PREreview of The HIV-1 restriction factor RPRD2 does not inhibit transcription of HIV-1 or endogenous retroelements

Date:15OCT26

Reviewer: Brandon Fonseca (PhD candidate, Summers lab, UMBC)

Preprint: The HIV-1 restriction factor RPRD2 does not inhibit transcription of HIV-1 or endogenous retroelements

Overview

This paper examines regulation by the protein Regulation of Nuclear Pre-mRNA Domain Containing 2 (RPRD2), which was previously shown to be a restriction factor for HIV-1 as well as regulator for cellular transcription. This dual involvement was of interest to the authors’ goal of illuminating how HIV-1 transcription is regulated within the cells. Using a GFP reporter construct, they found RPRD2 localized at the location of transcription and integration in nuclear speckles. Using immunoprecipitation pulldown assays, they showed that it interacted with both the HUSH complex which regulates retroelements and PAF1, a known restriction factor. Despite all this, when they began testing if RPRD2 can affect HIV-1 transcription they found no evidence to support that it plays a role in that process. They first tested whether RPRD2 regulated LTR promoters using a GFP reporter and siRNA knockdowns and found it did, but only if the promoter was in a plasmid. When they repeated this process with an integrated LTR promoter, the GFP signal remained the same regardless of RPRD2 depletion. These findings will save time and resources in future studies by eliminating RPRD2 as a possible target for study within the broader question of endogenous regulation of HIV-1 and how RPRD2 plays into latency. This paper identified and systematically tested RPRD2 activity in the cell regarding HIV-1 transcription and contributes to the larger field in a way that not many do: it presents findings that disprove their hypothesis.

Below are major and minor recommendations after having read this paper.

Major:

1. The authors used computer modeling to investigate binding. This was done by first predicting binding interfaces, then introducing a glutamic acid mutation in place of a positively charged residue followed by predicting binding free energy. Figure 1D shows that the predicted energies increase overall and suggests the importance of a positive charge. This could lead to a false positive in this situation by placing a negative charge near actual nucleic acid binding sites. This could disrupt interactions through like-charge repulsion that does not exist in the wild-type protein. Instead of using a highly negative glutamate, a neutral or uncharged residue such as methionine or leucine would test if this positive charge was required for binding without the confounding factor of the RNA-Glutamate repulsion. Since this was done with predictive software, would this mutation be possible to include?

Minor:

1. Using computational methods, the authors predicted that certain mutations would inhibit RNA binding of RPRD2 (figure 1.E-F) but did not provide a physical assay such as an EMSA. It would improve the argument that the modeling process was accurate if the actual mutations could be introduced to RPRD2 and show that RNA binding affinity is affected.

2. Closing out the paper the authors said, “future endeavors should instead focus on its ability to disrupt reverse transcription as the likely principal mechanism”. This is missing the context of where and when reverse transcription takes place. Recently, Li et al. 2020 (DOI:10.1073/pnas.2019467118) found that the capsid remains intact just before integration, which suggests reverse transcription takes place within the capsid. It is important for the authors to discuss how the protein RPRD2 could inhibit HIV-1 reverse transcription given the evidence that this process occurs within the intact HIV-1 capsid.

3. For figure 1A (top) the text states “RBDmap highlighted two RNA binding sites within the C-terminal region of RPRD2” and shows four peaks above what appears to be a cutoff. This data seems to show four predicted binding locations, two of which are in the NTD which is not mentioned in the text. Further, consider removing the mention of the LM as it could distract the reader who might think the data is intended to evaluate its prediction performance.

4. The presentation of the Western blots for Figures 1G and 4F could be improved for better readability. Figure 1G would benefit from a slight rotation to align the lanes with the page. Figure 4F is tightly cropped, making it challenging to understand the full context; either reducing the crop or zooming in less will add clarity to the protein bands being evaluated.

5. The core question seems to be whether RPRD2 regulates nascent transcription of HIV-1 provirus. The section “RPRD2 does not regulate integrated HIV LTR-driven transcription” answers this by showing that RPRD2 depletion does not affect the expression of an integrated LTR promoter. With this question answered, the following two sections on LINEs and the interferon response address a different question. Clearly stating the rationale at the start of the next sections as to why LINEs make good analogs for HIV-1, or the interferon response would help prevent any confusion for the reader.

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.