Skip to main content
NCBI home page
Molecular Therapy logoLink to Molecular Therapy
. 2015 Aug 25;23(11):1759–1771. doi: 10.1038/mt.2015.128

Huntingtin Haplotypes Provide Prioritized Target Panels for Allele-specific Silencing in Huntington Disease Patients of European Ancestry

Chris Kay 1, Jennifer A Collins 1, Niels H Skotte 1, Amber L Southwell 1, Simon C Warby 2, Nicholas S Caron 1, Crystal N Doty 1, Betty Nguyen 1, Annamaria Griguoli 3, Colin J Ross 1, Ferdinando Squitieri 3, Michael R Hayden 1,*
PMCID: PMC4817952  PMID: 26201449

Abstract

Huntington disease (HD) is a dominant neurodegenerative disorder caused by a CAG repeat expansion in the Huntingtin gene (HTT). Heterozygous polymorphisms in cis with the mutation allow for allele-specific suppression of the pathogenic HTT transcript as a therapeutic strategy. To prioritize target selection, precise heterozygosity estimates are needed across diverse HD patient populations. Here we present the first comprehensive investigation of all common target alleles across the HTT gene, using 738 reference haplotypes from the 1000 Genomes Project and 2364 haplotypes from HD patients and relatives in Canada, Sweden, France, and Italy. The most common HD haplotypes (A1, A2, and A3a) define mutually exclusive sets of polymorphisms for allele-specific therapy in the greatest number of patients. Across all four populations, a maximum of 80% are treatable using these three target haplotypes. We identify a novel deletion found exclusively on the A1 haplotype, enabling potent and selective silencing of mutant HTT in approximately 40% of the patients. Antisense oligonucleotides complementary to the deletion reduce mutant A1 HTT mRNA by 78% in patient cells while sparing wild-type HTT expression. By suppressing specific haplotypes on which expanded CAG occurs, we demonstrate a rational approach to the development of allele-specific therapy for a monogenic disorder.

Introduction

Huntington disease (HD, (MIM 143100)) is the most common monogenic movement disorder, and among the most common diseases resulting from a specific mutation.1,2 HD is caused by an expanded CAG repeat in exon 1 of the Huntingtin gene (HTT), molecularly defined by more than 35 tandem CAG triplets in one copy of the HTT gene.3,4,5 Expanded CAG triplets encode similarly repetitive glutamine residues in the HTT protein, leading to multiple downstream pathogenic effects and selective neuropathology.6 The progressive symptoms of HD typically manifest in adulthood, deteriorating to death approximately 15–20 years after onset.7,8 Despite more than 20 years of research since discovery of the causative mutation, there are no therapies at present to alter the course of HD.

The defined monogenic cause of HD, and its consequent gain-of-function toxicity, allow suppression of HTT as a therapeutic strategy.9 Preclinical studies have shown reversal of HD phenotypes by inducible or exogenous silencing of transgenic mutant HTT.10,11,12,13 However, reagents which suppress both wild-type HTT and mutant HTT may have detrimental long-term consequences. Constitutive loss of murine homolog Hdh is embryonic lethal and reduced Hdh expression is associated with developmental and cognitive deficits.14,15,16,17 Postnatal inactivation of Hdh leads to neurodegenerative phenotypes, suggesting that adult reduction of HTT may be poorly tolerated over long-term treatment.18 Wild-type HTT has also been shown to be protective against toxic effects of mutant HTT in a dose-dependent manner.19 Preferential silencing of the mutant HTT allele, preserving normal HTT expression, may therefore minimize loss-of-function effects and yield greater therapeutic benefit than nonspecific suppression of both HTT copies. Forthcoming trials will establish whether transient suppression of HTT is tolerated in the human brain, but the long-term safety and efficacy of nonspecific HTT suppression in HD patients remains unclear versus allele-specific strategies.

Single-nucleotide polymorphism (SNP)-targeted silencing of mutant HTT, acting to degrade a mutant transcript bearing a specific target allele, has achieved potent reduction of mutant HTT with negligible effect on normal HTT expression.20,21,22 Careful structure-activity studies of antisense oligonucleotides (ASOs) suggest that suppression of wild-type HTT transcript may be avoided with SNP-targeted reagents given appropriate preclinical screens.21,22,23 However, specificity of therapy requires a patient to be heterozygous for the target allele and for that allele to be present on the CAG-expanded copy of HTT. A crucial question in the development of SNP-targeted reagents is therefore the identification of target alleles with the greatest heterozygosity in the HD patient population, allowing allele-specific therapy in the greatest number of patients. Various transcribed HTT polymorphisms have been genotyped in local patient cohorts,20,24,25,26 but phased heterozygosity estimates are unavailable for large numbers of alleles across diverse patient groups, which are required to prioritize development of allele-specific therapeutics. Heterozygosity of a given target allele can vary considerably between patient populations. The Δ2642 codon deletion present in exon 58 of HTT has been targeted for selective HTT silencing in vitro by siRNA,27,28 but the frequency of this polymorphism among HD chromosomes varies from 59% in an American cohort26 to 18.6% in Italy.29 No study has examined the phased heterozygosity and haplotype relationship of all common alleles in the HTT transcript and it therefore remains unclear which HTT polymorphism, or combination of polymorphisms, would offer treatment for the greatest number of patients across different ancestry groups.

The goal of this study was to identify the most efficient and useful combination of targets for allele-specific HTT suppression in the greatest number of HD patients of European ancestry. We present the first comprehensive investigation of haplotypes spanning the entire ~170 kb HTT gene sequence using all common (>1% MAF) SNPs from the 1000 Genomes Project30,31 and dense SNP genotyping in 1285 HD patients and relatives from Canada, Sweden, France, Italy, and Finland. We reveal a regular genetic architecture across HTT, enabling rational, haplotype-specific targeting of mutant HTT in the greatest portion of HD patients of European descent. We identify multiple novel targets for silencing mutant HTT and show potent, selective silencing of the mutant transcript using ASOs complementary to a biallelic indel (insertion–deletion polymorphism) on the primary HD haplotype. Our work provides a general target selection strategy for dominant genetic disorders occurring on specific founder haplotypes, and provides a prioritized panel of targets for allele-specific therapeutic approaches to HD.

Results

SNPs across HTT represent specific gene-spanning haplotypes

In order to determine the frequency and heterozygosity of all common allele-specific HTT targets relative to one another, we sought to establish haplotypes for a large number of common polymorphisms across the gene region. Various partial haplotypes have been constructed across HTT, which overlap ambiguously due to low marker density in each study. We previously genotyped 91 SNPs across the HTT gene region,20 of which 63 are present at greater than 1% frequency in European populations.31 Of these 63 common SNPs, 51 are located between the start of the HTT 5'UTR and the end of the 3'UTR (chr4:3076408-3245687, GRCh37). In total, 527 Canadian HD patients and 305 control relatives from the UBC HD Biobank were genotyped and phased at all 63 SNPs for this study. Using patterns of familial segregation, we reconstructed gene-spanning haplotypes at all 63 SNPs for 293 unrelated CAG-expanded chromosomes (CAG > 35) and 829 control chromosomes (CAG ≤ 35) from Canadian individuals of European ancestry. Annotation of dense 63-SNP HTT haplotypes replicated haplogroup assignments using 22 tagging SNPs (tSNPs) across the HTT gene region,24,32 and revealed that recombination between common haplotypes rarely occurs within the transcribed HTT gene sequence (Figure 1a). For example, the A3 haplotype is frequently associated with an extragenic 5' crossover with the C1 haplotype, whereas no common haplotype recombines C1 within the HTT gene sequence. Only 9/283 (3.2%) HD chromosomes and 25/829 (3.0%) control chromosomes in our Canadian cohort represent intragenic recombinants of gene-spanning HTT haplotypes, confirming that recombination within HTT is rare. Analysis of pairwise linkage disequilibrium (LD) between genotypes of all 63 SNPs in 1,664 phased haplotypes from Canadian HD patients and controls reveals a ~170 kb region of high LD (D' > 0.9) from rs762855 to rs362303 (chr4: 3074795-3242307), indicating a haplotype block of exceedingly low recombination across the entire transcribed HTT sequence (Supplementary Figure S1).

