A novel lineage-tracing mouse model for studying early MmuPV1 infections

Decision letter after peer review:

Thank you for submitting your article "A novel lineage-tracing mouse model for MmuPV1 infection enables in vivo studies in the absence of cytopathic effects" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Diane Harper as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Megan E Spurgeon (Reviewer #1); Neil D Christensen (Reviewer #2); Sheila Graham (Reviewer #3).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

Specifically comments and suggestions from reviewer 1 meet the concerns of all reviewers.

Additional Experiments/Data/Analyses:

1. One of the major pieces of evidence that the MmuPV1-lox-Cre-lox virus can infect murine skin and ultimately establish a productive viral infection is the expression of viral transcripts by RNA in situ hybridization provided in Figure 3. It has been noted by multiple groups that the RNAscope assay using MmuPV1-specific probes can detect MmuPV1 viral DNA. In order to distinguish between viral DNA versus RNA transcripts, treatment with DNAse and/or RNAse can be performed (for example, see Xue et al., 2017 and Spurgeon and Lambert, 2019). The images in Figure 3 are blurry, but it does appear that there is ample nuclear signal in some of the tissues (usually indicative of viral DNA). Because the presence of MmuPV1 viral gene transcription is being used as a readout for active infection and is therefore such a foundational piece of evidence supporting their infection model system, the authors should make an effort to perform DNAse/RNAse treatment in their RNAscope assays to verify the signal is, in fact, due to the presence of MmuPV1 viral transcripts.

2. The production of infectious virus following MmuPV1-lox-Cre-lox delivery and recircularization would provide strong validation of their infection model and act as convincing evidence for its experimental utility. Assays such as L1 protein detection in infected tissues, isolation of viral particles from swabs/tissues, or using swab samples to infect the skin of immunodeficient mice would provide such key data.

3. For the skin swab and subsequent PCR experiment used to detect MmuPV1 L1 and Cre loss shown in Figure 3E, mock-infected control swabs should be included. The gel image quality could also be improved, if possible. In its current form, it is difficult to agree with their conclusion that they detect Cre loss by Day 5.

4. The authors performed flow cytometry to quantify Ki67 positive cells and measured an increase in YFP+ (infected) cells. Direct visualization of Ki67-positive cells by either IF/IHC in the regions of infected skin where they detect RNAscope-positive cells (those shown in Figure 3) would greatly strengthen and complement their data that infected cells exhibit higher proliferation dynamics shown in Figure 4.

Recommendations for improving the writing/presentation:

1. Regarding the title and the words "in the absence of cytopathic effect": This topic is hardly revisited in the Results/Discussion, nor is any rigorous histopathological analysis performed to exclude the presence of cytopathic effect. While their rationale that this will facilitate the study of subclinical infections is legitimate, this phrase does not seem to add anything to the impact of the report or accurately describe the findings discussed therein.

2. Conclusions and Significance (Lines 436-441): Length restrictions being appreciated, discussing a few of the 'outstanding questions in papillomavirus biology' that the authors believe their new infection model can help address would help bolster and convey the importance of their findings.

3. It would be useful to more clearly indicate in the Results and Figure Legends which genetic background is being used in each experiment. It is unclear where C57BL/6 versus FVB mice were used.

4. Lines 69-72: while there are broad similarities between HPVs and MmuPV1 with respect to genome structure, encoded proteins, etc., they are actually quite different at the level of nucleotide sequence (HPV16 vs. MmuPV1 only ~ 50% similar). Likewise, HPV and MmuPV1 E6 and E7 proteins are less than 50% similar at an amino acid level (please refer to Joh et al., J Gen Virol, 2011). While MmuPV1 is certainly a useful tool for studying PV infection and pathogenesis, as is the focus in this report, this text overstates the similarities between HPV and MmuPV1 and should be amended to more accurately reflect the differences.

5. Lines 78-80: It may be worth noting here that subclinical MmuPV1 infections have been observed by several research groups at several anatomical sites. Including this information would strengthen the use of the MmuPV1 infection model to evaluate these understudied types of infection.

6. The images in some of the figures (particularly Figure 2B-C and 3A-B) are blurry and out of focus.

7. The PCR gel image in Figure 3E is very hard to see and is therefore not as convincing as it could be.

8. Line 88: Please provide rationale for infecting tail skin versus other sites of skin (dorsal, ear, etc).

9. Line 136: Please provide technical information regarding the isolation of MEFs or provide a citation.

10. Lines 292-293: The sentence "…only immunocompromised mice are susceptible to the mouse papilloma virus" is not accurate. There are several reports of immunocompetent strains of mice that are susceptible to MmuPV1 infection and disease. It may be more accurate to state that immunocompromised mice are more susceptible.

11. Reference issues:

a. Line 66: Please refer to the original paper that described the discovery of MmuPV1 (Ingle et al., Vet Pathol, 2011) instead of the review article currently cited.

b. Line 77: There is also an inducible mouse model of HPV oncogene expression (see Bottinger et al., Carcinogenesis 2020).

c. Line 77: There are certainly more novel infection models than the Wei et al., 2020 cited here that could be included to accurately reflect the state and scope of MmuPV1 infection models.

