Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.
Learn more at Review Commons
Reply to the reviewers
*Reviewer #1 (Evidence, reproducibility and clarity (Required)):
*The study examined the mechanisms behind the nuclear transport of capsid proteins of various flaviviruses. The study used mass spectrometry to identify the interaction partners of JEV capsid protein and found Importin 7 as the top hit. After validating this interaction with IP-western blotting, using IPO7 knock-out cells they showed that the nuclear accumulation of capsid is dependent on IPO7. Moreover, they also observed nearly 10-folds reduction in titre of virus produced from knock out cells without reduction in virus replication or particle assembly.
The study needs improvements to bring it to publication standards. Some overaarching problems include, all capsid localization studies being done with GFP-tagged capsid, and not wild type capsid produced during authentic infection, lack of quantitation of most of the localization data and not showing capsid localization from infection experiments in knock out cells, and no in-depth analysis of the potential mechanisms behind the observed reduction in titre in knock out cells etc.
Thank you for your constructive comments. We have sincerely answered all of them, as shown below. We hope you are satisfied with our additional data and the revised manuscript.
The major comments are
Fig 1B: Please add quantitation and statistical analyses of the ratio of nuclear and cytoplasmic capsid protein of all different capsids used. Also include western blot to prove that there is no cleavage between Capsid and GFP and the green signal indeed comes from the fusion protein. Ideally you should use capsid alone instead of a fusion protein for at least selected few constructs to prove that the Capsid-GFP behaves identical to Capsid alone.
Following the reviewer’s comments, we have added quantification and statistical data in Figure 1D. We have added CBB data and western blot data in Figures 1B and S1. Because recombinant proteins of low molecular weights were artificially translocated into the nucleus through diffusion, less than 20 kDa proteins are typically used as GFP or GST fusion proteins for the IJ and PM experiments. Instead of IJ and PM experiments, we have added data on the translocation of the non-tagged core using IFA and its statistical data in Figure 1A. Although in vitro data on the translocation of capsid protein differ somewhat from IFA data, the data on nuclear translocation of core proteins are consistent across different experiments.
Fig 1C: It is unclear from the figure legends the WT JEV capsid means GFP-Capsid or Capsid alone. You should clearly state the GFP part if the construct includes GFP. Quantitation and statistics are missing and the information on how many independent experiments were performed is also not included in the figure legend.
Following the reviewer’s suggestion, we have described that the JEV proteins fused GFP as follows: “AcGFP-JEVCoreWT or AcGFP-JEVCoreGP/AA” (Line. 771). We added quantification and statistical analysis as shown in Figure 1E. IJ and PM experiments were performed three times independently and described in the legend of Figure 1 in the revised manuscript (Lines 773–774).
Fig 2B: Quantitation and statistics are missing. Ideally, the data need to be reproduced with Capsid alone instead of Capsid-GFP. A positive control is needed for the activity of Bimax to prove that the drug was working in the assay.
We have added quantitative and statistical data in the revised Figure 2B. As mentioned above, capsid alone is potentially translocated into the nucleus artificially using the IJ and PM assay. Bimax binds to importin alpha but not importin beta, specifically inhibiting the importin alpha/beta pathway. The RanGTP mutant binds to the importin beta family, including importin beta 1, and widely inhibits importin beta-dependent nuclear import. These inhibitors are well-characterized and recognized in the field. We cited the following reference: Tsujii et al., JBC, 2015.
Fig 2C: How do you reconcile the IP mass spectrometry data that Importin b1 is the second strongest hit with the lack of IP interaction you observed in fig 2C?
As shown in Figure 2C, importin b1 does not interact with the JEV core. Importin b1 is the most abundant member of the importin beta family. Thus, it might be a non-specific interaction between importin b1 and the JEV core. Therefore, we excluded importin b1 from further analyses. We added a sentence to explain why importin b1 was excluded on Line 145.
Fig 3C: How many independent confirmations of this experiment was performed?
All IJ and PM experiments were performed thrice independently. We described this in the legend of Figure 3 in the revised manuscript (Line; 794).
Fig 4A and B: Add quantitation for the western blot. 4A-D Include data on the number of biological repetitions. 4C-D: Add quantitation and statistical analyses of the ratio of nuclear and cytoplasmic capsid protein.
