Gene expression and neuronal activity in schizophrenia: a study of polyadenylated mRNA in the hippocampal formation and cerebral cortex

Changed relative to what? Housekeeping genes and normalization strategies in human brain gene expression studies

Many studies in biological psychiatry compare the abundance of individual messenger RNAs between cases and control subjects or, more recently, between genotype groups. Most utilize some form of normalization procedure, usually expressing the transcript(s) of interest relative to that of a housekeeping gene or genes (also called reference genes), to overcome various sources of experimental error. Indeed, normalization is such a standard procedure that its purpose, principles, and limitations are sometimes overlooked, and some papers lack sufficient information as to its implementation. Here, we review the rationales for normalization and argue that in well-conducted psychiatric gene expression studies using human brain tissue, it is reducing intersubject variability rather than experimental error that is the major benefit of normalization. We also review the conceptual and empirical basis for the category of housekeeping genes-i.e., genes with a ubiquitous and invariant expression. We conclude that the evidence is against any such simple categorization and that a more pragmatic, less dogmatic, approach to the selection and implementation of reference genes is required, which takes into account the particular issues that pertain to human brain tissue studies. This pragmatism extends to the issue of whether normalization should be to one or multiple reference genes. We end by making several recommendations toward a more flexible, transparent, and comprehensive approach to data presentation and analysis. We illustrate the review with examples from studies of schizophrenia and mood disorder.

Variations in differential gene expression patterns across multiple brain regions in schizophrenia

Large-scale gene expression studies in schizophrenia (SZ) have generally focused on the dorsolateral prefrontal cortex. Despite a wealth of evidence implicating multiple other brain regions in the disease, studies of other brain regions have been less frequent and have rarely been performed in the same subjects. We analyzed postmortem gene expression in the frontal, cingulate, temporal, parietal and occipital cortices (Brodmann areas 8, 10, 44, 46, 23/31, 24/32, 20, 21, 22, 36/28, 7 and 17, respectively) as well as in the hippocampus, caudate nucleus and putamen of persons with schizophrenia and control subjects (N's = 13) using Affymetrix GeneChip microarrays. Under identical data filtering conditions, the superior temporal cortex (BA22) of schizophrenia subjects showed the maximal number of altered transcripts (approximately 1200) compared to controls. Anterior and posterior cingulate cortices (BA23/31, 24/32) and the hippocampus followed the superior temporal cortex with two-times lower numbers of altered transcripts. The dorsolateral prefrontal cortex (BA46), a frequent target of SZ-associated studies, showed substantially fewer altered transcripts (approximately 33). These regional differences in differentially expressed genes could not be accounted for by factors such as total numbers of genes expressed or the filtering conditions and criteria used for identification of differentially expressed genes. These findings suggest that the temporal and cingulate cortices and the hippocampal formation represent brain regions of particular abnormality in SZ and may be more susceptible to the disease process(es) than other regions thus far studied.

Decreased expression of vesicular glutamate transporter 1 and complexin II mRNAs in schizophrenia: further evidence for a synaptic pathology affecting glutamate neurons

Synaptic protein gene expression is altered in schizophrenia. In the hippocampal formation there may be particular involvement of glutamatergic neurons and their synapses, but overall the profile remains unclear. In this in situ hybridization histochemistry (ISHH) study, we examined four informative synaptic protein transcripts: vesicular glutamate transporter (VGLUT) 1, VGLUT2, complexin I, and complexin II, in dorsolateral prefrontal cortex (DPFC), superior temporal cortex (STC), and hippocampal formation, in 13 subjects with schizophrenia and 18 controls. In these areas, VGLUT1 and complexin II are expressed primarily by excitatory neurons, whereas complexin I is mainly expressed by inhibitory neurons. In schizophrenia, VGLUT1 mRNA was decreased in hippocampal formation and DPFC, complexin II mRNA was reduced in DPFC and STC, and complexin I mRNA decreased in STC. Hippocampal VGLUT1 mRNA declined with age selectively in the schizophrenia group. VGLUT2 mRNA was not quantifiable due to its low level. The data provide additional evidence for a synaptic pathology in schizophrenia, in terms of a reduced expression of three synaptic protein genes. In the hippocampus, the loss of VGLUT1 mRNA supports data indicating that glutamatergic presynaptic deficits are prominent, whereas the pattern of results in temporal and frontal cortex suggests broadly similar changes may affect inhibitory and excitatory neurons. The impairment of synaptic transmission implied by the synaptic protein reductions may contribute to the dysfunction of cortical neural circuits that characterises the disorder.

