Sage Journals: Discover world-class research

Abstract

Background:

With an estimated 80% heritability, molecular genetic research into schizophrenia has remained inconclusive. Recent large-scale, genome-wide association studies only identified a small number of susceptibility genes with individually very small effect sizes. However, the variable expression of the phenotype is not well captured in diagnosis-based research as well as when assuming a ‘heterogenic risk model’ (as apposed to a monogenic or polygenic model). Hence, the expression of susceptibility genes in response to environmental factors in concert with other disease-promoting or protecting genes has increasingly attracted attention.

Method:

The current review summarises findings of microarray gene expression research with relevance to schizophrenia as they emerged over the past decade.

Results:

Most findings from post mortem, peripheral tissues and animal models to date have linked altered gene expression in schizophrenia to presynaptic function, signalling, myelination, neural migration, cellular immune mechanisms, and response to oxidative stress consistent with multiple small effects of many individual genes. However, the majority of results are difficult to interpret due to small sample sizes (i.e. potential type-2 errors), confounding factors (i.e. medication effects) or lack of plausible neurobiological theory.

Conclusion:

Nevertheless, microarray gene expression research is likely to play an important role in the future when investigating gene/gene and gene/environment interactions by adopting a neurobiologically sound theoretical framework.

Keywords

Animal models blast cells hippocampus lymphocytes molecular genetics post mortem brain tissue prefrontal cortex RNA temporal cortex transcript

Introduction

Schizophrenia is a low prevalence disorder affecting 0.7% to 1.4% of the population in a wide variety of geographic regions across the world (Jablensky et al., 1992). The heritability is high and estimated to contribute up to 83% (Canon et al., 1998; Lewis and Lieberman, 2000). However, the molecular genetics have not been identified despite recent large-scale genome-wide association studies involving up to 27,000 individuals (International Schizophrenia Consortium, 2009; Shi et al., 2009; Stefansson et al., 2009). The largest genome-wide association study into schizophrenia to date confirmed two (6p21.32-p22.1 and 18q21.2) and identified five novel loci (1p21.3, 2q32.3, 8p23.2, 8q21.3 and 10q24.32-q24.33) when investigating 51,695 individuals (Ripke et al., 2011). A main obstacle is the clinical phenomenology of schizophrenia which overlaps with other psychotic, affective and personality disorders. Hence, a ‘heterogenic risk model’ (as apposed to a monogenic or polygenic model) probably best accounts for the complex clinical syndromes and diagnostic subtypes of schizophrenia as well as the variable expression of symptoms during the course of illness (Galter et al., 2007; Jablensky, 1995).

Biomarkers or ‘endophenotypes’ of the disorder have been proposed to facilitate more targeted neurogenetic research (McGuffin et al., 1987). Promising findings are emerging from neurocognitive, neurophysiological and brain imaging research revealing a pattern of regional grey and white matter pathology that is also associated with the defining clinical, neurocognitive and pathophysiological features of the disorder (Andreasen et al., 2007; Tsuang et al., 2005).

Within a neurodevelopmental framework, the evolving phenotype is dependent on environmental factors which are impacting on the normal trajectory of brain development. This ‘gene by environment interaction’ is captured by gene expression; that is, the transcription of a gene’s DNA information into a mRNA copy to direct the synthesis of proteins via the genetic code. The transcription process itself reflects a major cellular mechanism of adaptation of cell function in its immediate environment at any given point in time. This also implies that gene expression data can capture mal-adaptation by identifying a signature of over and/or under-expressed quantities of RNA that is associated with a disease. Hence, studies into gene expression have gained substantial interests in schizophrenia research since it holds the promise of identifying a signature of a pathological mechanism at the molecular level.

The current review summarises some of the findings of microarray gene expression research with relevance to schizophrenia as they emerged over the past decade. Given the nature of this new technology – which is often applied without strong theoretical considerations – the studies will be reviewed by the sources of tissue (i.e. port mortem, peripheral tissues and animal models) rather than a theoretical framework. Nevertheless, the authors will attempt to discuss the findings in the context of patho-aetiological models as well as study limitations.

Microarray gene chip methodology

With recent technological advances it is now possible to analyse the expression of the entire genome (or ‘transcriptome’) in a single experiment. Historically, research has been limited to small numbers of genes when using, for instance, northern blot, in situ hybridisation, or polymerase chain reaction (PCR) methods (Deb et al., 1999; Yurov et al., 2001). On the other hand, ‘microarrays’ provide a platform for the simultaneous screening of a large number of genes from a variety of in vivo and post mortem tissues, including the examination of gene expression in single cells.

Microarrays were originally developed from ‘DNA microarrays’ containing complementary DNAs (cDNAs) for particular genes spotted onto nylon membranes that were then used in combination with radioactive probes to determine the levels of mRNA for those genes. In the mid-1990s, cDNAs were spotted onto glass microscope slides and this, coupled with fluorescent probes and improvements in scanning technologies, saw the development of DNA microarrays that significantly increased the number of genes that could be analysed in one single experiment. Over the past decade these technologies rapidly developed and expanded to allow the detection and characterisation of many genetic features, including single nucleotide polymorphisms (SNPs), methylation status, alternative splicing of genes, the actual gene sequence, the actual number of gene copies referred to as copy number variations (CNVs; inherited, somatic, or de novo), as well as, for instance, loss of heterozygosity in cancer cells.

Most of the microarrays are produced by photolithographic in situ oligonucleotide synthesis on a solid surface (such as a glass microscope slide), microscopic beads and plastic- or silicon-based chips, and all are sometimes referred to as ‘gene chips’. Today’s microarrays can feature a density > 750,000 probes for genes or features per array. While this technology now provides an efficient way of screening vast numbers of genes, it also poses the risk of false positive findings due to the high number of tests (i.e. each individual gene comparison constitutes a single statistical test), which are commonly performed without a priori hypothesis. However, this method certainly has its place when aiming for discovery as long as the results are independently reproducible and the findings are critically evaluated with reference to other data (often post hoc) and in the context of existing theoretical models.

One of the early analytic approaches for DNA microarrays was based on computing dendrograms of pairwise comparisons of gene expression (e.g. an individual index patient versus a matched healthy control subject) for a number of such paired comparisons within a single experiment. The resulting dendrogram represents the pattern of gene expression differences between patient and control data. Importantly, the measure taken here is the relative difference of gene expression in each individual pair as measured by a dye marker (e.g. patient sample red and control sample green) which indicates the relative expression difference within the red-green spectrum; that is, a higher reading in the red spectrum indicates a higher expression in the patient sample relative to the matched control sample by x-fold and vice versa. Pooled samples (i.e. establishing a relative measure for the average of a sample) have also been used; however, this lacks some control over potentially confounding variables, such as clinical features, duration of illness, age, sex, medication status, post mortem interval, for example, which can be used as an independent variable (e.g. old versus young, male versus female, etc.) when computing dendrograms.

Earlier methods include Serial Analysis of Gene Expression or SAGE, introduced by Velculescu et al. (1995) to describe the transcriptome in diseases like cancer in a wide range of organisms. More recent technological advances also allow for absolute expression measures which, in turn, can be used to calculate correlations. For instance, when comparing U133Plus2.0 Affymetrix microarray gene expression data of renal allograft biopsies with reverse transcriptase PCR (RT-PCR) data, Allanach et al. (2008) reported significant correlations of most RT-PCR probes with the corresponding microarray data. Nevertheless, microarray data should always be confirmed by another method, such as real-time RT-PCR, which offers a well-accepted quantitative measure of gene expression.

