The ligandome describes the peptide composition bound to the major histocompatibility complex (MHC) I and II presented on the cellular surface (Freudenmann et al., 2018). Once being identified as targets by the immune system, the peptides in the MHC-I are the contact point between cytotoxic CD8+ T cells and the tumor cells. Thus, the knowledge of those peptides is a key point in designing therapeutic cancer vaccines to generate and stimulate specific antitumor adaptive immune responses. Moreover, the interest in identifying and exploiting these targets gained momentum following the breakthrough of the immune checkpoint inhibitors (ICIs) as it became clear that ICI treatment can unleash the specific antitumor T-cell responses against these immunogenic candidate targets.(Bassani-Sternberg et al., 2016). Indeed, the ICI therapy activates a preexisting antitumor immune response with immune cell infiltration in the cancer lesions, defined as ‘hot’ tumors; instead, tumors not infiltrated with immune cells are called ‘cold.’ As a result, the response rate to the ICI therapy can vary from 40% to 70% to 10% to 25% either due to the lack of immune cell infiltration into the tumor or other immunosuppressive mechanisms in the tumor microenvironment (TME) (Schoenfeld and Hellmann, 2020; Wieder et al., 2018). Currently, there is an urgent need to find a way to turn ‘cold’ tumors to ‘hot’ ones, making the ICI therapies more effective. In this context, the development of effective peptide-based cancer vaccines for therapeutic approaches is facing two main challenges: the criteria to select peptides able to elicit an immune response and the use of an adjuvant to increase the antitumor immune response of the immunizing peptides. In this article, to overcome these issues, we have developed a pipeline that covers the diverse developmental stages of therapeutic cancer vaccines, moving from the isolation of the MHC-I-restricted tumor peptides, to the selection and screening of target candidates until the generation of an oncolytic cancer vaccine. First, we selected the known murine immunogenic tumor model CT26, allowing the study of the antitumor response (Lechner et al., 2013). We investigated the MHC-I antigen landscape of CT26 applying state-of-the-art immunopeptidome and mass spectrometric methodologies. The immunopeptidome profile was carefully analyzed and found to be qualitatively in line with already published dataset; the result list of peptides was then investigated through two approaches: RNAseq and HEX software. The latter is a tool that identifies tumor antigens similar to pathogen antigens, exploiting the cross-mimicry and cross-reactive T cells for clinical applications (Jacopo, 2020). The peptides derived from those analyses were then investigated in vivo by pre-immunizing mice with the adjuvant poly:(IC) and the peptides; the splenocytes were then harvested and functional characterization was performed by interferon-γ enzyme-linked immunospot (ELISpot), deconvoluting the single-peptide immunogenicity. For the last part of our pipeline, after the functional characterization, the selected peptides were used to generate an oncolytic cancer vaccine. To take full advantage of viral immunogenicity to induce a specific antitumor T-cell response, we used our previously developed platform, PeptiCRAd, based on an oncolytic adenovirus (OAd) coated with immunogenic tumor antigen peptides (Capasso et al., 2016; Feola et al., 2018). The peptide candidates in this study were tested in our PeptiCRAd platform, which here consisted of a conditionally replicating OAd armed with two immune-activating ligands, the ligand for cluster of differentiation 40 (CD40L) and the ligand for tumor necrosis factor receptor superfamily member 4 (OX40L), named VALO-mD901 (Ylösmäki et al., 2021). Intratumoral administration of PeptiCRAd coated with the peptides selected based on our pipeline controlled the tumor growth in CT26 tumor-bearing mice. Additionally, we observed that the specific antitumor immune activation generated in the primary tumor could be extended to a second tumor lesion in a phenomenon known as ‘abscopal effect.’ Thus, we developed and validated a pipeline moving from the isolation of the peptides to the selection of the target candidates until the combination of these in our PeptiCRAd platform, showing the efficacy in a preclinical model of colon cancer on to the primary tumor and distant lesions.

To the best of our knowledge, the described pipeline covers for the first time all the stages of a personalized therapeutic cancer vaccine development, starting from the isolation of MHC-I-restricted peptides derived from the primary tumor to their analysis in silico and in vivo to identify the best target candidates. Finally, an OAd was coated with these peptides to generate an effective therapeutic cancer vaccine. The pipeline can be translated to personalized cancer treatment in relevant clinical application as the OAd can be easily coated with the unique repertoire of patient-specific tumor peptides profile, a prerequisite for personalized therapy.

Cytotoxic antitumor CD8+ T cells (CTLs) recognize peptides typically of 8–10 amino acids within the MHC-I complex expressed on the cellular surface, and therefore, the knowledge of these peptides is the key to design T-cell-based therapeutic cancer vaccines; indeed, their efficacy relies mostly on the choice of the antigenic peptides (Hollingsworth and Jansen, 2019). These peptides should be highly immunogenic, expressed exclusively on the cancer cells to avoid on-target off-tumor toxicity and tailored to the patient’s specific tumor ligandome landscape. However, only a fewer if any of the tumor antigens meet those characteristics, making it very difficult to generate peptide-based vaccination technologies. Thus, the isolation and identification of MHC-I peptides and the subsequent selection criteria are of utmost importance in creating those vaccines. To fulfill these needs, we conceived a pipeline that comprises all the steps considered essential for an optimal development of a therapeutic cancer vaccine.

We decided to identify and isolate peptides directly from the MHC-I complexes, exploiting state-of-the-art immunoprecipitation and mass spectrometric methodologies as the direct elution and analysis of MHC-I-restricted peptides is so far the most reliable and used approach in studies of ligandome landscape (Freudenmann et al., 2018), identifying naturally processed and presented tumor epitopes that could generate clinically relevant antitumor responses. Even though computational algorithms can take into account the entire MHC complex presentation machinery (e.g., proteasomal cleavage, transporter-associated antigen processing [TAP] transport, binding motif) to predict relevant T-cell epitopes, the lack of validated and homogenous datasets makes the process difficult and less reproducible (Feola et al., 2020; Soria-Guerra et al., 2015). These considerations prompted us to adopt direct MHC-I immunoaffinity purification as a first step of our pipeline.

Moreover, to develop and further validate our proof-of-concept pipeline, the choice of the tumor model needed to meet specific requirements. First, we wanted a tumor model that expresses sufficient levels of MHC-I complexes, granting a fruitful recovery of peptides from the cellular surface. Indeed, the overall idea was to obtain a conspicuous list of peptides in order to later on challenge our prioritizing and selection criteria. Secondly, to test the selected candidates for their antitumoral efficacy profile, the preclinical model should have beneficial immunogenic features, in particular T-cell infiltration into the TME, allowing a better study of the immune modulation upon treatment administration. Based on that, the colon tumor model CT26 was selected as it showed high expression level of MHC-I complexes as demonstrated in our flow cytometry analysis and for being a widely used and characterized tumor model for developing and testing immunotherapeutic concepts in vivo (Taylor et al., 2019; Castle et al., 2014). As expected, the immunoaffinity purification generated a long list of peptides, containing more than 8000. Before moving forward in our pipeline, we carefully analyzed the quality of the produced dataset to ensure the solidity of our list and examine it for the presence of contaminants. The analysis demonstrated that the eluted peptides resembled a typical ligandome profile and therefore they could be considered as true MHC-I ligands. The strength and reliability of the ligandome dataset are critically important for the following steps as they influence the subsequent results.

