network analysis (WGCNA).\\u003c/b\\u003e The WGCNA R package was employed to develop co-expression networks and modules in healthy individuals and AD patients to identify key gene modules associated with AD. The top 50% of genes with the highest variance were applied to subsequent WGCNA analyses to ensure the accuracy of qualitative results. A total of 10 co-expression modules with different colors were obtained using a dynamic cutting algorithm, and a heatmap of the topological overlap matrix (TOM) was also presented, in which gene dendrograms and module colors were generated based on gene dissimilarity. The global gene expression profiles in each module are represented by module signature genes. Then an appropriate set of genes for each module is filtered by an algorithm to intersect with the set of differential genes to show the correlation.\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eGSVA analysis.\\u003c/b\\u003e GSVA enrichment analysis was performed in the \\\"GSVA\\\" R package to elucidate the signaling pathways that are differentially enriched between different groups of CRGs.[\\u003cspan citationid=\\\"CR23\\\" class=\\\"CitationRef\\\"\\u003e23\\u003c/span\\u003e]. The files \\\"c2.cp. Kegg.v7.4.symbols\\\" and \\\"c5.go.bp.v7.5.1.symbols\\\" were obtained from the MSigDB web database for further GSVA analysis. The \\\"limma\\\" R package was utilized to compare the GSVA scores between different clusters of CRGs to identify the differentially expressed pathways and biological functions.\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eNomogram model development and validation.\\u003c/b\\u003e A nomogram model was developed for the clinical diagnosis of AD in the \\\"RMS\\\" R package. Each predicted gene received a score, and the \\\"total score\\\" represented the sum of the above predictor scores. Calibration curves and DCA were used to estimate the predictive power of the nomogram model.\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eMachine learning and prognostic modeling.\\u003c/b\\u003e Models such as the Random Forest Model (RF), Support Vector Machine Model (SVM), Generalized Linear Model (GLM), and Extreme Gradient Boosting (XGB) were developed using the differentially expressed genes of two different clusters of CRGs and the intersecting genes of AD differentially expressed genes. The best machine learning model was identified based on the highest ROC values of the above models; accordingly, the SVM and the first five important variables were considered the key predictive genes associated with AD. Finally, the diagnostic value of the model was validated by assessing its accuracy by using the GSE33000 dataset.\\u003c/p\\u003e \\u003cp\\u003e\\u003cb\\u003eAnimal sources and construction of AD models.\\u003c/b\\u003e Thirty male SD rats, SPF-grade, 8 months old, weighing 220\\u0026ndash;240 g, were purchased from Guangdong Viton Lihua Laboratory Animal Technology Co. to be used as subjects. The subjects were housed, modeled, and administered in an SPF-grade laboratory at the Animal Experimentation Center of Youjiang Medical College of Nationalities. The SPF-grade laboratory was equipped with a 12 h/12 h alternating light and dark barrier system, and its ambient temperature and relative humidity were 25℃ and 60\\u0026ndash;70%, respectively. In addition, the subjects were fed until December for subsequent experiments. They also passed the Right River College of National Medicine's ethical review and examination for the use of experimental animals (Ethical Review Approval No. 2023040101), and all relevant guidelines were properly followed throughout the experiment.\\u003c/p\\u003e \\u003cp\\u003eThe subjects were randomly divided into the model, saline, and normal control groups (10 rats in each group). After weighing the subjects, they were anesthetized with isoflurane (RWD, CAT. NO.hsc-454-5) using a Reward R500 general-purpose small animal anesthesia machine. They were then placed on a sterile surgical table, and the head was fixed with a stereotaxic apparatus to monitor the respiratory rate. If the respiratory rate remained stable, a stereotaxic injection was performed after the subject's head was fixed appropriately. Subjects in the model group were treated with 1 \\u0026micro;l of Aβ\\u003csub\\u003e25\\u0026minus;35\\u003c/sub\\u003e on each side of the hippocampus (Glpbio, CAT. NO.GP10082)[\\u003cspan citationid=\\\"CR24\\\" class=\\\"CitationRef\\\"\\u003e24\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR25\\\" class=\\\"CitationRef\\\"\\u003e25\\u003c/span\\u003e], whereas those in the saline group were injected an equal amount of normal saline into each side of the hippocampus. Those in the normal control group received no surgical treatment. In addition, 100,000 units/g/day of Penicillin was injected into each subject for 7 days after surgery to prevent postoperative infections.\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eMorris water maze (MWM) test.\\u003c/b\\u003e The MWM test was performed using the DigBehv-MG model device, Shanghai Gizo Software Technology Co., to assess the spatial learning and memory of subjects[\\u003cspan citationid=\\\"CR26\\\" class=\\\"CitationRef\\\"\\u003e26\\u003c/span\\u003e]. The localization and navigation test lasted for 4 days, and each subject was trained 4 times per day. The subjects were submerged from the entry point that faces the pool wall, and the time it took them to find the platform within one minute was recorded (escape latency). If a subject failed to find the platform within one minute, the subject was instructed to stand on the platform and remain there for 10 seconds in order to achieve the training purpose; the escape latency for that subject was considered to be one minute. Two days after the localization and navigation tests finished, the platform in the pool was removed and the furthest quadrant of the original platform was taken as the entry point. Then the subjects in both groups were placed into the water while facing the wall from the entry point, and the escape latency was recorded within one minute. Moreover, the subjects were placed into the water from the entry point toward the wall of the pool, and the swimming time and the number of times they crossed the original platform within one minute were recorded. All subjects were acclimatized overnight before the experiments.\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eRNA extraction and qRT-PCR to verify the differential expression of CRGs mRNA in AD rat model.