Group II intron splicing in wheat mitochondria during embryo-to-seedling development

Abstract

Plant mitochondrial transcripts undergo complex RNA processing, including intron removal and C-to-U editing. My focus has been on the excision of the highly-structured group II introns, mainly located in NADH dehydrogenase ( nad) genes encoding for complex I of the respiratory chain. Their splicing mechanism is expected to be biochemically similar to that of nuclear spliceosomal introns whereby lariats are generated via a two-step transesterification pathway utilizing an unpaired adenosine (near the 3' end of the intron) as the first attacking nucleophile. However, certain plant mitochondrial introns deviate from classical group II features, one example being nad1 intron 2. To better understand the splicing mechanisms of aberrant plant mitochondrial introns, I examined their excised physical forms in germinating wheat embryos by employing strategies such as RNase H analysis, RT-PCR and cRT-PCR (RNA circularized with RNA ligase) to distinguish between lariat, linear and fully-circularized RNA species. For nad1 intron 2, which lacks the bulged adenosine, I found a mixed population of precisely full-length linear molecules, as well as circular ones that included blocks of non-encoded nucleotides. Interestingly, in the case of the cox2 intron, which contains a bulged adenosine, lariats were not predominant. Instead, I observed full-length linear introns that possessed non-encoded 3' terminal adenosines, as well as heterogeneous circular introns that lacked 3' nucleotide stretches. These mixed populations of excised introns suggest the use of multiple splicing mechanisms in vivo including hydrolytic splicing and potentially the use of alternative nucleophiles. Based on northern blot analysis, my examination of high molecular weight precursors which contained one or multiple group II introns (nad7, nad1, cox2, rps3) suggests that transcription is outpacing the RNA processing events during wheat germination (6 hr to 2 day stages of development). This contrasted with abundant mRNAs, for both the respiratory chain genes (nad7, cox1, cox2, atp6) and ribosomal protein genes ( rps2, rps3, rps7), in respective dormant embryos and 6 day seedlings. When cox2 precursors were examined, they were found to be incompletely edited in germinating embryos when compared to seedlings at both splice junctions and at internal exon sites. This supports the model that editing and splicing machinery are rate-limiting during the early stages of seedling development. Although fully-spliced mRNAs showed complete editing at expected sites at all developmental stages, including in dormant seeds, exon editing positions up to 6 nt away from unspliced introns were unedited in nad4 and nad7, while nad2 and nad5 counterparts were fully edited. This suggests an inter-relationship between splicing and editing at junction sites that would include (1) an unspliced intron sterically hindering editing machinery, or (2) an editing recognition motif generated only when the intron is excised and consecutive exons are joined together. It will be of future interest to elucidate the plant mitochondrial nuclear-encoded proteins and/or small RNAs involved in group II intron splicing and editing.