An integrative systems-biology approach defines mechanisms of Alzheimer's disease neurodegeneration

Abstract

Despite years of intense investigation, the mechanisms underlying neuronal death in Alzheimer's disease, remain incompletely understood. To define relevant pathways, we conducted an unbiased, genome-scale forward genetic screen for age-associated neurodegeneration in Drosophila. We also measured proteomics, phosphoproteomics, and metabolomics in Drosophila models of Alzheimer's disease and identified Alzheimer's genetic variants that modify gene expression in disease-vulnerable neurons in humans. We then used a network model to integrate these data with previously published Alzheimer's disease proteomics, lipidomics and genomics. Here, we computationally predict and experimentally confirm how HNRNPA2B1 and MEPCE enhance toxicity of the tau protein, a pathological feature of Alzheimer's disease. Furthermore, we demonstrated that the screen hits CSNK2A1 and NOTCH1 regulate DNA damage in Drosophila and human stem cell-derived neural progenitor cells. Our study identifies candidate pathways that could be targeted to ameliorate neurodegeneration in Alzheimer's disease.

Fig. 1. Overview of analytical framework for multi-omic integration to study the biological processes underlying neurodegeneration.

We performed a forward genetic screen for age-associated neurodegeneration in Drosophila. We measured proteomics, phosphoproteomics and metabolomics in amyloid β (gold) and tau (purple) models of Alzheimer’s disease and performed an eQTL meta-analysis of previous Alzheimer’s disease studies. We used a network integration model to integrate these new data with previously published human proteomics, human genetics, human lipidomics, and Drosophila modifiers of tau-mediated neurotoxicity. We tested hypotheses inferred from subnetwork analysis of the network model in Drosophila and human iPSC-derived neural progenitor cells. Created in BioRender. Leventhal, M. (2025) https://BioRender.com/y28s838.

Fig. 2. Gene expression of neurodegeneration screen hits declines with chronological age in the human brain.

a Geometric mean expression in transcripts per million (TPM) of neurodegeneration screen hits (neurodegeneration genes, orange) and all protein-coding genes in the Genotype-Tissue Expression (GTEx) shows that the expression of neurodegeneration screen hits declines with age in human brain tissues (all protein-coding genes: two-sided t-test Benjamini-Hochberg FDR-adjusted p = 2.91*10⁻⁴, neurodegeneration screen hits: two-sided t-test Benjamini-Hochberg FDR-adjusted p = 1.14*10^q5). There is a significant difference in the slopes of the trends between age and gene expression for neurodegeneration screen hits and all protein-coding genes (all protein-coding genes: R = 0.12, neurodegeneration screen hits: R = 0.15, p = 7.38*10⁻⁶). Regression lines indicate the relationship between age and TPM with a 95% confidence interval (standard error of the mean). The mixed effects regression analysis controlled for post-mortem interval, sex, ethnicity, and tissue of origin. b Gene set enrichment plot showing that the set of age-associated neurodegeneration genes has reduced expression with respect to age. Vertical lines indicate rank of neurodegeneration screen hits by their association between gene expression and age determined by mixed-effects regression analysis coefficients. c Proportion of genes that have significant associations between gene expression and age relative to the set of all protein-coding genes (blue, n = 20438) or the set of age-associated neurodegeneration genes (orange, n = 163). Data is presented as proportion of differentially expressed genes relative to total with 95% binomial confidence intervals of the estimated proportion of genes with a significant association with age. Asterisk indicates tissues with a FDR-adjusted one-tailed hypergeometric test p-value less than 0.01. Binomial test p-values after FDR correction: Amygdala=7.61*10⁻², Caudate=2.15*10⁻², Anterior Cingulate cortex=5.49*10⁻¹, Nucleus Accumbens=3.32*10⁻², Cerebellar Hemisphere=1.48*10⁻⁴, Cerebellum=1.74*10⁻⁴, Frontal Cortex=3.17*10⁻⁴, Hypothalamus=1.24*10⁻², Hippocampus=1.85*10⁻³, Cortex=1.85*10⁻³, Substantia Nigra=1.30*10⁻¹. d Proportion of protein-coding genes (blue, n = 17926) and age-associated neurodegeneration genes (orange, n = 169) that are differentially expressed between Alzheimer’s disease (AD) and control in excitatory neurons in single-nucleus RNA-seq. Data is presented as proportion of differentially expressed genes relative to total with 95% binomial confidence intervals. Source data for (a) is provided in the source data file.