Figure 1.

Figure 1

Open in a new tab

SNPs across HTT represent specific gene-spanning haplotypes. (a) Single-nucleotide polymorphisms (SNPs) across HTT belong to gene-spanning haplotypes representing three major haplogroups A, B, and C. Transcribed (intragenic) SNPs are shaded gray. The primary Huntington disease (HD) haplotype, A1, is defined by rs362307. The secondary HD haplotype A2 is defined by rs2798235 and rs363080. One SNP (rs2298969) distinguishes the A haplogroup from the B and C haplogroups (black box). In our panel, 8 of 51 intragenic SNPs exclusively distinguish the A and B haplogroups from the C haplogroup (bold). (b) Pairwise linkage disequilibrium (LD) of SNP genotypes (r2) reveals a complex haplotype structure across the HTT gene region (black, r2 = 1; shades of gray, 1 > r2 > 0; white, r2 = 0). Alleles present on similar haplotypes are in high pairwise LD across HTT, such as A2-defining rs2798235 and rs363080. Haplotype A1-defining SNP rs362307 is not in LD with any other variant in our initial 63-SNP panel. Positions correspond to GRCh37. Representative extragenic crossover sequences are colored according to the most likely originating haplotype.

In contrast, stringent pairwise LD by correlation coefficient (r2) reveals a punctuated pattern of SNP disequilibrium within the HTT haplotype block, reflecting a diversity of haplotypes spanning the gene locus despite the absence of recombination (Figure 1b). Strikingly, SNPs in high pairwise correlation within HTT mark specific intragenic haplotypes. For example, rs2798235 and rs363080 represent unique markers of the A2 haplotype and are found in near-perfect pairwise correlation (r2 = 0.98), whereas both SNPs show low pairwise correlation with all other variants in our common 63-SNP panel. The observed pattern of LD across HTT indicates that SNPs within HTT tag specific gene-spanning haplotypes encompassing the entire transcribed HTT sequence. Among Canadian subjects, 95.8% (271/283) of HD chromosomes and 95.9% (795/829) of control chromosomes conform to 20 specific nonrecombinant haplotypes at 51 common intragenic SNPs and at exon 1 CCG repeat length (Supplementary Figure S2). A and B haplogroups are always marked by 7 CCG in the Canadian cohort and C haplotypes are always 8, 9, or 10 CCG.

A1, A2, and A3 are the most common gene-spanning HD haplotypes

We next determined the most frequent gene-spanning haplotypes occurring on HD chromosomes. Among 283 unrelated Canadian HD chromosomes, 48.1% (136/283) are found on the A1 haplotype marked by rs362307, 32.2% (91/283) are found on closely related A2a or A2b, and 12.0% (34/283) are found on A3 (Figure 2a). In total, 92.2% (261/283) of Canadian HD chromosomes are found on A1, A2, or A3 haplotypes spanning HTT. Among control chromosomes, only 8.0% are A1, 16.4% are A2a or A2b, and 13.1% are A3. Haplotypes A4 and A5, each present on 6.3% of control chromosomes, are never observed on Canadian HD chromosomes. Notably, A1 and A2a represent the most genetically distant haplotypes within the A haplogroup, despite representing the most frequent HD haplotypes. Haplogroup B is a distinct genetic lineage in 5.3% of HTT controls, present on only 3/283 HD chromosomes in the Canadian cohort (1.1%). Haplogroup C is a complex collection of haplotypes constituting nearly half of unrelated control chromosomes (42.6%), but is found on only 3.2% of HD chromosomes. The most common intragenic haplotype among all annotations is C1, present on 29.8% of unrelated control chromosomes in the Canadian cohort.

Figure 2.

Figure 2

Open in a new tab

A1, A2, and A3a are the most common gene-spanning HD haplotypes. (a) In the UBC HD BioBank, the expanded CAG repeat (CAG > 35) is found on the A1 haplotype in 48.1% of phased, unrelated Canadian Huntington disease (HD) chromosomes. Intragenic recombinant haplotypes (X) are rare (3.0% of controls) whereas >95% of HTT haplotypes show no evidence of recombination within the transcribed gene sequence. C1 is the most common intragenic haplotype among Canadian control chromosomes, representing 29.8%. (b) The most common HD haplotype, A1, is uniquely defined by three transcribed polymorphisms in high pairwise linkage disequilibrium (LD) across HTT. The 4 bp indel rs72239206 represents a novel polymorphism associated with the CAG expansion (bold). The second most common HD haplotype, A2, is defined by five intragenic single-nucleotide polymorphisms (SNPs), three of which are novel (bold). HD-associated A3a, the third most common HD haplotype, is specifically marked by the novel SNP rs113407847. (c) Pairwise LD plot (r2) of A1 and A2 haplotype-defining polymorphisms as calculated from 700 phased haplotypes of European Caucasians.

Identification of all defining intragenic alleles on HD-associated haplotypes

Mutant HTT is enriched for gene-spanning A1 and A2 haplotypes relative to controls. This suggests that alleles found exclusively on these haplotypes are potential targets for allele-specific silencing of mutant HTT. To determine all polymorphisms uniquely found on the most frequent HD-associated haplotypes (A1 and A2), we identified all chromosomes in the 1000 Genomes Project with defining SNPs for these haplotypes. In total, 2297 intragenic polymorphisms are annotated across HTT (chr4:3076408-3245687, GRCh37) in the 1000 Genomes phase 1 data set across all ethnicities.31

In our 51-SNP HTT panel, A1 is uniquely defined by rs362307, a [T/C] SNP present in exon 67 and the 3'UTR of HTT. Among all 1000 Genomes control chromosomes, 3.5% (76/2166) carry this SNP. Among the 76 chromosomes bearing rs362307[T], 75 (98.7%) carry the glutamic acid deletion known as Δ2642 (rs149109767) and 74 (97.4%) carry a novel 4 bp intronic deletion (rs72239206). Among all 2166 chromosomes, including those 76 bearing rs362307[T], Δ2642 is present on 77 (3.5%) and rs72239206 is present on 83 (3.8%). Therefore 97.4% (75/77) of chromosomes with Δ2642 and 89.2% (74/83) of those with rs72239206 also carry rs362307[T]. Both polymorphisms are therefore highly sensitive proxy markers of rs362307 (Figure 2b). No other SNPs were strongly associated with rs362307. The next most common SNP also present on at least 90% of the A1 chromosomes was found nonspecifically on 633/2,166 chromosomes in the 1000 Genomes set. Among SNPs less frequent than rs362307 but specific to A1 chromosomes, rs187059132 is present on 32/76. Variants rs362307, rs149109767, and rs72239206 are therefore highly specific for the A1 haplotype, having high pairwise correlation (r2 > 0.9) with each other but with no other SNPs in HTT. 1000 Genomes phase 1 chromosomes bearing all three A1-defining polymorphisms are found exclusively on individuals of European or Admixed European ethnicity (Supplementary Table S1). Each individual variant is rare or absent in other ethnic groups, in agreement with the reported absence of Δ2642 and rs362307 in individuals of East Asian and black South African ancestry.32,33,34

The A2 haplotype, comprised of closely related subtypes A2a and A2b, is uniquely tagged by rs2798235 and rs363080 in our 51-SNP HTT panel. 100 chromosomes in 1000 Genomes phase 1 carry rs363080, of which 98 carry rs2798235. The latter SNP is exclusively found on chromosomes tagged by rs363080. High pairwise correlation between these two markers is similarly observed in direct genotyping of our Canadian HTT chromosomes (r2 = 0.98, Figure 1a). In 1000 Genomes, three additional intragenic SNPs—rs363107, rs362313, and rs2530595—are found on 100, 99, and 99 of chromosomes bearing rs363080, respectively, and are likewise present only on these chromosomes. All five polymorphisms are present on 98% of chromosomes bearing any of the five variants, and therefore represent specific markers of the A2 haplotype (Figure 2b).