\\u003c/b\\u003e The day after the end of the trans-spatial memory ability test, the subjects were anesthetized to remove their brain tissue; the tissue samples were weighed and then they were stored in a refrigerator at -80\\u0026deg;C. To extract RNA from the brain tissue samples, 30 mg of weighed brain tissue samples were taken from each group. Then, the centrifuge tubes were filled with the appropriate amount of TRIzol reagent (Thermo Fisher Scientific, CAT. NO.15596026) to extract the total RNA samples based on the mass of the weighed brain tissue samples. The Third Generation Variable Speed Tissue Mill (TGrinder, China) was used to grind the brain tissue samples in the centrifuge tubes until the brain tissue was totally dissolved in the TRIzol reagent. Then they were left on ice for 5 minutes for subsequent operations. RNA was reverse-transcribed into cDNA using ToloScript All-in-one RT EasyMix for qPCR kit (TOLOBIO, CAT. NO.22107), and the results were analyzed in a Light Cycle 96 real-time fluorescence PCR instrument (Roche, Germany) using SYBR Green qPCR Mix kit (TOLOBIO, CAT. NO.22204-1). The mRNA expression of the genes related to cuproptosis was determined in a Light Cyle 96 real-time fluorescence quantitative PCR instrument (Roche, Germany). GAPDH was the endogenous control gene, and the primers used for amplification were as follows.\\u003c/p\\u003e \\u003cdiv id=\\\"Sec3\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eGAPDH-F:5'-GACATGCCGCCTGGAGAAAC-3 '\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec4\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eGAPDH-R :5 '-AGCCCAGGATGCCCTTTAGT-3 '\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec5\\\" class=\\\"Section4\\\"\\u003e \\u003ch2\\u003eDLD-F:5 '-CATTTCAATCGGCTGTCTCA-3 '\\u003c/h2\\u003e \\u003cp\\u003eDLD-R:5 '-CAAGCATGTTCCTCCTAGTGTT-3 '\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec6\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eGLS-F:5 '-CTGAACGAGAAAGTGGAGACCGAA-3 '\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec7\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eGLS-R:5 '-TGGGCAGAAACCGCCATTAG-3 '\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec8\\\" class=\\\"Section4\\\"\\u003e \\u003ch2\\u003ePDHB-F:5 '-GCATTTGAACTTCCCACAGA-3 '\\u003c/h2\\u003e \\u003cp\\u003ePDHB-R:5 '-TTCCCTCCTTAGACAATACAGC-3 '\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec9\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eDBT-F:5'-ACGTGTGCTCTCTGTGGGGTTATC-3 '\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec10\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eDBT-R:5 '-GCTGTCAAACTGAGACACCG-3 '\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec11\\\" class=\\\"Section4\\\"\\u003e \\u003ch2\\u003eFDX1-F:5 '-TGGTGAAACGCTAACGACCA-3 '\\u003c/h2\\u003e \\u003cp\\u003eFDX1-R:5 '-CAAGCCAAAGTCCCCTCACA-3 '\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eProtein extraction and WB verification of differential expression of CRGs in rat model of AD.\\u003c/b\\u003e A sample of 50 mg of brain tissue was taken from each group. A third-generation variable-speed tissue grinder (TGrinder, China) was used to grind the brain tissue samples in centrifuge tubes until they were completely dissolved in the RIPA lysate. They were then lysed on ice for 30 minutes, centrifuged at 12,000 g for 30 minutes at 4\\u0026deg;C, and the supernatant was collected. All of these steps were done under the mass of the weighed rat brain tissue samples. The protein concentration was determined by the BCA protein assay. Separation gels were prepared according to the molecular weight of the target protein. The samples were boiled in the 4\\u0026times;protein-sampling buffer for 5 minutes and uploaded onto the gel. The upper gel was electrophoresed at 80 V for 30 minutes, while the lower gel was electrophoresed at 120 V for 90 minutes. A current of 250 mA was applied for 80 minutes for the membrane transfer process. After the transfer was completed, the membrane was sealed with protein-free rapid closure solution (1x) for 20 minutes, washed three times with TBST (20xTBST buffer 50 ml\\u0026thinsp;+\\u0026thinsp;950 ml purified water) for 10 min each time, and incubated overnight at 4\\u0026deg;C with the following primary antibodies: Anti-GAPDH, 36 kDa, 1:5000( Proteintech, CAT. NO.10494-1-AP); Anti-KGA/GAC (GLS). 65kDa, 1:5000( Proteintech, CAT. NO.12855-1-AP); Anti-DLD,56kDa, 1:5000( Proteintech, CAT. NO.16431-1-AP); Anti-PDHB,34kDa,1:5000 ( Proteintech, CAT. NO.14744-1-AP); Anti-DBT,53kDa,1:3000 ( Proteintech, CAT. NO.12451-1-AP). Then the membrane was washed three times with TBST for 10 minutes each time. The secondary antibody was incubated at a dilution ratio of 1:5000 for 1 hour at room temperature, and then it was washed three times with TBST for 10 min each time. Each membrane was placed in a fully automated chemiluminescence gel imager for color development after 200 \\u0026micro;l of the ultrasensitive luminescence solution (A: B formulated at 1:1) (Monad, CAT. NO.PW30701S) was prepared. GAPDH was selected as the endogenous control protein.\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eCell culture and AD model development.\\u003c/b\\u003e SH-SY5Y cells were obtained from Wuhan Prosperity Life Science Co. Ltd, China, okadaic acid (OA) for AD cell modeling was purchased from MCE[\\u003cspan citationid=\\\"CR27\\\" class=\\\"CitationRef\\\"\\u003e27\\u003c/span\\u003e], and the cells were purchased from MCE. SH-SY5Y cells were incubated in a cell culture containing 5% C0\\u003csub\\u003e2\\u003c/sub\\u003e at 37\\u0026deg;C. The medium used for the culture of cells was a combination of Dulbecco Modified Eagle medium (DMEM/F-12) with F-12, 14% Sigma Australia Fetal Bovine Serum (FBS), and a 1% Penicillin-Streptomycin Solution mixture. Dimethyl sulfoxide was used to prepare a 4 \\u0026micro;mol/L OA master mix, and the resulting solution was then stored at 4\\u0026deg;C. Using the above complete medium, SH-SY5Y cells were diluted to a density of 1\\u0026times;10\\u003csup\\u003e5\\u003c/sup\\u003e mL and inoculated with 200 \\u0026micro;L per well in 96-well plates. Each well received a full medium supplemented with OA solutions at concentrations of 0, 10, 20, 40, and 80 nmol/L after 24 hours. There were four replicates for every concentration. To compare the effects of different OA concentrations on the proliferation of SH-SY5Y cells 24 hours later, CCK-8 reagent (20 \\u0026micro;L) was added to each well and incubated for 2 hours in exposure to 5% CO\\u003csub\\u003e2\\u003c/sub\\u003e at 37\\u0026deg;C. The absorbance values at 450 nm were measured using a BIOBASE-EL10A enzyme marker (BIOBASE, Shandong, China)[\\u003cspan citationid=\\\"CR28\\\" class=\\\"CitationRef\\\"\\u003e28\\u003c/span\\u003e]. The experiment was conducted three times, and the optimal drug concentration of 25 nM for OA-induced AD in this cell model was found by monitoring the cell growth under a microscope and calculating the absorbance values. Then 200 \\u0026micro;l of SH-SH5Y cells was added to each well of 96-well plates, resulting in a density of 1\\u0026times;10\\u003csup\\u003e5\\u003c/sup\\u003e mL, and the plates were incubated for 24 hours. The cell model was then incubated for 2 hours at densities 0 and 25 nM. The medium was replaced with another one containing 0 and 25 nM of OA and incubated again for 24 hours. The cells were incubated for 24 hours with 20 \\u0026micro;l of OA per well. Then 20 \\u0026micro;l of CCK-8 reagent was added to each well, and the plates were incubated for 2 hours in exposure to 5% C0\\u003csub\\u003e2\\u003c/sub\\u003e at 37\\u0026deg;C. Absorbance values at 450 nm were determined using the BIOBASE-EL10A enzyme marker. The experiment was repeated three times in four replicates to obtain the cell survival index at a drug concentration of 25 nM.