Alzheimer's disease: early alterations in brain DNA methylation at ANK1, BIN1, RHBDF2 and other loci

Abstract

We used a collection of 708 prospectively collected autopsied brains to assess the methylation state of the brain's DNA in relation to Alzheimer's disease (AD). We found that the level of methylation at 71 of the 415,848 interrogated CpGs was significantly associated with the burden of AD pathology, including CpGs in the ABCA7 and BIN1 regions, which harbor known AD susceptibility variants. We validated 11 of the differentially methylated regions in an independent set of 117 subjects. Furthermore, we functionally validated these CpG associations and identified the nearby genes whose RNA expression was altered in AD: ANK1, CDH23, DIP2A, RHBDF2, RPL13, SERPINF1 and SERPINF2. Our analyses suggest that these DNA methylation changes may have a role in the onset of AD given that we observed them in presymptomatic subjects and that six of the validated genes connect to a known AD susceptibility gene network.

Figure 1. Summary of the genome-wide brain DNA methylation scan for NP burden and its validation using two sets of brain RNA data

Each sector of this diagram presents summary results of the three different analyses within a chromosome. The perimeter of this circular figure presents the physical position along each chromosome (in Mb). The cytogenetic bands of each chromosome are presented in the first circle, with the centromere highlighted in red. The next circle (green) reports the density of CpG probes successfully sampled by the Illumina beadset that are present in a given genomic segment. The blue circle reports the results of the DNA methylation scan: using a “–log(p-value)” scale, we report the results for each of the 71 associated CpG found in 60 independent differentially methylated regions (DMR) from the analysis relating DNA methylation levels to NP burden. Similarly, the first red circle reports the –log(p-value) for the 71 CpGs in the replication analysis. The large lightly colored circle reports the names of genes found within 50 kb of each associated CpG (light blue letters). The ABCA7 and BIN1 regions, which harbor AD susceptibility alleles, are highlighted in red letters. The subset of the genes with differential mRNA expression in AD in the Mayo clinic dataset are shown in black. The next, red circle reports the results of the association of RNA expression level of these genes to a diagnosis of AD in the Mayo clinic dataset. The central circle reports the set of validated CpGs who also have nearby genes whose expression is altered.

Figure 2. Extent of differences in methylation levels at associated CpGs and regional distribution of associations.

Two of the most AD-associated probes - from the MCF2L (a) and ANK1 (b) differentially methylated regions, Table 1 - are selected to illustrate the increase in methylation levels seen, on average, with a diagnosis of AD in 82% of the CpGs that meet our threshold of significance. Each panel reports data for one CpG, which is listed at the top of the panel. The panel itself is a smoothed histogram presenting the distribution of DNA methylation values at that CpG for subjects classified as having a neuropathologic diagnosis of AD (case, red, n=460) and those subjects that do not meet these diagnostic criteria (control, green, n=263). The scale is truncated at a methylation level of 0.6 since there are no values beyond this point; a methylation value of “1.0” means that all CpG in the sample are methylated. We see that the distribution of AD subjects is statistically significantly different from that of the control subjects. However, the two distributions overlap, and the absolute difference between the two distributions is modest. (c) Regional association plot around cg22883290 in the BIN1 differentially methylated region (DMR) that has previously been associated with AD susceptibility in genome-wide association studies. Each diamond represents on CpG tested in this region; the x-axis reports the physical distance across the region. The y-axis on the left reports the magnitude of the p-value in the analysis relating DNA methylation to NP burden. The horizontal dotted blue line highlights the threshold of significance for this analysis. The vertical blue line reports the density of CpG probes at a given point; values (probes/kb) are reported on the y-axis on the right. The extent to which DNA methylation level at a given CpG correlates with the level of DNA methylation of the best CpG (cg22883290) is reported using the r² value and visualized using a red,high/white, low scale (see upper right corner of each panel). Finally, above the diagram of the genes found in this DNA segment, we report the chromatin state of the region, as assessed in healthy, unimpaired older individuals with minimal AD-related pathology. The chromatin state is derived in 200 bp bins, and the color key is presented in the upper left corner of each panel. Overall, the BIN1 gene appears to be in an open, transcribed conformation in healthy, older dorsolateral prefrontal cortex, and the associated CpG appears to be located in a region just 3’ to the gene, which is largely in a conformation found on the periphery of actively transcribed regions. (d) Regional association plot around the RHBDF2 DMR, centered on cg13076843, which meets our threshold of significance. Here, we have an associated CpG that is found in close proximity to two genes, and our RNA analyses suggest that it is RHBDF2 that is the target of the DMR since its expression is altered relative to AD (Table 2).