HD also commonly occurs on A3. In our 51-SNP panel, the A3 haplotype is defined by intragenic markers of the A haplogroup in the absence of SNPs specific for the other A haplotypes (Figure 1a). One hundred and nineteen A3 haplotypes were identified out of 738 control chromosomes of European ancestry (16.1%). No identifying SNPs were found that uniquely encompass all 119 A3 chromosomes. However, a specific subtype SNP was observed on 45.4% (54/119) of A3 chromosomes—rs113407847—designating A3a (Figure 2b). In the 738 European individuals, rs113407847 is found only in the subset of A3 haplotypes. Despite common association with HD, no SNPs specific to both A1 and A3 were found, except when shared with other, non-HD associated A haplotypes.

Validation of polymorphisms specific for the HTT A1 haplotype

To validate our in silico association of all three A1-defining polymorphisms from whole genome sequencing data in 1000 Genomes, Δ2642 and rs72239206 were directly genotyped and phased to the CAG repeat in HTT chromosomes previously genotyped for rs362307. These comprised all Canadian HD chromosomes genotyped with the original 63-SNP panel as well as other previously haplotyped samples from various ethnic groups. In total, 454 phased, unrelated HD chromosomes and 652 unrelated control chromosomes were successfully genotyped and phased to the CAG repeat at rs362307, Δ2642, and rs72239206. Pairwise LD of direct genotyping data indicates that all three polymorphisms are present in HD and control chromosomes in near-perfect LD (r2 > 0.99) and that all three minor alleles are highly enriched on HD chromosomes versus controls (Supplementary Table S2).

Validation of HD haplotype-specific polymorphisms in distinct HD patient populations

Marker studies of the Δ2642 codon deletion suggest that the frequency of the A1 haplotype varies considerably between Caucasian HD patient populations (Supplementary Figure S3). A key question following our definition of specific gene-spanning HD haplotypes was therefore to determine the distribution of these haplotypes among different patient populations of European ancestry. A revised SNP panel was designed to include the prior 63-SNP panel as well as novel defining A1, A2, and A3a SNPs derived from the 1000 Genomes Project. Using this revised panel, we genotyped 120 Swedish, 76 French, and 209 Italian HD family members, derived from respective countries of origin. Haplotypes were reconstructed and phased to CAG repeat size, as for our Canadian Caucasian cohort. All common 63-SNP haplotypes found in the Canadian Caucasian cohort were replicated by genotyping of the European HD cohorts using our revised panel. All three A1 variants and all five A2 variants conformed to high expected pairwise correlation in direct genotyping of the European cohorts with the revised panel (Figure 2c). Among all European patients, the CAG expansion on A3 was found exclusively in phase with the unique A3a-identifying SNP rs113407847, but not on A3b lacking this SNP, suggesting that A3a is a disease-associated haplotype. Direct genotyping of rs113407847 in Canadian HD A3 chromosomes similarly revealed that the CAG expansion occurs almost exclusively on A3a when present on A3 (31 of 34 A3 Canadian HD chromosomes) (Figure 2a). Common HD-associated haplotypes A1, A2, and A3a therefore share uniform sets of defining markers in ethnically distinct European HD patient cohorts, implying deep ancestral relationship of these disease-associated haplotypes across different European populations.

Frequency of HD haplotypes differs by European ancestry

While intragenic haplotypes of HTT are consistently A1, A2, and A3a across Caucasian HD chromosomes, our direct genotyping reveals striking differences in the frequency of specific HD-associated haplotypes among CAG-expanded chromosomes in different European populations (Figure 3).

Figure 3.

Figure 3

Open in a new tab

In four European Huntington disease (HD) cohorts, distinct distributions of HD haplotypes are observed. As in Canada, A1 is the most frequent HD haplotype in Finland (60%) and Sweden (51%), and France (45%). In contrast, A2 is the most frequent HD haplotype in Italy (58%).

Among unrelated Swedish HD chromosomes, 51% (26/51) are found on A1, similar to our previously genotyped Canadian HD cohort (P = 0.7616, Fisher's Exact). The frequency of A2 among Swedish HD is comparatively lower than in Canada (18 versus 32%, P = 0.0455) and A3a is more frequent (28 versus 11% in Canadians, P = 0.0033). French HD chromosomes are also most frequently A1 (45 versus 48%, P = 0.7654) with A2 present at similar frequencies and A3a more common than in Canadian HD (A2, P = 0.1957; A3a, P = 0.0256). In contrast, Italian HD chromosomes are predominantly found on A2 (58%, P < 0.0001 versus Canadian), with a much smaller proportion of HD on A1 versus the Canadian cohort (19%, P < 0.0001) and a similar proportion of A3a (7%, P = 0.2647). In a small set of Finnish HD families, haplotyped with our original 63-SNP panel, all unrelated disease chromosomes are A1 (6/10, 60%) or A2 (4/10, 40%). Despite differences in specific haplotype frequency between our Canadian and European cohorts, >90% of HD chromosomes are found on A1, A2, and A3a haplotypes in all four populations of Northern European ancestry and in 84% of Italian HD chromosomes.

HTT haplotypes on control chromosomes also differ between European populations, though less dramatically than CAG-expanded chromosomes. The A haplogroup trends toward higher frequency in Italian controls versus Canadian (P = 0.0597), but is found at similar frequency among Swedish (P = 0.6838) and French control chromosomes (P = 0.8073). A1 occurs at statistically similar frequencies in all four control cohorts, whereas A2 occurs at higher frequency among Italian controls than in Canadian (24 versus 16%, P = 0.0185) or Swedish controls (13%, P = 0.0203), mirroring its elevated frequency among Italian HD chromosomes.

A1, A2, and A3a haplotypes are target panels to optimize population coverage for allele-specific HTT silencing

High pairwise correlation of specific haplotype-defining polymorphisms allows for targeting of the A1 and A2 haplotypes as a selective HTT silencing strategy. As all three A1 markers are present in near-perfect LD, targeting any single A1 polymorphism will allow allele-specific HTT silencing in a nearly equal number of HD patients heterozygous for this haplotype. Heterozygosity of A1 in HD patients, when phased to the CAG expansion, is highest in Sweden (47%), Canada (44%), and France (43%), but much lower in Italy (15%), suggesting greater utility in patients of Northern European ancestry (Figure 4).

Figure 4.

Figure 4

Open in a new tab

Cumulative treatment coverage of Huntington disease (HD) patients by A1, A2, and A3a allele-specific HTT silencing targets in four patient populations of European ancestry. Targeting one A1 allele (among three) and one A2 allele (among five) would permit selective HTT silencing treatment in 68% of HD patients in the Canadian cohort. In combination, A1 + A2 + A3a targets would permit selective silencing of an average of 80% of HD patients across all four populations.

An estimated 98% of patients with HD on A2 (phased to rs363080) will have all five A2 targets present. The percent of patients heterozygous for A2, phased to the CAG expansion, range from to 18% in Sweden to 43% in Italy, suggesting a greater utility in Southern European populations. Tertiary targeting of rs113407847 would allow treatment of patients bearing the CAG expansion on A3a, ranging from a maximum of 27% of patients in Sweden to only 5% in Italy.

On average across all four patient populations, the A1 haplotype may allow selective silencing treatment of 37% of HD patients (Table 1), with targeting of the A2 haplotype permitting treatment of an additional 27%. Since no patient will share both A1 and A2 targets on the same disease chromosome, percentages of patients treatable with each haplotype target are additive. Across all four cohorts, a maximum of 64% of patients will be treatable with two allele targets representing A1 and A2, rising to approximately 70% if Finnish HD chromosomes are included. In total, targeting three specific polymorphisms representing A1, A2, and A3a may allow selective silencing treatment of ~80% of HD patients overall from the Canadian, Swedish, French, and Italian patient populations. Defining SNPs of these HD-associated haplotypes therefore represent panels of targets that could achieve ~80% patient treatment by allele-specific HTT silencing strategies.

