May 7, 2014. Genetic variation in the sodium channel gene SCN2A contributes to cognition in people with schizophrenia, according to a paper published online April 9 in JAMA Psychiatry. Led by Daniel Weinberger at the Lieber Institute for Brain Development, Baltimore, Maryland, the study reports that two common variants in SCN2A were associated with "g," a measure of overall cognitive ability, in several schizophrenia samples, but not in controls. Brain imaging and postmortem data also associated SCN2A with altered brain function in regions important to cognition. Together, the results highlight sodium channels in the cognitive impairments of schizophrenia and suggest them as future treatment targets.
Though cognition is heritable, genomewide association studies (GWAS) in healthy participants have not turned up much in the way of its genetic contributors. The new findings, however, show some hits in the context of schizophrenia. Homing in on an isolated feature, such as cognition, as an intermediate phenotype may also be a way to better understand the genetic complexity of schizophrenia (Meyer-Lindenberg and Weinberger, 2006). This approach may reveal larger genetic signals that could be swamped out in a study of the disease itself, and it may help make sense of the growing list of genes so far associated with schizophrenia (see SRF related news report and SRF news report).
The new study found a clear signal at two linked single nucleotide polymorphisms (SNPs) within the sodium channel gene SCN2A, which encodes the alpha-2 subunit of a voltage-gated sodium ion channel that controls action potential initiation and propagation. Mutations in SCN2A have been implicated in a number of disorders, including autism (Sanders et al., 2012), intellectual disability (Rauch et al., 2012), and epilepsy (Meisler et al., 2005).
GWAS of g
First author Dwight Dickinson and colleagues began by screening DNA from a discovery cohort of 339 people with schizophrenia from the Clinical Brain Disorders Branch (CBDB)/National Institute of Mental Health (NIMH) Study of Schizophrenia Genetics sample. Participants had been tested on various measures of cognition, including verbal memory, working memory, and processing speed; scores were combined to obtain a composite measure of general cognition, or g. In the schizophrenia group, the researchers found a robust, genomewide-significant association between g and two linked SNPs within introns of the SCN2A gene. These SNPs explained 10.4 percent of the variance in g in this group—a whopping effect size for common variants—one that might reflect the uniformity of the sample, which was restricted to people with IQs over 70 and who were not substance abusers. For one of these variants, g varied with allele status: Those homozygous for the C allele were better off cognitively than those with one C and one T, who in turn performed better than T allele homozygotes.
Among controls, however, the story was different. No SNPs attained genomewide significance, and a SCN2A SNP highlighted in the schizophrenia group showed a weak trend in the opposite direction between allele status and g in controls. The SCN2A SNPs did not associate with case-control status, either. In 147 unaffected siblings, the researchers found a similar association between SCN2A genotype and g, which strengthens the idea that the SNPs reflect something about cognition in schizophrenia, rather than something about the disorder that affects cognition, such as hallucinations or medications.
Testing the SCN2A SNPs in three other independent schizophrenia samples also turned up significant associations with g, though these explained less g variation. Combining the discovery with the replications in schizophrenia samples (n = 983), researchers discovered that the SCN2A genotype explained on average 3.0 percent of variance in g.
SCN2A genotype also varied with direct measures of brain function. Dickinson and colleagues examined functional magnetic resonance imaging (fMRI) from 397 controls and 87 schizophrenia participants from the CBDB/NIMH sample who had performed a working memory task that engages prefrontal cortex. TT homozygotes with schizophrenia showed greater activation during this task—considered an indicator of a not very efficient brain working harder to achieve the same level of performance; in contrast, among controls, TT homozygotes showed the least amount of activation, a sign of maximal efficiency. Again, SCN2A genotype conferred different effects, depending on whether a person had schizophrenia or not.
Sequencing RNA from postmortem prefrontal cortex samples also revealed significantly lower levels in schizophrenia samples in two out of three SCN2A transcripts tested compared to control samples. Again, there was a difference in the effect of SCN2A genotype between cases and controls: In schizophrenia, T carriers expressed less SCN2A mRNA than CC homozygotes did, whereas the opposite was true for controls.
Together, the findings argue for a special relationship between SCN2A genotypes and cognition in schizophrenia. It may be that SCN2A effects on cognition only emerge in the context of other genetic risk factors for schizophrenia, i.e., epistasis. This echoes work by the same group that has found distinct differences in how one gene variant behaves depending on the identity of another gene variant (Papaleo et al., 2014; see also SRF related conference report). Though individual genetic contributors to cognition may be easily blurred by environmental influences that shape a person’s mental abilities over years of development, these may be made clearer in schizophrenia.—Michele Solis.
Dickinson D, Straub RE, Trampush JW, Gao Y, Feng N, Xie B, Shin JH, Lim HK, Ursini G, Bigos KL, Kolachana B, Hashimoto R, Takeda M, Baum GL, Rujescu D, Callicott JH, Hyde TM, Berman KF, Kleinman JE, Weinberger DR. Differential Effects of Common Variants in SCN2A on General Cognitive Ability, Brain Physiology, and messenger RNA Expression in Schizophrenia Cases and Control Individuals. JAMA Psychiatry. 2014 Apr 9. Abstract