Huntington's Disease Pathogenesis: Two Sequential Components

Abstract

Historically, Huntington's disease (HD; OMIM #143100) has played an important role in the enormous advances in human genetics seen over the past four decades. This familial neurodegenerative disorder involves variable onset followed by consistent worsening of characteristic abnormal movements along with cognitive decline and psychiatric disturbances. HD was the first autosomal disease for which the genetic defect was assigned to a position on the human chromosomes using only genetic linkage analysis with common DNA polymorphisms. This discovery set off a multitude of similar studies in other diseases, while the HD gene, later renamed HTT, and its vicinity in chromosome 4p16.3 then acted as a proving ground for development of technologies to clone and sequence genes based upon their genomic location, with the growing momentum of such advances fueling the Human Genome Project. The identification of the HD gene has not yet led to an effective treatment, but continued human genetic analysis of genotype-phenotype relationships in large HD subject populations, first at the HTT locus and subsequently genome-wide, has provided insights into pathogenesis that divide the course of the disease into two sequential, mechanistically distinct components.

Keywords: Huntington disease; genetic association; genetics; genotype-phenotype correlation; modifier gene; trinucleotide repeat expansion.

Fig.1

Relationship of age at motor onset with CAG repeat. A. Age at onset of motor signs estimated by raters (Y-axis) is compared to the size of uninterrupted CAG repeat (X-axis) for subjects with inherited CAG sizes of 40–55 repeats. Each circle represents a HD subject participating our recent onset modifier GWA study [37]. B. Age at onset of motor signs for the subset of HD subjects who inherited 43 CAGs (∼15.5% of the total data set) is plotted to show the wide variability in clinical manifestation due to factors other than CAG repeat length.

Fig.2

Continuous and dichotomous phenotypes used in GWA analysis. Three steps were taken to identify genetic modifiers in HD: 1) Phenotyping: For each HD subject, age at onset corrected for inherited CAG repeat length (i.e., residual age at onset) was calculated by subtracting the age at onset predicted for that individual (based on their CAG length in comparison with a large population of HD subjects) from the age at onset observed for that individual. 2) Genotyping: Genomic DNA samples were analyzed to determine genetic variations genome-wide, and subsequently used for genotype imputation using a large reference population in order to increase the number of SNPs available for analysis. 3) Association analysis: A statistical model was built to explain residual age at onset (continuous phenotype variable) as a function of a test SNP to judge significance in association between phenotype and genotype (continuous analysis). As a complementary approach, HD subjects with onset extremely earlier or later than their expected age at onset were identified based on residual age at onset, and for each test SNP, the allele frequencies were compared between the early and late groups (dichotomous analysis).

Fig.3

Age at motor onset correlates best with uninterrupted CAG repeat length. The data from Figure 1 are replotted noting those individuals whose CAG size measure from PCR fragment-size genotyping was corrected after sequencing to account for the loss of the CAA interruption or duplication of the CAACAG segment in a small minority of subjects (red and green circles, respectively). The black line represents the age at onset to CAG length relationship predicted from all subjects. The dotted red and green lines respectively represent the age at onset to CAG length relationship for the rare CAA interruption or CAACAG duplication subjects based on the inaccurate CAG size from genotyping while the solid red and green lines show the result after correction of these CAG sizes by sequencing. The differences between these solid red and green lines and the black line might reflect subtler differences in the properties of uninterrupted CAG repeats depending on their surrounding sequence context or the presence of a linked modifier locus on these chromosome 4s. Those subjects with an uninterrupted CAG length of 43 are highlighted to permit comparison of polyQ length, showing by example of filled circles that CAA loss (red) and CAACAG duplication (green) subjects differ by 4 Qs, with those possessing the longer 47Q segment having later onset than those possessing the shorter 43Q segment. Subjects represented by grey circles at this CAG size all possess 45Qs.

Fig.4

Continuous and dichotomous GWA analysis of European-ancestry HD subjects. The levels of significance for each test SNP (circles) across the genome are mirrored for both continuous phenotype (top portion) and dichotomous phenotype (bottom portion) for ease of comparison. Numbers in the middle of the plot represent chromosomes, and horizontal lines indicate genome-wide significant p-values. The loci harboring genes involved in DNA maintenance/repair process are labeled only in the upper panel while those loci harboring genes not known to be central to these processes are labeled only in the lower panel.

Fig.5

Modifier alleles are not predictive of age at onset in any given HD individual. For each of three modifier haplotypes (5AM1 and 5AM3 at MSH3 and 15AM2 at FAN1), a plot is shown relating the residual age at onset of HD subjects who inherited an uninterrupted CAG repeat length of 43 units, grouped by their genotype at the modifier locus, i.e., the number of copies of the SNP minor allele (0 = homozygous major allele; 1 = heterozygous; 2 = homozygous minor allele) that tags that particular haplotype (rs701383, rs1650742, and rs8034856, respectively) [37, 41]. Each HD subject is denoted by a blue circle with the population size (‘N’) for each genotype shown beside the distribution of residual age at onset values for that genotype. The dashed line is the trend line that reveals the influence of the modifier haplotype on age at onset, colored red for an onset-hastening effect (5AM1) and green for an onset-delaying effect (5AM3 and 15AM2), detectable and highly significant at the population level but not predictive of onset age in any given HD individual.

Fig.6

HD pathogenesis consists of two sequential components: somatic CAG expansion that results in cellular damage after a threshold length is reached. To illustrate the concept of a two component model of HD pathogenesis, which likely also applies to at least some other repeat disorders, hypothetical plots are shown for somatic expansion in the average target cell for two different starting alleles, uninterrupted CAG repeats of 45 units (orange), in the fully penetrant size range, and 38 units (blue), in the partially penetrant size range. Depicted are plots in the absence of a modifier (solid line) and in the presence of a strong onset-delaying modifier (dashed line), a weak onset-hastening modifier (dash-dotted line) or a strong onset-hastening modifier (dotted line). Somatic expansion the CAG repeat causes it to cross a threshold length (denoted by the dashed purple line) and trigger damaging consequences once in the range shaded as light purple. The CAG repeat inherited as 45 units crosses the threshold line at an early age that can be shifted earlier or later by modifiers of somatic CAG expansion while somatic expansion of the 38 CAG repeat crosses the threshold only late in life, except that in the presence of the strong onset-delaying modifier, the somatic CAG length in the average cell never exceeds the critical threshold to trigger damage during the lifetime of the subject.

Fig.7

GeM-HD Euro 9K website. The opening page of the GeM-HD Euro 9K website, which can be accessed through HDinHD (https://www.hdinhd.org/), provides links to regional association plots by gene or SNP, to the University of California at Santa Cruz Genome Browser with a custom track for the GWA data, to a summary data download and to a utility that performs Gene Set Enrichment Analysis for user-specified custom gene sets.

Fig.8

GeM-HD Euro 9K provides regional association plots by gene or SNP. Regional association plots can be generated by entering a gene symbol (case-sensitive) or SNP rs ID as explained in the text.

Fig.9

GeM-HD Euro 9K performs Gene Set Enrichment Analysis (GSEA) for user-provided gene sets. GSEA analysis of the GWA association data can be performed for any custom gene set by entering the genes in the top left box and selecting a number of permutations using the slider below it. The null distribution will be returned with an indication of the significance of enrichment of the user-defined gene set noted by a red triangle.