Reviewer #1 (Recommendations for the authors):

Additional Experiments/Data/Analyses:

1. One of the major pieces of evidence that the MmuPV1-lox-Cre-lox virus can infect murine skin and ultimately establish a productive viral infection is the expression of viral transcripts by RNA in situ hybridization provided in Figure 3. It has been noted by multiple groups that the RNAscope assay using MmuPV1-specific probes can detect MmuPV1 viral DNA. In order to distinguish between viral DNA versus RNA transcripts, treatment with DNAse and/or RNAse can be performed (for example, see Xue et al., 2017 and Spurgeon and Lambert, 2019). The images in Figure 3 are blurry, but it does appear that there is ample nuclear signal in some of the tissues (usually indicative of viral DNA). Because the presence of MmuPV1 viral gene transcription is being used as a readout for active infection and is therefore such a foundational piece of evidence supporting their infection model system, the authors should make an effort to perform DNAse/RNAse treatment in their RNAscope assays to verify the signal is, in fact, due to the presence of MmuPV1 viral transcripts.

2. The production of infectious virus following MmuPV1-lox-Cre-lox delivery and recircularization would provide strong validation of their infection model and act as convincing evidence for its experimental utility. Assays such as L1 protein detection in infected tissues, isolation of viral particles from swabs/tissues, or using swab samples to infect the skin of immunodeficient mice would provide such key data.

3. For the skin swab and subsequent PCR experiment used to detect MmuPV1 L1 and Cre loss shown in Figure 3E, mock-infected control swabs should be included. The gel image quality could also be improved, if possible. In its current form, it is difficult to agree with their conclusion that they detect Cre loss by Day 5.

4. The authors performed flow cytometry to quantify Ki67 positive cells and measured an increase in YFP+ (infected) cells. Direct visualization of Ki67-positive cells by either IF/IHC in the regions of infected skin where they detect RNAscope-positive cells (those shown in Figure 3) would greatly strengthen and complement their data that infected cells exhibit higher proliferation dynamics shown in Figure 4.

Recommendations for improving the writing/presentation:

1. Regarding the title and the words "in the absence of cytopathic effect": This topic is hardly revisited in the Results/Discussion, nor is any rigorous histopathological analysis performed to exclude the presence of cytopathic effect. While their rationale that this will facilitate the study of subclinical infections is legitimate, this phrase does not seem to add anything to the impact of the report or accurately describe the findings discussed therein.

2. Conclusions and Significance (Lines 436-441): Length restrictions being appreciated, discussing a few of the 'outstanding questions in papillomavirus biology' that the authors believe their new infection model can help address would help bolster and convey the importance of their findings.

3. It would be useful to more clearly indicate in the Results and Figure Legends which genetic background is being used in each experiment. It is unclear where C57BL/6 versus FVB mice were used.

4. Lines 69-72: while there are broad similarities between HPVs and MmuPV1 with respect to genome structure, encoded proteins, etc., they are actually quite different at the level of nucleotide sequence (HPV16 vs. MmuPV1 only ~ 50% similar). Likewise, HPV and MmuPV1 E6 and E7 proteins are less than 50% similar at an amino acid level (please refer to Joh et al., J Gen Virol, 2011). While MmuPV1 is certainly a useful tool for studying PV infection and pathogenesis, as is the focus in this report, this text overstates the similarities between HPV and MmuPV1 and should be amended to more accurately reflect the differences.

5. Lines 78-80: It may be worth noting here that subclinical MmuPV1 infections have been observed by several research groups at several anatomical sites. Including this information would strengthen the use of the MmuPV1 infection model to evaluate these understudied types of infection.

6. The images in some of the figures (particularly Figure 2B-C and 3A-B) are blurry and out of focus.

7. The PCR gel image in Figure 3E is very hard to see and is therefore not as convincing as it could be.

8. Line 88: Please provide rationale for infecting tail skin versus other sites of skin (dorsal, ear, etc).

9. Line 136: Please provide technical information regarding the isolation of MEFs or provide a citation.

10. Lines 292-293: The sentence "…only immunocompromised mice are susceptible to the mouse papilloma virus" is not accurate. There are several reports of immunocompetent strains of mice that are susceptible to MmuPV1 infection and disease. It may be more accurate to state that immunocompromised mice are more susceptible.

11. Reference issues:

a. Line 66: Please refer to the original paper that described the discovery of MmuPV1 (Ingle et al., Vet Pathol, 2011) instead of the review article currently cited.

b. Line 77: There is also an inducible mouse model of HPV oncogene expression (see Bottinger et al., Carcinogenesis 2020)

c. Line 77: There are certainly more novel infection models than the Wei et al., 2020 cited here that could be included to accurately reflect the state and scope of MmuPV1 infection models.

Reviewer #2 (Recommendations for the authors):

Some items for the authors consideration:

1. The figures include GFP labels for detection of fluorescence – this should be YFP based on expression in the transgenic mice.