The hippocampus in schizophrenia: a review of the neuropathological evidence and its pathophysiological implications

This paper puts the case for the hippocampus as being central to the neuropathology and pathophysiology of schizophrenia. The evidence comes from a range of approaches, both in vivo (neuropsychology, structural and functional imaging) and post mortem (histology, morphometry, gene expression, and neurochemistry). Neuropathologically, the main positive findings concern neuronal morphology, organisation, and presynaptic and dendritic parameters. The results are together suggestive of an altered synaptic circuitry or "wiring" within the hippocampus and its extrinsic connections, especially with the prefrontal cortex. These changes plausibly represent the anatomical component of the aberrant functional connectivity that underlies schizophrenia. Glutamatergic pathways are prominently but not exclusively affected. Changes appear somewhat greater in the left hippocampus than the right, and CA1 is relatively uninvolved compared to other subfields. Hippocampal pathology in schizophrenia may be due to genetic factors, aberrant neurodevelopment, and/or abnormal neural plasticity; it is not due to any recognised neurodegenerative process. Hippocampal involvement is likely to be associated with the neuropsychological impairments of schizophrenia rather than with its psychotic symptoms.

Quantitative and qualitative changes in gene expression patterns characterize the activity of plaques in multiple sclerosis

Multiple sclerosis (MS) is a complex autoimmune disorder of the CNS with both genetic and environmental contributing factors. Clinical symptoms are broadly characterized by initial onset, and progressive debilitating neurological impairment. In this study, RNA from MS chronic active and MS acute lesions was extracted, and compared with patient matched normal white matter by fluorescent cDNA microarray hybridization analysis. This resulted in the identification of 139 genes that were differentially regulated in MS plaque tissue compared to normal tissue. Of these, 69 genes showed a common pattern of expression in the chronic active and acute plaque tissues investigated (Pvalue<0.0001, rho=0.73, by Spearman's rho analysis); while 70 transcripts were uniquely differentially expressed (> or = 1.5-fold) in either acute or chronic active tissues. These results included known markers of MS such as the myelin basic protein (MBP) and glutathione S-transferase (GST) M1, nerve growth factors, such as nerve injury-induced protein 1 (NINJ1), X-ray and excision DNA repair factors (XRCC9 and ERCC5) and X-linked genes such as the ribosomal protein, RPS4X. Primers were then designed for seven array-selected genes, including transferrin (TF), superoxide dismutase 1 (SOD1), glutathione peroxidase 1 (GPX1), GSTP1, crystallin, alpha-B (CRYAB), phosphomannomutase 1 (PMM1) and tubulin beta-5 (TBB5), and real time quantitative (Q)-PCR analysis was performed. The results of comparative Q-PCR analysis correlated significantly with those obtained by array analysis (r=0.75, Pvalue<0.01, by Pearson's bivariate correlation). Both chronic active and acute plaques shared the majority of factors identified suggesting that quantitative, rather than gross qualitative differences in gene expression pattern may define the progression from acute to chronic active plaques in MS.

MOR1 receptor mRNA expression in human brains of drug-related fatalities—a real-time PCR quantification

The expression of the human micro-opiate receptor (MOR1) in post mortem human brain tissue was examined using real-time PCR technology. Tissue samples from 11 fatalities due to opiate overdose and five normal subjects with different causes of death were analysed in order to elucidate whether chronic opiate abuse is followed by a regulation of MOR1 expression. In each case nine selected brain regions (thalamus, caudate nucleus, hypothalamus, ventral tegmentum, hippocampus, amygdala, frontal cortex, nucleus accumbens, putamen) were evaluated. The MOR1-mRNA level was determined relative to the housekeeping gene beta2-microglobulin. While in most regions the MOR mRNA levels in the brain of addicts were not different from the control group-with varying levels between 0 and 15% of housekeeping gene level-in the brains of three drug-related fatalities an enormous increase was encountered in the thalamus where the MOR-mRNA level amounted for up to 10,000% of the measured housekeeping gene level. The results obtained by toxicological hair analysis in the group of drug-related fatalities indicate that the enormous thalamic MOR1-expression is primarily found in individuals who died from acute heroin overdose but did not show signs of a substantial chronic administration of the drug. Further studies have to be performed to evaluate if the observed MOR1-mRNA up-regulation in the thalamus in a subpopulation of acute lethal intoxications mirrors a state of functional hypersensitivity associated with the occurrence of death.