Of note, besides the identification of MHC-I peptides, the main issue is dealing with the prioritization of the peptides among thousands of possible candidates. In this context, we followed two parallel directions. First, we adopted a more conservative approach that consisted of analyzing the RNAseq expression level of the respective source proteins. In particular, based on the definition of TAA as an antigen overexpressed in malignant cells compared to healthy tissue, we considered the colon Balb/c as the reference normal tissue since CT26 is an undifferentiated colon carcinoma induced by the carcinogen N-nitroso-N-methylurethane (NMU) (Castle et al., 2014). In this sense, the selected peptides used as therapeutic cancer vaccine should evoke specific antitumor CTLs able to recognize and eradicate tumor cells, avoiding damages to normal colon tissue. Moreover, syngeneic mTECs express most of the known genes, and they are the site of T-cell selection to induce central tolerance to MHC peptides coded by their vast transcriptome. We assumed that the breaking of tolerance could most likely happen if the source proteins of the selected peptides from our dataset were overrepresented in the CT26 cell line compared to the mTEC. To ensure a more accurate selection of the candidates, we focused the choice on peptides that meet both criteria of (1) source protein overexpressed compared to normal colon and mTEC and (2) of being true MHC-I strong ligands. The second parallel approach represents the main novelty introduced in our pipeline, and it consisted of selecting peptides based on their similarity to antigen pathogen by exploiting HEX, a tool previously developed in our laboratory and successfully validated both in preclinical and clinical settings (Jacopo, 2020). The main idea relies on the intrinsic degeneracy of the TCR, defined as the ability of a single TCR to recognize more than one antigen, generating a phenomenon known as cross-reactivity. This property is an essential feature to broaden the breadth of the T-cell repertoire, and, for instance, it allows antiviral memory CD8+ T cells generated by prior infections to recognize unrelated viruses, as demonstrated in several studies in human and murine models (Welsh and Selin, 2002; Brehm et al., 2004). We thought that the same concept could be applied to cancer antigens that have similarities to viral antigens. We are aware that in this work we used mice naïve to viral infections, and therefore, no memory CD8+ T-cell cross-reactivity could be exploited. However, translated into a real clinical setting, our approach will have the added value of exploiting the cross-reactivity of preexisting viral CD8+ T cells to enhance the antitumor response. Applying the aforementioned in silico analysis, the number of candidates was shortened, making it feasible to further functionally characterize the list of the peptides in an ELISpot assay.

After the selection of candidates to exploit in a vaccine platform was done, we employed the peptides in our previously developed platform named PeptiCRAd, an oncolytic adenovirus coated with polyK-modified peptides (Capasso et al., 2016; Feola et al., 2018; Ylösmäki et al., 2021; Tähtinen et al., 2020). Indeed, after the FDA approval of T-VEC, a herpes virus encoding GM-CSF (Ribas et al., 2018) for the treatment of melanoma, the use of oncolytic viruses (OVs) has been extensively explored in cancer immunotherapy (Kaufman et al., 2016; Harrington et al., 2019; Macedo et al., 2020; Kottke et al., 2021). OVs are naturally occurring or genetically modified viruses able to infect and replicate in cancer cells; the OVs induce a systematic immune response, involving both innate and adaptive immune response. Moreover, the antigen spread following the viral burst acts as in situ cancer vaccine but is often not enough to generate a specific antitumor adaptive immune response, instead generating mainly an antiviral T-cell responses (Ylösmäki and Cerullo, 2020). To overcome this limitation, we decided to combine the immunogenicity of the OVs with the antitumor specificity of the peptides, generating an oncolytic cancer vaccine. Thus, to challenge our pipeline and investigate whether our selection criteria could actually be used to identify relevant candidates for cancer treatment, we decorated the OAd VALO-mD901 with the selected peptides to treat immunocompetent CT26 tumor-bearing mice. To understand whether our technology could actually evoke a systemic antitumor immune response, we engrafted two tumors for each mouse, both right and left flanks, and then treated only the tumor on the right flank. Of note, VALO-mD901 is encoding murine CD40L and OX40L under cytomegalovirus (CMV) promoter allowing transgene expression in murine cells. Stimulation of innate (due to CD40L) and adaptive (due to OX40L) immune cells explained the local antitumor activity in virus-injected tumors observed in our results and the lack of efficacy in the distant lesions. Contrarily, PeptiCRAd 1 (virus coated with peptide 1) treatment slowed the tumor growth of both the treated and untreated lesions, highlighting the generation of a systemic tumor-specific immune response. Additionally, in order to benchmark our pipeline, we employed as positive control peptide 7, an immunodominant peptide in CT26 cell line named AH1 (gp70_423–431); however, AH1 elicits variable antitumor response because T cells are suboptimally activated. To this end, we employed the heterolytic version named AH1-5, which differs from the AH1 peptide in a single amino acid (val to ala in position 5) that has been reported to protect mice against subsequent tumor challenge with CT26 when employed in a prophylactic setting (Jordan et al., 2010). Based on that and in the absence of other positive candidate controls, we reasoned that AH1-5 could be used in our work as an internal control. Nevertheless, in our work, we used a therapeutic approach treating well-established tumors instead of a prophylactic setting as showed in Jordan et al., 2010, and in conclusion our data showed that AH1-5 was actually ineffective against well-established tumors.

The overall results demonstrated the feasibility of applying the described pipeline for the generation of a tailored therapeutic cancer vaccine. We have addressed all the main issues universally recognized as challenges in the field with main focus on the prioritization and selection criteria among thousands of peptide candidates. Additionally, we adapted quick ‘plug-and-play’ technology based on decorating an OV with the selected peptides. The nature of this technology opens the possibility of a fast generation of tailored therapeutic cancer vaccines in future clinical applications where personalized therapies represent one of the main goals for a successful treatment. From a clinical application point of view, the integration of the ligandome and transcriptome analysis could benefit from the fast selection of peptides done with the HEX software. Indeed, recent data suggest that MHC-I-restricted peptides homologous to viral peptides are strongly immunogenic and offer a reliable source of candidates for cancer vaccine design. Our approach will capitalize on preexisting cross-reactive T cells (Zitvogel et al., 2021; Fluckiger et al., 2020), facilitating the peptide selection.

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD026463.

1. 1. Feola S
   2. Chiaro J
   3. Martins B
   4. Russo S
   5. Fusciello M
   6. Ylösmäki E
   7. Bonini C
   8. Ruggiero E
   9. Hamdan F
   10. Feodoroff M
   11. Antignani G
   12. Viitala T
   13. Pesonen S
   14. Grönholm M
   15. Branca RMM
   16. Lehtiö J
   17. Cerullo V

   (2021) PRIDE

   ID PXD026463. A novel immunopeptidomic-based pipeline for the generation of personalized oncolytic cancer vaccines.

   https://www.ebi.ac.uk/pride/archive/projects/PXD026463

1. 1. Laumont CM
   2. Vincent K
   3. Hesnard L
   4. Audemard E
   5. Bonneil E
   6. Laverdure JP

   (2018) NCBI Gene Expression Omnibus

   ID GSE111092. Noncoding regions are the main source of targetable tumor-specific antigens.

   https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE111092
2. 1. Proquin H
   2. Jetten MJ
   3. Jonkhout MCM
   4. Garduño-Balderas LG

   (2018) NCBI Gene Expression Omnibus

   ID GSE92563. Transcriptomics effects of ingestion of food additive titanium dioxide (E171) in the colon of BALB/c mice.

   https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE92563

1. 1. Adamopoulou E
   2. Tenzer S
   3. Hillen N
   4. Klug P
   5. Rota IA
   6. Tietz S
   7. Gebhardt M
   8. Stevanovic S
   9. Schild H
   10. Tolosa E
   11. Melms A
   12. Stoeckle C

   (2013) Exploring the MHC-peptide matrix of central tolerance in the human thymus

   Nature Communications 4:2039.

   https://doi.org/10.1038/ncomms3039

   • PubMed
   • Google Scholar
2. 1. Andreatta M
   2. Lund O
   3. Nielsen M

   (2013) Simultaneous alignment and clustering of peptide data using a Gibbs sampling approach

   Bioinformatics (Oxford, England) 29:8–14.

   https://doi.org/10.1093/bioinformatics/bts621

   • PubMed
   • Google Scholar
3. 1. Andreatta M
   2. Nielsen M

   (2016) Gapped sequence alignment using artificial neural networks: application to the MHC class I system

   Bioinformatics (Oxford, England) 32:511–517.

   https://doi.org/10.1093/bioinformatics/btv639

   • PubMed
   • Google Scholar
4. 1. Andreatta M
   2. Alvarez B
   3. Nielsen M

   (2017) GibbsCluster: unsupervised clustering and alignment of peptide sequences

   Nucleic Acids Research 45:W458–W463.

   https://doi.org/10.1093/nar/gkx248

   • PubMed
   • Google Scholar
5. 1. Bassani-Sternberg M
   2. Bräunlein E
   3. Klar R
   4. Engleitner T
   5. Sinitcyn P
   6. Audehm S
   7. Straub M
   8. Weber J
   9. Slotta-Huspenina J
   10. Specht K
   11. Martignoni ME
   12. Werner A
   13. Hein R
   14. H Busch D
   15. Peschel C
   16. Rad R
   17. Cox J
   18. Mann M
   19. Krackhardt AM

   (2016) Direct identification of clinically relevant neoepitopes presented on native human melanoma tissue by mass spectrometry

   Nature Communications 7:13404.

