High Avidity HLA-A2-Restricted CD8+ T Cells against the Wilms Tumor Protein (WT1) Can Be Isolated Only from HLA-A2 Negative Donors Not Subjected to HLA-A2-Mediated Thymic Deletion

The anti-tumor effect of donor lymphocyte infusions (DLI) after HLA-matched allogeneic stem cell transplantation (allo-SCT) can be mediated by donor T cells recognizing minor histocompatibility antigens (mHag) in the context of self HLA on the malignant cells of the recipient. However, T cells recognizing self antigens (Ag) that are overexpressed in malignant cells like the Wilms tumor protein (WT1) have also been proposed to contribute to the anti-tumor reactivity both after allo- and autologous SCT. WT1 is expressed in various types of cancer, including hematological malignancies at higher levels compared to their non-malignant counterparts. Different groups have isolated CD8+ HLA-A2 restricted WT1-specific T cells after HLA-matched transplantation and/or WT1 vaccination. Although these T cells stain with HLA-A2/WT1 peptide tetrameric complexes and are capable of recognizing peptide loaded HLA-A2+ target cells, recognition of primary leukemic cells could hardly be demonstrated. Stauss et al demonstrated recognition of endogenously processed Ag by WT1 specific T cells. However, these CD8 T cells were isolated in a situation where responder and stimulator cells were mismatched for HLA-A2. Based on these data we hypothesized that high avidity HLA-A2 restricted WT1 specific T cells may be deleted during thymic selection in HLA-A2+ donors. As a result, it is likely that only low avidity HLA-A2 restricted WT1 specific CD8+ T cells can be isolated from HLA-A2 positive donors, whereas it might be possible to isolate high avidity HLA-A2 restricted WT1 specific T cells from HLA-A2 negative donors due to the absence of the HLA restriction element in the thymus irrespective of the local WT1 expression. To test this hypothesis, we induced immune responses against the HLA-A2 binding WT1-derived peptide RMFPNAPYL using either HLA-A2+ or A2- donor T cells as responder cells. In case of HLA-A2+ donors (n=3) we stimulated CD45RO depleted donor T cells with autologous monocyte-derived antigen-presenting cells (APCs) loaded with 1E-6M WT1 peptide in the presence of 5ng/mL IL-7. At day 10 we either enriched tetramer-positive cells using immunomagnetic beads (MACS) or restimulated the responses with peptide-loaded autologous PBMC. At day 20 tetramer positive cells were isolated in bulk cultures and single cell/well by flowcytometric cell sorting. When using HLA-A2 negative donor T cells as responder cells, we used HLA-A2+ stimulator cells from an unrelated donor matched for all other class I alleles. The similar protocol was used for the induction of the immune response. At day 10 CD4+ T cells were depleted and the WT1 specific T cells were either enriched using tetramers and MACS or restimulated with peptide loaded HLA-A2+ PBMC. Using these approaches we were capable of isolating CD8+ T cell clones staining brightly with the A2/WT1 tetramer and expressing different T cell receptors (TCRs). HLA-A2+ WT1 specific T cells recognized only HLA-A2+ target cells loaded with 1E-8-1E-6M peptide both in IFNg release assays and cytotoxicity assays. Especially peptide-loaded target cells with an APC phenotype like EBVs were recognized efficiently, whereas normal peripheral blood cells were only recognized when loaded with 1E-6M peptide. In contrast, the HLA-A2 negative WT1-specific T cells efficiently recognized normal peripheral blood cells upon loading of <1E-9M WT1 peptide, indicating that these T cells express TCRs with a higher affinity for the WT1 peptide compared to the HLA-A2+ WT1-specific T cells. Strikingly, EBV-LCL as well as TAP-deficient T2 cells were recognized by the HLA-A2 negative T cells even in the absence of exogenously loaded WT1 peptide. This might be due to the efficient recognition by the high avidity T cells of low levels of WT1 presented in HLA-A2 processed from the endogenous WT1 expression. However, due to the absence of HLA-A2 mediated thymic selection in HLA-A2 negative donors, cross-recognition of other peptides in HLA-A2 may also occur. In conclusion, high avidity WT1 specific CD8+ T cells could be exclusively isolated from HLA-A2 negative donor cells. These high avidity T cells are capable of recognizing target cells expressing very low levels of WT1. These T cell responses will gain more insight into the correlation of WT1 expression in different normal and malignant cell types and the recognition by high avidity WT1 T cells, and the clinical applicability of T cells directed against self antigens.