Decision letter | Splicing repression allows the gradual emergence of new Alu-exons in primate evolution

Open accessCopyright infoDownload PDFDownload figures

Splicing repression allows the gradual emergence of new Alu-exons in primate evolution

Decision letter

Affiliation details

UCL Institute of Neurology, United Kingdom; MRC-Laboratory of Molecular Biology, United Kingdom; Institute de Biologie de l’ENS (IBENS), CNRS UMR 8197, France; University College London Genetics Institute, United Kingdom; Goethe University Frankfurt, Germany; Institute of Molecular Biology (IMB), Germany
Benjamin J Blencowe, Reviewing editor, University of Toronto, Canada

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

Thank you for submitting your article "Splicing repression and NMD control the emergence of Alu-exons" for consideration by eLife. Your article has been favorably evaluated by James Manley (Senior Editor) and three reviewers, one of whom, Ben Blencowe (Reviewer #3), is a member of our Board of Reviewing Editors. The following individuals involved in review of your submission have agreed to reveal their identity: Manuel Irimia (Reviewer #1); Kristen W Lynch (Reviewer #2).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.


In their submitted manuscript Ule and colleagues investigate mechanisms underlying the process of exonization that involve Alu repeat-derived sequences. Extending their previous work published in Cell demonstrating that hnRNPC has a widespread role in silencing Alu exons through binding to adjacent, cryptic 3´splice site polyU-tracts, the authors now show that nonsense mediated mRNA decay (NMD) can additionally act to suppress Alu exonization in a subset of transcripts, some of which are also suppressed by hnRNPC. However, a substantial proportion of Alu exon-containing transcripts that are predicted to be substrates for NMD do not appear to be targeted by this process, and many of these Alu exons are found adjacent to retained introns. The authors further show that ancient Alu exons are typified by having polyU tracts of reduced length but may often be associated with NMD activity, as supported by the observation that these transcripts have relatively reduced expression levels across human tissues.

Overall, this is an interesting and timely study that sheds light on mechanisms governing the suppression of potentially deleterious cryptic exons, the evolution of Alu-derived expressed exons, and a mechanism that may function to modulate mRNA expression across human tissues. Overall the authors' analyses are carefully performed and the conclusions mostly seem appropriate. However, it is requested that the authors address the following main points in a revised manuscript.

Essential revisions:

1) In the first part of the manuscript the authors deplete hnRNPC and assess, using RNA-Seq analysis of cytoplasmic and nuclear fractions, steady-state changes of transcripts that do – or do not – contain Alu exons. Differentially expressed transcripts are enriched for Alu-exons and there is an overall trend for down-regulation of cytoplasmic mRNA expression of genes that harbour these exons, which is validated for a subset of genes using RT-(q)PCR assays. However, hnRNPC depletion leads to differential expression of a large number of genes that lack Alu exons, indicating a more widespread role for hnRNPC in regulating steady-state mRNA levels independent of the presence of Alu-exons. To strengthen the conclusion that reduction in gene expression is specifically caused by increased Alu exon inclusion upon hnRNPC knockdown, rather than through alternative mechanisms, it would be useful to separately assess differential expression levels for those genes that contain cryptic, alternative, or constitutive Alu exons, where genes in the latter group should not be affected. Similarly, the authors should test whether there is a significant negative correlation between increased Alu exon inclusion and expression. This could be assessed in the UPF1-knockdown data where changes in splicing can be more accurately monitored.

2) Also related to point 1, an alternative explanation to the results of Figure 1A it that lowly expressed genes better tolerate Alu-exonization. In this regard, it would be informative to determine whether the expression of Alu-exon-free orthologs of these genes in mouse and other primates also have lower expression. If they do not, this would further support the authors' proposal.

3) The authors examine the role of NMD and show that two thirds of transcripts with hnRNPC-sensitive Alu exons are refractory to NMD (as assessed by UPF1 knockdown), even though the majority of these cases are computationally predicted to be NMD targets. The authors next show evidence that a significant fraction of these transcripts contain retained introns adjacent to the Alu-exons. They conclude that Alu-exon transcripts containing intron retention events can be readily exported and translated in the cytoplasm. However, their interpretation is based on RT-(q)PCR analysis of only a few transcripts, whereas it is unclear whether it applies more generally.

In particular, it is apparent that the majority of Alu exon-containing transcripts that are sensitive to hnRNPC knockdown show an increased nuclear:cytoplasmic ratio upon knockdown, while displaying reduced steady-state levels (lower left quadrant, above diagonal, in Figure 1—figure supplement 4 panel C). It is therefore recommended that the authors globally assess changes in the nuclear:cytoplasmic ratios of transcripts that include Alu-exons, with or without one or both adjacent retained introns, upon hnRNPC knockdown, using reads that span exon-intron junctions. While the authors' data may not be consistent with a "block" to nuclear RNA export that results in nuclear accumulation of Alu-exon transcripts as they claim, it is possible that intron retention in these transcripts leads to nuclear sequestration and subsequent nuclear turnover, and that such effects could in turn lead to reduced cytoplasmic levels of Alu-exon transcripts.

In relation to the above point, it is also unclear from the authors' data that retained intron-containing transcripts are associated with polysomes, beyond the one example that is analyzed. The authors should also not refer to polysome association as translation. Such a conclusion would require evidence at the peptide level to confirm that the corresponding protein is being produced.

4) The authors provide convincing data that Alu (and other) exons with flanking U-tracts are responsive to hnRNP C-depletion, but this doesn't rule out possible roles of other proteins. To what extent can the authors conclude that the U-tracts are functioning through hnRNP C versus altered recruitment of U2AF65 (further strengthening the splice sites), and/or allowing buffering by other U-tract binding hnRNP-like proteins (e.g. TIA). Some of the concern regarding the specificity of hnRNP C function is addressed by the author's earlier work, but it would be helpful to clarify this point in the text.