   https://doi.org/10.1038/ncomms13404

   • PubMed
   • Google Scholar
6. 1. Bassani-Sternberg M

   (2018) Mass Spectrometry Based Immunopeptidomics for the Discovery of Cancer Neoantigens

   Methods in Molecular Biology (Clifton, N.J.) 1719:209–221.

   https://doi.org/10.1007/978-1-4939-7537-2_14

   • PubMed
   • Google Scholar
7. 1. Brehm MA
   2. Selin LK
   3. Welsh RM

   (2004) CD8 T cell responses to viral infections in sequence

   Cellular Microbiology 6:411–421.

   https://doi.org/10.1111/j.1462-5822.2004.00390.x

   • PubMed
   • Google Scholar
8. 1. Capasso C
   2. Hirvinen M
   3. Garofalo M
   4. Romaniuk D
   5. Kuryk L
   6. Sarvela T
   7. Vitale A
   8. Antopolsky M
   9. Magarkar A
   10. Viitala T
   11. Suutari T
   12. Bunker A
   13. Yliperttula M
   14. Urtti A
   15. Cerullo V

   (2016) Oncolytic adenoviruses coated with MHC-I tumor epitopes increase the antitumor immunity and efficacy against melanoma

   Oncoimmunology 5:e1105429.

   https://doi.org/10.1080/2162402X.2015.1105429

   • PubMed
   • Google Scholar
9. 1. Castle JC
   2. Loewer M
   3. Boegel S
   4. de Graaf J
   5. Bender C
   6. Tadmor AD
   7. Boisguerin V
   8. Bukur T
   9. Sorn P
   10. Paret C
   11. Diken M
   12. Kreiter S
   13. Türeci Ö
   14. Sahin U

   (2014) Immunomic, genomic and transcriptomic characterization of CT26 colorectal carcinoma

   BMC Genomics 15:190.

   https://doi.org/10.1186/1471-2164-15-190

   • PubMed
   • Google Scholar
10. 1. Feola S
    2. Capasso C
    3. Fusciello M
    4. Martins B
    5. Tähtinen S
    6. Medeot M
    7. Carpi S
    8. Frascaro F
    9. Ylosmäki E
    10. Peltonen K
    11. Pastore L
    12. Cerullo V

    (2018) Oncolytic vaccines increase the response to PD-L1 blockade in immunogenic and poorly immunogenic tumors

    Oncoimmunology 7:e1457596.

    https://doi.org/10.1080/2162402X.2018.1457596

    • PubMed
    • Google Scholar
11. 1. Feola S
    2. Chiaro J
    3. Martins B
    4. Cerullo V

    (2020) Uncovering the Tumor Antigen Landscape: What to Know about the Discovery Process

    Cancers 12:E1660.

    https://doi.org/10.3390/cancers12061660

    • PubMed
    • Google Scholar
12. 1. Fluckiger A
    2. Daillère R
    3. Sassi M
    4. Sixt BS
    5. Liu P
    6. Loos F
    7. Richard C
    8. Rabu C
    9. Alou MT
    10. Goubet A-G
    11. Lemaitre F
    12. Ferrere G
    13. Derosa L
    14. Duong CPM
    15. Messaoudene M
    16. Gagné A
    17. Joubert P
    18. De Sordi L
    19. Debarbieux L
    20. Simon S
    21. Scarlata C-M
    22. Ayyoub M
    23. Palermo B
    24. Facciolo F
    25. Boidot R
    26. Wheeler R
    27. Boneca IG
    28. Sztupinszki Z
    29. Papp K
    30. Csabai I
    31. Pasolli E
    32. Segata N
    33. Lopez-Otin C
    34. Szallasi Z
    35. Andre F
    36. Iebba V
    37. Quiniou V
    38. Klatzmann D
    39. Boukhalil J
    40. Khelaifia S
    41. Raoult D
    42. Albiges L
    43. Escudier B
    44. Eggermont A
    45. Mami-Chouaib F
    46. Nistico P
    47. Ghiringhelli F
    48. Routy B
    49. Labarrière N
    50. Cattoir V
    51. Kroemer G
    52. Zitvogel L

    (2020) Cross-reactivity between tumor MHC class I-restricted antigens and an enterococcal bacteriophage

    Science (New York, N.Y.) 369:936–942.

    https://doi.org/10.1126/science.aax0701

    • PubMed
    • Google Scholar
13. 1. Freudenmann LK
    2. Marcu A
    3. Stevanović S

    (2018) Mapping the tumour human leukocyte antigen (HLA) ligandome by mass spectrometry

    Immunology 154:331–345.

    https://doi.org/10.1111/imm.12936

    • PubMed
    • Google Scholar
14. 1. Gattinoni L
    2. Speiser DE
    3. Lichterfeld M
    4. Bonini C

    (2017) T memory stem cells in health and disease

    Nature Medicine 23:18–27.

    https://doi.org/10.1038/nm.4241

    • PubMed
    • Google Scholar
15. 1. Harrington K
    2. Freeman DJ
    3. Kelly B
    4. Harper J
    5. Soria JC

    (2019) Optimizing oncolytic virotherapy in cancer treatment

    Nature Reviews. Drug Discovery 18:689–706.

    https://doi.org/10.1038/s41573-019-0029-0

    • PubMed
    • Google Scholar
16. 1. Hollingsworth RE
    2. Jansen K

    (2019) Turning the corner on therapeutic cancer vaccines

    NPJ Vaccines 4:7.

    https://doi.org/10.1038/s41541-019-0103-y

    • PubMed
    • Google Scholar
17. Preprint

    1. Jacopo H

    (2020) Viral Molecular Mimicry Influences the Antitumor Immune Response in Murine and Human Melanoma

    medRxiv.

    https://doi.org/10.1101/2020.09.09.20191171

    • Google Scholar
18. 1. Jordan KR
    2. McMahan RH
    3. Kemmler CB
    4. Kappler JW
    5. Slansky JE

    (2010) Peptide vaccines prevent tumor growth by activating T cells that respond to native tumor antigens

    PNAS 107:4652–4657.

    https://doi.org/10.1073/pnas.0914879107

    • PubMed
    • Google Scholar
19. 1. Kallio MA
    2. Tuimala JT
    3. Hupponen T
    4. Klemelä P
    5. Gentile M
    6. Scheinin I
    7. Koski M
    8. Käki J
    9. Korpelainen EI

    (2011) Chipster: user-friendly analysis software for microarray and other high-throughput data

    BMC Genomics 12:507.

    https://doi.org/10.1186/1471-2164-12-507

    • PubMed
    • Google Scholar
20. 1. Kaufman HL
    2. Kohlhapp FJ
    3. Zloza A

    (2016) Oncolytic viruses: a new class of immunotherapy drugs

    Nature Reviews. Drug Discovery 15:660.

    https://doi.org/10.1038/nrd.2016.178

    • PubMed
    • Google Scholar
21. 1. Kottke T
    2. Tonne J
    3. Evgin L
    4. Driscoll CB
    5. van Vloten J
    6. Jennings VA
    7. Huff AL
    8. Zell B
    9. Thompson JM
    10. Wongthida P
    11. Pulido J
    12. Schuelke MR
    13. Samson A
    14. Selby P
    15. Ilett E
    16. McNiven M
    17. Roberts LR
    18. Borad MJ
    19. Pandha H
    20. Harrington K
    21. Melcher A
    22. Vile RG

    (2021) Oncolytic virotherapy induced CSDE1 neo-antigenesis restricts VSV replication but can be targeted by immunotherapy

    Nature Communications 12:1930.

    https://doi.org/10.1038/s41467-021-22115-1

    • PubMed
    • Google Scholar
22. 1. Laumont CM
    2. Vincent K
    3. Hesnard L
    4. Audemard É
    5. Bonneil É
    6. Laverdure JP
    7. Gendron P
    8. Courcelles M
    9. Hardy MP
    10. Côté C
    11. Durette C
    12. St-Pierre C
    13. Benhammadi M
    14. Lanoix J
    15. Vobecky S
    16. Haddad E
    17. Lemieux S
    18. Thibault P
    19. Perreault C

    (2018) Noncoding regions are the main source of targetable tumor-specific antigens

    Science Translational Medicine 10:470.

    https://doi.org/10.1126/scitranslmed.aau5516

    • PubMed
    • Google Scholar
23. 1. Lechner MG
    2. Karimi SS
    3. Barry-Holson K
    4. Angell TE
    5. Murphy KA
    6. Church CH
    7. Ohlfest JR
    8. Hu P
    9. Epstein AL

