Evaluating the comparability of gene expression in blood and brain

Abstract

The availability of an accessible tissue whose gene expression profile is similar to more inaccessible CNS tissues has the potential to advance research in neuropsychiatric disorders. We conducted secondary data analysis of transcriptional profiling of 79 human tissues for 33,698 genes using the Affymetrix U133A microarray augmented with a custom microarray (Affymetrix GNF1H), which were produced by the Genomics Institute of the Novartis Research Foundation (http://symatlas.gnf.org). Our analyses suggested that: (a) on a transcriptome level, whole blood shares significant gene expression similarities with multiple CNS tissues; (b) the median non-parametric correlation between transcripts present in both whole blood and CNS was around 0.5; (c) this correlation of 0.5 was intermediate relative to all tissues in the Novartis data set--less than for the maximum achievable value of 0.85, less than a set of immune tissues (0.64), comparable to a heterogeneous set of somatic tissues (0.57) but greater than muscle (0.48) and peripheral nervous system tissues (0.36); (d) about half of a set of candidate genes relevant to schizophrenia were expressed in both whole blood and prefrontal cortex; and (e) the expression levels of many classes of biologically relevant processes were not significantly different between whole blood and prefrontal cortex. These analyses suggest that gene expression in whole blood is neither perfectly correlated and useful nor perfectly uncorrelated and useless with gene expression in multiple brain tissues. This suggests that the cautious and thoughtful use of peripheral gene expression may be a useful surrogate for gene expression in the CNS when it has been determined that the relevant gene is expressed in both.

Dexamethasone Stimulated Gene Expression in Peripheral Blood is a Sensitive Marker for Glucocorticoid Receptor Resistance in Depressed Patients

“…Although gene expression profiles in peripheral blood cells are not likely to reflect profiles in neuronal cell populations, there is an overlap between gene activity measures in brain and peripheral blood. Sullivan et al (2006) found a median correlation of 0.5 between transcripts expressed in both whole blood and the central nervous system and a study by (Rollins et al (2010) showed that over 4000 transcripts are co-expressed in blood cells and postmortem brain tissue. Additionally, glucocorticoids have effects on peripheral blood cells and the hypothalamus-pituitary adrenal (HPA) axis, which is perturbed during depressive episodes of many patients (Gladkevich et al, 2004).…”

Are retinoids potential therapeutic agents in disorders of social cognition including autism?

“…One group of biomarkers for neuropsychiatric disorders that shows considerable promise is blood gene expression profiling [108][109][110][111][112][113]. Most of the studies to date have focused on human lymphocytes gene expression profiling, comparison between illness groups and normal controls, and cross-matching with human postmortem brain gene expression data.…”

Dysbindin-1 and NRG-1 gene expression in immortalized lymphocytes from patients with schizophrenia

“…The gene expression profile of whole blood has been shown to have moderate degree of similarity to that of the central nervous system among 79 human tissues. 21 For these reasons, immortalized lymphocytes are good tools for determining the effect of genetic risks or drug treatment on gene expressions. Disrupted in schizophrenia 1 and a serotonin transporter polymorphism have been reported to have effect on their gene expressions in immortalized lymphocytes from bipolar disorder patients.…”