Single‐cell analyses reveal SARS‐CoV‐2 interference with intrinsic immune response in the human gut

Abstract Exacerbated pro‐inflammatory immune response contributes to COVID‐19 pathology. However, despite the mounting evidence about SARS‐CoV‐2 infecting the human gut, little is known about the antiviral programs triggered in this organ. To address this gap, we performed single‐cell transcriptomics of SARS‐CoV‐2‐infected intestinal organoids. We identified a subpopulation of enterocytes as the prime target of SARS‐CoV‐2 and, interestingly, found the lack of positive correlation between susceptibility to infection and the expression of ACE2. Infected cells activated strong pro‐inflammatory programs and produced interferon, while expression of interferon‐stimulated genes was limited to bystander cells due to SARS‐CoV‐2 suppressing the autocrine action of interferon. These findings reveal that SARS‐CoV‐2 curtails the immune response and highlights the gut as a pro‐inflammatory reservoir that should be considered to fully understand SARS‐CoV‐2 pathogenesis.


20th Jan 2021 1st Editorial Decision
Thank you again for sending us your manuscript "Single-cell analyses reveal SARS-CoV-2 int erference wit h int rinsic immune response in the human gut " and the reviews your received at the other journal. I think that the reviews seem thorough and const ruct ive and we can use them rat her than reviewing the st udy from scrat ch. In line wit h the comment s of the reviewers, I think that the scRNA-seq analyses are relevant for exploring the response to SARS-CoV-2 in the int est ine. We would be happy to consider the st udy and I would therefore invit e you to submit it once you have complet ed the revisions addressing the issues raised. Please include a det ailed point by point response, so that we can easily evaluat e the changes.
In summary, the main issues raised were mostly technical and refer to the need to: -provide further evidence that the used organoids have properly differentiated -better support that the infected vs uninfected cells have been correctly determined -clarify/better support that annotation and clustering processes -better support the conclusions regarding transcriptional repression We think that it would be important to address the technical issues delineated above. Given the technical nature of the concerns I think that we would need to consult an external expert (e.g. by a member of our Edit orial Advisory Board or an expert in the field) and ask them to look at the performed revisions.

9th Feb 2021 1st Authors' Response to Reviewers
Specific answers to the editor:

1)
provide further evidence that the used organoids have properly differentiated We have now added a panel (Fig. S1A) to show that our two dimensional organoids fully differentiate and have clarified this in the methods section. We have also added references that also use a two dimensional approach with differentiated organoids for infectious disease research.

2)
better support that the infected vs uninfected cells have been correctly determined We have now better justified why our approach of removing a baseline of virus counts in the Targeted scRNAseq allows us to correctly determine the infected cells. In response to the reviewer suggestion, we evaluated SoupX a and showed this approach is not suitable to correct the detection of contaminating viral transcript in our datasets, as this falsely enriches small cells (please see specific answer to reviewers).

3)
clarify/better support that annotation and clustering processes We added figure S3, here we clarify the different steps we used for the clustering and annotation of our colon and ileum organoids and we showed the top markers differentially expressed in each cluster. Importantly, we also show the correlation between the cell types from mock sample and infected sample and the association between our annotations and the previously reported annotation in the equivalent tissue cell atlas. Finally, with respect to the presence of immature enterocyte 2, we now better refer to our previous work where we have clearly identified immature enterocyte 2 in human ileum biopsy, further suggesting that we have properly annotated our data. There is currently a lack of consensus in the field with respect to the precise identity and subclusters of cells expected to be found in intestinal organoids. This is mostly due to the fact that there just a handful of reports describing scRNAseq of human intestinal organoids. To further justify our clustering we would like to direct you on the ressources website of Alex Shalek lab (world leading expert in scRNAseq at MIT) where one could find a UMAP of intestinal organoids clearly showing immature enterocytes (the shalek lab refer to them as early enterocyte). http://shaleklab.com/resource/covid-19-resources/?

4)
better support the conclusions regarding transcriptional repression We have clarified in the text that our experiments reporting inhibition of IFN-mediated signaling using different multiplicity of infection ( Figure 6) were analyzed by normalizing to a housekeeping gene (which will then compensate in case of global transcription repression). Most importantly, we have added the Fig.S2 and S5C showing that the modulation of markers we observed is not a consequence of a global down-regulation of gene expression modulated by the virus, as can be seen by the similar number of detected transcripts of treated and untreated cells.