    (2013) Immunogenicity of murine solid tumor models as a defining feature of in vivo behavior and response to immunotherapy

    Journal of Immunotherapy (Hagerstown, Md 36:477–489.

    https://doi.org/10.1097/01.cji.0000436722.46675.4a

    • PubMed
    • Google Scholar
24. 1. Löffler MW
    2. Kowalewski DJ
    3. Backert L
    4. Bernhardt J
    5. Adam P
    6. Schuster H
    7. Dengler F
    8. Backes D
    9. Kopp H-G
    10. Beckert S
    11. Wagner S
    12. Königsrainer I
    13. Kohlbacher O
    14. Kanz L
    15. Königsrainer A
    16. Rammensee H-G
    17. Stevanović S
    18. Haen SP

    (2018) Mapping the HLA Ligandome of Colorectal Cancer Reveals an Imprint of Malignant Cell Transformation

    Cancer Research 78:4627–4641.

    https://doi.org/10.1158/0008-5472.CAN-17-1745

    • PubMed
    • Google Scholar
25. 1. Macedo N
    2. Miller DM
    3. Haq R
    4. Kaufman HL

    (2020) Clinical landscape of oncolytic virus research in 2020

    Journal for Immunotherapy of Cancer 8:e001486.

    https://doi.org/10.1136/jitc-2020-001486

    • PubMed
    • Google Scholar
26. 1. Nielsen M
    2. Lundegaard C
    3. Worning P
    4. Lauemøller SL
    5. Lamberth K
    6. Buus S
    7. Brunak S
    8. Lund O

    (2003) Reliable prediction of T-cell epitopes using neural networks with novel sequence representations

    Protein Science 12:1007–1017.

    https://doi.org/10.1110/ps.0239403

    • PubMed
    • Google Scholar
27. 1. Ribas A
    2. Dummer R
    3. Puzanov I
    4. VanderWalde A
    5. Andtbacka RHI
    6. Michielin O
    7. Olszanski AJ
    8. Malvehy J
    9. Cebon J
    10. Fernandez E
    11. Kirkwood JM
    12. Gajewski TF
    13. Chen L
    14. Gorski KS
    15. Anderson AA
    16. Diede SJ
    17. Lassman ME
    18. Gansert J
    19. Hodi FS
    20. Long GV

    (2018) Oncolytic Virotherapy Promotes Intratumoral T Cell Infiltration and Improves Anti-PD-1 Immunotherapy

    Cell 174:1031–1032.

    https://doi.org/10.1016/j.cell.2018.07.035

    • PubMed
    • Google Scholar
28. 1. Schoenfeld AJ
    2. Hellmann MD

    (2020) Acquired Resistance to Immune Checkpoint Inhibitors

    Cancer Cell 37:443–455.

    https://doi.org/10.1016/j.ccell.2020.03.017

    • PubMed
    • Google Scholar
29. 1. Schuster H
    2. Shao W
    3. Weiss T
    4. Pedrioli PGA
    5. Roth P
    6. Weller M
    7. Campbell DS
    8. Deutsch EW
    9. Moritz RL
    10. Planz O
    11. Rammensee HG
    12. Aebersold R
    13. Caron E

    (2018) A tissue-based draft map of the murine MHC class I immunopeptidome

    Scientific Data 5:180157.

    https://doi.org/10.1038/sdata.2018.157

    • PubMed
    • Google Scholar
30. 1. Soria-Guerra RE
    2. Nieto-Gomez R
    3. Govea-Alonso DO
    4. Rosales-Mendoza S

    (2015) An overview of bioinformatics tools for epitope prediction: implications on vaccine development

    Journal of Biomedical Informatics 53:405–414.

    https://doi.org/10.1016/j.jbi.2014.11.003

    • PubMed
    • Google Scholar
31. 1. Tähtinen S
    2. Feola S
    3. Capasso C
    4. Laustio N
    5. Groeneveldt C
    6. Ylösmäki EO
    7. Ylösmäki L
    8. Martins B
    9. Fusciello M
    10. Medeot M
    11. Tagliamonte M
    12. Chiaro J
    13. Hamdan F
    14. Peltonen K
    15. Ranki T
    16. Buonaguro L
    17. Cerullo V

    (2020) Exploiting Preexisting Immunity to Enhance Oncolytic Cancer Immunotherapy

    Cancer Research 80:2575–2585.

    https://doi.org/10.1158/0008-5472.CAN-19-2062

    • PubMed
    • Google Scholar
32. 1. Taylor MA
    2. Hughes AM
    3. Walton J
    4. Coenen-Stass AML
    5. Magiera L
    6. Mooney L
    7. Bell S
    8. Staniszewska AD
    9. Sandin LC
    10. Barry ST
    11. Watkins A
    12. Carnevalli LS
    13. Hardaker EL

    (2019) Longitudinal immune characterization of syngeneic tumor models to enable model selection for immune oncology drug discovery

    Journal for Immunotherapy of Cancer 7:328.

    https://doi.org/10.1186/s40425-019-0794-7

    • PubMed
    • Google Scholar
33. 1. Wang Y
    2. Qiu F
    3. Xu Y
    4. Hou X
    5. Zhang Z
    6. Huang L
    7. Wang H
    8. Xing H
    9. Wu S

    (2021) Stem cell-like memory T cells: The generation and application

    Journal of Leukocyte Biology 110:1209–1223.

    https://doi.org/10.1002/JLB.5MR0321-145R

    • PubMed
    • Google Scholar
34. 1. Welsh RM
    2. Selin LK

    (2002) No one is naive: the significance of heterologous T-cell immunity

    Nature Reviews. Immunology 2:417–426.

    https://doi.org/10.1038/nri820

    • PubMed
    • Google Scholar
35. 1. Wieder T
    2. Eigentler T
    3. Brenner E
    4. Röcken M

    (2018) Immune checkpoint blockade therapy

    The Journal of Allergy and Clinical Immunology 142:1403–1414.

    https://doi.org/10.1016/j.jaci.2018.02.042

    • PubMed
    • Google Scholar
36. 1. Ylösmäki E
    2. Cerullo V

    (2020) Design and application of oncolytic viruses for cancer immunotherapy

    Current Opinion in Biotechnology 65:25–36.

    https://doi.org/10.1016/j.copbio.2019.11.016

    • PubMed
    • Google Scholar
37. 1. Ylösmäki E
    2. Ylösmäki L
    3. Fusciello M
    4. Martins B
    5. Ahokas P
    6. Cojoc H
    7. Uoti A
    8. Feola S
    9. Kreutzman A
    10. Ranki T
    11. Karbach J
    12. Viitala T
    13. Priha P
    14. Jäger E
    15. Pesonen S
    16. Cerullo V

    (2021) Characterization of a novel OX40 ligand and CD40 ligand-expressing oncolytic adenovirus used in the PeptiCRAd cancer vaccine platform

    Molecular Therapy Oncolytics 20:459–469.

    https://doi.org/10.1016/j.omto.2021.02.006

    • PubMed
    • Google Scholar
38. 1. Yu G
    2. Wang LG
    3. Han Y
    4. He QY

    (2012) clusterProfiler: an R package for comparing biological themes among gene clusters

    OMICS 16:284–287.

    https://doi.org/10.1089/omi.2011.0118

    • PubMed
    • Google Scholar
39. 1. Zhang Y
    2. Joe G
    3. Hexner E
    4. Zhu J
    5. Emerson SG

    (2005) Host-reactive CD8+ memory stem cells in graft-versus-host disease

    Nature Medicine 11:1299–1305.

    https://doi.org/10.1038/nm1326

    • PubMed
    • Google Scholar
40. 1. Zitvogel L
    2. Perreault C
    3. Finn OJ
    4. Kroemer G

    (2021) Beneficial autoimmunity improves cancer prognosis

    Nature Reviews. Clinical Oncology 18:591–602.

    https://doi.org/10.1038/s41571-021-00508-x

    • PubMed
    • Google Scholar

Decision letter after peer review:

Thank you for submitting your article "A novel immunopeptidomic-based pipeline for the generation of personalized oncolytic cancer vaccines" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by Carla Rothlin as the Senior Editor. The following individual involved in the review of your submission has agreed to reveal their identity: Etienne Caron (Reviewer #2).

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:

1) Improving the data analyses with proper controls.

2) Testing in additional setting, such as human samples or MCA sarcomas for fulfilling the claim on personalised format and validating the power of this platform.