\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eRNA extraction and qRT-PCR to verify the differential expression of CRGs mRNA in SH-SY5Y cells.\\u003c/b\\u003e Total RNA samples were extracted using TRIzol reagent. To this end, TRIzol reagent was added to cell samples of the model group, which were induced with 25 nM of OA, and cell samples of the control group, which received no treatment. The subsequent operation was the same as the animal part. GAPDH was selected as the endogenous control gene, and the primers were amplified as follows.\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec12\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eGAPDH-F:5'-CAGGAGGCATTGCTGATGAT-3 '\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec13\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eGAPDH-R:5 '-GAAGGCTGGGGGCTCATTT-3 '\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec14\\\" class=\\\"Section4\\\"\\u003e \\u003ch2\\u003eDLD-F:5 '-GTGATTTACACACACACCCTGA-3 '\\u003c/h2\\u003e \\u003cp\\u003eDLD-R:5 '-GTCTGTCGATTTCTGCCCAA-3 '\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec15\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eGLS-F:5 '-CTGAGCCCTGAAGCAGTTCG-3 '\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec16\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eGLS-R:5 '-AGGAGACCAGCACATCATACCC-3 '\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec17\\\" class=\\\"Section4\\\"\\u003e \\u003ch2\\u003ePDHB-F:5 '-GTGTCTGGCTTGGTGCGGAG-3 '\\u003c/h2\\u003e \\u003cp\\u003ePDHB-R:5 '-ACCTTGTATGCCCCATCATACTG-3 '\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec18\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eDBT-F:5'-CCATTGCATTTGCTCGTGGA-3 '\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec19\\\" class=\\\"Section4\\\"\\u003e \\u003ch2\\u003eDBT-R:5 '-ACCCTGCTCAGTATCCATTGC-3 '\\u003c/h2\\u003e \\u003cp\\u003eFDX1-F:5 '-CTTGTTCAACCTGTCACCTCATCT-3 '\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec20\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eFDX1-R:5 '-CAGCCACTGTTTCAGGCACTC-3 '\\u003c/h2\\u003e \\u003cp\\u003e \\u003cb\\u003eProtein extraction and WB verification of differential expression of CRGs in SH-SY5Y cells.\\u003c/b\\u003e A protein blotting assay was performed to assess the expression of proteins in different groups of cells, with GAPDH as the endogenous reference protein. The culture flasks were washed with pre-cooled PBS, and the cells were collected by using a cell spatula and then centrifuged in a 15-ml tube at 1200 rpm for 5 minutes. The centrifuge tube was filled with 100:1:1 RIPA tissue/cell lysate, protein phosphatase inhibitor mixture, and protease inhibitor and then lysed on ice for 30 minutes. Subsequent operations were the same as those in the animal part.\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eStatistical results.\\u003c/b\\u003e The obtained data were processed and statistically analyzed in R -4.2.1, SPSS-25, and GraphPad Prism-9. Analysis of variance (ANOVA) or t-test was used to determine whether there were significant differences between subgroups in DEG mRNA levels. The t-test was employed to compare RNA and protein expression levels, cell viability, and other variables. In addition, the MWN was analyzed using repeated-measures ANOVA. Differences at P\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05 were considered statistically significant.\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e\\u003cp\\u003e\\u003cstrong\\u003eEthical statement\\u003c/strong\\u003e\\u003cstrong\\u003e. \\u0026nbsp;\\u003c/strong\\u003eThe study is reported in accordance with ARRIVE guidelines.\\u003c/p\\u003e\"},{\"header\":\"Results\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003eIdentification of differentially expressed genes\\u003c/strong\\u003e\\u003cstrong\\u003e.\\u003c/strong\\u003e\\u003cstrong\\u003e\\u0026nbsp; \\u0026nbsp; \\u0026nbsp;\\u003c/strong\\u003eThis study analyzed gene expression array data from 78 human-derived temporal regression samples from the GEO database; Fig. 1. A presents the volcano plots of the 78 DEGs. A total of five differentially expressed CRGs were identified in this study, including one upregulated and four downregulated expressed genes(Figure 1. B). Figure 1. C shows the heatmap of the five differentially expressed related genes.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eBiological function analysis of cuproptosis-related core genes\\u003c/strong\\u003e\\u003cstrong\\u003e.\\u003c/strong\\u003e\\u003cstrong\\u003e\\u0026nbsp; \\u0026nbsp; \\u0026nbsp;\\u003c/strong\\u003eThe obtained differentially expressed CRGs were analyzed for their biological functions. GO analysis showed that the differentially expressed CRGs were mainly enriched in the cellular catabolic/metabolic processes of amino acids, and the cellular fractions were mainly enriched in the mitochondrial matrix oxidoreductase complex. In terms of molecular functions, the oxidoreductase activity, which acts on the donor\\u0026apos;s aldehyde or carbonyl group and the acceptor\\u0026apos;s NAD or NADP, as well as the oxidoreductase activity, which acts on the donor\\u0026apos;s aldehyde or carbonyl groups, were both affected (Fig. 2. A). KEGG enrichment analyses revealed that differentially expressed CRGs were mainly involved in lipoic acid metabolism, the citric acid cycle (TCA cycle), propionic acid metabolism, pyruvate metabolism, valine, leucine and isoleucine degradation, glycolysis, and carbon metabolism (Fig. 2. B).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u0026nbsp;\\u003cstrong\\u003eCorrelation analysis of cuproptosis-related core genes\\u003c/strong\\u003e\\u003cstrong\\u003e.\\u003c/strong\\u003e\\u003cstrong\\u003e\\u0026nbsp; \\u0026nbsp; \\u0026nbsp;\\u003c/strong\\u003eSubsequently, the obtained CRGs were correlated to explore whether they play a synergistic role in the progression of AD. The results indicated that some of the genes, such as PDHB and DLD, exhibit synergistic effects on A. However, DLAT and the remaining four genes showed significant antagonistic effects. In addition, further investigation of the correlation patterns of these CRGs revealed that both DLAT and DLD were significantly correlated with other modulators (Figure 2. C, D).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eImmune infiltration cell analysis of cuproptosis-related core genes\\u003c/strong\\u003e\\u003cstrong\\u003e.\\u003c/strong\\u003e\\u003cstrong\\u003e\\u0026nbsp; \\u0026nbsp; \\u0026nbsp;\\u003c/strong\\u003eImmune infiltration analysis was performed based on the CIBERS0RT algorithm, and the results showed differences between AD patients and healthy individuals in the proportion of 22 infiltrating immune cell types. The heatmap shows the distribution of the 22 infiltrating immune cell types (Figure 3. A). The results indicated that AD patients exhibited higher levels of infiltration in monocytes and neutrophils compared to controls (Figure 3. B). The correlation between the five differentially expressed CRGs and infiltrating immune cells demonstrated that DLD was positively correlated with dendritic-activated cells, M2 macrophages, and mast-activated monocytes, whereas PDHB was basically the same as DLD in the expression of immune correlations. Moreover, GLS showed a positive correlation with dendritic-activated cells (Figure 3. C). These results suggest that CRGs may be a key factor in regulating the molecular and immune infiltration status in AD patients.