Mutant huntingtin protein decreases with CAG repeat expansion: implications for therapeutics and bioassays

Abstract

Huntington’s disease is an inherited neurodegenerative disorder caused by a CAG repeat expansion that encodes a polyglutamine tract in the huntingtin (HTT) protein. The mutant CAG repeat is unstable and expands in specific brain cells and peripheral tissues throughout life. Genes involved in the DNA mismatch repair pathways, known to act on expansion, have been identified as genetic modifiers; therefore, it is the rate of somatic CAG repeat expansion that drives the age of onset and rate of disease progression. In the context of an expanded CAG repeat, the HTT pre-mRNA can be alternatively processed to generate the HTT1a transcript that encodes the aggregation prone and highly pathogenic HTT1a protein. This may be a mechanism through which somatic CAG repeat expansion exerts its pathogenic effects, as the longer the CAG repeat, the more HTT1a and HTT1a is produced. The allelic series of knock-in mouse models, HdhQ20, HdhQ50, HdhQ80, HdhQ111, CAG140 and zQ175 with polyglutamine expansions of 20, 50, 80, 111, 140 and ∼190, can be used to model the molecular and cellular consequences of CAG repeat expansion within a single neuron. By western blot of cortical lysates, we found that mutant HTT levels decreased with increasing CAG repeat length; mutant HTT was only 23 and 10% of wild-type levels in CAG140 and zQ175 cortices, respectively. To identify the optimal bioassays for detecting the full-length HTT and HTT1a isoforms, we interrogated the pairwise combinations of seven well-characterized antibodies on both the ‘homogeneous time-resolved fluorescence’ and ‘Meso Scale Discovery’ platforms. We tested 32 assays on each platform to detect ‘full-length mutant HTT’, HTT1a, ‘total mutant HTT’ (full-length HTT and HTT1a) and ‘total full-length HTT’ (mutant and wild type). None of these assays recapitulated the full-length mutant HTT levels as measured by western blot. We recommend using isoform- and species-specific assays that detect full-length mutant HTT, HTT1a or wild-type HTT as opposed to those that detect more than one isoform simultaneously. Our finding that as the CAG repeat expands, full-length mutant HTT levels decrease, while HTT1a and HTT1a levels increase has implications for therapeutic strategies. If mutant HTT levels in cells containing (CAG)₂₀₀ are only 10% of wild-type, HTT-lowering strategies targeting full-length HTT at sequences 3ʹ to Intron 1 HTT will predominantly lower wild-type HTT, as mutant HTT levels in these cells are already depleted. These data support a therapeutic strategy that lowers HTT1a and depletes levels of the HTT1a protein.

Landles et al. used mouse models to investigate what happens to huntingtin RNA and protein levels as the mutant CAG repeat expands in brain cells of people with Huntington’s disease throughout life. They show that the full-length mutant HTT protein decreases, whereas the aggregation prone and pathogenic HTT1a protein increases.

Graphical Abstract

Graphical Abstract.

Introduction

Huntington’s disease is an inherited neurodegenerative disorder that manifests with motor, cognitive and psychiatric impairments.¹ It is caused by an expanded CAG repeat in exon 1 of the huntingtin (HTT) gene that results in an expanded glutamine tract in the huntingtin (HTT) protein.² Repeats of (CAG)₄₀ or more, as measured in blood, are fully penetrant, individuals with (CAG)₃₅ or less remain unaffected,³ whereas individuals with expansions of around (CAG)₆₅ or more develop symptoms in childhood or adolescence.⁴ The CAG repeat is somatically unstable, expanding with age in specific brain regions.^5-7 Given that several DNA mismatch repair pathway genes, known to modify somatic CAG repeat expansion, have been identified as genetic modifiers of Huntington’s disease, this phenomenon is considered to drive the age of onset and rate of disease progression.^8-11  HTT pre-mRNAs with expanded CAG repeats can be alternatively processed to generate the HTT1a transcript that encodes the aggregation-prone and highly pathogenic HTT1a protein.^12,13 The levels of HTT1a increase with increasing CAG repeat length and therefore may be a mechanism through which somatic CAG expansion exerts its pathogenic effects.

It is essential that the levels of HTT protein isoforms can be measured in disease models and clinical samples to allow the interpretation of mechanistic studies and pre-clinical and clinical interventions. This is particularly important for potential therapies that target HTT directly through HTT-lowering approaches.¹⁴ Assays that detect soluble or aggregated HTT isoforms have been developed by utilizing pairs of anti-HTT antibodies. For cell lysates or tissue extracts from Huntington’s disease models, assays have been developed using homogeneous time-resolved fluorescence (HTRF; also termed TR-FRET),^15,16 Meso Scale Discovery (MSD)^17,18 and amplified luminescent proximity homogeneous assay (AlphaLISA)¹⁹ platforms. More sensitive single molecule counting assays have been needed to detect HTT in CSF.^20,21

In this study, we investigated the effect of the Huntington’s disease mutation on mutant HTT protein levels as well as the comparative performance of HTRF and MSD assays designed to detect soluble HTT isoforms. We used cortical lysates from heterozygous and homozygous HdhQ20, HdhQ50, HdhQ80, HdhQ111, CAG140 and zQ175 knock-in mice, together with their wild-type littermates at 11 weeks of age and included lysate from YAC128 transgenic mice, to control for the detection of HTT levels above wild-type. This allelic series of knock-in mice can be used to model changes in mutant HTT levels in response to large somatic CAG expansions within a single neuron.

We used western blotting to estimate full-length wild-type and mutant HTT levels in cortical lysates and then assessed the ability of 12 HTRF and 12 MSD assays to measure ‘full-length mutant HTT’, 2 HTRF and 2 MSD assays to measure HTT1a, 6 HTRF and 6 MSD assays to measure ‘total mutant HTT’ (full-length and HTT1a) and 12 HTRF and 12 MSD assays to measure ‘total full-length HTT’ (mutant and wild type). The western blots demonstrated that mutant HTT levels in lysates from homozygous HdhQ20 and HdhQ50 mice were equivalent to wild type. However, as the CAG repeat in the knock-in alleles of the HdhQ80, HdhQ111, CAG140 and zQ175 lines increased, full-length mutant HTT levels in the homozygous knock-in mice decreased to ∼10% of wild type in zQ175 cortices. In contrast, HTT1a protein levels increased. This finding has significant implications for HTT-lowering therapeutic interventions, and our data stress that it is essential to understand the effect of polyQ length on the performance of HTT bioassays before using them for mechanistic or therapeutic applications in the pre-clinical or clinical setting.