Table 1. Percent of Huntington disease patients treatable by different haplotype targets in four populations of European ancestry.

graphic file with name mt2015128t1.jpg

Open in a new tab

An alternative strategy is targeting SNPs specific for the A haplogroup that includes haplotypes A1, A2, and A3a. In populations of European ancestry, greater than 90% of HD patients carry an expanded CAG repeat on the A haplogroup and would be treatable if heterozygous with a B or C haplotype on the normal chromosome. In the Canadian, Swedish, and French cohorts, we find that a higher percentage of patients are heterozygous in phase with expanded CAG for the A haplogroup than for either A1 or A2 haplotypes (Table 1). Only in the Italian cohort does coverage with A2 exceed that of the A haplogroup given only one allele target. However, when two targets are considered cumulatively, only A1 and A2 alleles in combination allow treatment of the majority of patients in all four populations. Further, if one excludes patients eligible for A1 and A2 treatment, A3a is heterozygous in the largest proportion of European patients remaining. While defining polymorphisms of the A haplogroup present attractive heterozygosity rates due to deep ancestry of constituent A1, A2, and A3a haplotypes, targeting these specific, mutually exclusive disease haplotypes individually would enable treatment of more patients overall.

Both the 63-SNP Canadian data and the 1000 Genomes reference haplotypes indicate that the A haplogroup is defined by a single SNP, rs2298969. In contrast, 24 SNPs distinguish the A and B haplogroups from the C haplogroup (Table 2), offering many more potential targets for development. Averaged across all four cohorts, rs2298969 would offer treatment for 45% of patients, whereas each of 24 SNPs present on both A and B haplogroups are expected to be heterozygous in 42%. As with the A haplogroup, the overall percent of patients treatable with A1 and A2 in combination exceeds any corresponding combination of AB haplogroup alleles with either A1 or A2. The percent of patients treatable with A1, A2, and any AB target is similar to that achievable with A1, A2, and the A haplogroup (72% in both cases), but still fewer than the maximum percent treatable with A1, A2, and A3a in combination (80%).

Table 2. Prioritized intronic and exonic target alleles specific for the most common Huntington disease haplotypes and haplogroups.

graphic file with name mt2015128t2.jpg

Open in a new tab

ASOs targeting ΔACTT (rs72239206) potently and selectively silence A1 HTT mRNA and protein in human cells

Among all genotyped HD patients in this study, A1 is the most frequently heterozygous haplotype in cis with the expanded CAG repeat. Targeting the A1 haplotype is also crucial to achieving maximum cumulative patient coverage given two or three allele targets. The defining A1 markers, rs362307, Δ2642, and rs72239206, therefore represent key sites for achieving allele-specific treatment in the greatest number of HD patients of European ancestry. Both rs362307 and Δ2642 are found in mature mRNA, have known association with the CAG repeat expansion, and have been investigated as targets of siRNA-mediated selective HTT silencing.25,28 Small interfering RNA (siRNA) designed against the Δ2642 allele induced 51% reduction in mutant HTT mRNA in transfected juvenile HD fibroblasts, with ~20% off-target reduction of wild-type HTT mRNA.28 Potency and selectivity of siRNA against rs362307 has been shown using reporter constructs, but not in the context of full-length HTT.25

In addition to offering a novel A1 target not previously associated with HD, we hypothesized that targeting of the 4 bp rs72239206 indel sequence may offer greater selectivity than discrimination by a SNP, and sought to evaluate the potential of rs72239206 as a selective mutant HTT silencing target. Unlike rs362307 and Δ2642, rs72239206 is located in an intron (intron 22 of HTT) and is therefore only targetable by agents complementary to unspliced pre-mRNA. ASOs can induce RNAse H-mediated degradation of complementary pre-mRNA as well as mRNA,35 and we therefore designed ASO sequences incorporating a gapmer design with locked nucleic acid (LNA) wings and phosphorothioate linkages complementary to the rs72239206 deletion sequence (Figure 5a).

Figure 5.

Figure 5

Open in a new tab

ASOs targeting ΔACTT (rs72239206) potently and selectively silence A1 HTT mRNA and protein in human cells. (a) Design of antisense oligonucleotide (ASO) gapmers selectively targeting mutant HTT A1 mRNA at the ΔACTT sequence. (b) Transfection of patient-derived lymphoblasts (44/18 CAG) with ΔACTT-complementary ASOs selectively reduces mutant HTT mRNA. Patient lymphoblasts transfected with 5-9-5, 5-7-5, and 4-7-4 LNA gapmers show dose-dependent reduction of mutant HTT mRNA relative to untreated controls, falling to 21.5% mutant HTT mRNA at the highest 4-7-4 dose. Wild-type HTT mRNA levels do not fall below untreated levels at any tested dose of 5-7-5 or 4-7-4 LNA gapmer. (c) Dose-dependent reduction of mutant HTT protein relative to untreated controls, sparing wtHTT at all tested 5-7-5 and 4-7-4 LNA gapmer doses. ** and *** represent P = 0.01 and P = 0.001 by analysis of variance with Bonferroni post hoc. LNA, locked nucleic acid.

ASOs are passively taken up by neurons in primary culture.20 In the absence of transgenic HTT neurons bearing rs72239206, we sought to test the silencing potential of our LNA gapmers by passive uptake in human HD lymphoblasts bearing the A1 haplotype (GM03620, CAG 59/18). Remarkably, A1 HTT is selectively silenced in human lymphoblasts grown with rs72239206-targeted LNA gapmers in media, suggesting that lymphoblasts also passively take up ASO in culture (Supplementary Figure S4). On the basis of these preliminary experiments, we sought to examine dose-dependent knockdown of A1 HTT mRNA and protein in HD patient lymphoblasts of typical CAG length using active transfection to maximize effective dose.

Transfection of adult human A1/C1 lymphoblasts (GM02176, CAG 44/18) with a 5-9-5 LNA gapmer complementary to rs72239206 resulted in potent HTT mRNA silencing (Figure 5b), but only minimal discrimination between A1 and C1 transcripts (11% A1 and 29% C1 HTT mRNA remaining at the highest dose versus untreated cells). As expected, reduction of the DNA gap by two nucleotides to a 5-7-5 configuration improved selectivity but reduced potency,23 with the HTT A1 transcript reduced to 29% of untreated mRNA levels at the highest dose versus 92% HTT C1 control. Shortening this molecule to a 4-7-4 LNA gapmer design further improved selectivity for the HTT A1 transcript, reducing A1 HTT mRNA to 22% of untreated levels at the highest dose while sparing HTT C1 mRNA at untreated levels (Figure 5b). Western blot analysis using allelic separation of CAG 44/18 bands revealed similar reduction at the protein level for all three LNA gapmer designs, inducing dose-dependent reduction of mutant HTT with 5-7-5 and 4-7-4 gapmers while sparing normal HTT (Figure 5c). Targeting the A1-specific rs72239206 deletion sequence with complementary ASOs can therefore potently and selectively silence mutant HTT mRNA and protein in patient-derived cells that are genetically representative of HD patients bearing the A1 haplotype.

Targeting the rs72239206 deletion site is efficacious and tolerated in vivo

In the absence of transgenic mice bearing the rs72239206 deletion in cis with expanded CAG, in vivo silencing of A1 HTT mRNA and protein could not be directly evaluated. However, the wild-type analog of the 5-9-5 LNA gapmer, designed against reference sequence that includes the four bases deleted in A1, also elicited potent reduction of human HTT in brains of YAC128 mice bearing transgenic full-length mutant HTT (Supplementary Figure S5). The rs72239206 deletion site is therefore accessible to ASO-mediated HTT mRNA silencing in vivo.