2. Evidence for recircularization of MmuPV1 in this model appears to be indirect (viral RNA detection). Future studies could determine whether there is evidence of integration and production of infectious virions (containing circularized viral DNA).

3. Maybe correlate/discuss some of the findings in published RNAseq analyses of various other MmuPV1 (and HPV) infections (both in vitro raft culture studies and tissues).

4. Some typos: l-259 "resulting in YFP expression"; l-84 "MmuPV1"; l-481, Complete reference.

5. L-292 suggests that "only immunocompromised mice are susceptible to the mouse papillomavirus". However, it is clear that all mice are susceptible – it is just that most of the disease in such mice is subclinical and quickly cleared. Perhaps best to indicate that immunocompetent mice are resistant to clinical disease and/or persistence.

6. Interestingly, some of the early work by Campo's group with BPV and HPV identified E5 as a viral protein that down-regulated MHC1 (e.g. Oncogene. 2002 Jan 10;21(2):248-59. doi: 10.1038/sj.onc.1205008. Down-regulation of MHC class I by bovine papillomavirus E5 oncoproteins G Hossein Ashrafi, Emmanouella Tsirimonaki, Barbara Marchetti, Philippa M O'Brien, Gary J Sibbet, Linda Andrew, M Saveria Campo). MmuPV1 is noted to not have an E5 protein, however, other viral proteins impact MHC1 as shown here. Note also that Lamberts group conducted some interesting studies using an E5-expressing transgenic mouse with MmuPV1 infections (Virology. 2020 Feb;541:1-12. doi: 10.1016/j.virol.2019.12.002. Epub 2019 Dec 5. The human papillomavirus 16 E5 gene potentiates MmuPV1-Dependent pathogenesis Alexandra D Torres, Megan E Spurgeon, Andrea Bilger, Simon Blaine-Sauer, Aayushi Uberoi, Darya Buehler, Stephanie M McGregor, Ella Ward-Shaw, Paul F Lambert).

Reviewer #3 (Recommendations for the authors):

The introduction is a bit simplistic and could do with much more explanation to clarify for the general readership of the journal. For example, I don't agree that all HPV are commensal since they have an exquisitely tight relationship with the host epithelium and some are pathogenic. In this resepct it would be useful to discuss cutaneous versus mucosa PV infection since we now know of a huge number of commensal γ HPVs. HPVs colonise only epithelia, not "human tissues", which is too general a term. We understand how papillomaviruses enter, replicate and spread in epithelia so this aspect of their biology is understood. We do not know precisely how the life cycle is initiated from a single infected stem cell. So it is important to carefully discuss what is and is not known. Infection of the basal layer of the epithelium via tissue damage may not be necessary for infection: e.g. infection of single layer columnar cells e.g. in the endocervix is likely (see Doorbar and Griffin https://pubmed.ncbi.nlm.nih.gov/30974183/). HPV6/11 laryngeal papillomas and genital warts do show cytopathic effects.

Throughout, the paper is written in quite a superficial manner and lacks proper discussion of relevant literature. There are almost no points of discussion in the paper that are fully explored. Some of the figures/data are hard to see, or to decipher, and could be improved. There is a lack of breadth of data which means that the authors conclusions are not fully substantiated.

For example:

Figure 1 is hard to understand for the Cre non-specialist. A diagram of the plasmids that relate to the predicted band on the gel would be useful and well as annotation of the plasmid diagram in (A) with larger and clearer fonts for the primers. In (D) why is the E4 band the same size and the GAPDH band? Positions of the primers on the MmuPV1 genome should be given in a table. Why is the E6 band so faint compared to the L1 band when transcriptome analyses have shown that E6 is much more highly expressed than L1 (see https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1006715#). Since this is puzzling, these data should be backed up with qRT-PCR data.

UV irradiation of the mice may not affect the innate immune response but as this paper deals with validation of a model data showing the effects of UV on the adaptive and innate immune response in the mice is required.

In Figure 2B there is clear evidence of clonal expansion of MmuPV1 infected cells. However, I don't understand why K14, a basal cell marker is expressed in the upper epithelial layer, apparently in all tissues. Moreover, why is K14 staining colocalising with GFP at Day 3 (which could be explained by de-differentiation of the MmuPV1 cells) but not at subsequent days? It is expected that PV infected cells show decreased expression of differentiation markers such as K10 or involucrin and staining with at least these markers should be carried out to properly characterise the infection.

In Figure 3 the stainings for viral RNAs are quite indistinct. Perhaps removing haematoxylin staining and showing higher magnification would help. H and E staining for each tissue and +/- DNAse in the RNAScope would also be required. The bands for Cre loss and MmuPV1 DNA in Figure 3E are too faint to assess.

The investigation of MHC class I on the cell surface is interesting but the authors do not properly explain the background to this study, namely that HPV E5 is the main mediator of MHC downregulation. HPV E7 had been reported to also contribute. Of course, MmuPV1 does not express E5 but neither (to my knowledge) has MmuPV1 E7 been shown to regulate MHC class 1. The finding that cells surrounding the recombinant MmuPV1-containing cells is indicative, but more work would be necessary to firm up this potentially important conclusion.