Materials and methods

Ethics statement

All procedures were performed in accordance with the Animals (Scientific Procedures) Act, 1986, and approved by the University College London Ethical Review Process Committee.

Mouse breeding and maintenance

Heterozygous, homozygous and wild-type littermates were imported from the CHDI Foundation colonies at the Jackson Laboratory (Bar Harbor, ME, USA) for the following knock-in lines on a C57BL/6J background: HdhQ20, HdhQ50, HdhQ80 and HdhQ111,²² CAG140²³ and zQ175^24,25 and housed at University College London (UCL) until 11 weeks of age. YAC128 mice²⁶ were bred at UCL by pairing YAC128 males with C57BL/6J females (Charles River, the Netherlands) and tissues collected at 9 weeks of age. Mouse husbandry and health monitoring were as previously described.²⁷ Animals were kept in individually ventilated cages with Aspen Chips 4 Premium bedding (Datesand) and environmental enrichment comprising chew sticks and a play tunnel (Datesand). All mice had constant access to water and food (Teklad global 18% protein diet, Envigo, Limburg, the Netherlands). Temperature was regulated at 21 ± 1°C, and mice were kept on a 12-h light/dark cycle. The facility was barrier maintained, and quarterly non-sacrificial (Federation of European Laboratory Animal Science Associations) screens found no evidence of pathogens. Tissues were harvested, snap frozen in liquid nitrogen and stored at −80°C.

Genotyping and CAG repeat sizing

YAC128 mice were genotyped as previously described.²⁸ The polyQ repeat of 125 glutamines in YAC128 mice is encoded by (CAG)₂₃(CAA)₃CAGCAA(CAG)₈₀(CAA)₃CAGCAA(CAG)₁₀CAACAG, which is stable on germline transmission.²⁹ For imported mice, tail biopsies were collected upon sacrifice for genotype confirmation. The HdhQ20, HdhQ50, HdhQ80 and HdhQ111 mice were genotyped as previously described for HdhQ20²⁸ and the CAG140 and zQ175 mice as described for zQ175.²⁸ CAG repeat sizing had been performed by the CHDI Foundation before shipment to UCL, and CAG lengths are summarized in Supplementary Table 1.

Antibodies

Antibodies are summarized in Supplementary Table 2, and their epitope locations on the HTT protein are shown in Fig. 1A.

Figure 1.

Full-length mutant HTT levels decrease with increasing CAG repeat length. (A) Schematic showing the epitope locations of the HTT antibodies used in this study. Details of immunogens and antibody epitopes are given in Supplementary Table 2. (B) Western blots of full-length HTT, as detected by D7F7 in cortical lysates from wild-type, heterozygous and homozygous mice (n = 3/gender/genotype) for each of the knock-in mouse lines: HdhQ20, HdhQ50, HdhQ80, HdhQ111, CAG140 and zQ175. The DM15 antibody against HDAC4 was used as a loading control. The full-length blots and their quantification are shown in Supplementary Figs 4–9 and Supplementary Tables 16–21. (C) Quantification of the levels of total full-length HTT (wild-type and mutant) from the western blots shown in B. (D) Western blot of full-length HTT, as detected by D7F7 in cortical lysates from wild-type and YAC128 transgenic mice (n = 3/gender/genotype). DM15 antibody against HDAC4 was used as a loading control. The full-length blot and its quantification are shown in Supplementary Fig. 10 and Supplementary Table 22. Statistical analysis was by Student’s t-test and one-way ANOVA with Bonferroni post hoc correction. The test statistic, degrees of freedom and P-values for the ANOVA are provided in Supplementary Table 6. Error bars are mean ± standard error of the mean (SEM). *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001. aa, amino acid, polyQ, polyglutamine, polyP, polyproline, WT, wild type, HET, heterozygote, HOM, homozygote, TG, transgenic.

Western blotting

Cortical lysates were prepared in ice-cold radio immunoprecipitation assay (RIPA) buffer [50 mM tris(hydroxymethyl)aminomethane (TRIS) pH 8.0, 150 mM NaCl, 1% NP40, 0.5% Na-cholate and 0.1% sodium dodecyl sulphate (SDS)], NP40 buffer (50 mM TRIS pH 7.5, 150 mM NaCl and 1% NP40) or 2-(4-(2-hydroxyethyl)-1-piperazinyl)-ethanesulfonic acid (HEPES) buffer [50 mM HEPES pH 7.0, 150 mM NaCl, 10 mM ethylenediaminetetraacetic acid (EDTA), 1% NP40, 0.5% Na-cholate and 0.1% SDS] with complete protease inhibitors (Roche). Proteins were denatured in Laemmli loading buffer as previously described³⁰; 25 µg of protein was separated by 7.5% SDS-polyacrylamide acrylamide gel electrophoresis (Criterion, Bio-Rad) and blotted onto nitrocellulose membranes. Primary antibody dilution was 1:1000 for HTT (D7F7) and 1:1000 for HDAC4 (DM15), blots were immunoprobed in phosphate buffered saline (PBS), 0.1% TWEEN 20 (PBST) or TRIS-buffered saline, 0.1% TWEEN 20 (TBST) and detected by chemiluminescence, as described previously.³¹ Quantification of western blots was performed using the Image Lab software (Bio-Rad).

HTT bioassay protein lysate preparation

A 10% (w/v) total protein cortical homogenate was prepared in ice-cold bioassay buffer (PBS, 1% Triton-X-100) as previously described (n = 6/genotype).²⁸ Lysates were snap frozen and used for assays the following day.

HTRF assay

Cortical homogenates to a final volume of 10 μL were pipetted in triplicate into a 384-well (pure white, low volume, conical) proxiplate (Greiner Bio-One). Lysate dilutions and antibody concentrations are summarized in Supplementary Table 3. For HTRF assays, antibodies were added per well in 5 μL HTRF detection buffer [50 mM NaH₂PO₄, 0.2 M KF, 0.1% bovine serum albumin, 0.05% Tween-20] with complete protease inhibitor cocktail tablets (Roche). Plates were incubated for 3 h on an orbital shaker (250 rpm) at room temperature, before reading on an EnVision (Revvity) plate reader using optimized HTRF detection parameters as described previously.²⁸

MSD conjugation assay

Cortical homogenates to a final volume of 18 μL were pipetted in triplicate into 384-well, 1-spot custom printed MSD plates. MSD plates were run exactly according to the manufacturer’s recommended conditions. Lysate dilutions are summarized in Supplementary Table 4, and all incubations were performed at room temperature on a Jitterbug shaker (Boekel Scientific) at 1000 rpm. Wells were blocked by incubating overnight in 45 μL 3% bovine serum albumin in PBS. HTT proteins were then captured for 3 h at the default MSD anti-HTT capture concentration of ∼2 mg/mL ± 15% per well and detected for 1.5 h using 2.5 μg/mL anti-HTT Sulpho-tag, except for MW1 Sulpho-tag that was used at 1.5 μg/mL. Between each step, wells were washed three times in PBS-Tween buffer (PBS, 0.05% Tween 20). For detection, 45 μL MSD GOLD read buffer was added per well and read on an MSD plate reader (MSD).

