Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 Jul 23;28(4):1103-1116.e4.
doi: 10.1016/j.celrep.2019.06.073.

Integrating Gene and Protein Expression Reveals Perturbed Functional Networks in Alzheimer's Disease

Saranya Canchi  1 , Balaji Raao  2 , Deborah Masliah  2 , Sara Brin Rosenthal  3 , Roman Sasik  3 , Kathleen M Fisch  3 , Philip L De Jager  4 , David A Bennett  5 , Robert A Rissman  6
Affiliations

Integrating Gene and Protein Expression Reveals Perturbed Functional Networks in Alzheimer's Disease

Saranya Canchi et al. Cell Rep. .

Abstract

Asymptomatic and symptomatic Alzheimer's disease (AD) subjects may present with equivalent neuropathological burdens but have significantly different antemortem cognitive decline rates. Using the transcriptome as a proxy for functional state, we selected 414 expression profiles of symptomatic AD subjects and age-matched non-demented controls from a community-based neuropathological study. By combining brain tissue-specific protein interactomes with gene networks, we identified functionally distinct composite clusters of genes that reveal extensive changes in expression levels in AD. Global expression for clusters broadly corresponding to synaptic transmission, metabolism, cell cycle, survival, and immune response were downregulated, while the upregulated cluster included largely uncharacterized processes. We propose that loss of EGR3 regulation mediates synaptic deficits by targeting the synaptic vesicle cycle. Our results highlight the utility of integrating protein interactions with gene perturbations to generate a comprehensive framework for characterizing alterations in the molecular network as applied to AD.

Keywords: Alzheimer’s disease; Louvain algorithm; clustering; gene-protein networks; network analysis; protein-protein interaction; transcriptional regulators; transcriptome.

PubMed Disclaimer

Conflict of interest statement

DECLARATION OF INTERESTS

The authors declare no competing interests.

Figures

Figure 1.
Figure 1.. Global Gene Expression Patterns Highlight Differences in Alzheimer’s Disease (AD) Relative to Non-demented Controls (NDCs)
(A) Heatmap of all differential expressed genes (N = 1,722) clustered by phenotype In the prefrontal cortex. (B) Most differentially expressed mitochondrial genes are downregulated In AD, but a small proportion are upregulated relative to NDCs. (C) Nuclear encoded oxidative phosphorylation (OXPHOS) genes are dysregulated in AD.
Figure 2.
Figure 2.. Composite Gene-Protein Interaction Network Reveals Significant Clustering in AD
(A) AD subnetwork with cluster ID mapped to node color. The subnetwork is composed of genes differentially expressed in AD compared with NDCs and overlaid onto the GIANT brain-specific interactome. Node positions are fully determined from the spring layout. Clusters are determined using the Louvain modularity maximization algorithm. (B) Distributions of log fold change values in each cluster, as well as in the entire AD subnetwork. Distribution curves were determined using a kernel-density estimation function (see also Figure S1). (C) AD subnetwork with node positions biased by cluster membership. Node colors are mapped to the log fold change between AD and NDC. Node size represents the significance (FDR) of the gene’s differential expression in AD compared with NDC.
Figure 3.
Figure 3.. Heatmaps of Top Enriched Pathways from Each Functionally Distinct Cluster
(A–D) Enriched pathways correspond to (A) neurotransmitter-related signaling in cluster 0, (B) DNA repair and cell-cycle control in cluster 1, (C) immune response in cluster 2, and (D) metabolism and bioenergetics in cluster 4. No enriched pathways were identified for cluster 3 (see also Figures S2–S6).
Figure 4.
Figure 4.. Transcriptional Regulatory Network in AD
Four candidate transcription factors (TFs) identified in AD, along with all differentially expressed targets in the AD subnetwork, are displayed in a circular layout, ordered by cluster membership. TFs are displayed inside the circle as triangles. The log fold change between AD and NDC is mapped to the node color. BPTF does not belong to any cluster. The TF subnetwork highlights the connections between TFs and targets, along with connections between targets (see also Figure S7).
Figure 5.
Figure 5.. Protein Expression of TGIF and EGR1 in Human Prefrontal Cortex Tissue Samples
(A) Representative images of prefrontal cortex from NDCs and AD brains immunohistochemically stained for TGIF1 and EGR1, which is a direct target of EGR3. (B and C) Bar graphs of relative area fraction for (B) EGR1-stained cells and (C) TGIF1-stained cells are consistent with the RNA-level expression changes. Data are represented as mean ± SD for n = 5 per group. Each tissue section contributed to an average of 20 image fields. Scale bar is 50 μm. *p < 0.05, **p < 0.01.
Figure 6.
Figure 6.. Loss of EGR3 Regulation Mediates Synaptic Deficits by Targeting the Synaptic Vesicle Cycle
Genes names in blue imply downregulation in AD relative to NDCs, and direct targets of EGR3 are shown in bold. Reduced proton gradient (V-ATPASE) results in inefficient neurotransmitter uptake. Downregulation of key essential proteins, including SNAP25, imply reduced vesicle docking at the presynaptic membrane. Inefficient disassociation of actively fused vesicles, along with reduced clathrin (CLTC), leads to decreased vesicle recycling. This creates a negative feedback loop and, after multiple cycles, leads to eventual depletion in reserve vesicles. Downregulation of EGR3 results in reduced content and number of vesicle fusion activities, priming the cells toward ineffective synaptic transmission.
See this image and copyright information in PMC

References

    1. Adlard PA, Tran BA, Finkelstein DI, Desmond PM, Johnston LA, Bush AI, and Egan GF (2014). A review of β-amyloid neuroimaging in Alzheimer’s disease. Front. Neurosci 8, 327. - PMC - PubMed
    1. Alazami AM, Patel N, Shamseldin HE, Anazi S, Al-Dosari MS, Alzahrani F, Hijazi H, Alshammari M, Aldahmesh MA, Salih MA, et al. (2015). Accelerating novel candidate gene discovery in neurogenetic disorders via whole-exome sequencing of prescreened multiplex consanguineous families. Cell Rep. 10, 148–161. - PubMed
    1. An H-J, Cho G, Lee J-O, Paik S-G, Kim YS, and Lee H (2013). Higd1a interacts with Opa1 and is required for the morphological and functional integrity of mitochondria. Proc. Natl. Acad. Sci. USA 110, 13014–13019. - PMC - PubMed
    1. Atamna H, and Frey WH 2nd. (2007). Mechanisms of mitochondrial dysfunction and energy deficiency in Alzheimer’s disease. Mitochondrion 7, 297–310. - PubMed
    1. Bennet AM, Reynolds CA, Eriksson UK, Hong MG, Blennow K, Gatz M, Alexeyenko A, Pedersen NL, and Prince JA (2011). Genetic association of sequence variants near AGER/NOTCH4 and dementia. J. Alzheimers Dis 24, 475–84. - PMC - PubMed

Publication types

LinkOut - more resources