SRRM2, a potential blood biomarker revealing high alternative splicing in Parkinson's disease

Abstract

Background: Parkinson's disease (PD) is a progressive neurodegenerative disorder that affects about five million people worldwide. Diagnosis remains clinical, based on phenotypic patterns. The discovery of laboratory markers that will enhance diagnostic accuracy, allow pre-clinical detection and tracking of disease progression is critically needed. These biomarkers may include transcripts with different isoforms.

Methodology/principal findings: We performed extensive analysis on 3 PD microarray experiments available through GEO and found that the RNA splicing gene SRRM2 (or SRm300), sereine/arginine repetitive matrix 2, was the only gene differentially upregulated among all the three PD experiments. SRRM2 expression was not changed in the blood of other neurological diseased patients versus the healthy controls. Using real-time PCR, we report that the shorter transcript of SRRM2 was 1.7 fold (p = 0.008) upregulated in the substantia nigra of PDs vs controls while the longer transcript was 0.4 downregulated in both the substantia nigra (p = 0.03) and amygdala (p = 0.003). To validate our results and test for the possibility of alternative splicing in PD, we performed independent microarray scans, using Affymetrix Exon_ST1 arrays, from peripheral blood of 28 individuals (17 PDs and 11 Ctrls) and found a significant upregulation of the upstream (5') exons of SRRM2 and a downregulation of the downstream exons, causing a total of 0.7 fold down regulation (p = 0.04) of the long isoform. In addition, we report novel information about hundreds of genes with significant alternative splicing (differential exonic expression) in PD blood versus controls.

Conclusions/significance: The consistent dysregulation of the RNA splicing factor SRRM2 in two different PD neuronal sources and in PD blood but not in blood of other neurologically diseased patients makes SRRM2 a strong candidate gene for PD and draws attention to the role of RNA splicing in the disease.

Figure 1. Heat map and venn diagrams of differentially expressed genes overlapping in multiple PD studies.

A. Shown are the fold change expression levels of genes differentially expressed in at least 2/3 PD studies: (1) up in both the substantia nigra (SN) and 4-wk rotenone-treated cells (RT), (2) down in substantia nigra and blood, (3) down in 4-wk rotenone-treated cells and blood. Arrows point to selected upregulated (red) and downregulated (blue) genes. B. Differentially expressed genes determined by our analysis of 3 different microarray experiments on PD, each from a different tissue source, were compared to find common downregulated (left) and upregulated (right) transcripts. A 1.2 fold change and p<.05 significance cut-offs were used. No multiple correction was employed.

Figure 2. Alternate isoforms of SRRM2 with different number of exons and different expression levels in postmortem PD brain.

A. Two main splice variants of SRRM2 differ at their 3′ end. The longer SRRM2 isoform contains 15 exons and the shorter isoform contains 11 exons. B. SRRM2 isoforms were differentially expressed in postmortem PD brain regions. The shorter transcript of SRRM2 was 1.7 fold (p = 0.008) upregulated in the SN of PDs versus controls while the longer transcript was 0.4 fold downregulated in both the SN (p = 0.03) and Amygdala (p = 0.003) of PDs versus controls.

Figure 3. Differential Exonic expression in PD Blood.

Splicing Analysis reveals significant differential exonic expression within 218 genes (p<0.05 with Benjamini Hochberg FDR correction). Shown here are SRRM2, ADIPOR1, TKT, and SLC4A1, all four genes were also differentially expressed in the Scherzer blood data (p<0.05 with no multiple correction). Each probeset represents an exon. Note that in SRRM2, the 5′ exons are upregulated while the downstream exons are downregulated in PD patients versus controls.

Figure 4. Differential gene expression in 2 PD blood studies.

A. Shown are top 10 genes differentially expressed by at least 2.5 fold in PD blood versus controls according to analysis of our new expression arrays (p<0.05 with no multiple correction). B. 35 differentially expressed transcripts by at least 1.2 fold overlap between our new PD blood profiling and the Scherzer blood dataset (p<0.05 with no multiple correction).