3) A clear therapeutic advantage of exploiting molecular mimicry and tumor pathogen cross-reactive T cells over conventional TSA.

Reviewer #2 (Recommendations for the authors):

In this study, Feola et al., describe an immunopeptidomic-based pipeline to identify new tumor antigens, which are used in a vaccine platform to treat mice bearing established colon tumors. This study is technical rather than conceptual and aims to exploit viral molecular mimicry and tumor pathogen cross-reactive T-cells in cancer vaccine development. Please find below my concerns.

1) The most important aspect of this technical paper is the assessment of the therapeutic potential of the cancer vaccines to control tumor growth in vivo (Figure 7). However, the results are confusing and difficult to interpret, raising doubts about the overall approach.

i) In fact, the PeptiCRAd4 (ValomD901+gp70423-431 peptide) is the positive control since gp70423-431 is known to be immunodominant according to the authors. It is unclear how well this positive control was supposed to perform, and based on the results, tumor growth is hardly controlled using this immunodominant peptide.

ii) The uncoated adenovirus (ValomD901; no peptide), which I assumed was used as a negative control, outperformed PeptiCRAd3 and PeptiCRAd4 in terms of controlling tumor growth. This is confusing and suggest that the efficiency of such cancer vaccines can be MHCI-peptide independent.

iii) Considering point (i) and (ii), would other controls be more appropriate? How should the readers interpret such results? It seems to me that the final outcome is unpredictable and difficult to interpret. In other words, the current work gives the impression that if we are lucking, we may get relatively good tumor growth control, but it might also get worse than a placebo/negative control. The immediate impact of the current vaccine platform is therefore limited.

2) The viral molecular mimicry is very interesting but difficult to fully appreciate in this manuscript since very little information is provided about the HEX software and viral molecular mimicry in general. For instance, what is the state-of-the-art in exploiting molecular mimicry and tumor pathogen cross-reactive T-cells in cancer vaccine development? This should be introduced. Also, I have the impression that the opportunity that is missed in this paper to increase its impact would be the comparison of the newly identified HEX software-derived peptides versus previously reported TSA from the same tumor cell line. The immunopeptidome of CT26 cells has been characterized in Laumont et al., (2018) (PMID: 30518613). In Laumont et al., mutated TSA (mTSA), aberrantly expressed TSA (aeTSA) and conventional tumor-associated antigen (TAA) have been identified. Assessing the therapeutic potential of those TSA/TAA versus tumor antigens similar to pathogen antigens in the context of the vaccine platform described by the authors would be quite informative and may have the potential to further put forth the relevance of exploiting viral molecular mimicry in cancer immunotherapy.

In summary, I think this is a nice technical paper (with a lot of work!) but I am not convinced about the immediate impact of the current pipeline. The pipeline would benefit many other researchers in the field of cancer immunotherapy if the authors and/or other groups would show (1) a clear therapeutic advantage of exploiting molecular mimicry and tumor pathogen cross-reactive T cells over conventional TSA, and (2) a better mechanistic understanding of pathogen-cross reactive T cells in cancer.

Reviewer #3 (Recommendations for the authors):

The authors should test this system against spontaneous cancers, such as carcinogen-induced tumor models (i.e., MCA). By taking a resected tumor from one of the multiple induced lesions to run it through the pipeline, followed by efficacy testing of the generated vaccine, would have been a great approach to validate their pipeline and to achieve their goal for the study.

Alternatively, utilization of transplantable tumors with aggressive metastatic characteristics, such as the 4T1 breast cancer model, would have been an ideal model as well (PMID: 27584043).

This model allows resection of the primary tumor from the mammary gland, which extends the survival of the challenged mice by a month, during which the resected tumors can be used through the pipeline to develop the "personalized" oncolytic vaccines for testing their efficacy in each mouse.

By using such a system, the authors can use the cell lines as they did with CT26 while being able to use in vivo grown primary tumors as the starting material to truly mimic the clinical setting. As metastasized tumors may undergo additional immunoediting differently in each mouse, it may result in diminished expression of the peptide that was isolated from in vitro grown cancer cell line and may render generated vaccines ineffective.

Furthermore, while the authors initially demonstrated the ability of the vaccines to induce IFNγ responses in vivo by ELISPOT, they do not assess the changes in the immune responses in the treated mice beyond the surface markers of CD4 and CD8 T cells. It lacks the analysis of how the vaccines may or may not have altered the effector functions of the tumor-infiltrating immune cells. Previous studies by others have demonstrated that the induction of Tcm- or Tscm-like CD8 T cells in the peripheral lymphoid tissues is associated with effective immunotherapy responses. In order to demonstrate the long-term protection of the mice, such immunophenotyping and functional analysis may be required.

Reviewer #2 (Recommendations for the authors):

In this study, Feola et al., describe an immunopeptidomic-based pipeline to identify new tumor antigens, which are used in a vaccine platform to treat mice bearing established colon tumors. This study is technical rather than conceptual and aims to exploit viral molecular mimicry and tumor pathogen cross-reactive T-cells in cancer vaccine development. Please find below my concerns.

1) The most important aspect of this technical paper is the assessment of the therapeutic potential of the cancer vaccines to control tumor growth in vivo (Figure 7). However, the results are confusing and difficult to interpret, raising doubts about the overall approach.

i) In fact, the PeptiCRAd4 (ValomD901+gp70423-431 peptide) is the positive control since gp70423-431 is known to be immunodominant according to the authors. It is unclear how well this positive control was supposed to perform, and based on the results, tumor growth is hardly controlled using this immunodominant peptide.

We thank the reviewer for giving us the possibility to discuss further and clarify the use of PeptiCRAd4 (ValomD901+gp70423-431 peptide) as “positive” control in the animal

experiment.

The epitope AH1 used as positive control is a known immunodominant antigen of CT26 derived from gp70, an endogenous viral envelope glycoprotein of murine leukemia virus (1, 2). Based on that, we decided to use it as “positive” control. In agreement with previous studies (1, 3, 4) the peptide AH1 elicits variable anti-tumor response as T cell are sub-optimally activated by AH1 when used in a vaccination setting. However, in Jordan et al., (5) a mutated version of the epitope which differs from the AH1 peptide in a single amino acid (val to ala in position 5, named AH1-5), was used to vaccinate mice. AH1-5 protected against subsequent tumor challenge with CT26. Accordingly, we reasoned that AH1-5 could be used in our work as positive internal control to benchmark our pipeline, in absence of other suitable positive controls. Nevertheless, in our work we were proposing a therapeutic approach to treat well-established tumors; instead Jordan et al., (5) were using the AH1-5 peptide in a prophylactic setting. Additionally, in Jordan et al., (5) , the challenge experiment employed 5x10⁴ cells subcutaneously injected, whereas in our pipeline we subcutaneously injected 0.6x10⁶ cells on the left and 1x10⁶ cells on the right flank for each mouse. In conclusion, our data showed that AH1-5 was ineffective against established tumors, and we hope that our results could guide other researchers in the future in the selection of immunodominant antigen for experiments performed with CT26 tumor model.

To clarify the results, we have added the following sentence in the Discussion section

“Additionally, in order to benchmark our pipeline, we employed as positive control peptide 7, an immunodominant peptide in CT26 cell line named AH1 (gp70423–431); however, AH1 elicits variable anti-tumor response because T cell are sub-optimally activated. To this end, we employed the heterolytic version named AH1-5 which differs from the AH1 peptide in a single amino acid acid (val to ala in position 5) and that has been reported to protect mice against subsequent tumor challenge with CT26 when employed in a prophylactic setting (31). Based on that and in absence of other positive candidate controls, we reasoned that AH1-5 could be used in our work as an internal control. Nevertheless, in our work we are using a therapeutic approach treating well-established tumors instead of a prophylactic setting as showed in Jordan et al., (31)

and in conclusion our data showed that AH1-5 was actually ineffective against well-established tumors.”

We hope that this aspect is now better explained in our manuscript.

ii) The uncoated adenovirus (ValomD901; no peptide), which I assumed was used as a negative control, outperformed PeptiCRAd3 and PeptiCRAd4 in terms of controlling tumor growth. This is confusing and suggest that the efficiency of such cancer vaccines can be MHCI-peptide independent.

