Identification and Validation of Metabolism-Related Genes in Alzheimer’s Disease

doi:10.21203/rs.3.rs-2954667/v1

Download PDF

Research Article

Identification and Validation of Metabolism-Related Genes in Alzheimer’s Disease

https://doi.org/10.21203/rs.3.rs-2954667/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 15 Sep, 2023

Read the published version in Journal of Translational Medicine →

You are reading this latest preprint version

Background：Due to its heterogeneity, the pathogenic mechanisms underlying Alzheimer's disease (AD) are not yet fully elucidated. Emerging evidence has demonstrated the critical role of metabolism in the pathophysiology of AD. This study explored the metabolism-related signature genes of AD and precisely identified AD molecular subclasses.

Methods: The AD datasets were obtained from the NCBI GEO, and metabolism-relevant genes were downloaded from a previously published compilation. Consensus clustering was utilized to identify AD subclasses. We evaluated the clinic characteristics, correlations with metabolic signatures and immune infiltration of the AD subclasses. Feature genes were screened by WGCNA and processed for GO and KEGG pathway analysis. Furthermore, we used three machine learning algorithms to further narrow down the selection of feature genes. Finally, we identified the diagnostic value and expression of feature genes using dataset and RT-PCR analysis.

Results: Three subclasses of AD were identified and designated as MCA, MCB, and MCC. MCA had high AD progression signatures and maybe a high-risk subgroup compared to the other two groups. MCA displayed high glycolysis, fructose and galactose metabolism, whereas citrate cycle and pyruvate metabolism were decreased, associating with high immune infiltration. Conversely, MCB is chiefly involved in the citrate cycle and exhibits elevated expression of immune checkpoint genes. Through WGCNA, a set of 101 metabolic genes were discovered to having the strongest association with the poor progression of AD. Ultimately, the application of machine learning algorithms enabled us to successfully pinpoint eight feature genes. Employing the nomogram based on the eight feature genes translates to distinct clinical benefits for the patients. As indicated by the datasets and RT-PCR analysis, these eight distinctive genes are intimately linked to the advancement of the AD.

Conclusion: Metabolic dysfunction is correlated with AD. Hypothetical molecular subclasses based on metabolic genes may provide new insights for individualized therapy of AD. The metabolic feature genes most robust correlation with the advancement of AD were GFAP, CYB5R3, DARS, KIAA0513, EZR, KCNC1, COLEC12 and TST.

Alzheimer’s disease

metabolic subclass

immune infiltration

characteristic genes

machine learning

As the most prevalent type of dementia, Alzheimer's disease (AD) impacts over 50 million individuals globally[1]. The primary pathological characteristics of AD are the buildup of amyloid-β (Aβ) plaques and intraneuronal neurofibrillary tangles (NFTs) [2]. Aβ arises from successive enzymatic breakdown of Amyloid Precursor Protein by β-secretase and γ-secretase[3]. Despite decades of research, the pathogenic mechanism of AD remains unclear, and current treatments are unsatisfactory, let alone curative[4]. Therefore, early diagnosis and intervention are crucial for AD patients, which may offer more significant benefits. However, diagnosing AD has long been challenging, and current biomarkers are inadequate to provide personalized genetic-level treatments. Molecular subtypes can better uncover heterogeneity among AD patients and bridge the gap towards targeted AD therapies.

Mounting evidence suggests that AD is a wide-ranging metabolic disorder characterized by disrupted glycolipid and energy metabolism. These metabolic abnormalities may contribute to the severity of AD neuropathology and the eventual manifestation of AD symptoms[5–9]. These findings emphasize the significant role of metabolism in AD and elevate its prominence in AD research. It is an urgent need to explore metabolism-related signature genes of AD and precisely identify AD molecular subclasses.

In this study, we integrated eight AD datasets, comprising a total of 737 patients, into a single dataset for clustering analysis based on metabolic genes. Through consensus clustering, we identified three distinct subclasses within AD, and the three subclasses were designated as Metabolism Correlated A(MCA) subgroup, B(MCB) subgroup and C(MCC) subgroup. Subsequently, we conducted an evaluation of clinical characteristics, correlations with metabolic signatures, immune infiltration patterns, and prognostic implications for these AD subclasses. Furthermore, employing the “WGCNA” R package, we identified the module that most associated with poor progression of AD and performed functional enrichment analysis on the genes within this module. To further narrow down the selection of feature genes, three machine learning algorithms were utilized, including Support Vector Machines (SVM), LASSO regression and Random Forest (RF). Through this process, we successfully identified eight core genes exhibiting outstanding diagnostic potential and serving as promising therapeutic targets for AD.

2.1 Data collection and processing

Gene expression data of AD patients were sourced from the NCBI GEO database (https://www.ncbi.nlm.nih.gov/geo/) [10]. The eight obtained datasets were as follows: GSE48350, GSE5281, GSE28146, GSE122063, GSE118553, GSE8442201(GSE84422 includes three subsets respectively. GSE8442201 annotated by GPL570), GSE132903, GSE106241. A detailed description of these datasets can be found in Table S1. We performed data filtering, background correction, log2 transformation, and normalization procedures on these datasets. In addition, we merged the datasets and applied batch correction using the Combat method from the "sva" package.

