Expression analysis of novel striatal-enriched genes in Huntington disease

Abstract

Selective degeneration of striatal neurons is a pathologic hallmark of Huntington disease (HD). The exact mechanism(s) behind this specific neurodegeneration is still unknown. Expression studies of diseased human post-mortem brain, as well as different mouse models exhibiting striatal degeneration, have demonstrated changes in the expression of many important genes with a large proportion of changes being observed in the striatal-enriched genes. These investigations have raised questions about how enrichment of particular transcripts in the striatum can lead to its selective vulnerability to neurodegeneration. Monitoring the expression changes of striatal-enriched genes during the course of the disease may be informative about their potential involvement in selective degeneration. In this study, we analyzed a Serial Analysis of Gene Expression (SAGE) database (www.mouseatlas.org) and compared the mouse striatum to 18 other brain regions to generate a novel list of striatal-enriched transcripts. These novel striatal-enriched transcripts were subsequently evaluated for expression changes in the YAC128 mouse model of HD, and differentially expressed transcripts were further examined in human post-mortem caudate samples. We identified transcripts with altered expression in YAC128 mice, which also showed consistent expression changes in human post-mortem tissue. The identification of novel striatal-enriched genes with altered expression in HD offers new avenues of study, leading towards a better understanding of specific pathways involved in the selective degeneration of striatal neurons in HD.

Figure 1.

Selection algorithm identified novel striatal-enriched transcripts. We defined ‘striatal-enriched’ transcripts as transcripts with dominant patterns of expression in striatum compared to 18 other brain regions. Accordingly, selection criteria were designed to maximize detection of previously known striatal-enriched genes. Enrichment of a SAGE tag in the striatum relative to other regions was determined by using a P-value score ≥ 13 at postnatal day (dpn) 84 and mean ratio of ≥2.5 at all available time points. These two filtering steps reduced the number of tags to 187 consisting of tag-sequences mapping to both known and unknown regions of the genome. Striatal enrichment seen in Allen Institute for Brain Science (AIBS) in situ database images and quantitative real-time PCR were also used for further filtration of the SAGE tags. These selection criteria led to the identification of 34 novel striatal-enriched tags which corresponded to 23 known genes, 4 uncharacterized genes and 2 novel splice variants of known genes.

Figure 2.

Examples of qRT–PCR results validating SAGE data. qRT–PCR was used for validation of 40 novel striatal-enriched transcripts obtained from SAGE and AIBS analysis. qRT–PCR experiments for these transcripts revealed three groups of candidate transcripts: (i) transcripts whose expression was shown to be significantly higher in the striatum; (ii) transcripts that show a trend towards striatal enrichment and (iii) transcripts that showed lower expression levels in the striatum compared with at least one other tissue. A cross comparison between the expression levels in striatum, cortex, hippocampus, cerebellum, liver and muscle using qRT–PCR revealed Gm705 expansion, a previously unknown striatal-enriched transcript (A), and Tmem158, a transcript not previously known to be striatal-enriched (B) were significantly higher in the striatum; Ap1s1, a transcript selected by SAGE showed a trend towards striatal enrichment (C); for transcript Trom1, striatal-enrichment was not validated using qRT–PCR (D). Overall, out of 40 transcripts, 45% showed a significantly higher level in striatum, 37% showed a trend towards striatal enrichment and 17% showed lower levels of expression in striatum than at least one other tissue. Error bars represent SEM. Significant differences were determined by a one-way ANOVA analysis using a threshold value of P < 0.05. The qRT–PCR results for rest of the candidate transcripts are in Supplementary Material, Figure S2.

Figure 3.

Five striatal-enriched genes respond to pathogenesis in the YAC128 mouse model of HD. Relative quantification of the mRNA levels of 34 novel striatal-enriched transcripts in the striatum of 12-month-old YAC128 mice revealed the greatest mRNA changes for Ap1s1 (A) Cd4 (B) and Indo (C). These transcripts were also tested at 3 months and only Indo mRNA expression exhibited a significant up-regulation at both 12 months and 3 months of age (C). As expected, well-known striatal transcripts, Ppp1r1b and Gnal showed significant changes at 12 months (D and E). Tmem158 is an example of a striatal-enriched gene with no alteration in expression level in the YAC128 mice (F). Error bars represent SEM. Significant differences were determined by student's t-test (unpaired; two-tailed) using a threshold value of P < 0.05.

Figure 4.

Alteration in expression levels of striatal-enriched transcripts in postmortem caudate of HD subjects. Transcripts that were differentially expressed in the YAC128 mouse model of HD were tested in human postmortem caudate samples from HD patients and age-matched controls. PPP1R1B (A) and AP1S1 (C) showed significant down-regulation in HD subjects compared with controls, similar to expression changes observed in the YAC128 mouse model. CD4 (B) was significantly up-regulated in HD subjects compared with controls, opposite to the change observed in the YAC128 mouse model. Error bars represent SEM. Significant differences were determined by student's t-test (unpaired; two-tailed) using a threshold value of P < 0.05.