The etiology and pathophysiology of schizophrenia remain unknown. A parallel transcriptomics, proteomics and metabolomics approach was employed on human brain tissue to explore the molecular disease signatures. Almost half the altered proteins identified by proteomics were associated with mitochondrial function and oxidative stress responses. This was mirrored by transcriptional and metabolite perturbations. Cluster analysis of transcriptional alterations showed that genes related to energy metabolism and oxidative stress differentiated almost 90% of schizophrenia patients from controls, while confounding drug effects could be ruled out. We propose that oxidative stress and the ensuing cellular adaptations are linked to the schizophrenia disease process and hope that this new disease concept may advance the approach to treatment, diagnosis and disease prevention of schizophrenia and related syndromes.

No abstract

Severe psychiatric disorders such as schizophrenia, bipolar disorder and major depressive disorder are brain diseases of unknown origin. No biological marker has been documented at the pathological, cellular, or molecular level, suggesting that a number of complex but subtle changes underlie these illnesses. We have used proteomic technology to survey postmortem tissue to identify changes linked to the various diseases. Proteomics uses two-dimensional gel electrophoresis and mass spectrometric sequencing of proteins to allow the comparison of subsets of expressed proteins among a large number of samples. This form of analysis was combined with a multivariate statistical model to study changes in protein levels in 89 frontal cortices obtained postmortem from individuals with schizophrenia, bipolar disorder, major depressive disorder, and non-psychiatric controls. We identified eight protein species that display disease-specific alterations in level in the frontal cortex. Six show decreases compared with the non-psychiatric controls for one or more diseases. Four of these are forms of glial fibrillary acidic protein (GFAP), one is dihydropyrimidinase-related protein 2, and the sixth is ubiquinone cytochrome c reductase core protein 1. Two spots, carbonic anhydrase 1 and fructose biphosphate aldolase C, show increase in one or more diseases compared to controls. Proteomic analysis may identify novel pathogenic mechanisms of human neuropsychiatric diseases. Molecular Psychiatry (2000) 5, 142-149.

Transcript changes and altered pathways in schizophrenia prefrontal cortex. (a) Mitochondria are the most affected cellular components at the transcript level in schizophrenia. Cellular localization of the significantly altered genes (both up-and downregulated) that passed RMA and filtering procedures were analyzed and visualized using GO Surfer (http://biosun1.harvard.edu/complab/gosurfer/). Branches and nodes represent pathways containing greater than five genes. Significantly altered 'cellular components' (Po0.05) are highlighted in red and their categories are indicated. (b) Metabolic categories found to be most significantly altered at the transcript level. EASE (http://david.niaid.nih.gov/david/ease.htm) was used for pathway analysis of microarray results and to determine significantly up-and/or downregulated GO biological processes and KEGG metabolic pathways. (c) Hierarchical clustering tree of schizophrenia (vertical blue lines) and controls (vertical gray lines) microarray chips on the basis of 59 significantly altered genes related to energy metabolism and oxidative stress. Drug-naive schizophrenia patients are denoted by ** (n ¼ 7), while minimally treated patients are marked by * (o6000 lifetime fluphenazine units; n ¼ 4). Note that the schizophrenia group appears to fall into two subclusters with respect to lowered transcript expression as indicated by the prominent blue shading. For more information on this topic, please see the article by Prabakaran et al on pp 684-697.

Based on the epidemiological finding that individuals with schizophrenia tend to be born in winter/spring when compared to the general population, we examined (1) the strength and timing of this effect in Northern Hemisphere sites, and (2) the correlation between the season of birth effect size and latitude. Studies were located via electronic data sources, published citations, and letters to authors. Inclusion criteria were that studies specify the diagnostic criteria used, that studies specify the counts of schizophrenia and general population births for each month, and that subjects and the general population be drawn from the same birth years and catchment area. We extracted data from eight studies based on 126,196 patients with schizophrenia and 86,605,807 general population births and drawn from 27 Northern Hemisphere sites. Comparing winter/spring versus summer/autumn births, we found a significant excess for winter/spring births (pooled odds ratio = 1.07; 95% confidence interval 1.05, 1.08; population attributable risk = 3.3%). There was a small but significant positive correlation between the odds ratios for the season of birth comparison and latitude (r = 0.271,p < 0.005). Furthermore, the shape of the seasonality in schizophrenia births varied by latitude band. These variations may encourage researchers to generate candidate seasonally fluctuating exposures.