Tumor Implantation Site of Syngeneic Oral Cancer Models Differentially Induces Site-Dependent Local and Systemic Immunosuppression

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Recommendation: Major revision

The manuscript presents interesting comparative immunological profiling; however, major concerns limit its rigor and translational relevance. The experimental design lacks sufficient justification for model selection and sample size, while statistical analyses appear underpowered and inconsistently applied. Critical controls for batch effects in spectral flow cytometry and validation of gating strategies are insufficiently described. Mechanistic interpretation of immunosuppressive pathways remains largely correlative without functional validation. Furthermore, conclusions overstate clinical relevance given species differences and limited longitudinal validation and reporting remains inadequate

1. Why was the ROC1 model (poorly immunogenic) selected instead of including a more immunogenic comparator (e.g., MOC1) to validate generalizability?
2. The hypothesis suggests tdLN-mediated immune dampening—where is the functional evidence directly linking tdLN changes to systemic immunosuppression?
3. Given that orthotopic models are considered more clinically relevant, why is there no validation against human HNSCC datasets?
4. The study focuses on immune composition—why were functional tumor control assays (e.g., response to immunotherapy) not included?
5. The study relies heavily on spectral flow cytometry—how were batch effects and panel drift controlled across timepoints?
6. Tumor size differs between models at endpoints—how do you control for size-dependent immune variation as a confounder?
7. Multiple comparisons are performed extensively—how was false discovery rate controlled beyond Tukey correction?
8. The increase in CD4⁺ T-cells in heterotopic tumors is highlighted—why is functional relevance (e.g., Th1 vs Treg) not dissected?
9. The study reports increased PD-1 and PD-L1 expression—why were checkpoint blockade experiments not performed to validate functional relevance?
10. How do you explain the divergence between local (tumor) and systemic (blood/spleen) immune responses mechanistically?

Author Response

1. Why was the ROC1 model (poorly immunogenic) selected instead of including a more immunogenic comparator (e.g., MOC1) to validate generalizability?
   

   Thank you for your feedback, we have included a few sentences explaining why we utilized ROC1 over other models. “mEER and MOC models have been well-characterized, including exploration of tumor site-specific differences (1-3). However, existing models, although frequently utilized, do not fully recapitulate the heterogeneity of head and neck cancer and the human smoking signature. Furthermore, study of poorly immunogenic models is critical because these are of greatest relevance to the immunotherapy-resistant microenvironment of most head and neck cancers. In this study, we utilized an HPV-negative, poorly-immunogenic oral cavity cancer cell line, ROC1. ROC1 is a carcinogen-derived oral cancer cell line that has tobacco mutation signatures relevant to the study of head and neck cancers (23). The ROC1 line was generated by exposing mice to the carcinogen 4-nitroquinoline-1 oxide (4-NQO), which is reflective of the human smoking signature, and is resistant to traditional immunotherapy (23, 24).”

   1. Dorta-Estremera S, Hegde VL, Slay RB, Sun R, Yanamandra AV, Nicholas C, et al. Targeting interferon signaling and CTLA-4 enhance the therapeutic efficacy of anti-PD-1 immunotherapy in preclinical model of HPV(+) oral cancer. J Immunother Cancer. 2019;7(1):252.
   2. Kansal V, Kinney BLC, Schmitt NC. Characterization of the tumor microenvironment in the mouse oral cancer (MOC1) model after orthotopic implantation in the buccal mucosa. Head Neck. 2024;46(5):1056-62.
   3. Kono M, Saito S, Rokugo M, Egloff AM, Uppaluri R. Enhanced oral versus flank lymph node T cell response parallels anti-PD1 efficacy in head and neck cancer. Oral Oncol. 2024;152:106795.
   4. Shi Y, Xie T, Wang B, Wang R, Cai Y, Yuan B, et al. Mutant p53 drives an immune cold tumor immune microenvironment in oral squamous cell carcinoma. Commun Biol. 2022;5(1):757.
   5. Shi Y, Xie TX, Leach DG, Wang B, Young S, Osman AA, et al. Local Anti-PD-1 Delivery Prevents Progression of Premalignant Lesions in a 4NQO-Oral Carcinogenesis Mouse Model. Cancer Prev Res (Phila). 2021;14(8):767-78.
      
