Key genes and pathways in asparagine metabolism in Alzheimer’s Disease: a bioinformatics approach

1 1 Key genes and pathways in asparagine metabolism in 2 Alzheimer’s Disease: a bioinformatics approach 3 Xiaoqian Lan 1, Guangli Feng2, Qing Li3, Yuting Shi2, Shiyi Qin1, Lianmei Zhong2* 4 1. Department of Neurology, The First Affiliated Hospital of Kunming Medical University, 5 Kunming, China 6 2. Department of Neurology, Xuanwu Hospital, Capital Medical University, Beijing, China 7 3. Yunnan Key Laboratory of Stem Cell and Regenerative Medicine, School of Rehabilitation, 8 Kunming Medical University, Kunming, China 9 * Corresponding author 10 E-mail: [email protected] (LZ) 11 Abstract 12 Background: Asparagine (Asn) metabolism is essential for maintaining cellular homeostasis and 13 supporting neuronal energy demands. Recent studies have suggested its dysregulation may contribute 14 to Alzheimer’s disease (AD) pathogenesis; however, the specific genes and regulatory mechanisms 15 involved remain incompletely understood. 16 Methods: Four publicly available microarray datasets (GSE5281, GSE29378, GSE36980, and 17 GSE138260) were utilized to investigate genes with differential expression between control and AD .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 2 18 samples. Asparagine metabolism-related genes (AMGs) were retrieved from the GeneCards database, 19 and their intersection with DEGs yielded candidate asparagine metabolism-related differentially 20 expressed genes (AMG-DEGs). Functional enrichment analysis (Gene Set Enrichment Analysis, Gene 21 Ontology and Kyoto Encyclopedia of Genes and Genomes), protein–protein interaction (PPI) network 22 analysis, and centrality scoring identified hub genes. Regulatory mechanisms were investigated 23 through construction of competing endogenous RNA and transcription factor networks. Potential 24 therapeutic compounds were predicted via drug–gene enrichment and evaluated using molecular 25 docking simulations. 26 Results: Thirty-nine AMG-DEGs were identified and found to be enriched in neurodevelopmental, 27 synaptic transmission, and inflammatory signaling pathways. PPI analysis and centrality screening 28 revealed seven hub genes (HPRT1, GAD2, TUBB3, GFAP, CD44, CCL2, and NFKBIA). Regulatory 29 network analysis highlighted specific miRNAs, long non-coding RNAs, and transcription factors 30 involved in their modulation. Drug screening and docking identified Bathocuproine disulfonate, DL- 31 Mevalonic acid, and Phenethyl isothiocyanate as promising compounds with strong binding affinities 32 to hub proteins. 33 Conclusion: This study comprehensively maps the dysregulation of asparagine metabolism in 34 Alzheimer’s disease and reveals a set of hub genes and regulatory elements potentially involved in 35 disease progression. The predicted therapeutic compounds provide a foundation for further 36 experimental validation and may contribute to the development of novel metabolism-targeted strategies 37 for AD treatment. 38 Keywords: Alzheimer’s disease; asparagine metabolism; drug screening; molecular docking; 39 Bioinformatics .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 3 40 1. Introduction 41 Alzheimer’s disease (AD) is a chronic neurodegenerative condition marked by progressive 42 cognitive impairment, memory deterioration, behavioral abnormalities, and eventual loss of 43 autonomy(1, 2). Pathologically, AD is characterized by the deposition of extracellular β-amyloid (Aβ) 44 plaques, the formation of intracellular neurofibrillary tangles consisting of hyperphosphorylated tau, 45 and extensive synaptic dysfunction(3). Due to increasing global longevity, AD has become a leading 46 cause of morbidity among older adults, with over 50 million cases reported worldwide—a number 47 projected to triple by 2050(4, 5). This growing burden underscores the imperative for novel therapeutic 48 strategies and effective preventive measures. 49 Emerging evidence suggests that metabolic disturbances are key contributors to AD pathogenesis 50 beyond classical amyloid and tau pathology(6). Patients with AD often exhibit systemic dysregulation 51 of glucose utilization, lipid processing, and amino acid metabolism(7), which may trigger 52 neuroinflammation, synaptic breakdown, and vascular impairment(8). These metabolic disruptions can 53 modulate transcriptional programs through regulators such as NF-κB and Nrf2(9), and may also alter 54 non-coding RNA networks, further exacerbating neuronal dysfunction and disease progression(10). 55 Among multiple metabolic abnormalities, altered amino acid metabolism—especially involving 56 asparagine (Asn)—has drawn increasing research attention. Asn, synthesized by asparagine synthetase 57 and catabolized by asparaginase, is implicated in cellular immunity(11) and diverse neuronal functions, 58 including bioenergetics, neurotransmitter regulation, and protein post-translational modification(12). 59 Observations from both sporadic and autosomal-dominant AD brain tissues demonstrate disrupted Asn 60 homeostasis, suggesting a common metabolic signature associated with disease progression(13). 61 Clinical studies have shown that plasma Asn concentrations are markedly decreased in AD patients 62 and positively correlate with neurodegeneration-associated markers such as neurofilament light chain, .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 4 63 particularly among individuals with elevated cerebral Aβ load(14). Mechanistically, Asn can be 64 converted to aspartate via transamination, thereby entering the tricarboxylic acid (TCA) cycle and 65 contributing to cellular energy metabolism(15). Metabolomics studies have corroborated these 66 findings, showing significantly decreased Asn levels in the plasma of AD patients, which may 67 contribute to an energy-deficient and neurotoxic microenvironment(16). Asn is also essential for 68 protein N-glycosylation, a key post-translational modification process. Depletion of Asn may disrupt 69 glycosylation events(17); notably, aberrant glycosylation at the Asn368 site of tau protein has been 70 associated with impaired synaptic signaling and reduced plasticity, potentially intensifying AD-related 71 neuropathology(18). Although asparagine endopeptidase (AEP) does not participate directly in Asn 72 metabolism, many of its substrate proteins include Asn residues, suggesting its activity may be 73 modulated by Asn availability(19). AEP has been implicated in the proteolytic cleavage of both 74 amyloid precursor protein and tau, leading to the generation of neurotoxic fragments(20, 21) and 75 influences microglial activation(21), and inflammatory cascades that impair neuronal integrity and 76 cognitive performance(22). Although existing studies support a link between altered Asn metabolism 77 and AD, the molecular mechanisms underlying this association remain poorly understood. In 78 particular, the regulatory interactions of asparagine metabolism-related genes (AMGs)—especially 79 those involved in neuroinflammation and synaptic integrity—and their potential as therapeutic targets 80 require in-depth, systematic investigation. 