We have added quantification data, as shown in Figures 4A and 4B. All experimental results shown in Figures 4A, 4B, 4C, and 4D were performed thrice independently, as described in the legend of Figure 4 of the revised manuscript (Lines; 810-812).
Fig 5B. This data should be shown in the context of infection with untagged Capsid at least for 1-2 viruses. This is a serious drawback of the present study as there is no clear evidence presented that the native capsid protein in an infection context depend on importin 7 for nuclear accumulation and behave similar to the GFP-Capsid constructs being used.
Following the reviewer’s concerns, we used an un-tagged JEV and DENV core to examine core translocation in WT or IPO7KO Huh7 cells. As shown in Figures 5C and 5D and their quantitative data, nuclear translocation of JEV and DENV core protein was inhibited in IPO7KO Huh7 cells. We tested the translocation of core protein upon infection with DENV as shown in Figure 5F. Although we could not examine ZIKV infection because we could not find appropriate antibodies against the ZIKV core, these data are consistent in that nuclear translocation of flavivirus core protein largely depends on IPO7.
Fig 5 A-D: Two repetitions are insufficient; a minimum of three biological repeats and statistical analysis need to be included. 5E-F: You cannot do statistics on two repeats, need minimum of three repeats to perform statistical analysis. 5G-H: I presume three repetitions based on the data points shown, this should be clearly stated in the figure legend.
We repeated three independent experiments, shown in Figures 5A and 5C-5F, and indicated them on Lines 823. We have added statistical data in Figures 5B-5F. We have corrected the statement of biological repeats in Figures 6A and 6B (Lines; 843-844).
Fig 5E-G: Taking the data of 5E and 5G together it seems Importin 7 functions as the level of particle release and not particle assembly or maturation. Have you checked for the specific infectivity of the particles released from knock out cells to determine the reason behind the reduction in virus titre? You could look at the prM maturation by furin cleavage to check it this is altered in the IPO7 knock out cells.
We determined the ratio of infectious titer per 103 copies of viral RNA in Figure 6F. The proportion of infectious viruses targeting extracellular JEV RNA was decreased in IPO7KO cells. Simultaneously, no difference was observed in the proportion of infectious viruses targeting intracellular JEV RNA between WT and IPO7KO cells. Although we could not find appropriate antibodies against the JEV core, we checked prM expression using the DENV virus. The expression of prM was slightly increased in JEV-infected IPO7-KO Huh7 cells (Figure S3D). This result suggests that the efficiency of prM cleavage by furin was partially involved in the impairment of infectious virus release in IPO7KO Huh7 cells.
Fig 5H: Have you checked if the observation regarding intracellular RNA levels in 5F is applicable to these viruses as well.
We checked the intracellular RNA levels of DENV and ZIKV-infected cells. In contrast to JEV, intracellular ZIKV or DENV RNA showed no difference in IPO7-KO Huh7 cells (Figure 6H). We discuss it in Discussion section (Lines; 269-271)
Fig 6: The figure legend "Data are representative of two (A, B) independent experiments and are presented as the mean {plus minus} SD of three independent experiments (C)" is confusing. The sentence should be reworded to state the repetitions separately for independent experiments. Fig 6C should show original titres and not percentages.
We have corrected Figure legends according to the reviewer’s comments. We have showed the original titers in Figures 6C and 6E.
Fig 7B: This experiment should be performed in IPO7 knock out cells to confirm that the observed reduction of core mutant is mainly contributed from its lack of interaction with IPO7 and not from any other confounding factors.
Following the reviewer’s suggestion, we performed SRIP experiments for GP/AA mutation using IPO7KO Huh7 cells. As shown in Figure 7C, the SRIPs harboring WT core were impaired in IPO7KO Huh7 cells; no difference was observed in the SRIPs harboring GP/AA mutations in WT and IPO7KO cells. These results suggest that IPO7-dependent nuclear translocation of core protein is important for the viral release.
