Schizophrenia susceptibility genes converge on interlinked pathways related to glutamatergic transmission and long-term potentiation, oxidative stress and oligodendrocyte viability

Abstract

Over 130 genes have been associated with schizophrenia in genetic studies. None of these has reached a sufficient level of confidence to be accepted as a universal susceptibility gene and problems of replicability suggest that many may be false positives. Nevertheless, these genes can be grouped into distinct families related to glutamate transmission (in particular related to NMDA receptor function), the control of synaptic plasticity, dopaminergic transmission, oxidative stress, glutathione and quinone metabolism and oligodendrocyte viability. These families mirror the processes disrupted in the schizophrenic brain and certain gene families can be linked together to form a clearly defined signalling cascade involved in the phenomenon of NMDA receptor-dependent long-term potentiation and synaptic plasticity, that may be interconnected with oligodendrocyte and oxidative stress-related pathways. Many of the protein products of these genes interact with each other, forming complex integrated networks. Certain high-interest genes (for example DISC1, NRG1, COMT) may exert multiple effects on different areas of these pathways, while others exert more specific effects on certain branches. The convergence of a large number of genes on a definable signaling network raises the possibility of numerous interactions between gene candidates, and suggests that a targeted multigenic pathway approach would be useful in gene association studies.

Introduction

Schizophrenia is influenced both by genes and the environment. Prenatal exposure to famine or beri-beri (Davis and Bracha, 1996), tick infestation and Lyme disease (Brown, 1994) influenza or rubella (Brown, 2006) have all been associated with an increased risk of developing schizophrenia in later life. These observations, together with reports of increased ventricular volume and decreased cortical volume in later life in schizophrenia (with no evidence of an ongoing neuronal lesion process) (Johnstone et al., 1976, Wright et al., 2000) have contributed largely to neurodevelopmental hypotheses of schizophrenia (Rapoport et al., 2005).

Current hypotheses suggest that schizophrenia is characterised by hypoglutamatergic and hyperdopaminergic function. It is believed that the disease is related to over stimulation of subcortical DRD2 dopamine receptors, hypoactivity of frontal cortical DRD1 dopamine receptors and reduced prefrontal glutamatergic activity(Goldman-Rakic et al., 2004, Laruelle et al., 2003). There is also evidence for synaptic changes in many brain regions, reflecting a reorganisation of synaptic inputs and reductions in dendritic length, spine density and arborisation of receptive cells. These changes have been observed in the prefrontal and temporal cortex, hippocampus and caudate nucleus (Black et al., 2004, Garey et al., 1998, Glantz and Lewis, 2000, Kung et al., 1998, Rosoklija et al., 2000, Selemon and Goldman-Rakic, 1999) and constitute a major feature of schizophrenic pathology. Decreases in glutamate decarboxylase expression (GAD67/GAD1) in subsets of parvalbumin containing GABAergic neurones have also been observed in the frontal cortex (Lewis et al., 2004). Schizophrenia is also characterised by astrocytic malfunction and oligodendrocyte cell loss. A summary of multiple studies of prefrontal cortical tissue from schizophrenic brains from the Stanley consortium highlighted a consensus for a decrease in GFAP levels and oligodendrocyte number, a feature also observed in major depression and bipolar disorder (Knable et al., 2001, Uranova et al., 2001, Uranova et al., 2004). Apoptosis and necrosis of oligodendrocytes has been reported in the frontal cortex and caudate nucleus in both schizophrenia and bipolar disorder (Uranova et al., 2001).

Decreases in oligodendrocyte density of ∼ 20–30% have been observed in various frontal cortical regions in schizophrenia (Hof et al., 2002, Hof et al., 2003, Uranova et al., 2004) (Vostrikov et al., 2004). The immunoreactivity of the oligodendrocyte-associated markers 2′,3′-cyclic nucleotide 3′-phosphodiesterase (CNP) and myelin-associated glycoprotein (MAG)is reduced by a similar magnitude (Flynn et al., 2003). Microarray studies have also demonstrated a generalised reduction in the expression of oligodendrocyte and myelinisation associated gene messages in the schizophrenic brain (Hakak et al., 2001, Katsel et al., 2005a).

Schizophrenia is also characterised by a large reduction in cerebral and CSF glutathione levels in life (Do et al., 2000). Levels of 5-S-cysteinyl DOPAC and 5-S-cysteinyl dopamine, the end products of dopamine-derived quinone metabolism are increased in the schizophrenic brain (Carlsson et al., 1994) and a number of publications suggest that free radicals and oxidative stress are implicated in schizophrenia pathology (Mahadik and Scheffer, 1996, Reddy and Yao, 1996, Yao et al., 2001). Recently, a microarray and proteomics study showed that many of the genes and proteins whose expression is modified in the schizophrenic brain are related to glutathione and oxidative stress pathways (Prabakaran et al., 2004).

In parallel with the research that has led to these hypotheses and observations, genetic studies have identified over 130 putative (and highly contested) susceptibility genes that may predispose to schizophrenia in some populations. These genes are the combined result of directed research related to the major schizophrenia hypotheses and of gene selection following genome-wide scans of chromosomes with closely spaced polymorphic markers. This review outlines the function of some of these genes and suggests that they may be organised into a signaling network whose dysfunction may play a key role in schizophrenia.