ABSTRACT: In chronic viral infections such as HIV-1, HBV and HCV, CD8+ T cells may become ”exhausted”, characterized by progressive functional deficiency. Recently, the underlying mechanisms have been characterized by molecular profiling of virus-specific CD8+ T cells. In contrast, only little is known of self/tumor-specific CD8+ T cells from cancer patients, likely because they are much more difficult to assess. For the first time, we determined the molecular profile of human tumor-specific CD8+ T cells, upon sorting of Melan-A/MART-1 specific T cells directly ex vivo from 19 melanoma patients after vaccination with peptide and CpG. For comparison, we sorted protective T cells specific for the two herpes viruses EBV and CMV. In peripheral blood, we found multiple features of functional effector T cells, with only small but nevertheless significant differences between the three T cell populations, resulting in clean clustering according to antigen specificity. In contrast, Melan-A/MART-1 specific T cells obtained from tumor-infiltrated lymph nodes (TILNs) expressed multiple genes associated with T cell exhaustion, compatible with the known functional deficiencies of T cells in melanoma metastases. We show that individual T cells simultaneously expressed multiple inhibitory receptors implied in functional impairment. Together, the data indicate that in circulation, human tumor-specific T cells have the potential to become competent effector cells. In the tumor environment, however, T cells are exhausted and fail to control malignant disease. Our novel resource data identify mechanisms of functional T cell deficiency in melanoma patients, providing a rational basis for the improvement of immune therapy. 52 samples were measured in two sets of experiments. 4 self/tumor-specific samples from blood were replicated between the first and second sets. Set 1 (32 samples, all from blood): 13 naive samples, 10 EBV-specific samples, 9 self/tumor-specific samples. Set 2 (20 samples): 7 CMV-specific samples from blood, 7 self/tumor-specific samples from TILN, 6 self/tumor-specific samples from blood.