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCharacterization of epithelial cell aggregation in AD patients\\u003c/strong\\u003e\\u003cstrong\\u003e.\\u003c/strong\\u003e\\u003cstrong\\u003e\\u0026nbsp; \\u0026nbsp; \\u0026nbsp;\\u003c/strong\\u003eTo elucidate the expression patterns associated with cuproptosis in AD, 78 samples were grouped based on the expression profiles of the five CRGs using a consistent clustering algorithm. The number of clusters was most stable when the k-value was set to 2 (k=2) and the CDF curves fluctuated within the smallest range of the common index (0.2 to 0.6) (Fig. 4. A). When the k-value ranged between 2 and 9, the area under the CDF curve exhibited a difference between the two CDF curves (k and k-I) (Fig. 4. B). Clustering the k-values ranging from 2 to 9 showed a concordance score of \\u0026gt;0.9 across subtypes only when the k-value was equal to 2 (Fig. 4.c).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCharacterization of differentiation and immune infiltration of regulatory factors in two clusters of cuproptosis\\u003c/strong\\u003e\\u003cstrong\\u003e.\\u003c/strong\\u003e\\u003cstrong\\u003e\\u0026nbsp; \\u0026nbsp; \\u0026nbsp;\\u003c/strong\\u003eCombined with the heatmap of the shared matrix, 78 AD patients were finally grouped into two clusters, including Cluster 1 (n=46) and Cluster 2 (n=32). To compare the molecular features of the clusters, the expression differences of the five CRGs between Cluster 1 and Cluster 2 were first thoroughly assessed (Fig. 5. A). Different expression profiles of CRGs were observed between the two cuproptosis intoxication patterns, with Cluster 1 exhibiting high expression levels of DBT and Cluster 2 characterized by increased expression of DLD, FDX1, GLS, and PDHB (Figure 5. B). The immune infiltration analysis revealed a difference between the clusters in the immune infiltration profile (Figure 5. C); the percentage of macrophage M1 was higher in Cluster 1, whereas the percentage of dendritic activated cells was higher in Cluster 2 (Figure 5.D).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eDevelopment of a co-expression networks\\u003c/strong\\u003e\\u003cstrong\\u003e.\\u003c/strong\\u003e\\u003cstrong\\u003e\\u0026nbsp; \\u0026nbsp; \\u0026nbsp;\\u003c/strong\\u003eThe WGCNA algorithm was used to develop co-expression networks and modules in healthy individuals and AD patients to identify key gene modules associated with AD. After calculating the variance of each gene expression in GSE 109887, the top 25% of genes with the highest variance were selected for further analysis. Co-expressed gene modules were identified when the value of soft power was set to 11 and the scale-free R2 was equal to 0.5 (Figure 6. A). A total of 10 gene modules with different colors were obtained using the dynamic cutting algorithm, and then a heatmap of the topological overlap matrix (TOM) was drawn (Fig. 6. B-D). These genes were then used in each of the ten color modules in turn to examine the degree of similarity and proximity of module-clinical feature (control and AD) co-expression; the turquoise module demonstrated the strongest correlation with AD (Fig. 6. E). In addition, a positive association was observed between the turquoise module and module-associated genes (Fig. 6. F).\\u003c/p\\u003e\\n\\u003cp\\u003eA similar approach was also employed to further analyze the differentially expressed genes between the two cuproptosis clusters. \\u0026beta;=19 and R2=0.8 were screened as the most suitable soft threshold parameters for developing the scale-free network (Fig. 7. A). The heatmaps demonstrated the TOM of all module-related genes (Fig. 7. B-D). An analysis of the relationship between modules and clinical features (Clusters 1 and 2) revealed a strong correlation between turquoise modules and AD clusters (Fig. 7. E). The correlation analysis also showed a significant relationship between the turquoise module genes and selected modules (Fig. 7. F).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eIdentification of specific genes and GSEA\\u003c/strong\\u003e\\u003cstrong\\u003e.\\u003c/strong\\u003e\\u003cstrong\\u003e\\u0026nbsp; \\u0026nbsp; \\u0026nbsp;\\u003c/strong\\u003eA total of 134 related genes were identified by analyzing the crossover between the modular genes associated with AD (Fig. 8. A). GSVA analysis was used to further investigate the functional differences between the two clusters in terms of cluster-specific DEGs. Cluster 1 was found to be enhanced in non-small cell lung cancer, glycosaminoglycan biosynthesis of keratan sulfate, basal cell carcinoma, Kegel\\u0026apos;s chronic granulocytic leukemia, primary immunodeficiency, drug metabolism of cytochrome P450, isobiotic metabolism by cytochrome P450, and Notch signaling activity (Fig. 8. B). Therefore, it was concluded that it may be involved in various immune responses as well as metabolism.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eDevelopment and evaluation of machine learning models\\u003c/strong\\u003e\\u003cstrong\\u003e.\\u003c/strong\\u003e\\u003cstrong\\u003e\\u0026nbsp; \\u0026nbsp; \\u0026nbsp;\\u003c/strong\\u003eTo identify specific genes with high diagnostic value, four validated machine learning models were developed based on the intersection of 134 cluster-specific DEG and AD differentially expressed genes. These four models were Random Forest Model (RF), Support Vector Machine Model (SVM), Generalized Linear Model (GLM), and Extreme Gradient Boosting (XGB). The \\u0026quot;DALEX\\u0026quot; software package was employed to plot the residual distributions of each model in the test dataset. The SVM and RF models presented relatively low residuals (Fig. 9. A, B). The top 15 significant feature variables of each model were then ranked based on the root mean square error (RMSE) (Fig. 9. C). The discriminative performance of the four machine learning algorithms in the test dataset was evaluated by calculating the subject operating characteristic (ROC) curves based on a 5-fold cross-validation. The SVM showed the highest area under the ROC curve (AUC); the AUC of other models was as follows: GLM=0.585; SVM=0.803; RF=0.761; and XGB=0.624 (Figure 9.D). Combined with these results, the SVM was shown to distinguish well between patients belonging to different clusters. Finally, the top five SVM variables (CACNG3, TNS1, TCEAL2, FAM65C NPTN, and THY1) were selected as predictor genes for further analysis.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eNomogram construction and validation of predictive modeling accuracy\\u003c/strong\\u003e\\u003cstrong\\u003e.