Limits...
DNA methylation analysis of the autistic brain reveals multiple dysregulated biological pathways.

Nardone S, Sams DS, Reuveni E, Getselter D, Oron O, Karpuj M, Elliott E - Transl Psychiatry (2014)

Bottom Line: In BA10 we found a very significant enrichment for genomic areas responsible for immune functions among the hypomethylated CpGs, whereas genes related to synaptic membrane were enriched among hypermethylated CpGs.Finally, we determined the epigenetic dysregulation of the gene HDAC4, and we confirm that the locus encompassing C11orf21/TSPAN32 has multiple hypomethylated CpGs in the autistic brain, as previously demonstrated.Our data suggest a possible role for epigenetic processes in the etiology of ASD.

View Article: PubMed Central - PubMed

Affiliation: Faculty of Medicine, Bar Ilan University, Safed, Israel.

ABSTRACT
Autism spectrum disorders (ASD) are a group of neurodevelopmental conditions characterized by dysfunction in social interaction, communication and stereotypic behavior. Genetic and environmental factors have been implicated in the development of ASD, but the molecular mechanisms underlying their interaction are not clear. Epigenetic modifications have been suggested as molecular mechanism that can mediate the interaction between the environment and the genome to produce adaptive or maladaptive behaviors. Here, using the Illumina 450 K methylation array we have determined the existence of many dysregulated CpGs in two cortical regions, Brodmann area 10 (BA10) and Brodmann area 24 (BA24), of individuals who had ASD. In BA10 we found a very significant enrichment for genomic areas responsible for immune functions among the hypomethylated CpGs, whereas genes related to synaptic membrane were enriched among hypermethylated CpGs. By comparing our methylome data with previously published transcriptome data, and by performing real-time PCR on selected genes that were dysregulated in our study, we show that hypomethylated genes are often overexpressed, and that there is an inverse correlation between gene expression and DNA methylation within the individuals. Among these genes there were C1Q, C3, ITGB2 (C3R), TNF-α, IRF8 and SPI1, which have recently been implicated in synaptic pruning and microglial cell specification. Finally, we determined the epigenetic dysregulation of the gene HDAC4, and we confirm that the locus encompassing C11orf21/TSPAN32 has multiple hypomethylated CpGs in the autistic brain, as previously demonstrated. Our data suggest a possible role for epigenetic processes in the etiology of ASD.

Show MeSH

Related in: MedlinePlus

DNA methylation changes in autistic cerebral cortex regions. (a, b) Heat map of differentially methylated CpG sites between autism and control cohorts in prefrontal cortex (a) and cingulate gyrus (b). Scaled methylation values are color-coded according to the legend on the left. The dendrogram delineates the hierarchical clustering based on the >5% differentially methylated CpG sites after FDR (<0.05) correction. The top bar indicates the disease status: red, autism; black, control. The bottom bars evidence further variables for each sample: sex (gray, male; black, female), comorbidity of seizures (green, autism case with seizure disorder; red, autism case without seizure disorder; black, control), age, pH assessment and postmortem interval (PMI). The corresponding scale for the quantitative variables is shown on the left. (c, d) Pie charts depicting the neighborhood and genomic locations of CpGs represented in the microarray (middle) and those differentially methylated between controls and autism in BA10 (left) and BA24 (right). Only CpG sites >5% differentially methylated at an FDR <0.05 were plotted. (e) Diagram depicting the number of CpGs showing significant methylation differences between BA10 (blue) and BA24 (red) in control samples (top) and autism samples (bottom) at FDR <0.05 and difference of methylation >5%. (f) Scatter plot illustrating the second principle component (PC2; y axis) describing the divergence in methylation between two cortical areas, BA10 and BA24, in autism and control groups. BA, Brodmann area; FDR, false discovery rate.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4203003&req=5

