Preclinical evaluation of stereopure antisense oligonucleotides for allele-selective lowering of mutant HTT

Abstract

Huntington's disease (HD) is an autosomal dominant disease caused by the expansion of cytosine-adenine-guanine (CAG) repeats in one copy of the HTT gene (mutant HTT, mHTT). The unaffected HTT gene encodes wild-type HTT (wtHTT) protein, which supports processes important for the health and function of the central nervous system. Selective lowering of mHTT for the treatment of HD may provide a benefit over nonselective HTT-lowering approaches, as it aims to preserve the beneficial activities of wtHTT. Targeting a heterozygous single-nucleotide polymorphism (SNP) where the targeted variant is on the mHTT gene is one strategy for achieving allele-selective activity. Herein, we investigated whether stereopure phosphorothioate (PS)- and phosphoryl guanidine (PN)-containing oligonucleotides can direct allele-selective mHTT lowering by targeting rs362273 (SNP3). We demonstrate that our SNP3-targeting molecules are potent, durable, and selective for mHTT in vitro and in vivo in mouse models. Through comparisons with a surrogate for the nonselective investigational compound tominersen, we also demonstrate that allele-selective molecules display equivalent potency toward mHTT with improved durability while sparing wtHTT. Our preclinical findings support the advancement of WVE-003, an investigational allele-selective compound currently in clinical testing (NCT05032196) for the treatment of patients with HD.

Keywords: HTT; Huntington’s disease; MT: Oligonucleotides: Therapies and Applications; SELECT-HD; SNP3; WVE-003; allele-selective silencing; mutant huntingtin; phosphoryl guanidine; stereopure antisense oligonucleotide; wild-type huntingtin.

Graphical abstract

Figure 1

Rp linkages direct the rate, position, and selectivity of RNase H-mediated cleavage (A) Rates of RNase H-catalyzed RNA cleavage for oligonucleotide-RNA duplexes containing stereorandom or stereopure molecules. The mean percentage of full-length RNA remaining is plotted with respect to time (minutes) for the indicated molecule (n = 3). Data are shown as mean ± SD. Data were fit to nonlinear, one-phase decay using GraphPad Prism software. (B) RNase H cleavage maps in RNA surrogate (RNA-1) when duplexed with the indicated oligonucleotide. Arrows indicate the sites of cleavage. The length of the arrows represents the percentage of product derived from cleavage at that site. The 10-min time point was used to generate cleavage maps. Legend for oligonucleotide cartoons is shown to the right. (C–E) RNase H cleavage activities are shown for the stereorandom molecule HTT-1 (C), stereopure molecule HTT-2 (D), and stereopure molecule HTT-3 (E). The mean percentage of full-length mHTT (RNA-A) and wtHTT (RNA-G) surrogate RNA remaining is plotted with respect to time (minutes) for the indicated oligonucleotide-RNA duplex. The SNP is yellow. Data are presented as mean ± SE, n = 3. PS stereochemistry is indicated by ˆ symbols (Rp = red; Sp = blue; stereorandom = black).

Figure 2

PS stereochemistry can promote allele-selective activity at HTT SNP3 Schematic depicting allele-selectivity strategy. The SNP is depicted in both images. In the mHTT complex (yellow SNP), the oligonucleotide is complementary to the RNA. In the wtHTT complex (orange SNP), there is mismatch between that oligonucleotide and the RNA. (A) RNase H cleavage activity (top panels) is shown for stereorandom molecule HTT-1233 (B) and stereopure molecules HTT-956 (C), HTT-957 (D), and HTT-958 (E) where the SNP position varies between position 10 (P10) and position 12 (P12) when read from the 5′-end of the oligonucleotide. For all top panels, the mean percentage of full-length mHTT (RNA-A) and wtHTT (RNA-G) surrogate RNA remaining is plotted with respect to time (minutes) for the indicated antisense oligonucleotide-RNA duplex. Data are presented as mean ± SD. In cases where error bars are not shown, the error was smaller than the symbol. For all bottom panels, the RNA surrogates (depicted in green, 5′-3′) with the SNP shown in yellow are illustrated with respect to each oligonucleotide (shown 3′-5′).

Figure 3

SNP3-targeting oligonucleotides selectively and dose-dependently decreases mHTT in vitro Schematic representation of oligonucleotides is shown in (A). RNase H experiments performed with synthetic RNA substrates corresponding to mHTT and wtHTT transcripts. Data are presented as mean ± SEM, n = 3. Mean percentages of the indicated full-length RNA substrate remaining over time are plotted for HTT-1526 (B) and HTT-1533 (C). Patient-derived motor neurons homozygous for SNP3 were treated with HTT-1526, HTT-1533, or HTT-2723 at varying levels up to 60-μM concentration for 7 days. Dose-response curves for HTT normalized to TUBB RNA measured using the bDNA assay are shown in (D). Data shown as mean ± SD, n = 4. (E) Patient-derived motor neurons heterozygous for SNP3 were treated with NTC, HTT-2723, HTT-1526, or HTT-1533 (1.6 μM, 8 μM, or 40 μM). Data depict HTT RNA remaining for wtHTT allele (blue bar) and mHTT allele (red bar). Next-generation sequencing of PCR amplicons covering the SNP3 region was performed to quantify the amount of each allele remaining after treatment. Total HTT lowering was quantified by qPCR and normalized to HPRT1. All treatment conditions were then normalized to mock treatment. Data shown as mean ± SD (mock, n = 30; HTT-2723, n = 2; HTT-1526, n = 2, HTT-1533, n = 2). ∗∗∗p ≤ 0.001, ∗∗∗∗p ≤ 0.0001. p values were calculated via White-adjusted three-way ANOVA followed by two-tailed post hoc tests comparing mutant and wild-type expression per molecule/dose allowing unequal variance.