We agree with the reviewer´s concern regarding the use of ValomD901 as negative control that outperformed PeptiCRAd3 and PeptiCRAd4 in terms of controlling tumor growth. To better clarify this point, we added the following sentence: “Of note, VALO-mD901 is encoding murine CD40L and OX40L under CMV promoter allowing transgene expression in murine cells. Stimulation of innate (due to CD40L) and adaptive (due to OX40L) immune cells explained the local anti-tumor activity in virus-injected tumors observed in our results and the lack of efficacy in the distant lesions” (Line 400-404).

iii) Considering point (i) and (ii), would other controls be more appropriate? How should the readers interpret such results? It seems to me that the final outcome is unpredictable and difficult to interpret. In other words, the current work gives the impression that if we are lucking, we may get relatively good tumor growth control, but it might also get worse than a placebo/negative control. The immediate impact of the current vaccine platform is therefore limited.

We share the reviewer´s concern regarding the outcome of the in vivo experiment. In this manuscript, our main goal is providing a pipeline for the identification of candidate peptides and subsequently generation of cancer oncolytic vaccines. The overall idea is translating the described pipeline in human application for precise medicine treatments. According to reviewers´ suggestion, we looked for a more appropriate control and we selected two peptides previously identified in a ligandome done with CT26 as reported in Laumont et al., (6). We have repeated the animal experiment to compare the antitumor activity of PeptiCRAd1(PC1 group in the main manuscript, here indicated as PeptiCRAd) to the ValomD901 coated with the Laumont peptide (Laumont). The results confirmed the tumor growth control in both injected and not injected lesions upon PetiCRAd1 treatment and the extension of the anti-tumor response is comparable to Laumont´s group activity. See Author response image 1.

Author response image 1

Download asset Open asset

2) The viral molecular mimicry is very interesting but difficult to fully appreciate in this manuscript since very little information is provided about the HEX software and viral molecular mimicry in general. For instance, what is the state-of-the-art in exploiting molecular mimicry and tumor pathogen cross-reactive T-cells in cancer vaccine development? This should be introduced. Also, I have the impression that the opportunity that is missed in this paper to increase its impact would be the comparison of the newly identified HEX software-derived peptides versus previously reported TSA from the same tumor cell line. The immunopeptidome of CT26 cells has been characterized in Laumont et al., (2018) (PMID: 30518613). In Laumont et al., mutated TSA (mTSA), aberrantly expressed TSA (aeTSA) and conventional tumor-associated antigen (TAA) have been identified. Assessing the therapeutic potential of those TSA/TAA versus tumor antigens similar to pathogen antigens in the context of the vaccine platform described by the authors would be quite informative and may have the potential to further put forth the relevance of exploiting viral molecular mimicry in cancer immunotherapy.

We thank the reviewer for his/her suggestion and we agree that the viral molecular mimicry is an important element to exploit for cancer immunotherapeutic approaches as also previously shown in Chiaro et al., (7) and we are happy to discuss it further. The integration of T-cell cross reactivity in cancer immunotherapy has been slow and primarily focused on avoiding off-target effects. As results, peptide selection criteria avoided similarity to self (8) rather than selecting tumor peptides based on similarity to viral ones to recall crossreactive viral T cells at the tumor site. Nevertheless, emerging data highlighted that MHC-I associated peptides homologous to viral ones are strongly immunogenic; however, to the best of our knowledge, very few works have reported that pathogen specific CD4+ and CD8+ T cells attack malignant lesions. For instance, in Fluckiger et al., the authors showed that microbiota specific T cells cross-react with tumor associated antigens across different murine tumor types (9). Furthermore, in renal and lung cancer patients, microbiota cross-reactive T cells correlated with better PD1 blockade therapy response and in melanoma patients’ microbial peptides crossreactive to tumor antigen were also reported (9). Moreover, an interesting paper focuses on analysing MHC-I peptides derived from endogenous retroelements (ERE) in 16 Blymphoblastoid cell lines; here the authors confirmed the “viral” origin of EREs, demonstrating the similarity between ERE and viral peptides. As consequence of that, the ERE are more immunogenic compared to canonical MHC-I peptide, representing attractive targets for developing cancer vaccines (10). Overall, tumor antigens homologous to pathogen peptides offer an interesting source of candidates for cancer vaccine design, that could benefit future cancer immunotherapeutic approaches.

Additionally, we thank the reviewer for giving us the opportunity of assessing the

therapeutic potential of the TSA/TAA described in Laumont at al. (6) versus tumor antigens similar to pathogen antigens in the context of our vaccine platform. To this end, we repeated the animal experiment, comparing the anti-tumor activity of the best group reported in our manuscript (PC1 group in the main manuscript, here indicated as PeptiCRAd) with a PeptiCRAd version containing Laumont peptides (Laumont). The results confirmed tumor growth control in both injected and not injected lesions upon PeptiCRAd treatment and the extension of the antitumor response was comparable to Laumont´s group activity (Author response image 1). Thus, TSA/TAA described in Laumont were able to induce anti-tumor immune response at the same extent of our candidate peptides similar to viral pathogens. However, we are clearly limited by the animal model in exploring viral pre-existing T cells as the murine immune system is naive to pathogens, lacking memory pathogens T cell that could be re-engaged as an anti-tumor response. In contrast, in human setting, our approach will have the added value of exploiting the T cell memory repertoire to enhance anti-cancer immune response.

In summary, I think this is a nice technical paper (with a lot of work!) but I am not convinced about the immediate impact of the current pipeline. The pipeline would benefit many other researchers in the field of cancer immunotherapy if the authors and/or other groups would show (1) a clear therapeutic advantage of exploiting molecular mimicry and tumor pathogen cross-reactive T cells over conventional TSA, and (2) a better mechanistic understanding of pathogen-cross reactive T cells in cancer.

Reviewer #3 (Recommendations for the authors):

The authors should test this system against spontaneous cancers, such as carcinogen-induced tumor models (i.e., MCA). By taking a resected tumor from one of the multiple induced lesions to run it through the pipeline, followed by efficacy testing of the generated vaccine, would have been a great approach to validate their pipeline and to achieve their goal for the study.

Alternatively, utilization of transplantable tumors with aggressive metastatic characteristics, such as the 4T1 breast cancer model, would have been an ideal model as well (PMID: 27584043).

This model allows resection of the primary tumor from the mammary gland, which extends the survival of the challenged mice by a month, during which the resected tumors can be used through the pipeline to develop the "personalized" oncolytic vaccines for testing their efficacy in each mouse.

By using such a system, the authors can use the cell lines as they did with CT26 while being able to use in vivo grown primary tumors as the starting material to truly mimic the clinical setting. As metastasized tumors may undergo additional immunoediting differently in each mouse, it may result in diminished expression of the peptide that was isolated from in vitro grown cancer cell line and may render generated vaccines ineffective.

We agree with the reviewer´s comments regarding the need of validating our pipeline for a future personalized medicine application. Thus, we are glad to discuss further the idea of using either a carcinogen-induced tumor models (such as MCA) or a transplantable tumor with aggressive characteristics (the reviewer is wisely suggesting the murine triple negative breast cancer model 4T1). We selected this latter as we have the cell line available in the lab, and we are familiar with the subcutaneous 4T1 model in mice (13). Furthermore, we checked the publication (14) suggested here by the reviewer to plan our experiment. In this work the authors developed a simple orthotopic model of breast cancer. The 4T1 cells are first injected in the mammary fat pad and between 21-30 days the tumor is established. At this time, the tumor is removed to extend the mouse life span, but the metastasis has already spread, offering an opportunity to exploit it as model of metastasis in breast cancer (14).

Even though the described model allows the researchers to examinate the metastases

burden upon different treatment, the survival of the mice is reported been around 1 month after the tumor removal (14). When we planned the experiment, we realized that the ligandome landscape investigation performed on the resected tumors would have taken at least 1 week. After that, the analysis in vitro and in vivo of the candidate peptides would have taken an additional month. Indeed, once selected in vitro the candidates, the synthesis of the peptides could have taken up to one month. Despite the time schedule is still acceptable in human scale, in the proposed murine tumor model, it is unpracticable. Looking forward to finding a murine model to validate our truly personalized cancer vaccine approach.

Furthermore, while the authors initially demonstrated the ability of the vaccines to induce IFNγ responses in vivo by ELISPOT, they do not assess the changes in the immune responses in the treated mice beyond the surface markers of CD4 and CD8 T cells. It lacks the analysis of how the vaccines may or may not have altered the effector functions of the tumor-infiltrating immune cells. Previous studies by others have demonstrated that the induction of Tcm- or Tscm-like CD8 T cells in the peripheral lymphoid tissues is associated with effective immunotherapy responses. In order to demonstrate the long-term protection of the mice, such immunophenotyping and functional analysis may be required.