With the availability of different suppliers of microarray chips (e.g. Affymetrix, Illumina, Agilent, etc.), inter-laboratory and cross-platform reliability testing has become possible. Such testing suggests excellent reliability between laboratories and chips (Mao et al., 2007). However, when comparing the two high-resolution microarray platforms using DNA copy number variance as an indicator on a collection of established human melanoma cell lines, Greshock et al. (2007) reported that Affymetrix’s single nucleotide polymorphism microarrays offered better detection of dose-dependant changes in gene expression (also replicated by Mao et al., 2007) while Agilent’s 60-mer oligonucleotide microarray with probe design optimised for genomic hybridization seems to offer higher sensitivity and specificity (Greshock et al., 2007; Agilent Technologies Inc, 2001).

As such, microarray gene chip technology has matured and developed into an important tool of molecular genetic research. The technology is constantly improving as well as the analytical tools capable of identifying the potential genetic signatures in large data sets. Notwithstanding, critical evaluation of each individual finding is still paramount, last but not least due to the statistical pitfalls of potential type-2 errors.

Post mortem brain findings

An important target of investigations into gene expression is post mortem brain tissue, but one potential confounding factor is RNA degradation. To address this, Catts et al. (2005) quantified RNA over various post mortem intervals and observed good stability within the first 6 hours and acceptable RNA levels for up to 48 hours post mortem, but a significant decline thereafter. Hence, post mortem studies are nowadays generally required to cite RNA Integrity Number (RIN) as an index for potential RNA degradation. Brain pH is a significant factor that affects the quality of the RNA obtained from port mortem tissue (Bahn et al., 2001; Halim et al., 2008). Hence, post mortem intervals and tissue pH are generally reported and commonly matched when comparing patient and control data. Other commonly employed matching criteria are age, sex, co-morbidity and medication status. However, post mortem interval should not be viewed as a constant measure across cases as the environment after death may vary across cases with complex impact on levels of autolysis.

Most gene expression studies targeted grey matter while some studies investigated gene expression in immunohistochemically characterised single cells (Bahn et al., 2001). The latter approach is of particular interest since it is specific to individual cell types (Pietersen et al., 2009).

Common brain regions include: (1) the prefrontal cortex which is implicated in planning complex cognitive behaviours, decision-making and social cognition (Goldman-Rakic and Selemon, 1997; Pomarol et al., 2010; Yang and Raine, 2009); (2) the temporal lobes, including entorhinal cortex, which serves as the main interface between hippocampus and neocortex (Hargreaves et al., 2005); and (3) the hippocampus which forms a part of the limbic system and plays an important role in memory function and spatial information processing (Hargreaves et al., 2005; Harrison, 2004). All three brain regions have been implicated in schizophrenia (Pantazopoulos et al., 2010; Sivagnanasundaram et al., 2003) as well as interconnecting thalamic relay nuclei. For instance, Chu et al. (2009) screened the genome of laser-captured micro-dissected neurones of the primary thalamic relay nucleus of the dorsolateral prefrontal cortex and found IGF1-mTOR-, AKT-, RAS-, VEGF-, Wnt- and immune-related signalling, and eIF2- and proteasome-related genes were uniquely dysregulated in schizophrenia (n = 15) compared to patients with major depressive or bipolar disorder or healthy control subjects (n = 15 each).

Animal models suggest that down-regulation of oligodendrocyte transcripts in medial prefrontal cortex – thereby affecting myelinisation – is associated with impaired executive function that is commonly seen in patients with schizophrenia (Gregg et al., 2009). The prefrontal Brodmann’s area (BA) 9 was also the first human brain region to be investigated by microarray analysis by Mirnics et al. (2000). This study consisted of two groups of schizophrenia and control subjects with a total n of 11 for each matched pair that were matched for age, sex, post mortem interval, brain pH and brain storage time. Gene expression was determined using microarrays containing over 7000 features. The authors observed a significant decrease in expression of genes involved in prefrontal presynaptic function and confirmed using in situ hybridisation that the presynaptic function genes coding for N-ethylmaleimide-sensitive factor and synapsin II were consistently down-regulated. The group later also reported that the transcript-encoding regulator of G-protein signalling 4 (RGS4) was the most consistently and significantly decreased in the prefrontal cortex (BA9) in schizophrenia (n = 6) when compared to post mortem tissues of matched controls (n = 6) or patients diagnosed with major depressive disorder (n = 10) (Mirnics et al., 2001). Decreased RGS4 expression was also verified across three cortical areas of 10 subjects with schizophrenia with quantitative in situ hybridization. These findings suggest multiple small but synergistic effects on gene expression, linked to impaired nerve terminal function in schizophrenia.

Further support for this notion is derived from several findings such as the reduced expression of the synaptophysin gene in layer II stellate neurons in entorhinal cortex (Hemby et al., 2002). Synaptophysin is an integral membrane protein of small synaptic vesicles in the brain and endocrine cells and appears to be involved in organising membrane components and it therefore is easy to see how these changes could cause dysregulation of the synapse in schizophrenia. More recently, consistent expression changes were also identified in gene sets associated with synaptic vesicle recycling, transmitter release and cytoskeleton dynamics by Maycox et al. (2009) in the anterior prefrontal cortex tissue of BA10 when using high-density microarrays (30,000 features) to investigate two large schizophrenia cohorts.

A year after the first microarray study in schizophrenia, reduced expression of five oligodendrocyte and myelin-related genes were identified by Hakak et al. (2001) in prefrontal cortex (BA46) in a cohort of 12 schizophrenia patients versus matched controls. Altered expression of myelination-related genes have also been reported in other brain areas (Tkachev et al., 2007) such as CNP, MAG, OLIG2 and ErbB3 which were down-regulated throughout neocortex, hippocampus, caudate nucleus and putamen (Katsel et al., 2005). These findings are of particular relevance to schizophrenia, given that myelination is a key process of brain maturation.

In a well-controlled study of 14 schizophrenia brains, Arion et al. (2007) identified over-expression of immune chaperone function-related genes in the same cortical region. The authors used a study-specific DNA microarray platform with long oligonucleotides and multiple probes and replicates consisting of 1800 genes. The authors carefully matched their 14 patients on sex, age, post mortem interval, brain tissue pH, RNA integrity number and tissue storage time with an equal number of healthy control subjects. Another study indicated higher activity of histone de-actylase 1 in BA10 and 46 of prefrontal cortex suggestive of transcriptional repression due to increased lysine de-acetylisation in histone and non-histone proteins (Sharma et al., 2008). Kim et al. (2007) initially identified changed expression of 70 genes of which PCR confirmed a significant down-regulation of phospholipid scramblase 4 (a gene coding for a protein involved in apoptosis and blood coagulation) and empty spiracles homolog 2 transcription factor, which is involved in neurodevelopment. However, these expression changes were only observed in a cohort of suicide completers.

