Abstract 4031

Following HLA-matched allogeneic stem cell transplantation (SCT) donor T cells may recognize polymorphic antigens specifically expressed on cells of the recipient, designated as minor histocompatibility antigens (MiHA). It has been hypothesized that hematopoiesis-restricted MiHA may give rise to a specific graft versus leukemia (GvL) effect without graft versus host disease (GvHD), whereas ubiquitously expressed MiHA can be targets for both GvL and GvHD. We and others have observed that during an overt GvL response, associated with the occurrence of T cells specific for hematopoiesis-restricted MiHA, may coincide with limited GvHD. Therefore, we investigated whether allo-reactive T cells recognizing hematopoiesis-restricted MiHA on leukemic cells may induce collateral damage to surrounding non-hematopoietic tissues. To test this, we stimulated MiHA-specific cytotoxic T lymphocytes (CTLs) in vitro with MiHA positive hematopoietic targets and analyzed whether this resulted in lysis of surrounding MiHA negative primary human fibroblasts seeded in the same well. Using chromium release assays, quantitative FACS analysis and fluorescence microscopy, we demonstrated that MiHA-specific CTLs (CD8 n=3, CD4 n=2) and allo-reactive CTLs upon proper activation by MiHA-positive hematopoietic stimulator cells (EBV-LCL/ DCs), induced 80% (60–100%) bystander cytotoxicity to the surrounding fibroblasts at a 5/1 T cell/fibroblast-ratio within 24 hrs. T cell receptor (TCR)-specific T cell activation by the hematopoietic cells was pivotal for the induction of this collateral damage since no bystander kill was observed when MiHA-specific CTLs were exposed to MiHA negative stimulators. Next, we investigated the influence of the strength of T cell activation by stimulating the CTLs with MiHA-negative stimulator cells loaded with increasing concentrations of the relevant peptide resulting in increasing T cell activation measured by IFNg ELISA and expression of the T cell activation marker CD137. We demonstrated that this increase in the strength of T cell activation is correlated with an increased severity of collateral damage. Moreover, we showed also that T cell activation in a cell free system using PHA (0,5ug/ml) or antiCD3/CD28 beads resulted in induction of collateral damage of the surrounding non-hematopoietic cells illustrating that the MiHA-negative fibroblasts are not attacked due to cross-presentation of HLA/peptide complexes from apoptotic hematopoietic cells. Next, we further investigated the prerequisites for collateral damage induction to occur. Using a transwell system, we demonstrated that direct T cell-fibroblast interaction was required. Furthermore, ICAM-1 expression on the non-hematopoietic targets, induced by pro-inflammatory cytokines which were secreted by the activated CTLs, was demonstrated to be crucial for strong T cell-fibroblast interaction. To investigate the contribution of the different T cell effector mechanisms, we retrovirally overexpressed the anti-apoptotic proteins c-FLIP, Bcl-2 or PI-9 in the fibroblasts, thereby inhibiting the death receptor-, mitochondrial-, or granzyme-B (GzB)-mediated pathways, respectively. High expression of c-FLIP or Bcl-2 did not inhibit induction of collateral damage. In contrast, increased PI-9 expression in the fibroblasts partially inhibited induction of collateral damage. Moreover, blockade of Ca2+-dependent secretion of perforin by EGTA, could also inhibit collateral damage suggesting that the main killing pathway of collateral damage is via release of GzB- and perforin-containing granules by the T cells upon TCR-triggering. Misdirection of the release of cytotoxic granules towards the MiHA negative non-hematopoietic targets may induce collateral damage. In conclusion, these data suggest that hematopoiesis-restricted T cells actively participating in an overt GvL response can induce GvHD by collateral damage to MiHA-negative non-hematopoietic targets by misdirection of their cytotoxic execution machinery when they are in close proximity of the non-hematopoietic tissue.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution