Skip to main content
2025 BMES Annual Meeting Main banner
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.
ViewAttendees 2

Bioinformatics, Computational and Systems Biology

Bioinformatics, Computational and Systems Biology - Poster Session C & D

S32 - Integrative multiomics reveals mechanistic sex differences in late-onset sporadic Alzheimer’s disease

Friday, October 10, 2025
10:00 AM - 11:00 AM PDT
Location: Exhibit Hall F, G & H

Introduction: : Late-onset sporadic Alzheimer’s disease (LOSAD) is a highly heterogeneous disease caused by heritability in around 60-80% of all AD cases from different variants such as clusterin (CLU), triggering receptor expressed on myeloid cells 2 (TREM2) and apolipoprotein E (APOE) [1]. It is usually characterized by progressive neurological decline, memory loss as well as endpoint clinical neuropathological changes such as Aβ plaques, tau tangles and cognitive dysfunction [2]. Previous studies in large cohorts of postmortem LOSAD patient brains have provided some insights into the disease progression as well as molecular subtypes of LOSAD related to neuroinflammation, synaptic loss related to signaling and function, immune activity, mitochondria organization and then remodeling of neuronal lineage and the chromatin landscape. Hallmark endotypes in one-third of LOSAD patients were also discovered such as increased inflammation and decreased synaptic signaling [1,3]. Although this is the case, there have been very limited studies looking at the mechanistic sex differences of LOSAD. To better understand this, large cohorts of postmortem LOSAD patient-derived brains, particularly in the dorosolateral prefrontal cortex (DLPFC), a neocortical region that is the center of cognitive circuitry [4] have been used to study whether endotypes (i.e. disease subtypes) do arise in female and male patients.

Materials and

Methods: : We used a publicly available dataset from the Synapse database derived from postmortem patient-derived brains after patient death; we specifically looked at the dorsolateral prefrontal cortex (DLPFC) which primarily controls working memory and cognitive flexibility [5] that are mainly impaired during LOSAD progression from two aging studies: The Religious Orders Study (ROS) and the Memory and Aging Project (MAP) coming from the same team of investigators at the Rush Alzheimer’s Disease Center (RADC), combined as the ROSMAP study. RNA-seq and ChIP-seq for LOSAD patients were carried out to characterize the gene expression and histone modification H3K9ac changes, respectively. Assignment of new cluster LOSAD groups using an unsupervised machine learning approach called agglomerative hierarchical clustering (AGNES) via hcut [6], differential gene expression and enrichment analysis using limma [7] and fast gene set enrichment analysis (fGSEA) [8] respectively, co-expression module detection using CEMitool [9] coupled with hypergeometric enrichment of module genes using tmodHGtest [10] and then pathway and TF activity identification using decoupleR [11] have been performed for the RNA-seq data. Differential peak region analysis, TF footprinting and functional enrichment in promoter-only regions are being performed for the ChIP-seq data using csaw [12] with TMM normalization from edgeR [15], quantile normalization and differential analysis from limma [7], TF footprinting using HINT [16, 17] and then enrichment from ChIPenrich [13]. Finally, estimation of differential TF activity as either repressors or activators using a tool called diffTF [14] for LOSAD will be performed after integrating histone acetylation data (ChIP-seq) with gene expression data (RNA-seq).

Results, Conclusions, and Discussions:: Our approach shows findings in identifying endotypes and their associated transcriptional regulators for four diseased female and male patient-derived cluster groups compared to healthy subjects using gene expression data; repression of neuron lineage (REST, NFE2L1), cell cycle (E2F1, MYC), non-ectoderm lineage (TCF3), mitochondrial energy production (NRF1), synaptic function (CREB1, EGR1) and estrogen signaling (ESR1) except for neuroinflammation (IRF1, STAT3) were found in LOSAD females whereas a modest activation of similar endotypes were found except for neuron lineage, non-ectoderm lineage, and synaptic function for LOSAD males. Other findings related to differential peak binding analysis via histone modification H3K9ac changes and the correlation between gene expression and histone modification for key disease LOSAD endotypes with their regulators will be further discussed. We have demonstrated so far that LOSAD could be severely impacted more due to the severity of gene expression changes observed in postmortem patient-derived brains of females compared to males where gene expression changes are deemed more modest instead. Furthermore, we can mechanistically propose that since there could be decreased estrogen related activity happening more in the females, there could be no protection against cellular senescence (i.e. a state of irreversible cell-cycle arrest), thus cell cycling is disrupted. This in turn can induce mitochondrial dysfunction and promote accumulation of amyloid-β plaques, which elevates neuroinflammation more in the female compared to male LOSAD brain. These preliminary mechanistic findings will be further validated with both histone acetylation and multiomics data integration analyses and discussed.