81 This study employed bioinformatics approaches to systematically analyze the characteristics of 82 Asn metabolic dysregulation in AD. By integrating multiple public databases, we successfully 83 identified asparagine metabolism-related differentially expressed genes (AMG-DEGs). Furthermore, 84 we elucidated how these genes participate in the molecular mechanisms underlying AD pathogenesis 85 by regulating key biological processes, including neuroinflammation, synaptic dysfunction, and energy 86 metabolism. Through functional enrichment analyses, such as Gene Set Enrichment Analysis (GSEA), .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 5 87 Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG), we systematically 88 characterized the biological functions of Asn metabolism-related genes. Additionally, by constructing 89 protein-protein interaction (PPI) and gene regulatory networks, we identified several hub genes that 90 play critical roles throughout the progression of AD. Notably, through drug screening and molecular 91 docking simulations, we identified multiple high-affinity small molecules that target key pathways 92 involved in Asn metabolism, providing significant theoretical insights and potential drug candidates 93 for the development of novel AD therapeutic strategies based on Asn metabolism modulation. 94 Fig 1. Research design flow chart. 95 2. Materials and methods 96 2.1 Datasets and preprocessing 97 Microarray datasets associated with AD, including GSE5281, GSE29378, GSE36980, and 98 GSE138260, ere accessed from the Gene Expression Omnibus (GEO) repository 99 (https://www.ncbi.nlm.nih.gov/geo/). Corresponding platform annotation files were obtained to 100 facilitate probe-to-gene symbol mapping. Clinical metadata including age, sex, and group assignment 101 were also extracted. A list of AMGs was retrieved from the GeneCards human gene database 102 (https://www.genecards.org/) by searching for the keyword 'asparagine metabolism'. Genes with a 103 relevance score > 6 and that were protein-coding were selected for analysis. A total of 1,294 AMGs 104 were compiled for subsequent analyses. Data preprocessing was performed on each dataset, including 105 background correction, log2 transformation, and quantile normalization of the raw expression values. 106 Gene-level analysis was performed by mapping probe IDs to gene symbols, with the average 107 expression value applied when multiple probes corresponded to a single gene. The data utilized in this 108 study were publicly accessible, and no experimental work was carried out by the authors. .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 6 109 2.2 Identification of DEGs 110 Expression data were processed in R version 4.4.1 using the limma and sva packages. Batch 111 effects were first assessed via principal component analysis (PCA), which allowed for the identification 112 of significant inter-dataset variability. Genes with an adjusted p-value 0.585 were considered significant. The DEGs from all datasets were intersected 114 to identify common AD-related genes. DEG visualization was performed using the pheatmap package 115 for heatmap generation and ggplot2 for volcano plot visualization. Venn diagram analyses were 116 conducted using the VennDiagram package to intersect DEGs with the list of AMGs, ultimately 117 identifying 39 candidate AMG-DEGs. 118 2.3 Functional enrichment analyses: GSEA, GO and KEGG. 119 GSEA was performed using the clusterProfiler package in R, based on the reference gene set 120 collections c5.go.Hs.symbols.gmt. The DEGs were ranked by their log2FC, and enrichment scores 121 were computed for each gene set using the GSEA function. A threshold of an adjusted p-value < 0.05 122 was considered statistically significant. 123 GO and KEGG pathway enrichment analyses were performed using the “clusterProfiler” R 124 package to explore cellular components (CC), molecular functions (MF), biological processes (BP), 125 and signaling pathways associated with the intersecting AMG-DEGs. GO and KEGG enrichment 126 analyses were conducted with the enrichGO and enrichKEGG functions, respectively, using the 127 parameters: p-value < 0.05 and adjusted p-value < 0.05 as the significance thresholds. 128 2.4 Construction of the PPI network .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 7 129 To explore the interactions among the 39 AMG-DEGs, a PPI network was generated using the 130 Retrieval of Interacting Genes (STRING) database (https://string-db.org/) for gene interaction 131 retrieval, with a confidence score cutoff > 0.4. Unrelated nodes were removed to improve network 132 clarity. The network was visualized using Cytoscape software (version 3.10.0), and hub modules were 133 identified using the MCODE plugin with the following parameters: K-core = 2, node score cutoff = 134 0.2, degree cutoff = 2 and maximum depth = 100. 135 2.5 Hub gene identification and functional interaction analysis 136 The cytoHubba plugin in Cytoscape was employed to identify core AMG-DEGs using the 137 Maximal Clique Centrality algorithm. Additionally, Gene Multiple Association Network Integration 138 Algorithm (GeneMANIA) (http://www.genemania.org/) was used to construct gene co-expression 139 networks, providing insight into potential functional relationships between hub genes. 140 2.6 Construction of the competing endogenous RNA (ceRNA) regulatory 141 network 142 MiRNA-mRNA interactions targeting hub genes were predicted using multiple databases, 143 including miRanda (http://www.microrna.org/), miRTarBase (http://mirtarbase.cuhk.edu.cn/), miRDB 144 (http://mirdb.org/), and TargetScan (http://www.targetscan.org/). Long non-coding RNAs (lncRNAs)– 145 miRNA interactions were obtained via SpongeScan (http://spongescan.rc.ufl.edu/). All predicted 146 interactions were visualized in Cytoscape to generate the final ceRNA network, integrating lncRNA– 147 miRNA and miRNA-mRNA regulatory axes. 148 2.7 Transcription factor regulatory network construction .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 8 149 Transcription factor (TF) regulating the identified hub genes were retrieved from the TRRUST 150 database (https://www.grnpedia.org/trrust/). The TF-mRNA regulatory network was then visualized 151 using Cytoscape to elucidate potential upstream regulatory mechanisms. 152 2.8 Drug enrichment analysis 153 Candidate drugs targeting the hub genes were identified from the Drug-Gene Interaction Database 154 (DGIdb, https://www.dgidb.org/). Gene–compound association enrichment was performed using Drug 155 Signatures Database (DSigDB) (https://maayanlab.cloud/DSigDB/), and results were visualized using 156 the enrichplot package in R. 157 2.9 Molecular docking 158 Molecular docking was employed to assess the binding affinity between predicted small-molecule 159 compounds and their target proteins. The top five candidate drugs—ranked by adjusted p-values from 160 drug enrichment analysis—included DL-Mevalonic acid (MVA), Bathocuproine disulfonate (BCS), 161 Phenethyl isothiocyanate (PEITC), MELAMINE, and CHLOROBENZENE. Corresponding protein 162 structures were downloaded from the RCSB Protein Data Bank (PDB, https://www.rcsb.org/) for CD44 163 (PDB ID: 1UUH), CCL2 (PDB ID: 4USP), and TUBB3 (PDB ID: 5IJ0). The protein structure of 164 NFKBIA was predicted using the AlphaFold Protein Structure Database 165 (https://www.alphafold.ebi.ac.uk/entry/P25963). 3D molecular docking simulations were performed 166 with CB-Dock2 (https://cadd.labshare.cn/cb-dock2/php/index.php), and Vina score (binding energy ≤ 167 −5.0 kcal/mol) were used to prioritize ligand–receptor interactions. 168 3. Results .