Both the Activation Kinetics and the Frequency of Regulatory T Cells Determine the Ability to Generate Primary Anti-Tumor and Pathogen- Specific Immune Responses from a Naïve Donor T Cell Repertoire

Allogeneic stem cell transplantation (allo-SCT) is commonly used in the treatment of patients with hematological malignancies. However, the impaired immune reconstitution after allo-SCT is coincided by an increased incidence of viral complications in the first six months after the transplant, especially in seropositive patients transplanted with a stem cell graft from a seronegative donor. Infusions of unmodified donor T cells (DLI) can lead to the induction of anti-tumor immune responses caused by donor T cells recognizing minor histocompatibility antigens (mHag) on the recipient cells, but is often coincided by the induction of graft versus host disease. Moreover, the number of patients in which profound and sustained anti-tumor responses are induced is limited, most likely due to the inadequate induction of a primary immune response. A more effective and specific control of the malignant disorder or the infectious complications may be accomplished by the administration of in-vitro selected tumor-specific or virus-specific T cells. Although the in-vitro induction of primary immune responses against both leukemic cells and defined epitopes derived from mHag, tumor antigens (Ag), and viral Ag has been shown to be feasible by us and others, the robustness of the procedure remains limited, hampering large scale clinical application. We hypothesized that the major complication in the in-vitro induction of primary immune responses is the unfavorable balance between the number and activation of antigen-specific precursor T cells (Tprec), estimated to be <0.001% of the naïve donor T cells, and the number of regulatory T cells (Treg; 0.5–5%) capable of inhibiting the activation and/or expansion of the Ag-specific Tprec locally at the site of the priming of the immune response, e.g. the antigen presenting cell (APC). In accordance with this hypothesis, depletion of CD45RO+ responder T cells, containing the majority of Treg, increased the ability to induce primary immune responses against leukemic cells, mHag and viral epitopes, most likely due to the relative depletion of Treg and the relative enrichment of Ag-specific Tprec. Therefore, we subsequently analyzed the activation kinetics of naïve donor T cells and CD4+CD25+CD127-FoxP3+ Treg by analyzing the co-expression of a panel of activation markers including CD137 (4-1BB), CD69, HLA-DR, and CD154 (CD40L) upon activation of the different T cell populations with anti-CD3/28 stimulating antibodies. Whereas it took 2–4 days to obtain maximal activation of all naïve CD4 and CD8 T cells, the population of Treg were uniformly activated already after 24 hours, resulting in expression of CD137 and CD69 on the whole population, and HLA-DR and CD154 on part of the Treg. Similar activation of the Treg was seen upon coculture with autologous mature monocyte-derived APC. In addition, the APC promoted the survival of Treg, even in the absence of exogenous IL-2. Next, we determined the inhibitory capacity of unstimulated Treg and Treg stimulated with autologous APC or CD3/28 antibodies. As responder cells (Tresp) CD25 depleted donor T cells labeled with CFSE were used counterstained for the expression CD137 and CD69 to allow simultaneous analysis of the inhibition of Tresp activation and proliferation in response to CD3/28 stimulation. Treg populations were added in frequencies ranging from 0.05–150%. Whereas on average a 1/1 ratio (100%) of unstimulated Treg was necessary to obtain 50% inhibition of CD3/28 induced Tresp activation and proliferation, activation of the Treg by autologous APC significantly increased the inhibitory potential of the Treg resulting in 50% inhibition of Tresp activation at a median frequency of 10% added Treg (n=8, p=0.001). CD3/28 pre-stimulation further increased the inhibitory potential of the Treg resulting in 50% Tresp inhibition at the lowest ratios tested. In conclusion, these data demonstrate that Treg cells are activated by autologous APC with much faster kinetics compared to naïve T cells. Moreover, this activation further increased their inhibitory capacity. This may shift the balance locally at the APC site further towards Treg-induced inhibition. Relative depletion of Treg from the Tresp populations by depletion of CD45RO+ cells can shift the numeric balance towards the activation of Ag-specific Tprec cells facilitating the reproducible induction of virus- and tumor-specific T cells for adoptive transfer.