Novel vaccination strategies for CD4+ T cell immunotherapy of melanoma

Immunotherapy has emerged as a standard treatment modality in melanoma and many other cancers. While a lot is known about the anti-tumoral effector functions of CD8+ T cells, CD4+ T cells remain less well understood in cancer immunotherapy. In the current work, it was hypothesized that melanocyte antigen-specific CD4+ T cells can control the growth of melanomas as efficiently as corresponding CD8+ T cells but differ in the way they recognize antigen and exert their effector functions against tumor cells in the tissue microenvironment. It has been previously shown by the Tüting lab that a single administration of an adenovirus vector expressing the melanocytic antigen gp100 can promote effective expansion of adoptively transferred gp100-specific Pmel-1 TCR transgenic CD8+ T cells and cause regression of established melanomas in syngeneic mice. Here, a similar therapy protocol was established for Trp1-specific TCR transgenic CD4+ T cells. For this, the adenoviral vaccine vector Ad-GTY expressing both gp100 and Trp1 epitopes was first generated. Ad-GTY could expand adoptively transferred Trp1 CD4+ T cells in vivo, albeit less efficiently when compared to Pmel-1 CD8+ T cells. Nevertheless, a Trp1 CD4+ T cell ACT protocol with Ad-GTY showed significant anti-tumor efficacy and could control the growth of HCmel12 melanomas. The recombinant MVA virus vector MVA-PMTP that also expressed both the gp100 and Trp1 epitopes was generated to evaluate prime-boost vaccine strategies. However, MVA-PMTP was only able to re-expand CD8+ T cells but not CD4+ T cells. Moreover, the Ad-MVA prime-boost vaccination strategy did not significantly increase the therapeutic efficacy of the ACT protocols. Following Trp1 CD4+ ACT escaping melanoma cells frequently down-regulated melanocytic antigen expression and acquired a dedifferentiated phenotype presumably due to therapy-induced inflammation. As shown previously with CD8+ T cells, this also represented a major limitation of targeting melanocytic antigens with antigen-specific CD4+ T cells. Experiments using HCmel12 Trp1 antigen loss variants generated with CRISPR-Cas9 genome editing techniques revealed that the control of tumor growth by Trp1 CD4+ T cells is antigen-specific. Experiments with mixtures of HCmel12 control and Trp1 knockout cells demonstrated that Trp1 CD4+ cells can exert significant bystander killing and that immunoselection for irreversible genetic antigen loss is dominant over reversible phenotypic antigen loss for immune escape of melanoma cells. HCmel12 Ciita loss variants were also generated with CRISPR-Cas9 genome editing techniques. Unlike unmodified HCmel12 cells they fail to upregulate MHC class II and therefore cannot be directly recognized by Trp1 CD4+ T cells. Experiments revealed that direct MHC class II restricted recognition of melanoma cells by Trp1 CD4+ T cells was not required for tumor growth control in vivo. This suggested an important role for indirect stimulation of Trp1 CD4+ T cells by APC in the tumor microenvironment. Likely, Trp1 CD4+ T cells indirectly control melanoma growth in the tumor microenvironment through Th1 associated cytokines such as IFN-γ and TNF-α. Future studies will have to address the spatial location of Trp1 CD4+ T cells in the tumor microenvironment, their interaction with other immune cells and the role of Th1-associated cytokines for their anti-tumor efficacy. Combining T cell therapies with signal transduction inhibitors or checkpoint inhibitors to counteract mechanisms of therapy resistance and immune escape in mouse models will help to delineate strategies for more effective treatment of melanoma patients that should be tested in the clinic