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 9 169 3.1 Identification of AMG-DEGs in AD 170 To identify candidate genes associated with asparagine metabolism in AD, we analyzed four 171 publicly available microarray datasets (GSE5281, GSE29378, GSE36980, and GSE138260). Clinical 172 information including age, sex, and sample group distributions for each dataset is summarized in Table 173 1. PCA was performed to evaluate batch effects. Before correction, samples clustered separately by 174 dataset, indicating strong batch effects (Fig 2A). After batch effect adjustment, the samples showed 175 improved clustering consistency (Fig 2B), confirming effective normalization. Differential expression 176 analysis revealed a total of 363 DEGs between AD and control samples, with 156 genes significantly 177 upregulated and 207 downregulated (Fig 2C, S1 Fig). A Venn diagram analysis identified 39 genes 178 overlapping between DEGs and AMGs (Fig 2D). The expression patterns of these 39 overlapping 179 genes were visualized using a heatmap, which demonstrated distinct clustering between control and 180 AD groups (Fig 2E). The complete list of these AMG-DEGs is provided in S1 Table. 181 Table 1 Microarrays datasets clinical characteristics. Dataset GSE5281 GSE29378 GSE36980 GSE138260 Groups Control AD Control AD Control AD Control AD Numbe r 74 87 32 31 47 33 19 17 Age 73.30±18.3 5 79.53±6.8 4 81.65±6.7 7 76.64±8.9 6 78.09±9.3 5 91.70±5.8 3 64.31±17.1 5 79.82±9.3 4 Male 53 50 22 16 22 15 9 7 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 10 Female 21 37 10 15 25 18 9 10 182 183 Fig 2. Expression profile analysis of overlapping genes between DEGs and AMGs. 184 (A–B) PCA for batch effect assessment: (A) Before batch correction; (B) After batch correction. 185 (C) Volcano plot displaying the results of DEGs analysis in AD, with red and blue dots representing 186 upregulated and downregulated genes, respectively. (D) Venn diagram illustrating the intersection 187 between DEGs and AMGs. (E) Heatmap of AMG-DEGs, with red and blue representing high and low 188 expression levels, respectively. 189 3.2 Functional enrichment analyses 190 GSEA revealed that in control samples, energy production and neurotransmission-related 191 pathways were enriched, including ATP synthesis coupled electron transport, GABA ergic synapse, 192 inner mitochondrial membrane protein complex, mitochondrial protein containing complex and 193 synaptic vesicle membrane (Fig 3A, S2 Table). In contrast, AD samples exhibited enrichment in 194 pathways associated with positive regulation of vasculature development, regulation of epithelial cell 195 differentiation, regulation of vasculature development, collagen containing extracellular matrix and 196 growth factor binding (Fig 3B, S2 Table). 197 To further understand the biological roles of the 39 AMG-DEGs, we conducted GO and KEGG 198 pathway enrichment analyses. The GO analysis revealed that these genes were significantly enriched 199 in BP such as response to steroid hormones, regulation of neurogenesis, and nervous system 200 development. For CC, the AMG-DEGs were found to be associated with the neuronal cell body, 201 GABAergic synapses, and clathrin-coated vesicle membranes (Fig 3C-D, Table 2), emphasizing their .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 11 202 involvement in maintaining neuronal function and synaptic communication. In terms of MF, the 203 enriched terms included carbon-carbon lyase activity, carboxy-lyase activity, and pyridoxal phosphate 204 (PLP) binding, all of which are essential for energy and neurotransmitter regulation (Fig 3C-D). KEGG 205 pathway enrichment revealed that AMG-DEGs were significantly involved in pathways like 206 GABAergic synapses (hsa04727), alanine, aspartate, and glutamate metabolism (hsa00250), and IL- 207 17 signaling (hsa04657). Interestingly, pathways related to butanoate metabolism (hsa00650) and β- 208 alanine metabolism (hsa00410) were found to be downregulated, while inflammation-related pathways, 209 including IL-17 signaling and rheumatoid arthritis (hsa05323), were upregulated (Fig 3E-F). 210 Fig 3. Functional enrichment analyses. 211 (A-B) GSEA: (A) GO terms enriched in the control group; (B) GO terms enriched in the AD group; 212 (C) GO enrichment bar plot showing the top 10 significantly enriched MF, CC, and BP. (D) GO 213 enrichment bubble plot. (E) Circular plot of KEGG pathway enrichment analysis. (F) KEGG pathway 214 enrichment bubble plot. 215 Table 2. GO and KEGG enrichment analysis of AMG‐DEGs. Term ID Description GeneRatio p.Value GO:0048545 response to steroid hormone 8/39 3.67E-07BP GO:0043649 dicarboxylic acid catabolic process 3/39 5.42E-06 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 12 GO:0050767 regulation of neurogenesis 7/39 1.39E-05 GO:0009065 glutamine family amino acid catabolic process 3/39 2.05E-05 GO:0016188 synaptic vesicle maturation 3/39 2.57E-05 GO:0098982 GABA-ergic synapse 5/39 4.50E-07 GO:0030665 clathrin-coated vesicle membrane 5/39 7.11E-06 GO:0043025 neuronal cell body 7/39 3.99E-05 GO:0030662 coated vesicle membrane 5/39 4.79E-05 CC GO:0030136 clathrin-coated vesicle 5/39 6.41E-05 GO:0016831 carboxy-lyase activity 3/39 4.94E-05MF GO:0016830 carbon-carbon lyase activity 3/39 1.68E-04 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 13 GO:0030170 pyridoxal phosphate binding 3/39 2.21E-04 GO:0070279 vitamin B6 binding 3/39 2.33E-04 GO:0008503 benzodiazepine receptor activity 2/39 2.35E-04 hsa04727 GABAergic synapse 4/34 4.02E-04 hsa00250 Alanine, aspartate and glutamate metabolism 3/34 4.08E-04 hsa00430 Taurine and hypotaurine metabolism 2/34 2.01E-03 hsa05120 Epithelial cell signaling in Helicobacter pylori infection 3/34 2.74E-03 KEGG hsa00650 Butanoate metabolism 2/34 5.07E-03 216 3.3 Construction of PPI network and module analysis 217 A PPI network was constructed for the 39 AMG-DEGs using the STRING database with a 218 confidence score threshold of >0.4. The network was visualized in Cytoscape, which revealed a 219 network consisting of 34 nodes and 78 edges, with 15 upregulated genes highlighted in pink and 19 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 14 220 downregulated genes in green (Fig 4A). The MCODE plugin was used to identify key network modules, 221 which revealed two significant clusters. Module 1 contained 7 hub genes and 21 interactions, which 222 were primarily enriched in genes related to GABAergic neurotransmission, such as SST, GAD1, and 223 GAD2. This suggests a central role for these genes in neurotransmitter regulation (Fig 4B). Module 2 224 was composed of immune-related genes, including CCL2, CXCR4, and CD44, which are implicated in 225 immune response and inflammatory signaling (Fig 4C). These results highlight the involvement of 226 AMG-DEGs in both neurotransmission and immune regulation, suggesting that disrupted pathways in 227 AD may affect both neural and inflammatory processes. 228 Fig 4. PPI network of AMG-DEGs and functional subclusters. 229 (A) PPI network constructed from AMG-DEGs. (B-C) Subclusters extracted from the PPI network. 230 Red nodes indicate upregulated genes, green nodes indicate downregulated genes. 