30 Oct 2014
October 31, 2014. The strongest and most replicable genetic signal for schizophrenia emanates from a region of chromosome 6 containing genes for the major histocompatibility complex (MHC) molecules important for immunity, as well as many non-immune related genes. Pinpointing the source of these signals is not straightforward, however, because over 200 genes lie within the suspicious 10 Mb region, and because high linkage disequilibrium gives high correlations between individual SNP signals.
But researchers may finally be making some inroads into MHC's terra incognita, according to two presentations on Wednesday, October 15. Semanti Mukherjee of the Feinstein Institute for Medical Research, Manhasset, New York, presented the results from conditional analyses of the region in 38 cohorts of European ancestry that are part of the latest Psychiatric Genomics Consortium (PGC) schizophrenia sample (almost 30,000 cases and over 35,000 controls). By statistically taking into account the correlations between individual SNP signals to find the truly independent ones, Mukherjee reported three signals: one coming from a region containing GPX6 and histone genes; one near SYNGAP1, a gene already implicated in intellectual disability; and one from the HLA class II locus, which encodes proteins used to establish immunity.
In a later session, Steve McCarroll of the Broad Institute, Cambridge, Massachusetts, reported an extreme form of structural variation in this HLA locus. Focusing on the complement component 4 (C4) gene, which encodes a protein involved in the complement system that activates immune responses, he reported that, in Europeans, it has four different structures (C4A long, C4A short, C4B long, C4B short) that vary in their copy number. Using new techniques to figure out the precise number of C4 copies in each person, McCarroll reported that the C4 structures were associated with two to five different haplotypes, or patterns of SNPs in the region. This meant that the exact C4 structure was not taggable by any single SNP, but could be deduced by the haplotype. He also found that the number of C4 copies influenced C4 expression in postmortem brain tissue and schizophrenia risk. The role of C4 in the brain—for example, immune, neurodevelopmental, and/or something else—remains unclear.
The curious case of schizophrenia and rheumatoid arthritis
Two other presentations on the same day took up the epidemiological finding that rheumatoid arthritis (RA) is rarer in people with schizophrenia than expected by chance. People with schizophrenia get RA at 29 percent of the rate found in the general population, despite their high rates of smoking, which is a risk factor for RA. Clues like this might help identify genetic variants involved in both disorders. A meta-analysis of studies investigating the prevalence of RA in schizophrenia compared to controls confirmed this relationship, reported Jack Euesden of King's College London, UK. Looking at 10 studies, he found an odds ratio of 0.48, meaning that schizophrenia protected a person from RA. Using the polygenic risk score, which adds the contributions of a person's schizophrenia risk alleles, Euesden did not find a relationship with RA occurrence, however, leading him to suggest that environmental factors rather than genetic ones might underlie the association.
Naomi Wray of the University of Queensland, Brisbane, Australia, looked at the relationship between schizophrenia and RA in a different way. Looking at the SNP genotypes of over 60,000 people with either schizophrenia or RA, or controls, she found a negative relationship between schizophrenia and RA. This meant that, judging by their pattern of SNPs across the genome, people with schizophrenia were genetically less similar to those with RA than they were to controls. This negative correlation seemed mostly based in SNPs tagging coding and regulatory regions of the genome. The dual action of a single SNP—protective for one disease but a risk factor in another—suggests that common variants may have very different roles in different tissues, with different selection pressures acting in different contexts.
To look for a genetic footprint of immune system dysfunction in schizophrenia, Jennie Pouget of the Centre for Addiction and Mental Health, Toronto, Canada, presented results from her "hypothesis-driven" GWAS, in which she categorized SNPs as immune or non-immune, depending on whether they tagged over 900 genes involved in the immune response. Though this revealed an enrichment of immune SNPs in psoriasis, a known immune-related disease, an enrichment was not found in schizophrenia. Pouget suggested that the immune system link to schizophrenia may reflect events downstream of genetic factors.—Michele Solis.