Acknowledgements and/or References (Optional): : Acknowledgements: We would like to thank the numerous participants of the ROS and MAP studies for their generous donations of their brains. In addition, we thank Dr. Shakti Gupta (UC San Diego) for his assistance with calculating the top variable genes as well as Dr. Philip De Jager (Columbia University) and Dr. Hans-Ulrich Klein (Columbia University) for their assistance with bulk ChIP-seq data processing.

References:
[1] Neff RA, Wang M, Vatansever S, Guo L, Ming C, Wang Q, et al. Molecular subtyping of Alzheimer’s disease using RNA sequencing data reveals novel mechanisms and targets. Science Advances. 7:eabb5398.

[2] Caldwell AB, Anantharaman BG, Ramachandran S, Nguyen P, Liu Q, Trinh I, et al. Transcriptomic profiling of sporadic Alzheimer’s disease patients. Molecular Brain. 2022;15:83.

[3] Caldwell AB, Liu Q, Zhang C, Schroth GP, Galasko DR, Rynearson KD, et al. Endotype reversal as a novel strategy for screening drugs targeting familial Alzheimer’s disease. Alzheimers Dement. 2022

[4] De Jager PL, Ma Y, McCabe C, Xu J, Vardarajan BN, Felsky D, et al. A multi-omic atlas of the human frontal cortex for aging and Alzheimer’s disease research. Sci Data. 2018;5:180142.

[5] Kaplan JT, Gimbel SI, Harris S. Neural correlates of maintaining one’s political beliefs in the face of counterevidence. Sci Rep. 2016;6:39589.

[6] hcut function - RDocumentation [Internet]. [cited 2025 Jul 28]. Available from: https://www.rdocumentation.org/packages/factoextra/versions/1.0.7/topics/hcut

[7] Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Research. 2015;43:e47.

[8] Korotkevich G, Sukhov V, Budin N, Shpak B, Artyomov MN, Sergushichev A. Fast gene set enrichment analysis [Internet]. 2021 Feb p. 060012. Available from: https://www.biorxiv.org/content/10.1101/060012v3

[9] Russo PST, Ferreira GR, Cardozo LE, Bürger MC, Arias-Carrasco R, Maruyama SR, et al. CEMiTool: a Bioconductor package for performing comprehensive modular co-expression analyses. BMC Bioinformatics. 2018;19:56.

[10] Zyla J, Marczyk M, Domaszewska T, Kaufmann SHE, Polanska J, Weiner J 3rd. Gene set enrichment for reproducible science: comparison of CERNO and eight other algorithms. Bioinformatics. 2019;35:5146–54.

[11] Badia-i-Mompel P, Vélez Santiago J, Braunger J, Geiss C, Dimitrov D, Müller-Dott S, et al. decoupleR: ensemble of computational methods to infer biological activities from omics data. Bioinformatics Advances. 2022;2:vbac016.

[12] Reske JJ, Wilson MR, Chandler RL. ATAC-seq normalization method can significantly affect differential accessibility analysis and interpretation. Epigenetics & Chromatin. 2020;13:22.

[13] Welch RP, Lee C, Imbriano PM, Patil S, Weymouth TE, Smith RA, et al. ChIP-Enrich: gene set enrichment testing for ChIP-seq data. Nucleic Acids Res. 2014;42:e105.

[14] Berest I, Arnold C, Reyes-Palomares A, Palla G, Dindler-Rasmussen K, Giles H, et al. Quantification of differential transcription factor activity and multiomics-based classification into activators and repressors: diffTF. Cell Reports. 2019;29:3147–59. 103.

[15] Robinson MD, McCarthy DJ, Smyth GK. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010;26:139–40.

[16] E. G. Gusmao, M. Allhoff, M. Zenke, I. G. Costa, Analysis of computational footprinting methods for DNase sequencing experiments. Nat. Methods 13, 303–309 (2016).

[17] Z. Li, M. H. Schulz, T. Look, M. Begemann, M. Zenke, I. G. Costa, Identification of transcription factor binding sites using ATAC-seq. Genome Biol. 20, 45 (2019).