مراجعة PREreview لMitochondrial RNA processing promotes translation by resolving structured precursor RNAs

In this manuscript, Nuessmeier et al. from the Churchman laboratory investigated the role of non-canonical processing in the post-transcriptional regulation of mitochondrial mRNAs (mt-mRNAs). The authors provide evidence that while FASTKD4, a member of the Fas-activated serine/threonine kinase family, is required for processing at both the ND5-CYB and ATP6/8-CO3 non-canonical junctions, processing by this factor is specifically required to promote translation of CYB transcripts by resolving a structural barrier within precursor transcripts. This conclusion is well supported by western blot analysis and RNA-seq data showing the impact of loss of FASTKD4 on OXPHOS subunits at the protein and transcript levels. Additionally, the authors use a metabolic labeling assay and mitoribosome profiling to show how FASTKD4 regulates the translation and stability of mt-mRNAs. By comparing the effects of FASTKD4 depletion on the steady-state abundance, RNA turnover, translational efficiency, and RNA structure of ATP6/8-CO3 and ND5-CYB, the authors uncover evidence that a cis-element within unprocessed ND5-CYB transcripts leads to a decrease in CYB translation upon loss of FASTKD4. While their hypothesis is strengthened by DMS-MaPseq data and secondary structure predictions, which agree with previously reported data, this evidence does not directly test their model that the structural element in the unprocessed ND5-CYB transcript impedes mitoribosome loading. However, such a hypothesis is very challenging to directly test. Thus, it is recommended that the authors soften their language, such as in the title, to avoid sounding conclusive.

The authors acknowledge that there are several outstanding questions about non-canonical processing, including the mechanism of FASTKD4-mediated processing of non-canonical junctions. It has previously been shown that FASTKD4 preferentially promotes processing at the ND5-CYB junction while FASTKD5 preferentially aids 5’ CO1 processing. The authors’ work not only supports this previous observation but also proposes that structural determinants within precursor transcripts underlie the specificity of FASTKD4 for ND5-CYB processing. The findings of this work can further the field’s understanding of how mt-mRNA processing can regulate mitochondrial translation which may have important consequences for respiratory function and mitochondrial diseases.

Major points

1. The main claim of this preprint is centered on the author’s hypothesis that processing by FASTKD4 removes a structural element near the start codon of unprocessed CYB transcripts to promote mitoribosome loading. The authors support this hypothesis by documenting the presence of a secondary structure near the CYB start codon using DMS-MaPseq and secondary structure predictions (Fig. 4). However, the authors do not directly test whether the secondary structure blocks translation. Given how challenging it would be to directly test this hypothesis, I recommend the authors tone down their language and discuss alternative hypotheses for the improved translation of CYB, such as loss of a sequence motif or altered binding of a translation activator. As a future direction to test their model, the authors could consider either removing the structured sequence from the mt-DNA with a mitochondrial base editing strategy or use an in vitro mitochondrial translation system to test the impact of the secondary structure on translation of the CYB precursor.

 Minor points

1. In addition to CYB, the authors also observe a decrease in protein abundance of COX1 (Fig. 1B). In the discussion, they speculate that this may be an indirect consequence of reduced CO3 expression and defects in complex IV assembly rather than impaired CO1 transcript processing. However, evidence for these effects is not shown. To test this hypothesis, the authors can provide a western blot of CO3 protein levels to show that there is reduced CO3 expression upon loss of FASTKD4. Additionally, they can perform assays to show that complex IV assembly is impaired in FASTKD4 KO cells. These experiments would strengthen their argument as well as provide broader understanding of the functional consequences of loss of FASTKD4.

2. In Fig. S2C, the authors observe decreased stability of both ATP6/8 and CO3 transcripts in FASTKD4 KO cells. However, it is unclear why CO3 transcript stability decreases to a much greater degree than ATP6/8 upon loss of FASTKD4. If the unprocessed transcript is unstable, ATP6/8 would be expected to be destabilized to a similar extent. However, the data presented would suggest that FASTKD4 plays a role in preferentially stabilizing the CO3 transcript. It would be helpful for the authors to discuss a possible explanation for this observation in the text in order to aid reader interpretation.

3. The authors show a western blot of clonal FASTKD4 KO cell lines in Fig. S2A. However, a faint band is still present in each of the KO lanes and thus they may not be considered true clonal knockout cells. While this does not affect the validity or interpretation of their data since the protein is, at a minimum, very strongly depleted, it would be beneficial for the authors to comment on why this band is observed (e.g. antibody recognition of a non-specific band or contamination of clones with wild-type cells) for transparency.

4. The authors perform sucrose gradient fractionation of mitochondrial polysomes to show that mitoribosome assembly is unaffected by loss of FASTKD4 (Fig. S3A). A nice complement to this experiment would be to visualize mitorobosomes by monitoring UV absorbance. In addition to mitoribosome assembly, this analysis could also provide insight on how global mitochondrial translation is affected upon loss of FASTKD4

Relevant Expertise

The author works in the RNA biology field and is familiar with gene expression assays (western blot, qPCR, ribosome profiling, RNA-seq etc.) but has limited knowledge and expertise on mitochondrial translation and DMS-MaPseq.

I reviewed this preprint as part of the Transparent and Accountable Peer Review training program. This report was written with the support of my research mentor.

Competing interests

The authors declare that they have no competing interests.

Use of Artificial Intelligence (AI)

The authors declare that they did not use generative AI to come up with new ideas for their review.