RNA extraction and real-time quantitative PCR

Cortical tissue was homogenized in Qiazol lysis reagent in lysing matrix tubes D (MP Biomedicals) three times for 30 s using the Fastprep-24 (MP Biomedicals). Total RNA was extracted, DNase I treated and reverse transcribed with oligodT₁₈ (Invitrogen) for Htt1a and random hexamers (Invitrogen) for full-length Htt as previously described.²⁷

Primers and probes for the full-length Htt and Htt1a transcripts (Supplementary Table 5), and the custom Taqman assays for mouse reference genes (Atp5b, Sdha and Eif4a2) were from Thermo Fisher Scientific. Reactions were performed in triplicate. Each 12 μL reaction consisted of 3 μL of 1:10 diluted cDNA, TaqMan Fast Advanced Mix (Applied Biosystems) and primer probe mix distributed into 384-well thin-walled white PCR plates (Bio-Rad). Plates were sealed with Microseal ‘B’ seals (Bio-Rad), centrifuged at 800 × g for 30 s and amplified in a Bio-Rad CFX Opus 384 Real time PCR machine as follows: 40 s at 95°C and 40× (7 s at 95°C, 20 s at 60°C). Cq values deviating by 0.5 from the mean were excluded from further analysis. Data for genes of interest were normalized to the geometric mean of the reference genes, and the 2^−ΔCt2 was calculated.³²

Statistical analysis

Data were screened for outliers using Grubb’s test, and outliers were removed before between-group comparisons. All datasets were tested for a normal Gaussian distribution (Shapiro–Wilk test). Statistical analysis was performed using either Student’s t-test or one-way ANOVA with Bonferroni post hoc test. All statistical analyses were performed, and graphs were prepared using GraphPad Prism (v10).

Results

We set out to determine the effect of increasing CAG repeat length on levels of mutant HTT and to investigate how best to detect HTT protein isoforms by HTRF and MSD assays. All analyses were performed using fresh cortical lysates from homozygous, heterozygous and wild-type mice from HdhQ20, HdhQ50, HdhQ80, HdhQ111, CAG140 and zQ175 knock-in mouse colonies at 11 weeks of age [in total, two sets of n = 3 mice/gender/genotype for all western, quantitative real-time PCR (qPCR) and bioassay data]. All knock-in lines had been generated by replacing wild-type mouse exon 1 Htt with a mutant version of human exon 1 HTT.^22,23 Cortical lysates were also prepared from transgenic YAC128 mice and their wild-type littermates at 9 weeks of age (n = 6/genotype), to control for the detection of mutant HTT in addition to endogenous wild-type levels.

The epitope locations for the anti-HTT antibodies are illustrated in Fig. 1A. The D7F7 antibody was used to detect full-length HTT levels on western blots. For HTRF and MSD assays, pairwise antibody combinations were selected that would be predicted to detect ‘full-length mutant HTT’ (MW1 or 4C9 with MAB5490, MAB2166 or D7F7) and HTT1a (2B7 or MW1 with MW8), ‘total mutant HTT’ (HTT1a and full-length mutant HTT) (2B7, MW1 and 4C9) and ‘total full-length HTT’ (mutant and wild type; 2B7, MAB5490, MAB2166 and D7F7). Depending on the antibody location, the full-length HTT assays will also detect either all, or a subset of, proteolytic HTT fragments. For each antibody pair, the two antibodies were used in both donor and acceptor orientations for HTRF and in both capture and detection orientations for MSD.

Mutant full-length HTT levels decrease with increasing CAG repeat length

For western blot experiments, we began by preparing cortical lysates in three different detergent-containing lysis buffers: RIPA, NP40 and HEPES from the same wild-type and zQ175 mice (n = 2/genotype). Blots were immunoprobed for HTT with D7F7 in either PBST or TBST. The signals were marginally stronger with less background for the RIPA lysates and for blots immunoprobed in PBST (Supplementary Fig. 1).

Cortical lysates were prepared in RIPA buffer from wild-type, heterozygous and homozygous mice for each of the knock-in colonies, separated by 7.5% SDS-polyacrylamide acrylamide gel electrophoresis, and blots were probed with D7F7 for full-length HTT and with DM15 for HDAC4 as a loading control (Fig. 1B). Expanded polyQ tracts retard HTT migration, and consequently, mutant HTT could be distinguished from wild type for all lines except for HdhQ20 (Fig. 1B). Blots were quantified (Fig. 1C; Supplementary Fig. 2 and Supplementary Table 14) to compare total levels of HTT (wild-type and mutant HTT) in cortical lysates between wild-type, heterozygous and homozygous mice for each line (Fig. 1C; Supplementary Fig. 2A). The levels of mutant and wild-type HTT were equivalent in the HdhQ20 and HdhQ50 lines. In contrast, comparison of the wild-type and homozygous mutant signals indicated that the level of total HTT in the HdhQ80, HdhQ111, CAG140 and zQ175 cortical lysates had decreased with increasing CAG/polyQ length (Fig. 1C; Supplementary Fig. 2A). HTT signals for each knock-in line could be compared within a blot but not between blots. Therefore, to better demonstrate the reduction in mutant HTT among the HdhQ111, CAG140 and zQ175 lines, homozygote lysates were separated on the same gel (Supplementary Fig. 3 and Supplementary Table 15). The level of mutant HTT in the HdhQ111, CAG140 and zQ175 lines was ∼50, 23 and 10% of wild-type HTT, respectively (Supplementary Fig. 3C and D).