2.2 Identification of AD subclasses

For subsequent consensus clustering[11], we utilized a previously published compilation of 2,752 metabolism-relevant genes[12], which encode all known metabolic enzymes and transporters in humans. Our aim was to classify the AD samples into distinct subclasses using consensus clustering. We set the maximum number of clusters to 5 and applied a filter based on a cluster consensus score threshold of > 0.8.

2.3 Gene set variation analysis (GSVA)

GSVA represents an unsupervised and nonparametric approach to gene set enrichment that estimates the score attributed to a particular pathway or signature based on transcriptomic data[13]. In this study, we acquired 84 metabolism-relevant gene signatures from previously published research [12]. By utilizing the GSVA R package, we calculated 120 scores for each sample corresponding to these 84 metabolism signatures.

2.4 Evaluation of Immune Infiltration

Various algorithms are employed to assess the status of immune infiltration. The "XCELL" package was employed to quantify the relative abundance of immune and stromal cells between AD subclasses utilizing the gene expression profiles. In this research, the EPIC[14], ssGSEA[15], quanTIseq[16], TIMER[17], CIBERSORT[18], MCPCounter[19], XCELL[19] and ESTIMATE[20] algorithms were used to calculate the ESTIMATE score cells and the abundance of immune.

2.5 Weighted correlation network analysis (WGCNA)

We utilized the WGCNA package to establish a WGCNA network aimed at identifying gene modules associated with three subclasses and clinical characteristics in AD patients[21]. To determine the optimal soft threshold power, we employed the scale-free topology standard. Subsequently, we generated a weighted adjacency matrix and the transformation of a topological overlap matrix. Hierarchical clustering tree analysis was employed to screen modules containing more than 50 genes. Each module was visually represented using an arbitrary color. The module eigengene represented each of the distinct modules. The traits examined in this study included AD subclasses and several clinical features, such as NFT, braak, ect.

2.6 Functional enrichment analysis

The R package "clusterProfiler"[22] was used to perform Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis on the functions and pathways of hub genes in the cyan module.

2.7 Machine learning

Stable and robust features play a crucial role in forecasting the onset and advancement of AD. We developed three machine learning models: RF, LASSO regression and SVM. The RF algorithm, known for its effectiveness and popularity, utilizes a majority voting approach to combine decision trees, resulting in superior precision and rapid autonomous learning across diverse datasets. The LASSO regression algorithm, a well-established linear prediction method, makes predictions based on regression coefficients and has been extensively utilized in various fields over the past decades [23]. The SVM algorithm, a widely used machine learning technique, projects input data into a higher-dimensional feature space by mapping a kernel function, facilitating easier classification compared to the original feature space[24]. Through an iterative learning process, SVM converges to the optimal hyperplane that maximizes interclass span. These models of machine learning were built based on earlier study[25].

2.8 Establishment and assessment of a nomogram

The combined dataset comprised a total of 1,262 samples, featuring 525 normal samples and 737 AD samples. These samples were randomly partitioned into a testing dataset (20%, N = 252) and a training dataset (80%, N = 1,010). The identified feature genes were utilized to develop a nomogram using the “rms” package in training set. The effectiveness of the nomogram was assessed separately in both the test set and the training set. Calibration curves were employed to assess the predictive performance of the nomogram model. Ultimately, the clinical value of the model was assessed through decision curve analysis (DCA) and by examining the area under the curve (AUC) values.

2.9 Assessment of the diagnostic significance of feature genes in AD

To assess the discriminative capability of the signature genes in distinguishing between non-AD controls and AD, we employed eight datasets: GSE5281, GSE48350, GSE118553, GSE28146, GSE122063, GSE132903, GSE8442201, and GSE1297. The diagnostic performance of these signature genes was visualized by plotting the AUC using the R package of “pROC”.

2.10 Animals

The male P301S transgenic mice and wild-type C57BL/6J male mice were housed under standard laboratory conditions and maintained at an artificial 12 h/12 h light/dark cycle. Food and water were available ad libitum. All animal experiments were reviewed and approved by the Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology.

2.11 Real-Time Quantitative PCR

Total RNA was extracted from cortices of P301S transgenic mice and WT mice by utilizing TRIzol. The mRNA was reverse transcribed to cDNA using reverse transcription kits (Takara, Japan) according to the manufacturer's instructions. The cDNA and primers, along with ChamQ SYBR qPCR Master Mix (Vazyme, China), were mixed in the PCR plate, after which the transcription level of GFAP, CYB5R3, DARS, KIAA0513, EZR, KCNC1, COLEC12 and TST were analyzed by the StepOnePlus real-time PCR System. All experiments were repeated three times, and the primer sequences are displayed in Table S2.

2.12 Statistical analysis

The statistical analyses were conducted using R language (version 4.2.0). The comparison between groups was conducted using Wilcoxon test. A value of P < 0.05 was deemed to be statistically significant.

3.1 Consensus clustering identifies three subclasses in AD

The flowchart systematically describes our study (Fig. 1). Based on the previously reported 2,752 metabolism-related genes, consensus clustering classified 743 AD samples’ gene expression profiles after removing batch effect (Fig. 2A, B) into distinct subclasses. We categorized them into two to five subclasses (refer to Fig. S1). After comprehensive consideration, k = 3 was determined to be the best number of clusters. When k = 3, the consensus matrix heatmap retains clear and distinct boundaries (Fig. 2D).