Fig. 3. Multi-omic changes in human Alzheimer’s disease patients and model systems.

a Schematic depicting the identification of eGenes from laser-capture microdissection of temporal cortex pyramidal neuron-enriched populations from 75 individuals including 42 human Alzheimer’s disease (AD) and 33 healthy control patients and identification of eGenes. b The eQTL associated with the eGene HLA-DRB1 is highlighted in red and overlaps with DNA binding motifs of MEF2B, CUX1 and ATF2 derived from ENCODE ChIP-seq and FIMO-detected motifs. Grey horizontal bars indicate ChIP-seq binding regions and the black horizontal bars indicate where the DNA-binding motif is located. c Dot plots showing the negative log₁₀ FDR-adjusted hypergeometric test p-values for enriched GO terms in proteins that are significantly upregulated or significantly downregulated in both Drosophila models of tau and amyloid β, only differentially abundant in Drosophila models of amyloid β (Amyloid β only), or only differentially abundant in Drosophila models of tau (Tau only). d Heat maps depict the log2 fold changes between Aβ_1-42 transgenic flies (Amyloid β) or tau^R406W (Tau) transgenic flies with controls for (d) proteins or (e) phosphoproteins that were hits in the age-associated neurodegeneration screen. An asterisk indicates whether the comparison was significant at an FDR threshold of 0.1. The columns of all heatmaps were clustered by hierarchical clustering.

Fig. 4. Network integration of Alzheimer’s disease multi-omics and novel genetic screening data identifies subnetworks characterized by hallmarks of neurodegeneration and processes previously not implicated in Alzheimer’s disease.

a Network integration of human and Drosophila multi-omics for Alzheimer’s Disease highlights subnetworks enriched for proteins belonging to known gene ontologies. Each subnetwork is represented by a pie chart, which indicates the proportion of nodes represented by a given data type. Edge width is determined by the number of interactions between nodes within or with another subnetwork and colored by one of the involved subnetworks. Each pie chart is labeled by the enriched biological process by hypergeometric test (FDR-adjusted p-value less than 0.1). b A subnetwork enriched for postsynaptic activity. Nodes belonging to the annotated process are highlighted in yellow. Also in this subnetwork are metabolites associated with postsynaptic activity such as acetylcholine. c Phosphorylated tau, APOE, and APP-processing proteins interact with each other and are in a subnetwork enriched for NOTCH signaling-associated genes. Members of the NOTCH signaling pathway are highlighted in yellow.

Fig. 5. Network integration of Alzheimer’s disease multi-omics and novel genetic screening data reveals biological processes associated with tau-mediated neurotoxicity.