The HTT signal in YAC128 cortical lysates was ∼120% of wild-type HTT; mutant HTT levels were ∼40% of that arising from an endogenous wild-type allele (Fig. 1D; Supplementary Fig. 2A and C). Quantification of the wild-type HTT signals in heterozygous knock-in lysates confirmed that these were ∼50% of that in wild-type mice and the level of wild-type HTT in YAC128 lysates was at endogenous levels (Supplementary Fig. 2B). Quantification of the mutant HTT signals in homozygous knock-in lysates confirmed that these were approximately double that detected in the heterozygous lysates for all lines (Supplementary Fig. 2C).

The reduction in mutant HTT cannot be attributed to decreased Htt transcript levels

We prepared cortical cDNA from wild-type, heterozygous and homozygous mice from the knock-in lines and from YAC128 and wild-type littermates. qPCR indicated that full-length Htt levels in heterozygous and homozygous HdhQ20, HdhQ50, HdhQ80 and HdhQ111 cortices were comparable with wild type but decreased in CAG140 and zQ175 (Fig. 2A). This reduction was confirmed by comparing full-length Htt reads in the previously published RNA-seq data³³ from the allelic series of knock-in mice; mRNA levels being reduced by 14 and 20% of wild-type in heterozygous CAG140 and zQ175 cortex, respectively (Fig. 2B). In contrast, Htt1a levels increased with increasing CAG repeat length as measured by qPCR (Fig. 2C) and by extracting data from the previously published RNA-seq dataset (Fig. 2D).³³ The levels in homozygous mice were approximately double those in their heterozygous counterparts (Fig. 2C).

Figure 2.

Changes in Htt transcript levels do not account for the reduction in mutant HTT levels in the knock-in mouse lines. (A) qPCR for full-length Htt transcript levels in cDNA prepared from the cortex of wild-type, heterozygous and homozygous knock-in mice and from YAC128 and wild-type mice (n = 3/gender/genotype). Cortical full-length Htt mRNA levels were comparable among wild-type, heterozygous and homozygous mice for the HdhQ20, HdhQ50, HdhQ80 and HdhQ111 lines. There was a reduction in full-length Htt in the heterozygous and homozygous CAG140 and zQ175 mice. This qPCR assay amplifies mouse Htt and therefore does not detect the human HTT transgene in YAC128 mice but demonstrates that wild-type Htt levels were not altered. (B) Determination of comparative full-length Htt levels in wild-type and heterozygous HdhQ20, HdhQ80, HdhQ111, CAG140 and zQ175 cortex from RNA-Seq datasets.³³ There was no difference in the level of full-length Htt levels between wild-type and heterozygous HdhQ20, HdhQ80 and HdhQ111 mice. Htt levels were reduced by ∼14 and 20% of wild-type levels in the CAG140 and zQ175 heterozygotes, respectively. (C) qPCR for Htt1a transcript levels in cDNA prepared from the cortex of wild-type, heterozygous and homozygous knock-in mice (n = 3/gender/genotype). The level of HTT1a increased with increasing CAG repeat length and homozygous levels were approximately twice that in heterozygotes. (D) Determination of comparative HTT1a levels in wild-type and heterozygous HdhQ20, HdhQ80, HdhQ111, CAG140 and zQ175 cortex from RNA-Seq datasets.³³ Statistical analysis was Student’s t-test or one-way ANOVA with Bonferroni post hoc correction. The test statistic, degrees of freedom and P-values for the ANOVA are provided in Supplementary Table 7. Error bars are mean ± SEM. *P ≤ 0.5, **P ≤ 0.01, ***P ≤ 0.001. WT, wild type, HET, heterozygote, HOM, homozygote, HK, housekeeping gene, RNA-Seq, RNA sequence, TG, transgenic, TPM, transcripts per million.

Characterization of assays designed to detect ‘full-length mutant HTT’ (excluding the HTT1a protein)

The MW1 antibody recognizes an expanded polyQ tract, and, since it does not detect endogenous mouse HTT, can be used to develop assays selective for mutant HTT.^15,28 Pairings between MW1 and antibodies C-terminal to HTT1a (MAB5490, MAB2166 and D7F7) will therefore detect ‘full-length mutant HTT’, as well as some proteolytic fragments dependent on which of these three antibodies has been chosen. We performed HTRF with MW1 as the donor antibody (Fig. 3A) and MSD with MW1 as the capture antibody (Fig. 3B) paired with MAB5490, MAB2166 or D7F7, as well as the reciprocal assays of HTRF with MW1 as the acceptor antibody (Fig. 3C) and MSD with MW1 as the detector antibody (Fig. 3D) paired with MAB5490, MAB2166 or D7F7. In all cases, the assays were specific for mutant HTT (Fig. 3). Because MW1 binds to the polyQ tract, the assay signal might be expected to increase with increasing polyQ length. This occurred for all assays for polyQ sizes from Q20 to Q80, but from Q80 or Q111 onwards the signals continued to increase, levelled out or decreased (Fig. 3). In all cases, the signal was greater in homozygous when compared with heterozygous lysates. Therefore, the failure of the signal to continue to increase in heterozygotes with increasing polyQ length was not due to limiting antibody concentrations (Fig. 3). It was only in the HdhQ50 cortex that, for some assays, the signal in homozygotes was double that obtained for the heterozygotes. The pattern by which the signal changed in response to increasing polyQ length was relatively comparable between HTRF and MSD assays, except for MW1-MAB2166 and MAB5490-MW1. The different assays gave very variable and inconsistent signals for the level of ‘full-length mutant HTT’ in the YAC128 cortex when compared with the knock-in lines, with assays utilizing MAB5490 giving the greatest signal (Fig. 3). In summary, the non-linear relationship between the length of the polyQ repeat and MW1 antibody binding may reflect conformational changes in very long polyQ tracts. These assays cannot be used to compare ‘full-length mutant HTT’ levels in tissues or biofluids containing HTT with variable polyQ lengths.

Figure 3.

PolyQ-length dependence of HTRF and MSD assays that use MW1 and detect soluble full-length mutant HTT. Antibody pairings of MW1-MAB5490, MW1-MAB2166 and MW1-D7F7 were tested by HTRF (A) and MSD (B) and of MAB5490-MW1, MAB2166-MW1 and D7F7-MW1 were tested by HTRF (C) and MSD (D) using cortical lysates from wild-type, heterozygous and homozygous HdhQ20, HdhQ50, HdhQ80, HdhQ111, CAG140 and zQ175 mice at 11 weeks of age and YAC128 and wild-type littermates at 9 weeks of age (n = 3/gender/genotype). The x-axis legends are the same as in Fig. 2A. Statistical analysis was Student’s t-test (for YAC128) or one-way ANOVA with Bonferroni post hoc correction per mouse line. The test statistic, degrees of freedom and P values for the ANOVA are provided in Supplementary Table 8. Error bars are mean ± SEM. *P ≤ 0.5, **P ≤ 0.01, ***P ≤ 0.001. WT, wild type, HET, heterozygote, HOM, homozygote, TG, transgenic.

