We aimed at exploring potential pathophysiological processes across psychotic disorders, applying metabolomics in a large and well-characterized sample of patients and healthy controls.

Patients with schizophrenia and bipolar disorders (N = 212) and healthy controls (N = 68) had blood sampling with subsequent metabolomics analyses using electrochemical coulometric array detection. Concentrations of 52 metabolites including tyrosine, tryptophan and purine pathways were compared between patients and controls while controlling for demographic and clinical characteristics. Significant findings were further tested in medication-free subsamples.

Significantly decreased plasma concentrations in patients compared to healthy controls were found for 3-hydroxykynurenine (3OHKY, p = 0.0008), xanthurenic acid (XANU, p = 1.5×10−5), vanillylmandelic acid (VMA, p = 4.5×10−5) and metanephrine (MN, p = 0.0001). Plasma concentration of xanthine (XAN) was increased in the patient group (p = 3.5×10−5). Differences of 3OHKY, XANU, VMA and XAN were replicated across schizophrenia spectrum disorders and bipolar disorders subsamples of medication-free individuals.

Although prone to residual confounding, the present results suggest the kynurenine pathway of tryptophan metabolism, noradrenergic and purinergic system dysfunction as trait factors in schizophrenia spectrum and bipolar disorders. Of special interest is XANU, a metabolite previously not found to be associated with bipolar disorders.

Inflammation and immune activation have been implicated in the pathogenesis of severe mental disorders and cardiovascular disease (CVD). Despite high level of comorbidity, many studies of the immune system in severe mental disorders have not systematically taken cardiometabolic risk factors into account.

We investigated if inflammatory markers were increased in schizophrenia (SCZ) and affective (AFF) disorders independently of comorbid CVD risk factors. Cardiometabolic risk factors (blood lipids, body mass index and glucose) and CVD-related inflammatory markers CXCL16, soluble interleukin-2 receptor (sIL-2R), soluble CD14 (sCD14), macrophage inhibitory factor and activated leukocyte cell adhesion molecule (ALCAM) were measured in n = 992 patients (SCZ, AFF), and n = 647 healthy controls. We analyzed the inflammatory markers before and after controlling for comorbid cardiometabolic risk factors, and tested for association with psychotropic medication and symptom levels.

CXCL16 (p = 0.03) and sIL-2R (p = 7.8 × 10−5) were higher, while sCD14 (p = 0.05) were lower in patients compared to controls after controlling for confounders, with significant differences in SCZ for CXCL16 (p = 0.04) and sIL-2R (p = 1.1 × 10−5). After adjustment for cardiometabolic risk factors higher levels of sIL-2R (p = 0.001) and lower sCD14 (p = 0.002) remained, also in SCZ (sIL-2R, p = 3.0 × 10−4 and sCD14, p = 0.01). The adjustment revealed lower ALCAM levels (p = 0.03) in patients. We found no significant associations with psychotropic medication or symptom levels.

The results indicate that inflammation, in particular enhanced T cell activation and impaired monocyte activation, are associated with severe mental disorders independent of comorbid cardiometabolic risk factors. This suggests a role of novel pathophysiological mechanisms in severe mental disorders, particularly SCZ.

Common variants in the Vaccinia-related kinase 2 (VRK2) gene have been associated with schizophrenia, but the relevance of its encoded protein VRK2 in the disorder remains unclear.

To identify potential differences in VRK2 gene expression levels between schizophrenia, bipolar disorder, psychosis not otherwise specified (PNOS) and healthy controls.

VRK2 mRNA level was measured in whole blood in 652 individuals (schizophrenia, n = 201; bipolar disorder, n = 167; PNOS, n = 61; healthy controls, n = 223), and compared across diagnostic categories and subcategories. Additionally, we analysed for association between 1566 VRK2 single nucleotide polymorphisms and mRNA levels.

We found lower VRK2 mRNA levels in schizophrenia compared with healthy controls (P<10–12), bipolar disorder (P<10–12) and PNOS (P = 0.0011), and lower levels in PNOS than in healthy controls (P = 0.0042) and bipolar disorder (P = 0.00026). Expression quantitative trait loci in close proximity to the transcription start site of the short isoforms of the VRK2 gene were identified.

Altered VRK2 gene expression seems specific for schizophrenia and PNOS, which is in accordance with findings from genome-wide association studies. These results suggest that reduced VRK2 mRNA levels are involved in the underlying mechanisms in schizophrenia spectrum disorders.