Long-Term Persistence of Functionally Altered GPI-Anchor Negative T-Cell Subsets After Alemtuzumab-Based T-Cell Depleted Hematopoietic Stem Cell Transplantation.

Abstract 2438

Poster Board II-415

The anti CD52 antibody alemtuzumab is frequently used for in vivo T cell depletion (TCD) in the context of allogeneic hematopoietic stem cell transplantation (HSCT). We have recently demonstrated the persistence of CD52-negative T-cell subsets in patients after HSCT following alemtuzumab-mediated TCD (Meyer, Wagner et al., Bone Marrow Transplantation 2009). The loss of CD52 among lymphocytes was exclusively related to T cells and was more prominent in CD4 compared to CD8 T cells. CD8-depleted donor-lymphocyte infusions (DLI) increased the percentage of CD52-positive CD4 T cells. In patients who did not receive DLI, CD52-negative T cells were detected in significant proportions of up to 40% found even more than 3 years after transplantation. We therefore investigated the regulation as well as the functional consequences of a loss of CD52-expression in T cells of our patients. Peripheral blood T cells of patients with CD52-negative T cells after more than 12 months post allogeneic HSCT following TCD with high-dose alemtuzumab (100 mg) were sorted according to their expression of CD52. RT-PCR showed no difference in CD52 mRNA expression of CD52-positive compared to negative T cells. Since transcriptional regulation was therefore unlikely and CD52 is a glycosylphosphatidylinositol (GPI)-anchored protein, we stained for the presence of further GPI-anchored molecules such as CD55 and CD59 on peripheral blood lymphocytes of our patients. We found that the CD52-negative T cells had also lost expression of CD55 and CD59, whereas CD52-positive cells remained positive for these antigens. We then directly labeled the GPI-anchors using FLAER (fluorescent aerolysin) and thereby confirmed that the loss of CD52 was correlated with a reduced density of the GPI-anchors in the cell-membrane. However, our patients did not exhibit clinical signs of paroxysmal nocturnal hemoglobinuria (PNH), which is in line with the finding that the loss of GPI-anchors was only related to T cells. With the aim to characterize the functional impact of the reduced GPI-anchor density on T cells, we separated CD52-negative from CD52-positive T cells by flow cytometry. The subpopulations were expanded in vitro using low-dose IL2, OKT3, and allogeneic feeder-cells. CD52 expression remained unaltered in CD52-negative as well as CD52-positive cultures for more than 6 weeks. In contrast, when purified T cells of healthy donors were treated with alemtuzumab in vitro (10 μg/mL, 4 h), we only observed a transient down-regulation of the antigen. The growth-kinetics of the non-specifically stimulated T cell cultures did not differ between the CD52-positive and the negative cultures. Yet, when we expanded T cells of a cytomegalovirus (CMV)-positive patient, transplanted from a CMV-positive donor, by subsequent stimulation with overlapping peptides of CMV-pp65, only the proliferation of CD52-positive T cells increased after the addition of peptides. We furthermore applied CD52-positive as well as CD52-negative CD4 and CD8 T cells derived from the antigen-independent culture of this patient in an IFN-gamma ELISPOT assay with autologous dendritic cells (DC) loaded with overlapping peptides of CMV-pp65 and IE1. CMV-specific IFN-gamma spot-production was only evident in the CD52-positive populations. We also conducted IFN-gamma secretion-assays on ex vivo T cells stimulated with autologous DC loaded with CMV-peptides and found a reduced antigen-specific IFN-gamma production in CD52-negative CD4 and CD8 T cells. In addition, we analyzed IFN-gamma secretion of T cells following allogeneic stimulation with DC of a healthy individual and again detected lower levels of IFN-gamma production by CD52-negative compared to CD52-positive T cells. In summary, we demonstrated that the permanent loss of CD52 in a proportion of reconstituting T cells after alemtzumab-based TCD is associated with a loss of GPI-anchors in the cellular membrane. Our data suggest that this loss correlates with reduced T-cell effector-functions in response to antigen-specific stimulation. In addition to a better understanding of the role of alemtuzumab-mediated TCD on T cell reconstitution, further comparison of functional responses in different T-cell subsets in association with the presence or absence of GPI-anchors might help to explore the impact of GPI-anchors and GPI-anchored molecules in this context.