Asymmetrical reductions of hippocampal NMDAR1 glutamate receptor mRNA in the psychoses

The psychotomimetic properties of NMDA glutamate receptor antagonists suggest there may be disease related changes of this receptor in schizophrenia. Using in situ hybridisation histochemistry (ISHH), we measured mRNA for the obligatory NMDAR1 subunit of the NMDA glutamate receptor in post-mortem samples of hippocampus from schizophrenics, depressives, bipolar patients and normal controls. A significant main effect of diagnosis was observed in the dentate gyrus (ANOVA, p = 0.004) and a trend in the CA3 region (ANOVA, p = 0.06), with all psychiatric groups having reduced NMDAR1 mRNA levels compared to normal controls. In contrast to the affectively ill groups, the reductions in schizophrenics were more pronounced in the left side compared to the right. Expression of poly A mRNA also showed left-sided losses in the dentate gyrus in schizophrenia but reductions in NMDAR1 remained significant when expressed as a ratio of poly A. The findings confirm a recent report of reduced hippocampal NMDAR1 mRNA in schizophrenia. However, our new evidence suggests that this is a feature of both affective and schizophrenic disorders and that schizophrenia is distinguished from the others by left-sided reductions in hippocampal NMDAR1 gene expression.

Gene expression profiling in the post-mortem human brain--no cause for dismay

Global expression profiling techniques such as microarray technology promise to revolutionize biology. Soon it will be possible to investigate alterations at the transcript level of the entire human genome. There is great hope that these techniques will at last shed light on the pathological processes involved in complex neuropsychiatric disorders such as schizophrenia. These scientific advances in turn have re-kindled a great interest and demand for post-mortem brain tissue. Good quality post-mortem tissue undoubtedly is the fundamental prerequisite to investigate complex brain disorders with molecular profiling techniques. In this review we show that post-mortem brain tissue can yield good quality mRNA and intact protein antigens which allow the successful application of traditional molecular biology methods as well as novel profiling techniques. We also consider the use of laser-capture microdissection on post-mortem tissue. This recently developed technique allows the experimenter to explore the molecular basis of cellular function at the single cell level. The combination of laser-capture microdissection with high throughput profiling techniques offers opportunities to obtain precise genetic fingerprints of individual neurons allowing comparisons of normal and pathological states.

Molecular characterization of schizophrenia viewed by microarray analysis of gene expression in prefrontal cortex

Microarray expression profiling of prefrontal cortex from matched pairs of schizophrenic and control subjects and hierarchical data analysis revealed that transcripts encoding proteins involved in the regulation of presynaptic function (PSYN) were decreased in all subjects with schizophrenia. Genes of the PSYN group showed a different combination of decreased expression across subjects. Over 250 other gene groups did not show altered expression. Selected PSYN microarray observations were verified by in situ hybridization. Two of the most consistently changed transcripts in the PSYN functional gene group, N-ethylmaleimide sensitive factor and synapsin II, were decreased in ten of ten and nine of ten subjects with schizophrenia, respectively. The combined data suggest that subjects with schizophrenia share a common abnormality in presynaptic function. We set forth a predictive, testable model.

Hippocampal synaptic pathology in schizophrenia, bipolar disorder and major depression: a study of complexin mRNAs

Complexin (cx) I and cx II are synaptic proteins preferentially expressed by inhibitory and excitatory hippocampal neurons respectively. We previously reported decreased hippocampal formation cx mRNA and protein expression in schizophrenia, with a greater loss of cx II than cx I. The present in situ hybridization study was both an attempt at replication, and an extension to include bipolar and unipolar mood disorders, using sections from the Stanley Foundation brain series. In schizophrenia, both mRNAs were decreased in some hippocampal subfields, especially CA4, but were preserved in subiculum. The cx II/cx I mRNA ratio was unchanged. In bipolar disorder, the mRNAs were reduced in CA4, subiculum and parahippocampal gyrus, with the deficit in subiculum being diagnostically specific. No alterations in cx mRNAs were found in major depression. Treatment of rats with antipsychotics (haloperidol or chlorpromazine) for 2 weeks had no effect on hippocampal cx mRNAs. These data replicate the finding of decreased cx I and cx II expression in the hippocampus in schizophrenia and show a similar or greater abnormality in bipolar disorder. Non-replication of the cx II > cx I mRNA loss in schizophrenia means that the hypothesis of a preferential involvement of excitatory connections was not supported. The results extend the emerging evidence that altered circuitry may be a component of the neuroanatomy of both schizophrenia and bipolar mood disorder.

Synaptophysin gene expression in schizophrenia. Investigation of synaptic pathology in the cerebral cortex

BACKGROUND

Decreased expression of proteins such as synaptophysin in the hippocampus and prefrontal cortex in schizophrenia is suggestive of synaptic pathology. However, the overall profile of changes is unclear.

AIMS

To investigate synaptophysin gene expression in the cerebral cortex in schizophrenia.