We thank the reviewers for suggesting this interesting immunophenotyping and functional analysis. Following the reviewer´s suggestions, we have further investigated the T cell population in peripheral lymphoid tissue (spleens) harvested from the mice. According to the reviewer´s suggestion, we have examined the induction of T cm (T central memory) and T scm (T stem cell memory like) CD8 in spleens harvested from Mock, VALO-mD901 and PeptiCRAd1 treated groups. The T scm population has been reported as novel memory T cell subgroup (CD44low CD62L+) characterized by high amount of stem cell antigen 1 (SCA-1) and capable of self-renew (15-17). This latter characteristic makes the T scm CD8 T a desirable population to develop upon vaccination (15). In line with that, we investigated the expression of SCA-1 in CD44low CD62L+ CD8+T cells according to the gating strategy reported in Figure8—figure supplement 2A; we observed that SCA-1 showed a tendency to be upregulated in spleens from mice treated with PeptiCRAd1(Figure8—figure supplement 2B), suggesting the possible generation of a long-term CD8 T cell population in mice upon oncolytic vaccine treatment. No differences were instead observed in CD44+ CD62L+ (T central memory) (Figure8—figure supplement 2 C) or CD44+ CD62L- (T effector memory) ( Figure8—figure supplement 2D) compartments.

This is an interesting feature that we should investigate in a long-term experiment. In

the animal setting here reported, we were limited in our time window due to ethical human endpoint for the mice bearing tumors. However, given the interesting results, the data were added in the main manuscript as Supplementary Figure ("Figure8—figure supplement 2").

The following sentence has been added in the Discussion section as well:

“According to previous studies, the effectiveness of the immune therapy is associated with the induction of Tcm (T central memory) or Tscm (T stem cell memory)-like CD8+ T cells in the peripheral lymphoid tissues. The T scm population has been reported as novel memory T cell subgroup (CD44low CD62L+) characterized by high expression amount of stem cell antigen 1 (SCA-1) and capable of self-renew (14-16). This latter characteristic makes the T scm T CD8+ cells a desirable population to develop upon vaccination (14). In line with that, we investigated the expression of SCA-1 in CD44low CD62L+ CD8+T cells in spleens harvested from Mock, VALO-mD901 and PeptiCRAd1 treated groups according to the gating strategy reported in Figure 8—figure supplement 1A we observed that SCA-1 showed a tendency to be upregulated in spleens from mice treated with PeptiCRAd1 (Figure 8—figure supplement 2B), suggesting the possible generation of a long-term CD8+ T cell population in mice upon oncolytic vaccine treatment. No differences were instead observed in CD44+ CD62L+ (T central memory) (Figure 8—figure supplement 2C) or CD44+ CD62L- (T effector memory) (Figure 8—figure supplement 2D) compartments.”.

References

1. Kershaw MH, Hsu C, Mondesire W, Parker LL, Wang G, Overwijk WW, et al. Immunization against endogenous retroviral tumor-associated antigens. Cancer Res. 2001;61(21):7920-4.

2. Casares N, Lasarte JJ, de Cerio AL, Sarobe P, Ruiz M, Melero I, et al. Immunization with a tumor-associated CTL epitope plus a tumor-related or unrelated Th1 helper peptide elicits protective CTL immunity. Eur J Immunol. 2001;31(6):1780-9.

3. McWilliams JA, Sullivan RT, Jordan KR, McMahan RH, Kemmler CB, McDuffie M, et al. Age-dependent tolerance to an endogenous tumor-associated antigen. Vaccine. 2008;26(15):1863-73.

4. Slansky JE, Rattis FM, Boyd LF, Fahmy T, Jaffee EM, Schneck JP, et al. Enhanced antigen specific antitumor immunity with altered peptide ligands that stabilize the MHC-peptide-TCR complex. Immunity. 2000;13(4):529-38.

5. Jordan KR, McMahan RH, Kemmler CB, Kappler JW, Slansky JE. Peptide vaccines prevent tumor growth by activating T cells that respond to native tumor antigens. Proc Natl Acad Sci U S A. 2010;107(10):4652-7.

6. Laumont CM, Vincent K, Hesnard L, Audemard E, Bonneil E, Laverdure JP, et al. Noncoding regions are the main source of targetable tumor-specific antigens. Sci Transl Med. 2018;10(470).

7. Jacopo Chiaro HHEK, Thomas Whalley, Cristian Capasso, Mikaela Grönholm, Sara Feola, Karita D. Peltonen, Firas S. Hamdan, Micaela M. Hernberg, Siru Mäkelä, Hanna Karhapää, Paul E. Brown, Beatriz Martins, Manlio Fusciello, Erkko Ylösmäki, Anna S. Kreutzman, Satu M. Mustjoki ,Barbara Szomolay, Vincenzo Cerullo. Viral Molecular Mimicry Influences the Antitumor Immune Response in Murine and Human Melanoma. medRxiv. 2020.

8. Antunes DA, Rigo MM, Freitas MV, Mendes MFA, Sinigaglia M, Lizee G, et al. Interpreting T-Cell Cross-reactivity through Structure: Implications for TCR-Based Cancer Immunotherapy. Front Immunol. 2017;8:1210.

9. Fluckiger A, Daillere R, Sassi M, Sixt BS, Liu P, Loos F, et al. Cross-reactivity between tumor MHC class I-restricted antigens and an enterococcal bacteriophage. Science. 2020;369(6506):936-42.

10. Larouche JD, Trofimov A, Hesnard L, Ehx G, Zhao Q, Vincent K, et al. Widespread and tissue-specific expression of endogenous retroelements in human somatic tissues. Genome Med. 2020;12(1):40.

11. Chong C, Marino F, Pak H, Racle J, Daniel RT, Muller M, et al. High-throughput and Sensitive Immunopeptidomics Platform Reveals Profound Interferongamma-Mediated Remodeling of the Human Leukocyte Antigen (HLA) Ligandome. Mol Cell Proteomics. 2018;17(3):533-48.

12. Kubiniok P, Marcu A, Bichmann L, Kuchenbecker L, Schuster H, Hamelin DJ, et al. Understanding the Constitutive Presentation of MHC Class I Immunopeptidomes in Primary Tissues. iScience. 2022:103768.

13. Feola S, Capasso C, Fusciello M, Martins B, Tahtinen S, Medeot M, et al. Oncolytic vaccines increase the response to PD-L1 blockade in immunogenic and poorly immunogenic tumors. Oncoimmunology. 2018;7(8):e1457596.

14. Paschall AV, Liu K. An Orthotopic Mouse Model of Spontaneous Breast Cancer Metastasis. J Vis Exp. 2016(114).

15. Gattinoni L, Speiser DE, Lichterfeld M, Bonini C. T memory stem cells in health and disease. Nat Med. 2017;23(1):18-27.

16. Wang Y, Qiu F, Xu Y, Hou X, Zhang Z, Huang L, et al. Stem cell-like memory T cells: The generation and application. J Leukoc Biol. 2021;110(6):1209-23.

17. Zhang Y, Joe G, Hexner E, Zhu J, Emerson SG. Host-reactive CD8+ memory stem cells in graft-versus-host disease. Nat Med. 2005;11(12):1299-305.