3.2 Clinical characteristics of AD subclasses

The gamma-secretase activity in MCA was notably higher compared to MCB and MCC (P < 0.001 or P < 0.05; Fig. 3A). Compared with MCB, the beta-secretase activity, NFT and braak were elevated in MCA and MCC (P < 0.01; Fig. 3B, D, E). The PH in MCC were higher than MCA and MCB (P < 0.05; Fig. 3F). The three subclasses have a greater proportion of women than proportion of men (Fig. 3H). Furthermore, the percentages of both one and two APOE 4 alleles were significantly higher in MCA than in MCB and MCC (Fig. 3I). As for the age and the alpha-secretase activity, there was no difference between the three subclasses (Fig. 3C, G). The tissue origin of MCA, MCB and MCC was shown in Fig. S2.

3.3 Association between AD subclasses and metabolism-associated signatures

Given that the AD subclasses were established on the basis of metabolism genes, we conducted further investigations to determine whether the different subclasses had varying metabolic signatures. Initially, 84 metabolism processes were measured utilizing the “GSVA” R package. Next, we performed differential analysis to identify subclass-specific metabolic signatures, which were identified as signatures with a greater GSVA score in the relevant subclasses. Results displayed that only MCA and MCB exhibited distinct metabolism signatures, and the numbers were 40 and 30, respectively, whereas MCC had little distinct metabolism signatures. Notably, 7 of the 40 distinct metabolism signatures in MCA were associated with carbohydrate metabolism (Fig. 4).

3.4 Association between AD subclasses and immune infiltration

To further investigate the characteristics of subclasses, ESTIMATE algorithm was applied to compute the immune and stromal score. The immune scores were noted to display a marked variance across the three groups, whereby MCA demonstrated a higher immune score compared to MCB and MCC. (P < 0.0001; Fig. 5B). Furthermore, MCA exhibited higher stromal score than MCB and MCC (P < 0.0001; Fi.5B). Given the observed variance in immune scores among subclasses, an investigation into immune infiltration was conducted with the aim of characterizing their immunological landscape. We quantified the infiltration levels of 24 microenvironment cell types and surveyed the samples for the expression of immune checkpoints (Fig. 5A). We observed significantly higher expression of several immune checkpoints in MCB, which could potentially serve as targets for immunotherapy, including CD274 (PDL1), PDCD1 (PDL2; Fig. 5C). In addition, significant difference was observed between MCA and other two subclasses, with higher abundance of 18 immune cell populations (regulatory T cells, CD4 + T cells, nature B cells, memory B cells, activated Dendritic cells, M1 Macrophages, activated NK cells, memory CD4 + T cells, activated Mast cells, resting NK cells, M0 macrophages, M2 macrophages, Eosinophils, resting Dendritic cells, resting Mast cells, Neutrophils, Endothelial cells and Fibroblasts) for MCA compared with MCB or MCC(Fig. 5D). Of note, MCA showed high infiltration with endothelial cells and fibroblasts (Fig. 5D). Therefore, we quantified various types of CAFs and observed that MCA exhibited an enrichment of all distinct subtypes of fibroblasts. Furthermore, MCA exhibited a depletion of normal fibroblasts. (Fig. 5E).

3.5 Identification of AD feature genes and functional enrichment analysis

We conducted the WGCNA using a merged dataset to identify primary modules associated with the poor progression of AD. When the soft threshold was set to 4, the scale-free network and connectivity exhibited maximum efficiency. (Fig. 6A). Using a hierarchical clustering algorithm, the clustering tree was classified into six gene modules that were each assigned a unique color (Fig. 6B). Among the given modules, the cyan module consisting of 3284 genes demonstrated the strongest positive correlation with MCA (R = 0.49) and a series of AD-related high-risk indicators, such as NFT (R = 0.51), braak (R = 0.32), gamma-secretase activity (R = 0.3), amyloid-beta 42 (R = 0.28) and alpha-secretase activity (R = 0.25) (Fig. 6C). Therefore, the cyan module was identified as the primary module, from which hub genes were extracted using the selection criteria of cor.MM > 0.7 and cor.GS > 0.4 (Fig. 6D). In addition, we performed GO and KEGG enrichment analysis according to the hub genes aforementioned (Fig. 6E, F). KEGG enrichment analysis suggested that various synapse including GABAergic synapse, Glutamatergic synapse, and Dopaminergic synapse, and synaptic transmission-related signaling pathways including the calcium signaling pathway, adrenergic signaling and synaptic vesicle cycle signaling pathway were closely associated with these hub genes (Fig. 6E, Table. S3). GO enrichment analysis showed that these hub genes were predominantly enriched in the regulation of cell morphogenesis, actin filament organization, actin filament bundle assembly, actin filament bundle organization and cell − matrix adhesion (Fig. 6F, Table. S4). These results displayed the critical roles of genes.

