Integrative brain transcriptome analysis links complement component 4 and HSPA2 to the APOE ε2 protective effect in Alzheimer disease

Mechanisms underlying the protective effect of apolipoprotein E (APOE) ε2 against Alzheimer disease (AD) are not well understood. We analyzed gene expression data derived from autopsied brains donated by 982 individuals including 135 APOE ɛ2/ɛ3 carriers. Complement pathway genes C4A and C4B were among the most significantly differentially expressed genes between ɛ2/ɛ3 AD cases and controls. We also identified an APOE ε2/ε3 AD-specific co-expression network enriched for astrocytes, oligodendrocytes and oligodendrocyte progenitor cells containing the genes C4A, C4B, and HSPA2. These genes were significantly associated with the ratio of phosphorylated tau at position 231 to total Tau but not with amyloid-β 42 level, suggesting this APOE ɛ2 related co-expression network may primarily be involved with tau pathology. HSPA2 expression was oligodendrocyte-specific and significantly associated with C4B protein. Our findings provide the first evidence of a crucial role of the complement pathway in the protective effect of APOE ε2 for AD.

Plaque (CERAD) scores reflect the density of neuritic plaques: 1 (definite), 2 (probable), 3 (possible, and 4 (none); Braak stages represent the spatial distribution of neurofibrillary tangles: 0 (none), I-II (transentorhinal and entorhinal cortex), III-IV (hippocampal and neighboring limbic areas), and V-VI (neocortical areas). All traits were rank-transformed for age of death and sex. Expression differences (β estimate represent the magnitude of effect) for each trait were calculated using a linear regression model separately in each dataset and the results were combined by meta-analysis.
Supplementary Table 8. Association of expression of top-ranked differentially expressed genes among APOE ε2/ε3 subjects with neurofibrillary tangles (Braak stage) and neuritic plaques (plaque score) in the ROSMAP dataset Expression differences (β estimate represent the magnitude of effect) for each trait were calculated using a linear regression model separately in each dataset and the results were combined by meta-analysis.  Expression of genes listed were nominally associated (P<0.05) in both traits in the meta-analysis. Expression differences (β estimate represent the magnitude of effect) for each trait were calculated using a linear regression model separately in each dataset and the results were combined by meta-analysis. * Genes with low expression (mean < 2) in oligodendrocytes, oligodendrocyte progenitor cells, and astrocytes were excluded.

Supplementary
1 Log2 fold change of expression levels (LogFC) and P value (P) were calculated between AD and control brains from single nuclei RNA sequencing data from ROSMAP.
Ast: astrocyte; Oli: oligodendrocytes; OPC: oligodendrocyte progenitor cells; NA: not available due to small number of single cells showing expression (average expression < 2).
P values significant after false discovery rate (FDR) adjustment are highlighted in bold.  Table S1 and Supplemental Table S11.

Supplementary Figure 2. Volcano Plots for Differential Gene Expression Analysis in the ROSMAP Sample
Among Subjects with APOE Genotypes ɛ2/ɛ3 (panel A), ɛ3/ɛ3 (panel B) and ɛ3/ɛ4 (panel C), and in the Total Sample (panel D). Measure of differential expression between AD cases and controls expressed as Log2 fold change is shown on the x-axis and significance level expressed as the -log P-value is shown on the y-axis. Results for each gene are represented by colored dots and are restricted to those with | log2 fold-change | > 0.3 or p-value < 0.01. Genes that were surpassed the p-value and log2 fold-change thresholds are shown in red, genes that surpassed the p-value threshold only are shown in blue, genes that surpassed the log2 fold-change threshold only are shown in green, and genes that did not surpass either threshold are shown in grey.

Supplementary Figure 3. Volcano Plots for Differential Gene Expression Analysis in the MAYO Sample Among Subjects with APOE Genotypes ɛ2/ɛ3 (panel A), ɛ3/ɛ3 (panel B) and ɛ3/ɛ4 (panel C), and in the Total Sample (panel D).
Measure of differential expression between AD cases and controls expressed as Log2 fold change is shown on the x-axis and significance level expressed as the -log P-value is shown on the y-axis. Results for each gene are represented by colored dots and are restricted to those with | log2 fold-change | > 0.3 or p-value < 0.01. Genes that were surpassed the p-value and log2 fold-change thresholds are shown in red, genes that surpassed the p-value threshold only are shown in blue, genes that surpassed the log2 fold-change threshold only are shown in green, and genes that did not surpass either threshold are shown in grey.

Supplementary Figure 4. Volcano Plots for Differential Gene Expression Analysis in the FHS/BUADC Sample Among Subjects with APOE Genotypes ɛ2/ɛ3 (panel A), ɛ3/ɛ3 (panel B) and ɛ3/ɛ4 (panel C), and in the Total Sample (panel D).
Measure of differential expression between AD cases and controls expressed as Log2 fold change is shown on the x-axis and significance level expressed as the -log P-value is shown on the yaxis. Results for each gene are represented by colored dots and are restricted to those with | log2 fold-change | > 0.3 or p-value < 0.01. Genes that were surpassed the p-value and log2 fold-change thresholds are shown in red, genes that surpassed the p-value threshold only are shown in blue, genes that surpassed the log2 fold-change threshold only are shown in green, and genes that did not surpass either threshold are shown in grey. Cell-type characterization of brain tissue obtained from 24 AD cases and 24 controls in the ROSMAP Study. Plot shows a two-dimensional t-distributed stochastic neighbor embedding (t-SNE) projection of all annotated cells derived from single nuclei RNA sequencing (snRNA-seq) data.