Frataxin fracas

Recently, Carvajal et al. stated that the frataxin gene mutated in Friedreich's ataxia (FRDA) and the neighbouring STM7 gene form a single transcriptional unit. They concluded that the product encoded by the STM7 gene, phosphatidylinositol-4-phosphate-5-kinase (PI4P5K), is defective in FRDA. STM7lies 5' (centromeric) to the frataxin gene, in the same orientation. Its major 3' exon (exon 16) is 20-40 kb from frataxin exon 1 (unpublished results). The small intronless PRKACG gene is localized between the two, on the opposite strand. No STM7-like sequence is localized 5' to the frataxin homologue in Caenorhabditis elegans and Saccharomyces cerevisiae, suggesting that the two genes initially evolved in separate locations. The first frataxin exon contains a CpG island and a translation initiation site preceded by in frame stop codons, good indicators for the 5' end of a gene. Carvajal et al. used primers from frataxin and STM7 in nested reverse-transcription PCR (RT-PCR) and generated several cloned products; three of these from placental RNA, skipped exon 16 of STM7 and exon 1 of frataxin and had variable exon composition. In two cases these included a PRKACG fragment. Another PCR product, from adult atrium RNA, included frataxin exon 1, but its open reading frame terminated before the frataxin sequence. Such heterogeneous STM7-frataxin products are absent from natural cDNAs (that is, cDNAs not obtained by forced exon-connection PCR): i) none of 13 cloned human and mouse frataxin cDNAs and expressed sequence tags (ESTs), nor three retro-transcribed pseudogenes (one human and two mouse) contain any STM7 sequence; ii) none of the 30 cloned human and mouse STM7 ESTs, nor the recently identified mouse cDNA homologue of STM7 contain any sequence corresponding to frataxin. In northern blots, common STM7-frataxin bands are not detectable. Carvajal et al. reported, but did not show, a similar to 1.3-kb band hybridizing to exons 5, 12 and 15 of STM7. Contrary to their conclusion, this cannot be the major frataxin transcript, as the latter is almost entirely accounted for by cloned frataxin cDNAs, leaving insufficient space for three (or more) STM7 exons. Our RNAse protection assays did not reveal transcripts containing exon 2 of frataxin spliced to any exon other than exon 1, at least in lymphoblasts.