Figure 2. Extent of differences in methylation levels at associated CpGs and regional distribution of associations.

Two of the most AD-associated probes - from the MCF2L (a) and ANK1 (b) differentially methylated regions, Table 1 - are selected to illustrate the increase in methylation levels seen, on average, with a diagnosis of AD in 82% of the CpGs that meet our threshold of significance. Each panel reports data for one CpG, which is listed at the top of the panel. The panel itself is a smoothed histogram presenting the distribution of DNA methylation values at that CpG for subjects classified as having a neuropathologic diagnosis of AD (case, red, n=460) and those subjects that do not meet these diagnostic criteria (control, green, n=263). The scale is truncated at a methylation level of 0.6 since there are no values beyond this point; a methylation value of “1.0” means that all CpG in the sample are methylated. We see that the distribution of AD subjects is statistically significantly different from that of the control subjects. However, the two distributions overlap, and the absolute difference between the two distributions is modest. (c) Regional association plot around cg22883290 in the BIN1 differentially methylated region (DMR) that has previously been associated with AD susceptibility in genome-wide association studies. Each diamond represents on CpG tested in this region; the x-axis reports the physical distance across the region. The y-axis on the left reports the magnitude of the p-value in the analysis relating DNA methylation to NP burden. The horizontal dotted blue line highlights the threshold of significance for this analysis. The vertical blue line reports the density of CpG probes at a given point; values (probes/kb) are reported on the y-axis on the right. The extent to which DNA methylation level at a given CpG correlates with the level of DNA methylation of the best CpG (cg22883290) is reported using the r² value and visualized using a red,high/white, low scale (see upper right corner of each panel). Finally, above the diagram of the genes found in this DNA segment, we report the chromatin state of the region, as assessed in healthy, unimpaired older individuals with minimal AD-related pathology. The chromatin state is derived in 200 bp bins, and the color key is presented in the upper left corner of each panel. Overall, the BIN1 gene appears to be in an open, transcribed conformation in healthy, older dorsolateral prefrontal cortex, and the associated CpG appears to be located in a region just 3’ to the gene, which is largely in a conformation found on the periphery of actively transcribed regions. (d) Regional association plot around the RHBDF2 DMR, centered on cg13076843, which meets our threshold of significance. Here, we have an associated CpG that is found in close proximity to two genes, and our RNA analyses suggest that it is RHBDF2 that is the target of the DMR since its expression is altered relative to AD (Table 2).

Figure 2. Extent of differences in methylation levels at associated CpGs and regional distribution of associations.

Two of the most AD-associated probes - from the MCF2L (a) and ANK1 (b) differentially methylated regions, Table 1 - are selected to illustrate the increase in methylation levels seen, on average, with a diagnosis of AD in 82% of the CpGs that meet our threshold of significance. Each panel reports data for one CpG, which is listed at the top of the panel. The panel itself is a smoothed histogram presenting the distribution of DNA methylation values at that CpG for subjects classified as having a neuropathologic diagnosis of AD (case, red, n=460) and those subjects that do not meet these diagnostic criteria (control, green, n=263). The scale is truncated at a methylation level of 0.6 since there are no values beyond this point; a methylation value of “1.0” means that all CpG in the sample are methylated. We see that the distribution of AD subjects is statistically significantly different from that of the control subjects. However, the two distributions overlap, and the absolute difference between the two distributions is modest. (c) Regional association plot around cg22883290 in the BIN1 differentially methylated region (DMR) that has previously been associated with AD susceptibility in genome-wide association studies. Each diamond represents on CpG tested in this region; the x-axis reports the physical distance across the region. The y-axis on the left reports the magnitude of the p-value in the analysis relating DNA methylation to NP burden. The horizontal dotted blue line highlights the threshold of significance for this analysis. The vertical blue line reports the density of CpG probes at a given point; values (probes/kb) are reported on the y-axis on the right. The extent to which DNA methylation level at a given CpG correlates with the level of DNA methylation of the best CpG (cg22883290) is reported using the r² value and visualized using a red,high/white, low scale (see upper right corner of each panel). Finally, above the diagram of the genes found in this DNA segment, we report the chromatin state of the region, as assessed in healthy, unimpaired older individuals with minimal AD-related pathology. The chromatin state is derived in 200 bp bins, and the color key is presented in the upper left corner of each panel. Overall, the BIN1 gene appears to be in an open, transcribed conformation in healthy, older dorsolateral prefrontal cortex, and the associated CpG appears to be located in a region just 3’ to the gene, which is largely in a conformation found on the periphery of actively transcribed regions. (d) Regional association plot around the RHBDF2 DMR, centered on cg13076843, which meets our threshold of significance. Here, we have an associated CpG that is found in close proximity to two genes, and our RNA analyses suggest that it is RHBDF2 that is the target of the DMR since its expression is altered relative to AD (Table 2).

