Predictive Network Analysis Identifies JMJD6 and Other Novel Key Drivers in Alzheimer Disease
Merchant, J. P.; Zhu, K.; Henrion, M. Y. R.; Zaidi, S. S. A.; Branden, L.; Moein, S.; Alamprese, M. L.; Pearse, R. V.; Bennett, D. A.; Ertekin-Taner, N.; Young-Pearse, T. L.; Chang, R.
Show abstract
Despite decades of genetic studies on late onset Alzheimers disease (LOAD), the molecular mechanisms of Alzheimers disease (AD) remain unclear. Furthermore, different cell types in the central nervous system (CNS) play distinct roles in the onset and progression of AD pathology. To better comprehend the complex etiology of AD, we used an integrative approach to build robust predictive (causal) network models which were cross-validated over multiple large human multi-omics datasets in AD. We employed a published method to delineate bulk-tissue gene expression into single cell-type gene expression and integrated clinical and pathologic traits of AD, single nucleotide variation, and deconvoluted gene expression for the construction of predictive network models for each cell type in AD. With these predictive causal models, we are able to identify and prioritize robust key drivers of the AD-associated network state. In this study, we focused on neuron-specific network models and prioritized 19 predicted key drivers modulating AD pathology. These targets were validated via shRNA knockdown in human induced pluripotent stem cell (iPSC) derived neurons (iNs), in which 10 out of the 19 neuron-related targets (JMJD6, NSF, NUDT2, YWHAZ, RBM4, DCAF12, NDRG4, STXBP1, ATP1B1, and FIBP) significantly modulated levels of amyloid-beta and/or phosphorylated tau peptides in the postmitotic iNs. Most notably, knockdown of JMJD6 significantly altered the neurotoxic ratios of A{beta}42 to 40 and p231-tau to total tau, indicating its potential therapeutic relevance to both amyloid and tau pathology in AD. Molecular validation by RNA sequencing (RNAseq) in iNs further confirmed the network structure, showing significant enrichment in differentially expressed genes after knockdown of the validated targets. Interestingly, our network model predicts that these 10 key drivers are upstream regulators of REST and VGF, two recently identified key regulators of AD pathogenesis.
Matching journals
The top 6 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- A public resource of single cell transcriptomes and multiscale networks from persons with and without Alzheimer's disease 96%
- Molecular Signatures of Resilience to Alzheimer's Disease in Neocortical Layer 4 Neurons 96%
- Cell-type-specific Alzheimer’s disease polygenic risk scores are associated with distinct disease processes in Alzheimer’s disease 96%
Similar papers in this journal
- Natural genetic variation determines microglia heterogeneity in wild-derived mouse models of Alzheimer's disease 95%
- Applying high-resolution spatial transcriptomics to characterise the amyloid plaque cell niche in Alzheimer's Disease 94%
- Sustained TREM2 stabilization accelerates microglia heterogeneity and Abeta pathology in a mouse model of Alzheimer s disease 94%
Similar papers in this journal
- Genome-wide consensus transcriptional signatures identify synaptic pruning linking Alzheimer's disease and epilepsy 96%
- Integrative Brain Transcriptome Analysis Links Complement Component 4 and HSPA2 to the APOE ε2 Protective Effect in Alzheimer Disease 94%
- Alzheimer's Disease Patient Brain Extracts Induce Multiple Pathologies in Vascularized Neuroimmune Organoids for Disease Modeling and Drug Discovery 94%
Similar papers in this journal
- A microglia clonal inflammatory disorder in Alzheimer's Disease 95%
- Distinctive Whole-brain Cell-Types Predict Tissue Damage Patterns in Thirteen Neurodegenerative Conditions 94%
- The gene expression landscape of the human locus coeruleus revealed by single-nucleus and spatially-resolved transcriptomics 93%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.