Discussion

The translation of allele-specific HTT silencing to therapeutic application requires clarity as to which transcribed SNPs are the most useful targets in the HD patient population. The frequency of specific polymorphic targets is known to vary between clinical cohorts, while secondary and tertiary targets that maximize the total number of patients treatable have been incompletely described. Our study provides the first comprehensive heterozygosity estimates across the HTT transcript in multiple patient populations, identifying specific allele targets of highest priority for development of selective antisense therapies. We have fully described the most common gene-spanning haplotypes relevant for selective suppression of mutant HTT in patients of European ancestry—A1, A2, and A3a—and identify all common polymorphisms specific for these haplotypes. In four different patient populations, these gene-spanning haplotypes represent panels of allele-specific targets that would achieve treatment of the greatest proportion of HD patients. We show that as few as three gene silencing reagents targeting the A1, A2, and A3a haplotypes may offer allele-specific HTT silencing therapy for 80% of all patients of European descent. A1 may be silenced using one of three defining polymorphisms, and A2 using one of five defining polymorphisms. If only one allele target can be chosen for development, silencing the A haplogroup by rs2298969 may offer treatment in the greatest proportion of patients. But when two targets are considered additively, A1 and A2 targets in combination allow for treatment of the majority of patients in all four major populations evaluated in this study. When three targets are considered, no combination of intragenic polymorphisms allows for selective silencing in a greater proportion of cases than defining polymorphisms of the A1, A2, and A3a haplotypes. A1 and A2 haplotypes therefore represent sets of priority targets for preclinical evaluation of allele-specific HTT silencing reagents, with rs113407847 a priority tertiary candidate. However, due to RNA structure or kinetics of pre-mRNA splicing, not all SNP targets are available for ASO-mediated silencing.22 Should A3a-defining rs113407847 fail as a target, our results suggest that rs2298969 or any of 24 SNPs present on the A and B haplogroups would then represent tertiary targets for treating the next greatest proportion of patients.

Only four of our 34 priority alleles are located within exons: two specific to the A1 haplotype (rs149109767 and rs362307), one specific to the A2 haplotype (rs2530595), and one on the A and B haplogroups (rs362331) (Table 2). For allele-specific antisense strategies restricted to mature mRNA, such as siRNA, our data suggest that targeting either exonic A1 allele (rs149109767 or rs362307) in combination with the A2-specific allele rs2530595 will enable treatment of the most patients. It is notable, however, that rs149109767 is a codon deletion within a tandem GAG glutamic acid repeat, limiting effective design of short oligonucleotide sequences that can span and distinguish the mutant and wild-type alleles at this site. Therefore rs362307 remains a high-priority SNP for the development of A1-specific antisense reagents acting against mature mRNA, with rs2530595 representing a secondary exonic target specific for the A2 haplotype.

A polyadenylated exon 1 transcript of HTT containing expanded CAG repeats is translated into an N-terminal polyglutamine fragment in HD patients.36 Pathogenicity of mutant HTT exon 1 in N-terminal fragment models of HD raises the concern as to whether suppression of the exon 1 transcript, in addition to suppression of full-length mutant HTT, will be necessary to achieve therapeutic outcomes. No HD haplotype-defining SNP among our 34 target alleles occurs within exon 1 of HTT, although we note association of the CCG 7 repeat allele in exon 1 with the disease-associated A haplogroup (Supplementary Figure S2). The mutant exon 1 transcript is found in multiple transgenic models of HD, including full-length YAC128 and BACHD, yet therapeutic rescue of HD phenotypes has been observed in these models by ASO-mediated suppression of mutant HTT outside exon 1.11 This suggests that suppression of full-length mutant HTT is sufficient for phenotypic reversal in full-length murine models, despite continued production of exon 1 transcript containing expanded CAG repeats. Indeed, we detect no change in exon 1 transcript levels in YAC128 brains when full-length HTT mRNA expression is abolished by ASO treatment (data not shown). Exogenous suppression of full-length mutant HTT by ASO is also sufficient for rescue of transcriptional abnormalities in YAC128 mice.37 While the pathogenic impact of the exon 1 transcript remains unclear in HD patients, the efficacious reversal of HD phenotypes by suppression of full-length HTT in transgenic models suggests that reduction of the exon 1 transcript may not be required to elicit therapeutic benefits of ASO treatment.

Expansion of the CAG repeat has been shown to occur on multiple haplotypes in different Caucasian populations.24,26 Here we demonstrate that three intragenic HTT haplotypes, identical across four different populations of European ancestry, account for approximately 90% of HD chromosomes across these groups. This suggests that haplotypes on which repeated CAG expansion events occur are ancestral to all individuals of European descent, and may perhaps be shared by other related populations. The Δ2642 codon deletion (rs149109767), identified here as an exclusive marker of the A1 haplotype, has been observed in HD patients and controls from India38 whereas A1 is entirely absent among both HD and control chromosomes of black South Africans and East Asians where prevalence of HD is dramatically lower.32,33 This suggests that association of the A1 haplotype with HD may occur in all populations of Indo-European ancestry, spanning South Asia, Europe, and American populations of European descent. The frequency of HD on A1, A2, and A3a haplotypes requires detailed haplotype analysis in patient populations from the Middle East, Central Asia, South Asia, and Africa to evaluate the global therapeutic impact of these targets. The high prevalence of expanded CAG on A1, A2, and A3a among all patient populations of European descent, and the presence of these haplotypes in other ancestrally related populations, suggests that these haplotypes may allow allele-specific silencing in the maximum proportion of patients worldwide.

In all populations, there remains a portion of HD patients not eligible for SNP-selective silencing due to HTT haplotype homozygosity. 15% of Italian HD patients in our study are homozygous at all SNPs genotyped, whereas only 2 and 4% are homozygous in the French and Swedish cohorts, respectively (Table 1). On average, 8% of the patients are ineligible for allele-specific therapy due to the absence of heterozygosity at any common SNP in HTT. Only a small proportion of HD patients of European ancestry (<10%) are therefore likely to benefit from more than three allele-specific targets, given that ~80% of patients of European descent are heterozygous for A1, A2, and A3a and ~8% are homozygous at all SNPs. It is nonetheless theoretically possible to achieve treatment of all patients heterozygous for at least one SNP, constituting 92% on average between the four patient populations examined.

Among all patients lacking heterozygous HTT SNPs in this study, all but one are homozygous for specific target alleles of A1, A2, or A3a. ASOs specifically designed for nonselective HTT silencing have shown short-term tolerability and efficacy in transgenic models,11,39 and nonselective treatment of patients homozygous for disease-associated haplotypes may be possible using antisense reagents developed for selective silencing in heterozygous patients. Alternatively, targeting both SNP alleles at one polymorphic site, as suggested for rs7685686,22 would enable selective treatment of approximately half (45.9%) of European HD patients in our study while also allowing the possibility of nonselective treatment of all remaining patients. In vivo efficacy testing of allele-specific silencing reagents, as currently underway in humanized 97/18 mice,21,40 will also crucially inform the future prospects of allele-specific therapy for HD. Should these model studies yield robust therapeutic outcomes, the potential to safely treat over three quarters of the global HD patient population with only three allele-specific drugs, or half of patients selectively and half nonselectively with only two drugs, may present attractive strategies for development.

Antisense reagents designed only for nonselective suppression of HTT, complementary to the HTT transcript but not selective for disease-associated alleles, are also being developed as a therapeutic approach to HD. However, it remains unclear whether adult loss of HTT protein can be adequately compensated in humans over the long-term duration of HTT suppressive therapy. The protective effects of wild-type HTT in various models of HD suggest that specific reduction of mutant HTT may offer a safer and more efficacious alternative to indiscriminate suppression of both wild-type and mutant alleles in patients. Although transient suppression of Hdh appears to be tolerated in mice, adverse effects of human HTT suppression may emerge during clinical trials or long-term treatment. Options for selective HTT suppression should be considered given these potential risks. Any adverse outcomes of long-term total HTT suppression will also need weighed against the cost of developing allele-specific drugs that may be safer but useful only in a subset of genetically eligible patients. The expense of developing secondary and tertiary targets for selective treatment must further be considered versus the number of additional patients that each new target may treat.

The disease-associated alleles identified in this study provide a prioritized list for validation and preclinical screening of allele-specific HTT silencing compounds. One ASO targeting a single allele may treat a maximum of 45% of patients across all patients of European ancestry, whereas three haplotype-specific ASOs targeting A1, A2, and A3a may enable selective treatment of a maximum of 80%. If only one allele-specific drug can be developed, our data supports targeting the A haplogroup to selectively treat the most patients. However, if total HTT suppression results in poor patient outcomes, necessitating selective suppression of mutant HTT in order to achieve clinical benefit, our data recommends targeting the A1 haplotype as the necessary first step to achieving maximum allele-specific treatment of the patient population until multiple antisense drugs can be developed for each disease-associated haplotype.

