A human mitochondrial poly(A) polymerase mutation reveals the complexities of post-transcriptional mitochondrial gene expression.
Bottom Line: The addition of LRPPRC/SLIRP, a mitochondrial RNA-binding complex, enhanced activity of the wild-type mtPAP resulting in increased overall tail length.The LRPPRC/SLIRP effect although present was less marked with mutated mtPAP, independent of RNA secondary structure.We conclude that (i) the polymerase activity of mtPAP can be modulated by the presence of LRPPRC/SLIRP, (ii) N478D mtPAP mutation decreases polymerase activity and (iii) the alteration in poly(A) length is sufficient to cause dysregulation of post-transcriptional expression and the pathogenic lack of respiratory chain complexes.
Affiliation: Wellcome Trust Centre for Mitochondrial Research, Institute for Ageing and Health.Show MeSH
Related in: MedlinePlus
Mentions: The p.N478D mtPAP mutation did not affect either the stability or the location of the enzyme necessitating further investigations to characterize the effect of this mutation on mtPAP. The crystal structure of this non-canonical poly(A) polymerase has been solved (23) revealing that the protein dimerises but lacks a classic RNA-binding domain. The region encompassing p.N478, however, was disordered in the crystal. To determine whether the p.N478 mutation influences the dimer formation, we overexpressed and purified both the wild-type and mutant enzyme for use in in vitro studies. Both wild-type and mutant mtPAP were isolated to electrophoretic purity and subjected to gel filtration analysis. As shown in Supplementary Material, Fig. S3, both mutant and wild-type mtPAP form dimers indicating that the p.N478D mutation does not affect the overall fold or the oligomeric state of the protein. The mutation, therefore, must affect mtPAP function directly, because the N478 region does not contribute to the dimer interface (23). We subsequently compared polyadenylation activities of the two enzymes in vitro. Initially, we generated long RNA substrates [terminal 277 nt of MTND3 or 248 nt of RNA14, with or without oligo(A8)]. The wild-type enzyme extended from both templates irrespective of the oligoadenylation status. The mutant protein, however, was unable to discernibly extend unadenylated MTND3 substrate (Fig. 4A). Using such a large substrate under these electrophoresis conditions, however, meant that it was not possible to determine whether p.N478D was entirely inactive or only able to oligoadenylate, as could be seen for mt-mRNA in the mutant cell lines (Fig. 1A). The absence of any activity would suggest that another poly(A) polymerase would have to be present to add the oligo(A) modification. We therefore designed a number of short (40 nt) RNA substrates to mimic either internal sections of MTCO1 and MTND3 or the native 3′ end of RNA14 post-processing. The latter substrate was synthesized with or without an oligo(A8) 3′ terminal extension to establish whether mtPAP had a different preference for (i) correctly processed 3′ termini compared with any free 3′ end and (ii) to see whether mtPAP only used pre-oligoadenylated transcripts as a template. Saturating levels of wild-type mtPAP were able to extend ∼40 nt from the 3′ terminus of both the internal MTND3 sequences (Supplementary Material, Fig. S4) and also correctly processed RNA14 with or without oligo(A8) ends (Fig. 4B lanes 2 and 6). Thus, the wild-type mtPAP demonstrated no preference for 3′ terminal sequence of the RNA substrate. In addition to the main population of extended products, there was a second, much longer population. The ratio of this product (Fig. 4C lane 6, *) to the main product varied between experiments. In contrast, the mutant mtPAP showed severely compromised activity, where the extension achieved on either template, with or without oligo(A8), was only 5–10 nt (RNA14 shown in Fig. 4B lanes 3 and 7). The mutant protein was not able to generate longer extensions even in prolonged -h incubations. To determine whether the p.N478D mutation affected polymerase processivity, it was first necessary to identify whether mtPAP was distributive or processive under these in vitro conditions. The extension assay was therefore performed under conditions where the enzyme/RNA substrate ratio fell below 1 : 1. Figure 4C shows that the wild-type enzyme does not fully extend the subset of RNA substrate to which it has bound, indicating that under these in vitro conditions, it is distributive, thus precluding any effect of the mutation on enzyme processivity.Figure 4.
Affiliation: Wellcome Trust Centre for Mitochondrial Research, Institute for Ageing and Health.