In a cross-study analysis of seven gene expression microarrays, Choi et al. (2008) identified 110 differentially expressed transcripts across various brain regions in 163 individuals with and without psychosis. In dorsolateral prefrontal cortex, PCR confirmed up-regulated metallothionein genes (which code for proteins with an affinity to heavy metals and glucocorticoids), while neuropeptide-related genes were found to be down-regulated. In an interesting study, Narayan et al. (2008) investigated duration of illness effects on gene expression in BA46 in three cohorts of matched pairs of schizophrenia and control subjects. The first cohort had a duration of illness of less than 4 years (n = 8 matched pairs), the second cohort between 7 and 18 years (n = 14 matched pairs) and the third of more than 28 years (n = 8 matched pairs). Early in the progression of schizophrenia, dysfunction to genes involved in gene transcription and RNA processing, vesicle-mediated transport functions, and metal ion binding was identified whilst long-term illness was associated with inflammation and immune responses amongst other various biological functions (Kasai et al., 2003). This suggested that gene expression changes as the disease progresses.

Tang et al. (2009) reported that normal aging was significantly linked to abnormalities in pathways related to synaptic function, cell-cycle/DNA damage and apoptosis. In contrast, aging in schizophrenia was significantly associated with fatty acid and steroid metabolism, but not with those functions associated with normal aging when investigating post mortem BA46 tissue from 29 schizophrenia and 30 healthy subjects aged from 19 to 81 years. Investigating the same brain region in 27 schizophrenia subjects and 27 matched healthy controls, Narayan et al. (2009) found differential expression of genes, particularly related to glycosphingolipid/sphingolipid metabolism and N- and O-linked glycan biosynthesis in schizophrenia. Expression decreases of seven genes associated with these pathways (i.e. UGT8, SGPP1, GALC, B4GALT6, SPTLC2, ASAH1 and GAL3ST1) were validated by quantitative PCR in schizophrenia subjects with short-term illness while only one of these genes (i.e. SPTLC2) was differentially expressed in chronic schizophrenia. Moreover, the expression of five of these genes was also significantly positively correlated with age in schizophrenia but not in control subjects. The authors concluded that a disruption of the sphingolipid metabolism in the early phase of illness could result in widespread downstream effects involving myelination and oligodendrocyte function, while the age-related effects may represent an adaptive response to disease progression or pharmacotherapy. The latter notion, however, was not supported by their data showing no correlation of gene expression levels with medication doses.

Torkamani et al. (2010) investigated post mortem BA9 and BA46 tissue from a large sample of 101 subjects and also confirmed differential age-related gene expression. In particular, the expression if genes related to developmental processes (i.e. neurite outgrowth, neuronal differentiation and dopamine-related cellular signalling) decreased with aging in normal subjects but not in schizophrenia subjects.

Temporal lobe neuropathology appears to be associated with some of the positive symptoms of schizophrenia (e.g. Kasai et al., 2003; Sivagnanasundaram et al., 2003), thus making the temporal cortex another prime target of molecular genetic research along with the prefrontal cortex. To this end, Hemby et al. (2002) reported decreased expression of genes associated with G-protein, glutamate and N-methyl-d-aspartate (NMDA) function in entorhinal cortex from post mortem tissue. Indeed, NMDA dysfunction is highly relevant to schizophrenia psychopathology by mediating positive and negative symptoms of the disorder (Jentsch and Roth, 1999; Sivagnanasundaram et al., 2003). Aston et al. (2004) explored expression of 12,000 genes in post mortem middle temporal gyrus tissue that was collected from the brains of 12 schizophrenia patients and 14 matched healthy control subjects. The authors observed significantly reduced expression of genes associated with myelination (i.e. MAG, PLLP [TM4SF11], PLP1 and ERBB3). They also observed altered expression of genes involved in neurodevelopment (i.e. TRAF4, Neurod1 and histone deacetylase 3), the circadian pacemaker gene (PER1) and several other regulator genes of chromatin function and signalling mechanisms. Also, genes related to immune function have been recently reported to be differentially expressed in post mortem BA22 tissue of schizophrenia patients (Schmidt et al., 2011).

More recently, Bowden et al. (2008) identified altered expression of genes involved in neurotransmission (linked to presynaptic functions), myelination and neurodevelopment when investigating post mortem brain tissue derived from the superior temporal gyrus of seven schizophrenia patients versus seven closely matched healthy control subjects. Interestingly, similarly altered expression profiles were also found in peripheral blood lymphocytes of schizophrenia patients (Bowden et al., 2006; see also below).

Subcortical brain structures that have been implicated in schizophrenia include the hippocampus and amygdala. Along with volume reduction and impaired glutamate neurotransmission, impaired function of the hippocampus has been linked to psychopathology and cognitive impairment in schizophrenia (e.g. Goff and Coyle, 2001; Heckers, 2001; Phillips et al., 2003). Chung et al. (2003) observed increased expression of chondrex (or YKL-40, which codes extracellular matrix glycoprotein involved in cell growth and migration), histamine-releasing factor, HERC2 and heat-shock 70. The histamine-releasing factor gene has been linked to negative symptoms, impaired learning and memory deficits in schizophrenia (Chung et al., 2003; Goff and Coyle, 2001) but was not confirmed by real-time RT-PCR and neither were HERC2 and heat-shock 70. Increased expression of chondrex, on the other hand, was confirmed, suggesting altered neuronal migration during neurodevelopment. In the amygdala, up-regulation of genes involved in presynaptic vesicle release (i.e. piccolo, RIMS2 and RIMS3) in the cytomatrix active zone as well as changes in genes involved in neurotransmission and myelination were observed in subjects with schizophrenia (Weidenhofer et al., 2006). In their follow-up study, Weidenhofer et al. (2009) showed that antipsychotic drug treatment was not likely to be the cause of the up-regulation of the genes involved in presynaptic release.

With hundreds of changes in gene expression being observed in different brain regions in schizophrenia, researchers more recently started to look for the mechanism(s) that might cause such widespread dysregulation of gene expression. One possibility that is gaining momentum is some evidence for increased global expression of microRNAs, which are involved in the regulation of translation of schizophrenia-related mRNA in superior temporal gyrus and dorsolateral prefrontal cortex in schizophrenia (Beveridge et al., 2010). Thus, one or several microRNA could be causing major changes in gene expression in schizophrenia.

So far, the review has summarised the evidence for an array of altered gene expression in various brain regions implicated in schizophrenia (see Table 1 for summary). Taking these post mortem findings together, these genes are likely to affect synaptic processes such as vesicle release and neurotransmitter signalling (reviewed by Eastwood and Harrison, 2001) as well as neurodevelopmental processes such as myelination and neural migration. These are all very plausible changes that could lead to schizophrenia. However, post mortem brain research is limited by the availability of suitable tissue – thus, usually resulting in small sample sizes – and it also relies on the quality and accuracy of existing medical records of deceased tissue donors. The use of peripheral tissues and cell samples for sources of RNA like blood has been investigated as an alternative to post mortem brain tissue. The choice of peripheral tissue may address some of these limitations but also raises the question of whether non-neural cells can inform about the brain (Matigian et al., 2008).

Table 1.

