A Bioinformatics Approach Toward Unravelling the Synaptic Molecular Crosstalk Between Alzheimer's Disease and Diabetes

Abstract

Background: Increasing evidence links impaired brain insulin signaling and insulin resistance to the development of Alzheimer's disease (AD).

Objective: This evidence prompted a search for molecular players common to AD and diabetes mellitus (DM).

Methods: The work incorporated studies based on a primary care-based cohort (pcb-Cohort) and a bioinformatics analysis to identify central nodes, that are key players in AD and insulin signaling (IS) pathways. The interactome for each of these key proteins was retrieved and network maps were developed for AD and IS. Synaptic enrichment was performed to reveal synaptic common hubs.

Results: Cohort analysis showed that individuals with DM exhibited a correlation with poor performance in the Mini-Mental State Examination (MMSE) cognitive test. Additionally, APOE ɛ2 allele carriers appear to potentially be relatively more protected against both DM and cognitive deficits. Ten clusters were identified in this network and 32 key synaptic proteins were common to AD and IS. Given the relevance of signaling pathways, another network was constructed focusing on protein kinases and protein phosphatases, and the top 6 kinase nodes (LRRK2, GSK3B, AKT1, EGFR, MAPK1, and FYN) were further analyzed.

Conclusion: This allowed the elaboration of signaling cascades directly impacting AβPP and tau, whereby distinct signaling pathway play a major role and strengthen an AD-IS link at a molecular level.

Keywords: Alzheimer’s disease; apolipoprotein E; insulin; leucine-rich repeat serine-threonine protein kinase-2; type 2 diabetes mellitus.

Fig. 5

Signaling pathways towards AβPP phosphorylation. Graphical view of signaling pathways of the top 6 proteins with higher betweenness centrality values from Fig. 4 towards AβPP using the kinector tool. 1) LRRK2 signaling pathway towards AβPP. 2) GSK3B signaling pathway towards AβPP. 3) AKT1 signaling pathway towards AβPP. 4) EGFR signaling pathway towards AβPP. 5) ERK2 signaling pathway towards APP. 6) FYN signaling pathway towards AβPP.

Fig. 6

Signaling pathways towards tau phosphorylation. Graphical view of signaling pathways of the top 6 proteins with higher betweenness centrality values from Fig. 4 towards tau using the kinector tool. 1) LRRK2 signaling pathway towards tau. 2) GSK3B signaling pathway towards tau. 3) AKT1 signaling pathway towards tau. 4) EGFR signaling pathway towards tau. 5) ERK2 signaling pathway towards tau. 6) FYN signaling pathway towards tau.

Fig. 1

Synaptic AD/IS coincident network. Representation of all proteins retrieved from the interception between Alzheimer’s disease (AD), insulin signaling (IS), and the synapse. Community cluster (GLay) from ClusterMaker 2.0 was used to create the communities. Node size increases with higher values of betweenness centrality. A) Each community/cluster is highlighted in a different color. Key AD proteins are marked with a black circumference, key IS proteins with a dark grey circumference, and nodes common to both with a red circumference. B) Network with kinases filled in red, catalytic subunits of protein phosphatases filled in green and genetic risk factors for AD marked with a black circumference. Clusters are numbered left to right, top panel (clusters 1 to 4) and bottom panel (clusters 5 to 10), several proteins (102) are not integrated into clusters.

Fig. 2

Cluster analysis. A) Absolute number of nodes with characteristics as indicated for each cluster. B) Different node characteristics represented as a percentage of nodes within each cluster and as a total of the synaptic AD and IS coincident network. Cluster 9 and 10 were excluded since no relevant nodes were present. OC, outside of the clusters.

Fig. 3

Gene ontology analyses. Graphical representation of the distribution of the genes in the ‘Synaptic AD/IS coincident network’ with respect to cellular component, biological process, and molecular function. Top 12 hits are indicated.

Fig. 4

AD/IS synaptic kinase and phosphatase network. Color code obeys the communities presented in Fig. 1A. Node size increases with higher values of betweenness centrality. Kinases are highlighted with a dark red border and catalytic subunits of protein phosphatases are highlighted with a green border.