3.6 Selection of AD feature genes

We conducted three different machine learning algorithms to screen the potential biomarkers of AD. Through the LASSO regression algorithm, above hub genes were narrowed down to 51 variables (Fig. 7A, B). Through the SVM-REF algorithm, we identified a subset of 86 features among the hub genes (Fig. 7C, D). Through the RF algorithm, we identified the top 20 features (Fig. 7E, F). The overlapping genes among LASSO, RF and SVM-REF algorithm (GFAP, CYB5R3, PMP2, DARS, KIAA0513, ITGB8, ENAH, EZR, RIN2, KCNC1, FOXO1, COLEC12, TST, AKR1C3, TSPO and ANTXR2) were ultimately selected for further study (Fig. 7G). Finally, we used logistic regression to filter out the 8 feature genes (GFAP, CYB5R3, DARS, KIAA0513, EZR, KCNC1, COLEC12 and TST, p < 0.05) in above 16 overlapping genes.

3.7 Verification of feature genes

Develop a nomogram model for AD diagnosis utilizing the eight feature genes (GFAP, CYB5R3, DARS, KIAA0513, EZR, KCNC1, COLEC12 and TST)(Fig. 8A). Next, calibration curve was utilized to assess the predictive capabilities of the nomogram model in the training dataset and testing dataset. The calibration curve showed that a small error between the actual risk and the predicted risk of AD, implying high accuracy of the nomogram model in predicting AD༈Fig. 8B༉. DCA showed that the “nomogram” curve was higher than the line representing the “intervention for none” curve, “intervention for all” curve and all single gene cures, suggested that the patients could benefit from the nomogram model at high risk threshold from 0 to 1 and the clinical benefit of the nomogram model was higher than the single gene curve༈Fig. 8C༉. Subsequently, the ROC curve analysis was employed to evaluate the diagnostic capability of each feature gene in predicting the progression of AD in the internal datasets. The AUC values in the training dataset were 0.788 for nomogram model, 0.729 for GFAP, 0.692 for EZR, 0.656 for COLEC12, 0.652 for KIAA0513, 0.698 for CYB5R3, 0.560 for DARS, 0.558 for KCNC1 and 0.557 for TST (Fig. 8D). The AUC values of ROC curves in the testing set were 0.770 for nomogram model, 0.708 for GFAP, 0.698 for EZR, 0.677 for COLEC12, 0.692 for KIAA0513, 0.575 for CYB5R3, 0.566 for DARS, 0.566 for KCNC1 and 0.584 for TST (Fig. 8D). In addition, eight single validation datasets: GSE5281, GSE48350, GSE118553, GSE28146, GSE122063, GSE132903, GSE8442201, GSE28146 and GSE1297, were used to further validate the diagnostic efficacy of these eight feature genes (Fig. 8E-L). To some extent, these results also suggest that the eight genes may have a significant role in the pathogenesis of AD.

3.8 Validation of feature genes expression

We utilized the GEO database to elucidate gene expression in AD patients. AD patients exhibited a significant reduction in the expression levels of KIAA0513 and KCNC1 expression compared to normal controls, while that of other genes was markedly increased (Fig. 9B). We further verified the expression of these eight feature genes using RT-PCR analysis. Consistent with the GEO dataset results from brain tissue samples, the expression levels of GFAP, CYB5R3, DARS, EZR, COLEC12, and TST were significantly higher in AD mice, whereas KIAA0513 exhibited significant downregulation (Fig. 9A). In contrast, KCNC1 expression did not differ between AD and normal groups.

AD is a neurodegenerative disease in which Aβ and NFT aggregation leads to loss of synapses, neuronal death and subsequent memory impairment. There is a large heterogeneity of pathogenesis among patients with AD, and biomarkers of AD progression still need to be refined[26, 27]. Accordingly, suitable AD subtypes and more powerful biomarkers is crucial for the precision diagnosis and therapy in AD.

Accumulating evidence suggested that the occurrence and advancement of AD is closely related to substance and energy metabolism, and the metabolism of glucose, lipids and energy has an important impact on AD[28–30]. The energy of brain mainly depends on glucose metabolism, which is metabolized to ATP via glycolysis, the tricarboxylic acid (TCA) cycle, and the electron transport chain[31]. The glucose metabolism is remarkably damaged in AD brain. The attenuated ATP production due to inefficient glucose utilization can accompany with the signal transduction breakdown, ionic pump dysfunction, neurotransmission failure, ultimately leading to neuronal degeneration and death[28]. Lipids are also involved in the pathology of AD[32]. As the strongest genetic risk factor for AD, the apolipoprotein E ε4 (APOE4) drives metabolic dysregulation in astrocytes and microglia, leading to cholesterol accumulation, decreased neuronal excitability and neuroinflammation[33, 34]. Restoring metabolic homeostasis can exert prominent neuroprotective effect[35]. Despite accumulating studies have validated the pathological mechanisms of metabolism in AD, the exact biological functions of metabolism in AD have not been completely illustrated, especially the biomarkers associated with metabolism and the function of metabolism in the regulation of AD immunity.

In this study, to identify AD subclass associated with metabolic processes, AD classification was built based on the metabolic genes from previous publications. Three distinct AD subclasses (MCA, MCB, and MCC) were identified. We explored the clinical features, metabolic signatures and immune infiltration of each AD subclass. Results demonstrated that MCA exhibited specific metabolic signatures and accompanied with high AD progression signatures (β-secretase activity, γ-secretase activity, NFT, braak and AD-risk gene APOE4).