Gene expression studies into post mortem brain tissue in schizophrenia and mood disorders

Publication	Source of tissue	Sample characteristics	Key findings
Mirnics et al., 2000	Prefrontal cortex (BA9)	11 matched pairs of schizophrenia and control subjects	N-ethylmaleimide-sensitive factor and synapsin II down-regulated

Mirnics et al., 2001	Prefrontal cortex (BA9)	6 schizophrenia patients, 6 matched controls and 10 healthy controls; replication in another 10 schizophrenia patients	Decreased transcript encoding regulator of G-protein signalling 4 gene in schizophrenia

Hakak et al., 2001	Prefrontal cortex (BA46)	12 medicated and 4 unmedicated schizophrenia patients and 12 controls	Reduced expression of five oligodendrocyte and myelin-related genes

Hemby et al., 2002	Entorhinal cortex (BA28/34)	8 schizophrenia patients and 9 matched controls	Reduced expression of synaptophysin gene in layer II stellate neurons

Chung et al., 2003	Hippocampus	7 schizophrenia patients and 8 controls	Increased expression of the chondrex gene in schizophrenia

Aston et al., 2004	Temporal cortex (BA21)	12 schizophrenia and 14 matched controls	Significantly reduced expression of genes associated with myelination (i.e. MAG, PLLP [TM4SF11], PLP1 and ERBB3), neurodevelopment (i.e. TRAF4, Neurod1 and histone deacetylase 3) and the circadian pacemaker gene (PER1)

Katsel et al., 2005	Prefrontal, temporal and cingulate cortex, hippocampus, caudate nucleus, putamen (BA8, 10, 44, 46, 23/31, 24/32, 20, 21, 22, 36/28, 7 & 17)	13 schizophrenia patients and 13 controls	Significantly down-regulated expression of 29 genes associated with oligodendrocyte function and myelination, particularly in hippocampus and cingulate cortex

Weidenhofer et al., 2006	Amygdala	11 matched pairs of schizophrenia patients and controls	Up-regulated expression of genes in schizophrenia involved in the cytomatrix active zone, regulating membrane exocytosis 2, regulating membrane exocytosis 3 and piccolo; in vitro analysis suggests antipsychotic drug treatment was unlikely to have caused the changes in RIMS2, RIMS3 and piccolo expression observed in the amygdala in schizophrenia (Weidenhofer et al., 2009)

Tkachev et al., 2007	Prefrontal cortex (BA9)	15 matched pairs of schizophrenia patients and controls	Up-regulation of N-acetylaspartate-related genes (ASPA, FOLH1, PGCP and GRM3) in schizophrenia resulting in dysregulation of various enzymes involved in N-acetylaspartate metabolism

Arion et al., 2007	Prefrontal cortex (BA46)	14 matched pairs of schizophrenia patients and controls	Expression changes in synaptic, oligodendrocyte and signal transduction genes as well as immune/chaperone transcript up-regulation of SERPINA3, IFITM1, IFITM2, IFITM3, CHI3L1, MT2A, CD14, HSPB1, HSPA1B and HSPA1A in the schizophrenia sample

Benes et al., 2007	Hippocampus	7 schizophrenia, 7 bipolar and 7 matched control subjects	IL-1β, (GRIK2/3), TGF-β2, TGF-βR1, histone deacetylase 1 (HDAC1), death-associated protein (DAXX) and cyclin D2 (CCND2) were significantly up-regulated in schizophrenia, whereas PAX5, Runx2, LEF1, TLE1 and CCND2 were significantly down-regulated in bipolar disorder

Bowden et al., 2008	Temporal cortex (BA21)	7 schizophrenia patients and 7 matched controls	Altered expression of genes involved in neurotransmission, neurodevelopment and presynaptic function and confirming previously reported changes in schizophrenia for PCLO, Ascbg1, FXYD1 and RGS4

Sharma et al., 2008	Prefrontal cortex	16 schizophrenia, 3 schizoaffective, 18 bipolar and 27 matched controls	Confirming increased expression of histone deactylase 1 expression in schizophrenia

Choi et al., 2008	Prefrontal cortex (BA46/10, 8/9 & 6)	Cross-study analysis of 7 gene expression microarrays that include both psychosis and non-psychosis subjects of > 400 microarray samples (163 individual subjects) on 3 different Affymetrix microarray platforms	Up-regulated metallothioneins (MT1E, MT1F, MT1H, MT1K, MT1X, MT2A and MT3) and down-regulated neuropeptides (SST, TAC1 and NPY) in BA46 of 56 psychosis versus 49 non-psychosis patients

Narayan et al., 2008	Prefrontal cortex (BA46)	26 schizophrenia patients and 29 matched controls	Most pronounced altered gene expression in the early phase of illness (< 5 years since diagnosis of schizophrenia) affecting metal ion binding, RNA processing and vesicle-mediated transport, whereas long-term illness was associated with genes involved in inflammation, stimulus-response and immune functions; SAMSN1, CDC42BPB, DSC2 and PTPRE were consistently dysregulated in schizophrenia

Maycox et al., 2009	Prefrontal cortex (BA9 & 10)	28 schizophrenia patients and 23 matched controls from two cohorts	49 genes associated with synaptic vesicle recycling, transmitter release and cytoskeletal dynamics show same direction of significant disease-related change in expression across the two cohorts of schizophrenia patients

Tang et al., 2009	Prefrontal cortex (BA46)	29 schizophrenia patients and 30 matched controls	Aging in schizophrenia was significantly associated with fatty acid and steroid metabolism, but not with those functions associated with normal aging

Narayan et al., 2009	Prefrontal cortex (BA46)	27 schizophrenia patients and 27 matched controls	Decreased expression of UGT8, SGPP1, GALC, B4GALT6, SPTLC2, ASAH1 and GAL3ST1 in schizophrenia with short-term illness; only SPTLC2 confirmed for chronic disease state

Torkamani et al., 2010	Prefrontal cortex (BA46 & 9)	30 schizophrenia patients and 30 matched controls (BA46); 19 schizophrenia patients and 26 matched controls (BA9)	No normal age-related decreases in gene expression related to developmental processes, including neurite outgrowth, neuronal differentiation and dopamine-related cellular signalling in schizophrenia

Beveridge et al., 2010	Prefrontal (BA9) and temporal cortex (BA21)	36 schizophrenia patients and 30 matched controls	Significant schizophrenia-associated increase in global microRNA miR-15 family expression in both brain regions targeting previously implicated schizophrenia-linked genes (i.e. RGS4, GRM7, GRIN3A, HTR2A, RELN, VSNL1, DLG4, DRD1 and PLXNA2)

Schmitt et al., 2011	Temporal cortex (BA22)	10 schizophrenia patients and 10 matched controls	Significant down-regulation of immune-modulatory genes (i.e. PTGER4, ILS, EDG3, ILIA, LPL and CFD) in schizophrenia

Gene expression findings from peripheral tissues

In contrast to post mortem brain tissue, peripheral blood mononuclear cells (PBMC) can be easily collected from large cohorts for gene expression studies. It also has the advantage that patients can be phenotyped and followed-up longitudinally. PBMC gene expression studies have targeted putative risk factors for schizophrenia as well as potential biomarkers of treatment response and prognosis (Bartolomucci et al., 2010; Huang et al., 2008; Kennedy, 1996; Tsuang et al., 2005; Yao et al., 2008). There have also been two types of studies conducted, those using fresh PBMCs and those that have used transformed PBMCs into immortalised lymphoblastoid cell lines (LCLs). These cell transformations, however, may introduce a confound on gene expression profiles, thus making it potentially difficult to separate genuine disease effects from disease by cell culture interactions.