Reviewer #1 (Significance (Required)):
While the authors could convincingly demonstrate the interaction between capsid and IPO7, how that interaction results in the observed reduction in viral titre is largely unexplored. As all the localization data used a GFP-tagged capsid outside an infection context, this reviewer is not confident that all the reported observations will hold in an infection setting. This need to be urgently addressed to rise the confidence about the observation. The current data is insufficient to confidently attribute the change in titre to the interaction between capsid and IPO7 and the capsid localization to the nucleus. Knocking out IPO7 could have pleotropic effects independent of capsid nuclear accumulation that could lead to the observed titre reduction. This need to be addressed further before linking both these phenotypes. Certain key experiments needed to address these questions are currently missing. While the interaction of Capsid with IPO7 is certainly intriguing, the implications of this interaction on virus biology needed further investigation before clear conclusions can be drawn regarding this observation.
Reviewer #2 (Evidence, reproducibility and clarity (Required)):
Summary: In this study Itoh and colleagues investigate the mechanism, role and impact of the nuclear localization of the flavivirus core protein. The import of the core protein has long been observed and investigated and herein the authors use some novel approaches to identify potential cellular binding partners that facilitate nuclear import. Via proteomics and biochemical approaches they determine that importin-7 plays a crucial role in the import of the core protein that appears to be conserved across Flavivirus members. In general the findings and conclusions are sound but there are some significant omissions and caveats that warrant further investigation.
Major comments:
- one of the major caveats of the study is that the flavivirus NS5 protein also translocates to the nucleus in an Importin-alpha/beta dependent manner. Therefore how can the authors discount any impact of preventing NS5 import, in addition to core, on virus and SRIP replication and production. Some discussion, if not additional experiments are required here ie. NS5 localization in the KO cells during virus infection
We examined the localization of NS5 using IPO7KO Huh7 cells. As shown in Figure S2D and S2E, we confirmed that IPO7 was not involved in the nuclear localization of NS5.
- the localization is predominantly nucleolus rather that nucleoplasm when compared to the SV40 NLS. What are the sequence differences between the flavivirus proteins that potentially could account for this? A protein known to localize solely to the cytoplasm should also be used eg. NS1 or NS3.
The JEV core does not contain a consensus nucleolar localization signal. Nuclear localization of NS5 depended on importin-α similar to the SV40 NLS, while flavivirus core proteins were independent of importin-α. Gly42 and Pro43 are critical amino acids for the nuclear localization of the core protein, as shown in Figures 1C and 1D. The Gly42 to Pro43 of core proteins were well-conserved in the core proteins of the Flaviviridae family.
- controls for Figure 2? Ie. a protein known to be inhibited by Bimax but not the RanGTP mutant and vice versa.
Bimax binds to importin alpha but not importin beta and specifically inhibits the importin alpha/beta pathway. The RanGTP mutant binds to the importin beta family, including importin beta 1, and widely inhibits importin beta-dependent nuclear import. These inhibitors are well-characterized and recognized in the field. Therefore, we have cited the following references: Tsujii et al., JBC, 2015.
- Fig 5. Difference with WNV and DENV in nucleoplasm localization but also WNV still appeared to have Core in the nucleus in the KO cells
We agree with the reviewer’s comment about differences in nuclear localization among the viruses using the IJ assay. We have added new data to examine the localization of the DENV core after DENV infection. Nucleolar localization of the DENV core following DENV infection was observed, as shown in Figure 5F. Therefore, differences in nucleoplasm or nucleolar localization among different viruses shown in Figure 1C and Figure 5B might be artifacts of recombinant proteins. One possibility is that the localization of core proteins using IJ assay was detected by anti-GFP antibodies. Although purified GFP-core proteins, as shown in Figure 1B and S1, were observed as a single band of fusion proteins, core proteins of WNV and DENV might be cleaved during IJ experiments, and GFP alone might be detected at nucleoplasm, as shown in Figure 5B. Because our study focused on the nuclear translocation of flavivirus core proteins, the detailed localization of each core protein in the nucleus will be studied in the future.
- Fig 5C still has substantial JEV and DENV core but not WNV and ZIKV. Why is the DENV and WNV localization pattern different to Fig 5B?