9th Mar 2021 1st Revision -Editorial Decision
Thank you again for submit ting your work to Molecular Syst ems Biology. We have now heard back from the reviewer who was asked to evaluat e your st udy. As we previously discussed, the reviewer was given access to the manuscript and your responses to the reviewers' comment s from the other journal. They were asked to evaluat e whet her the reviewers' concerns have been adequat ely addressed, and to assess the suit abilit y of the revised st udy for publicat ion keeping in mind the edit orial crit eria of Molecular Syst ems Biology. As you will see below, the reviewer is overall support ive. However, they recommend some modificat ions, which we would ask you to perform in a revision.
Most issues raised refer to the need to include furt her discussions and clarificat ions in order to present the pot ent ial limit at ions of the st udy in a more balanced way, given that it is performed in organoids and not in pat ient s. Regarding the ident ificat ion of infect ed vs byst ander cells, an issue that was prominent in the previous report s, the reviewer st rongly recommends including support ing dat a using an ort hogonal met hod, to bet ter support the relat ed conclusions. On a more edit orial level, we would ask you to address the following. In this manuscript , the aut hors st udy the int est inal tropism of SARS-CoV-2 by following the infect ion in a 2D model of primary human int est inal organoids. They st udied the relat ionship bet ween the levels of Ace2 expression wit h suscept ibilit y and found a lack of correlat ion bet ween these feat ures. They furt her describe differences in induced proinflammat ory gene expression signat ures bet ween infect ed and byst ander cells.
I appreciat ed the discussion of the inappropriat e usage of ACE2 expression as the only basis for conject ures about the infect abilit y of cell types or organs. While Ace2 expression is easily measurable by scRNAseq, many virologist s (Vincent Racaniello amongst ot hers) have warned that a recept or is necessary but not sufficient for infect ion. The int ernal cellular environment has to play a permissive role and further external factors are involved.
Several groups have used intestinal organoids to study SARS-CoV2 infection in the intestinal epithelium. While I certainly appreciate the feasibility of this approach, I'm not fully convinced of its relevance to COVID19 pathophysiology in humans. The weaknesses of the model and its application in this study have been raised by the previous round of reviews and has been partially addressed by the authors. I understand that studying the intestinal response to SARS-CoV2 at single-cell resolution in a more relevant context would be logistically challenging due to timing and biosafety issues and lies beyond the scope of the current manuscript. I am not opposed to the publication of this manuscript in MSB but I strongly recommend a discussion of the weaknesses and limitations of the used model system in the discussion section and discourage uncritical extrapolation of the findings to human disease.