\\u003c/strong\\u003e\\u003cstrong\\u003e\\u0026nbsp; \\u0026nbsp; \\u0026nbsp;\\u003c/strong\\u003eA nomogram was developed to estimate the risk of cuproptosis in 78 AD patients to further generalize the RF model to clinical applications (Fig. 10. A). Correction curves and decision curve analysis (DCA) were then employed to evaluate the predictive efficiency of the nomogram model. The calibration curve showed that there was a small error difference between the actual AD aggregation risk and the predicted risk (Fig. 10. B). Since the nomogram\\u0026rsquo;s calibration curve confirmed its high accuracy, it can be used as a basis for clinical decisions (Fig. 10. C). The gene prediction model developed in this study was validated on two external tissue datasets, and the ROC curves showed that it performed well in the GSE 33000 dataset, with an AUC of 0.837 (Fig. 10.D). These results suggest the outstanding diagnostic power of the prediction model.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCharacterization of CRGs expression in animal models of AD\\u003c/strong\\u003e\\u003cstrong\\u003e.\\u003c/strong\\u003e\\u003cstrong\\u003e\\u0026nbsp; \\u0026nbsp; \\u0026nbsp;\\u003c/strong\\u003eAfter the MWM confirmed the model group\\u0026apos;s modeling, hippocampal tissue was collected for qRT-PCR and Western blot experiments to verify CRG expression. No animals perished during the MWM after modeling ten rats in each group, and the analysis revealed that five modeling attempts were successful for the model group (n=5). The MWM was used to test the effect of A\\u0026beta;\\u003csub\\u003e25-35\\u003c/sub\\u003e on cognitive function in SD rats. Figure 11(A) shows the distance traveled by each group of rats in the MWM. The 4-day localization cruise experiment showed that the mean swimming speed of rats in the AD group (treated with A\\u0026beta;\\u003csub\\u003e25-35\\u003c/sub\\u003e) reduced and the avoidance latency increased (Figure 11. B-C, *p\\u0026lt;0.05, **p\\u0026lt;0.01, *p\\u0026lt;0.001). The spatial exploration experiment also indicated that the rats in the model group traversed the platform less frequently than those in the control and saline groups (Fig. 11.D). These results suggest that the memory function was impaired in the AD group. The qRT-PCR showed that the relative expression of DLD, FDX1, GLS, and PDHB mRNA in the model group was lower than that in the control and saline groups, whereas the expression level of DBT mRNA in the model group was higher than that of the other two groups (Fig. 11. E). Furthermore, the relative expression level of DLD, GLS, and PDHB proteins showed a reduction and that of DBT showed an increase in the model group (Figure 11. F). These results indicated that the expression of DLD, FDX1, GLS, and PDHB was downregulated and the expression of DBT was upregulated in the AD model group, which supports the bioinformatic analysis results.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCharacterization of CRGs expression in AD cell models\\u003c/strong\\u003e\\u003cstrong\\u003e.\\u003c/strong\\u003e\\u003cstrong\\u003e\\u0026nbsp; \\u0026nbsp; \\u0026nbsp;\\u003c/strong\\u003eThe modeling drug concentration of 25 nM was determined based on LC50 after OA action on SH-SY5Y cells at various concentrations was observed (Fig. 12. A). Then a drug concentration of 25 nM was added to the cells for 24 hours, and cell viability was determined to investigate the effects of this concentration on the cell proliferation rate (Figure 12. B). The AD model was then created using this drug concentration. According to the qRT-PCR results, the expression of DLD, FDX1, GLS, and PDHB was lower and the expression of DBT was higher in the cells of the drug group than that of the normal group (Fig. 12. C). Western blot results also showed that the DLD, GLS, and PDHB content decreased and the DBT content increased in the drug group (Figure 12.D). These findings support the results of the bioinformatics analysis and the animal model.\\u003c/p\\u003e\"},{\"header\":\"Discussion\",\"content\":\"\\u003cp\\u003eThe pathogenesis and pathologic characteristics of AD are complex. Apart from the primary pathological characteristics such as neurofibrillary tangles, senile plaques, and neuronal loss, AD patients exhibit signs of mitochondrial dysfunction, including changed mitochondrial morphology, decreased activity of cytochrome oxidase (respiratory chain complex IV), and decreased glucose metabolism[29]. Metal ions can cause neurodegeneration in AD patients by upsetting the mitochondrial functional cycle, which has been suggested to be closely associated with mitochondrial respiration by cuproptosis ions and cuproptosis, a type of cell death with cuproptosis dependence[30,31]. Since cuproptosis ions induce A\\u0026beta; aggregation and the generation of reactive oxygen species (ROS), they may play a role in the progression of AD[32]. For example, A\\u0026beta; aggregation-mediated oxidative stress leads to mitochondrial dysfunction and accelerates the progression of AD[33,34]. Ying Zhang et al. suggested that cuproptosis-activated oxidative stress may affect memory function in AD patients[35]. The findings of Alexander Pilozzi et al. also verified the effects of cuproptosis redox on AD amyloid pathology[36]. There is a close relationship between AD and cuproptosis in mitochondrial function and the oxidative respiratory chain. This is consistent with the enrichment analysis results in this study. For example, GO and KEGG analyses showed that CRGs, i.e., FDX1, DLD, and PDHB, are closely related to the oxidative respiratory chain and play a role in mitochondrial functions. This is consistent with the findings of Starkov et al. and Yang Weiguang et al[37,38]. It can be hence assumed that there are similarities between the mechanisms of AD and cuproptosis.\\u003c/p\\u003e\\n\\u003cp\\u003eIt has been shown that immune cells in the human brain, such as microglia and astrocytes, are involved in AD pathogenesis[39]. Several studies have investigated the effects of cuproptosis ions and cuproptosis on the infiltration of immune cells[40,41]. The correlation analysis of CRGs of infiltrating immune cells in this study showed that almost all genes were strongly associated with dendritic cell activation and monocytes. Other studies have also proven the significantly higher infiltration of primitive B cells and neutrophils in the hippocampus of AD patients[42,43]. This is consistent with the results of the immune infiltration analysis for cuproptosis core genes. Subsequent experiments revealed the high similarity between DLD and PDHB in the infiltration of immune cells, supporting the strong positive correlation between these two variables, as shown by the correlation analysis.