We appreciate the reviewer’s suggestion; we re-checked all our data presented in Figure 5B and other data shown in Figure 5B. We quantified the ratio of nuclear localization as shown in the right of Figure 5B. Our quantification data showed that the nuclear transport of all core proteins used in this study was dependent on IPO7. In contrast, Figure 5A shows that nuclear translocation of WNV core protein is partially dependent on IPO7. This discrepancy might be explained that nuclear translocation of WNV core protein might be regulated by several nuclear carriers. We described this in discussion section (Line; 250-254).
- Fig 5F, does the KO also restrict NS5 from entering the nucleus and could this then results in increase polymerase activity confined to the cytoplasm resulting in more viral RNA?
Following the reviewer’s suggestion, we examined NS5 localization during viral infection and plasmid transfection, as shown in Figure S2D and S2E. Previous data regarding the nuclear localization of NS5 depended on importin-α. Our data are consistent with previous reports that IPO7 was not involved in the nuclear localization of NS5. In contract to JEV, we also confirm that intracellular ZIKV or DENV RNA showed no difference in WT and IPO7-KO Huh7 cells (Figure 6H). As described in the discussion, other factors, such as antiviral factors, might be involved in IPO7-mediated nuclear transports in JEV infected cells (Line; 269-271).
- Why was WNV infection not performed in Fig 5H? What where the viral tires compared to for the relative % values?
Because our institution does not have a BSL3 facility, we could not use WNV. Following the reviewer’s comment, we showed viral titers in Figure 6G.
- Fig 6B, still a significant amount of core present in the nucleolus. Also WT cells have (almost?) no cytoplasmic staining for core where this could be clearly observed in the WT cells in Fig 5D. Why the difference?
Plasmid transfection of AcGFP-Core WT showed that almost all core proteins were located in the nucleus. We assumed that AcGFP might influence nuclear exports of core proteins or the efficiency of nuclear transports as shown in other data of in vitro experiments. However, our finding that IPO7 was involved in the nuclear transport of core proteins is consistent.
- In Fig 7B, D and E, when were the SRIPs collected and what was the time period after subsequent infection?
Following the reviewer’s comments, we have added more details on SRIP experiments in Materials & Methods (Line; 521-523).
- In Fig 7C was the luciferase measured from the initial transfection and how did it correlate with RNA production? A 15-fold increase in replicon RNA actually seems quite low over a 48h period
Because large amounts of in vitro-transcribed replicon RNA were injected into cells in this experiment, we observed that significant amounts of luciferase values were detected after 4 h. However, the 15-fold enhancement in luciferase value was consistent with previous reports (PMID: 30413742, PMID: 17024179). We have added references in the revised manuscript.
- quantitation is required throughout all of the experimental IFA data provided
Following reviewer comments, we have quantified all IFA data and showed their results.
Reviewer #2 (Significance (Required)):
The nuclear translocation of flavivirus protein has long been studied and it has been observed that the core, NS5 (RNA polymerase) and potentially the NS3 (helicase/protease) proteins all translocate the nucleus. Importin alpha and beta have been shown to facilitate this process. The authors aim to extend this to identify importin-7 as a major cellular factor enabling nuclear translocation. Overall the experiments have been performed well but there is a lack of quantitation for many of the results an suitable controls are required.
I am a researcher in the field of flavivirus replication
Reviewer #3 (Evidence, reproducibility and clarity (Required)):
In the presented study the authors identified and mechanistically investigated how Flaviviruses including Japanese encephalitis virus (JEV), Dengue virus (DENV), and Zika virus (ZIKV) commonly use importin-7 (IPO7), an importin-β family protein, as a cellular carrier protein to facilitate nuclear core protein translocation. The authors evaluated how the production of infectious viruses is regulated by IPO7 using cellular infection models including IPO7-deficient knockout cells. In the submitted manuscript, the authors provide evidence that IPO7 facilitates viral core protein import into the nucleus of infected cells, which is essential for effective Flavivirus replication.
Taken together, the study is interesting to a broader readership with interest in molecular virology, and its findings are informative for potential future targeting of IPO7 to affect flavivirus replication using small molecule drugs. The manuscript is well-written and easy to follow, the methods are appropriate, the structure is logical, and statistical analysis is adequate.