Specific comments:
The proportions of cell types and states in intestinal organoids do not fully reflect the tissue of origin. Growth in the Sato medium results in strongly stem-cell biased conditions, the here employed differentiation protocol seems to induce more mature cells but (as mentioned by the previous reviewer 1) the lack of goblet cell capture in the colon organoids is worrying. The authors argue that their scRNAseq dataset of human intestinal organoids is consistent with the (limited) existing literature, yet this does not mean that organoids are an appropriate model for the study of cell tropism and will faithfully reflect the tissue of origin. This should be discussed.
Along similar lines: I am not convinced by the use of Cyp3a4 expression as a single qPCR target to justify "full" differentiation of the cultures. In their previous Cell Reports study the authors have used SI as marker for differentiation, why do they use Cyp3a4 here? There is a continuum of gene expression along the Crypt-Villus trajectory in ileal enterocytes (or crypt bottom to top in colonocytes, respectively), either the authors should validate the expression of true top markers or they should mention in the discussion the remaining uncertainties regarding the in vivo relevance of the present results (since organoids are likely not reflecting all enterocyte states).
The identification of infected vs bystander cells is of importance in this manuscript. Both previous reviewers have questioned the applied approach with a hard threshold. I am not fully convinced by the novel explanations (I assume the authors referred to Figure S4 and not S3 as mentioned in their response to the reviewers). I understand the problems with the proposed SoupX approach, the resulting data should be presented in the supplementary figure. To clarify the issue and to increase confidence in the applied method I suggest an orthogonal measurement of the number of infected cells by quantifying fluorescently labeled infected organoids (as in Fig 3f) and use scRNAseq on the second half of the same sample for comparison. Figure 7: the depiction of a human gut implies in vivo relevance of the presented findings. I have not seen evidence that the proposed interaction between infected and bystander cells is relevant outside of organoid cultures. I have also missed in vivo validation of the existence/relevance of the described organoid immature enterocyte states 1 and 2, so I ask the authors to replace the gut depiction with an organoid scheme and to mention this in the discussion.  Most issues raised refer to the need to include further discussions and clarifications in order to present the potential limitations of the study in a more balanced way, given that it is performed in organoids and not in patients. Regarding the identification of infected vs bystander cells, an issue that was prominent in the previous reports, the reviewer strongly recommends including supporting data using an orthogonal method, to better support the related conclusions.
We have now modified our discussion to avoid over interpretation of our organoid model and highlighted its limitations. The orthogonal approach (immunostaining) was already performed in the original version but we did not include a quantification of the number of infected cells. In the revised version, we have included this quantification (Fig EV1 and EV3) and are now commenting in the text that the numbers of infected cells are matching between the single cell experiments and our immunofluorescence. Importantly we have also highlighted that the immunostaining samples were performed in parallel to the sequencing.
Response to the reviewer comments: We thank the reviewer for their comments and suggestions. We believe this new version is now strengthened, more clear and also less speculative in our conclusion to the implication of our findings to the "real" human gut. In short, we have addressed the main concern of the reviewer and have updated our text to show the 21st Mar 2021 2nd Authors' Response to Reviewers limitations of the organoids system compared to "real" primary human intestinal tissue. While we would have loved to do this study with primary patient material, as the reviewer mentioned this was out of the scope of this project. We have actively tried to assess biopsies of infected patients through our gastroenterologist collaborator. While a great number of patients had mild to severe diarrhea, we concluded that their gastroenteric symptoms did not justify the risks associated with the invasive process of taking a biopsy as many of the patients were on blood thinners and were at a high risk for these procedures.
Reviewer #1: In this manuscript, the authors study the intestinal tropism of SARS-CoV-2 by following the infection in a 2D model of primary human intestinal organoids. They studied the relationship between the levels of Ace2 expression with susceptibility and found a lack of correlation between these features. They further describe differences in induced proinflammatory gene expression signatures between infected and bystander cells.
I appreciated the discussion of the inappropriate usage of ACE2 expression as the only basis for conjectures about the infectability of cell types or organs. While Ace2 expression is easily measurable by scRNAseq, many virologists (Vincent Racaniello amongst others) have warned that a receptor is necessary but not sufficient for infection. The internal cellular environment has to play a permissive role and further external factors are involved.
Several groups have used intestinal organoids to study SARS-CoV2 infection in the intestinal epithelium. While I certainly appreciate the feasibility of this approach, I'm not fully convinced of its relevance to COVID19 pathophysiology in humans. The weaknesses of the model and its application in this study have been raised by the previous round of reviews and has been partially addressed by the authors. I understand that studying the intestinal response to SARS-CoV2 at single-cell resolution in a more relevant context would be logistically challenging due to timing and biosafety issues and lies beyond the scope of the current manuscript. I am not opposed to the publication of this manuscript in MSB but I strongly recommend a discussion of the weaknesses and limitations of the used model system in the discussion section and discourage uncritical extrapolation of the findings to human disease.
We thank the reviewer for their comments. We agree that these models are important as they allow us to address cell lineage specific differences but they do have their limitations. The discussion has now been updated to reflect these limitations.

