Targeting RNA G-quadruplex with repurposed drugs blocks SARS-CoV-2 entry

The rapid emergence of SARS-CoV-2 variants of concern, the complexity of infection, and the functional redundancy of host factors, underscore an urgent need for broad-spectrum antivirals against the continuous COVID-19 pandemic, with drug repurposing as a viable therapeutic strategy. Here we report the potential of RNA G-quadruplex (RG4)-targeting therapeutic strategy for SARS-CoV-2 entry. Combining bioinformatics, biochemical and biophysical approaches, we characterize the existence of RG4s in several SARS-CoV-2 host factors. In silico screening followed by experimental validation identify Topotecan (TPT) and Berbamine (BBM), two clinical approved drugs, as RG4-stabilizing agents with repurposing potential for COVID-19. Both TPT and BBM can reduce the protein level of RG4-containing host factors, including ACE2, AXL, FURIN, and TMPRSS2. Intriguingly, TPT and BBM block SARS-CoV-2 pseudovirus entry into target cells in vitro and murine tissues in vivo. These findings emphasize the significance of RG4 in SARS-CoV-2 pathogenesis and provide a potential broad-spectrum antiviral strategy for COVID-19 prevention and treatment.


inhibitory effect of TPT and BBM on SARS-CoV infection in mouse models in vivo.
This study not only expands the existence of RG4 in SARS-CoV-2 host factors, but also provides a novel, to my knowledge, RG4-targeting strategy for COVID-19 prevention and therapy. This manuscript is well designed, and the data are solid. Overall, this timely and novel study is of interest and importance to the pathophysiology and treatment of COVID-19, which could offer new insights to beat the ongoing pandemic.

Response:
We are very grateful to the reviewer's encouraging and thoughtful comments.

Major comments
1. It has clearly showed that the mutations of RG4 sites can increase the protein levels of exogenous host factors, including ACE2, AXL, FURIN, and TMPRSS2 (Fig. 3A). These data are critical to demonstrate that the formation of RG4 structure can repress the expression of these host factors.
However, it remains unclear whether the increase in protein level is due to the difference of transfected plasmid or not. In this regard, the authors should provide the mRNA levels of these host factors after transfection.
Response: Thank the reviewer for this crucial comment. As suggested, we have determined the effect of plasmid transfection on mRNA expression. It showed that transfection of G4WT and G4MUT plasmids into H1299 cells led to a comparable increase in Ace2, Axl, Furin, and Tmprss2 mRNAs (

Both TPT and BBM showed considerable inhibition on ACE2-G4WT
expression (Fig. 3B). However, an inhibitory effect was also seen on ACE2-G4MUT expression (Fig. 3C). What is the potential explanation? It is better to perform additional experiments to strengthen this conclusion, if applicable.

Response:
We appreciate the reviewer for this critical comment. In addition to the dominant PQS-2302, PQS-1682 also has the possibility to form RG4 ( Fig.   1C and S1C Fig). In the original version, only PQS-2302 was mutated in the ACE2-G4MUT plasmid, which may be responsible for the observed inhibition.
In this regard, we have generated a new ACE2-G4MUT plasmid, in which both PQS-2303 and PQS-1682 were mutated with synonymous substitution ( Figure below). Western blot analysis showed that TPT and BBM had no effect on

BBM showed a better effect on SARS-CoV-2 infection than TPT in cultured
cells, but seemed to be less than TPT in mouse models. The authors should appropriately discuss this point.
Response: Thank the reviewer for this important comment. We totally agree with the reviewer on the different effectiveness in vitro and in vivo. A possible explanation is that the concentration of TPT and BBM used in cultured cells comparatively high, which is a safety threshold without inducing cell death.
While the concentration in mouse models was determined with reference to other experimental animal studies, which may not represent the most efficacious dosage in the present study. It is also important to note that, in mouse models, BBM showed a better effect than TPT at the early stage (day 4), but less at the late stage (day 8) (Fig 5C and 5D). This difference might be due to individual variation in the rate of drug absorption and metabolism in the animals, which is more complicated than that in cultured cells. By following this critical comment, we have added a few words into the DISCUSION section as follows: "Interestingly, comparative effects of BBM and TPT in vitro and in vivo appeared to be different. That is, BBM was more effective for inhibiting pseudovirus entry in cultured cells (Fig 4), but was less in mouse models than TPT (Fig 5). This observation might be associated with the differences of these two drugs in optimized dose, and the rate of drug absorption and metabolism in vivo, which awaits further investigation. Taken together, these results strongly suggest a potent inhibitory role of TPT and BBM in SARS-CoV-2 pseudovirus entry." (Page 17, line 8)

Recent findings concerning BBM and TPT on COVID-19 should be
introduced in more details, and appropriately discussed with the results presented in this study.
Response: Thank the reviewer for this constructive comment. Following the suggestion, we have added a few sentences to introduce the background and discuss our results and recent findings concerning BBM/TPT on COVID-19 in the DISCUSSION section (Page 19, line 2). 10. S3 Fig B: "Kiney" should be "Kidney".