Author details

1. Sara Feola

   1. Drug Research Program (DRP) ImmunoViroTherapy Lab (IVT), Division of Pharmaceutical Biosciences, Faculty of Pharmacy, Viikinkaari 5E, University of Helsinki, Helsinki, Finland
   2. Helsinki Institute of Life Science (HiLIFE), Fabianinkatu 33, University of Helsinki, Helsinki, Finland
   3. Translational Immunology Program (TRIMM), Faculty of Medicine Helsinki University, postal code Haartmaninkatu 8, University of Helsinki, Helsinki, Finland

   Contribution

   Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing - original draft, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-4012-4310

2. Jacopo Chiaro

   1. Drug Research Program (DRP) ImmunoViroTherapy Lab (IVT), Division of Pharmaceutical Biosciences, Faculty of Pharmacy, Viikinkaari 5E, University of Helsinki, Helsinki, Finland
   2. Helsinki Institute of Life Science (HiLIFE), Fabianinkatu 33, University of Helsinki, Helsinki, Finland
   3. Translational Immunology Program (TRIMM), Faculty of Medicine Helsinki University, postal code Haartmaninkatu 8, University of Helsinki, Helsinki, Finland

   Contribution

   Data curation, Investigation, Writing – review and editing

   Contributed equally with

   Beatriz Martins

   Competing interests

   No competing interests declared

3. Beatriz Martins

   1. Drug Research Program (DRP) ImmunoViroTherapy Lab (IVT), Division of Pharmaceutical Biosciences, Faculty of Pharmacy, Viikinkaari 5E, University of Helsinki, Helsinki, Finland
   2. Helsinki Institute of Life Science (HiLIFE), Fabianinkatu 33, University of Helsinki, Helsinki, Finland
   3. Translational Immunology Program (TRIMM), Faculty of Medicine Helsinki University, postal code Haartmaninkatu 8, University of Helsinki, Helsinki, Finland

   Contribution

   Conceptualization, Data curation, Writing – review and editing

   Contributed equally with

   Jacopo Chiaro

   Competing interests

   No competing interests declared

4. Salvatore Russo

   1. Drug Research Program (DRP) ImmunoViroTherapy Lab (IVT), Division of Pharmaceutical Biosciences, Faculty of Pharmacy, Viikinkaari 5E, University of Helsinki, Helsinki, Finland
   2. Helsinki Institute of Life Science (HiLIFE), Fabianinkatu 33, University of Helsinki, Helsinki, Finland
   3. Translational Immunology Program (TRIMM), Faculty of Medicine Helsinki University, postal code Haartmaninkatu 8, University of Helsinki, Helsinki, Finland

   Contribution

   Investigation, Methodology, Writing – review and editing

   Competing interests

   No competing interests declared

5. Manlio Fusciello

   1. Drug Research Program (DRP) ImmunoViroTherapy Lab (IVT), Division of Pharmaceutical Biosciences, Faculty of Pharmacy, Viikinkaari 5E, University of Helsinki, Helsinki, Finland
   2. Helsinki Institute of Life Science (HiLIFE), Fabianinkatu 33, University of Helsinki, Helsinki, Finland
   3. Translational Immunology Program (TRIMM), Faculty of Medicine Helsinki University, postal code Haartmaninkatu 8, University of Helsinki, Helsinki, Finland

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

6. Erkko Ylösmäki

   1. Drug Research Program (DRP) ImmunoViroTherapy Lab (IVT), Division of Pharmaceutical Biosciences, Faculty of Pharmacy, Viikinkaari 5E, University of Helsinki, Helsinki, Finland
   2. Helsinki Institute of Life Science (HiLIFE), Fabianinkatu 33, University of Helsinki, Helsinki, Finland
   3. Translational Immunology Program (TRIMM), Faculty of Medicine Helsinki University, postal code Haartmaninkatu 8, University of Helsinki, Helsinki, Finland

   Contribution

   Conceptualization, Data curation, Formal analysis, Investigation, Methodology

   Competing interests

   No competing interests declared

7. Chiara Bonini

   Experimental Hematology Unit, Division of Immunology, Transplantation and Infectious Diseases, IRCCS San Raffaele Scientific Institute, University Vita-Salute San Raffaele, Milan, Italy

   Contribution

   Methodology, Validation

   Competing interests

   No competing interests declared

8. Eliana Ruggiero

   Experimental Hematology Unit, Division of Immunology, Transplantation and Infectious Diseases, IRCCS San Raffaele Scientific Institute, University Vita-Salute San Raffaele, Milan, Italy

   Contribution

   Methodology, Validation

   Competing interests

   No competing interests declared

9. Firas Hamdan

   1. Drug Research Program (DRP) ImmunoViroTherapy Lab (IVT), Division of Pharmaceutical Biosciences, Faculty of Pharmacy, Viikinkaari 5E, University of Helsinki, Helsinki, Finland
   2. Helsinki Institute of Life Science (HiLIFE), Fabianinkatu 33, University of Helsinki, Helsinki, Finland
   3. Translational Immunology Program (TRIMM), Faculty of Medicine Helsinki University, postal code Haartmaninkatu 8, University of Helsinki, Helsinki, Finland

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

10. Michaela Feodoroff

    1. Drug Research Program (DRP) ImmunoViroTherapy Lab (IVT), Division of Pharmaceutical Biosciences, Faculty of Pharmacy, Viikinkaari 5E, University of Helsinki, Helsinki, Finland
    2. Helsinki Institute of Life Science (HiLIFE), Fabianinkatu 33, University of Helsinki, Helsinki, Finland
    3. Translational Immunology Program (TRIMM), Faculty of Medicine Helsinki University, postal code Haartmaninkatu 8, University of Helsinki, Helsinki, Finland
    4. Digital Precision Cancer Medicine Flagship (iCAN), University of Helsinki, Helsinki, Finland
    5. Institute for Molecular Medicine Finland, FIMM, Helsinki Institute of Life Science (HiLIFE), University of Helsinki, Helsinki, Finland

    Contribution

    Investigation, Methodology

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0002-6094-9838

11. Gabriella Antignani

    1. Drug Research Program (DRP) ImmunoViroTherapy Lab (IVT), Division of Pharmaceutical Biosciences, Faculty of Pharmacy, Viikinkaari 5E, University of Helsinki, Helsinki, Finland
    2. Helsinki Institute of Life Science (HiLIFE), Fabianinkatu 33, University of Helsinki, Helsinki, Finland
    3. Translational Immunology Program (TRIMM), Faculty of Medicine Helsinki University, postal code Haartmaninkatu 8, University of Helsinki, Helsinki, Finland

    Contribution

    Data curation, Methodology

    Competing interests

    No competing interests declared

12. Tapani Viitala

    Pharmaceutical Biophysics Research Group, Drug Research Program, Faculty of Pharmacy, University of Helsinki, Helsinki, Finland

    Contribution

    Data curation, Visualization, Writing – review and editing

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0001-9074-9450

13. Sari Pesonen

    Valo Therapeutics Oy, Helsinki, Finland

    Contribution

    Data curation, Visualization, Writing - original draft, Writing – review and editing

    Competing interests

    is an employee and a shareholder at VALO Therapeutics

14. Mikaela Grönholm

    1. Drug Research Program (DRP) ImmunoViroTherapy Lab (IVT), Division of Pharmaceutical Biosciences, Faculty of Pharmacy, Viikinkaari 5E, University of Helsinki, Helsinki, Finland
    2. Helsinki Institute of Life Science (HiLIFE), Fabianinkatu 33, University of Helsinki, Helsinki, Finland
    3. Translational Immunology Program (TRIMM), Faculty of Medicine Helsinki University, postal code Haartmaninkatu 8, University of Helsinki, Helsinki, Finland
    4. Digital Precision Cancer Medicine Flagship (iCAN), University of Helsinki, Helsinki, Finland

    Contribution

    Data curation, Writing - original draft, Writing – review and editing

    Competing interests

    No competing interests declared

15. Rui MM Branca

    Science for Life Laboratory, Department of Oncology-Pathology, Karolinska Institutet, Solna, Sweden

    Contribution

    Data curation, Methodology, Writing – review and editing

    Competing interests

    No competing interests declared

16. Janne Lehtiö

    Science for Life Laboratory, Department of Oncology-Pathology, Karolinska Institutet, Solna, Sweden

    Contribution

    Methodology, Writing – review and editing

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0002-8100-9562

17. Vincenzo Cerullo

    1. Drug Research Program (DRP) ImmunoViroTherapy Lab (IVT), Division of Pharmaceutical Biosciences, Faculty of Pharmacy, Viikinkaari 5E, University of Helsinki, Helsinki, Finland
    2. Helsinki Institute of Life Science (HiLIFE), Fabianinkatu 33, University of Helsinki, Helsinki, Finland
    3. Translational Immunology Program (TRIMM), Faculty of Medicine Helsinki University, postal code Haartmaninkatu 8, University of Helsinki, Helsinki, Finland
    4. Digital Precision Cancer Medicine Flagship (iCAN), University of Helsinki, Helsinki, Finland
    5. Science for Life Laboratory, Department of Oncology-Pathology, Karolinska Institutet, Solna, Sweden
    6. Department of Molecular Medicine and Medical Biotechnology, Naples University “Federico II”, Pansini, Italy

    Contribution

    Conceptualization, Funding acquisition, Project administration, Writing - original draft, Writing – review and editing

    For correspondence

    vincenzo.cerullo@helsinki.fi

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0003-4901-3796

• 2,996

  Page views

• 588

  Downloads

• 9

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.