231 3.4 Identification and characterization of core AMG-DEGs 232 Using the UpSetR package and 10 centrality metrics in the cytoHubba plugin, we identified the 233 top 20 genes per metric and obtained the intersection, resulting in 7 consensus hub genes: GFAP, CCL2, 234 NFKBIA, TUBB3, GAD2, CD44, and HPRT1 (Fig 5A, S3 Table). Table 3 summarizes their full names 235 and functional annotations. GeneMANIA analysis of these hub genes revealed a complex co- 236 expression network dominated by co-expression interactions (80.85%), with co-localization 237 relationships accounting for the remaining 19.15% (Fig 5B). Key pathways identified as significantly 238 enriched included responses to bacterial molecules and lipopolysaccharides, regulation of leukocyte 239 cell-cell adhesion, cellular reactions to biotic stimuli, as well as negative regulation of apoptotic 240 signaling, among others. 241 Fig 5. Identification of key hub genes among AMG - DEGs. .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 15 242 (A) UpSet plot showing the intersection of seven hub genes identified by 10 different computational 243 algorithms. (B) GeneMANIA analysis of core AMG-DEGs and their co-expressed genes. 244 Table 3. The details of the core AMG-DEGs. Gene symbol Full name Function of genea GFAP Glial Fibrillary Acidic Protein A class-III intermediate filament serves as a cell-specific marker, distinguishing astrocytes from other glial cells during central nervous system development. GAD2 Glutamate Decarboxylas e 2 Catalyzes the production of GABA. CD44 CD44 Molecule (IN Blood Group) CD44 plays a key role in immune regulation, cell adhesion and migration, and signal transduction. CCL2 C-C Motif Chemokine Ligand 2 The CCL2 gene encodes a protein that acts as a ligand for the CCR2 receptor. By activating CCR2, it triggers calcium ion influx and chemotactic responses, specifically recruiting monocytes and basophils (but not neutrophils or eosinophils). .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 16 It may play a role in atherosclerosis by promoting monocyte migration into arterial walls. NFKBI A NFKB Inhibitor Alpha The NFKBIA gene encodes a protein that inhibits NF-κB dimeric complexes (e.g., RELA/p65 and NFKB1/p50) by masking their nuclear localization signals, trapping them in the cytoplasm. Upon immune or inflammatory stimulation, NFKBIA is phosphorylated and degraded, releasing NF-κB to translocate into the nucleus and activate target gene transcription. HPRT1 Hypoxanthin e Phosphoribos yltransferase 1 Catalyzes the conversion of guanine to guanosine monophosphate and hypoxanthine to inosine monophosphate. Transfers the 5-phosphoribosyl group from 5- phosphoribosylpyrophosphate to the purine. Essential for purine nucleotide synthesis via the purine salvage pathway. TUBB3 Tubulin Beta 3 Class III The TUBB3 gene encodes β-tubulin, a core component of microtubules that regulates dynamic assembly (GTP-bound state promotes growth while GDP-bound triggers disassembly). It is essential for axon guidance and maintenance by modulating microtubule dynamics (e.g., dorsal root ganglion axon projection) and mediates axon .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 17 repulsion through interaction with the Netrin-1/UNC5C signaling pathway. 245 aGene functional annotations were obtained from GeneCards (https://www.genecards.org/). 246 3.5 Construction of a ceRNA regulatory network 247 To further elucidate the post-transcriptional regulatory mechanisms associated with AD, we 248 constructed a ceRNA regulatory network (Fig 6A). In this network, HPRT1, CD44, and CCL2 emerged 249 as central nodes, suggesting their potential roles in asparagine metabolism-related ceRNA regulation. 250 lncRNAs including RP4-539M6.22, CITF22-1A6.3, LA16c-306A4.2, and SNHG14 were predicted to 251 regulate the expression of HPRT1 mRNA by competitively binding to hsa-miR-130a-3p. Similarly, 252 CTC-459F4.1 was found to potentially modulate HPRT1 through competition for hsa-miR-576-5p. In 253 the regulation of CCL2, lncRNAs LINC01043, GNG12-AS1, and RP3-470B24.5 were identified as 254 ceRNAs via their interaction with hsa-miR-1-3p. Moreover, lncRNAs GS1-251I9.3, CTC-457E21.1, 255 and RP11-486P11.1 were predicted to regulate CD44 expression through competition for hsa-miR- 256 130b-5p. These lncRNA-miRNA-mRNA regulatory axes may play critical roles in the pathogenesis of 257 AD. 258 3.6 Construction of a transcription factor regulatory network 259 A transcriptional regulatory network consisting of 9 nodes and 12 interactions was established 260 based on known transcription factor (TF)-target relationships. Our analysis revealed that CCL2, a key 261 pro-inflammatory cytokine, is regulated by multiple TFs, including NFKB1, STAT3, SP1, NFIC, and 262 RELA. Likewise, the expression of GFAP is modulated by NFKB1, STAT3, NFIC, and RELA. NFKBIA 263 was found to be jointly regulated by NFKB1 and RELA, while CD44 is transcriptionally controlled by .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 18 264 SP1. These results highlight NFKB1, STAT3, SP1, NFIC, and RELA as core transcriptional regulators 265 of AMG-DEGs in AD (Fig 6B). 266 Fig 6. Regulatory networks of ceRNAs and transcription factors. 267 (A) ceRNA regulatory network: green hexagons represent miRNAs, blue rhomboids represent 268 lncRNAs, and red circles represent mRNAs. (B) TF regulatory network: green rectangles represent TF, 269 red ellipses represent target genes. 270 3.7 Identification of potential therapeutic compounds 271 To explore therapeutic strategies targeting key AMG-DEGs, we conducted drug enrichment 272 analysis. The top candidate compounds with the highest statistical significance included MVA, BCS, 273 PEITC, MELAMINE, and CHLOROBENZENE (Fig 7A). A drug-gene interaction network was 274 constructed to further elucidate the molecular mechanisms underlying these compounds (Fig 7B). The 275 network revealed potential interactions between the identified compounds and the hub genes, 276 suggesting that these molecules may exert therapeutic effects by modulating inflammation- or 277 neurofunction-related pathways involved in AD pathogenesis. 278 Fig 7. Drug enrichment analysis and gene-drug interaction network. 279 (A) Bar plot of drug enrichment results. (B) Drug-gene interaction network illustrating the potential 280 associations between identified drugs and target genes. 281 3.8 Molecular docking analysis 282 To validate the therapeutic potential of the candidate compounds, molecular docking analysis was 283 performed using CB-Dock, and the binding affinities were evaluated using Vina scores (Table 4, Fig .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 19 284 8). Among the compounds, BCS exhibited the strongest binding affinity, particularly with CD44, 285 showing a Vina score of -9.3 kcal/mol (Fig 8C), outperforming MVA (-5.7 kcal/mol, Fig 8A). 286 Additionally, BCS demonstrated high binding affinities with CCL2 and NFKBIA (Vina scores: -7.6 287 and -7.1 kcal/mol, respectively; Fig 8B, D), supporting its potential as a multi-target modulator. PEITC 288 showed a favorable binding affinity to TUBB3 (-6.2 kcal/mol) with a relatively large cavity volume 289 (1608 ų, Fig 8E), although its interaction with CCL2 and NFKBIA was relatively weak (Vina scores: 290 -4.2 and -4.4 kcal/mol, respectively; S2 Fig). In contrast, MELAMINE and CHLOROBENZENE 291 exhibited poor binding capacities across all docking analyses, with the highest Vina scores of -4.2 and 292 -3.8 kcal/mol, respectively, suggesting limited therapeutic potential (S2 Fig). 293 Fig 8. Molecular docking results of small-molecule compounds with target proteins. 