In summary, we show that HTT is defined by a gene-spanning haplotype block in populations of European descent, and that specific sets of SNPs define gene-spanning haplotypes in both HD patients and controls. To our knowledge, this is the first annotation of dense haplotypes encompassing the HTT gene using whole-genome sequencing data. We identify and validate all polymorphisms specific for the three most common HD haplotypes, comprising >90% of HD chromosomes in four distinct populations of European ancestry. The defining polymorphisms of these haplotypes constitute optimal targets for development of allele-specific silencing compounds, with all other transcribed SNPs providing treatment for fewer patients of European descent. Fine-scale HTT haplotype analysis in other patient populations will be necessary to guide a global allele-specific targeting strategy, particularly where the number of patients may be high, to further clarify the clinical impact worldwide. Targetable HTT haplotypes revealed by this study represent a crucial step toward that objective, and toward safe gene silencing treatment of the greatest number of HD patients.

Materials and Methods

Genotyping and haplotype assignment in Canadian subjects. Ninety-one SNPs were genotyped in 858 Canadian Caucasian HD patients and relatives using the Illumina GoldenGate genotyping array and BeadXpress platform (Illumina, San Diego, CA). Genotypes were called using Illumina GenomeStudio software, and 91-SNP haplotypes reconstructed using PHASE v2.1. Haplotypes were manually annotated, then phased to CAG repeat length and confirmed for sequence identity by familial relationship. 28/91 SNPs in our original panel are rare or occur predominantly in non-Caucasian ethnic groups, leaving 63 SNPs of >1% frequency in European populations (1000 Genomes). Fifty-one of these 63 common SNPs occur within the HTT gene sequence, and were used for annotation of intragenic haplotypes within the extended 63 SNP haplotype.

Analysis of HTT haplotypes in 1000 Genomes. Variant call files encompassing the HTT gene region (GRCh37 3034088-3288007, ± 50 kb of HTT gene, SHAPEIT haplotypes) were downloaded from the 1000 Genomes Project Consortium (phase 1)31 using the Data Slicer tool and analyzed in the R statistical computing environment. 2,166 phased haplotypes of chromosome 4 were available from 1,083 individuals. Chromosomes bearing the intragenic A1 haplotype were identified using rs362307, a previously defined tSNP.24 SNPs present on at least 90% of rs362307 chromosomes (76 with rs362307[T], and therefore at least 70 of 76) but less than 100 of all 2166 chromosomes were identified as candidate A1 markers for further analysis. A2 chromosomes were similarly identified using rs2798235 and rs363080, the defining A2 markers from manual 63-SNP haplotype annotation of the Canadian Caucasian cohort. Discovery of linked A2 variants followed a similar strategy as for linked A1 variants. A3 chromosomes were identified among A haplogroup chromosomes by exclusion of all chromosomes bearing any specific A haplotype-defining SNPs in our 63-SNP panel. A1, A2, and A3 subtype markers were defined as any SNP present on a subset of each haplotype, but on no other chromosomes in the complete 1000 Genomes data set. Following identification of all A1, A2, and A3 variants, phased genotypes of intragenic HTT SNPs present at ≥5% EUR frequency in the 1000 Genomes phase 1 data set were extracted from all 738 European chromosomes and manually annotated in comparison to our directly genotyped 63-SNP haplotype data.

Genotyping and haplotype assignment in European subjects. Two hundred Swedish, 100 French, 33 Finnish, and 291 Italian HD family members were identified from the UBC HD BioBank and in cooperation with IRCCS Neuromed in Pozzilli, Italy. All Swedish, French, and Finnish samples were collected in their respective countries of origin for HD gene mapping studies in the 1990s.41 Of these samples, 120 Swedish, 76 French, 22 Finnish, and 209 Italian family members were identified as phaseable for haplotype and CAG repeat length. All haplotype-defining 63 SNPs genotyped in the Canadian Caucasian cohort were genotyped in the selected European samples, with addition of six novel A1 and A1 subtype SNPs, five novel A2 SNPs, and one novel A3 subtype SNP. European samples were additionally genotyped at 15 SNPs not present in the 63 SNP panel but necessary for reconstruction of haplotypes inferred in prior 4p16.3 locus genotyping.26 Haplotypes in European samples were reconstructed with PHASE v2.1 and manually annotated as for the Canadian Caucasian cohort.

Direct genotyping of HTT A1 variants. A1 markers rs149109767 and rs72239206 are biallelic indels, and were genotyped by fragment analysis in phaseable samples from the UBC HD BioBank with 63-SNP haplotype data. Genotypes of rs149109767 and rs72239206 were phased to SNP haplotype and CAG repeat length by familial relationship. In total, 454 phased, nonredundant HD chromosomes and 652 nonredundant control chromosomes were directly genotyped and phased to CAG repeat length. PCR products containing rs149109767 were amplified using dye-labeled del2642F (6FAM-GCTGGGGAACAGCATCACACCC) and del2642 R (CCTGGAGTTGACTGGAGACTTG). Products containing rs72239206 were amplified with delACTT 3F (GAGGATTGACCACACCACCT) and dye-labeled delACTT 3R (HEX-ATGTGGCCATTTGACACGATA). Primers were multiplexed for ease of genotyping, and PCR products analyzed by ABI 3730xl BioAnalyzer with GeneMapper software (Applied Biosystems, Foster City, CA).

Design of A1-targeted ASOs. Locked nucleic acid gapmer ASOs targeting the mutant ΔACTT (rs72239206) and reference sequence were designed in-house and synthesized by Exiqon on a fee-for-service basis (Exiqon, Woburn, MA). Oligos were resuspended in 1X TE and stored at -20C between transfection experiments.

Passive transfection of HD patient cells with A1-targeted ASOs. Human HD lymphoblasts previously haplotyped as A1/C1 (Applied Biosystems; GM03620, CAG 59/18) were cultured in 2ml complete RPMI media (500,000 cells in 15% fetal bovine serum + 1% pen-strep) with 78, 312, or 1,250 nmol/l of ASO. Cells were incubated 120 hours, and harvested for western blot analysis as described previously.22 Anti-non-muscle myosin IIA (ab24762, Abcam, Cambridge, MA) immunoblotting was used as a loading control.

Allele-specific HTT mRNA quantification. Human HD lymphoblasts previously haplotyped as A1/C1 (GM02176, CAG 44/18) were cultured in complete RPMI media (15% fetal bovine serum + 1% pen-strep). 5 × 106 cells were transfected by electroporation using the Amaxa Nucleofector Kit C for each ASO dose (5,000, 1,250, 312, and 78 nmol/l) in 100 μl nucleofector solution (Lonza, Basel, Switzerland). Transfected cells were cultured for 24 hours, and half of each culture pelleted for RNA extraction, cDNA synthesis, and allele-specific qPCR. Remaining cell culture was propagated for protein analysis at 72 hours. A1 and C1 HTT mRNA transcript was quantified in quadruplicate for each dose of each experiment using an allele-specific TaqMan probe designed to rs362331 (C___2231945_10, Applied Biosystems)) and normalized to glyceraldehyde 3-phosphate dehydrogenase (4333764F). Each transfection experiment was performed three times, with two transfection replicates for each dose in each experiment (n = 4–6 for each data point).