Major comments:
- It is unclear why the authors specifically used Ala substitution at Gly42 anb Pro43 to obtain the abolishment of nuclear core protein localization. It would be helpful to put this into more context and explain the approach.
Mutations of Gly42 and Pro43 to Ala were previously reported and characterized by the same research group (PMID: 15731239). Following the reviewer’s comment, we have added more details of GP mutations in the text (Lines 66–70).
- In Figure 4, the authors claim that the binding between IPO7 and RPS7 is disrupted upon the addition of RanGTPQ69L. This is not clearly evident from the pulldown experiment and should be proven experimentally with additional experiments (e.g. by using an imaging approach) to underline the statement that the binding mode of IPO7 to the JEV core protein is similar to that of RPS7. Loading controls for pulldown blots should be added.
As described in response to the comment by reviewer#2 regarding Figure 2, the RanGTPQ69L mutant inhibits the interaction between the importin beta family, including IPO7 and its substrates, by directly binding to importin beta proteins. For the benefit of readers without knowledge of the typical Ran-dependent nuclear transport mechanism, we have described its effects with several cited references (Dickmanns et al., 1996; Tachibana et al., 2000). We referred to a study that showed that IPO7 transports RPL proteins, including RPS7 (Jäkel and Görlich, 1998). The data in Figures 4A and 4B demonstrate that adding RanGTPQ69L remarkably reduces the binding of IPO7 to the Core proteins and that the effect is more robust than that for RPS7. We believe that these results are experimentally valid, indicating that nuclear transport of Core proteins by IPO7 is achieved through a typical Ran-dependent pathway.
- Most methods used are presented logically but require some more details so that they can be reproduced. In particular, the difference between Figure 4 E and 4H is confusing. What is the difference? Is 4E showing intracellular viral titers and 4H infectious viral titers in the supernatant of cells? Clarification needed. Put relevance of these experiments in context of the hypothesis.
We apologize for the confusion regarding the data in Figures 5E and 5H (we assume). These data were derived from the same experiments, except for the time-course data presented in Figure 5E. We have removed Figure 5E to simplify our results.
- Identical phenotypes induced by IPO7 knockout in a number of HuH7 clones are shown in Figures 6A to 6C. This data does not add to the overall understanding and should be moved to supplementary figures. Why are 293T cells used in experiments shown in Figure 6D and 6E? What is the relevance of kidney cells to Flavirius infections?
Following the reviewer’s comments, we have moved Figure 6 to supplementary figures. We used 293T cells because of efficient JEV propagation and gene-deficient efficiency. We wanted to demonstrate that our data are not Huh7-dependent through experiments in 293T cells.
- Prior studies are referenced appropriately, however, in a recent study it was demonstrated that IPO7 is stabilized upon Epstein-Barr Virus infection and that IPO7 presence is required for the survival of host cells (Yang YC, Front Microbiol. 2021 Feb 16;12:643327. doi: 10.3389/fmicb.2021.643327).
We deeply appreciate the publications in these fields. Following the reviewer’s comment, we have cited these references.
This important study about the physiological relevance of IPO7 during viral infections has not been cited by Itoh and colleagues in the presented study. However, the results of the uncited study are very relevant to the provided manuscript, since Itoh and colleagues are using IPO7 knockout cells to investigate its function in Flavivirus core protein nuclear import. Hence, the authors should perform cell survival and cellular fitness experiments to demonstrate that observed phenomena of reduced viral replication and virus export in IPO7 knockout cells are independent of compromised cellular fitness due to IPO7 deficiency.
We evaluated cellular fitness between WT and IPO7KO Huh7 cells using PI (Propidium Iodide) staining through flow cytometry. As shown in Figure S2F, no differences were observed in cell viability between WT and IPO7KO Huh7 cells. It suggests that viral titers reduced in IPO7KO Huh7 cells are not involved in cellular fitness.
Minor comments:
- Describing Figure 3B, the authors state that they focused on IPO7 among the core binding proteins belonging to the importin-b family, because IPO7 "was identified the most peptides" in the mass spectrometry approach. This requires a more detailed explanation. Also, an explanation of why HEK293T cells were used for this approach and not HuH7 cells, as used predominately in most parts of the study, would provide more clarity to the reader.