      
2. The hypothesis suggests tdLN-mediated immune dampening—where is the functional evidence directly linking tdLN changes to systemic immunosuppression? 

   Thank you for your comment, we have rephrased the hypothesis from “We thus hypothesized that there are location-based TIME differences between implantation site and results in a dampened immune response through the tdLN. These differences may be further extended to the cells present in peripheral blood and spleen.” to “We thus hypothesized that there are location-based TIME differences between implantation site and results in a dampened immune response that are extended to the tdLN, peripheral blood and spleen.”
   
   

3. Given that orthotopic models are considered more clinically relevant, why is there no validation against human HNSCC datasets?

   Thank you for your feedback, while we agree with the reviewer that comparing with human HNSCC datasets would provide more clinical relevance, it is beyond the scope of the paper. In this manuscript, we compare the immune microenvironment of preclinical tumor implantation sites for consideration of model selection. In preclinical studies, there is a shift to utilizing orthotopic models for clinical relevance^1-3, therefore we wanted to compare between orthotopic and heterotopic sites to determine if there were differences between the immune composition of ROC1 tumors.

   1. Dorta-Estremera S, Hegde VL, Slay RB, Sun R, Yanamandra AV, Nicholas C, et al. Targeting interferon signaling and CTLA-4 enhance the therapeutic efficacy of anti-PD-1 immunotherapy in preclinical model of HPV(+) oral cancer. J Immunother Cancer. 2019;7(1):252.
   2. Kansal V, Kinney BLC, Schmitt NC. Characterization of the tumor microenvironment in the mouse oral cancer (MOC1) model after orthotopic implantation in the buccal mucosa. Head Neck. 2024;46(5):1056-62.
   3. Kono M, Saito S, Rokugo M, Egloff AM, Uppaluri R. Enhanced oral versus flank lymph node T cell response parallels anti-PD1 efficacy in head and neck cancer. Oral Oncol. 2024;152:106795.
      
      
4. The study focuses on immune composition—why were functional tumor control assays (e.g., response to immunotherapy) not included?

   Thank you for your comment, while we agree implantation site immunotherapy responses would provide additional tumor context, our focus was comprehensively profiling this new head and neck model in the orthotopic and heterotopic sites. We have previously published immunotherapy responses utilizing the ROC1 orthotopic site with a biomaterial system.¹

   1. Swain JWR, Molina AH, Sunga GM, Chew-Martinez D, Dharmaraj N, Perez AC, et al. A multidomain peptide-liposome composite for controlled release of a cyclic dinucleotide in oral cancer. J Control Release. 2026:114730.
      
      
5. The study relies heavily on spectral flow cytometry—how were batch effects and panel drift controlled across timepoints?

   Thank you for your feedback, we controlled batch effects and panel drift across timepoints by utilizing the Flow Cytometry and Cellular Imaging Core Facility for the Cytek Aurora equipment maintance and quality control, sample processing remains consistent, running single-stained controls, and prepare master mix of titrated antibody that are consistent across the runs. When preparing for the 28-marker panel, we ran FMO-controls to confirm the gating strategy.
   
   

6. Tumor size differs between models at endpoints—how do you control for size-dependent immune variation as a confounder?

   Thank you for your comment. We controlled for size by directly comparing tumors of the same tumor area, we did not analyze tumors at endpoints.
   
   

7. Multiple comparisons are performed extensively—how was false discovery rate controlled beyond Tukey correction?

   Thank you for your feedback, the Benjamini, Krieger, Yekutieli method was also done to control the false discovery rate in multiple comparison correction. This analysis resulted in discoveries (based on corrected p-values, or “q-values”) in the same populations as that done with the Tukey correction and, therefore, does not change the significance or the way the data is interpreted.
   
   

8. The increase in CD4⁺ T-cells in heterotopic tumors is highlighted—why is functional relevance (e.g., Th1 vs Treg) not dissected?