294 (A) Molecular docking of DL-Mevalonic acid with CD44 protein. (B) Docking result of Bathocuproine 295 disulfonate with NFKBIA protein. (C) Docking result of Bathocuproine disulfonate with CD44 protein. 296 (D) Docking result of Bathocuproine disulfonate with CCL2 protein. (E) Docking result of Phenethyl 297 isothiocyanate with TUBB3 protein. 298 Table 4. Binding affinity and pocket volume of drug-protein complexes identified by molecular 299 docking. Drug Protein Vina score (kcal/mol) Cavity volume (ų) DL-Mevalonic acid CD44 -5.7 121 DL-Mevalonic acid CCL2 -4.1 82 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 20 DL-Mevalonic acid NFKBIA -3.7 285 Bathocuproine disulfonate CD44 -9.3 308 Bathocuproine disulfonate CCL2 -7.6 102 Bathocuproine disulfonate NFKBIA -7.1 72 Phenethyl isothiocyanate CCL2 -4.2 107 Phenethyl isothiocyanate NFKBIA -4.4 72 Phenethyl isothiocyanate TUBB3 -6.2 1608 MELAMINE CCL2 -4.2 107 MELAMINE NFKBIA -4.2 285 CHLOROBENZENE CCL2 -3.5 102 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 21 CHLOROBENZENE NFKBIA -3.8 72 300 301 Discussion 302 AD remains one of the most widespread neurodegenerative conditions globally, particularly 303 affecting the aging population. Despite recent advances in AD diagnosis and treatment, the disease 304 remains incurable, with current therapeutic strategies primarily aimed at alleviating symptoms rather 305 than halting disease progression(23). Identifying new molecular targets involved in the core processes 306 of AD is increasingly important. This study offers fresh insights into how Asn metabolism may 307 contribute to AD progression. Our results emphasize the role of key genes and regulatory mechanisms 308 that impact neuroinflammation, synaptic function, and energy balance in AD. 309 GSEA in this study revealed impairments in energy metabolism, synaptic dysfunction, and 310 vascular dysfunction of AD, which aligns with previous reports(24, 25). GO analysis further supported 311 this, indicating that genes involved in Asn metabolism are primarily engaged in neurogenesis, nervous 312 system development, and synaptic plasticity—processes tightly linked to cognitive decline in AD(26, 313 27). In particular, the CC enrichment results showed significant AMG-DEGs localization in neuronal 314 cell bodies, GABAergic synapses, and clathrin-coated vesicle membranes, indicating that asparagine 315 metabolism may impact synaptic signaling and vesicle transport. Previous studies have demonstrated 316 that GABAergic synapses are essential for inhibitory neural regulation(28), and disruptions in this 317 system have been associated with synaptic abnormalities in AD and other neurodegenerative 318 diseases(29, 30). Interestingly, MF enrichment analysis revealed a significant association with carbon- 319 carbon lyase activity and pyridoxal phosphate binding. Pyridoxal phosphate is a crucial cofactor for 320 neurotransmitter biosynthesis(31), indicating that alterations in Asn metabolism may disrupt .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 22 321 neurotransmitter balance in AD. KEGG pathway analysis revealed a downregulation in butyrate and 322 β-alanine metabolism pathways, while inflammation-related pathways, such as the IL-17 signaling 323 pathway, were upregulated. This inverse regulation between metabolism and inflammation suggests 324 an imbalance in the immune-metabolic axis in AD. Although previous studies demonstrated the 325 dynamic interplay between neuroinflammation and disturbances in lipid, glucose, and amino acid 326 metabolism in AD(32, 33), our study identified asparagine metabolism as a previously unrecognized 327 key node within the immune-metabolic axis. This metabolite-specific mechanism not only supports the 328 concept of the immune-metabolic axis but also provides further evidence for how discrete metabolic 329 dysregulation can modulate neuroinflammation, thus refining the theoretical framework of immune- 330 metabolism in AD. 331 Among the seven hub genes identified in our study, GFAP, CD44, CCL2, and NFKBIA were 332 upregulated, while GAD2, TUBB3, and HPRT1 were downregulated. These genes play pivotal roles in 333 inflammation, synaptic stability, and neurotransmitter regulation, which are critical to AD 334 pathophysiology. The upregulation of neuroinflammatory markers like GFAP(34), CD44(35), and 335 CCL2(36) in AD corroborates findings from prior studies linking chronic inflammation to disease 336 progression(35, 37-39). Upregulation of CCL2(40) and NFKBIA(41) is associated with the IL-17 337 signaling pathway, further validating the results from the KEGG pathway analysis. Furthermore, the 338 downregulation of GAD2, which encodes glutamate decarboxylase responsible for GABA synthesis, 339 is consistent with the known imbalance in excitatory-inhibitory neurotransmission in AD(42, 43). Our 340 findings show that both TUBB3 and HPRT1 are downregulated in AD. Research has shown that 341 downregulation of TUBB3 may destabilize microtubules, leading to Tau hyperphosphorylation and 342 neuronal structural damage in AD(44). Additionally, consistent with extensive literature on energy 343 metabolism imbalances in AD(45, 46), our results suggest that the downregulation of HPRT1 could .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 23 344 impair purine metabolism, leading to insufficient energy supply and mitochondrial dysfunction in 345 neurons. 346 The regulatory networks constructed in this study also highlighted the complexity of post- 347 transcriptional regulation in AD. The ceRNA network analysis suggested that multiple lncRNAs and 348 miRNAs regulate hub genes such as HPRT1, CD44, and CCL2. These interactions may modulate the 349 inflammatory and energy metabolism in AD. Although previous studies have shown that under 350 ischemic conditions, SNHG14 exacerbates neuronal damage through the miR-182-5p/BINP3 signaling 351 axis(47), our study identifies a novel mechanism in AD, wherein SNHG14 modulates the miR-130a- 352 3p/HPRT1 pathway to influence purine metabolism. This differential regulatory pattern suggests that 353 SNHG14 may act as a 'metabolic switch', maintaining energy homeostasis by regulating mitochondrial 354 autophagy during acute injury, while modulating energy supply through the purine metabolism 355 pathway in chronic neurodegenerative conditions. The TF network revealed the involvement of NFKB1, 356 STAT3, and RELA in regulating these pathways, emphasizing their potential as therapeutic targets in 357 modulating the immune response and synaptic plasticity. Our analysis revealed that CD44 is uniquely 358 regulated by the transcription factor SP1, suggesting that CD44 may influence inflammatory responses 359 through an independent pathway(48). Notably, NFKBIA, a negative regulator of NF-κB signaling, 360 appears to act as a compensatory mechanism to suppress excessive inflammation. However, this 361 inhibition may be insufficient to counterbalance the chronic neuroinflammation observed in AD 362 progression(49). 363 The potential therapeutic compounds identified through drug enrichment and molecular docking 364 analysis, including BCS, MVA, and PEITC, show promising results in targeting key genes and 365 pathways involved in AD. Molecular docking results revealed that Bisulfate-derived compound BCS 366 exhibits a strong binding affinity to CD44, CCL2, and NFKBIA proteins, suggesting that BCS may .