Graft versus host disease (GVHD) in allogeneic stem cell transplantation (SCT) and graft rejection is caused by alloreactive T-cells. Alloreactivity can be exerted by naïve as well as by memory T-cells. Persistent latent viral infections, like those with herpes viruses, have a profound impact on the repertoire of memory T-cells. This implies that virus specific memory T-cells are also potentially alloreactive. Previously it has been shown that virus specific T-cell clones can cross react against allo-HLA. We investigated the frequency of alloreactivity mediated by virus specific T-cells. Mixed lymphocyte reactions, previously used to determine precursor frequencies of alloreactive T-cells, give an underestimation of the total frequency of alloreactive T-cells, due to limited number of allo-HLA alleles tested in this system. Therefore, in this study multiple CD8+ virus specific T-cells lines and clones were tested for alloreactivity against almost all frequent HLA class I and II alleles. From different healthy individuals we derived CD8+ virus specific T-cell lines, specific for Epstein Barr virus (EBV), Cytomegalovirus (CMV), Varicella Zoster virus (VZV) and Influenza virus (Flu) which were restricted to different HLA molecules. The generation of the T-cell lines and clones was performed by bulk sorting and single cell sorting, based on staining with viral peptide/MHC complex specific tetramers. The viral specificity of the expanded lines and clones was confirmed by tetramer staining and cytotoxicity and cytokine production assays. Polyclonality of the T-cell lines and monoclonality of the T-cell clones was confirmed by TCR Vβ analysis. Next, the T-cell lines and clones were screened for alloreactivity by testing against a panel of 29 different EBV transformed LCLs, together covering almost all frequent HLA class I and II molecules. 90% of tested virus specific T-cell lines and 40% of virus specific T-cell clones were found to be alloreactive, recognizing at least one of the allo-HLA alleles. For several lines and clones the specific recognized allo-HLA molecule was further identified using a panel of HLA typed target cells in combination with HLA specific blocking antibodies. Additionally, single HLA antigen expressing cell lines were used as target cells. Thus far we found EBV EBNA3A specific, HLA-A3 restricted T-cell clones to recognize HLA-A31. A CMV pp50 specific, HLA-A1 restricted T-cell line recognized HLA-A68. One VZV IE62 specific, HLA-A2 restricted clone showed recognition of HLA-B57, while another clone with the same specificity but with a different TCR Vβ recognized HLA-B55. An EBV BMLF specific, HLA-A2 restricted T-cell line showed recognition of HLA-A11. Finally an EBV BRLF specific, HLA-A3 restricted clone recognized HLA-A2. Our results show that a high percentage of virus specific T-cells can exert alloreactivity against allo- HLA molecules. Previously it was assumed that virus specific T-cells are not alloreactive against foreign HLA, allowing safe application of virus specific T-cell lines derived from HLA disparate donors in patients without the risk of inducing GVHD. Our data indicate that applying virus specific T-cell lines over HLA barriers does give a significant risk of GVHD and suggest that lines should be tested for alloreactivity against patient specific HLA alleles prior to application. A substantial part of the memory T-cell pool consists of virus specific T-cells, which are dominated by a limited repertoire of virus specific T-cell clones, present in high frequencies. Thus, virus specific T-cells recognizing allo-HLA alleles may also play an essential role in graft rejection.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author