The 4C9 antibody is specific for the human polyproline-rich region in HTT exon 1 and can also be used for mutant HTT specific assays as the knock-in lines have a humanized exon 1.^16,28 To develop ‘full-length mutant HTT’ assays, we performed HTRF with 4C9 as the donor antibody (Fig. 4A) and MSD with 4C9 as the capture antibody (Fig. 4B) paired with MAB5490, MAB2166 or D7F7 as well as the reciprocal assays of HTRF with 4C9 as the acceptor antibody (Fig. 4C) and MSD with 4C9 as the detector antibody (Fig. 4D) paired with MAB5490, MAB2166 or D7F7. In all cases, the assays were specific for mutant HTT (Fig. 4). The pattern by which the signal changed in response to increasing polyQ length was relatively comparable between HTRF and MSD assays for any antibody pairing. The heterozygous and homozygous signals decreased with increasing polyQ tract length for all assays, albeit to differing extents. The decrease in the heterozygote signal could not be due to limiting antibody concentrations as a greater signal was obtained for the homozygotes, and for many of the assays, the signal in the homozygote cortices was approximately double that in the heterozygotes (Fig. 4). Western blot analysis had indicated that full-length mutant HTT levels were the same as wild type in HdhQ20 and HdhQ50 cortex but decreased with increasing CAG repeat length from HdhQ80 onwards (Fig. 1C). This change in mutant HTT levels was not precisely reflected by any of these assays. In some cases, the level of mutant HTT increased from HdhQ20 to HdhQ50, and in many cases, there was no difference between HdhQ50 and HdhQ80. Once again, different assays gave very variable signals for the level of full-length mutant HTT in the YAC128 cortex when compared with the knock-in lines, with assays utilizing MAB5490 giving the greatest signal (Fig. 4). The length of the polyQ repeat may influence the binding of the 4C9 antibody, because of the proximity of polyQ repeat and the 4C9 epitope.

Figure 4.

PolyQ-length dependence of HTRF and MSD assays that use 4C9 and detect soluble full-length mutant HTT. Antibody pairings of 4C9-MAB5490, 4C9-MAB2166 and 4C9-D7F7 were tested by HTRF (A) and MSD (B) and of MAB5490-4C9, MAB2166-4C9 and D7F7-4C9 were tested by HTRF (C) and MSD (D) using cortical lysates from wild-type, heterozygous and homozygous HdhQ20, HdhQ50, HdhQ80, HdhQ111, CAG140 and zQ175 mice at 11 weeks of age and YAC128 and wild-type littermates at 9 weeks of age (n = 3/gender/genotype). The x-axis legends are the same as in Fig. 2A. Statistical analysis was Student’s t-test (for YAC128) or one-way ANOVA with Bonferroni post hoc correction per mouse line. The test statistic, degrees of freedom and P-values for the ANOVA are provided in Supplementary Table 9. Error bars are mean ± SEM. *P ≤ 0.5, **P ≤ 0.01, ***P ≤ 0.001. WT, wild type, HET, heterozygote, HOM, homozygote, TG, transgenic.

Assays that detect the HTT1a protein (soluble or aggregated)

We have previously shown that MW8 is a neo-epitope antibody for the C-terminus of HTT1a under western blotting conditions³⁰ and can be used to generate HTT1a-specific HTRF, MSD and AlphaLISA assays.²⁸ To compare the effect of polyQ length on the detection of soluble HTT1a, we performed HTRF with 2B7 as the donor antibody (Fig. 5A) and MSD with 2B7 as the capture antibody (Fig. 5B) paired with MW8, as well as HTRF with MW1 as the donor antibody (Fig. 5C) and MSD with MW1 as the capture antibody (Fig. 5D) paired with MW8. For HTRF, the 2B7-MW8 assay (Fig. 5A) was much more sensitive than MW1-MW8 (Fig. 5C). For MSD, the 2B7-MW8 assay was very weak and not viable (Fig. 5B), with MW1-MW8 providing a better alternative (Fig. 5D). HTT1a levels increase with increasing CAG repeat (Fig. 5A–D) as would be predicted by the increasing levels of the Htt1a transcript (Fig. 2C) from which it is translated. The level of soluble HTT1a in the YAC128 cortex was comparable with that in an HdhQ80 heterozygous mouse (Fig. 5A–D).

Figure 5.

PolyQ-length dependence of HTRF and MSD assays that detect soluble and aggregated HTT1a. For soluble HTT1a, 2B7-MW8 was tested by HTRF (A) and MSD (B), and MW1-MW8 was tested by HTRF (C) and MSD (D) using cortical lysates from wild-type, heterozygous and homozygous HdhQ20, HdhQ50, HdhQ80, HdhQ111, CAG140 and zQ175 mice at 11 weeks of age and YAC128 and wild-type littermates at 9 weeks of age (n = 3/gender/genotype). For aggregated HTT1a, 4C9-MW8 was tested by HTRF (E) and MSD (F) using cortical lysates from wild-type, heterozygous and homozygous HdhQ20, HdhQ50, HdhQ80, HdhQ111, CAG140 and zQ175 mice at 11 weeks of age, wild-type and zQ175 heterozygous mice at 26 weeks of age and wild-type and YAC128 mice at 9 and 26 weeks of age (n = 3/gender/genotype). The x-axis legends are the same as in Fig. 2A. Statistical analysis was Student’s t-test (for YAC128) or one-way ANOVA with Bonferroni post hoc correction per mouse line. The test statistic, degrees of freedom and P-values for the ANOVA are provided in Supplementary Table 10. Error bars are mean ± SEM. *P ≤ 0.5, **P ≤ 0.01, ***P ≤ 0.001. WT, wild type, HET, heterozygote, HOM, homozygote, TG, transgenic.

