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 3' 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.

Keywords: HTT1a protein; Huntington’s disease; huntingtin bioassay; knock-in mouse models of Huntington’s disease; mutant huntingtin protein.

Graphical Abstract

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.

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. 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. 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.

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.

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.

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.

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.

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.

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.