Molecular Mechanisms and Timing of Cortical Immune Activation in Schizophrenia

The role of the immune system in psychiatric disorders has been garnering increased attention (1). In the absence of established classic immune-mediated neuropathology such as that seen in autoimmune or infectious disorders, the interpretations of these findings have evoked considerable controversy and debate. Previously, members of the Lewis team at the University of Pittsburgh reported immune system activation including higher mRNA levels of interferon-induced transmembrane protein (IFITM) in the dorsolateral prefrontal cortex from schizophrenia cases (2). In their follow-up study in this issue, Volk et al. (3) explore a basis for this increased gene expression and seek to evaluate whether a prenatal immune-related insult, as opposed to a postnatal exposure to infectious agents, causes the observed immune gene expression increase in the prefrontal cortex of adults diagnosed with schizophrenia.

The authors identify several genes that are significantly overexpressed in dorsolateral prefrontal cortex gray matter of schizophrenia subjects in comparison to matched controls. Specifically, they report significant increases in mRNA levels for interleukin-6 (IL-6) and interferon-β, both of which induce the expression of IFITM. Interestingly, Schnurri-2 mRNA is significantly lowered, which decreases IFITM levels by inhibiting NF-κB, a well-known transcription regulator of the immune gene network. The authors’ findings collectively amount to a significant increase in IFITM in dorsolateral prefrontal cortex gray matter of subjects diagnosed with schizophrenia. Of note, it is well known that IFITM and other immune genes reported here constitute an immune response to a host of viral infections, including influenza A, dengue, and West Nile virus (4). This raises questions as to the clinical, etiological, and/or functional relevance of IFITM expression in schizophrenia. The increased expression of interferon-β and NF-κB makes matters more complicated, since these are master regulators of many genes. One of the basic functions of components of the immune system is to protect the host against pathogens. However, they may also play a separate role distinct from immunological responses. Of particular interest is a report on the major histocompatibility complex I and its potential role in synaptic plasticity and development (5).

The series of experiments detailed by Volk et al. in this latest article are carefully crafted and well designed. Regarding postmortem brain case-control matching, the authors pay careful attention to parameters such as general demographic characteristics, postmortem interval, RNA integrity number, and pH (known to be altered by antipsychotic treatment). However, in a postmortem sample, it is difficult to assess other potentially confounding factors that can influence immune system activation, such as a medical history of severe viral infections, including hepatitis, HIV immunogenicity, influenza, and so on. Given their medical and socioeconomic status—many are homeless or institutionalized—individuals with schizophrenia have a higher likelihood of exposure to a myriad of infectious agents.

There has been a report on significant effects of antipsychotic treatment on the immune gene network (6), which means that postmortem studies of schizophrenia must undertake the almost impossible task of controlling for treatment-effect confounds on gene expression in the brain. That said, Volk et al. do address antipsychotic treatment effects in this study and report no significant effect of olanzapine or haloperidol on target gene expression in prefrontal cortex of chronically (17–27 months) treated monkeys (Macaca fascicularis). However, the chronic effects of antipsychotic treatment over decades may produce stable changes in human brain neurochemistry and structure that are difficult to replicate, even in nonhuman primates. While an alternative approach may include looking for correlations between gene expression and estimates of lifetime antipsychotic doses or antipsychotic levels at the time of death, this approach also has its limitations.

The immune system is extremely complex, with a highly interconnected group of genes that are subject to expression alterations by a number of genetic and epigenetic factors. The most tractable of these factors that remain unchanged in a changeable environment are genetic variants that are associated with disease and possibly with downstream target gene expression. Recent genome-wide association studies have identified significant single-nucleotide polymorphisms (SNPs) within the major histocompatibility complex (MHC) in association with schizophrenia risk (7, 8). However, because of genetic diversity, extensive linkage disequilibrium, and the highly variable nature of immune gene expression regulation (9), no particular genes or transcripts have been proposed as targets within the MHC region. One way to tackle this dilemma would be to divide the immune genetic network into functional gene clusters and then check for expression quantitative trait loci associations between schizophrenia risk SNPs and single/co-expressed gene targets within these clusters in nonpsychiatric subjects free of treatment and substance abuse confounds. Then the differential expression of these SNP-derived gene targets across psychiatric diagnoses could be tested. Lending support to this approach, genetic variants with the MHC region have already been connected to functional outcomes such as hippocampal volume and episodic memory performance (10) that are important to consider in complex psychiatric disorders.

When evaluating changes in the immune system in schizophrenia, both the cause and the timing of the changes are important considerations. Are they due to a pre- or perinatal trigger, such as maternal infection, or do they occur later in life, perhaps as a result of abnormal regulation of the immune system, as in multiple sclerosis? Volk et al. correctly point out that a potential trigger for immune activation includes exposure to infectious diseases at hospitals or mental institutions. Schizophrenia patients are commonly housed in group homes, nursing homes, shelters, and state hospitals, where there is a virtually continuous exposure to innumerable infectious agents. Potential pathogens include viruses that can induce expression of immune system genes, including NFKB and IFITM. The authors attempt to address this issue in a mouse model of immune system activation. Their findings suggest that prenatal immune insults, at least in mice, are not responsible for the changes in gene expression that they observed in schizophrenia. This leads to the suggestion that the immunological changes in postmortem schizophrenia brain tissue occur well after birth, which needs to be reconciled with data implicating a neurodevelopmental origin for this disorder.

The epidemiological evidence also suggests a link between infection, autoimmune illness, and schizophrenia. Before the onset of schizophrenia, a serious infection resulting in hospitalization raises the risk of schizophrenia by 60%, pre-existing autoimmune illness raises risk by 29%, and the two factors combined double the risk for developing schizophrenia (11). Among other studies, a recent serological study reported that C-reactive protein in the pregnant mother, consistent with an inflammatory process, could potentially involve the fetus and synergistically interact with risk of the fetus developing schizophrenia later on in life (12, 13).

In summary, the Volk et al. study offers an important contribution to intriguing questions surrounding the involvement of the immune system in the pathogenesis of schizophrenia. It also contributes to our understanding of prenatal infectious triggers and postnatal immune system activation. While the authors provide interesting evidence for the latter (ruling against prenatal effects), the relationship between immune gene expression alterations and schizophrenia etiology remains inconclusive. The complex and variable nature of immune gene expression regulation would suggest a more tractable approach using risk genetic variant association with gene clusters within the immune system followed by a differential gene expression study by diagnosis.