Immune-mediated inflammation is normally greatly reduced in the brain, a phenomenon called immune privilege [23
]. Immune mediated inflammation can have severe consequences for the brain, which have limited capacity for regeneration. We here show that four genes involved in inflammatory response are up-regulated in post-mortem frontal cortex samples from schizophrenic subjects irrespective of neuroleptic treatment. That inflammatory factors may be involved in the disease has been scientifically discussed since the "epidemic" appearance of psychotic disorders in 1845 [24
]. For example, signs of inflammation and microglia activation in post-mortem brains, a dysfunctional blood brain barrier, enhanced cytokine levels in the CSF preceding psychotic episodes, and signs of T cell system activation, all point to parallels between schizophrenia and autoimmune/infectious disorders. Other indirect evidence to support that early inflammation may contribute to the development of schizophrenia include the facts that maternal infections increase the risk of the disease [25
], mice exposed to cytokines in the uterus show behavioural alterations relevant for schizophrenia [26
], and an increased incidence of the disease during the winter months [29
Our results give additional support to the inflammatory hypothesis for schizophrenia and provide four biological markers for the disease likely to be directly involved in inflammatory processes, namely GBP1, SERPINA3, IFITM2, and IFITM3. GBP1
, or Interferon-gamma-inducible human guanylate binding protein 1, belongs to the family of dynamin-related large GTP-binding proteins. It is specifically induced by interferon gamma [30
], and is a marker for monitoring cytokine induced phenotype of endothelial cells during inflammation [31
is a member of the serine protease inhibitor superfamily. It is required for the regulation of leukocyte proteases released during an inflammatory response [32
], and its plasma levels increase more than four-fold in response to inflammation [33
]. IFITM2 and -3
are members of a family of genes encoding interferon-induced transmembrane proteins. Interestingly, mouse ifitm1 and -3, which are expressed in primordial germ cells, also have important and specific roles for cell development [34
We observed coordinated increases in the expression levels of these four genes in the brain of the schizophrenic subjects. It is unlikely that our results are driven by sporadic infectious diseases late in life, as schizophrenic subjects showed significantly elevated mRNA levels also when the effect of infectious death cause was accounted for in the analysis. Although alterations in these immune-related transcripts are not sufficient evidence for inflammation in schizophrenia, and these genes are not classic indicators of chronic or acute inflammation, our results indicate a significant alteration of inflammation-related pathways. Thus, modifications in inflammatory-related responses in the brain appear to be a common characteristic for the progression of schizophrenia for a large group of patients.
A recent theory postulates that schizophrenia may primarily be a consequence of vascular inflammation in the brain [14
]. The theory proposes that abnormalities arise because genetically modulated inflammatory reactions damage the micro-vascular system in response to environmental agents such as infections, hypoxia, and physical trauma. To evaluate this hypothesis, we examined whether inflammatory cytokines can induce the expression of the four genes in a human endothelial cell line. Our in-vitro
experiments demonstrated that GBP1, SERPINA3, IFITM2, and IFITM3, could indeed be induced by cytokine stimulation in endothelial cells. GBP1 is known as a key mediator of the anti-proliferative effect of inflammatory cytokines in endothelial cells [35
]. The elevated mRNA levels in schizophrenic subjects may thus reflect changes in endothelial brain cells, with possible long-term consequences for the micro-vascular system during the progression of the disease.
Intriguingly, we also observed expression of inflammation related genes in oligodendroglial cells and induction by cytokines in these cells. Our results indicate that oligodendroglial cells may indeed contain receptors for several inflammatory cytokines and that the expression of several inflammation-related genes can be induced by these cytokines. It has recently been shown that inflammation can modify myelination levels on transplanted oligodendrocyte precursors [36
]. Also, oligodendrocyte precursor responses are dependent on the presence of cytokines [37
]. Thus, the observed increased expression of GBP1, IFITM2, IFITM3 and SERPINA3 in schizophrenic subjects may affect myelin producing cells, offering a possible link between the inflammatory and the myelin hypothesis of the disease.
agents appear to have little or no effect on symptoms of schizophrenia [38
]. On the other hand, anti-inflammatory
agents have been used for the treatment of one patient, with improvement of negative symptoms [40
]. Our results may encourage future clinical trials, aimed to determine whether anti-inflammatory agents in combination with traditional neuroleptics can be used as an effective treatment for positive and/or negative symptoms of schizophrenia.