Identification and quantification of FXN antisense transcript 1 (FAST-1) in friedreich ataxia

Abstract

Friedreich ataxia (FRDA) is a lethal autosomal recessive neurodegenerative disorder caused by expanded GAA repeats in the FXN gene, resulting in local epigenetic changes and reduced expression of the mitochondrial protein frataxin. The disease is characterised by neurodegeneration of large sensory neurons of the dorsal root ganglia and spinocerebellar tracts. It has been recently reported that a novel frataxin antisense transcript, FAST-1, is overexpressed in FRDA patient derived fibroblasts. However, the lack of fundamental information about FAST-1 gene such as size, sequence, length and its origin has hindered the understanding of its interactions with FXN gene. Therefore, I proposed to investigate these characteristics of FAST-1 in a panel of FRDA cells and mouse models. Firstly, using Northern blot hybridisation with small and large riboprobes, I identified two bands with different sizes (~500 bp and 9 kb size), representing potential FAST-1 transcripts. Then to confirm the exact size and the location of the FAST-1 gene, I performed 5’- and 3’ RACE experiments, followed by cloning and sequencing. This analysis resulted in identification of the 5’- and 3’-ends of FAST-1, which mapped to nucleotide positions ‘-359’ and ‘164’ of the FXN gene, giving the total length of FAST-1 as 523 bp size. Strikingly, the full-length 523 bp FAST-1 transcript also corresponds to one of the Northern blotting results where I identified a band at approximately 500 bp size, indicating that the Northern blotting may have correctly identified the same full-length FAST-1 transcript. Subsequently, by optimising number of experimental parameters within our lab, I developed a robust qRT-PCR method to quantify FAST-1 expression levels. Using this technique, I analysed the expression pattern of this antisense transcript in various FRDA cell lines and mouse models. I confirmed the original finding of increased FAST-1 levels in human FRDA fibroblasts, and further quantified FAST-1 levels in FRDA mouse model cell lines and tissues. However, no consistently altered patterns of FAST-1 expression were identified in relation to FXN expression. Therefore, either they are not directly connected, as originally reported by De Biase et al., or their relationship varies between cell and tissue types. Lastly, improved understanding of epigenetic changes in FRDA and growing evidence on long-gene regulation led me to study the ‘neighbouring genes’ rather than just focusing on the FXN gene. Therefore, I studied a region of approximately 750 kb on both sides of the FXN and quantified the expression levels of two genes (PGM5 and PIP5K1β) on 5’- end and four genes (TJP2, FAM189A2, APBA1 and PTAR1) on 3’- end of FXN gene in human primary fibroblasts. I found that PGM5 and PIP5K1β genes, located at 5’- end of the FXN genes, were downregulated in FRDA fibroblasts and these findings coincide with the recent epigenetic changes identified in FRDA, where significant enrichment of gene repressive histone marks and increased DNA methylation were shown in upstream region of GAA repeats in intron 1 of the FXN gene. Out of four genes that were studied in the 3’- end of the FXN gene, only one gene (APBA1) was downregulated, which suggests that there are fewer repressive epigenetic marks downstream of the GAA repeat.