In 2006, Vawter et al. (2006) reported gene expression profiles from LCLs obtained from a multiplex schizophrenia pedigree containing five individuals with schizophrenia and nine unaffected family members. A total of nine genes were altered by array analysis, but only in two of those the neuropeptide Y receptor Y1 gene and the human guanine nucleotide-binding regulatory protein Go-alpha were confirmed to be significantly decreased by RT-PCR.

Following on from this were three studies of fresh PBMCs. The first used PBMCs from 28 controls, 30 individuals with schizophrenia and 16 with bipolar disorder (Tsuang et al., 2005). This study showed that gene expression profiling of RNA from PBMCs could distinguish between schizophrenia and bipolar disorder. This was supported by the second study using PBMCs from sib-pairs with schizophrenia (n = 66 siblings) and bipolar disorders (n = 10 siblings) that also detected more than 2000 genes with altered expression in schizophrenia. Interestingly, some of these genes were from functional groups involved in processes such as neurotransmission and presynaptic function (Middleton et al., 2005). In the third study, Bowden et al. (2006) observed altered expression of 18 brain-related genes in PBMC derived from 14 drug-treated schizophrenia patients versus matched healthy control subjects. These studies suggested that gene expression changes in PBMCs might reflect changes in the brain. Indeed, Sullivan et al. (2006) compared gene expression profiles from PBMCs and various brain tissues, including the prefrontal cortex, and found significant overlap in the expression of genes in the blood and the brain which, however, does not mean that these peripheral gene expression changes also translate into potential effects on brain function.

The use of fresh PBMCs still does not mitigate antipsychotic drug treatment. In order to do this, studies need to be conducted in treatment-naïve participants. To this end, when comparing PBMC gene expression of 13 drug treatment-naïve schizophrenia patients with matched healthy control volunteers, Zvara et al. (2005) found PCR-confirmed increased expression of the inwardly rectifying potassium channel (Kir2.3) and dopamine-3 receptor genes. Craddock et al. (2007) also showed prominent transcript changes in genes involved with cell-cycle machinery, intracellular signalling, oxidative stress and metabolism in schizophrenia patients (compared to controls) in a sample of six schizophrenia patients versus six control subjects by using human whole-genome microarray on freshly isolated T-cell-derived RNA. Subsequent chromosomal location analysis of altered genes showed susceptibility clusters at 1p36, 1q42 and 6p22 which had been previously implicated with schizophrenia (Sivagnanasundaram et al., 2003).

Another line of research has focused on undifferentiated cell lines or blast cells derived from olfactory neuroepithelium showing dysregulated cell-cycle and phosphatidylinositol signalling in schizophrenia and bipolar I disorder, respectively (McCurdy et al., 2006). More recently, Glatt et al. (2009) investigated alternatively spliced genes as potential biomarkers in peripheral blood mononuclear cells of 13 schizophrenia, nine bipolar and eight healthy control subjects. Their preliminary finding suggests significant interactions between diagnostic group and exon identity, with 33 genes showing differential splicing patterns between schizophrenia and bipolar disorder. The group also reported dysregulated ubiquitin proteasome pathways for psychosis and bipolar diagnostic groups in two independent samples (Bousman et al., 2010).

This line of research, however, stills lacks consistency and has not produced exclusive biomarkers or other signatures that may assist with diagnosis (see Table 2 for summary). This is mostly due to the heterogeneous nature of schizophrenia and the small sample sizes, which are prone to false-positive results. The few findings reported here also show little overlap with post mortem reports, making extrapolations to impaired brain functions difficult. Nevertheless, gene expression studies of peripheral tissues gain momentum in schizophrenia research but are very much dependent on systematic research of transient (e.g. medication effects) and more stable illness effects (i.e. associated with susceptibility genes). Importantly, the in vivo nature of this type of research allows for longitudinal tissue collection in parallel with observing the clinical phenotype.

Table 2.

Gene expression studies into peripheral blood tissues in schizophrenia and bipolar disorder

Publication	Cell type	Sample characteristics	Key findings
Tsuang et al., 2005	Leukocytes	25 schizophrenia, 7 bipolar disorder and 10 healthy controls subjects with microarray RNA and real-time PCR data	8 putative biomarker genes (APOBEC3B, ADSS, ATM, CLC, CTBP1, DATF1, CXCL1 and S100A9) discriminated schizophrenia, bipolar disorder and control samples

Middleton et al., 2005	Leukocytes	33 sex and age-matched discordant sibling pairs for schizophrenia and 5 pairs for bipolar disorder	Most significant increases in schizophrenia-associated expression in genes involved in immune and/or inflammatory function (e.g. CD14 antigen, chemokine receptor 1)

Zvara et al., 2005	Lymphocytes	13 drug-naïve/drug-free schizophrenia patients	Genes for dopamine receptor D2 (DRD2) and the inwardly rectifying potassium channel (Kir2.3) were over-expressed

Vawter et al., 2006	Lymphocytes	Multiplex pedigree study of 5 haplotype-positive individuals with schizophrenia, 5 unaffected haplotype-negative controls	AGA expression levels showed suggestive linkage to multiple markers in the haplotype; GALNT7 expression levels showed linkage to regulatory loci at 4q28.1 and in the haplotype region at 4q33−35.1

Bowden et al., 2006	Lymphocytes	14 schizophrenia and 14 matched controls	18 genes with brain-related functions were altered in schizophrenia versus controls; 4 of which, endothelial differentiation gene 2 (Edg-2), ezrin-radixin-moesin phosphoprotein 50 (EBP50), myc-associated zinc finger protein (MAZ) and tumor necrosis factor receptor 2 (TNFR2), were confirmed by relative real-time PCR

McCurdy et al., 2006	Olfactory neuroepithelium	10 schizophrenia, 8 bipolar and 9 matched control subjects	Genes associated with vesicle-mediated transport, cell proliferation, neurogenesis, apoptosis and inositol phosphate metabolism/phosphatidylinositol signalling were most significantly dysregulated in schizophrenia

Matigian et al., 2008	Lymphoblasts and fibroblasts	16 schizophrenia patients and 14 matched controls	No significantly different expression in schizophrenia versus controls

Yao et al., 2008	Blood mononuclear cells	30 first-admission schizophrenia patients and 26 healthy control subjects	Increased chemokine (C-X-C motif) ligand 1 (CXCL1) expression in schizophrenia confirming findings reported by Tsuang et al. (2005)

Glatt et al., 2009	Blood mononuclear cells	13 schizophrenia, 9 bipolar disorder and 8 healthy control subjects	Significant interaction between diagnostic group and exon identity, with 33 genes showing differential splicing patterns between schizophrenia and bipolar disorder samples

Bousman et al., 2010	Blood mononuclear cells	24 schizophrenia, 23 bipolar disorder, 25 healthy control subjects from two independent samples (United States and Taiwan)	Dysregulated ubiquitin proteasome pathways for psychosis and bipolar diagnostic groups in two independent samples which also correlate with positive symptoms (UBE2K, SIAH2 and USP2) (Bousman et al., 2010)

Animal models research

With the advances of genetically modified animals, the effects of genome alterations on gene expression have become an important area of schizophrenia research. For instance, 22q11.2 deletion syndrome probably accounts for approximately 1% of all schizophrenia patients (Bassett and Chow, 2008) whereas up to 35% of the individuals with the deletion are likely to develop schizophrenia.

