Suppression of Huntington's Disease Somatic Instability by Transcriptional Repression and Direct CAG Repeat Binding

Abstract

Huntington's disease (HD) arises from a CAG expansion in the huntingtin (HTT) gene beyond a critical threshold. A major thrust of current HD therapeutic development is lowering levels of mutant HTT mRNA (mHTT) and protein (mHTT) with the aim of reducing the toxicity of these product(s). Human genetic data also support a key role for somatic instability (SI) in HTT's CAG repeat - whereby it lengthens with age in specific somatic cell types - as a key driver of age of motor dysfunction onset. Thus, an attractive HD therapy would address both mHTT toxicity and SI, but to date the relationship between SI and HTT lowering remains unexplored. Here, we investigated multiple therapeutically-relevant HTT-lowering modalities to establish the relationship between HTT lowering and SI in HD knock-in mice. We find that repressing transcription of mutant Htt (mHtt) provides robust protection from SI, using diverse genetic and pharmacological approaches (antisense oligonucleotides, CRISPR-Cas9 genome editing, the Lac repressor, and virally delivered zinc finger transcriptional repressor proteins, ZFPs). However, we find that small interfering RNA (siRNA), a potent HTT-lowering treatment, lowers HTT levels without influencing SI and that SI is also normal in mice lacking 50% of total HTT levels, suggesting HTT levels, per se, do not modulate SI in trans. Remarkably, modified ZFPs that bind the mHtt locus, but lack a repressive domain, robustly protect from SI, despite not reducing HTT mRNA or protein levels. These results have important therapeutic implications in HD, as they suggest that DNA-targeted HTT-lowering treatments may have significant advantages compared to other HTT-lowering approaches, and that interaction of a DNA-binding protein and HTT's CAG repeats may provide protection from SI while sparing HTT expression.

Figure 1:. Chronic ASO treatment reduces somatic instability in mutant Htt’s CAG repeat and reduces Htt transcriptional rate in the livers of Htt^Q111/+ mice.

A) Overview of peripheral ASO administration mouse cohorts. B) Chronic ASO treatment reduces levels of both wildtype (63% reduction; p = 0.001) and mutant HTT (54% reduction; p = 0.086) in the liver, as assayed with MSD at 14 months of age following 12 months of ASO treatment (N = 7–8 per arm). C) Total Htt mRNA, as assayed by qRT-PCR, is also reduced by chronic ASO treatment (94% reduction; p<0.0001). D) Exemplar traces of the size distribution of PCR products from a CAG-spanning PCR reaction. The top panel arises from a mouse treated with saline, while the bottom one is treated with Htt-targeted ASO. E) Somatic instability is reduced by chronic peripheral dosing of HTT ASO, regardless of length of treatment (39%, 35%, 51% reductions at 7.5, 10 and 14 months, respectively; p = 0.0002, p < 0.0001, p < 0.0001). F) Chronic ASO treatment reduces the levels of Htt pre-mRNA, as quantified by qRT-PCR with primer pairs that span the indicated exon-intron boundaries, with reduction increasing downstream of the ASO binding site (exon 2-intron 2: 44% reduction, p = 0.035; exon 37-intron 38: 66%, p = 0.009; exon 66-intron 66: 69%, p = 0.002). G) Overview of locations of the ASO target and exon/intron primer pairs. For box and whisker plots, each data point is shown, and the horizontal lines indicate the 25th, 50th and 75th percentiles of the data, with vertical lines indicating the range, with outliers detached from the vertical lines.

Figure 2:. Partial lac repressor-mediated transcriptional repression of mHtt reduces somatic instability in Htt’s CAG repeat in the livers of Htt^Q140 mice.

A) Schematic of study design indicating when mice in 3 cohorts were treated with IPTG, reducing LacR-mediated repression of the Htt^Q140 allele. B) Overview of the mechanisms of action of the lac repressor system. C) Protein knockdown at sacrifice: wtHTT is unaffected by the lac repressor (p = 0.52), while mHTT levels are reduced (46%, p < 0.0001). D) Knockdown of total Htt mRNA from liver from replicate mouse experiment (27%, p = 0.018; NB, only mHTT is repressed, so this non-allele selective mRNA assay represents an under-estimate of mHTT-selective knockdown). E) Somatic instability is reduced via partial LacR-mediated repression of mHtt, with greater suppression of instability given longer suppression durations. F) Exemplar traces of repressed and unrepressed mice. For box and whisker plots, each data point is shown, and the horizontal lines indicate the 25th, 50th and 75th percentiles of the data, with vertical lines indicating the range, with outliers detached from the vertical lines.

Figure 3:. CRISPR/Cas9-mediated deletion of the mHtt promoter region reduces SI in Htt’s CAG repeat tract.

A) Cartoon indicating the location of the gRNA pair referred to as “QPro,” spanning an approximately 1,200bp region, that targets unique sequences in the humanized Htt^Q111 allele that are not present in mouse Htt. B) AAV8-delivered treatment with QPro results in robust, allele-selective lowering of HTT protein in the livers of Htt^Q111+; Cas9 mice. C) Treatment with AAV8::QPro at 2.5 months of age results in reduction of expansion index at 8 months of age. D) Exemplar traces of CAG tract expansions in untreated and treated animals.

Figure 4:. siRNA-mediated repression of Htt does not alter its somatic instability or transcriptional rate in Htt^Q111/+ mice.

A) Schematic of study design. B) Chronic (10 mg/kg/month) siRNA treatment reduces levels of both wildtype (74% reduction; p < 0.0001) and mutant HTT (55% reduction; p < 0.0001) in the liver, as assayed with MSD at 6 months of age following 3.5 months of treatment (N = 10 per arm). C) Total Htt mRNA, as assayed by mouse Htt qRT-PCR, is also reduced by siRNA treatment (64% reduction; N = 10, p < 0.0001). D) Htt pre-mRNA levels are not impacted by siRNA treatment (NS). E) Chronic siRNA treatment does not change SI in HTT’s CAG tract (NS).

Figure 5:. Genetic loss of wtHTT does not influence somatic instability.

A-B) Western blot demonstrating the lack of wildtype HTT in Htt^Q111/− mice, and no alteration in mHTT levels. C) Instability in Htt’s CAG repeat is not altered in Htt^Q111/− mice compared to Htt^Q111/+ mice in either the striatum or liver.

Figure 6:. ZFP treatment reduces HTT’s transcriptional rate and blocks CAG SI, while delta-KRAB lowers SI but spares transcription.

A) Schematic of ZFP constructs: the DNA-binding domain (DBD) binds CAG repeats, while the KRAB domain induces repressive chromatin. B) Schematic of treatment paradigm - each subject mouse is treated with two viruses - one in each hemisphere. C) Htt transcript levels as measured by bDNA assay decrease after ZFP treatment (full ZFP vs KRAB-only, 20% reduction, p = 0.0002) but are unaffected by DBD-only treatment (NS). NB, this is a non-allele selective assay, and so likely underestimates the amount of allele-selective lowering of mHtt. D) Treatment with the complete ZFP constructs leads to a marked reduction in exon-1 only mHtt (Htt1a) levels as measured by bDNA assay (50% reduction, p < 0.0001), but are unaffected by DBD-only treatment (NS). E) Somatic instability of mHtt’s CAG tracts is reduced (70%, p < 0.0001) by the full ZFP construct, with the DBD-only construct resulting in an intermediate reduction (42%, p < 0.0001). F) Exemplar traces from each treatment group.

Figure 7/. graphical abstract: Summary of our findings of the impact of HTT-lowering treatments on somatic instability.