MCA was chiefly involved in gene signatures of carbohydrate metabolism and lipid metabolism. The carbohydrate metabolism of MCA mainly involves glycolysis, fructose, mannose and galactose metabolism, whereas citrate cycle and pyruvate metabolism were decreased compared to the other two subgroups, indicating the TCA cycle impaired and loss of glucose utilization(thereby reducing ATP production). Meanwhile, the lipid metabolism of MCA mainly involves fatty acid degradation, which is likely to be a result of low ATP production prompting a shift in energy metabolism to ketogenic pathway. These metabolic disorders affect the energy supply of neurons in the brain. Furthermore, previous studies have also confirmed that the mitochondrial ATP-synthase α subunit is lipoxidized distinctly and the ATP-synthase activity was obviously reduced in the entorhinal cortex in AD patients compared to controls[36]. Analysis of clinical features and metabolic signatures showed that high APOE4-carrying, NFT accumulation and significant metabolic disorders were observed in the MCA subgroup, thus presenting a poorer prognosis༎Immune infiltration analysis suggested that MCA showed an augmented immune score and relatively higher immune cell infiltration levels. A significant change in immune cells ratio was found in AD subclasses, MCA displayed higher infiltration levels of regulatory T cells (Tregs), CD4 + T cells, memory CD4 + T cells, B cells, activated Dendritic cells, Macrophages and Neutrophils, etc, which were consistent with the previous studies[37–39]. Meanwhile, MCA also showed stromal score and high infiltration with endothelial cells and fibroblasts. Immune checkpoints that represent potential targets for immunotherapy, such as CD274 (PDL1), PDCD1 (PDL2), were mainly increased in the MCB.

Recent studies have displayed that defects in glucose and lipid metabolism occur early and before significant cognitive decline[40], suggesting that metabolism-related molecular markers may contribute greatly to the diagnosis and treatment of AD. We used the combined dataset of AD patients to construct co-expression networks via WGCNA. The cyan module was found to be positively correlated with MCA and “A/T/N” system such as NFT, further supporting our above speculation that the MCA subgroup maybe a high-risk subgroup for AD. Functional enrichment analysis revealed that the hub genes in cyan module were primarily enriched in cellular morphological regulation and synapse-related functions and pathways. The impaired TCA cycle in the MCA above mentioned is the main function of the mitochondria in the cell. Metabolic disorders may lead to mitochondrial dysfunction, inadequate energy supply, and massive reactive oxygen species release, inducing oxidative stress and calcium regulation imbalance, ultimately triggering neuronal apoptosis and synaptic loss[8].

Recently, various machine learning algorithms has been extensively employed to predict new biomarkers and offer new insights for disease pathogenesis, owing to its outstanding performance in diagnosis[41, 42]. Therefore, we used three machine algorithms to further narrow down the range of hub genes. Eight hub genes were finally identified, including GFAP, CYB5R3, DARS, KIAA0513, EZR, KCNC1, COLEC12 and TST. GFAP is a marker of astrogliosis. Recently, Shen et al. showed that plasma GFAP is significantly elevated from the preclinical stage of AD and is a promising diagnostic and predictive biomarker to distinguish AD from controls and non-AD dementia[43]. The CYB5R3 gene encodes cytochrome b5 reductase 3, which is essential for reductive reactions such as cholesterol biosynthesis, fatty acid elongation, methemoglobin reduction and drug metabolism[44]. CYB5R3 expression was found to be elevated in human cortex in an AD proteomics study[45]. As the aspartyl-tRNA synthetase, DARS missense mutations result in a significant pattern of hypomyelination, motor abnormalities, and cognitive impairment[46]. Bioinformatics analysis suggested that the reduction of KIAA0513 serve as a potential biomarker for early diagnosis of AD[47]. EZR, which is a member of the ERM (ezrin-radixin-moesin) protein family, has been recognized as a regulator of adhesion signal pathways. EZR plays a key role in promoting the invasion and metastasis of malignant tumors[48]. KCNC1 encodes a subunit of the Kv3 voltage-gated potassium channels and associated with a variety of human diseases, including ataxia, epilepsy and developmental delay[49]. COLEC12 encodes a member of the C-lectin family, which is a scavenger receptor that plays a crucial role in the binding and clearance of Aβ[50]. TST is an enzyme that is widely distributed in both prokaryotes and eukaryotes, playing a crucial role in mitochondrial function [51]. These findings and our findings are concordant in indicating that overexpression of GFAP, CYB5R3, DARS, EZR, COLEC12, and TST, and decreased KIAA0513, and KCNC1 may predict poor prognosis in AD patients. In addition, the nomogram model, calibration curves, DCA and ROC curves verified the satisfactory diagnostic ability of these eight signature genes.

In this study, we classified ADs from the perspective of metabolism for the first time. The screening and validation of feature genes provide potential molecular targets for further exploring the metabolic mechanism of AD. Whereas, there are still some limitations in our research. Firstly, the feature genes were only validated in AD mice, and clinical data support was lacking. Secondly, KCNC1 showed inconsistent results in the GEO database and AD mice, possibly due to the small sample size. Finally, the mechanism of metabolism regulation in AD still needs to be further investigated in vivo and in vitro, which will be the focus of our subsequent studies.