We focused on IPO7 because it had the highest number of detected peptides, and we found that the second most detected peptide, IPOB1, did not bind to JEV core proteins as shown in Figure 2C. Therefore, we included the lack of interaction between IPO7 and IPOB1 as part of the rationale.
- In Figures 4E and 4F, colour coding is missing.
We have indicated color coding in this data. Thank you for your comments.
Reviewer #3 (Significance (Required)):
The provided manuscript 'Importin-7-dependent nuclear localization of the Flavivirus core protein is required for infectious virus production' by Itoh and colleagues investigates a topic with important scientific relevance. The presented study builds on previous findings by the authors where they have demonstrated that Flavivirus core protein nuclear localization is actually conserved among Flaviviridae and represents a potential target for broad-range antiviral small molecule drugs (Tokunaga et al., Virology, 2020 Feb;541:41-51). However, our understanding of Flavivirus core protein nuclear localization during viral replication and how the processes could potentially be targeted using novel therapeutic drugs remains elusive. Here, the provided manuscript addresses a mechanistic investigation of how the Flavivirus core protein is actually translocated from the cytoplasm to the nucleus of infected cells. The study is informative particularly for virologists with expertise in Flavivirus replication.
However, from my point of view as a virologist investigating host-pathogen interactions with a strong interest in clinical translational, the manuscript requires a more careful evaluation and interpretation of some results of key experiments. In addition, some of the results need to be more precisely described for clearer understanding by a broader readership.
Reviewer #4 (Evidence, reproducibility and clarity (Required)):
Summary:
In the manuscript entitled "Importin-7-dependent nuclear localization of the Flavivirus core protein is required for infectious virus production", by combining proteomics, CRISPR/Cas9 gene KO, CLSM and standard virology techniques, Yumi Itoh report novel data concerning the involvement of IPO7 in the nuclear and nucleolar localization of Flaviviridae core nuclear and nucleolar localization and viral particle release. Surprisingly, IMPa/b1 inhibition via Bimax2 does not affect core nuclear transport, whereas both RanQ69L and WGA did so. The authors try to identify the cellular transporters involved in core nuclear import, and to this end performed a MS spec analysis of JEV core interactors, which yielded IPO7 as the most likely candidate. After confirming the result by Co-IP, the authors go on showing most core proteins require IPO7 for nuclear delivery using Huh7 and HEK7 IPO7-KO cells, with the exception of WNV core which was able to partially enter the nucleus. In such cells, upon infection, extracellular (but not intracellular) viral titers were strongly reduced, a phenotype which was observed with a JEV core mutant bearing the Gly42 and Pro43 to Ala substitutions in a previous study.
Major comments:
- The major conclusions of the study are:
1.IPO7 is the main driver of core nuclear transport
2.Core nuclear localization is somehow important for viral particle release
Both conclusions are well-supported by experimental evidence.
Methods are clear and precise, the study appears to have been produced with high quality standards, and so is the presentation of the results.
A few controls however should be added to increase the reliability of the results presented here (see below)
Since the authors attempt to link the phenotype observed on virus release upon IPO7 KO to defects on core nuclear import by making a parallelism with core GP/AA mutant, it would be important to know the behavior of such virus in Huh7 wt and Huh IPO7 KO cells. In other words, is GP/AA JEV released efficiently in Huh7 IPO7 KO cells?
We have added new data examining the propagation of the GP/AA JEV mutant in IPO7KO Huh7 cells (Figure 6F). Our new data showed that there were no differences in the propagation of the GP/AA mutant in WT and IPO7-KO Huh7 cells.
A similar approach can be applied to data shown in Figure 7 (effect on release on a capsid nuclear deficient mutant). This would help understand if IPO7 KO, viral release defects and core nuclear import are somehow linked.
We produced SRIPs harboring GP/AA core using WT and IPO7KO Huh7 cells and demonstrated that the number of infectious viruses produced by WT and IPO7KO Huh7 cells was the same (Figure 7C).
Minor comments:
INTRODUCTION
• “Flaviviruses...are mosquito-borne human pathogens"
What about tick borne encephalitis virus?
We have corrected it (Line; 43-44).