\\u003c/p\\u003e\\n\\u003cp\\u003eThe study results confirmed the acceptable diagnostic power of the AD risk prediction model developed based on five cuproptosis genes (including DLD, FDX1, GLS, PDHB, and DBT). FDX1 gene promotes lipoylation of pyruvate dehydrogenase (PDH) and \\u0026alpha;-ketoglutarate dehydrogenase (\\u0026alpha;-KDH), which in turn affect the TCA cycle. It has been further investigated in studies on neurodegenerative disorders and cuproptosis mortality[44,45]. On the other hand, the pyruvate dehydrogenase E1 subunit beta (PDHB) gene and the dihydrolipoic acid dehydrogenase (DLD) gene are involved in encoding the pyruvate dehydrogenase complex (PDC), which is essential for mitochondrial respiration[46] and is also closely related to the regulation of FDX1 and AD. The branched-chain alpha-ketoacid dehydrogenase complex (BCKD) is an endomitochondrial enzyme complex that influences the catabolism of the branched-chain amino acids isoleucine, leucine, and valine. It has been closely linked to the pathogenesis of AD. The dihydrolipoamide branched-chain transacylase E2 (DBT) gene encoding the transacylase (E2) subunit is also involved in the formation of this complex[47]. In contrast, the glutaminase (GLS) gene, an ammonia metabolism gene, has been reported to be associated with a wide range of glutamate signaling disorders, including AD[48]. An algorithmic analysis revealed a significant difference between the AD group and the normal group in the expression level of DLD, FDX1, GLS, and PDHB. The results showed that the DBT expression increased and the DLD expression decreased in the AD group. Considering the results of validation experiments performed on SH-SY5Y cells of SD rats, there is a need for further studies to investigate the exact mechanism of action. GSEA analysis showed all of these key genes to be associated with neurodegenerative diseases. The difference between Cluster 1 and Cluster 2 in the marker pathway activity also revealed a significant increase in Notch signaling pathway activity. Arunima Kapoor et al. and Perna et al. also reported a relationship between this signaling pathway and AD [49,50]. The accuracy of the AD diagnostic model developed based on CRGs was revalidated using external datasets, and then a nomogram was developed to assess the risk of AD using the CRGs score. The calibration curve and DCA confirmed the validity and the improved application of the graph.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eLimitations\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eSimilar to any other research project, this study faced some limitations. Firstly, there were not enough relevant AD samples or datasets for further analysis. Second, this study demonstrated the correlation between cuproptosis and AD, but the specific mechanism of action of differentially expressed genes in AD was not validated. Meanwhile the single method of modeling animal models needs to be improved.\\u003c/p\\u003e\"},{\"header\":\"Conclusions\",\"content\":\"\\u003cp\\u003eThis study investigated the complex relationship between cuproptosis and AD by identifying five core cuproptosis genes (DLD, FDX1, GLS, PDHB, and DBT) associated with AD. The in vivo and in vitro experiments also validated the expression level of these five core cuproptosis genes in AD models. In addition, a practical prognostic risk model was developed to assess the risk of cuproptosis and the pathological consequences of AD. The correlation between the five core cuproptosis genes and AD was validated by multiple data analyses and in vivo and ex vivo experiments. The study results revealed the downregulation of DLD, FDX1, GLS, and PDHB and the upregulation of DBT in AD patients, compared with healthy individuals. This differential expression is expected to provide new ideas for the diagnosis and prognostic monitoring of AD.\\u003c/p\\u003e\\n\"},{\"header\":\"Declarations\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003eData availability\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eTe\\u0026nbsp;data\\u0026nbsp;used\\u0026nbsp;to\\u0026nbsp;support\\u0026nbsp;the\\u0026nbsp;fndings\\u0026nbsp;of\\u0026nbsp;this\\u0026nbsp;study\\u0026nbsp;is\\u0026nbsp;available.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u0026nbsp;\\u003cstrong\\u003eAcknowledgments\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eWe thank the GEO and KEGG databases for providing the platform and contributors for uploading meaningful datasets.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eAuthor\\u0026nbsp;contributions\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eR.H, Z.X, M.Y.Q, C.Y.L, G.Y.W, Y.Y.W, Z.S.H, and M.Y.D participated in this study.\\u0026nbsp;R.H\\u0026nbsp;conducted the validation experiments and collected the experimental data, and worked with Z .X\\u0026nbsp;on the statistical analysis and paper writing. M.Y.Q, C.Y.L, G.Y.W and Y.Y.W assisted in the animal feeding and sampling process. Z.S.H\\u0026nbsp;and\\u0026nbsp;M.Y.D participated in the conceptualization of the whole article and revised the manuscript, and participated in the technical support. All authors approved the manuscript.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eFunding\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eProject of Enhancement of Basic Research Ability of Young and Middle-aged Teachers in Guangxi Universities (No.2023KY0554); Baise City Science and Technology Program Project (No.20211807, No.20224139).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCompeting\\u003c/strong\\u003e\\u003cstrong\\u003e\\u0026nbsp;\\u003c/strong\\u003e\\u003cstrong\\u003einterests\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eTe authors declare no competing interests.\\u003c/p\\u003e\"},{\"header\":\"References\",\"content\":\"\\u003col\\u003e\\n\\u003cli\\u003eG\\u0026oacute;nzalez de San Rom\\u0026aacute;n E, Manuel I, Giralt MT, Ferrer I, Rodr\\u0026iacute;guez-Puertas R. Imaging mass spectrometry (IMS) of cortical lipids from preclinical to severe stages of Alzheimer\\u0026apos;s disease. Biochim Biophys Acta Biomembr, 2017;1859(9 Pt B):1604-1614.\\u003c/li\\u003e\\n\\u003cli\\u003eKent SA, Spires-Jones TL, Durrant CS. The physiological roles of tau and A\\u0026beta;: implications for Alzheimer\\u0026apos;s disease pathology and therapeutics. Acta Neuropathol, 2020;140(4):417-447.\\u003c/li\\u003e\\n\\u003cli\\u003eRostagno AA. Pathogenesis of Alzheimer\\u0026apos;s Disease . Int J Mol Sci. 2022;24(1):107.\\u003c/li\\u003e\\n\\u003cli\\u003eDi Fede G, Catania M, Maderna E, Ghidoni R, Benussi L, Tonoli E, Giaccone G, Moda F, Paterlini A, Campagnani I, Sorrentino S, Colombo L, Kubis A, Bistaffa E, Ghetti B, Tagliavini F. Molecular subtypes of Alzheimer\\u0026apos;s disease. Sci Rep. 2018;8(1):3269.\\u003c/li\\u003e\\n\\u003cli\\u003eScheltens P, De Strooper B, Kivipelto M, Holstege H, Ch\\u0026eacute;telat G, Teunissen CE, Cummings J, van der Flier WM. Alzheimer\\u0026apos;s disease. Lancet. 2021;397(10284):1577-1590.\\u003c/li\\u003e\\n\\u003cli\\u003eKawamata H, Manfredi G. Import, maturation, and function of SOD1 and its copper chaperone CCS in the mitochondrial intermembrane space. Antioxid Redox Signal. 2010;13(9):1375-1384.\\u003c/li\\u003e\\n\\u003cli\\u003eScheiber I, Dringen R, Mercer JF. Copper: effects of deficiency and overload. Met Ions Life Sci. 2013;13:359-387.\\u003c/li\\u003e\\n\\u003cli\\u003eGromadzka G, Tarnacka B, Flaga A, Adamczyk A. Copper Dyshomeostasis in Neurodegenerative Diseases-Therapeutic Implications. Int J Mol Sci. 2020;21(23):9259.\\u003c/li\\u003e\\n\\u003cli\\u003eSayre LM, Perry G, Harris PL, Liu Y, Schubert KA, Smith MA. In situ oxidative catalysis by neurofibrillary tangles and senile plaques in Alzheimer\\u0026apos;s disease: a central role for bound transition metals. J Neurochem. 