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 24 367 alleviate the progression of AD through the inhibition of inflammatory pathways. Previous studies 368 have indicated that excessive copper in Alzheimer's disease may exacerbate Aβ aggregation and 369 oxidative stress(50). BCS, a copper chelator, has shown potential therapeutic value by restoring metal 370 ion balance and may therefore provide beneficial effects in the context of AD(51, 52). MVA, a key 371 intermediate in cholesterol biosynthesis, was significantly enriched in our analysis. Cholesterol 372 metabolism dysregulation has been implicated in AD pathogenesis(53, 54). Statins, which target this 373 pathway, have shown promise in AD prevention and treatment by reducing Aβ accumulation, 374 suppressing inflammation, enhancing vascular function, and modulating Tau phosphorylation(55). 375 Nonetheless, the optimal therapeutic window and specific statin formulations for AD remain under 376 debate. Intriguingly, we found that MVA exhibits strong binding affinity to CD44 proteins, a finding 377 not widely reported in previous studies. This suggests a potential link between MVA and CD44 378 signaling in AD, warranting further investigation. PEITC is a naturally occurring compound known 379 for its antioxidant and anti-inflammatory properties(56). Our docking analysis indicated significant 380 binding affinity between PEITC and TUBB3 proteins, suggesting that PEITC may help stabilize 381 microtubule structures and mitigate neuronal damage associated with AD. Although the binding 382 affinity of PEITC to CCL2 and NFKBIA proteins was relatively weaker, a considerable body of 383 evidence suggests that PEITC not only inhibits Aβ aggregation in AD(57) ， but also exerts its 384 therapeutic effects through antioxidant and anti-inflammatory mechanisms, thereby slowing the 385 disease's progression(58, 59). However, despite promising preclinical findings, clinical studies 386 investigating the efficacy of BCS, MVA, and PEITC in AD treatment are currently lacking, and further 387 validation of their practical effects is required. 388 This study comprehensively elucidated the potential involvement of asparagine metabolism in 389 AD pathogenesis and identified several key molecular targets. Nevertheless, some limitations should 390 be noted. First, our findings are primarily based on public datasets and bioinformatics predictions, .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 25 391 requiring further validation through experiments. Second, the clinical translational potential of our 392 molecular docking results needs to be further assessed. Future studies integrating single-cell 393 sequencing, metabolomics, and experimental validation are warranted to clarify the role of asparagine 394 metabolism in AD and evaluate its feasibility as a therapeutic target. 395 Conclusion 396 In conclusion, this study sheds light on the role of asparagine metabolism in AD. We identified 397 key genes associated with neuroinflammation, synaptic function, and energy metabolism, suggesting 398 they could be potential targets for therapy. Our analysis of regulatory networks also uncovered intricate 399 interactions between miRNAs, lncRNAs, and transcription factors that influence these genes. 400 Moreover, drug screening and molecular docking revealed several promising compounds that may 401 offer therapeutic benefits for AD. These findings indicate that modulating asparagine metabolism could 402 be a new strategy for AD treatment. However, further research is needed to validate these results 403 experimentally. 404 Acknowledgments 405 We gratefully acknowledge all contributors for their valuable participation in this study. 406 Funding Statement 407 This study was supported by the Yunnan Science and Technology Program (Grant/Award 408 Number: 202401AT070176). 409 Abbreviations .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 26 410 Aβ, β-amyloid; AD, Alzheimer’s disease; AEP, asparagine endopeptidase; AMGs, asparagine 411 metabolism-related genes; AMG ‐DEGs, asparagine metabolism ‐differentially expressed genes; 412 Asn, asparagine; BCS, Bathocuproine disulfonate; Betweenness, Betweenness Centrality; BottleNeck, 413 BottleNeck Algorithm; BP, biological process; CC, cellular component; ceRNA, competing 414 endogenous RNA; Closeness, Closeness Centrality; DEG, differentially expressed gene; Degree, 415 Degree Centrality; DGIdb, Drug-Gene Interaction Database; DMNC, Density of Maximum 416 Neighborhood Component; DSigDB, Drug Signatures Database; EcCentricity, Eccentricity Centrality; 417 GeneMANIA, Gene Multiple Association Network Integration Algorithm; GO, Gene Ontology; 418 KEGG, Kyoto Encyclopedia of Genes and Genomes; lncRNAs, Long non-coding RNAs; MCC, 419 Maximal Clique Centrality; MF, molecular function; MNC, Maximum Neighborhood Component; 420 MVA, DL-Mevalonic acid; PCA, principal component analysis; PEITC, Phenethyl isothiocyanate; PPI, 421 protein-protein interaction; Radiality, Radiality Centrality; Stress, Stress Centrality; STRING, Search 422 Tool for the Retrieval of Interacting Genes; TCA, tricarboxylic acid; TF, Transcription factor. 423 References 424 1. Niu Q, Li D, Zhang J, Piao Z, Xu B, Xi Y, et al. The new perspective of Alzheimer's Disease 425 Research: Mechanism and therapeutic strategy of neuronal senescence. Ageing Res Rev. 426 2024;102:102593. Epub 2024/11/21. doi: 10.1016/j.arr.2024.102593. PubMed PMID: 39566741. 427 2. Tenchov R, Sasso JM, Zhou QA. Alzheimer's Disease: Exploring the Landscape of Cognitive 428 Decline. ACS Chem Neurosci. 2024;15(21):3800-27. Epub 2024/10/11. doi: 429 10.1021/acschemneuro.4c00339. PubMed PMID: 39392435; PubMed Central PMCID: 430 PMC11587518. 431 3. Azargoonjahromi A. The duality of amyloid-β: its role in normal and Alzheimer's disease 432 states. Mol Brain. 2024;17(1):44. Epub 2024/07/18. doi: 10.1186/s13041-024-01118-1. PubMed 433 PMID: 39020435; PubMed Central PMCID: PMC11256416. 434 4. Wang S, Jiang Y, Yang A, Meng F, Zhang J. The Expanding Burden of Neurodegenerative 435 Diseases: An Unmet Medical and Social Need. Aging Dis. 2024. Epub 2024/11/21. doi: 436 10.14336/ad.2024.1071. PubMed PMID: 39571158. .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 27 437 5. Zhang Y, Li Y, Ma L. Recent advances in research on Alzheimer's disease in China. J Clin 438 Neurosci. 2020;81:43-6. Epub 2020/11/24. doi: 10.1016/j.jocn.2020.09.018. PubMed PMID: 439 33222956. 440 6. Peng Y, Gao P, Shi L, Chen L, Liu J, Long J. Central and Peripheral Metabolic Defects 441 Contribute to the Pathogenesis of Alzheimer's Disease: Targeting Mitochondria for Diagnosis and 442 Prevention. Antioxid Redox Signal. 2020;32(16):1188-236. Epub 2020/02/14. doi: 443 10.1089/ars.2019.7763. PubMed PMID: 32050773; PubMed Central PMCID: PMC7196371. 444 7. Xu L, Liu R, Qin Y, Wang T. Brain metabolism in Alzheimer's disease: biological 445 mechanisms of exercise. Transl Neurodegener. 2023;12(1):33. Epub 2023/06/27. doi: 446 10.1186/s40035-023-00364-y. PubMed PMID: 37365651; PubMed Central PMCID: PMC10294518. 447 8. Ezkurdia A, Ramírez MJ, Solas M. Metabolic Syndrome as a Risk Factor for Alzheimer's 448 Disease: A Focus on Insulin Resistance. Int J Mol Sci. 2023;24(5). Epub 2023/03/12. doi: 449 10.3390/ijms24054354. PubMed PMID: 36901787; PubMed Central PMCID: PMC10001958. 