Specific comments:
The proportions of cell types and states in intestinal organoids do not fully reflect the tissue of origin. Growth in the Sato medium results in strongly stem-cell biased conditions, the here employed differentiation protocol seems to induce more mature cells but (as mentioned by the previous reviewer 1) the lack of goblet cell capture in the colon organoids is worrying. The authors argue that their scRNAseq dataset of human intestinal organoids is consistent with the (limited) existing literature, yet this does not mean that organoids are an appropriate model for the study of cell tropism and will faithfully reflect the tissue of origin. This should be discussed.
We agree with the reviewer that organoids do not fully recapitulate the tissue and are limited to the differentiation media that they are exposed. Unlike in the body, organoids are kept in a high stem cell state which allows them to constantly proliferate. The Sato differentiation media that we use induces differentiation but the ratio of the absorptive vs secretory lineages is different from in vivo. This has been now discussed in the text.
Along similar lines: I am not convinced by the use of Cyp3a4 expression as a single qPCR target to justify "full" differentiation of the cultures. In their previous Cell Reports study the authors have used SI as a marker for differentiation, why do they use Cyp3a4 here? There is a continuum of gene expression along the Crypt-Villus trajectory in ileal enterocytes (or crypt bottom to top in colonocytes, respectively), either the authors should validate the expression of true top markers or they should mention in the discussion the remaining uncertainties regarding the in vivo relevance of the present results (since organoids are likely not reflecting all enterocyte states).
We have used several markers for the validation but only included a few. We have not updated the figure to include a few more makers (including SI) and have updated the discussion regarding the remaining uncertainties and limitations associated with organoids.
The identification of infected vs bystander cells is of importance in this manuscript. Both previous reviewers have questioned the applied approach with a hard threshold. I am not fully convinced by the novel explanations (I assume the authors referred to Figure S4 and not S3 as mentioned in their response to the reviewers). I understand the problems with the proposed SoupX approach, the resulting data should be presented in the supplementary figure. To clarify the issue and to increase confidence in the applied method I suggest an orthogonal measurement of the number of infected cells by quantifying fluorescently labeled infected organoids (as in Fig 3f) and use scRNAseq on the second half of the same sample for comparison.
We thank the reviewer for their comment as this was apparently unclear from our text. We have now updated the text to more clearly explain that all single cell experiments were also done in parallel with immunofluorescence and qPCR validation. Analyses of our single cell sequencing experiments revealed a similar number of infected cells compared to the direct determination of the intectivity by immunofluorescence as depicted in the time course of infection presented in Figure  EV1. We have realized that the quantification of the number of infected cells from our infected organoids was missing in our original submission. We have now added a panel in EV1 showing the percentage of infected cells and the percent of infected cells was also added to the single cell data in EV3. Additionally, we have also added the SoupX data and made it into a new supplementary figure (Appendix Figure 2). Figure 7: the depiction of a human gut implies in vivo relevance of the presented findings. I have not seen evidence that the proposed interaction between infected and bystander cells is relevant outside of organoid cultures. I have also missed in vivo validation of the existence/relevance of the described organoid immature enterocyte states 1 and 2, so I ask the authors to replace the gut depiction with an organoid scheme and to mention this in the discussion.
We thank the reviewer for this comment. It is true that organoids are not in vivo models and more work would need to be done to make precise statements about how infection impacts the gastrointestinal tract of COVID-19 patients. We have updated the figure and the text and toned down our statements in the discussion. We have checked the image and from our side the heatmaps appear ok. It seems that maybe something happened in the compression of the file when sent to reviewers. Let us know if this problem occurs again in this version and we will upload these individual figures separately to the MSB website.
We would like to thank the reviewer once more for their time. We do appreciate all comments as we feel that the manuscript has now been improved in readability, has a more balanced conclusion avoiding overstatement of our findings to the human gut and we do feel that discussing the limitation of organoids is also important for the field.
Looking forward to hearing from you Megan Stanifer and Steeve Boulant, on behalf of all co-authors.
29th Mar 2021 2nd Revision -Editorial Decision , Thank you for sending us your revised manuscript . We think that the performed revisions have sat isfact orily addressed the issues raised by the reviewer. As such, we are glad to inform you that your st udy can soon be accept ed for publicat ion, pending some minor modificat ions list ed below.

31st Mar 2021 3rd Authors' Response to Reviewers
The authors have made all requested editorial changes.

Data
the data were obtained and processed according to the field's best practice and are presented to reflect the results of the experiments in an accurate and unbiased manner. figure panels include only data points, measurements or observations that can be compared to each other in a scientifically meaningful way. graphs include clearly labeled error bars for independent experiments and sample sizes. Unless justified, error bars should not be shown for technical replicates. if n< 5, the individual data points from each experiment should be plotted and any statistical test employed should be justified the exact sample size (n) for each experimental group/condition, given as a number, not a range; Each figure caption should contain the following information, for each panel where they are relevant:

Captions
The data shown in figures should satisfy the following conditions: Source Data should be included to report the data underlying graphs. Please follow the guidelines set out in the author ship guidelines on Data Presentation.
Please fill out these boxes ê (Do not worry if you cannot see all your text once you press return) a specification of the experimental system investigated (eg cell line, species name).
We performed all experiments in triplicate and no pre-determination was performed. N/A. No animals were used for this study.

B-Statistics and general methods
the assay(s) and method(s) used to carry out the reported observations and measurements an explicit mention of the biological and chemical entity(ies) that are being measured. an explicit mention of the biological and chemical entity(ies) that are altered/varied/perturbed in a controlled manner. a statement of how many times the experiment shown was independently replicated in the laboratory.
Any descriptions too long for the figure legend should be included in the methods section and/or with the source data.
In the pink boxes below, please ensure that the answers to the following questions are reported in the manuscript itself. Every question should be answered. If the question is not relevant to your research, please write NA (non applicable). We encourage you to include a specific subsection in the methods section for statistics, reagents, animal models and human subjects.

definitions of statistical methods and measures:
a description of the sample collection allowing the reader to understand whether the samples represent technical or biological replicates (including how many animals, litters, cultures, etc.).