fig1: DNA methylation changes in autistic cerebral cortex regions. (a, b) Heat map of differentially methylated CpG sites between autism and control cohorts in prefrontal cortex (a) and cingulate gyrus (b). Scaled methylation values are color-coded according to the legend on the left. The dendrogram delineates the hierarchical clustering based on the >5% differentially methylated CpG sites after FDR (<0.05) correction. The top bar indicates the disease status: red, autism; black, control. The bottom bars evidence further variables for each sample: sex (gray, male; black, female), comorbidity of seizures (green, autism case with seizure disorder; red, autism case without seizure disorder; black, control), age, pH assessment and postmortem interval (PMI). The corresponding scale for the quantitative variables is shown on the left. (c, d) Pie charts depicting the neighborhood and genomic locations of CpGs represented in the microarray (middle) and those differentially methylated between controls and autism in BA10 (left) and BA24 (right). Only CpG sites >5% differentially methylated at an FDR <0.05 were plotted. (e) Diagram depicting the number of CpGs showing significant methylation differences between BA10 (blue) and BA24 (red) in control samples (top) and autism samples (bottom) at FDR <0.05 and difference of methylation >5%. (f) Scatter plot illustrating the second principle component (PC2; y axis) describing the divergence in methylation between two cortical areas, BA10 and BA24, in autism and control groups. BA, Brodmann area; FDR, false discovery rate.

Mentions: We performed filtering for high quality data which led to the exclusion of one sample, the BA24 region from individual AN04166, due to low signal-to-noise ratio in the sample. Following normalization procedures previously established for the Illumina 450 K methylation array (see Materials and Methods), we set cutoff parameters to detect differentially methylated CpG sites. We considered for further analyses only those CpGs that had a difference of at least 5% in DNA methylation after false discovery rate correction (false discovery rate <0.05). In addition, we removed probes that contained single-nucleotide polymorphisms that had a frequency >5% in human population, are cross-reactive or have previously been shown to display polymorphic binding (Supplementary Table 3).23 Applying these parameters, there were 5329 CpG sites differentially methylated between control and autism cohorts in BA10 (Figure 1a, Supplementary Table 5) and 10 745 in the BA24 (Figure 1b, Supplementary Table 6). Analysis of neighborhood location revealed that differentially methylated probes were more often found in areas of low CpG density, and rarely in CpG islands, compared with their representation on the microarray (Figure 1c). In addition, analysis of genomic location revealed a slight enrichment of differentially methylated probes in gene bodies and a significant decrease in transcription start site (TSS) here defined as TSS200 (Figure 1d).


DNA methylation analysis of the autistic brain reveals multiple dysregulated biological pathways.

Nardone S, Sams DS, Reuveni E, Getselter D, Oron O, Karpuj M, Elliott E - Transl Psychiatry (2014)

DNA methylation changes in autistic cerebral cortex regions. (a, b) Heat map of differentially methylated CpG sites between autism and control cohorts in prefrontal cortex (a) and cingulate gyrus (b). Scaled methylation values are color-coded according to the legend on the left. The dendrogram delineates the hierarchical clustering based on the >5% differentially methylated CpG sites after FDR (<0.05) correction. The top bar indicates the disease status: red, autism; black, control. The bottom bars evidence further variables for each sample: sex (gray, male; black, female), comorbidity of seizures (green, autism case with seizure disorder; red, autism case without seizure disorder; black, control), age, pH assessment and postmortem interval (PMI). The corresponding scale for the quantitative variables is shown on the left. (c, d) Pie charts depicting the neighborhood and genomic locations of CpGs represented in the microarray (middle) and those differentially methylated between controls and autism in BA10 (left) and BA24 (right). Only CpG sites >5% differentially methylated at an FDR <0.05 were plotted. (e) Diagram depicting the number of CpGs showing significant methylation differences between BA10 (blue) and BA24 (red) in control samples (top) and autism samples (bottom) at FDR <0.05 and difference of methylation >5%. (f) Scatter plot illustrating the second principle component (PC2; y axis) describing the divergence in methylation between two cortical areas, BA10 and BA24, in autism and control groups. BA, Brodmann area; FDR, false discovery rate.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4203003&req=5