2000;74(1):270-279.\\u003c/li\\u003e\\n\\u003cli\\u003eChen LL, Fan YG, Zhao LX, Zhang Q, Wang ZY. The metal ion hypothesis of Alzheimer\\u0026apos;s disease and the anti-neuroinflammatory effect of metal chelators. Bioorg Chem. 2023;131:106301.\\u003c/li\\u003e\\n\\u003cli\\u003eTsvetkov P, Coy S, Petrova B, Dreishpoon M, Verma A, AbdusamAD M, Rossen J, Joesch-Cohen L, Humeidi R, Spangler RD, Eaton JK, Frenkel E, Kocak M, Corsello SM, Lutsenko S, Kanarek N, Santagata S, Golub TR. Copper induces cell death by targeting lipoylated TCA cycle proteins . Science. 2022 Apr 22;376(6591):eabq4855].\\u003c/li\\u003e\\n\\u003cli\\u003eXie J, Yang Y, Gao Y, He J. Cuproptosis: mechanisms and links with cancers . Mol Cancer, 2023;22(1):46.\\u003c/li\\u003e\\n\\u003cli\\u003eSwerdlow RH. Mitochondria and Mitochondrial Cascades in Alzheimer\\u0026apos;s Disease. J Alzheimers Dis. 2018;62(3):1403-1416.\\u003c/li\\u003e\\n\\u003cli\\u003eMacdonald R, Barnes K, Hastings C, Mortiboys H. Mitochondrial abnormalities in Parkinson\\u0026apos;s disease and Alzheimer\\u0026apos;s disease: can mitochondria be targeted therapeutically?. Biochem Soc Trans. 2018;46(4):891-909.\\u003c/li\\u003e\\n\\u003cli\\u003eTang D, Chen X, Kroemer G. Cuproptosis: a copper-triggered modality of mitochondrial cell death. Cell Res. 2022;32(5):417-418.\\u003c/li\\u003e\\n\\u003cli\\u003eGarza NM, Swaminathan AB, Maremanda KP, Zulkifli M, Gohil VM. Mitochondrial copper in human genetic disorders. \\u003cem\\u003eTrends Endocrinol Metab\\u003c/em\\u003e. 2023;34(1):21-33.\\u003c/li\\u003e\\n\\u003cli\\u003eWang Q, Sun J, Chen T, Song S, Hou Y, Feng L, Fan C, Li M. Ferroptosis, Pyroptosis, and Cuproptosis in Alzheimer\\u0026apos;s Disease. ACS Chem Neurosci. 2023;14(19):3564-3587.\\u003c/li\\u003e\\n\\u003cli\\u003eLardenoije R, Roubroeks JAY, Pishva E, Leber M, Wagner H, Iatrou A, Smith AR, Smith RG, Eijssen LMT, Kleineidam L, Kawalia A, Hoffmann P, Luck T, Riedel-Heller S, Jessen F, Maier W, Wagner M, Hurlemann R, Kenis G, Ali M, Del Sol A, Mastroeni D, Delvaux E, Coleman PD, Mill J, Rutten BPF, Lunnon K, Ramirez A, van den Hove DLA.Alzheimer\\u0026apos;s disease-associated (hydroxy)methylomic changes in the brain and blood. \\u003cem\\u003eClin Epigenetics\\u003c/em\\u003e. 2019;11(1):164.\\u003c/li\\u003e\\n\\u003cli\\u003eNarayanan M, Huynh JL, Wang K, Yang X, Yoo S, McElwee J, Zhang B, Zhang C, Lamb JR, Xie T, Suver C, Molony C, Melquist S, Johnson AD, Fan G, Stone DJ, Schadt EE, Casaccia P, Emilsson V, Zhu J. Common dysregulation network in the human prefrontal cortex underlies two neurodegenerative diseases. Mol Syst Biol. 2014;10(7):743.\\u003c/li\\u003e\\n\\u003cli\\u003eRitchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, Smyth GK. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43(7):e47.\\u003c/li\\u003e\\n\\u003cli\\u003eVarma S. Blind estimation and correction of microarray batch effect. PLoS One. 2020;15(4):e0231446.\\u003c/li\\u003e\\n\\u003cli\\u003eNewman AM, Liu CL, Green MR, Gentles AJ, Feng W, Xu Y, Hoang CD, Diehn M, Alizadeh AA. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods. 2015;12(5):453-457.\\u003c/li\\u003e\\n\\u003cli\\u003eH\\u0026auml;nzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data [J]. BMC Bioinformatics, 2013, 14(7): 1471-2105.\\u003c/li\\u003e\\n\\u003cli\\u003eStepanichev MYu, Lazareva NA, Onufriev MV, Mitrokhina OS, Moiseeva YuV, Gulyaeva NV. Effects of doses of fragment (25-35) of beta-amyloid peptide on behavior in rats. Neurosci Behav Physiol. 1998;28(5):564-566.\\u003c/li\\u003e\\n\\u003cli\\u003eLi Z, Tong Q, Xu H, et al. Therapeutic Effects of TianDiJingWan on the A\\u0026beta; 25-35-Induced Alzheimer\\u0026apos;s Disease Model Rats. Evid Based Complement Alternat Med. 2015;2015:307350.\\u003c/li\\u003e\\n\\u003cli\\u003eOthman MZ, Hassan Z, Che Has AT. Morris water maze: a versatile and pertinent tool for assessing spatial learning and memory. Exp Anim. 2022;71(3):264-280.\\u003c/li\\u003e\\n\\u003cli\\u003eAmonruttanapun P, Chongthammakun S, Chamniansawat S. The effects of okadaic acid-treated SH-SY5Y cells on microglia activation and phagocytosis. Cell Biol Int. 2022 Feb;46(2):234-242.\\u003c/li\\u003e\\n\\u003cli\\u003eYang X, Zhong Y, Wang D, Lu Z. A simple colorimetric method for viable bacteria detection based on cell counting Kit-8. Anal Methods. 2021;13(43):5211-5215.\\u003c/li\\u003e\\n\\u003cli\\u003eZhu X, Smith MA, Perry G, Aliev G. Mitochondrial failures in Alzheimer\\u0026apos;s disease. \\u003cem\\u003eAm J Alzheimers Dis Other Demen\\u003c/em\\u003e. 2004;19(6):345-352.\\u003c/li\\u003e\\n\\u003cli\\u003eXie J, Yang Y, Gao Y, He J. Cuproptosis: mechanisms and links with cancers. Mol Cancer. 2023;22(1):46.\\u003c/li\\u003e\\n\\u003cli\\u003eMezzaroba L, Alfieri DF, Colado Sim\\u0026atilde;o AN, Vissoci Reiche EM. The role of zinc, copper, manganese and iron in neurodegenerative diseases. Neurotoxicology. 2019;74:230-241.\\u003c/li\\u003e\\n\\u003cli\\u003eHureau C, Faller P. Abeta-mediated ROS production by Cu ions: structural insights, mechanisms and relevance to Alzheimer\\u0026apos;s disease. Biochimie. 2009;91(10):1212-1217.\\u003c/li\\u003e\\n\\u003cli\\u003eButterfield DA, Boyd-Kimball D. Redox proteomics and amyloid \\u0026beta;-peptide: insights into Alzheimer disease. J Neurochem. 2019;151(4):459-487.\\u003c/li\\u003e\\n\\u003cli\\u003eJansen IE, van der Lee SJ, Gomez-Fonseca D, de Rojas I, Dalmasso MC, Grenier-Boley B, Zettergren A, Mishra A, Ali M, Andrade V, Bellenguez C, Kleineidam L, K\\u0026uuml;\\u0026ccedil;\\u0026uuml;kali F, Sung YJ, Tes\\u0026iacute; N, Vromen EM, Wightman DP, Alcolea D, Alegret M, Alvarez I, Amouyel P, Athanasiu L, Bahrami S, Bailly H, Belbin O, Bergh S, Bertram L, Biessels GJ, Blennow K, Blesa R, Boada M, Boland A, Buerger K, Carracedo \\u0026Aacute;, Cervera-Carles L, Chene G, Claassen JAHR, Debette S, Deleuze JF, de Deyn PP, Diehl-Schmid J, Djurovic S, Dols-Icardo O, Dufouil C, Duron E, D\\u0026uuml;zel E; EADB consortium; Fladby T, Fortea J, Fr\\u0026ouml;lich L, Garc\\u0026iacute;a-Gonz\\u0026aacute;lez P, Garcia-Martinez M, Giegling I, Goldhardt O, Gobom J, Grimmer T, Haapasalo A, Hampel H, Hanon O, Hausner L, Heilmann-Heimbach S, Helisalmi S, Heneka MT, Hern\\u0026aacute;ndez I, Herukka SK, Holstege H, Jarholm J, Kern S, Knapskog AB, Koivisto AM, Kornhuber J, Kuulasmaa T, Lage C, Laske C, Leinonen V, Lewczuk P, Lle\\u0026oacute; A, de Munain AL, Lopez-Garcia S, Maier W, Marqui\\u0026eacute; M, Mol MO, Montrreal L, Moreno F, Moreno-Grau S, Nicolas G, N\\u0026ouml;then MM, Orellana A, P\\u0026aring;lhaugen L, Papma JM, Pasquier F, Perneczky R, Peters O, Pijnenburg YAL, Popp J, Posthuma D, Pozueta A, Priller J, Puerta R, Quintela I, Ramakers I, Rodriguez-Rodriguez E, Rujescu D, Saltvedt I, Sanchez-Juan P, Scheltens P, Scherbaum N, Schmid M, Schneider A, Selb\\u0026aelig;k G, Selnes P, Shadrin A, Skoog I, Soininen H, T\\u0026aacute;rraga L, Teipel S; GR@ACE study group; Tijms B, Tsolaki M, Van Broeckhoven C, Van Dongen J, van Swieten JC, Vandenberghe R, Vidal JS, Visser PJ, Vogelgsang J, Waern M, Wagner M, Wiltfang J, Wittens MMJ, Zetterberg H, Zulaica M, van Duijn CM, Bjerke M, Engelborghs S, Jessen F, Teunissen CE, Pastor P, Hiltunen M, Ingelsson M, Andreassen OA, Clarim\\u0026oacute;n J, Sleegers K, Ruiz A, Ramirez A, Cruchaga C, Lambert JC, van der Flier W. Genome-wide meta-analysis for Alzheimer\\u0026apos;s disease cerebrospinal fluid biomarkers. Acta Neuropathol. 