Genetic analysis of the GRIK2 modifier effect in Huntington's disease

Abstract

Background: In Huntington's disease (HD), age at neurological onset is inversely correlated with the length of the CAG trinucleotide repeat mutation, but can be modified by genetic factors beyond the HD gene. Association of a relatively infrequent 16 TAA allele of a trinucleotide repeat polymorphism in the GRIK2 3'UTR with earlier than expected age at neurological onset has been suggested to reflect linkage disequilibrium with a functional polymorphism in GRIK2 or an adjacent gene.

Results: We have tested this hypothesis by sequencing all GRIK2 exons, the exon-flanking sequences and 3'UTR in several individuals who were crucial to demonstrating the modifier effect, as they showed much earlier age at neurological onset than would be expected from the length of their HD CAG mutation. Though ten known SNPs were detected, no sequence variants were found in coding or adjacent sequence that could explain the modifier effect by linkage disequilibrium with the 16 TAA allele. Haplotype analysis using microsatellites, known SNPs and new variants discovered in the 3'UTR argues against a common ancestral origin for the 16 TAA repeat alleles in these individuals.

Conclusion: These data suggest that the modifier effect is actually due to the TAA repeat itself, possibly via a functional consequence on the GRIK2 mRNA.

Figure 1

Genomic structure of GRIK2 and locations of microsatellites and SNPs. The relative positions of nine microsatellites chosen using the Human Genome Database and the UCSC Genome Browser are shown above a bar, oriented from centromere (cen) to telomere (tel), representing the chromosome 6q16.3 region containing GRIK2 [18, 33]. The GRIK2 gene, comprising 17 exons (numbered) is expanded below the bar, and the region from exon 11 through the 3'UTR is further expanded below. The annotation of the 3'UTR region of GRIK2 in human genome databases is incomplete, indicating only a shorter transcript associated with use of a polyadenylation signal upstream from the TAA repeat. Data from [29] indicate that two polyadenyation signals may be used, producing either a shorter mRNA, without the TAA repeat, or a longer mRNA that contains it. Coding sequences are shown in dark blue and UTRs are shown in light blue, with introns shown as a thin black bar. The locations of relevant microsatellites and SNPs are denoted by downward and upward-pointing triangles, respectively. SNPs located in the GRIK2 gene are denoted by letter: a: rs2852565, b: rs2786251, c: rs6922753, d: rs2518283, e: rs2243355, f: rs3213607, g: rs2227283, h: rs2852620, i: rs1034254, j: rs12198351, k: rs28383483, l: rs28383484. The last polymorphism is located at the edge of the region of extended homology with rodent mRNAs (see Figure 2) and may lie either at the extreme end of the human 3'UTR or immediately outside of it.

Figure 2

Evolutionary conservation of the GRIK2 3'UTR region. Plots comparing chimpanzee, dog, rat, mouse, chicken and frog orthologue sequences to the conserved 3'UTR region of human GRIK2 were generated using mVISTA [34-36]. For the human sequence, we chose chr6: 102623080–102624690, comprising 1610 bp sequence downstream from the stop codon, based upon the report of a polyadenylation signal at 1547 bp after the stop codon [29]. The relative location of the polymorphic TAA repeat within the human sequence is shown between vertical lines (from 787 to 828bp). Plots show % identity over a 100 bp window (vertical axis) of chimpanzee (chr5 (Nov 2003): 104613747-104615339), dog (chr12 (July 2004): 62913564-62915139), rat (chr20 (June 2003): 53258651-53260220), mouse (chr10 (May 2004): 49099305-49100874), chicken (chr3 (Feb 2004): 68531473-68532948), and frog (Xenopus tropicalis) (Scaffold-599 (Oct 2004): 27668-29079), respectively, relative to the human sequence (horizontal axis) [18].

Figure 3

GRIK2 3' haplotypes in HD patients with earlier than expected onset. Haplotypes determined for microsatellite and SNP (labeled as in Figure 1) markers from GRIK2 exon 11 through the 3'UTR are shown for the four HD patients with earlier than expected onset, three possessing a 16 TAA allele at D6S1028 and one possessing a rare 17 TAA allele. The 16 TAA allele chromosomes are distinguishable by other polymorphisms except in the fully sequenced 2.3 kb region between rs12198351 and rs28383484 (denoted by a red bar above this region of the haplotype) suggesting that the TAA allele is responsible for the modifier effect.