- " replication.... occur in the endoplasmic reticulum (ER)"
This sentence is a bit inaccurate. Flaviviridae RNA replication occurs in so-called viral replication factories, double membrane vesicles which are partly derived from the ER. see "PMID: 26958917".
We have corrected this sentence according to the reviewer’s comment (Line; 60-62).
- "it is known that some flavivirus core proteins are translocated from the cytoplasm into the nucleus"
o I think the first evidence of core in the nucleus dates back to 1989, and here it might be appropriate to cite the original reference: "PMID: 2471810".
o It might be worth mentioning that NS5 has also been reported in the nucleus (See "PMID: 28106839")
We have corrected the sentence according to the reviewer’s comment (Line; 63-65).
- "In the cytoplasm, NLS-containing proteins are recognized by importin-α "
o This is true only for classical NLSs, not every NLS binds IMPa, as the authors confirm in this study! Indeed, we have also PY-NLS, IPO7 specific NLSs, IPOb1 NLSs, etc. I therefore suggest rephrasing.
Thank you for pointing out the exact description of NLS. We agree with the reviewer’s comment that “NLS” includes all types of signal sequences, such as PY-NLS. To clearly distinguish between the CLASSICAL nuclear transport pathway by importin α/β1 and the various nuclear transport pathways by the importin β family, such as transportin, we refer to NLS as classical NLS (cNLS) in the document. We have modified the following sentence by adding “such as transportin” and “without importin-α.”
RESULTS
• Fig. 1.
o it is not clear what is new here, with respect to what has been already published. The authors should clearly differentiate novel findings from confirmatory results
Thank you for your suggestion. We would like to introduce our new assay using recombinant virus core proteins, as shown in Figures 1C and 1D. The data shown in Figure 1 are crucial for understanding our data in Figure 2, and we believe this figure is required for broad-ranging readers.
Fig. 2 and 4
o Proteins whose nuclear transport is dependent on IMPa/IMPb1 (such as SV40 NLS) are lacking here
Bimax binds to importin alpha but not to importin beta and specifically inhibits the importin alpha/beta pathway. The RanGTP mutant binds to the importin beta family, including importin beta 1, and widely inhibits importin beta-dependent nuclear import. These inhibitors are well-characterized and recognized in the field. Therefore, we have cited the following references: Tsujii et al., JBC, 2015.
- Fig.5
o It would be important to know the effect on total virus infectivity (intracellular + extracellular) and total viral RNA. It would also be important the effect on RNA replication by using a subgenomic viral replicon (with deletion of the env gene for example). The question here is if IPO7 depletion affects to any extent viral genome replication, and this is impossible to assess in a fully assembling system.
We determined the ratio of infectious titer per 103 copies of viral RNA in Figure 5D. The proportion of infectious viruses targeting extracellular JEV RNA was decreased in IPO7KO cells, and there was no difference in the proportion of infectious viruses targeting intracellular JEV RNA between WT and IPO7KO cells. We examined the effects of IPO7 on viral RNA replication of subgenomic replicon. We showed that the deficiency of IPO7 enhanced viral RNA replication as shown in Figure 7E. As described in the Discussion section, IPO7 may transport other factors possessing antiviral activity against flaviviruses. These data will be investigated in the future.
o Panels A-F legend is missing, consider adding it?
We have added more details to Figure 5A-5F following the reviewer’s suggestion.
- Fig.7
o I did not completely understand how NLuc is the readout here
To quantify RNA replication, we quantified Nluc values using a plate reader. We have added more details on the reporter assay in Materials and Methods (Line; 521-523).
o Also, I do not understand if the effect of GP/AA substitution of panel B has already been reported or if it is a novel finding
Previous reports regarding the effect of GP/AA substitution of JEV showed the impairment of infectious virus release. However, the SRIP assay was performed to examine the viral release step. Our detailed data showed that the lack of IPO7-mediated nuclear transport of core proteins impaired infectious viral release, and our new results using SRIPs harboring GP/AA core showed that the lack of nuclear transport of core proteins also impaired the release of infectious viruses. Our data strongly suggest that the lack of nuclear transport of core proteins influences the viral release.
- All CLSM figures lack quantification (Fn/c; Fno/n)
We have added quantitative data for IFA experiments in our revised manuscript.
