Resistance of Quiescent Cells to T Cell Attack Determines Both the Efficacy and Specificity of Adoptive T Cell Therapy.

Abstract

Although profound anti-leukemic immune responses can be induced with donor lymphocyte infusions in patients with relapsed or persistent leukemia after allogeneic stem cell transplantation, (late) relapses of the same disease develop regularly even in patients initially entering a complete remission. This suggests that a subpopulation of leukemic (precursor) cells with ultimate self-renewal capacity is capable of resisting T cell attack. We hypothesized that quiescent leukemic precursor cells can evade anti-leukemic therapy by their capacity to survive and persist in the presence of competent cytotoxic T cells. In addition, selectivity of cytotoxic T cells (CTLs) for target cells in active cell cycle in general may also explain why powerful immune responses directed against antigens that are broadly expressed on all tissues of the recipient, like the male-specific HY-antigens, do not necessarily result in severe damage to all tissues of the recipient. Therefore, we determined the efficacy of high affinity CTL clones directed against allo-HLA or minor histocompatibility antigens to kill normal and leukemic hematopoietic cells in dormancy and in active cell cycle, comprising normal and leukemic CD34+ precursor cells, normal B cells, T cells and monocytes, and activated B cells (EBV-LCL) and activated T cells (PHA blasts). Using a CFSE-based cytotoxicity assay allowing the analysis of susceptibility to lysis of specific cell types within a heterogeneous target cell population, we found that all activated target cells were very efficiently lysed, resulting in 60–90% lysis already after 4 hours of exposure to the CTL clones (E/T ratios 1/1–5/1). In contrast, target cells in relative dormancy including the non-proliferating CD34+ CML stem cell fraction, unmanipulated CD34 progenitor cells, and resting T and B cells were protected from CTL-induced cell death (0–20% lysis). Since normal expression of adhesion and HLA class I molecules was shown on these dormant cells, we investigated whether decreased avidity of the T cell/target cell interaction was underlying the poor susceptibility. Therefore, we artificially enhanced the avidity by exogenous loading of the target cells with saturating concentrations of the relevant peptide. This was sufficient to restore the sensitivity to levels comparable to activated target cells, suggesting that decreased avidity of the interaction between high affinity CTL and resting target cells plays a role in the resistance phenomenon. However, even after restoration of the high avidity interaction, a small population of (leukemic) target cells (0,1–10% of the total cell population) was capable of residing, suggesting that additional factors like resistance of quiescent target cells to one or more of the T cell effector mechanisms are involved. To analyze the influence of the sensitivity to T cell lysis of specific target cell types on the specificity of adoptive T cell therapy, we used non-hematopoietic target cells like mesenchymal stem cells and biliary epithelium cells as target cells. Alloreactive T cells showed also diminished capacity to lyse these target cells (10–20% lysis). The addition of inflammatory cytokines like TNF and interferons slightly increased the recognition. In conclusion, under steady state conditions potent allo immune responses may have limited activity against quiescent target cells. Therefore in order to cure the disease, specific activation strategies and/or prolonged persistence of specific T cells will be needed to achieve a potent anti-leukemic effect with controlled GVHD.