In conclusion, we determined the strong relationship between metabolic status and AD pathogenesis through comprehensive bioinformatics analysis. We divided ADs into three subclasses from the perspective of metabolism. We identified and verified eight feature genes, including GFAP, CYB5R3, DARS, EZR, COLEC12, and TST, which showed high expression, while KIAA0513 and KCNC1 displayed low expression. The diagnostic model built by eight genes displayed outstanding diagnostic value. This offers a study basis for accurate and early diagnosis of AD.

AD Alzheimer's disease

Aβ Amyloid-β

NFT Neurofibrillary tangles

GSVA Gene set variation analysis

SVM Support Vector Machines

RF Random Forest

WGCNA Weighted correlation network analysis

GO Gene Ontology

KEGG Kyoto Encyclopedia of Genes and Genomes

AUC Area under the curve

PCA Principal component analysis

DCA Decision curve analysis

TCA Tricarboxylic acid

APOE4 Apolipoprotein E ε4

Ethics approval and consent to participate

All animal experiments were reviewed and approved by the Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology (ACUC Number:3121).

Consent for publication

Not applicable.

Availability of data and materials

The datasets analysed during the current study are available in the GEO database (https:// www. ncbi. nlm. nih. gov/ geo/), openly available for free download.

Competing interests

The authors declare that they have no competing interests.

Funding

This study was supported by grant from the National Natural Science Foundation of China (NSFC Project, No. 81873734 and 81974200).

Author contributions

PL and XC were contributed to the conception and design of the study, processing of bioinformatics data, experimental validation and writing of the manuscript. CW was responsible for animal modeling. KL was responsible for the statistical work towards experimental data. YX and XC supervised the whole analysis and provided guidance and instructions. All authors contributed to the manuscript revision and approved the submitted version.

Acknowledgements

We thank all contributors to the GEO database.