Response:
We appreciate the reviewer for careful reviewing. We are so sorry for our carelessness. We have thoroughly checked the manuscript and corrected these typo errors. Therefore,

-either this study positions itself as a fundamental study aimed at highlighting biological mechanisms and characterising them (which it does very well and which is commendable), and part of the article should then be rewritten with this in mind, limiting itself to the results obtained -or it intends to position itself as a study of the efficacy of molecules with antiviral potential, and it is then necessary to carry out complementary in vitro and in vivo studies with SARS-CoV-2 viruses and an adapted animal model and methodology.
It is important to note that the results presented are not usable to infer drug efficacy, but also that the results presented are not suggestive at this stage of a very high inhibitory potential. The potential use of the drug studied to prevent infection would also deserve a lot of precaution, as the molecules mentioned are not harmless and the benefit/risk balance cannot be assessed from the data provided Response: We appreciate the reviewer for the favorable and insightful comments.

Major comments
No key new experiment if the paper is limited to the study of pseudoviruses and removes any reference to efficacy. If demonstration of efficacy is the objective:

-antiviral EC50 and EC90 determination in relevant cells (eg TMPRSS2 vero cells + primary explantation bronchial cells) using real virus and different variants -in vivo experiments in a relevant real virus mouse model or in hamsters
Response: We thank the reviewer very much for these thoughtful and crucial comments. Following these constructive suggestions, we have carefully modified the title, abstract, main text, and Figure legends to emphasize that our manuscript is a fundamental study highlighting biological mechanisms.
Basically, we have replaced "SARS-CoV-2 infection" with "SARS-CoV-2 pseudovirus entry", removed "drug efficacy" concerning our results, limited our conclusions to the results obtained, weakened the clinical potential of our findings, balanced the potential benefit/risk of these two chemicals, and so on.
We do hope these modifications considerable and acceptable for the reviewer.

Minor comments minor edits, eg simulation vs stimulation
Response: We appreciate the reviewer for careful reviewing. This error has been corrected.

Reviewer 3
The approach of developing antiviral therapy, but the experimental design and presented results are not sufficient to support the conclusions.

Response:
We appreciate the reviewer for her/his critical and insightful comments.

Major comments
1. In figure 3

In figure 3, no data actually prove that TPT and BBM indeed inhibit these proteins through targeting RNA G-quadruplex (RG4).
Response: Thank the reviewer for this thoughtful comment. As shown in Fig   3C and 3D, both TPT and BBM diminished the expression of ACE2, AXL, FURIN or TMPRSS2 induced by G4WT plasmids. However, this inhibition was abolished in cells transfected with G4MUT plasmids, in which guanines in Gtracts of target PQSs were substituted with adenines to eliminate RG4 formation with synonymous substitution (Fig 3A). Of note, the difference of G4WT and G4MUT plasmid is the former can form RG4 structure, but the latter cannot form RG4 structure due to synonymous substitution. Therefore, these results, together with the high-affinity binding of RG4 structures with TPT and BBM (Fig 2D-H), strongly suggest that the repression of ACE2, AXL, FURIN and TMPRSS2 by TPT and BBM depends on RG4 formation. Following this important comment, we have added a sentence in the RESULT section as follows: "which were designed to eliminate the RG4 formation with synonymous substitution" (Page 10, line 12), to improve the clarity.

As mentioned in the Introduction, the authors explained that both SARS-
CoV-2 and host factors contain RG4. It is evident that many other genes which may or may not relevant to SARS-CoV-2 could also contain RG4.
There is no experimental data provided in this study demonstrating which

RG4-containing factors (viral or host) actually (partially) mediated the antiviral effects of TPT and BBM. There is only association of TPT and BBM
treatment with protein levels of some host factors presented in figure 3.

Response:
We appreciate the reviewer for this crucial and insightful comment.
We fully agree with the reviewer that it is of importance to demonstrate the contribution of RG4-containing factors to TPT-and BBM-mediated effects. To address this critical concern, we have performed additional experiments, in which ACE2, AXL, FURIN and TMPRSS2-G4WT plasmids, or their corresponding G4MUT plasmids, were simultaneously transfected into hACE2-

Pseudovirus was used throughout the study. It is essential to be validated using infectious virus strains.
Response: Thank the reviewer for this important comment, which is also raised by Reviewer 2. We totally agree with the Reviewer that using infectious authentic virus will greatly strengthen the link between BBM/TPT and SARS-CoV-2. However, due to the extraordinary, limited sources and strict control of biological safety, authentic virus assay was extremely difficult for us to conduct at present. However, TPT, in line with another known RG4 stabilizers, has been shown excellent activity to protect against SARS-CoV-2 by using authentic virus in animal models (Cell, 2021, PMID: 33836156; Cell Discov, 2022, PMID: 36068208). These results are consistent with our data, supporting the potential of our strategy for overcoming SARS-CoV-2. Nevertheless, in combination with the suggestions regarding this issue from Reviewer 2, we have discussed these limitations in the revised manuscript (Page 19, line 14), and modified some sentences to emphasize our current manuscript as a fundamental study.
We hope this is considerable and acceptable for the reviewer.

Minor comments
The introduction section, especially the first paragraph can be shortened.
Response: Thank the reviewer for this thoughtful comment. Following the suggestion, we have simplified the introduction section, especially the first paragraph, in the revised manuscript.