450 9. Yue C, Fu Y, Zhao Y, Ou Y, Sun Y, Tan L. Association between Alzheimer’s disease and 451 metabolic syndrome: Unveiling the role of dyslipidemia mechanisms. Brain Network Disorders. 452 2025;1(1):21-7. doi: https://doi.org/10.1016/j.bnd.2024.10.006. 453 10. Lauretti E, Dabrowski K, Praticò D. The neurobiology of non-coding RNAs and Alzheimer’s 454 disease pathogenesis: Pathways, mechanisms and translational opportunities. Ageing Research 455 Reviews. 2021;71:101425. doi: https://doi.org/10.1016/j.arr.2021.101425 . 456 11. Chang H-C, Tsai C-Y, Hsu C-L, Tai T-S, Cheng M-L, Chuang Y-M, et al. Asparagine 457 deprivation enhances T cell antitumour response in patients via ROS-mediated metabolic and signal 458 adaptations. Nature Metabolism. 2025. doi: 10.1038/s42255-025-01245-6. 459 12. Lomelino CL, Andring JT, McKenna R, Kilberg MS. Asparagine synthetase: Function, 460 structure, and role in disease. J Biol Chem. 2017;292(49):19952-8. Epub 2017/11/01. doi: 461 10.1074/jbc.R117.819060. PubMed PMID: 29084849; PubMed Central PMCID: PMC5723983. 462 13. Novotny BC, Fernandez MV, Wang C, Budde JP, Bergmann K, Eteleeb AM, et al. 463 Metabolomic and lipidomic signatures in autosomal dominant and late-onset Alzheimer's disease 464 brains. Alzheimers Dement. 2023;19(5):1785-99. Epub 2022/10/18. doi: 10.1002/alz.12800. PubMed 465 PMID: 36251323; PubMed Central PMCID: PMC10106526. 466 14. Chatterjee P, Cheong YJ, Bhatnagar A, Goozee K, Wu Y, McKay M, et al. Plasma 467 metabolites associated with biomarker evidence of neurodegeneration in cognitively normal older 468 adults. J Neurochem. 2021;159(2):389-402. Epub 2020/07/18. doi: 10.1111/jnc.15128. PubMed 469 PMID: 32679614. 470 15. Alkan HF, Bogner-Strauss JG. Maintaining cytosolic aspartate levels is a major function of 471 the TCA cycle in proliferating cells. Mol Cell Oncol. 2019;6(5):e1536843. Epub 2019/09/19. doi: 472 10.1080/23723556.2018.1536843. PubMed PMID: 31528687; PubMed Central PMCID: 473 PMC6736317. 474 16. González-Domínguez R, García-Barrera T, Gómez-Ariza JL. Metabolite profiling for the 475 identification of altered metabolic pathways in Alzheimer's disease. J Pharm Biomed Anal. 476 2015;107:75-81. Epub 2015/01/13. doi: 10.1016/j.jpba.2014.10.010. PubMed PMID: 25575172. 477 17. Bradberry MM, Peters-Clarke TM, Shishkova E, Chapman ER, Coon JJ. N-glycoproteomics 478 of brain synapses and synaptic vesicles. Cell Rep. 2023;42(4):112368. Epub 2023/04/11. doi: 479 10.1016/j.celrep.2023.112368. PubMed PMID: 37036808; PubMed Central PMCID: PMC10560701. .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 28 480 18. Shimonaka S, Matsumoto SE, Elahi M, Ishiguro K, Hasegawa M, Hattori N, et al. Asparagine 481 residue 368 is involved in Alzheimer's disease tau strain-specific aggregation. J Biol Chem. 482 2020;295(41):13996-4014. Epub 2020/08/08. doi: 10.1074/jbc.RA120.013271. PubMed PMID: 483 32759167; PubMed Central PMCID: PMC7549045. 484 19. Tang TMS, Luk LYP. Asparaginyl endopeptidases: enzymology, applications and limitations. 485 Org Biomol Chem. 2021;19(23):5048-62. Epub 2021/05/27. doi: 10.1039/d1ob00608h. PubMed 486 PMID: 34037066; PubMed Central PMCID: PMC8209628. 487 20. Song M. The Asparaginyl Endopeptidase Legumain: An Emerging Therapeutic Target and 488 Potential Biomarker for Alzheimer's Disease. Int J Mol Sci. 2022;23(18). Epub 2022/09/24. doi: 489 10.3390/ijms231810223. PubMed PMID: 36142134; PubMed Central PMCID: PMC9499314. 490 21. Behrendt A, Bichmann M, Ercan-Herbst E, Haberkant P, Schöndorf DC, Wolf M, et al. 491 Asparagine endopeptidase cleaves tau at N167 after uptake into microglia. Neurobiology of Disease. 492 2019;130:104518. doi: https://doi.org/10.1016/j.nbd.2019.104518. 493 22. Meng X, Li B, Wang M, Zheng W, Ye K. Development of asparagine endopeptidase 494 inhibitors for treating neurodegenerative diseases. Trends in Molecular Medicine. 2025;31(4):359-72. 495 doi: https://doi.org/10.1016/j.molmed.2025.01.009. 496 23. Ogos M, Stary D, Bajda M. Recent Advances in the Search for Effective Anti-Alzheimer's 497 Drugs. Int J Mol Sci. 2024;26(1). Epub 2025/01/11. doi: 10.3390/ijms26010157. PubMed PMID: 498 39796014; PubMed Central PMCID: PMC11720639. 499 24. Wilson DM, 3rd, Cookson MR, Van Den Bosch L, Zetterberg H, Holtzman DM, Dewachter 500 I. Hallmarks of neurodegenerative diseases. Cell. 2023;186(4):693-714. Epub 2023/02/22. doi: 501 10.1016/j.cell.2022.12.032. PubMed PMID: 36803602. 502 25. Yuan Y, Zhao G, Zhao Y. Dysregulation of energy metabolism in Alzheimer's disease. J 503 Neurol. 2024;272(1):2. Epub 2024/12/02. doi: 10.1007/s00415-024-12800-8. PubMed PMID: 504 39621206; PubMed Central PMCID: PMC11611936. 505 26. Zhang X, Wei X, Mei Y, Wang D, Wang J, Zhang Y, et al. Modulating adult neurogenesis 506 affects synaptic plasticity and cognitive functions in mouse models of Alzheimer's disease. Stem Cell 507 Reports. 2021;16(12):3005-19. Epub 2021/12/04. doi: 10.1016/j.stemcr.2021.11.003. PubMed 508 PMID: 34861165; PubMed Central PMCID: PMC8693766. 509 27. Peng L, Bestard-Lorigados I, Song W. The synapse as a treatment avenue for Alzheimer's 510 Disease. Mol Psychiatry. 2022;27(7):2940-9. Epub 2022/04/22. doi: 10.1038/s41380-022-01565-z. 511 PubMed PMID: 35444256. 512 28. Lepeta K, Lourenco MV, Schweitzer BC, Martino Adami PV, Banerjee P, Catuara-Solarz S, 513 et al. Synaptopathies: synaptic dysfunction in neurological disorders - A review from students to 514 students. J Neurochem. 2016;138(6):785-805. Epub 2016/06/23. doi: 10.1111/jnc.13713. PubMed 515 PMID: 27333343; PubMed Central PMCID: PMC5095804. 516 29. Krueger-Burg D. Understanding GABAergic synapse diversity and its implications for 517 GABAergic pharmacotherapy. Trends Neurosci. 2025;48(1):47-61. Epub 2025/01/09. doi: 518 10.1016/j.tins.2024.11.007. PubMed PMID: 39779392. 519 30. Tang X, Jaenisch R, Sur M. The role of GABAergic signalling in neurodevelopmental 520 disorders. Nat Rev Neurosci. 2021;22(5):290-307. Epub 2021/03/28. doi: 10.1038/s41583-021- 521 00443-x. PubMed PMID: 33772226; PubMed Central PMCID: PMC9001156. .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 29 522 31. Bisello G, Longo C, Rossignoli G, Phillips RS, Bertoldi M. Oxygen reactivity with pyridoxal 523 5'-phosphate enzymes: biochemical implications and functional relevance. Amino Acids. 524 2020;52(8):1089-105. Epub 2020/08/28. doi: 10.1007/s00726-020-02885-6. PubMed PMID: 525 32844248; PubMed Central PMCID: PMC7497351. 526 32. Jung ES, Choi H, Mook-Jung I. Decoding microglial immunometabolism: a new frontier in 527 Alzheimer's disease research. Mol Neurodegener. 2025;20(1):37. Epub 2025/03/28. doi: 528 10.1186/s13024-025-00825-0. PubMed PMID: 40149001; PubMed Central PMCID: PMC11948825. 529 33. Chen H, Guo Z, Sun Y, Dai X. The immunometabolic reprogramming of microglia in 530 Alzheimerʼs disease. Neurochemistry International. 2023;171:105614. doi: 531 https://doi.org/10.1016/j.neuint.2023.105614. 532 34. Li D, Liu X, Liu T, Liu H, Tong L, Jia S, et al. Neurochemical regulation of the expression 533 and function of glial fibrillary acidic protein in astrocytes. Glia. 2020;68(5):878-97. Epub 534 2019/10/19. doi: 10.1002/glia.23734. PubMed PMID: 31626364. 535 35. Al-Dalahmah O, Sosunov AA, Sun Y, Liu Y, Madden N, Connolly ES, et al. The Matrix 536 Receptor CD44 Is Present in Astrocytes throughout the Human Central Nervous System and 537 Accumulates in Hypoxia and Seizures. Cells. 