Jurata et al. (2006) and Sivagnanasundaram et al. (2007) investigated a mouse model by deleting 1 million base pairs in a region of chromosome 16. The resulting Df1/+ mouse model carries a hemizygous deletion in a region which corresponds to the human 22q11.2 region. The heterozygous (Df1/+) mice display deficits in prepulse inhibition, learning and memory, consistent with the schizophrenia phenotype (Blundell et al., 2010).

Sivagnanasundaram et al. (2007) identified 159 differentially expressed genes in heterozygous (Df1/+) mice, with 12 genes mapping to the deleted region and expressed in hippocampus. Moreover, 15 genes involve signal transduction, synaptic plasticity, neuronal differentiation, microtubule assembly and a ubiquitin pathway relevant to hippocampus-mediated function as confirmed by PCR. Earlier, Jurata et al. (2006) reported significantly reduced expression of COMT and PRODH genes in hippocampal dentate granule neurons.

Rodents with ventral hippocampus lesions have also been introduced as an animal model for schizophrenia (Lipska, 2004). Wong et al. (2005) showed that treatment with the antipsychotic haloperidol normalises the expression of those genes where the expression has been altered by the lesion. Intrapartum viral infection has also been linked to a higher prevalence of schizophrenia (Wright et al., 1995). When infecting mice with influenza virus at day 18 of pregnancy, altered expression of immunoglobulin domain (Ig), secreted semaphoring 3A (Sema3a), transferrin receptor 2 (Trfr2) and Vldlr were identified by microarray and confirmed with PCR in hippocampus, prefrontal cortex and cerebellum (Fatemi et al., 2008, 2009).

Isolation rearing has also been introduced to study the effects of an environmental insult on neurodevelopment. Genes related to GABA neurotransmission and synapse structure (Gabra4, Nsf, Syn2 and Dlgh1) were found to be over-expressed in prefrontal cortex of affected rat pups along with reduced extracellular glutamate levels and under-expression of transcripts related to glutamatergic transmission and synapse integrity in the adult animals (Murphy et al., 2010).

Genetically modified animals have become an important tool when investigating the potential biological mechanisms of the pathogenesis of schizophrenia. Studying altered gene expression in these animal models is another logical step when considering the complexity of effects which are arising from single gene knock-outs. However, owing to the complex nature of the genetics of schizophrenia, single gene knock-out animal models are highly unlikely to be representative of the schizophrenia syndrome. Thus, to date, any animal model is limited to certain aspects of the disorder – such as individual aspects of micro or macro neuropathology or pathophysiology – with the complete clinical phenotype remaining an exclusive human phenomenon. Moreover, the effects of knocking out individual genes can also considerably differ between strains of mice; for instance, being potentially lethal in one strain and not having any discernable effect in another.

Moreover, gene expression changes have been reported across diagnostic categories. For instance, Kanazawa et al. (2008) reported up-regulated SELENBP1 in post mortem dorsolateral prefrontal cortex of patients diagnosed with schizophrenia and bipolar disorder whereby the expression of SELENBP1 was significantly correlated with the presence of psychosis across diagnoses. Hence, animal modelling may also need to take into account that syndrome-based diagnostic categories potentially represent a variable phenotype of a common pathological mechanism.

Conclusion

Microarray gene expression studies into schizophrenia are still inconsistent in their findings. While some findings have linked altered gene expression to presynaptic function, signalling, myelination, neural migration, cellular immune mechanisms and response to oxidative stress, the majority of results are difficult to interpret due to small sample sizes (i.e. potential type-2 errors), confounding factors (i.e. medication effects) or lack of plausible neurobiological theory. By complementing structural genetic research into the origin of schizophrenia, microarray gene expression studies are likely to play an important role in the future when investigating gene/gene and gene/environment interactions. For instance, longitudinal research into PBMC or other in vivo tissue holds the promise of identifying gene expression profiles that are associated with phenotype changes (e.g. acute versus chronic illness; prodromal versus established illness, etc.) and environmental factors (e.g. on and off medication treatment; contributing effects of substance abuse; etc.). Experimental animal models, on the other hand, are particularly useful when studying changes of gene expression in genetically modified animals, including the effects of environmental stressors on brain development. Post mortem brain research, however, relies on a scarce resource which should not be wasted when alternative methods such as in vivo tissue and animal models are available. Nevertheless, post mortem human tissue research will remain an important prerequisite to validate findings of in vivo and animal model research.

Footnotes

Funding

NK was supported by the World Health Organization, the Schizophrenia Research Institute and the University of Newcastle, and received infrastructure support from the Australian Schizophrenia Research Bank, the Hunter Medical Research Institute and New South Wales Health.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

References

Agilent Technologies Inc. (2001). Performance of Agilent Technologies 60 mer in situ synthesized oligonucleotide microarrays. Publication number: 5988–5063EN

Allanach

Mengel

Einecke

. (2008) Comparing microarray versus RT-PCR assessment of renal allograft biopsies: similar performance despite different dynamic ranges. American Journal of Transplantation 8: 1006–1015.

Andreasen

O’Leary

Flaum

. (2007) Hypofrontality in schizophrenia: distributed dysfunctional circuits in neuroleptic-naive patients. Lancet 349: 1730–1734.

Arion

Unger

Lewis

. (2007) Molecular evidence for increased expression of genes related to immune and chaperone function in the prefrontal cortex in schizophrenia. Biological Psychiatry 62: 711–721.

Aston

Jiang

Sokolov

(2004) Microarray analysis of postmortem temporal cortex from patients with schizophrenia. Journal of Neuroscience Research 77: 858–866.

Bahn

Augood

Ryan

. (2001) Gene expression profiling in the post-mortem human brain – no cause for dismay. Journal of Chemical Neuroanatomy 22: 79–94.

Bartolomucci

Pasinetti

Salton

(2010) Granins as disease-biomarkers: translational potential for psychiatric and neurological disorders. Neuroscience 170: 289–297.

Bassett

Chow

(2008) Schizophrenia and 22q11.2 deletion syndrome. Current Psychiatry Reports 10: 148–157.

Benes

Lim

Matzilevich

. Regulation of GABA cell phenotype in hippocampus of schizophrenics and bipolars. Proceedings of the National Academy of Sciences of the United States of America 104: 10,164–10,169.

10.

Beveridge

Gardiner

Carroll

. (2010) Schizophrenia is associated with an increase in cortical microRNA biogenesis. Molecular Psychiatry 15: 1176–1189.

11.

Blundell

Kaeser

Sudhof

. (2010) RIM1alpha and interacting proteins involved in presynaptic plasticity mediate prepulse inhibition and additional behaviors linked to schizophrenia. Journal of Neuroscience 30: 5326–5333.

12.

Bousman

Chana

Glatt

. (2010) Preliminary evidence of ubiquitin proteasome system dysregulation in schizophrenia and bipolar disorder: convergent pathway analysis findings from two independent samples. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics 153B: 494–502.

13.

Bowden

Scott

Tooney

(2008) Altered gene expression in the superior temporal gyrus in schizophrenia. BMC Genomics 9: 199.

14.