Authors' information

Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China

Estimation of the global prevalence of dementia. in 2019 and forecasted prevalence in 2050: an analysis for the Global Burden of Disease Study 2019. Lancet Public Health. 2022;7:e105–25.
DeTure MA, Dickson DW. The neuropathological diagnosis of Alzheimer's disease. Mol Neurodegener. 2019;14:32.
Yang Y, Arseni D, Zhang W, Huang M, Lövestam S, Schweighauser M, Kotecha A, Murzin AG, Peak-Chew SY, Macdonald J, et al. Cryo-EM structures of amyloid-β 42 filaments from human brains. Science. 2022;375:167–72.
Hodson R. Alzheimer's disease. Nature. 2018;559:1.
Poddar MK, Banerjee S, Chakraborty A, Dutta D. Metabolic disorder in Alzheimer's disease. Metab Brain Dis. 2021;36:781–813.
Kuehn BM. In Alzheimer Research, Glucose Metabolism Moves to Center Stage. JAMA. 2020;323:297–9.
Yu L, Jin J, Xu Y, Zhu X. Aberrant Energy Metabolism in Alzheimer's Disease. J Transl Int Med. 2022;10:197–206.
Peng Y, Gao P, Shi L, Chen L, Liu J, Long J. Central and Peripheral Metabolic Defects Contribute to the Pathogenesis of Alzheimer's Disease: Targeting Mitochondria for Diagnosis and Prevention. Antioxid Redox Signal. 2020;32:1188–236.
Varma VR, Oommen AM, Varma S, Casanova R, An Y, Andrews RM, O'Brien R, Pletnikova O, Troncoso JC, Toledo J, et al. Brain and blood metabolite signatures of pathology and progression in Alzheimer disease: A targeted metabolomics study. PLoS Med. 2018;15:e1002482.
Barrett T, Wilhite SE, Ledoux P, Evangelista C, Kim IF, Tomashevsky M, Marshall KA, Phillippy KH, Sherman PM, Holko M, et al. NCBI GEO: archive for functional genomics data sets–update. Nucleic Acids Res. 2013;41:D991–995.
Wilkerson MD, Hayes DN. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinformatics. 2010;26:1572–3.
Possemato R, Marks KM, Shaul YD, Pacold ME, Kim D, Birsoy K, Sethumadhavan S, Woo HK, Jang HG, Jha AK, et al. Functional genomics reveal that the serine synthesis pathway is essential in breast cancer. Nature. 2011;476:346–50.
Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013;14:7.
Racle J, de Jonge K, Baumgaertner P, Speiser DE, Gfeller D. Simultaneous enumeration of cancer and immune cell types from bulk tumor gene expression data. Elife 2017, 6.
Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES, Mesirov JP. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102:15545–50.
Finotello F, Mayer C, Plattner C, Laschober G, Rieder D, Hackl H, Krogsdam A, Loncova Z, Posch W, Wilflingseder D, et al. Molecular and pharmacological modulators of the tumor immune contexture revealed by deconvolution of RNA-seq data. Genome Med. 2019;11:34.
Li B, Liu JS, Liu XS. Revisit linear regression-based deconvolution methods for tumor gene expression data. Genome Biol. 2017;18:127.
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;12:453–7.
Becht E, Giraldo NA, Lacroix L, Buttard B, Elarouci N, Petitprez F, Selves J, Laurent-Puig P, Sautès-Fridman C, Fridman WH, de Reyniès A. Estimating the population abundance of tissue-infiltrating immune and stromal cell populations using gene expression. Genome Biol. 2016;17:218.
Yoshihara K, Shahmoradgoli M, Martínez E, Vegesna R, Kim H, Torres-Garcia W, Treviño V, Shen H, Laird PW, Levine DA, et al. Inferring tumour purity and stromal and immune cell admixture from expression data. Nat Commun. 2013;4:2612.
Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559.
Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics. 2012;16:284–7.
Bellocco R, Pagano M. Poisson regression. Nutrition. 1998;14:63–4.
Uddin S, Khan A, Hossain ME, Moni MA. Comparing different supervised machine learning algorithms for disease prediction. BMC Med Inform Decis Mak. 2019;19:281.
Lai Y, Lin X, Lin C, Lin X, Chen Z, Zhang L. Identification of endoplasmic reticulum stress-associated genes and subtypes for prediction of Alzheimer's disease based on interpretable machine learning. Front Pharmacol. 2022;13:975774.
Duara R, Barker W. Heterogeneity in Alzheimer's Disease Diagnosis and Progression Rates: Implications for Therapeutic Trials. Neurotherapeutics. 2022;19:8–25.
Cano A, Turowski P, Ettcheto M, Duskey JT, Tosi G, Sánchez-López E, García ML, Camins A, Souto EB, Ruiz A, et al. Nanomedicine-based technologies and novel biomarkers for the diagnosis and treatment of Alzheimer's disease: from current to future challenges. J Nanobiotechnol. 2021;19:122.
Butterfield DA, Halliwell B. Oxidative stress, dysfunctional glucose metabolism and Alzheimer disease. Nat Rev Neurosci. 2019;20:148–60.
Arnold M, Nho K, Kueider-Paisley A, Massaro T, Huynh K, Brauner B, MahmoudianDehkordi S, Louie G, Moseley MA, Thompson JW, et al. Sex and APOE ε4 genotype modify the Alzheimer's disease serum metabolome. Nat Commun. 2020;11:1148.
Zhang X, Tong T, Chang A, Ang TFA, Tao Q, Auerbach S, Devine S, Qiu WQ, Mez J, Massaro J, et al. Midlife lipid and glucose levels are associated with Alzheimer's disease. Alzheimers Dement. 2023;19:181–93.
Cunnane SC, Trushina E, Morland C, Prigione A, Casadesus G, Andrews ZB, Beal MF, Bergersen LH, Brinton RD, de la Monte S, et al. Brain energy rescue: an emerging therapeutic concept for neurodegenerative disorders of ageing. Nat Rev Drug Discov. 2020;19:609–33.
Markesbery WR, Kryscio RJ, Lovell MA, Morrow JD. Lipid peroxidation is an early event in the brain in amnestic mild cognitive impairment. Ann Neurol. 2005;58:730–5.
Tcw J, Qian L, Pipalia NH, Chao MJ, Liang SA, Shi Y, Jain BR, Bertelsen SE, Kapoor M, Marcora E, et al. Cholesterol and matrisome pathways dysregulated in astrocytes and microglia. Cell. 2022;185:2213–2233e2225.
Victor MB, Leary N, Luna X, Meharena HS, Scannail AN, Bozzelli PL, Samaan G, Murdock MH, von Maydell D, Effenberger AH, et al. Lipid accumulation induced by APOE4 impairs microglial surveillance of neuronal-network activity. Cell Stem Cell. 2022;29:1197–1212e1198.
Zheng J, Xie Y, Ren L, Qi L, Wu L, Pan X, Zhou J, Chen Z, Liu L. GLP-1 improves the supportive ability of astrocytes to neurons by promoting aerobic glycolysis in Alzheimer's disease. Mol Metab. 2021;47:101180.
Terni B, Boada J, Portero-Otin M, Pamplona R, Ferrer I. Mitochondrial ATP-synthase in the entorhinal cortex is a target of oxidative stress at stages I/II of Alzheimer's disease pathology. Brain Pathol. 2010;20:222–33.
Saresella M, Calabrese E, Marventano I, Piancone F, Gatti A, Alberoni M, Nemni R, Clerici M. Increased activity of Th-17 and Th-9 lymphocytes and a skewing of the post-thymic differentiation pathway are seen in Alzheimer's disease. Brain Behav Immun. 2011;25:539–47.
Song L, Yang YT, Guo Q, Zhao XM. Cellular transcriptional alterations of peripheral blood in Alzheimer's disease. BMC Med. 2022;20:266.
Kim K, Wang X, Ragonnaud E, Bodogai M, Illouz T, DeLuca M, McDevitt RA, Gusev F, Okun E, Rogaev E, Biragyn A. Therapeutic B-cell depletion reverses progression of Alzheimer's disease. Nat Commun. 2021;12:2185.
Wang J, Gu BJ, Masters CL, Wang YJ. A systemic view of Alzheimer disease - insights from amyloid-β metabolism beyond the brain. Nat Rev Neurol. 2017;13:703.
Lai Y, Lin P, Lin F, Chen M, Lin C, Lin X, Wu L, Zheng M, Chen J. Identification of immune microenvironment subtypes and signature genes for Alzheimer's disease diagnosis and risk prediction based on explainable machine learning. Front Immunol. 2022;13:1046410.
Li J, Zhang Y, Lu T, Liang R, Wu Z, Liu M, Qin L, Chen H, Yan X, Deng S, et al. Identification of diagnostic genes for both Alzheimer's disease and Metabolic syndrome by the machine learning algorithm. Front Immunol. 2022;13:1037318.
Shen XN, Huang SY, Cui M, Zhao QH, Guo Y, Huang YY, Zhang W, Ma YH, Chen SD, Zhang YR, et al. Plasma Glial Fibrillary Acidic Protein in the Alzheimer Disease Continuum: Relationship to Other Biomarkers, Differential Diagnosis, and Prediction of Clinical Progression. Clin Chem. 2023;69:411–21.
Rahaman MM, Reinders FG, Koes D, Nguyen AT, Mutchler SM, Sparacino-Watkins C, Alvarez RA, Miller MP, Cheng D, Chen BB, et al. Structure Guided Chemical Modifications of Propylthiouracil Reveal Novel Small Molecule Inhibitors of Cytochrome b5 Reductase 3 That Increase Nitric Oxide Bioavailability. J Biol Chem. 2015;290:16861–72.
Wang H, Dey KK, Chen PC, Li Y, Niu M, Cho JH, Wang X, Bai B, Jiao Y, Chepyala SR, et al. Integrated analysis of ultra-deep proteomes in cortex, cerebrospinal fluid and serum reveals a mitochondrial signature in Alzheimer's disease. Mol Neurodegener. 2020;15:43.
Fröhlich D, Suchowerska AK, Voss C, He R, Wolvetang E, von Jonquieres G, Simons C, Fath T, Housley GD, Klugmann M. Expression Pattern of the Aspartyl-tRNA Synthetase DARS in the Human Brain. Front Mol Neurosci. 2018;11:81.
Zhu M, Jia L, Li F, Jia J. Identification of KIAA0513 and Other Hub Genes Associated With Alzheimer Disease Using Weighted Gene Coexpression Network Analysis. Front Genet. 2020;11:981.
Xu J, Zhang W. EZR promotes pancreatic cancer proliferation and metastasis by activating FAK/AKT signaling pathway. Cancer Cell Int. 2021;21:521.
Li X, Zheng Y, Li S, Nair U, Sun C, Zhao C, Lu J, Zhang VW, Maljevic S, Petrou S, Lin J. Kv3.1 channelopathy: a novel loss-of-function variant and the mechanistic basis of its clinical phenotypes. Ann Transl Med. 2021;9:1397.
Nakamura K, Ohya W, Funakoshi H, Sakaguchi G, Kato A, Takeda M, Kudo T, Nakamura T. Possible role of scavenger receptor SRCL in the clearance of amyloid-beta in Alzheimer's disease. J Neurosci Res. 2006;84:874–90.
Buonvino S, Arciero I, Melino S. Thiosulfate-Cyanide Sulfurtransferase a Mitochondrial Essential Enzyme: From Cell Metabolism to the Biotechnological Applications. Int J Mol Sci 2022, 23.

