PREreview of Systematic mapping of chromatin dysregulation driven by viral transcriptional regulators at scale
- Published
- DOI
- 10.5281/zenodo.24375003
- License
- CC BY 4.0
In this manuscript, Corban and colleagues set out to gain a systematic understanding of how virally-encoded transcription factors alter host chromatin accessibility. They leverage a method previously used to study oncofusions, PROD-ATAC, to carefully dissect the role of viral transcription regulators (vTRs) using a defined locus in a human cell line. To enable studying many vTRs simultaneously, they use single-cell ATAC-seq and a vTR barcode to assign accessibility profiles to individual proteins. Their pool of proteins is broad and samples a diverse variety of viruses. They recapitulate many known effects of vTRs including those from EBV and HTLV. They also establish that many more vTRs affect genome accessibility despite few functional annotations suggesting they would. The authors also extend their PROD-ATAC approach to chemical genetics, comparing how vTRs affect chromatin accessibility after inhibition of key cellular pathways. Overall, this manuscript represents an impressive amount of work spanning a large taxonomic range that will be useful for the systems virology field and spur exciting questions for experimentation in native viral systems. It also serves to highlight the utility of the authors’ methodology. We have several suggestions for additional controls, ways to contextualize the data, and improve the clarity of presentation.
Major points:
1. The foundation of this study involves ectopic expression of virally-derived cDNAs, and a negative result could be due to one of several reasons including lack of sufficient expression. Are the vTRs are expressed at variable levels, either at the RNA or protein level? Especially for those that do not significantly alter chromatin accessibility, it would be worthwhile to discuss whether this could simply be due to lack of transcription or translation. I suggest selecting a few representative classes of behaviors and probe the RNA expression by RNA-seq or RT-qPCR and protein levels by Western blotting, flow cytometry, or immunofluorescence.
2. How many of the vTRs are likely to be nuclear and could thus have a direct effect, or are cytoplasmic and are therefore modulating accessibility indirectly through host factors? This could be addressed via immunofluorescence or biochemical fractionation on a representative subset. Regardless, it would be helpful to include a series of protein domain diagrams highlighting known or putative functional regions, along with an indication of whether these could predict subcellular localization.
Minor points:
1. In Figure 1D, it might be helpful to color code the peribunyavirus non-structural proteins mentioned in the text so that a reader can see them amongst the other points.
2. In referring to the PROD-ATAC results initially, the authors mention there being 137,491 high-quality nuclei, and 43,798 passed stringent filtering. The following paragraph suggests that LSI was performed on the larger set, but the figure legend suggests only 40,835 were used for UMAP. Grouping the same sets of nuclei will provide the reader with a better sense of the limitations of each analysis. It would also be helpful to specify which are being used more clearly.
3. In the Results section discussing Figure 2, it would be helpful to point out that in addition to viral co-factors, it is possible the cell type context (293T) might not reflect native viral biology and could obscure some findings.
4. In Figure 2A, the volcano plots show that many vTRs affect peak accessibility, but it is hard to make firm conclusions beyond that. Demonstrative browser tracks might be more helpful, as they could indicate how an important gene/locus changing could impact downstream gene expression.
5. The heatmap in Figure 2B would be easier to interpret if:
a. The vertical clustering were clearer (i.e., point out if there is some organization to the peaks being shown).
b. Similar to Major Point 2: Add an annotation indicating domains contained within each protein in addition to the functional indication. For example: do all of these contain DNA-binding domains? RNA? Transactivation domains?
6. The authors have established a powerful system for comparing the effects of vTR expression, but the rationale for some comparisons in Figure 4A (Tax v LMP1) and 4B (Orf10 v NS1; Orf10 v EBNA3B) is not obvious.
7. In all cases, the vTR expression data are compared to the empty control as the reference. In Figure 4G, indicating what the Empty control looks like would help contextualize the remaining tracks.
8. As the authors collected such an impressive amount of data, showing the distribution of results will help the reader appreciate how varied the effect sizes can be. In Figure 5A, is it possible to show the range of data points to have a better understanding of the distribution of cells?
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 author declares that they have no competing interests.
Use of Artificial Intelligence (AI)
The author declares that they did not use generative AI to come up with new ideas for their review.