Bowden

Weidenhofer

Scott

. (2006) Preliminary investigation of gene expression profiles in peripheral blood lymphocytes in schizophrenia. Schizophrenia Research 82: 175–183.

15.

Cannon

Kaprio

Lönnqvist

. (1998) The genetic epidemiology of schizophrenia in a Finnish twin cohort. Archives of General Psychiatry 55: 67–74.

16.

Catts

Fernandez

. (2005) A microarray study of post-mortem mRNA degradation in mouse brain tissue. Brain Research: Molecular Brain Research 138: 164–177.

17.

Choi

Elashoff

Higgs

. (2008) Putative psychosis genes in the prefrontal cortex: combined analysis of gene expression microarrays. BMC Psychiatry 8: 87.

18.

Chu

Liu

Kemether

(2009) Thalamic transcriptome screening in three psychiatric states. Journal of Human Genetics 54: 665–675.

19.

Chung

Tallerico

Seeman

(2003) Schizophrenia hippocampus has elevated expression of chondrex glycoprotein gene. Synapse 50: 29–34.

20.

Craddock

Lockstone

Rider

. (2007) Altered T-cell function in schizophrenia: a cellular model to investigate molecular disease mechanisms. PLoS One 2: e692.

21.

Deb

Klempan

O’Reilly

. (1999). Search for retroviral related DNA polymorphisms using RAPD PCR in schizophrenia. Biochimica et Biophysica Acta 1453: 216–220.

22.

Eastwood

Harrison

(2001) Synaptic pathology in the anterior cingulate cortex in schizophrenia and mood disorders. A review and a Western blot study of synaptophysin, GAP-43 and the complexins. Brain Research Bulletin 55: 569–578.

23.

Fatemi

Folsom

Reutiman

. (2009) Prenatal viral infection of mice at E16 causes changes in gene expression in hippocampi of the offspring. European Neuropsychopharmacology 19: 648–653.

24.

Fatemi

Reutiman

Folsom

(2008) Maternal infection leads to abnormal gene regulation and brain atrophy in mouse offspring: implications for genesis of neurodevelopmental disorders. Schizophrenia Research 99: 56–70.

25.

Galter

Westerlund

Belin

. (2007) DJ-1 and UCH-L1 gene activity patterns in the brains of controls, Parkinson and schizophrenia patients and in rodents. Physiology & Behavior 92: 46–53.

26.

Glatt

Chandler

Bousman

. (2009) Alternatively spliced genes as biomarkers for schizophrenia, bipolar disorder and psychosis: a blood-based spliceome-profiling exploratory study. Current Pharmacogenomics and Personalized Medicine 7: 164–188.

27.

Goff

Coyle

(2001) The emerging role of glutamate in the pathophysiology and treatment of schizophrenia. American Journal of Psychiatry 158: 1367–1377.

28.

Goldman-Rakic

Selemon

(1997) Functional and anatomical aspects of prefrontal pathology in schizophrenia. Schizophrenia Bulletin 23: 437–458.

29.

Gregg

Herring

Naydenov

. (2009) Downregulation of oligodendrocyte transcripts is associated with impaired prefrontal cortex function in rats. Schizophrenia Research 113: 277–287.

30.

Greshock

Feng

Nogueira

. (2007) A comparison of DNA copy number profiling platforms. Cancer Research 67: 10,173–10,180.

31.

Hakak

Walker

. (2001) Genome-wide expression analysis reveals dysregulation of myelination-related genes in chronic schizophrenia. Proceedings of the National Academy of Sciences of the United States of America 98: 4746–4751.

32.

Halim

Lipska

Hyde

(2008) An increased lactate levels and reduced pH in postmortem brains of schizophrenics: medication confounds. Journal of Neuroscience Methods 169: 208–213.

33.

Hargreaves

Rao

Lee

. (2005) Major dissociation between medial and lateral entorhinal input to dorsal hippocampus. Science 308: 1792–1794.

34.

Harrison

(2004) The hippocampus in schizophrenia: a review of the neuropathological evidence and its pathophysiological implications Psychopharmacology 174: 151–162.

35.

Heckers

(2001) Neuroimaging studies of the hippocampus in schizophrenia. Hippocampus 11: 520–528.

36.

Hemby

Ginsberg

Brunk

. (2002) Gene expression profile for schizophrenia: discrete neuron transcription patterns in the entorhinal cortex. Archives of General Psychiatry 59: 631–640.

37.

Huang

Agrawal

Giacomini

. (2008) Genetics of P450 oxidoreductase: sequence variation in 842 individuals of four ethnicities and activities of 15 missense mutations. Proceedings of the National Academy of Sciences of the United States of America 105: 1733–1738.

38.

International Schizophrenia Consortium, Purcell

Wray

. (2009) Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature 460: 748–752.

39.

Jablensky

(1995) Schizophrenia: recent epidemiological issues. Epidemiologic Reviews 17: 10–20.

40.

Jablensky

Satorius

Ernberg

. (1992) Schizophrenia: manifestations, incidence and course in different cultures. A World Health Organization ten-country study. Psychological Medicine Monograph Supplement 20: 1–97.

41.

Jentsch

Roth

(1999) The neuropsychopharmacology of phencyclidine: from NMDA receptor hypofunction to the dopamine hypothesis of schizophrenia. Neuropsychopharmacology 20: 201–225.

42.

Jurata

Gallagher

Lemire

. (2006) Altered expression of hippocampal dentate granule neuron genes in a mouse model of human 22q11 deletion syndrome. Schizophrenia Research 88: 251–259.

43.

Kanazawa

Chana

Glatt

. (2008) The utility of SELENBP1 gene expression as a biomarker for major psychotic disorders: replication in schizophrenia and extension to bipolar disorder with psychosis. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics 147B: 686–689.

44.

Kasai

Shenton

Salisbury

(2003) Progressive decrease of left superior temporal gyrus gray matter volume in patients with first-episode schizophrenia. American Journal of Psychiatry 160: 156–164.

45.

Katsel

Davis

Haroutunian

(2005) Variations in myelin and oligodendrocyte-related gene expression across multiple brain regions in schizophrenia: a gene ontology study. Schizophrenia Research 79: 157–173.

46.

Kennedy

(1996) Schizophrenia genetics: the quest for an anchor. American Journal of Psychiatry 153: 1513–1514.

47.

Kim

Choi

Baykiz

. (2007) Suicide candidate genes associated with bipolar disorder and schizophrenia: an exploratory gene expression profiling analysis of post-mortem prefrontal cortex. BMC Genomics 8: 413.

48.

Lewis

Lieberman

(2000) Catching up on schizophrenia: natural history and neurobiology. Neuron 28: 325–334.

49.

Lipska

(2004) Using animal models to test a neurodevelopmental hypothesis of schizophrenia. Journal of Psychiatry & Neuroscience 29: 282–286.

50.

McCurdy

Féron

Perry

. (2006) Cell cycle alterations in biopsied olfactory neuroepithelium in schizophrenia and bipolar I disorder using cell culture and gene expression analyses. Schizophrenia Research 82: 163–173.

51.

McGuffin

Farmer

Gottesman

(1987) Is there really a split in schizophrenia? The genetic evidence. British Journal of Psychiatry 150: 581–592.

52.

Mao

Young

(2007) The application of single nucleotide polymorphism microarrays in cancer research. Current Genomics 8: 219–228.