fig1: DNA methylation changes in autistic cerebral cortex regions. (a, b) Heat map of differentially methylated CpG sites between autism and control cohorts in prefrontal cortex (a) and cingulate gyrus (b). Scaled methylation values are color-coded according to the legend on the left. The dendrogram delineates the hierarchical clustering based on the >5% differentially methylated CpG sites after FDR (<0.05) correction. The top bar indicates the disease status: red, autism; black, control. The bottom bars evidence further variables for each sample: sex (gray, male; black, female), comorbidity of seizures (green, autism case with seizure disorder; red, autism case without seizure disorder; black, control), age, pH assessment and postmortem interval (PMI). The corresponding scale for the quantitative variables is shown on the left. (c, d) Pie charts depicting the neighborhood and genomic locations of CpGs represented in the microarray (middle) and those differentially methylated between controls and autism in BA10 (left) and BA24 (right). Only CpG sites >5% differentially methylated at an FDR <0.05 were plotted. (e) Diagram depicting the number of CpGs showing significant methylation differences between BA10 (blue) and BA24 (red) in control samples (top) and autism samples (bottom) at FDR <0.05 and difference of methylation >5%. (f) Scatter plot illustrating the second principle component (PC2; y axis) describing the divergence in methylation between two cortical areas, BA10 and BA24, in autism and control groups. BA, Brodmann area; FDR, false discovery rate.
Mentions: We performed filtering for high quality data which led to the exclusion of one sample, the BA24 region from individual AN04166, due to low signal-to-noise ratio in the sample. Following normalization procedures previously established for the Illumina 450 K methylation array (see Materials and Methods), we set cutoff parameters to detect differentially methylated CpG sites. We considered for further analyses only those CpGs that had a difference of at least 5% in DNA methylation after false discovery rate correction (false discovery rate <0.05). In addition, we removed probes that contained single-nucleotide polymorphisms that had a frequency >5% in human population, are cross-reactive or have previously been shown to display polymorphic binding (Supplementary Table 3).23 Applying these parameters, there were 5329 CpG sites differentially methylated between control and autism cohorts in BA10 (Figure 1a, Supplementary Table 5) and 10 745 in the BA24 (Figure 1b, Supplementary Table 6). Analysis of neighborhood location revealed that differentially methylated probes were more often found in areas of low CpG density, and rarely in CpG islands, compared with their representation on the microarray (Figure 1c). In addition, analysis of genomic location revealed a slight enrichment of differentially methylated probes in gene bodies and a significant decrease in transcription start site (TSS) here defined as TSS200 (Figure 1d).

Bottom Line: In BA10 we found a very significant enrichment for genomic areas responsible for immune functions among the hypomethylated CpGs, whereas genes related to synaptic membrane were enriched among hypermethylated CpGs.Finally, we determined the epigenetic dysregulation of the gene HDAC4, and we confirm that the locus encompassing C11orf21/TSPAN32 has multiple hypomethylated CpGs in the autistic brain, as previously demonstrated.Our data suggest a possible role for epigenetic processes in the etiology of ASD.

View Article: PubMed Central - PubMed

Affiliation: Faculty of Medicine, Bar Ilan University, Safed, Israel.

ABSTRACT
Autism spectrum disorders (ASD) are a group of neurodevelopmental conditions characterized by dysfunction in social interaction, communication and stereotypic behavior. Genetic and environmental factors have been implicated in the development of ASD, but the molecular mechanisms underlying their interaction are not clear. Epigenetic modifications have been suggested as molecular mechanism that can mediate the interaction between the environment and the genome to produce adaptive or maladaptive behaviors. Here, using the Illumina 450 K methylation array we have determined the existence of many dysregulated CpGs in two cortical regions, Brodmann area 10 (BA10) and Brodmann area 24 (BA24), of individuals who had ASD. In BA10 we found a very significant enrichment for genomic areas responsible for immune functions among the hypomethylated CpGs, whereas genes related to synaptic membrane were enriched among hypermethylated CpGs. By comparing our methylome data with previously published transcriptome data, and by performing real-time PCR on selected genes that were dysregulated in our study, we show that hypomethylated genes are often overexpressed, and that there is an inverse correlation between gene expression and DNA methylation within the individuals. Among these genes there were C1Q, C3, ITGB2 (C3R), TNF-α, IRF8 and SPI1, which have recently been implicated in synaptic pruning and microglial cell specification. Finally, we determined the epigenetic dysregulation of the gene HDAC4, and we confirm that the locus encompassing C11orf21/TSPAN32 has multiple hypomethylated CpGs in the autistic brain, as previously demonstrated. Our data suggest a possible role for epigenetic processes in the etiology of ASD.

Show MeSH
Related in: MedlinePlus