Retention of hexanucleotide repeat-containing intron in C9orf72 mRNA: implications for the pathogenesis of ALS/FTD

Abstract

Introduction: The most common forms of amyotrophic lateral sclerosis and frontotemporal dementia are caused by a large GGGGCC repeat expansion in the first intron of the C9orf72 gene. The repeat-containing intron should be degraded after being spliced out, however GGGGCC repeat-containing RNA species either accumulate in nuclear foci or are exported to the cytoplasm where they are translated into potentially toxic dipeptide repeat proteins by repeat-associated non-AUG-initiated (RAN) translation.

Results: In order to determine the mechanisms of repeat-containing intron misprocessing, we have analyzed C9orf72 transcripts in lymphoblasts from C9orf72 expansion carriers (n = 15) and control individuals (n = 15). We have identified polyadenylated C9orf72 RNA species retaining the repeat-containing intron and in which downstream exons are spliced correctly resulting in a C9orf72 mRNA with an enlarged 5'-UTR containing the GGGGCC repeats. Intron-retaining transcripts are produced from both wild-type and mutant alleles. Intron-retaining C9orf72 transcripts were also detected in brain with a 2.7 fold increase measured in the frontal cortex from heterozygous expansion carriers (n = 11) compared to controls (n = 10). The level of intron-retaining transcripts was increased 5.9 fold in a case homozygous for the expansion. We also show that a large proportion of intron 1-retaining C9orf72 transcripts accumulate in the nucleus.

Conclusions: Retention of the repeat-containing intron in mature C9orf72 mRNA can potentially explain nuclear foci formation as well as nuclear export of GGGGCC repeat RNA and suggests that the misprocessing of C9orf72 transcripts initiates the pathogenic process caused by C9orf72 hexanucleotide repeat expansions as well as provides the basis for novel therapeutic strategies.

Fig. 1

Intron 1 retention in C9orf72 transcripts in lymphoblasts. C9orf72 transcripts were analyzed in lymphoblasts from C9orf72 G₄C₂ expansion carriers and control individuals. a RT-PCR analysis of poly(A)⁺ RNA using primers spanning the 5’ splice site (left) or the 3’ splice site (right) of intron 1 demonstrating retention of intron 1in polyadenylated RNA. The position of the primers is indicated on the diagram above the gels. GAPDH was used as a loading control. b Correctly spliced transcripts detected in controls and expansion carrier cells using primers in exons 1a and 2. c Quantitative analysis of intron retention by real-time PCR. Levels of C9orf72 transcripts spliced or unspliced at the 5’ and 3’ end of intron 1 were determined by real-time qRT-PCR. Data are shown as means ± SEM. Each data point represents an individual case, n=15 (C9^-); 15 (C9⁺). No significant differences were observed between the C9⁻ and C9⁺ groups. d Sequencing of the 3’ PCR product demonstrates an exact intron 1-exon 2 boundary. C9⁻, controls; C9⁺, expansion carriers

Fig. 2

a Intron 1-retaining transcripts contain the G₄C₂ repeat sequence. PCR using primers spanning the G₄C₂ repeat region was performed on cDNA reversed transcribed from polyadenylated C9orf72 RNA from control and expansion carrier lymphoblasts. The two products of slightly different sizes detected in cells from normal individuals correspond to the two alleles; the single product generated from expansion carrier cells corresponds to the wild-type allele. b Intron 1-retaining transcripts are produced from both wild-type and mutant alleles. Sequence traces of C9orf72 overlapping the intron 1-exon 2 boundary and the rs10757668 SNP in cDNA (top panels) and genomic DNA (gDNA, bottom panels) prepared from lymphoblasts from two control individuals and two expansion carriers heterozygous for rs10757668. One of the control cases (top trace) is heterozygous for rs10757668. The sequence of the reverse strand is shown; the position of rs10757668 is indicated by the grey box. Intron 1-retaining transcripts contain both the C and T alleles showing that intron 1 retention occurs in RNA transcribed from the wild-type as well as the expanded allele. C9⁻, controls; C9⁺, expansion carriers

Fig. 3

Nuclear accumulation of intron 1-retaining C9orf72 transcripts. Partitioning of intron 1-retaining and correctly spliced C9orf72 polyadenylated transcripts was compared between nuclear and cytoplasmic fractions from lymphoblasts from G₄C₂ repeat expansion carriers and controls. a Poly(A)⁺ RNA was analyzed in nuclear and cytoplasmic fractions by RT-PCR as in Fig. 1. The nuclear long non-coding RNA, NEAT1, was used to confirm the purity of the fractions. GAPDH was used as a loading control. b Levels of nuclear and cytoplasmic unspliced and spliced C9orf72 transcripts were compared according to cell equivalent and the ratios nuclear/(nuclear + cytoplasmic) were calculated for each product. Data represent mean ± SEM, n = 6 (C9⁺); 7 (C9⁻), ** P < 0.01, *** P < 0.001. C9⁻, controls; C9⁺, expansion carriers

Fig. 4

Intron 1 retention in C9orf72 transcripts in the brain of C9orf72 expansion carriers. C9orf72 transcripts were analyzed in the frontal cortex from frontotemporal lobar degeneration (FTLD) or frontotemporal lobar degeneration with motor neuron disease (FTLD-MND) cases with confirmed C9orf72 hexanucleotide expansions and from control individuals. a RNA was analyzed by RT-PCR as in Fig. 1. GAPDH was used as a loading control. The level of C9orf72 transcripts unspliced at the 5’ and 3’ ends in an FTLD case homozygous for the C9orf72 G₄C₂ repeat expansion was markedly higher than in heterozygous cases (C9^(+/+), far right lane). b Quantitative analysis of intron retention by real-time PCR. Levels of C9orf72 transcripts spliced or unspliced at the 5’ and 3’ end of intron 1 were determined by real-time qRT-PCR in heterozygous expansion carriers (n = 11) and control cases (n = 10). Data are shown as means ± SEM. Each data point represents an individual case, ***P < 0.001, Mann–Whitney U test. C9⁻, controls; C9⁺, expansion carriers. ▲(+/+) indicates the values for the single homozygous case analyzed

Fig. 5

Model of expanded C9orf72 transcripts processing explaining the main pathological features of c9ALS/FTD. Retention of intron 1 generates a C9orf72 mRNA with an enlarged 5’-UTR containing the G₄C₂ repeat sequence. The majority of intron 1-retaining C9orf72 mRNA accumulates in the nucleus where it is targeted to a specific degradation pathway unable to process G₄C₂ RNA repeats that subsequently aggregate into foci. A small proportion of intron 1-retaining C9orf72 mRNA is exported to the cytoplasm for RAN translation into DPRs