Allele-specific HTT protein quantification. Transfected lymphoblast culture was harvested at 72 hours and pelleted for quantitative western blot analysis. Cells were pelleted by centrifugation at 250 g for 5 minutes at 4 °C and stored at −80 °C. Proteins were extracted by lysis with single detergent lysis buffer (50 mmol/l Tris, 150 mmol/l NaCl, 1% NP40, 40 mmol/l beta-glycerophosphate, 10 mmol/l NaF, 1 mmol/l Na3VO4, 1 mmol/l PMSF, 5 umol/l zVAD, 1X Roche Complete, pH 8.0) and 70 µg of total protein were resolved on 10% low-BIS acrylamide gels and transferred to 0.45 µm nitrocellulose membrane as previously described.22 Membranes were blocked with 5% milk in phosphate-buffered saline, and then blotted with anti-HTT antibody 2166 (Millipore) for detection of HTT. Anti-non-muscle myosin IIA (ab24762, Abcam, Cambridge, MA) immunoblotting was used as a loading control. Secondary antibodies, IR dye 800CW goat anti-mouse (610-131-007, Rockland, Limerick, PA)) and AlexaFluor 680 goat anti-rabbit (A21076, Life Technologies, Carlsbad, CA), were used for detection and membranes were scanned using the LiCor Odyssey Infrared Imaging system. Licor Image Studio Lite was used to quantify the intensity of the individual bands (n = 5–6 for each data point). Figure data are presented as mean ± SEM. Two-way analysis of variance with Bonferroni post hoc test was performed for each dose series and P values illustrated with ** and *** for P = 0.01 and P = 0.001, respectively. Representative images for HTT were chosen.

In vivo ASO treatment with ASOs. YAC128 HD model mice (Slow et al.42) were maintained under a 12-hour light:12-hour dark cycle in a clean facility and given free access to food and water. Experiments were performed with the approval of the animal care committee of the University of British Columbia. ASOs were delivered by intracerebroventricular injection as in Southwell et al.21 at the indicated doses diluted to a final volume of 10 µl in sterile phosphate-buffered saline. Four weeks later, brains were collected and sectioned in a 1 mm coronal rodent brain matrix (ASI Instruments, Warren, MI). The most anterior 2 mm section, containing mostly olfactory bulb, was discarded. The next most anterior 2 mm section, containing mostly cortex and striatum, was divided into hemispheres and lysed as previously described.21 Forty micrograms of total protein were used for allele-specific HTT protein quantification as above.

Ethics statement. Consent and access procedures were in accordance with institutional ethics approval for human research (UBC certificates H05-70532 and H06-70467). Publically available human lymphoblast cell lines were obtained from NIGMS Human Genetic Cell Repository at the Coriell Institute for Medical Research (https://catalog.coriell.org/1/NIGMS).

SUPPLEMENTARY MATERIAL Figure S1. Linkage disequilibrium across the HTT locus in the Canadian patient population, expressed as D' using the 63-SNP panel. Figure S2. All 20 common intragenic haplotypes from the Canadian Caucasian cohort including CCG repeat length (51 SNPs). Figure S3. A1 haplotype frequencies vary among HD and control chromosomes in various populations, as inferred from Δ2642 genotypes. Figure S4. Passive transfection of patient-derived lymphoblasts with rs72239206-complementary ASO selectively reduces mutant HTT protein. Figure S5. ICV injection of YAC128 mice with WT 5-9-5 LNA gapmer complementary to the rs72239206 major allele results in potent reduction of mutant HTT protein in vivo. Table S1. A1 allele counts and relative frequencies among HD and control chromosomes in the UBC HD Biobank and 1000 Genomes Phase I. Table S2. Direct genotyping of A1 haplotype-defining alleles in HD and control chromosomes from the UBC HD BioBank.

Acknowledgments

The authors thank the many HD patients and family members who donated DNA to the UBC HD BioBank to make this study possible. We also acknowledge Michelle Higginson and Nasim Massah for their assistance in genotyping. Funding for this study was provided by the Canadian Institutes of Health Research (CIHR: MOP-84438) and by Teva Pharmaceuticals Ltd. C.K. is funded by a Doctoral Research Award from CIHR. N.H.S. is funded by a postdoctoral fellowship from CIHR. A.L.S. is funded by postdoctoral fellowships from CIHR, the Huntington Society of Canada, and the Michael Smith Foundation for Health Research. We acknowledge Lega Italiana Ricerca Huntington e malattie correlate onlus (www.lirh.it) for assistance in collecting samples from Italy.

Supplementary Material

Supplementary Figures and Tables
Click here for additional data file. (1.6MB, doc)

