Background
Patients with schizophrenia have altered immune function [
1]. This, coupled with studies indicating the impact of maternal illness on offspring development [
2,
3], strongly suggests a link between the inflammasome and schizophrenia. As such, interleukin 6 (IL-6), which can function as both a pro- and anti-inflammatory cytokine- has been examined for its role in the development, course, symptomology, and applicability as a biomarker of schizophrenia.
Various lines of evidence suggest a compelling role for IL-6 in the underlying pathogenesis of schizophrenia. First and foremost, IL-6 and sIL-6R serum levels are significantly elevated in patients with schizophrenia [
4,
5], with implications for symptom severity [
6]. The single-nucleotide polymorphism (SNP) G/C at position -174 of the IL-6 gene may also result in increased susceptibility to schizophrenia, although this seems to be highly dependent on environmental factors [
7]. This SNP has also been linked to overall symptomology in schizophrenia [
8,
9]. Maternal immune activation models of neurodevelopmental insults have heralded insight into the role of IL-6 and schizophrenia. Maternally produced cytokines can both cross the placenta [
10] and induce an acute-phase reaction in response to immune challenges [
11]. A single maternal dose of IL-6 results in schizophrenic-like behavioral changes in rodent offspring [
12]. Furthermore, inhibiting IL-6 action in response to an immune challenge can normalize these behaviors in adult offspring [
12]. Thus, multiple lines of evidence link inflammatory factors, specifically IL-6, with mental illness and related symptomology.
In this study, we demonstrate that IL-6 mRNA levels from freshly extracted peripheral blood mononuclear cells (PBMC) could be a useful and easily clinically accessible biomarker for a diagnosis of schizophrenia. There is a general lack of correlation between cytokine serum levels and PBMC gene expression [
13]. As such, measurement of mRNA in PBMC is tissue specific and avoids interpretation regarding the source of serum levels of cytokines, which could emanate from the liver, muscle, adipose and other tissues. Additionally, we attempted to elucidate the role of IL-6 in disease onset and symptom presentation in schizophrenia.
Discussion
Our results demonstrate that IL-6 mRNA levels from freshly extracted PBMC are significantly upregulated in participants with schizophrenia when compared to healthy controls, thereby confirming our hypothesis. We also demonstrate that IL-6 mRNA levels are significantly elevated in patients who had an earlier age of onset, as well as those who demonstrated worse positive symptomatology. Finally, IL-6 levels may be a useful biomarker for a diagnosis of schizophrenia due to the predictive capabilities we demonstrate here.
Associations between a pro-inflammatory state and schizophrenia have been one of the more enduring findings of the field [
4]. Our finding of an elevation in levels of IL-6 is consistent with the majority of the literature [
6,
9,
22,
23]. One such study, by Frydecka, et al., [
23] found that serum IL-6 levels were significantly higher in participants with schizophrenia when compared to healthy controls. Additionally, in their sample, elevated IL-6 serum levels were associated with longer duration of illness. Our study complements and extends these findings. Importantly, IL-6 levels examined in this manuscript are mRNA from peripheral blood mononuclear cells. IL-6 levels measured in the serum, as in the Frydecka paper, can originate from multiple tissues, including adipose, muscle and liver, with particularly confounding effects given the well known metabolic disturbances in the schizophrenia population [
24,
25]. Thus the demonstrated elevated serum IL-6 levels could have originated at any number of locations, thereby making the true source, cause, and implications difficult to interpret. Additionally, the Frydecka et al., study found no associations between symptom presentation, as measured by the PANSS, and serum IL-6 levels, while we found a correlation with IL-6 mRNA levels and positive symptoms [
23]. These differences may be attributable to the differences in type of biochemical sample examined between our two studies, paving the way for further discussion as to the importance of the location of the IL-6 signal and its role in mental illness.
Due to the heterogeneity of prescribed psychotropic medication in our naturalistic study, all antipsychotic use was converted to Chlorpromazine (CPZ) units. The impact of psychotropic medication on the inflammasome in participants with schizophrenia is a dynamic and novel field. Levels of proinflammatory cytokines are significantly elevated in female patients treated with clozapine, when compared to healthy controls [
26]. However, these participants had significantly increased BMI levels when compared to normal controls, a phenotype associated with elevated cytokine levels, further complicating the interpretation. Additionally, treatment with olanzapine for 8 weeks reduces plasma IL-2 levels, but results in no changes in IL-6 or TNFa levels [
27]. Treatment with chlorpromazine inhibits TNFa, but upregulates peripheral IL-10 levels in mice [
28]. These findings all indicate that more research is needed regarding the uncontrollable variable of psychotropic medications play in the larger picture of inflammation and schizophrenia.
The association between IL-6 mRNA levels with age of onset and positive symptomatology is consistent with previous literature. In schizophrenia, serum IL-6 levels are positively correlated with PANSS Scores [
29], an association our data replicates and extends on by examining this correlation with PBMCs. In schizophrenia, age of illness onset is a crucial determinant of clinical outcomes [
30,
31], with a younger onset-age correlated with a worse disease prognosis and increased symptom presentation [
32‐
34]. IL-6 levels are associated with younger psychosis onset, longer duration of illness and worse illness presentation [
23], all indicators of worse disease course. IL-6 has been proposed to act as a “state marker” for schizophrenia, with elevated serum levels seen in first episode psychosis and acutely relapsed patients [
35]. These listed studies are all in line with our findings that the underlying pathophysiology of schizophrenia is characterized by a low-grade inflammatory state. It is important to the field to present multiple lines of consistent data regarding elevated levels of IL-6 in participants with schizophrenia in independent cohorts. Additionally, further studies need to be conducted examining IL-6 mRNA levels in participants with schizophrenia when compared to other psychiatric disorders, specifically depression.
Previous work has explored the use of IL-6 as a clinical predictor, but for infection in febrile neutropenia [
36], not schizophrenia. In schizophrenia research, the receiver operating characteristic analysis has been used to examine the predictive value of specific microRNAs not associated with the immune system [
37,
38]. Our work merges these two concepts to examine the predictive value of PBMC IL-6 mRNA levels in schizophrenia. Interestingly, evidence indicates that IL-6 levels are increased in schizophrenia-risk subjects, with highest levels demonstrated in those who transitioned from risk to acute psychosis [
39]. This suggests that IL-6 could serve as a pre-diagnostic tool for identifying individuals who are more likely to suffer from the disease. In the clinic, an especially useful aspect of the ROC is criterion; a cutoff that attempts to maximize the likelihood of a true positive (accurate diagnosis) while minimizing the chance of an incorrect diagnosis (false positive). In our data set, using a cutoff IL-6 mRNA value of 4.9 (arbitrary units) would result in a true positive rate of 53 %, while a false positive would occur 17 % of the time. If the cutoff value is increased, the chance of a true positive increases, while false positives decrease. For example, a cutoff of 6.0 in our sample would be five times more likely to make a correct diagnosis of schizophrenia based solely on IL-6 mRNA levels in PBMC. In the clinic, physicians would have access to much more patient data, and thus could adjust their criterion as needed to ensure an accurate diagnoses.
Acknowledgements
The authors would like to thank all the subjects who participated in this study.