Background Immunotherapy of cancer by checkpoint blockade has significantly improved the survival of melanoma patients. However, in patients with tumors that are poorly infiltrated by effector T cells the clinical results are not encouraging. Therefore, combination approaches that enhance pre-existing anti-tumor immunity and reset the patients‘ immunological status are urgently needed. In this study we used the tg(Grm1)EPv melanoma mouse model that reflects a non-immunogenic tumor microenvironment. In this mouse model, spontaneous melanoma development is driven by the ectopic expression of the metabotropic glutamate receptor-1 in melanocytes, which confers to them a hyperproliferative and anti-apoptotic phenotype. The same alteration has been shown to be present in 40% of melanoma patient samples. The aim of our study was to investigate whether enhancing dendritic cell (DC) numbers and function in the tg(Grm1)EPv mouse model could restore responsiveness to checkpoint blockade.
Material and Methods We used multicolor flow cytometry, gene expression analysis by RNA-seq and microarray to analyze tumors and tumor-draining lymph nodes (tdLN). With various immunological in vitro and in vivo assays we determined the functional role of DC in tumor immunity.
Results A loss of skin DC has previously been reported for primary melanoma lesions and we here show that melanoma progression in the tg(Grm1)EPv mouse model coincides with a gradual decrease in the skin cDC2 subset and an upregulation of the inhibitory ligands PD-L1 and galectin-9. Monotherapy with anti-PD-L1 could not delay tumor growth, suggesting that this is a good model to study resistance to checkpoint blockade. We hypothesized that by boosting DC numbers and function we would restore responsiveness to checkpoint blockade. By administering a treatment consisting of systemic Flt3L and intratumoral polyI:C/anti-CD40, we were able to rescue the numbers and function of skin cDC2. Analysis of the treated tumors by flow cytometry showed that the DC boost regimen led to an increased tumor infiltration of activated CD4+ and CD8+T cells. An in vitro T cell proliferation assay revealed that dermal cDC2 that had migrated to the tdLN, played a crucial role in this process, since these were able to cross-present endogenous gp100 antigen more efficiently than migratory Langerhans cells and dermal cDC1. CD4+ and CD8+T cells recruited in the tumors of the DC boost treated mice, expressed PD-1 and TIM-3. Therefore, combination therapy with checkpoint blockade of these molecules resulted in increased cytotoxic activity within the tumor and eventually delay of tumor growth.
Conclusions Our results demonstrate that skin DC shape the tumor microenvironment upon immunotherapy and thus, therapies that aim to enhance responsiveness to checkpoint blockade may well benefit from a component that boosts the numbers and the function of skin DC.
Disclosure Information N. Prokopi: None. C.H. Tripp: None. B. Tummers: None. J.C. Crawford: None. M. Efremova: None. K. Hutter: None. L. Bellmann: None. G. Cappellano: None. L. Boon: None. D. Ortner: None. Z. Trajanoski: None. S. Chen: None. T. de Gruijl: None. D.R. Green: None. P. Stoitzner: None.