   Thank you for your comment. Since this was the first immune profiling reported in both models, we focused on general immune cell populations, with a focus on cytotoxic and suppressive cell types. Our current analysis does include investigation of T-regs, but we did not include specific markers to dissect further subpopulations, such as T-helper cell types. Additionally, our tSNE analysis requires downsampling of each sample. Due to low frequencies, we were unable to correctly determine T-regs via tSNE and, therefore, did not emphasize this population.
   
   

9. The study reports increased PD-1 and PD-L1 expression—why were checkpoint blockade experiments not performed to validate functional relevance?

   Thank you for your feedback, while we agree checkpoint blockade experiments would validate functional relevance, our focus was comprehensively profiling this new head and neck model in the orthotopic and heterotopic sites. We have also previously published anti-PD-1 and anti-CTLA-4 treatments combined with STING agonist utilizing the ROC1 orthotopic site with a biomaterial system.¹

   1. Swain JWR, Molina AH, Sunga GM, Chew-Martinez D, Dharmaraj N, Perez AC, et al. A multidomain peptide-liposome composite for controlled release of a cyclic dinucleotide in oral cancer. J Control Release. 2026:114730.
      
      
10. How do you explain the divergence between local (tumor) and systemic (blood/spleen) immune responses mechanistically?

    Thank you for your comment, in our manuscript, we have describe several mechanisms throughout our Discussion (Section 4) that may result in the divergence between tumor (lines 549-624) and peripheral or systemic (lines 625-674) immune responses, such as the interplay between activation and exhaustion cell states and the upregulation of trafficking markers.

please also view the attached response

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This study performed characterization of tumor growth, mice survival and microenvironment in a mice model with ROC1 cell line, which is a  mouse oral squamous cell carcinoma cell line due to chemical damage similar to tabacco. The study used two types of mice, wild type c57bl6 and rag knockout. However, it is not clear in the results the number of mice per group and if results are of wild type or rag1-/-. It may be my mistake in reading the paper. Please confirm.

Additional comments:

(1) Line 121. Please describe the vitamin solution.

(2) Line 125. Could you please describe the properties of  wild-type C57BL/6 and Rag1-/- mice?

(3) Line 127. Please describe "digital calipers".

(4) Line 132. What is the criteria of deciding that endonpoint?

(5) What version of FlowJo was used?

(6) Please add the antibody list and catalog numbers of the flow cytometry antibodies.

(7) Please add version of GraphPad software.

(8) Line 65. Regarding "immunotherapy". Please define as immune therapy is a broad term.

(9) Line 95. Please describe the meaning of "MOC1" with more detail. Mouse oral squamous cell carcinoma cell line 1?

(10) As I understand ROC1 cell line is mutagenic cell line because of chemical reaction derived from mouse cell? What type of cell? Epithelial?

(11) Please describe the total number of mice per each group.

(12) As I understand both tumor area and survival was the same for heterotopic and orthotopic (figures 1a-b). Is this right? Please confirm.

(13) Data shows differences between orthotopic and heterotopic. Could you please specify for each figure what type of mice was used in each experiment (control wild-type or RAG1-/-). Of note, Mice homozygous for the Rag1tm1Mom mutation produce no mature T cells or B cells. Their phenotype can be described as a "non-leaky" immune deficiency. 

(14) In section 5, I understand that the conclusion is that location of tumor does not matter. However, I understand that you refere to wild-type mice?

Author Response

1. Line 121. Please describe the vitamin solution.

   Thank you for your feedback. The vitamin solution used is “MEM vitamins solution (1X)”. This was added to manuscript line 122.
   
   

2. Line 125. Could you please describe the properties of  wild-type C57BL/6 and Rag1-/- mice?

   Thank you for your feedback. We further described the difference between C57BL/6 and Rag1-/- in line 126-127: “wild-type immunocompetent C57BL/6 or Rag1-/- non-leaky immune deficient mice, lacking both mature T and B cells”.
   
   

3. Line 127. Please describe "digital calipers".

   Thank you for your comment. We use battery-operated digital calipers from Fisherbrand, we added this into the manuscript in line 129.
   