DISCUSSION
• "The nuclear entry of viral genomic DNA has been demonstrated to involve IPO7"
o It would be nice to know which viruses the authors are freeing to here
We have added the virus name and corresponding references.
- "While RNA viruses, including flaviviruses, are considered to replicate in the cytoplasm of mammalian cells, increasing evidence suggests nucleolar localization of the viruses "
o I suspect Rawlinson did not propose the viruses localize to the nucleolus, as this sentence seems to imply. Rather, a trafficking of viral proteins to nucleoli, to manipulate cell function, is more realistic. I suggest considering rephrasing.
We have corrected this sentence.
Reviewer #4 (Significance (Required)):
SECTION B - Significance
========================
- Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field.
As alluded to above, this work presents several advances of current knowledge in the field of viral proteins nuclear trafficking, and in Flavivirus biology. The finding of most core proteins depending on IPO-7 is novel and intriguing, and opens the question of what makes WNV core special. Indeed, this protein nuclear targeting is only partially inhibited in IPO7 deficient cells. The fact that the authors extend their findings to several Flaviviruses adds significance.
The role of nuclear core for virus release is also intriguing, but appears poorly characterized. In this respect a mechanistic explanation of the phenomenon would be highly desirable to increase the significance of the work presented here.
In this context I would have a few suggestions:
A) The authors performed MS spec on JEV core, this most likely resulted in a long list of "hits". However, they only report IMPb superfamily members. This is perfectly fine, since they focus at identifying partners responsible for nuclear import. However, it might be helpful for understanding the role of nuclear core. By comparing MS of wt core and GP/AA core, and or wt core in wt and IPO7KO cells, authors could identify core biding partners in the nucleus (in the nucleolus?) which are important for virus release. This could be subsequently addressed by knocking down these factors and study the effect on virus life cycle.
We appreciate the reviewer’s valuable comments. We did not perform MS analysis on GP/AA core protein and core protein using WT or IPO7KO Hun7 cells. To report IPO7-mediated core translocation simply, we would like to cite our manuscript focusing on IPO7. To clarify the importance of nuclear transport of core protein on the viral life cycle, we will perform wide-ranging proteomics.
- B) Further, the authors should try to address the role of core in the nucleus (and nucleolus). Does it interact with cellular/nucleolar proteins? Does it deliver viral RNA to sites of assembly? Does it interfere with rRNA synthesis? All these findings would be easily obtainable using the GP/AA virus and/or Huh7 KO cells, and tremendously increase the impact of the study, which at the moment is limited at points 1 and 2 in the first section of the current report.
Thank you for your valuable comments. We agree that we should clarify the roles of the nucleus or nucleolar localization of the core protein. We tested the effects of rRNA synthesis on JEV core expression. Our data showed that core protein expression slightly impaired the maturation of rRNA synthesis, as shown here. However, the core expression did not influence protein translation. We focused on the phase separation capacity of core protein localized in the nucleolar or nucleus. From our accumulating data, we hypothesized that the acquisition of phase separation capacity of core protein might be involved in an efficient virus release step. We hope that these data will be reported in the near future.
Overall, this work should be interesting for both cell biologists interested in trafficking of viral proteins, and virologists interested in virus-host interactions. The antiviral approach at the moment is a bit less convincing, but the manuscript might be interesting for scientists trying to develop new antiviral strategies. (In this context it might be worth reading and possible discussing the very recent paper from the Bartenschlager group "PMID: 37702492."
Also, I think that it would be worth discussing the recent discovery that a closely related virus belonging to the Hepacivirus genus within the Flaviviridae family, mediated re-localization of Nups to viral replication factories, where they are believed to control access to DMVs interior, thereby regulating virus replication and assembly. Could the core IPO7-interaction have any role in core delivery to DMVs? See "PMID: 26150811".
Thank you for your valuable comments. We have added several sentences in the Discussion section (Line; 297-305). We will investigate the role of nuclear transports in viral life cycles in the future.
Since I am a molecular virologist studying viral nucleocytoplasmic trafficking, virus-host interactions, and antiviral drug-discovery I think I have sufficient expertise for an informative and helpful revision of this work.