a The proteins coded by the neurodegeneration modifier HNRNPA2B1 and the eGene MEPCE interact with each other and have protein-protein interactions with modifiers of tau neurotoxicity. The interaction between HNRNPA2B1 and MEPCE is found in the subnetwork in Fig. 4 that is enriched for insulin signaling. b Knockdown of the Drosophila orthologs of HNRNPA2B1 (Hrb98DE) and MEPCE (CG1293) shows enhancement of the rough eye phenotype in flies expressing wild type human tau. c Quantification of rough eye severity. The scale reflects the extent of morphological disruption after human tau retinal expression (Methods). Statistical significance was measured using a one-way ANOVA with Tukey’s post-hoc correction and is indicated with an asterisk. Error bars are the standard error of the mean. Two independent RNAi constructs were used to knock down each gene. n = 8. Control is GMR-GAL4/+. Flies are one day old. The boxplots are defined where the line within the box depicts the median, the 25th-75th percentiles are the boundaries of the box and the whiskers depict 1.5 times the interquartile range. d Volcano plot depicting differential expression analysis by DeSeq2 of bulk RNA-seq after HNRNPA2B1 CRISPRi knockdown in NGN2 neural progenitor cells (Benjamini-Hochberg FDR-adjusted two-sided Wald test p-value < 0.1, absolute log₂ fold change > 1). Each dot represents a single gene. The horizontal dashed line indicates the negative log₁₀ FDR-adjusted p-value significance cut-off of 0.1 and the vertical dashed lines indicate the log₂ fold change cut-offs of 1 and −1. Red dots indicate genes that are significantly upregulated and blue dots indicate genes that are significantly downregulated. e Dot plot of the enriched pathways identified by gene set enrichment analysis of the RNA-seq data. The 10 pathways with the highest negative log₁₀ FDR-adjusted Gene Set Enrichment Analysis empirical p-value are plotted. The size of the dot indicates the proportion of genes that are part of the enriched pathway. The color of the dot represents the normalized enrichment score (NES), where blue indicates downregulation and red indicates upregulation. The x-position of the dot indicates the negative log₁₀ FDR-adjusted Gene Set Enrichment Analysis empirical p-value and the y-position is the corresponding, enriched pathway. Source data for (c) is provided in the source data file.

Fig. 6. Network analysis implicates neurodegeneration genes as regulators of the AD-associated biological process of DNA damage repair.

a NOTCH1 and CSNK2A1 interactors (highlighted in yellow) are involved in DNA damage repair. b Knockdown of Drosophila orthologs for NOTCH1(N) and CSNK2A1 (CKIIa and CKIIb lead to increased DNA damage in adult neurons as measured by increased pH2Av-positive foci (red, arrowheads); elav immunostaining (green) marks neurons; nuclei are identified with DAPI (blue). Scale bar = 5 µm. c Percent of pH2Av foci in control and knockdowns of CKIIa, CKIIb and N. Asterisks indicate p < 0.01 with a one-way binomial test after Benjamini-Hochberg FDR correction with 95% binomial confidence intervals (n = 6, N RNAi 1 p = 2.04*10⁻⁶, N RNAi 2 p = 5.87*10⁻¹², CKIIa p = 1.34*10⁻³¹, CKIIb p = 1.42*10⁻²⁹). d COMET assay shows increased DNA damage in human iPSC neural progenitor cells after inhibition of Casein Kinase 2 (CK2) by CX-4945 and inhibition of NOTCH cleavage. e Quantification of mean tail moments from panel A in arbitrary units. Asterisks indicate p < 0.01 by one-way ANOVA with Tukey’s Post-Hoc correction (p = 6.01*10⁻³, Compound E vs DMSO, p = 3.95*10⁻³, CX-4945 vs DMSO). Error bars = +/- SEM (DMSO n = 693, CX-4945 n = 793, Compound E n = 860). f Dot plots showing the normalized enrichment scores (NES) of selected, significantly enriched pathways after CSNK2A1 and NOTCH1 knockdown in NGN2 neural progenitor cells. Red and blue dots indicate positive (upregulation) and negative (downregulation) NES, respectively. g Representative immunofluorescence images of mature neurons in Drosophila brains show inappropriate cell cycle re-entry in postmitotic neurons as indicated by PCNA expression (red, arrow) following CKIIa knockdown. elav (green) identifies neurons. h) Quantification of PCNA expression in control brains and brains of Drosophila with knockdown of orthologs of CSNK2A1 (CKIIa and CKIIb). Asterisks indicate p < 0.01 by one-way ANOVA with Tukey’s Post-Hoc correction (p = 4.94*10⁻⁵, CKIIa vs control, p = 3.92*10⁻⁴, CkIIb vs control). n = 6. Control is elav-GAL4/ + ; UAS-Dcr-2/+. Flies are 10 days old. For the boxplots: the line marks the median, the 25th-75th percentiles are the boundaries of the box and the whiskers depict 1.5 times the interquartile range. The source data for (c, e, h) are provided in the source data file.