Figure 4

SNP3-targeting HTT-1533 performs better in patient iPSC-derived neurons than first-generation oligonucleotides Schematic representation of oligonucleotides is shown in (A). HTT-164 targets the U variant of SNP1 (rs362307). HTT-273 targets the U variant of SNP2 (rs362331). HTT-1533 targets the A variant of SNP3 (rs362273). Results of genotyping and phasing for iPS-109Q cells are shown in (B). The HTT gene contains both the CAG repeat sequence (yellow) and the SNP (blue box), which need to be in-phase for allele-selective silencing to work. The SNP is heterozygous (blue and green boxes), with the targeted mutant variant (blue) on same copy as the expanded CAG repeat (long yellow box). The wtHTT allele contains a non-expanded CAG tract (short yellow box) and the non-targeted variant at the SNP (green). Genotyping and phasing results for iPS-109Q cells are summarized. (C) Patient iPSC-derived neurons heterozygous for SNP1, SNP2, and SNP3 were treated with HTT-164, HTT-273, HTT-1533 or the corresponding NTC at increasing concentrations (1.6, 8.0, or 40 μM) for 7 days. Percentage of mHTT (red) and wtHTT (blue) transcript expression relative to mock-treated control cells is shown. Data shown as mean ± SEM, n = 3. Stats: three-way ANOVA with two-tailed post hoc comparisons of mHTT to wtHTT per molecule per dose assuming equal variance ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001.

Figure 5

SNP3-targeting oligonucleotides dose-dependently decrease mHTT in cortex and striatum of BACHD mice Dosing regimen for administration of PBS or oligonucleotide (12.5 μg, 25 μg, 50 μg, or 100 μg) to BACHD mice and sample collection 15 days after first dose is shown (A). The relative fold change of human mHTT to mouse Tubb3 as a percentage of PBS in the cortex (B) and striatum (C) is shown at 2 weeks after the first administration (n = 7–8 per treatment). Data shown as mean ± SD (∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001, ∗∗∗∗p ≤ 0.0001 vs. PBS). p values were calculated via one-way ANOVA followed by two-tailed post-hoc comparisons to PBS assuming equal variance. Dosing regimen for administration and sample collection 29 days after first dose is shown in (D). The relative fold change of human mHTT to mouse Tubb3 as a percentage of PBS in the cortex (E) and striatum (F) is shown at 4 weeks after the first administration (n = 7–8 per treatment). Data shown as mean ± SD (∗p ≤ 0.05, ∗∗p ≤ 0.01,∗∗∗p ≤ 0.001, ∗∗∗∗p ≤ 0.0001 vs. PBS). p values were calculated via one-way ANOVA followed by two-tailed post hoc comparisons to PBS assuming equal variance.

Figure 6

SNP3-targeting oligonucleotides durably reduce mHTT in the cortex and striatum of BACHD mice Dosing regimens and time of sample collection post-first dose for studies 1 and 2 are shown (A). Oligonucleotides or PBS were administered by ICV injection (3 × 100 μg). For study 1, the relative fold change of human mHTT to mouse Tubb3 as a percentage of PBS in the cortex (B) and striatum (C) is shown at 2, 4, 8, and 12 weeks after the first administration (n = 7–8 per treatment). For study 2, the relative fold change of human mHTT to mouse Tubb3 as a percentage of PBS in the cortex (D) and striatum (E) is shown at 2, 4, and 12 weeks after the first administration (n = 8 per treatment). For (B)–(E), data are shown as mean ± SD (∗p ≤ 0.05; ∗∗p ≤ 0.01; ∗∗∗p ≤ 0.001; ∗∗∗∗p ≤ 0.0001 compared with PBS). p values were calculated via two-way ANOVA (B–D), or White-adjusted two-way ANOVA (E) followed by two-tailed post-hoc comparisons to PBS per week with equal (B–D) or unequal (E) variance.

Figure 7

SNP3-targeting oligonucleotides selectively decrease mHTT RNA and protein in the CNS of Hu97/18 mice Dosing regimen and timing of sample collection post-first dose for administration of PBS or oligonucleotide (3 × 100 μg) to Hu97/18 mice is shown in (A). The relative fold change of human mHTT (red) or wtHTT (blue) transcript to mouse Tubb3 as a percentage of PBS in the cortex (B), striatum (C), and hippocampus (D) are shown for tissue collected 4, 8, and 12 weeks after the first administration (n = 6–8 per treatment). Data are shown as mean ± SD. ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001, ∗∗∗∗p ≤ 0.0001. p values were calculated via three-way ANOVA (B and C) or White-adjusted three-way ANOVA (D) followed by two-tailed post-hoc comparisons of wtHTT to mHTT at each time point with equal (B and C) or unequal (D) variance. Representative western blots from cortex collected 4, 8, and 12 weeks after the first administration are shown to the left (E). The fold change of human mHTT protein (yellow) or wtHTT protein (dark blue) with respect to mouse vinculin protein in the cortex is shown (n = 7–8) to the right (E). Data are shown as mean ± SD. p values were calculated via three-way ANOVA followed by two-tailed post-hoc comparisons of wtHTT and mHTT at each time point with equal variance, ∗p ≤ 0.05, ∗∗∗p ≤ 0.001. Uncropped western blots are shown in Figure S5.