53.

Matigian

McCurdy

Féron

(2008) Fibroblast and lymphoblast gene expression profiles in schizophrenia: are non-neural cells informative? PLoS One 3: e2412.

54.

Maycox

Kelly

Taylor

(2009) Analysis of gene expression in two large schizophrenia cohorts identifies multiple changes associated with nerve terminal function. Molecular Psychiatry 14: 1083–1094.

55.

Mexal

Berger

Adams

. (2006) Brain pH has a significant impact on human postmortem hippocampal gene expression profiles. Brain Research 1106: 1–11.

56.

Middleton

Pato

Gentile

. (2005) Gene expression analysis of peripheral blood leukocytes from discordant sib-pairs with schizophrenia and bipolar disorder reveals points of convergence between genetic and functional genomic approaches. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics 136B: 12–25.

57.

Mirnics

Middleton

Marquez

. (2000) Molecular characterization of schizophrenia viewed by microarray analysis of gene expression in prefrontal cortex. Neuron 28: 53–67.

58.

Mirnics

Middleton

Stanwood

. (2001) Disease-specific changes in regulator of G-protein signaling 4 (RGS4) expression in schizophrenia. Molecular Psychiatry 6: 293–301.

59.

Murphy

Ter Horst

Cassidy

. (2010) Temporal dysregulation of cortical gene expression in the isolation reared Wistar rat. Journal of Neurochemistry 113: 601–614.

60.

Narayan

Head

Gilmartin

. (2009) Evidence for disruption of sphingolipid metabolism in schizophrenia. Journal of Neuroscience Research 87: 278–288.

61.

Narayan

Tang

Head

. (2008) Molecular profiles of schizophrenia in the CNS at different stages of illness. Brain Research 1239: 235–248.

62.

Pantazopoulos

Woo

Lim

. (2010) Extracellular matrix-glial abnormalities in the amygdala and entorhinal cortex of subjects diagnosed with schizophrenia. Archives of General Psychiatry 67: 155–166.

63.

Phillips

Drevets

Rauch

. (2003) Neurobiology of emotion perception II: implications for major psychiatric disorders. Biological Psychiatry 54: 515–528.

64.

Pietersen

Lim

Woo

(2009) Obtaining high quality RNA from single cell populations in human postmortem brain tissue. Journal of Visualized Experiments (30), pii: 1444.

65.

Pomarol-Clotet

Canales-Rodríguez

Salvador

. (2010) Medial prefrontal cortex pathology in schizophrenia as revealed by convergent findings from multimodal imaging. Molecular Psychiatry 15: 823–830.

66.

Ripke

Sanders

Kendler

. (2011) Schizophrenia Psychiatric Genome-Wide Association Study (GWAS) Consortium Genome-wide association study identifies five new loci. Nature Genetics 43: 969-976.

67.

Schmitt

Leonardi-Essmann

Durrenberger

. (2011) Regulation of immune-modulatory genes in left superior temporal cortex of schizophrenia patients: a genome-wide microarray study. World Journal of Biological Psychiatry 12: 201–215.

68.

Sharma

Grayson

Gavin

(2008) Histone deactylase 1 expression is increased in the prefrontal cortex of Schizophrenia subjects; analysis of the National Brain Databank microarray collection. Schizophrenia Research 98: 111–117.

69.

Shi

Levinson

Duan

(2009) Common variants on chromosome 6p22.1 are associated with schizophrenia. Nature 460: 753–757.

70.

Sivagnanasundaram

Fletcher

Hubank

. (2007) Differential gene expression in the hippocampus of the Df1/+ mice: a model for 22q11.2 deletion syndrome and schizophrenia. Brain Research 1139: 48–59.

71.

Sivagnanasundaram

Mueller

Gubanov

. (2003) Genetics of schizophrenia: current strategies. Clinical Neuroscience Research 3: 5–16.

72.

Stefansson

Ophoff

Steinberg

. (2009) Common variants conferring risk of schizophrenia. Nature 460: 744–747.

73.

Sullivan

Fan

Perou

(2006) Evaluating the comparability of gene expression in blood and brain. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics 141B: 261–268.

74.

Tang

Chang

Lanigan

. (2009) Normal human aging and early-stage schizophrenia share common molecular profiles. Aging Cell 8: 339–342.

75.

Tkachev

Mimmack

Huffaker

. (2007) Further evidence for altered myelin biosynthesis and glutamatergic dysfunction in schizophrenia. International Journal of Neuropsychopharmacology 10: 557–563.

76.

Torkamani

Dean

Schork

. (2010) Coexpression network analysis of neural tissue reveals perturbations in developmental processes in schizophrenia. Genome Research 20: 403–412.

77.

Tsuang

Nossova

Yager

. (2005) Assessing the validity of blood-based gene expression profiles for the classification of schizophrenia and bipolar disorder: a preliminary report. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics 133B: 1–5.

78.

Vawter

Atz

Rollins

. (2006) Genome scans and gene expression microarrays converge to identify gene regulatory loci relevant in schizophrenia. Human Genetics 119: 558–570.

79.

Velculescu

Zhang

Vogelstein

. (1995) Serial analysis of gene expression. Science 270: 484–487.

80.

Weidenhofer

Bowden

Scott

. (2006) Altered gene expression in the amygdala in schizophrenia: up-regulation of genes located in the cytomatrix active zone. Molecular and Cellular Neuroscience 31: 243–250.

81.

Weidenhofer

Scott

Tooney

(2009) Investigation of the expression of genes affecting cytomatrix active zone function in the amygdala in schizophrenia: effects of antipsychotic drugs. Journal of Psychiatric Research 43: 282–290.

82.

Wong

Lipska

Likhodi

. (2005) Cortical gene expression in the neonatal ventral-hippocampal lesion rat model. Schizophrenia Research 77: 261–270.

83.

Wright

Takei

Rifkin

. (1995) Maternal influenza, obstetric complications, and schizophrenia. American Journal of Psychiatry 152: 1714–1720.

84.

Yang

Raine

(2009) Prefrontal structural and functional brain imaging findings in antisocial, violent, and psychopathic individuals: a meta-analysis. Psychiatry Research 174: 81–88.

85.

Yao

Schröder

Karlsson

(2008) Verification of proposed peripheral biomarkers in mononuclear cells of individuals with schizophrenia. Journal of Psychiatric Research 42: 639–643.

86.

Yurov

Vostrikov

Vorsanova

. (2001) Multicolour fluorescent in situ hybridization on post-mortem brain in schizophrenia as an approach for identification of low-level chromosomal aneuploidy in neuropsychiatric diseases. Brain & Development 23 (Suppl 1): S186–190.

87.

Zvara

Szekeres

Janka

. (2005) Over-expression of dopamine D2 receptor and inwardly rectifying potassium channel genes in drug-naive schizophrenic peripheral blood lymphocytes as potential diagnostic markers. Disease Markers 21: 61–69.

Finding the needle in the haystack: A review of microarray gene expression research into schizophrenia

Abstract

Background:

Method:

Results:

Conclusion:

Keywords

Introduction

Microarray gene chip methodology

Post mortem brain findings

Gene expression findings from peripheral tissues

Animal models research

Conclusion

Footnotes

Funding

Declaration of interest

References