   

4. Line 132. What is the criteria of deciding that endonpoint?

   Thank you for your feedback. We utilize the body conditioning score to determine whether the mice reach euthanasia criteria from their tumor size (mm), degree of necrosis, hydration, mobility, weight loss or gain, and overall appearance per our respective protocols from the Animal Welfare Committee (AWC-21-0059 approval date: 06/25/2021, AWC-24-0029 approval date: 04/26/2024), and the Center for Laboratory Animal Medicine and Care at The University of Texas Health Science Center at Houston and The University of Texas MD Anderson Cancer Center’s Institutional Animal Care and Use Committee (00002336-RN01 approval date: 01/17/2023).
   
   

5. What version of FlowJo was used?

   Thank you for your comment, we used FlowJo 10.10.0. We have updated the manuscript in line 156.
   
   

6. Please add the antibody list and catalog numbers of the flow cytometry antibodies.

   Thank you for your feedback. The antibody list, catalog numbers, and respective vendors are provided in the Supplemental Table 1.
   
   

7. Please add version of GraphPad software.

   Thank you for your comment. We used GraphPad Prism 10.6.1 software, we added this to the manuscript in line 171.
   
   

8. Line 65. Regarding "immunotherapy". Please define as immune therapy is a broad term.

   Thank you for your feedback, we added more definition to immunotherapy from line 66-69, “Immunotherapy is a type of cancer treatment that can program immune cells to target and recognize cancer cells in the primary tumor and the tumor-draining lymph nodes (tdLNs), thereby boosting immunological memory and reducing cancer recurrence.”
   
   

9. Line 95. Please describe the meaning of "MOC1" with more detail. Mouse oral squamous cell carcinoma cell line 1?

   Thank you for your comment, we added “mouse oral squamous cell carcinoma cell line 1” in line 96.
   
   

10. As I understand ROC1 cell line is mutagenic cell line because of chemical reaction derived from mouse cell? What type of cell? Epithelial?

    Thank you for your feedback, ROC1 was developed by our collaborator’s group Dr. Roberto Rangel. They generated syngeneic oral cavity squamous cell carcinoma models with defined Trp53 (p53) mutations and characterized their TIMEs. Their immunohistochemistry analysis using mesenchymal and epithelial markers revealed low levels of keratin-14 and high levels of vimentin expressed in ROC1.¹

    1. Shi Y, Xie T, Wang B, Wang R, Cai Y, Yuan B, Gleber-Netto FO, Tian X, Rodriguez-Rosario AE, Osman AA, Wang J, Pickering CR, Ren X, Sikora AG, Myers JN, Rangel R. Mutant p53 drives an immune cold tumor immune microenvironment in oral squamous cell carcinoma. Commun Biol. 2022 Jul 28;5(1):757. doi: 10.1038/s42003-022-03675-4. PMID: 35902768; PMCID: PMC9334280.
       
       
11. Please describe the total number of mice per each group.

    Thank you for your comment. The total number of mice per group was described in each figure legend as lowercase n (ex. n=10/timepoint).
    
    

12. As I understand both tumor area and survival was the same for heterotopic and orthotopic (figures 1a-b). Is this right? Please confirm.

    Thank you for your feedback. The tumor area and survival was similar for the heterotopic and orthotopic models.
    

13. Data shows differences between orthotopic and heterotopic. Could you please specify for each figure what type of mice was used in each experiment (control wild-type or RAG1-/-). Of note, Mice homozygous for the Rag1tm1Mom mutation produce no mature T cells or B cells. Their phenotype can be described as a "non-leaky" immune deficiency. 

    Thank you for your comment, the data from the main manuscript are all from wild-type C57BL/6 mice. The only Rag1-/- mice are from Figure S2. We clarified this in the figure legends of each of the main and supplemental figures.
    
    

14. In section 5, I understand that the conclusion is that location of tumor does not matter. However, I understand that you refere to wild-type mice?

    Thank you for your feedback. We have concluded that there are no differences between mean tumor growth and the survival from tumor-bearing wild-type C57BL/6 mice. In terms of the tumor immune microenvironment, tumor-draining lymph nodes, blood and spleen there are differences between the implantation site of the tumor from tumor-bearing wild-type C57BL/6 mice. These differences may need to be considered when selecting preclinical models.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Accept in present form