METHOD

The dorsolateral prefrontal (Brodmann area [BA] 9/46), anterior cingulate (BA 24), superior temporal (BA 22) and occipital (BA 17) cortex were studied in two series of brains, totalling 19 cases and 19 controls. Synaptophysin was measured by immunoautoradiography and immunoblotting. Synaptophysin messenger RNA (mRNA) was measured using in situ hybridisation.

RESULTS

Synaptophysin was unchanged in schizophrenia, except for a reduction in BA 17 of one brain series. Synaptophysin mRNA was decreased in BA 17, and in BA 22 in the women with schizophrenia. No alterations were seen in BA 9/46.

CONCLUSIONS

Synaptophysin expression is decreased in some cortical areas in schizophrenia. The alterations affect the mRNA more than the protein, and have an unexpected regional distribution. The characteristics of the implied synaptic pathology remain to be determined.

Detection and quantification of hippocampal synaptophysin messenger RNA in schizophrenia using autoclaved, formalin-fixed, paraffin wax-embedded sections

Most in situ hybridization histochemistry studies of messenger RNA in human brain have been carried out on frozen tissue. Recently, autoclaving has been reported to enable routinely processsed material to be used for in situ localization of messenger RNA. We have investigated whether autoclaving also permits in situ hybridization histochemistry to be used quantitatively. To do this, we targeted synaptophysin messenger RNA with a 35S-labelled oligonucleotide probe in autoclaved, formalin-fixed, paraffin wax-embedded sections of the hippocampal formation of 11 schizophrenics and 11 controls. We compared the results with those seen on frozen sections from adjacent blocks, which had been used previously to demonstrate a loss of the messenger RNA in schizophrenia. Synaptophysin messenger RNA was readily detected in the autoclaved sections. The hybridization signal correlated strongly with that seen in the frozen sections. We found a similar pattern and magnitude of decreased synaptophysin messenger RNA in schizophrenia in the autoclaved sections as we had in the frozen sections, including the selective preservation of synaptophysin messenger RNA in CA1. The reduction of synaptophysin messenger RNA was replicated when six subjects with schizophrenia not included in the earlier study were considered separately. We conclude that autoclaving renders formalin-fixed, paraffin wax-embedded sections of human brain suitable for quantitative in situ hybridization histochemistry. This has considerable implications, given the wider availability, better morphology and easier handling of fixed than frozen human brain tissue. Using this material, we confirmed the finding of decreased synaptophysin messenger RNA in the hippocampal formation in schizophrenia, furthering the evidence for synaptic pathology in this region in the disorder.

Hippocampal and cortical growth-associated protein-43 messenger RNA in schizophrenia

Growth-associated protein-43 is involved in maturational and plasticity-associated processes, and changes in growth-associated protein-43 expression are a marker of altered plasticity following experimental and neuropathological lesions. Using in situ hybridization, we have investigated growth-associated protein-43 mRNA in the medial temporal lobe and cerebral cortex in 11 normal subjects and 11 matched subjects with schizophrenia, a disorder in which perturbed neurodevelopment and aberrant plasticity are implicated. In the schizophrenia group, growth-associated protein-43 messenger RNA was decreased in the medial temporal lobe, primary visual cortex and anterior cingulate gyrus, but was unaltered in the superior temporal and dorsolateral prefrontal cortices. Correlations of growth-associated protein-43 messenger RNA signal between areas were stronger and more numerous in the schizophrenics than in the controls, suggesting a more global regulation of growth-associated protein-43 expression. Finally, the ratio of growth-associated protein-43 messenger RNA to synaptophysin messenger RNA--a putative index of the production of new synapses--was decreased in the medial temporal lobe in the schizophrenics. Our findings imply that neuronal plasticity as indexed by growth-associated protein-43 expression is impaired, and perhaps aberrantly regulated, in schizophrenia. The data support the emerging view that the disease pathophysiology is one which affects the hippocampal and cortical circuitry and that the abnormalities are reflected in the altered expression of specific neuronal genes.

Temporal cortex synaptophysin mRNA is reduced in Alzheimer's disease and is negatively correlated with the severity of dementia

We measured synaptophysin mRNA in neocortical tissue from 7 prospectively assessed, pathologically verified normal individuals, 17 subjects with Alzheimer's disease (AD), and 13 subjects with a non-AD dementia. In temporal cortex (Brodmann area 21), synaptophysin mRNA was decreased in AD and non-AD dementia groups compared to controls. The loss was also present relative to polyadenylated mRNA content. Synaptophysin mRNA signal correlated negatively with the degree of dementia and negatively with the pathological severity of AD. In occipital cortex (Brodmann area 17) there were no differences between groups nor clinicopathological correlations. These data extend the evidence for a regional synaptic pathology in AD which affects synaptic protein gene expression by temporal cortex neurons.