We have previously shown that the 4C9-MW8 HTRF assay is specific for aggregated HTT1a.³⁴ To investigate the comparable level of aggregated HTT1a, we ran the 4C9-MW8 HTRF and MSD assays on lysates from the same cortices used for the soluble HTT1a assays (Fig. 5E and F). The HTRF assay was more sensitive than the MSD assay. At 11 weeks of age, statistically significant levels of HTT1a aggregation could be detected in heterozygous HdhQ80 mice and homozygous HdhQ50 mice by HTRF, with the level in homozygous mice increasing more rapidly with polyQ length than that in heterozygous mice (Fig. 5E and F). Therefore, we would expect that the aggregation of HTT1a to have influenced the level of soluble HTT1a detected in Fig. 5A–D and may explain why the soluble HTT1a levels in homozygotes lysates were relatively comparable with those in heterozygotes lysates (Fig. 5A–D). The level of aggregation was similar in YAC128 cortex at 26 weeks of age to the cortex of heterozygous zQ175 mice at 11 weeks (Fig. 5E and F).

Characterization of assays designed to detect ‘total mutant HTT’ (full-length HTT and HTT1a)

Total mutant HTT, comprising HTT1a, full-length mutant HTT and all proteolytic fragments can be detected by pairing two antibodies that both recognize epitopes within exon 1 of HTT. To investigate the effect of polyQ length on these assays, we performed HTRF with 2B7 as the donor antibody (Fig. 6A) and MSD with 2B7 as the capture antibody (Fig. 6B) paired with MW1 or 4C9; HTRF with MW1 as the donor antibody (Fig. 6C) and MSD with MW1 as the capture antibody (Fig. 6D) paired with 2B7 or 4C9 and HTRF with 4C9 as the donor antibody (Fig. 6E) and MSD with 4C9 as the capture antibody (Fig. 6F) paired with 2B7 or MW1. In all cases, the assays were specific for mutant HTT (Fig. 6). The signal was predominantly greater in cortical lysates from homozygous when compared with heterozygous mice, indicating that the reductions in the heterozygous signals with CAG repeat length could not have resulted from limiting antibody concentrations. To ensure that this was also the case for the MW1-2B7 MSD assay in the HdhQ111, CAG140 and zQ175 lysates (Fig. 6D), this assay was run through a series of dilutions for zQ175 lysates, and the homozygous signals were found to be greater than those for the heterozygotes (Supplementary Fig. 11 and Supplementary Table 23).

Figure 6.

PolyQ-length dependence of HTRF and MSD assays that detect total soluble mutant HTT (full-length HTT and HTT1a). Antibody pairings of 2B7-MW1 and 2B7-4C9 were tested by HTRF (A) and MSD (B), of MW1-2B7 and MW1-4C9 was tested by HTRF (C) and MSD (D) and of 4C9-2B7 and 4C9-MW1 tested by HTRF (E) and MSD (F) using cortical lysates from wild-type, heterozygous and homozygous HdhQ20, HdhQ50, HdhQ80, HdhQ111, CAG140 and zQ175 mice at 11 weeks of age and YAC128 and wild-type littermates at 9 weeks of age (n = 3/gender/genotype). The x-axis legends are the same as in Fig. 2A. Statistical analysis was Student’s t-test (for YAC128) or one-way ANOVA with Bonferroni post hoc correction per mouse line. The test statistic, degrees of freedom and P-values for the ANOVA are provided in Supplementary Table 11. Error bars are mean ± SEM. *P ≤ 0.5, **P ≤ 0.01, ***P ≤ 0.001. WT, wild type, HET, heterozygote, HOM, homozygote, TG, transgenic.

Comparison of the relative signals for these ‘total mutant HTT’ assays with the patterns obtained for ‘full-length mutant HTT’ (Figs 3 and 4) and for HTT1a (Fig. 5) indicated that specific ‘total mutant HTT’ assays exhibited a preference for one of these isoforms. The most unexpected result was for the frequently used MSD assay, 2B7-MW1, which produced a pattern that tracked with HTT1a. The four HTRF and three other MSD assays utilizing MW1 described a pattern reminiscent of the ‘full-length mutant HTT’ assays (Fig. 3). The 4C9-2B7 HTRF and MSD assays also appear to track ‘full-length mutant HTT’, whereas the distinctive patterns of the 2B7-4C9 MSD and HTRF assays suggested that they are detecting both ‘full-length mutant HTT’ and HTT1a.

Characterization of assays designed to detect ‘total full-length HTT’ (mutant and wild type)

Total full-length HTT (mutant and wild-type) can be detected by pairing 2B7 with antibodies that are located C-terminal to exon 1 HTT. We performed HTRF with 2B7 as the donor antibody (Fig. 7A) and MSD with 2B7 as the capture antibody (Fig. 7B) paired with MAB5490, MAB2166 or D7F7, as well as the reciprocal assays of HTRF with 2B7 as the acceptor antibody (Fig. 7C) and 2B7 as the detector antibody (Fig. 7D) paired with MAB5490, MAB2166 or D7F7. All antibody pairings showed much more inter-sample variability in the MSD data than the HTRF data (Fig. 7). For the HTRF assays, the wild-type signals for endogenous HTT were relatively comparable between mouse lines (Fig. 7).

Figure 7.

PolyQ-length dependence of HTRF and MSD assays that use 2B7 and detect total full-length HTT (mutant and wild-type). Antibody pairings of 2B7-MAB5490, 2B7-MAB2166 and 2B7-D7F7 were tested by HTRF (A) and MSD (B) and of MAB5490-2B7, MAB2166-2B7 and D7F7-2B7 were tested by HTRF (C) and MSD (D) using cortical lysates from wild-type, heterozygous and homozygous HdhQ20, HdhQ50, HdhQ80, HdhQ111, CAG140 and zQ175 mice at 11 weeks of age and YAC128 and wild-type littermates at 9 weeks of age (n = 3/gender/genotype). The x-axis legends are the same as in Fig. 2A. Statistical analysis was Student’s t-test (for YAC128) or one-way ANOVA with Bonferroni post hoc correction per mouse line. The test statistic, degrees of freedom and P-values for the ANOVA are provided in Supplementary Table 12. Error bars are mean ± SEM. *P ≤ 0.5, **P ≤ 0.01, ***P ≤ 0.001. WT, wild type, HET, heterozygote, HOM, homozygote, TG, transgenic.

In general, the heterozygous and homozygous signals decreased with polyQ length, with the homozygote signals lower than the heterozygote, consistent with western blot data (Fig. 1). However, all HTRF assays indicated that the heterozygous and homozygous HdhQ50 lysates contained less full-length HTT than wild type, which was inconsistent. Also, none of the HTRF assays registered that the YAC128 lysate contained more full-length HTT than wild type and the three HTRF assays with 2B7 as donor, measured less HTT in YAC128 lysate than in wild type, which cannot be the case. An increase in total HTT in the YAC128 lysates was only apparent in the signals generated by the 2B7-MAB5490, MAB5490-2B7 and D7F7-2B7 MSD assays, but these assays also detected more full-length HTT in heterozygous and/or homozygous Hdh20 lysates when compared with wild type, which is again inconsistent with the western blot data. To investigate this further, the HTRF assays were repeated with 5 and 2.5% lysates. This demonstrated that the relative signals within and between lines were sensitive to lysate concentration (Supplementary Fig. 12 and Supplementary Table 24), in a manner that is difficult to interpret.

