Epigenetic promoter silencing in Friedreich ataxia is dependent on repeat length

Abstract

Objective: Friedreich ataxia (FRDA) is caused by an expanded GAA triplet-repeat (GAA-TR) mutation in the FXN gene. Patients are typically homozygous for expanded alleles containing 100 to 1,300 triplets, and phenotypic severity is significantly correlated with the length of the shorter of the 2 expanded alleles. Patients have a severe deficiency of FXN transcript, which is predominantly caused by epigenetic silencing of the FXN promoter. We sought to determine whether the severity of FXN promoter silencing is related to the length of the expanded GAA-TR mutation in FRDA.

Methods: Patient-derived lymphoblastoid cell lines bearing a range of expanded alleles (200-1,122 triplets) were evaluated for FXN transcript levels by quantitative reverse transcriptase polymerase chain reaction. FXN promoter function was directly measured by quantitative analysis of transcriptional initiation via metabolic labeling of newly synthesized transcripts in living cells.

Results: FXN transcriptional deficiency was significantly correlated with the length of the shorter of the 2 expanded alleles, which was noted both upstream (R(2) = 0.84, p = 0.014) and downstream (R(2) = 0.89, p = 0.002) of the expanded GAA-TR mutation, suggesting that FXN promoter silencing in FRDA is related to repeat length. A bilinear regression model revealed that length dependence was strongest when the shorter of the 2 expanded alleles contained <400 triplets. Direct measurement of FXN promoter activity in patients with expanded alleles containing <400 versus >400 triplets in the shorter of the 2 expanded alleles revealed a significantly greater deficiency in individuals with longer GAA-TR alleles (p < 0.05).

Interpretation: FXN promoter silencing in FRDA is dependent on the length of the expanded GAA-TR mutation.

Figure 1. FXN transcriptional deficiency in FRDA extends upstream and downstream of the expanded GAA-TR mutation

(A) Relevant portions of the FXN gene are depicted schematically, with the GAA-TR mutation in intron 1 and the FXN transcriptional start site (TSS) at −59 (position numbers are relative to the initiation codon, indicated as +1). FXN transcript was quantitatively measured both upstream (Ex1; in the vicinity of the TSS) and downstream (Ex3-Ex4) of the GAA-TR mutation. Solid lines depict the full-length FXN transcript and shorter predicted transcripts caused by defects in transcriptional elongation through the expanded GAA-TR mutation and by deficient transcriptional initiation due to FXN promoter silencing. Deficiency of transcript at both locations would suggest a defect in transcriptional initiation, and deficiency of only Ex3-Ex4 would suggest a defect in transcriptional elongation. (B) Quantitative RT-PCR showing significantly reduced amounts of FXN mRNA both upstream (Ex1) and downstream (Ex3-Ex4) of the expanded GAA-TR mutation in FRDA versus non-FRDA (CNTR) lymphoblastoid cells. Heterozygous carriers of the GAA-TR mutation (CARRIER) show transcript levels that are intermediate between FRDA and non-FRDA controls. Graphs represent the cumulative data from two complete experiments using three FRDA, three heterozygous carriers, and three non-FRDA (CNTR) lymphoblastoid cell lines, each assayed in triplicate. Error bars represent +/− SEM. ** = p<0.01, *** = p<0.001.

Figure 2. Repeat length-dependent FXN transcriptional deficiency in FRDA is modulated by the shorter of the two expanded GAA-TR alleles containing <400 triplets

Relative FXN transcript levels (Y axis) of FRDA patients (n = 16), homozygous for a wide range of expanded GAA-TR alleles (200 – 1122 triplets), plotted against the length of the shorter of the two expanded alleles (X axis). (A) A piecewise bilinear regression model showed strong correlation (R² = 0.84) between FXN transcript levels measured at Ex1 and the length of the shorter GAA-TR allele; highly significant correlation (p = 0.014) was seen with alleles shorter than the estimated breakpoint of 423 triplets (95% CI, 353 – 494 triplets). (B) A piecewise bilinear regression model showed strong correlation (R² = 0.89) between FXN transcript levels measured at Ex3-Ex4 and the length of the shorter GAA-TR allele; highly significant correlation (p = 0.002) was seen with alleles shorter than the estimated breakpoint of 405 triplets (95% CI, 346 – 464 triplets). Note: open circles depict the six cell lines that were used for assessing repeat length-dependent FXN promoter silencing in Fig. 3B.

Figure 3. FXN promoter silencing in FRDA is dependent on the length of the expanded GAA-TR mutation

(A) Quantitative RT-PCR of metabolically labeled nascent FXN transcript for the indicated incubation times (2h & 4h) is shown at Ex1, i.e., in the vicinity of the transcription start site, for FRDA patients (n = 3) and non-FRDA controls (CNTR; n = 3). FRDA cells showed 15–19 fold less nascent FXN transcript compared with non-FRDA cells at both time points assayed. The graph represents cumulative data from two complete experiments with each cell line assayed in triplicate. Error bars represent +/− SEM. *** = P<0.001. (B) Quantitative RT-PCR of metabolically labeled FXN nascent transcript for the indicated incubation times (2h & 4h) is shown at Ex1, for FRDA patients with the shorter of the two expanded GAA-TR alleles containing <400 triplets (n = 3) and >400 triplets (n = 3). FRDA cells with shorter GAA-TR alleles showed 2-fold more nascent FXN transcript compared with FRDA cells with longer GAA-TR alleles at both time points assayed. The graph represents cumulative data from two complete experiments with each cell line assayed in triplicate. Error bars represent +/− SEM. * = p<0.05, ** = p<0.01.