2024;13(2). Epub 2024/01/22. doi: 538 10.3390/cells13020129. PubMed PMID: 38247821; PubMed Central PMCID: PMC10814649. 539 36. Guo S, Zhang Q, Guo Y, Yin X, Zhang P, Mao T, et al. The role and therapeutic targeting of 540 the CCL2/CCR2 signaling axis in inflammatory and fibrotic diseases. Front Immunol. 541 2024;15:1497026. Epub 2025/01/24. doi: 10.3389/fimmu.2024.1497026. PubMed PMID: 39850880; 542 PubMed Central PMCID: PMC11754255. 543 37. Roveta F, Bonino L, Piella EM, Rainero I, Rubino E. Neuroinflammatory Biomarkers in 544 Alzheimer's Disease: From Pathophysiology to Clinical Implications. Int J Mol Sci. 2024;25(22). 545 Epub 2024/11/27. doi: 10.3390/ijms252211941. PubMed PMID: 39596011; PubMed Central 546 PMCID: PMC11593837. 547 38. Kölliker-Frers R, Udovin L, Otero-Losada M, Kobiec T, Herrera MI, Palacios J, et al. 548 Neuroinflammation: An Integrating Overview of Reactive-Neuroimmune Cell Interactions in Health 549 and Disease. Mediators Inflamm. 2021;2021:9999146. Epub 2021/06/24. doi: 550 10.1155/2021/9999146. PubMed PMID: 34158806; PubMed Central PMCID: PMC8187052. 551 39. Hu D, Mo X, Jihang L, Huang C, Xie H, Jin L. Novel diagnostic biomarkers of oxidative 552 stress, immunological characterization and experimental validation in Alzheimer's disease. Aging 553 (Albany NY). 2023;15(19):10389-406. Epub 2023/10/06. doi: 10.18632/aging.205084. PubMed 554 PMID: 37801482; PubMed Central PMCID: PMC10599743. 555 40. Ruiz de Morales JMG, Puig L, Daudén E, Cañete JD, Pablos JL, Martín AO, et al. Critical 556 role of interleukin (IL)-17 in inflammatory and immune disorders: An updated review of the 557 evidence focusing in controversies. Autoimmunity Reviews. 2020;19(1):102429. doi: 558 https://doi.org/10.1016/j.autrev.2019.102429. 559 41. Liu Y, Meng Y, Zhou C, Yan J, Guo C, Dong W. Activation of the IL-17/TRAF6/NF-κB 560 pathway is implicated in Aβ-induced neurotoxicity. BMC Neurosci. 2023;24(1):14. Epub 561 2023/02/25. doi: 10.1186/s12868-023-00782-8. PubMed PMID: 36823558; PubMed Central 562 PMCID: PMC9951515. .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 30 563 42. Chiu CQ, Barberis A, Higley MJ. Preserving the balance: diverse forms of long-term 564 GABAergic synaptic plasticity. Nat Rev Neurosci. 2019;20(5):272-81. Epub 2019/03/07. doi: 565 10.1038/s41583-019-0141-5. PubMed PMID: 30837689. 566 43. Koh W, Kwak H, Cheong E, Lee CJ. GABA tone regulation and its cognitive functions in the 567 brain. Nat Rev Neurosci. 2023;24(9):523-39. Epub 2023/07/27. doi: 10.1038/s41583-023-00724-7. 568 PubMed PMID: 37495761. 569 44. Radwitz J, Hausrat TJ, Heisler FF, Janiesch PC, Pechmann Y, Rübhausen M, et al. Tubb3 570 expression levels are sensitive to neuronal activity changes and determine microtubule growth and 571 kinesin-mediated transport. Cell Mol Life Sci. 2022;79(11):575. Epub 2022/10/31. doi: 572 10.1007/s00018-022-04607-5. PubMed PMID: 36309617; PubMed Central PMCID: PMC9617967. 573 45. Vinokurov AY, Soldatov VO, Seregina ES, Dolgikh AI, Tagunov PA, Dunaev AV, et al. 574 HPRT1 Deficiency Induces Alteration of Mitochondrial Energy Metabolism in the Brain. Mol 575 Neurobiol. 2023;60(6):3147-57. Epub 2023/02/22. doi: 10.1007/s12035-023-03266-2. PubMed 576 PMID: 36802322; PubMed Central PMCID: PMC10122629. 577 46. Sekine M, Fujiwara M, Okamoto K, Ichida K, Nagata K, Hille R, et al. Significance and 578 amplification methods of the purine salvage pathway in human brain cells. J Biol Chem. 579 2024;300(8):107524. Epub 2024/07/04. doi: 10.1016/j.jbc.2024.107524. PubMed PMID: 38960035; 580 PubMed Central PMCID: PMC11342100. 581 47. Deng Z, Ou H, Ren F, Guan Y, Huan Y, Cai H, et al. LncRNA SNHG14 promotes OGD/R- 582 induced neuron injury by inducing excessive mitophagy via miR-182-5p/BINP3 axis in HT22 mouse 583 hippocampal neuronal cells. Biol Res. 2020;53(1):38. Epub 2020/09/12. doi: 10.1186/s40659-020- 584 00304-4. PubMed PMID: 32912324; PubMed Central PMCID: PMC7488096. 585 48. Lee Y, Lee J, Jo D-G. A novel function of CD44 in the pathogenesis of Alzheimer’s disease. 586 Alzheimer's & Dementia. 2023;19(S21):e076285. doi: https://doi.org/10.1002/alz.076285. 587 49. Yu H, Lin L, Zhang Z, Zhang H, Hu H. Targeting NF-κB pathway for the therapy of diseases: 588 mechanism and clinical study. Signal Transduct Target Ther. 2020;5(1):209. Epub 2020/09/23. doi: 589 10.1038/s41392-020-00312-6. PubMed PMID: 32958760; PubMed Central PMCID: PMC7506548. 590 50. Cheignon C, Tomas M, Bonnefont-Rousselot D, Faller P, Hureau C, Collin F. Oxidative 591 stress and the amyloid beta peptide in Alzheimer’s disease. Redox Biology. 2018;14:450-64. doi: 592 https://doi.org/10.1016/j.redox.2017.10.014. 593 51. You H, Tsutsui S, Hameed S, Kannanayakal TJ, Chen L, Xia P, et al. Aβ neurotoxicity 594 depends on interactions between copper ions, prion protein, and N-methyl-D-aspartate receptors. 595 Proc Natl Acad Sci U S A. 2012;109(5):1737-42. Epub 2012/02/07. doi: 10.1073/pnas.1110789109. 596 PubMed PMID: 22307640; PubMed Central PMCID: PMC3277185. 597 52. Bulcke F, Santofimia-Castaño P, Gonzalez-Mateos A, Dringen R. Modulation of copper 598 accumulation and copper-induced toxicity by antioxidants and copper chelators in cultured primary 599 brain astrocytes. J Trace Elem Med Biol. 2015;32:168-76. Epub 2015/08/26. doi: 600 10.1016/j.jtemb.2015.07.001. PubMed PMID: 26302925. 601 53. Zeki AA, Yeganeh B, Kenyon NJ, Ghavami S. Editorial: New Insights into a Classical 602 Pathway: Key Roles of the Mevalonate Cascade in Different Diseases (Part II). Curr Mol Pharmacol. 603 2017;10(2):74-6. Epub 2017/04/26. doi: 10.2174/187446721002170301204357. PubMed PMID: 604 28440195; PubMed Central PMCID: PMC6018051. .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint 31 605 54. Varma VR, Büşra Lüleci H, Oommen AM, Varma S, Blackshear CT, Griswold ME, et al. 606 Abnormal brain cholesterol homeostasis in Alzheimer’s disease—a targeted metabolomic and 607 transcriptomic study. npj Aging and Mechanisms of Disease. 2021;7(1):11. doi: 10.1038/s41514- 608 021-00064-9. 609 55. Pappolla MA, Refolo L, Sambamurti K, Zambon D, Duff K. Hypercholesterolemia and 610 Alzheimer's Disease: Unraveling the Connection and Assessing the Efficacy of Lipid-Lowering 611 Therapies. J Alzheimers Dis. 2024;101(s1):S371-s93. Epub 2024/10/18. doi: 10.3233/jad-240388. 612 PubMed PMID: 39422957. 613 56. Coscueta ER, Sousa AS, Reis CA, Pintado MM. Phenylethyl Isothiocyanate: A Bioactive 614 Agent for Gastrointestinal Health. Molecules. 2022;27(3). Epub 2022/02/16. doi: 615 10.3390/molecules27030794. PubMed PMID: 35164058; PubMed Central PMCID: PMC8838155. 616 57. Jaafaru MS, Abd Karim NA, Enas ME, Rollin P, Mazzon E, Abdull Razis AF. Protective 617 Effect of Glucosinolates Hydrolytic Products in Neurodegenerative Diseases (NDDs). Nutrients. 618 2018;10(5). Epub 2018/05/09. doi: 10.3390/nu10050580. PubMed PMID: 29738500; PubMed 619 Central PMCID: PMC5986460. 620 58. Asif M, Kala C, Gilani S, Imam S, Taleuzzaman M, Naaz F, et al. Protective Effects Of 621 Isothiocyanates Against Alzheimer's Disease. Current Traditional Medicine. 2021;07. doi: 622 10.2174/2215083807666211109121345. 623 59. Ross IA. Neurodegenerative Diseases. In: Ross IA, editor. Plant-Based Therapeutics, Volume 624 2: The Brassicaceae Family. Cham: Springer Nature Switzerland; 2024. p. 261-314. 625 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 28, 2025. ; https://doi.org/10.1101/2025.04.25.650586doi: bioRxiv preprint