Total full-length HTT can also be detected by pairing two antibodies that both recognize epitopes located C-terminal to exon 1 HTT. We performed HTRF with MAB5490 as the donor antibody (Fig. 8A) and MSD with MAB5490 as the capture antibody (Fig. 8B) paired with MAB2166 or D7F7; HTRF with MAB2166 as the donor antibody (Fig. 8C) and MSD with MAB2166 as the capture antibody (Fig. 8D) paired with MAB5490 or D7F7 and HTRF with D7F7 as the donor antibody (Fig. 8E) and MSD with D7F7 as the capture antibody (Fig. 8F) paired with MAB5490 or MAB2166. The changing pattern of signals with respect to polyQ length was relatively consistent between the HTRF and MSD assays for any given antibody pair, but the inter-sample variability was greater for MSD (Fig. 8). None of the assays recapitulated the decrease in mutant HTT levels as detected by western blot; in general, the assays indicated that there was more full-length HTT in heterozygous and/or homozygous HdhQ50 lysates than wild-type, that the levels in the HdhQ80 and HdhQ111 heterozygotes and homozygotes were comparable with wild-type and that decreases in mutant HTT were only apparent in heterozygous and homozygous CAG140 and zQ175 lysates (Fig. 8). However, in all cases, on both platforms, more full-length HTT was detected in the YAC128 lysate than in wild type (Fig. 8). Biologically, we do not understand how small changes in the length of the polyQ tract, e.g. from 20 to 50 glutamines, might result in similar changes in the performance of these six HTRF FRET–based and six MSD enzyme-linked immunosorbent (ELISA)-based assays.

Figure 8.

PolyQ-length dependence of HTRF and MSD assays that use pairs of antibodies C-terminal to exon 1 HTT and detect total full-length HTT (mutant and wild-type). Antibody pairings of MAB5490-MAB2166, MAB5490-D7F7 were tested by HTRF (A) and MSD (B), of MAB2166-MAB5490 and MAB2166-D7F7 were tested by HTRF (C) and MSD (D) and of D7F7-MAB5490 and D7F7-MAB2166 were tested by HTRF (E) and MSD (F) using cortical lysates from wild-type, heterozygous and homozygous HdhQ20, HdhQ50, HdhQ80, HdhQ111, CAG140 and zQ175 mice at 11 weeks of age and YAC128 and wild-type littermates at 9 weeks of age (n = 3/gender/genotype). The x-axis legends are the same as in Fig. 2A. Statistical analysis was Student’s t-test (for YAC128) or one-way ANOVA with Bonferroni post hoc correction per mouse line. The test statistic, degrees of freedom and P-values for the ANOVA are provided in Supplementary Table 13. Error bars are mean ± SEM. *P ≤ 0.5, **P ≤ 0.01, ***P ≤ 0.001. WT, wild type, HET, heterozygote, HOM, homozygote, TG, transgenic.

Discussion

Somatic CAG repeat expansion can lead to neurons containing mutant HTT with expansions of hundreds of CAGs in the brains of mutation carriers.^6,7 The allelic series of Huntington’s disease knock-in mouse models, with repeat lengths ranging from (CAG)₂₀ to ∼(CAG)₁₉₀, provide a means by which the molecular and cellular consequences of somatic CAG repeat expansion can be modelled. Here, we have shown that as CAG repeat length increased, the level of cortical full-length mutant HTT protein decreased. Mutant HTT levels were equivalent to those of wild-type HTT in the cortex of mice with (CAG)₅₀, decreasing to ∼10% of wild-type HTT levels with expansions of (CAG)₁₉₀. In contrast, as the CAG repeat length increased, the production of the Htt1a transcript and HTT1a protein increased.

We used western blotting to quantify the levels of full-length mutant HTT in cortical tissues from knock-in mice with CAG repeat lengths of approximately 20, 50, 80, 111, 140 and 190 CAGs. We selected the stringent RIPA buffer for lysate preparation, to solubilize membrane bound and hard-to-solubilize proteins. Analysis of heterozygous mice showed that the Huntington’s disease mutation had no effect on wild-type HTT levels. However, while the level of mutant HTT in heterozygous mice was equivalent to wild type for HdhQ20 and HdhQ50 mice, this had decreased in the HdhQ80, HdhQ111, CAG140 and zQ175 lines. Cortical mutant HTT levels in homozygous knock-in mice were 50% of wild-type levels in HdhQ111, 23% in CAG140 and only 10% in zQ175 mice. The low HTT levels in homozygous zQ175 mice are consistent with the recovery of 13% homozygous pups (instead of 25%) from breeding heterozygous male with heterozygous female zQ175 mice (B. Callahan, personal communication). They are also consistent with data showing that mutant HTT levels are lower than wild type in juvenile Huntington’s disease brains.³⁵

This decrease in full-length mutant HTT levels has most likely occurred through a combination of mechanisms. First, our data indicate that the mutant Htt transcript was only reduced in the CAG140 and zQ175 cortices, and so this could not have accounted for the much greater reduction in mutant HTT protein observed in the HdhQ80, HdhQ111, CAG140 and zQ175 lines. Second, the translation of the mutant transcript could have been impaired; the translation of transcripts with long CAG repeats has been shown to be decreased due to ribosome stalling and/or collisions.³⁶ Third, if soluble mutant HTT had been recruited into HTT aggregates, it would not have been solubilized by RIPA buffer. However, we previously demonstrated that the HTT1a protein is recruited into aggregates in zQ175 brain regions and not full-length HTT.³⁴ Fourth, the full-length mutant protein could have undergone proteolysis to generate smaller HTT fragments not detected by D7F7. We have previously assessed the proteolytic fragments generated from mutant HTT in the HdhQ150 knock-in model, but these were not abundant and could not account for such dramatic reductions in mutant HTT levels.³⁰

It is essential that soluble and aggregated HTT isoforms can be detected in tissues and biofluids. In our previous study, we had interpreted the detection, by some assays, of much lower levels of mutant HTT in zQ175 striatum and cortex, when compared with HdhQ20, to be due to the expanded polyQ interfering with assay performance.²⁸ However, our western blot data suggest that this interpretation was not entirely correct.

