Immune characterization of pre-clinical murine models of neuroblastoma

www.nature.com/scientificreports OPEN Immune characterization of pre‑clinical murine models of neuroblastoma Emily R. Webb1,3, Silvia Lanati1, Carol Wareham1, Alistair Easton1,2,4, Stuart N. Dunn1, Tatyana Inzhelevskaya1, Freja M. Sadler1, Sonya James1, Margaret Ashton‑Key2, Mark S. Cragg1, Stephen A. Beers1,5 & Juliet C. Gray1,5* Immunotherapy offers a potentially less toxic, more tumor-specific treatment for neuroblastoma than conventional cytotoxic therapies. Accurate and reproducible immune competent preclinical models are key to understanding mechanisms of action, interactions with other therapies and mechanisms of resistance to immunotherapy. Here we characterized the tumor and splenic microenvironment of two syngeneic subcutaneous (NXS2 and 9464D), and a spontaneous transgenic (TH-MYCN) murine model of neuroblastoma, comparing histological features and immune infiltrates to previously published data on human neuroblastoma. Histological sections of frozen tissues were stained by immunohistochemistry and immunofluorescence for immune cell markers and tumor architecture. Tissues were dissociated by enzymatic digestion, stained with panels of antibodies to detect and quantify cancer cells, along with lymphocytic and myeloid infiltration by flow cytometry. Finally, we tested TH-MYCN mice as a feasible model for immunotherapy, using prior treatment with cyclophosphamide to create a therapeutic window of minimal residual disease to favor host immune development. Immune infiltration differed significantly between all the models. TH-MYCN tumors were found to resemble immune infiltration in human tumors more closely than the subcutaneous models, alongside similar GD2 and MHC class I expression. Finally, TH-MYCN transgenic mice were administered cyclophosphamide alone or in combination with an anti-GD2 or anti-4-1BB monoclonal antibody, which resulted in increase in survival in both combination therapies. The TH-MYCN transgenic mouse is a promising in vivo model for testing immunotherapy compounds and combination therapy in a preclinical setting. Neuroblastoma is one of the commonest childhood malignancies, accounting for 8% of pediatric cancers, and 15% of pediatric cancer deaths. It is an embryonal tumor, arising from progenitor cells of the sympathetic nervous system, most commonly in the adrenal glands1. Over 50% of patients with neuroblastoma are considered to have ‘high risk’ disease, because of adverse prognostic features such as amplification of the MYCN oncogene or metastatic spread. The outcome for these patients is poor (< 50% long term survival), despite intensive multi-modal therapies1–4. Immunotherapy is a potentially attractive alternative or additional treatment for these patients. A number of tumor antigens have been identified and endogenous anti-tumor immune responses are documented5,6. The majority of clinical immunotherapies for neuroblastoma have focused on targeting the disialoganglioside GD2, expressed on virtually all n euroblastomas7,8. Anti-GD2 mAbs have been shown to improve outcome in first line treatment for high risk neuroblastoma, and are now considered a key component of treatment9,10. Despite the improvement in outcome achieved with anti-GD2 mAbs, the many children with high-risk neuroblastoma still ultimately relapse and die from their disease11,12. Improving the efficacy of anti-GD2 antibody 1 Antibody and Vaccine Group, Centre for Cancer Immunology, University of Southampton Faculty of Medicine, Southampton General Hospital (MP127), Tremona Road, Southampton, Hampshire SO16 6YD, UK. 2Cellular Pathology, University Hospitals Southampton NHS Foundation Trust, Southampton SO16 6YD, UK. 3Present address: Edinburgh Cancer Research Centre, Institute of Genetics and Molecular Medicine, University of Edinburgh, Western General Hospital, Edinburgh EH4 2XU, UK. 4Present address: Department of Oncology, University of Oxford, Old Road Campus Research Building, Roosevelt Drive, Oxford OX3 7DQ, UK. 5These authors jointly supervised this work: Stephen A. Beers and Juliet C. Gray. *email: Scientific Reports | (2020) 10:16695 | https://doi.org/10.1038/s41598-020-73695-9 1 Vol.:(0123456789) www.nature.com/scientificreports/ Scientific Reports | Vol:.(1234567890) (2020) 10:16695 | https://doi.org/10.1038/s41598-020-73695-9 2 www.nature.com/scientificreports/ ◂Figure 1.  Tumor growth kinetics, survival and histological structure differ between subcutaneous and spontaneous neuroblastoma tumor models. (A) AJ or C57BL/6 mice were inoculated subcutaneously with NXS2 or 9464D (respectively) cells. Comparison of tumor growth kinetics of individual mice of NXS2 (grey) and 9464D (black) tumors, and average tumor growth. Tumor size is shown as mm2. (B) Survival of mice bearing either NXS2 (grey), 9464D (black) or TH-MYCN (dashed) tumors. NXS2 and 9464D survival was measured from date of inoculation. TH-MYCN survival is measure from date of birth. n = 5. (C) ex vivo tumors were frozen in OCT and stained by H&E. For NXS2 tumor: (i) homogeneous appearance of the tissue; (ii) collapsed blood vessels; (iii) Mitotic figures; (iv) pseudorosettes; (v) muscle fibers. (D) For 9464D tumor: (i) loosely packed appearance with thick capsule; (ii) Large dilated blood vessel and thick capsule around tumor edge; (iii–v) demonstration of ‘holes’ throughout tumor. (E) For TH-MYCN tumor: (i) transverse section shows high complexity of the tumor microenvironment with visibly enlarged vessels ( →) and tertiary lymphoid structures surrounding the tumor mass (*); (ii) islands of cancer cells divided by fibrous septa; (iii) tertiary lymphoid structure; (iv) border between tumor cells (bottom) and adrenal gland (top) with ganglion cells in it; (v) cross section of a nerve with embedded ganglion cells, surrounded by cancer cells, fat tissue and enlarged arteriole. Scale bars, 100 µm. therapy, as well as developing novel immunotherapies, is a key focus of efforts to improving survival. Many different immunotherapeutic approaches have shown pre-clinical efficacy, but relatively few have progressed after clinical trials. Pre-clinical modeling is essential in order to compare and prioritize different therapies, to understand mechanisms of action and resistance, and to identify potential biomarkers. Furthermore, it is likely that maximal benefit will be achieved by using immunotherapies in combination. The number of potential combinational therapies is vast, and careful pre-clinical assessment is vital to rationally guided trial design, particularly given the small potential patient population. Assessment of most immunotherapies requires immunocompetent in vivo models. Ideally these should be comparable in immunogenicity to human neuroblastoma, with similar endogenous anti-tumor immune responses and regulatory mechanisms. The number of such neuroblastoma models is limited. Immunocompetent models used include injected syngeneic cell lines, genetically modified murine models whic (...truncated)