Fig.S1.tif
Additional file 1: Figure S1. Consensus clustering matrix for k = 2–5.
Fig.S2.tif
Additional file 2: Figure S2. Box-plot of tissue original of AD subclasses.
TableS14.xlsx
Additional file 3: Table S1. GEO datasets information. Table S2.Primer sequences of feature genes. Table S3. KEGG pathway enrichment analyses of hub genes in cyan module. Table S4. GO enrichment analyses of hub genes in cyan module.

Download PDF

Journal Publication

published 15 Sep, 2023

Read the published version in Journal of Translational Medicine →

Reviewers agreed at journal
24 May, 2023
Reviewers invited by journal
24 May, 2023
Editor assigned by journal
20 May, 2023
First submitted to journal
18 May, 2023

You are reading this latest preprint version

Identification and Validation of Metabolism-Related Genes in Alzheimer’s Disease

Status:

Journal Publication

Version 1

Abstract

Figures

1 Background

2 Methods

2.1 Data collection and processing

2.2 Identification of AD subclasses

2.3 Gene set variation analysis (GSVA)

2.4 Evaluation of Immune Infiltration

2.5 Weighted correlation network analysis (WGCNA)

2.6 Functional enrichment analysis

2.7 Machine learning

2.8 Establishment and assessment of a nomogram

2.9 Assessment of the diagnostic significance of feature genes in AD

2.10 Animals

2.11 Real-Time Quantitative PCR

2.12 Statistical analysis

3 Results

3.1 Consensus clustering identifies three subclasses in AD

3.2 Clinical characteristics of AD subclasses

3.3 Association between AD subclasses and metabolism-associated signatures

3.4 Association between AD subclasses and immune infiltration

3.5 Identification of AD feature genes and functional enrichment analysis

3.6 Selection of AD feature genes

3.7 Verification of feature genes

3.8 Validation of feature genes expression

4 Discussion

5 Conclusion

Abbreviations

Declarations

References

Supplementary Files

Status:

Journal Publication

Version 1