References

  1. Fisher, ER and Hayden, MR (2014). Multisource ascertainment of Huntington disease in Canada: prevalence and population at risk. Mov Disord 29: 105–114. [DOI] [PubMed] [Google Scholar]
  2. Kay, C, Fisher, ER and Hayden, M. Epidemiology (in Huntington's Disease, 4th edn). Oxford University Press, New York, NY 2014. [Google Scholar]
  3. The Huntington's Disease Collaborative Research Group. (1993). A novel gene containing a trinucleotide repeat that is expanded and unstable on Huntington's disease chromosomes. Cell 72: 971–983. [DOI] [PubMed] [Google Scholar]
  4. Kremer, B, Goldberg, P, Andrew, SE, Theilmann, J, Telenius, H, Zeisler, J et al. (1994). A worldwide study of the Huntington's disease mutation. The sensitivity and specificity of measuring CAG repeats. N Engl J Med 330: 1401–1406. [DOI] [PubMed] [Google Scholar]
  5. Rubinsztein, DC, Leggo, J, Coles, R, Almqvist, E, Biancalana, V, Cassiman, JJ et al. (1996). Phenotypic characterization of individuals with 30-40 CAG repeats in the Huntington disease (HD) gene reveals HD cases with 36 repeats and apparently normal elderly individuals with 36-39 repeats. Am J Hum Genet 59: 16–22. [PMC free article] [PubMed] [Google Scholar]
  6. Zuccato, C, Valenza, M and Cattaneo, E (2010). Molecular mechanisms and potential therapeutical targets in Huntington's disease. Physiol Rev 90: 905–981. [DOI] [PubMed] [Google Scholar]
  7. Roos, RA, Hermans, J, Vegter-van der Vlis, M, van Ommen, GJ and Bruyn, GW (1993). Duration of illness in Huntington's disease is not related to age at onset. J Neurol Neurosurg Psychiatry 56: 98–100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Ross, CA, Aylward, EH, Wild, EJ, Langbehn, DR, Long, JD, Warner, JH et al. (2014). Huntington disease: natural history, biomarkers and prospects for therapeutics. Nat Rev Neurol 10: 204–216. [DOI] [PubMed] [Google Scholar]
  9. Southwell, AL, Skotte, NH, Bennett, CF and Hayden, MR (2012). Antisense oligonucleotide therapeutics for inherited neurodegenerative diseases. Trends Mol Med 18: 634–643. [DOI] [PubMed] [Google Scholar]
  10. Yamamoto, A, Lucas, JJ and Hen, R (2000). Reversal of neuropathology and motor dysfunction in a conditional model of Huntington's disease. Cell 101: 57–66. [DOI] [PubMed] [Google Scholar]
  11. Kordasiewicz, HB, Stanek, LM, Wancewicz, EV, Mazur, C, McAlonis, MM, Pytel, KA et al. (2012). Sustained therapeutic reversal of Huntington's disease by transient repression of huntingtin synthesis. Neuron 74: 1031–1044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Harper, SQ, Staber, PD, He, X, Eliason, SL, Martins, IH, Mao, Q et al. (2005). RNA interference improves motor and neuropathological abnormalities in a Huntington's disease mouse model. Proc Natl Acad Sci USA 102: 5820–5825. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Rodriguez-Lebron, E, Denovan-Wright, EM, Nash, K, Lewin, AS and Mandel, RJ (2005). Intrastriatal rAAV-mediated delivery of anti-huntingtin shRNAs induces partial reversal of disease progression in R6/1 Huntington's disease transgenic mice. Mol Ther 12: 618–633. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Nasir, J, Floresco, SB, O'Kusky, JR, Diewert, VM, Richman, JM, Zeisler, J et al. (1995). Targeted disruption of the Huntington's disease gene results in embryonic lethality and behavioral and morphological changes in heterozygotes. Cell 81: 811–823. [DOI] [PubMed] [Google Scholar]
  15. Duyao, MP, Auerbach, AB, Ryan, A, Persichetti, F, Barnes, GT, McNeil, SM et al. (1995). Inactivation of the mouse Huntington's disease gene homolog Hdh. Science 269: 407–410. [DOI] [PubMed] [Google Scholar]
  16. Zeitlin, S, Liu, JP, Chapman, DL, Papaioannou, VE and Efstratiadis, A (1995). Increased apoptosis and early embryonic lethality in mice nullizygous for the Huntington's disease gene homologue. Nat Genet 11: 155–163. [DOI] [PubMed] [Google Scholar]
  17. Auerbach, W, Hurlbert, MS, Hilditch-Maguire, P, Wadghiri, YZ, Wheeler, VC, Cohen, SI et al. (2001). The HD mutation causes progressive lethal neurological disease in mice expressing reduced levels of huntingtin. Hum Mol Genet 10: 2515–2523. [DOI] [PubMed] [Google Scholar]
  18. Dragatsis, I, Levine, MS and Zeitlin, S (2000). Inactivation of Hdh in the brain and testis results in progressive neurodegeneration and sterility in mice. Nat Genet 26: 300–306. [DOI] [PubMed] [Google Scholar]
  19. Leavitt, BR, Guttman, JA, Hodgson, JG, Kimel, GH, Singaraja, R, Vogl, AW et al. (2001). Wild-type huntingtin reduces the cellular toxicity of mutant huntingtin in vivo. Am J Hum Genet 68: 313–324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Carroll, JB, Warby, SC, Southwell, AL, Doty, CN, Greenlee, S, Skotte, N et al. (2011). Potent and selective antisense oligonucleotides targeting single-nucleotide polymorphisms in the Huntington disease gene / allele-specific silencing of mutant huntingtin. Mol Ther 19: 2178–2185. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Southwell, AL, Skotte, NH, Kordasiewicz, HB, Østergaard, ME, Watt, AT, Carroll, JB et al. (2014). In vivo evaluation of candidate allele-specific mutant huntingtin gene silencing antisense oligonucleotides. Mol Ther 22: 2093–2106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Skotte, NH, Southwell, AL, Østergaard, ME, Carroll, JB, Warby, SC, Doty, CN et al. (2014). Allele-specific suppression of mutant huntingtin using antisense oligonucleotides: providing a therapeutic option for all Huntington disease patients. PLoS One 9: e107434. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Østergaard, ME, Southwell, AL, Kordasiewicz, H, Watt, AT, Skotte, NH, Doty, CN et al. (2013). Rational design of antisense oligonucleotides targeting single nucleotide polymorphisms for potent and allele selective suppression of mutant Huntingtin in the CNS. Nucleic Acids Res 41: 9634–9650. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Warby, SC, Montpetit, A, Hayden, AR, Carroll, JB, Butland, SL, Visscher, H et al. (2009). CAG expansion in the Huntington disease gene is associated with a specific and targetable predisposing haplogroup. Am J Hum Genet 84: 351–366. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Pfister, EL, Kennington, L, Straubhaar, J, Wagh, S, Liu, W, DiFiglia, M et al. (2009). Five siRNAs targeting three SNPs may provide therapy for three-quarters of Huntington's disease patients. Curr Biol 19: 774–778. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lee, JM, Gillis, T, Mysore, JS, Ramos, EM, Myers, RH, Hayden, MR et al. (2012). Common SNP-based haplotype analysis of the 4p16.3 Huntington disease gene region. Am J Hum Genet 90: 434–444. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Rubinsztein, DC, Leggo, J, Goodburn, S, Barton, DE and Ferguson-Smith, MA (1995). Haplotype analysis of the delta 2642 and (CAG)n polymorphisms in the Huntington's disease (HD) gene provides an explanation for an apparent ‘founder' HD haplotype. Hum Mol Genet 4: 203–206. [DOI] [PubMed] [Google Scholar]
  28. Zhang, Y, Engelman, J and Friedlander, RM (2009). Allele-specific silencing of mutant Huntington's disease gene. J Neurochem 108: 82–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Novelletto, A, Persichetti, F, Sabbadini, G, Mandich, P, Bellone, E, Ajmar, F et al. (1994). Analysis of the trinucleotide repeat expansion in Italian families affected with Huntington disease. Hum Mol Genet 3: 93–98. [DOI] [PubMed] [Google Scholar]
  30. Abecasis, GR, Altshuler, D, Auton, A, Brooks, LD, Durbin, RM, Gibbs, RA et al. (2010). A map of human genome variation from population-scale sequencing. Nature 467: 1061–1073. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Abecasis, GR, Auton, A, Brooks, LD, DePristo, MA, Durbin, RM, Handsaker, RE et al. (2012). An integrated map of genetic variation from 1,092 human genomes. Nature 491: 56–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Baine, FK, Kay, C, Ketelaar, ME, Collins, JA, Semaka, A, Doty, CN et al. (2013). Huntington disease in the South African population occurs on diverse and ethnically distinct genetic haplotypes. Eur J Hum Genet 21: 1120–1127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Warby, SC, Visscher, H, Collins, JA, Doty, CN, Carter, C, Butland, SL et al. (2011). HTT haplotypes contribute to differences in Huntington disease prevalence between Europe and East Asia. Eur J Hum Genet 19: 561–566. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Almqvist, E, Spence, N, Nichol, K, Andrew, SE, Vesa, J, Peltonen, L et al. (1995). Ancestral differences in the distribution of the delta 2642 glutamic acid polymorphism is associated with varying CAG repeat lengths on normal chromosomes: insights into the genetic evolution of Huntington disease. Hum Mol Genet 4: 207–214. [DOI] [PubMed] [Google Scholar]
  35. Bennett, CF and Swayze, EE (2010). RNA targeting therapeutics: molecular mechanisms of antisense oligonucleotides as a therapeutic platform. Annu Rev Pharmacol Toxicol 50: 259–293. [DOI] [PubMed] [Google Scholar]
  36. Sathasivam, K, Neueder, A, Gipson, TA, Landles, C, Benjamin, AC, Bondulich, MK et al. (2013). Aberrant splicing of HTT generates the pathogenic exon 1 protein in Huntington disease. Proc Natl Acad Sci USA 110: 2366–2370. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Stanek, LM, Yang, W, Angus, S, Sardi, PS, Hayden, MR, Hung, GH et al. (2013). Antisense oligonucleotide-mediated correction of transcriptional dysregulation is correlated with behavioral benefits in the YAC128 mouse model of Huntington's disease. J Huntingtons Dis 2: 217–228. [DOI] [PubMed] [Google Scholar]
  38. Saleem, Q, Roy, S, Murgood, U, Saxena, R, Verma, IC, Anand, A et al. (2003). Molecular analysis of Huntington's disease and linked polymorphisms in the Indian population. Acta Neurol Scand 108: 281–286. [DOI] [PubMed] [Google Scholar]
  39. Boudreau, RL, McBride, JL, Martins, I, Shen, S, Xing, Y, Carter, BJ et al. (2009). Nonallele-specific silencing of mutant and wild-type huntingtin demonstrates therapeutic efficacy in Huntington's disease mice. Mol Ther 17: 1053–1063. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Southwell, AL, Warby, SC, Carroll, JB, Doty, CN, Skotte, NH, Zhang, W et al. (2013). A fully humanized transgenic mouse model of Huntington disease. Hum Mol Genet 22: 18–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Andrew, S, Theilmann, J, Almqvist, E, Norremolle, A, Lucotte, G, Anvret, M et al. (1993). DNA analysis of distinct populations suggests multiple origins for the mutation causing Huntington disease. Clin Genet 43: 286–294. [DOI] [PubMed] [Google Scholar]
  42. Slow, EJ, van Raamsdonk, J, Rogers, D, Coleman, SH, Graham, RK, Deng, Y et al. (2003). Selective striatal neuronal loss in a YAC128 mouse model of Huntington disease. Hum Mol Genet 12: 1555–1567. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Figures and Tables
Click here for additional data file. (1.6MB, doc)

Articles from Molecular Therapy are provided here courtesy of The American Society of Gene & Cell Therapy

RESOURCES