Novel Insights Into the Causes of Schizophrenia: Part 1

March 14, 2016
Brian Miller, MD, PhD, MPH
Volume 33, Issue 3

Groundbreaking work by a Harvard team represents a bold step in our journey toward understanding schizophrenia.

A journey of a thousand miles  begins with a single step.


Strategic Objective 1 of the National Institute of Mental Health (NIMH) Strategic Plan is to “define the mechanisms of complex behaviors,” including molecules and genomic factors.1 The Plan notes that “large, replicated genomic studies have revealed many common and rare variants associated with the most heritable conditions (eg, schizophrenia, bipolar disorder, autism). . . . However, we still cannot explain the root cause(s) of mental illnesses. The task now is to sort through the complex patterns of genomic variation to define and elucidate how these variations confer risk. This strategy should not only identify critical pathways and circuits but potential new therapeutic targets.”

[[{"type":"media","view_mode":"media_crop","fid":"46657","attributes":{"alt":"© Heather de Rivera/ McCarroll lab, Harvard Medical School","class":"media-image media-image-right","id":"media_crop_8330930862575","media_crop_h":"0","media_crop_image_style":"-1","media_crop_instance":"5429","media_crop_rotate":"0","media_crop_scale_h":"140","media_crop_scale_w":"200","media_crop_w":"0","media_crop_x":"0","media_crop_y":"0","style":"font-size: 13.008px; line-height: 1.538em; float: right;","title":"© Heather de Rivera/ McCarroll lab, Harvard Medical School","typeof":"foaf:Image"}}]]Newly published work by Sekar and colleagues2 from Steven McCarroll’s laboratory at Harvard Medical School represents tangible progress toward this NIMH strategic objective. This rigorous landmark study links together how a specific genomic variant confers increased risk of schizophrenia. Namely, the C4 gene, which encodes the complement protein C4, may increase schizophrenia risk by influencing synaptic pruning during critical periods of brain development. Furthermore, this work connects replicated, yet seemingly disparate, findings that schizophrenia is associated with immune genes (specifically those within the major histocompatibility complex [MHC] locus), has a usual age of onset in young adulthood, and is associated with cortical thinning and synapse loss.

The MHC locus and schizophrenia

Although the etiopathophysiology of schizophrenia is largely unknown, replicated pathological findings include progressive gray matter loss and decreased density of neuronal dendritic spines.3,4 Genome-wide association studies have identified more than 100 candidate loci associated with schizophrenia5: genetic markers across the MHC on chromosome 6, which plays critical roles in immune system function, show the strongest association.6 However, specific genes involved in the association between schizophrenia and the MHC have not been characterized.

In a series of provocative experiments, Sekar and colleagues2 discovered that the association between schizophrenia and the MHC locus involves many common, structurally distinct alleles of the C4 gene that encodes the complement protein C4. These C4 alleles affect expression of 2 different C4 proteins in the brain (C4A and C4B), and each allele was associated with schizophrenia risk in proportion to its effect on C4A expression. C4 is expressed by neurons (localized to dendrites, axons, and synapses) and is secreted. In a mouse model, C4 promoted synaptic elimination or “pruning” during the developmentally timed maturation of a neuronal circuit.

Human C4 exists as 2 different genes, C4A and C4B, and 1 to 3 of these genes are commonly found in tandem (called haplotypes) within the MHC locus. C4A and C4B also exist in both long (L) and short (S) forms (ie, C4AL, C4AS, C4BL, C4BS). The long form is distinguished by the insertion of a human endogenous retrovirus (HERV). Using polymerase chain reaction, the  team assessed the C4 gene status of 222 copies of human chromosome 6. Four common C4 haplotypes (C4AL-C4BL, C4AL-C4BS, C4AL-C4AL, and C4BS) were found on 90% of these 222 chromosomes.

Next, the researchers explored how C4 variation affected expression of C4 genes. They measured C4A and C4B RNA expression in 674 postmortem human brain samples from 245 donors. They found that:

C4A and C4B RNA expression increased in proportion to C4A and C4B copy number

• Expression levels were 2 to 3 times greater for C4A than for C4B

• The copy number of the C4 long form (which contains the HERV insertion) increased the ratio of C4A to C4B expression

Using statistical imputation techniques the researchers then evaluated the association between schizophrenia and C4 haplotypes in the extended MHC locus and predicted C4A and C4B expression levels based on C4 haplotypes, using the large Psychiatric Genomics Consortium (28,799 patients with schizophrenia and 35,986 controls from 40 cohorts in 22 counties). They found a peak at C4, where schizophrenia associated most strongly with the genetic predictor of C4A expression levels. Among the 4 common C4 haplotypes, the risk (odds ratio) of schizophrenia ranged from 1.00 (for C4BS) to 1.27 (for C4AL-C4AL). In post-mortem brain samples, the C4AL-C4AL was also associated with the highest levels of C4A expression.

Based on these results, the team measured C4A brain RNA expression levels in 35 patients with schizophrenia and 70 controls across 5 brain regions. They found a 1.4-fold greater C4A expression in patients with schizophrenia.

They then evaluated the distribution of C4 in the human brain, using immunohistochemistry. They found C4+ cells in both gray and white matter. The greatest density of C4+ cells was in the hippocampus. The pattern of C4+ reactivity suggested that C4 is produced by, or deposited on, neurons and synapses. Furthermore, cultured human primary cortical neurons expressed C4 RNA and secreted C4 protein.

Based on the association between C4 and schizophrenia, the presence of C4 at synapses, and the role of complement proteins in synapse elimination, the researchers used a mouse model to test the hypothesis that C4 is involved in synaptic pruning. It is known that synaptic projections from retinal ganglion cells onto thalamic relay neurons in the dorsal lateral geniculate nucleus undergo activity-dependent synaptic refinement.

C4 RNA was expressed in retinal ganglion cells and in the lateral geniculate nucleus purified from the retina during the period of synaptic pruning in mice. Furthermore, mice genetically deficient in C4 (C4-/-) had greater overlap between retinal ganglion cell inputs from the eyes than wild-type controls (C4+/+), and heterozygous mice (C4+/-) had an intermediate phenotype, indicating that C4 mediates synaptic pruning in the developing brain.

The bottom line? The association between schizophrenia and the MHC locus involves many common, structurally distinct alleles of the C4 gene, which encodes the complement protein C4 and plays a role in synaptic elimination or “pruning” during brain development.


Dr Miller is Associate Professor in the Department of Psychiatry at Georgia Regents University in Augusta, GA, and Schizophrenia Section Editor for Psychiatric Times. He reports no conflicts of interest concerning the subject matter of this article.


For a full list of references, see Part 2 of this article.