Table 1.
Genome-Wide Association Studies main features.
Figure 1.
Genotype imputation was carried out in the Tgen1 and NIA-LOAD/NCRAD datasets (asterisk). The meta-analysis was conducted using the inverse variance method in PLINK, after pruning bad genotypes and samples using standard quality control (QC) tests. The ADNI dataset was included for replication following similar QC procedures at this step. Meta-analysis (dark grey arrows) and ADNI associations (light grey arrows) results were annotated and single gene p-values calculated using the Gene-Wise (GW) method or the more stringent GATES procedure (threshold p<0.05). We next introduced this information to FPAN (from STRING database). Module search was performed 10 times, side by side with the permuted data and without genome-wide-significant (WGW) results, which served as internal controls. Significant sub-networks (white squares) were compared and assayed for gene ontology (GO) term and KEGG pathways enrichment to obtain the final overrepresented pathways associated with AD, inside each sub-network. Equal results between Meta and WGW analysis (“ = ”) that could not be obtained with the permuted control (“≠”) were expected.
Figure 2.
Genome wide meta-analysis results in AD.
Manhattan plot showing the p-values obtained in the meta-analysis. The end and beginning of a chromosome is denoted by the change of color pattern of the SNPs (black, grey and brown dots). Genome-wide significance threshold is denoted by a red line (5.0×10−8). The Y-axis has been truncated to show all associated SNPs inside the APOE loci and to improve visualization of suggestive associations.
Table 2.
Meta-analysis 25 top hits.
Figure 3.
(A) The number of significant SNs (size <50 and score>3) in Meta-GW (light green) and Meta-GATES (dark green) is shown compared with same values permuted across the FPAN: Meta-GW-Permuted (light grey) and Meta-GATES-Permuted (dark grey). (B) Score comparison of the top 10 SNs obtained in the corresponding module searches presented in (A). (C) The number of significant SNs in the replication step for ADNI-GW (light blue) and ADNI-GATES (dark blue) analysis, in comparison with their corresponding permuted controls: ADNI-GW-Permuted (light grey) and ADNI-GW-Permuted (dark grey). (D) Score comparison of the top 10 SNs obtained for each module searches presented in (C). Caped bar/points denote SD; Significant differences between real and permuted data observed in GW and GATES analysis are denoted by an asterisk and those between real and permuted data observed only in GW analysis are denoted by a plus sign (two-sided Student's t-test; p<0.01).
Table 3.
GO terms enriched in Meta-GW and Meta-GATES top 3 Sub-Networks.
Table 4.
Glutamate positive sub-networks in ADNI-GW analysis.
Figure 4.
Structure and relationships between glutamate signaling SNs overrepresented in AD.
SN gene composition (nodes) and interactions (edges) are shown for: Meta-SN1 in the upper left corner with green edges (which includes GW and GATES modules, GNAZ* gene only present in GW); ADNI-GW SN3 in the bottom left corner with dark blue edges; ADNI-GW SN4 in the bottom right corner with blue edges and ADNI-GW SN7 in the upper right corner with light blue edges. Genes shared by at least 2 SNs are located at the center in bold font and cross interactions between genes inside each module are denoted by light grey edges. Node color represents the OR behavior in a gradient from green to red values (i.e. green: OR<1; red OR>1; white: OR = 1), denoting protection and risk, respectively. Similarly, node size is proportional to the –log10 p-value obtained from the meta-analysis (if absent, node size is the minimum). Triangle shaped nodes marks genes belonging to the glutamate signaling pathway GO term (GO:0007215); Diamond shaped nodes denotes genes belonging to KEGG Glutamatergic synapse pathway (hsa04724); Squares Square shaped nodes denotes genes belonging to both gene ontology term GO:0007215 and KEGG hsa04724 pathway.
Figure 5.
Overrepresented glutamate signaling components in the functional synapse.
The original version of the hsa04724 KEGG pathway (Glutamatergic synapse) is shown. Black arrows denote direct molecular interaction or relation between gene products (green squares) or other types of molecules (unfilled circles), while black arrows with dashed lines denote an indirect effect between the each node. The relationship with other KEGG pathways is shown with the presence of white round rectangles. Gene symbols in components belonging to Meta-GW and META-GATES SN1 are denoted in red.
Figure 6.
Gene expression analysis of glutamate signaling components in selected human brain regions.
Heatmap and dendrogram of normalized expression levels of the 20 genes of interest displaying significant clustering in: (A) hippocampal formation (HIF); (B) hypothalamus (HY); (C) Dorsal Thalamus (DT); and (D) white matter (WM). Heatmaps were generated using normalized Z score gene-wise expression values, which were averaged from 6 brain donor individuals (ids. H0351.2001, H0351.2002, H0351.1009, H0351.1012, H0351.1015 and H0351.1016). Bright red and green color indicates high (Z>2) and low expression (Z<2). Highly correlated gene clusters (Euclidean distance correlation coefficient r>0.7) are denoted by colored lines in the dendrograms: green clusters, indicates low expression patterns; red clusters show high levels of expression of correlated genes. Gene expression patterns in the corresponding substructures are shown for HIF: Dentate Gyrus (DG); Cornu Ammonis 1 (CA1); Cornu Ammonis 2 (CA2); Cornu Ammonis 3 (CA3); Cornu Ammonis 4 (CA4) and Subiculum (S). For HY: Anterior Hypothalamic Area (AHA); Lateral hypothalamic Area (LHA); Paraventricular Nucleus of the Hypothalamus (PVH); Supraoptic Nucleus (SO); Lateral Hypothalamic Area, Mammillary Region (LHM); Mammillary Body (MB); Posterior Hypothalamic Area (PHA); Supramammillary Nucleus (SuM); Tuberomammillary Nucleus (TM); Preoptic Region (PrOR); Arcuate Nucleus of the Hypothalamus (ARH); Dorsomedial Hypothalamic Nucleus (DMH); Lateral Hypothalamic Area, Tuberal Region (LHT); Lateral Tuberal Nucleus (LTu); Perifornical Nucleus (PeF); Ventromedial Hypothalamic Nucleus (VMH). For DT: Anterior Group of Nuclei (DTA); Caudal Group of intralaminar Nuclei (ILc); Dorsal Lateral Geneiculate Nucleus (LGd); Lateral Group of Nuclei, Dorsal Division (DTLd); Lateral Group of Nuclei, Ventral Division (DTLv); Medial Geniculate Complex (MG); Medial Group of Nuclei (DTM); Posterior Group of Nuclei (DTP); Rostral Group of Intralaminar Nuclei (ILr). For WM: Cc: Corpus callosum; Cgb: Cingulum bundle.