2022;144(5):821-842.\\u003c/li\\u003e\\n\\u003cli\\u003eZhang Y, Zhou Q, Lu L, Su Y, Shi W, Zhang H, Liu R, Pu Y, Yin L. Copper Induces Cognitive Impairment in Mice via Modulation of Cuproptosis and CREB Signaling. Nutrients. 2023;15(4):972.\\u003c/li\\u003e\\n\\u003cli\\u003ePilozzi A, Yu Z, Carreras I, Cormier K, Hartley D, Rogers J, Dedeoglu A, Huang X. A Preliminary Study of Cu Exposure Effects upon Alzheimer\\u0026apos;s Amyloid Pathology. \\u003cem\\u003eBiomolecules\\u003c/em\\u003e. 2020;10(3):408.\\u003c/li\\u003e\\n\\u003cli\\u003eStarkov AA, Fiskum G, Chinopoulos C, Lorenzo BJ, Browne SE, Patel MS, Beal MF. Mitochondrial alpha-ketoglutarate dehydrogenase complex generates reactive oxygen species. J Neurosci. 2004;24(36):7779-7788.\\u003c/li\\u003e\\n\\u003cli\\u003eYang W, Guo Q, Wu H, Tong L, Xiao J, Wang Y, Liu R, Xu L, Yan H, Sun Z. Comprehensive analysis of the cuproptosis-related gene DLD across cancers: A potential prognostic and immunotherapeutic target. Front Pharmacol. 2023;14:1111462.\\u003c/li\\u003e\\n\\u003cli\\u003eRajesh Y, Kanneganti TD. Innate Immune Cell Death in Neuroinflammation and Alzheimer\\u0026apos;s Disease. Cells. 2022;11(12):1885.\\u003c/li\\u003e\\n\\u003cli\\u003eSuliman IH, Kim K, Chen W, et al. Metal-Based Nanoparticles for Cancer Metalloimmunotherapy. Pharmaceutics. 2023;15(7):2003.\\u003c/li\\u003e\\n\\u003cli\\u003eWang L, Cao Y, Guo W, Xu J. High expression of cuproptosis-related gene FDX1 in relation to good prognosis and immune cells infiltration in colon adenocarcinoma (COAD). J Cancer Res Clin Oncol. 2023;149(1):15-24.\\u003c/li\\u003e\\n\\u003cli\\u003eHuang LT, Zhang CP, Wang YB, Wang JH. Association of Peripheral Blood Cell Profile With Alzheimer\\u0026apos;s Disease: A Meta-Analysis. Front Aging Neurosci. 2022;14:888946.\\u003c/li\\u003e\\n\\u003cli\\u003eSong L, Yang YT, Guo Q; ZIB Consortium, Zhao XM. Cellular transcriptional alterations of peripheral blood in Alzheimer\\u0026apos;s disease. BMC Med. 2022;20(1):266.\\u003c/li\\u003e\\n\\u003cli\\u003eTsvetkov P, Coy S, Petrova B, Dreishpoon M, Verma A, Abdusamad M, Rossen J, Joesch-Cohen L, Humeidi R, Spangler RD, Eaton JK, Frenkel E, Kocak M, Corsello SM, Lutsenko S, Kanarek N, Santagata S, Golub TR. Copper induces cell death by targeting lipoylated TCA cycle proteins [published correction appears in Science. 2022 Apr 22;376(6591):eabq4855]. Science. 2022;375(6586):1254-1261.\\u003c/li\\u003e\\n\\u003cli\\u003eZhang Y, Zhou Q, Lu L, Su Y, Shi W, Zhang H, Liu R, Pu Y, Yin L. Copper Induces Cognitive Impairment in Mice via Modulation of Cuproptosis and CREB Signaling. Nutrients. 2023;15(4):972.\\u003c/li\\u003e\\n\\u003cli\\u003ePatel MS, Nemeria NS, Furey W, Jordan F. The pyruvate dehydrogenase complexes: structure-based function and regulation. J Biol Chem. 2014;289(24):16615-16623.\\u003c/li\\u003e\\n\\u003cli\\u003eHull J, Usmari Moraes M, Brookes E, Love S, Conway ME. Distribution of the branched-chain \\u0026alpha;-ketoacid dehydrogenase complex E1\\u0026alpha; subunit and glutamate dehydrogenase in the human brain and their role in neuro-metabolism. Neurochem Int. 2018;112:49-58.\\u003c/li\\u003e\\n\\u003cli\\u003eGriffin JWD, Liu Y, Bradshaw PC, Wang K. In Silico Preliminary Association of Ammonia Metabolism Genes GLS, CPS1, and GLUL with Risk of Alzheimer\\u0026apos;s Disease, Major Depressive Disorder, and Type 2 Diabetes. J Mol Neurosci. 2018;64(3):385-396.\\u003c/li\\u003e\\n\\u003cli\\u003eKapoor A, Nation DA. Role of Notch signaling in neurovascular aging and Alzheimer\\u0026apos;s disease. Semin Cell Dev Biol. 2021;116:90-97.\\u003c/li\\u003e\\n\\u003cli\\u003ePerna A, Marathe S, Dreos R, Falquet L, Akarsu Egger H, Auber LA. Revealing NOTCH-dependencies in synaptic targets associated with Alzheimer\\u0026apos;s disease. Mol Cell Neurosci. 2021;115:103657.\\u003c/li\\u003e\\n\\u003c/ol\\u003e\"}],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":true,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":false,\"hideJournal\":true,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":false,\"isAuthorSuppliedPdf\":false,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":false,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true},\"keywords\":\"Alzheimer's disease1, Cuproptosis2, immune infiltration3, machine learning model4, Nomogram5\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-3854023/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-3854023/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003eThis study aimed to validate the correlation between core cuproptosis genes (CRGs) and Alzheimer's disease (AD) from both bioinformatics and experimental perspectives and also to develop a risk prediction model. To this end, 78 human-derived temporal back samples were analyzed in GSE109887, and then the biological functions of the resulting CRGs were explored by cluster analysis, weighted gene co-expression network analysis (WGCNA), and similar methods to identify the best machine model. Moreover, a nomogram was developed to validate the model. The mRNA and protein expression of CRGs were validated using the SH-SY5Y cell model and SD rat animal model. The RT-qPCR and western blot results showed that the mRNA and protein expression content of DLD, FDX1, GLS, and PDHB decreased, and the DBT expression content increased in AD, which supported the bioinformatic analysis results. CRGs expression alterations affected the aggregation and infiltration of certain immune cells. The study results also confirmed the accuracy and validity of AD diagnostic models and nomograms. This study validated the correlation between five CRGs and AD, indicating a significant difference between AD patients and healthy individuals. Therefore, CRGs are expected to serve as relevant biomarkers for the diagnosis and prognostic monitoring of AD.\\u003c/p\\u003e\",\"manuscriptTitle\":\"Construction and validation of a bioinformatics-based screen for Cuproptosis-related genes and risk model for Alzheimer's disease\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2024-01-23 15:49:53\",\"doi\":\"10.21203/rs.3.rs-3854023/v1\",\"editorialEvents\":[{\"type\":\"communityComments\",\"content\":0}],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true}}],\"origin\":\"\",\"ownerIdentity\":\"e5bfc1ee-3a92-496d-9d7d-7292dfbab08d\",\"owner\":[],\"postedDate\":\"January 23rd, 2024\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"posted\",\"subjectAreas\":[],\"tags\":[],\"updatedAt\":\"2024-04-09T11:15:12+00:00\",\"versionOfRecord\":[],\"versionCreatedAt\":\"2024-01-23 15:49:53\",\"video\":\"\",\"vorDoi\":\"\",\"vorDoiUrl\":\"\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-3854023\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-3854023\",\"identity\":\"rs-3854023\",\"version\":[\"v1\"]},\"buildId\":\"qtupq5eGEP_6zYnWcrvyt\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}