February 15, 2014. There is genetic overlap between schizophrenia and multiple sclerosis (MS) that appears to be driven by immune system-related genes, according to a new study that mined genomewide association data from the Psychiatric Genomics Consortium (PGC). Published online January 28 in Molecular Psychiatry, the study was led by Ole Andreassen, University of Oslo, Norway, and Anders Dale, University of California, San Diego. The researchers leveraged the increased statistical power obtained by combining samples from the two diseases, allowing them to identify substantially more loci associated with schizophrenia than could be identified using the data from schizophrenia alone.
Schizophrenia shares many risk variants with other disorders, especially bipolar disorder (see SRF related conference report; SRF related news report). The existence of this overlapping risk, whereby one gene or variant contributes to more than one phenotype, is known as pleiotropy. Andreassen, Dale, and colleagues recently used a novel Conditional False Discovery Rate approach to improve the detection of variants associated with both schizophrenia and bipolar disorder (Andreassen et al., 2013).
Increasing evidence points to a role for immune system molecules in the neurobiology of schizophrenia (see SRF related conference report). Epidemiological studies have suggested that maternal infection during pregnancy, thought to stimulate an inflammatory response in the fetus, elevates risk for the illness. Supporting this idea, several single nucleotide polymorphisms (SNPs) in the major histocompatibility complex (MHC) of chromosome 6, a region containing over 200 genes involved in the immune system, have been tied to an elevated risk for schizophrenia (see SRF related news report). Genes in this region are some of the best replicated genomewide association study (GWAS) findings to date (see SRF related news report). In the current analysis, Andreassen, Dale, and colleagues used the Conditional False Discovery Rate approach to examine the pleiotropy between schizophrenia and MS, a disease strongly linked to the immune system.
The pleiotropic advantage
In traditional GWAS analysis, individual SNPs below a p value of 10-8 are considered significant. However, in a polygenic disorder such as schizophrenia, many disease-related SNPs will be impossible to identify using standard statistical tools and current sample sizes. The approach used in this study is based on the idea that pleiotropic SNPs (those that contribute to more than one illness) will be easier to identify because of the greater power afforded by combining two samples. In the current study, the researchers used summary statistic data from two prior GWAS to identify schizophrenia SNPs conditioned on the effect in MS.
By combining the schizophrenia (n = 21,856) and MS (n = 27,148) GWAS data, the researchers found that the Conditional False Discovery Rate approach was able to identify 21 independent loci associated with schizophrenia, only 10 of which have been previously identified by using standard GWAS analysis approaches. One locus was within the MHC region and encompassed many genes, while the other 20 were single-gene loci at other regions throughout the genome. Within the complex MHC locus, the significant SNPs identified accounted for the majority of the overlap between the two diseases, and removing the MHC-related SNPs diminished the enrichment of schizophrenia conditional on MS.
Interestingly, the directions of effect in these SNPs were opposite: Alleles that increased the risk of multiple sclerosis were associated with a decreased risk of schizophrenia, suggesting that although both diseases are associated with MHC loci, the risk profiles are very different. The authors caution that the low imputation rates of the study make it difficult to determine the exact MHC variants associated with schizophrenia. Further evidence of immune involvement in schizophrenia comes from the finding that many of the non-MHC loci identified are also related to the immune system functioning. However, some of the MHC loci finger genes such as NOTCH4 and TRIM26, which have neurobiological functions, making it difficult to conclude that the overlap between schizophrenia and MS is solely linked to the immune system.
Andreassen and colleagues did not find an association between MS and bipolar disorder (n = 16,731), consistent with the earlier finding that the genetic overlap between schizophrenia and bipolar disorder does not involve the MHC region. Removing the MHC loci from the analysis did not affect the enrichment of bipolar disorder conditioned on MS, or on the previously reported enrichment of schizophrenia conditioned on bipolar disorder, suggesting that MHC involvement among psychotic disorders is specific to schizophrenia and may be a way to differentiate between schizophrenia and bipolar disorder. MHC involvement also does not seem to play a role in other psychiatric illnesses, as the researchers found no genetic enrichment in major depressive disorder, autism spectrum disorder, and attention deficit/hyperactivity disorder conditioned on MS.
Apart from the implications for schizophrenia, the findings of the current study have broad methodological implications, conclude the authors. “Our findings also indicate that leveraging existing genomewide association studies for pleiotropic genes may represent an important avenue by which currently available resources can be optimized to identify more of the missing heritability of complex phenotypes.”—Allison A. Curley.
O A Andreassen, H F Harbo, Y Wang, W K Thompson, A J Schork, M Mattingsdal, V Zuber, F Bettella, S Ripke, J R Kelsoe, K S Kendler, M C O'Donovan, P Sklar, The Psychiatric Genomics Consortium (PGC) Bipolar Disorder and Schizophrenia Work Groups, The International Multiple Sclerosis Genetics Consortium (IMSGC), L K McEvoy, R S Desikan, B A Lie, S Djurovic, and A M Dale. Genetic pleiotropy between multiple sclerosis and schizophrenia but not bipolar disorder: differential involvement of immune-related gene loci. Mol Psychiatry. 2014 Jan 28. Abstract