January 30, 2014. The first genomewide survey of methylation patterns in schizophrenia turns up 25 suspect regions, some of which include genes linked to hypoxia. Published online January 8 in JAMA Psychiatry, the study examines blood samples from 759 people with schizophrenia and 738 controls to find differences in the methyl groups that, when attached to a segment of DNA, suppress expression of nearby genes. Led by Edwin van den Oord at Virginia Commonwealth University in Richmond, the results suggest that methylation patterns found in blood could record past environmental insults and may serve as useful biomarkers for schizophrenia.
Some methylation patterns found in blood cells may mirror those in the brain, according to a second, smaller study published in Translational Psychiatry on January 7. A team led by Joanne Voisey at Queensland University of Technology, Brisbane, Australia, found that 99 genes that were differently methylated in brain samples from people with schizophrenia compared to controls were on the same list of genes previously singled out as differently methylated in blood samples.
Together, the results point to methylation patterns as a rich source for variation that may hold clues to how genes and environment combine to boost risk for schizophrenia. Researchers have seized upon epigenetics, either in the form of methylation or histone modification, because it offers a way for experience to alter what is produced by the genome. The cells in which the methylation patterns are studied, however, may matter. If patterns found in blood match those found in the brain, they might provide some causal insights into the disease process. If they are specific to blood, however, they may still be a useful biomarker from an easy to obtain source.
Previous studies have found methylation differences in schizophrenia in a limited selection of genes (see SRF related news report), but, as reported from a conference last year (see SRF related conference report), the new study is the first to look at all methylation across the entire genome, termed the “methylome.” Using a design similar to genomewide association studies (GWAS) (see SRF genetics primer), the researchers systematically compared methylation sites between cases and controls to see if any were differently methylated. Unlike initial GWAS efforts, the resulting methylome-wide association study (MWAS) identified some definitive hits that were also replicated in a second group.
Hints of hypoxia
First author Karolina Aberg and colleagues began by extracting the methylated portions of the genome from their samples, then sequencing those parts. They checked the methylation status of each of 26,752,702 methylation sites, also known as “CpG” sites, for the sequence of cytosine and guanine that permits methylation. Because the methylation status of a single site can be highly correlated with adjacent methylation sites, the researchers collapsed adjacent sites into blocks, giving a total of 4,344,016 blocks, which they tested for association with schizophrenia.
In all, 139 of these were significantly associated with schizophrenia, with 112 of the sites residing within genes. Some of these genes, such as SMAD3 and ARNT, are affected by hypoxia, with their transcription varying with oxygen levels. Network analyses of all MWAS-significant genes found that they were enriched in hypoxia-related gene sets, as well as immune system genes, based on interactions between their protein products, the pathways to which they belonged, or whether they were targets of the same microRNAs.
Twenty-five of the 139 survived a conservative correction for multiple comparisons, and the top hit—FAM63B—was replicated in an independent cohort. FAM63B itself is regulated by different microRNAs, which the researchers noted have links to neuron differentiation and dopamine expression. RELN, a familiar candidate gene found to be overly methylated in schizophrenia (Grayson et al., 2005), also turned up as a significant hit in the MWAS, and this remained significant in their replication sample. Except for RELN, the sites showed decreased methylation in schizophrenia relative to controls.
Smoking and other lifestyle differences did not explain the results, leaving the researchers to speculate that methylation status over hypoxia-related genes reflected past environmental mishaps. These may contribute to schizophrenia risk, which is boosted by obstetric complications (e.g., Clarke et al., 2011; see SRF related news report). The methylation patterns over these hypoxia-related genes would not necessarily be the risk factor incarnate; rather, they would compose a signature of past, pathogenic events. Even if such a signature remained only in blood cells, the researchers argue it could still provide a useful biomarker for identifying subtypes of schizophrenia.
Alternatively, blood may be a good proxy for brain, as suggested by the RELN findings, which mimicked those in earlier studies of postmortem brain. This view was supported by Voisey’s study of a less extensive set of methylation sites in postmortem brain samples from 24 people with schizophrenia and 24 controls. When first author L. Wockner and colleagues scanned 485,000 CpG sites for methylation, they found differences in 2,929 genes between schizophrenia and controls. Some of these, such as NOS1, SOX10, and DTNBP1, have previous ties to schizophrenia, and 99 were in the list of 589 genes flagged by a previous methylation study in blood (Nishioka et al., 2012). These results suggest that, to some extent, methylation patterns in blood may mirror the situation in brain and may provide clues to causal factors in schizophrenia.—Michele Solis.
Aberg KA, McClay JL, Nerella S, Clark S, Kumar G, Chen W, Khachane AN, Xie L, Hudson A, Gao G, Harada A, Hultman CM, Sullivan PF, Magnusson PK, van den Oord EJ. Methylome-Wide Association Study of Schizophrenia: Identifying Blood Biomarker Signatures of Environmental Insults. JAMA Psychiatry. 2014 Jan 8. Abstract
Wockner LF, Noble EP, Lawford BR, Young RM, Morris CP, Whitehall VL, Voisey J. Genome-wide DNA methylation analysis of human brain tissue from schizophrenia patients. Transl Psychiatry. 2014 Jan 7. Abstract