In this study, we set out to determine which HTRF and MSD assays might best reflect the changes in mutant HTT levels with increasing CAG repeat length. Cortical tissues from two sets of age-matched knock-in mice from the allelic series were used for all the western blots, HTRF and MSD assays. As mutant HTT can aggregate in tissue lysates, even when stored at −80°C, lysates were always frozen after preparation and used the following day. HTRF and MSD assays were run according to the manufacturer’s recommended conditions. If the results suggested that antibody levels might be limiting, the assay was repeated with a series of lysate dilutions. For a given antibody pair, with a few exceptions, the overall pattern of signals across the allelic series was relatively comparable between HTRF and MSD, although there was often greater sample heterogeneity for MSD.

The validity of the HTRF and MSD assays at detecting full-length HTT was assessed by comparing their performance with our western blot data (Fig. 1), for which the protein had been denatured and all antibody-binding sites should have been accessible. We used two sets of assays to detect full-length mutant HTT, employing either the MW1 (Fig. 3) or 4C9 (Fig. 4) antibodies to impart mutant HTT specificity. We would avoid using MW1-based assays because of the non-linear relationship between polyQ length and signal intensity (Fig. 3). The 4C9-based assays better represented the changes in mutant HTT with increasing CAG repeat length; however, none of the assays replicated the western blot data precisely (Fig. 4). The choice of assay must be informed by the CAG repeat lengths in the samples that are to be compared. If these are comparable, the choice of assay is less critical. However, if they are known to vary, it is important to know how the assay performance is influenced by CAG repeat length, and the results interpreted with caution.

For the detection of mutant HTT, we would recommend using assays that are selective for either ‘mutant full-length HTT’ or HTT1a and avoid those that detect both isoforms. Comparison of the relative signal levels for the ‘total mutant HTT’ assays (Fig. 6) with the patterns obtained for ‘full-length mutant HTT’ (Figs 3 and 4) and for HTT1a (Fig. 5) indicated that specific ‘total mutant HTT’ assays preferentially detected one of these isoforms. The most unexpected result was for the frequently used MSD assay, 2B7-MW1, which produced a pattern that tracked with HTT1a; further experiments are needed to confirm this result.

We used two sets of assays to detect ‘total full-length HTT’ levels (wild type and mutant) (Figs 7 and 8). In summary, the HTRF assays that used the 2B7 antibody were the most promising (Fig. 7), although none of the 12 HTRF or 12 MSD total full-length HTT assays recapitulate the changes in total HTT with increasing polyQ length as measured by western blot (Fig. 1). Instead, we would recommend using species-specific assays, i.e. a ‘soluble mutant full-length HTT’ assay as discussed above and an ‘endogenous mouse HTT’ assay that we have previously published.³⁴

Establishing a set of antibody-based bioassays to detect soluble and aggregated HTT isoforms is complex. These assays are not quantitative, as in the absence of recombinant protein standards with polyQ lengths to match those in the tissue analyte, it is not possible to run standard curves. In order to apply them to any given mouse model or cell line, antibody concentrations and lysate dilutions must be optimized to ensure that an assay is operating in the linear part of the curve.²⁸ If performance of an assay can be influenced by polyQ length, it is not possible to compare HTT levels in samples containing HTT with different polyQ repeats. Interpreting the consequences of interventions that might alter polyQ length (e.g. targeting somatic CAG repeat expansion) on HTT levels must be treated with caution as a signal may change because the length of the polyQ repeat has been altered, not because the level of HTT has changed.

This work has significant implications for the development of assays to measure HTT proteins in human biofluids. HTRF and MSD assays are not sufficiently sensitive for use in human plasma or CSF, and instead, single molecule counting assays have been established for that purpose.²¹ Assays that distinguish the mutant HTT isoforms: full-length HTT and HTT1a, as well as assays that distinguish full-length mutant and wild-type HTT are ideally required. In knock-in mice, mutant and wild-type HTT can be distinguished without using antibodies to the polyQ tract, but this is not possible for human samples, and so options are more constrained. Therefore, whether a given assay preferentially selects a specific mutant HTT isoform and/or whether the assay performance is influenced by polyQ length should be determined. Both questions remain unanswered for the 2B7-MW1 single molecule counting assay that has been frequently used to measure total mutant HTT in CSF^21,37 and could be addressed by applying existing and future assays to diluted tissue lysates from the allelic series of knock-in mice. This is important, because, due to somatic CAG repeat expansion, the mutant HTT protein in CSF may contain polyQ repeats that are heterogeneous in length and may contain up to several hundred glutamine residues.

Our data demonstrate that as the CAG repeat increases the level of full-length mutant HTT decreases while that of the Htt1a transcript and the HTT1a protein increase. We have previously shown that the HTT1a is recruited into aggregates³⁴ and that in the striatum of zQ175 mice, further expansion of the CAG repeat is not required for the initiation of HTT aggregation and transcriptional dysregulation.³⁸ In heterozygous zQ175 mice, with ∼190 CAGs, cortical full-length mutant HTT is at ∼10% of wild-type levels, and the extent to which it contributes to the pathogenic process is not known. Single cell analyses from post-mortem Huntington’s disease brains have recently shown that somatic expansion occurs in the medium spiny neurons of the striatum⁷ and that expansion of CAG repeats beyond (CAG)₁₅₀ initiates a cell-autonomous toxicity that is executed through the further expansion of the repeat.³⁹ Our data predict that levels of full-length mutant HTT in these cells would be very low. If this was the case, therapeutic strategies to lower HTT levels by targeting sequences that are 3ʹ to exon 1 HTT will predominantly decrease wild-type HTT in cells with pathogenic expansions; allele-selective strategies will reduce further what are already very low levels of mutant HTT. Both approaches will leave the pathogenic HTT1a protein levels unchanged.

Supplementary material

Supplementary material is available at Brain Communications online.

Funding

This work was supported by grants from the CHDI Foundation and UK Dementia Research Institute, which receives its funding from Dementia Research Institute Ltd, funded by the UK Medical Research Council and Alzheimer’s Society and Alzheimer’s Research UK.

Competing interests

J.R.G. and K.B. are employees of Rancho BioSciences. The authors report no other competing interests.

Data availability

The authors confirm that all the data supporting the findings of this study are available within the article and in its Supplementary material. Raw data will be shared by the corresponding author on request.