Multiregional gene expression profiling identifies MRPS6 as a possible candidate gene for Parkinson's disease

Abstract

Combining large-scale gene expression approaches and bioinformatics may provide insights into the molecular variability of biological processes underlying neurodegeneration. To identify novel candidate genes and mechanisms, we conducted a multiregional gene expression analysis in postmortem brain. Gene arrays were performed utilizing Affymetrix HG U133 Plus 2.0 gene chips. Brain specimens from 21 different brain regions were taken from Parkinson's disease (PD) (n = 22) and normal aged (n = 23) brain donors. The rationale for conducting a multiregional survey of gene expression changes was based on the assumption that if a gene is changed in more than one brain region, it may be a higher probability candidate gene compared to genes that are changed in a single region. Although no gene was significantly changed in all of the 21 brain regions surveyed, we identified 11 candidate genes whose pattern of expression was regulated in at least 18 out of 21 regions. The expression of a gene encoding the mitochondria ribosomal protein S6 (MRPS6) had the highest combined mean fold change and topped the list of regulated genes. The analysis revealed other genes related to apoptosis, cell signaling, and cell cycle that may be of importance to disease pathophysiology. High throughput gene expression is an emerging technology for molecular target discovery in neurological and psychiatric disorders. The top gene reported here is the nuclear encoded MRPS6, a building block of the human mitoribosome of the oxidative phosphorylation system (OXPHOS). Impairments in mitochondrial OXPHOS have been linked to the pathogenesis of PD.

Figure 1

The regional brain sample tiling chart. Columns represent gene chips from individual subjects and rows illustrate gene chips per regional comparison. Data from a total of 499 gene chips (283 from 22 PD patients and 216 from 23 aged control subjects) represent a sample from 21 regions that passed both the RNA and microarray quality control.

Figure 2

RNA quality control (QC) parameters for the substantia nigra. The relationships between RNA QC (determined by the average of the 5′/3′ signal ratios of β-actin and GADPH across Plus 2.0 chips) and the postmortem interval (PMI) in SN show no effect of the PMI on the QC measurements.

Figure 3

Frequency pyramid illustrates the number of candidate genes differentially expressed in PD versus normal aged control subjects. Four genes topped the pyramid (20 out of 21 regions surveyed).

Figure 4

Comparison of MRPS6 gene expression in PD patients versus normal aged controls in selected regions: (A) Raw Affymetrix data show significant upregulation of MRPS6 gene expression in the caudate and the insula. (B) Quantitative real-time RT-PCR gave comparable results of MRPS6 expression in both regions. *p ≤ 0.001.