Figure 3. Distribution of CpGs associated (p<1×10⁻⁷) wih NP among 11 chromatin states found in mid-frontal cortex. (A) Chromatin map of the dorsolateral prefrontal cortex

Using data generated by the National Institute of Health's Epigenomic Roadmap effort, we assign each chromosomal segment to one of 11 discrete chromatin states that are listed in a column to the left of the figure. On the X-axis, we list the individual chromatin marks (antigens) targeted in each set of ChiP-Seq data generated from MAP subjects that were cognitively non-impaired at the time of death and have minimal pathology on neuropathological examination. The heatmap (white, low; blue, high) graphically displays the relative abundance of sequences found in a segment of DNA after immunoprecipitation for a particular histone mark. Each chromatin state has a unique complement of histone marks. (B) We use the chromatin map illustrated in panel A to identify the chromatin state within which each of the interrogated CpG dinucleotides are found. The histogram compares the distribution of chromatin states found at those 71 associated CpG dinucleotides whose methylation level is associated with neuritic plaques (Table 1) (pink bars) to the overall distribution of chromatin states found in all 415,848 CpG dinucleotides that were analyzed (blue bars).

Figure 3. Distribution of CpGs associated (p<1×10⁻⁷) wih NP among 11 chromatin states found in mid-frontal cortex. (A) Chromatin map of the dorsolateral prefrontal cortex

Using data generated by the National Institute of Health's Epigenomic Roadmap effort, we assign each chromosomal segment to one of 11 discrete chromatin states that are listed in a column to the left of the figure. On the X-axis, we list the individual chromatin marks (antigens) targeted in each set of ChiP-Seq data generated from MAP subjects that were cognitively non-impaired at the time of death and have minimal pathology on neuropathological examination. The heatmap (white, low; blue, high) graphically displays the relative abundance of sequences found in a segment of DNA after immunoprecipitation for a particular histone mark. Each chromatin state has a unique complement of histone marks. (B) We use the chromatin map illustrated in panel A to identify the chromatin state within which each of the interrogated CpG dinucleotides are found. The histogram compares the distribution of chromatin states found at those 71 associated CpG dinucleotides whose methylation level is associated with neuritic plaques (Table 1) (pink bars) to the overall distribution of chromatin states found in all 415,848 CpG dinucleotides that were analyzed (blue bars).

Figure 4. Genes identified in our DNA methylation screen connect to a network of known AD susceptibility genes

Using protein:protein interaction data, the DAPPLE algorithm evaluated the extent of connectivity among known AD genes (susceptibility and Mendelian genes) and the eight genes found in DMRs that are also differentially expressed relative to AD. The figure displays the results of an analysis allowing for one common interactor protein that is not known to be associated with AD. For example, RHBDF2 is displayed at the top of the figure in green and connects to PTK2B, a protein tyrosine kinase genetically associated with AD susceptibility which has a central role in this network. Interestingly, SERPINF1 and SERPINF2 connect to different elements of the amyloid component of the network (bottom left). Further, DIP2A connects the recently described PLD3 gene that has a rare AD susceptibility allele to SORL1, a gene with a common AD susceptibility allele, that connects to the amyloid precursor protein (APP). These interconnections are consistent with the reported effects of both PLD3 and SORL1 on amyloid biology and implicate DIP2A in the same process. Alternative figures presenting the network with all interacting proteins listed and the result of the network analysis with only the genetically associated loci are found in Supplementary Figure 6. The colored nodes are the proteins encoded by genes implicated in AD (genetic and epigenomic associations); the colors have no meaning. The connecting proteins not known to be associated AD are shown in gray.