Clonally-Restricted T-Lymphocytes in PigA Mutant Mice.

Somatic mutation in the X-linked phosphatydylinositol glycan class A (PIG-A) gene causes glycosyl phosphatidylinositol (GPI) anchor deficiency in hematopoietic stem and progenitor cells, in humans, a requirement for the development of the disease paroxysmal nocturnal hemoglobinuria (PNH). While progress has been made in understanding PNH and especially in treatment of intravascular hemolysis secondary to cell surface deficiency of CD59, why PIG-A mutant stem cells expand in the setting of immune-mediated bone marrow failure remains obscure. We produced a conditional PigA knock-out animal model (PigA^−/−) by cross-breeding mice carrying germline insertion of two lox sites flanking exon 6 of PigA gene with mice carrying the transgene Cre-recombinase driven by the human c-fes promoter. The resultant B6 Fes-cre PigA^flox (PigA^−/−) mice had PigA gene inactivation specifically in hematopoietic cells. We observed that GPI-deficient (GPI⁻) bone marrow (BM) and spleen cells from PigA^−/− mice contained much larger proportions of lymphocytes, especially CD8⁺ T cells, in comparison to GPI⁺ cells. The expansion of GPI⁻CD8⁺ T cells was not associated with any obvious hematological phenotype, and blood and BM cell counts were relatively normal in PigA^−/− mice. In comparison to GPI⁺ cells analyzed by microarray, GPI⁻ BM cells showed up-regulation in expression of genes important for immune function responses. Pathway analysis revealed that differentially-expressed genes were clustered in several groups related to immunological function, such as lymphocyte markers (CD8b1, CD8a, CD3e, CD3d, CD7, CD2, CD5, CD6, CD28, CD96, CD27), proteins related to T cell activation (Lck, Zap70, Fyn, Zeta, Lat, Traf1, Tcf7, Ctla2a/Ctla2b), TCR components (Tcr-beta-V13, Tcr-beta-V8.2, Tcr-alpha, Tcr-beta- J, Tcr-gamma), chemokines and C-C motifs (Ccl5, CXCR6, Ccr7), and molecules of the killer lectin-like receptor subfamily (Klrc1, Klrc2, Klra7, Klra8). We transplanted into lethally-irradiated recipients BM cells from PigA^−/− mice (pre-incubated with aerolysin to lyse GPI⁺ cells) or BM cells from normal PigA^+/+ donors. By microarray, transplanted GPI⁻ cells retained the phenotype of untransplanted GPI⁻ cells, with a much increased CD8⁺ T cell proportion and up-regulated immune function gene expression in comparison to transplanted normal BM cells. The enlarged GPI⁻CD8⁺ T cell pool had a significantly lower proportion of CD11a⁺ cells than did GPI⁺CD8⁺ T cells, suggesting that GPI⁻CD8⁺ T cells were generally less active. There was no difference in the proportion of CD44⁻ naive T cells between GPI⁻CD8⁺ and GPI⁺CD8⁺ T cells; GPI⁻CD8⁺ T cells were not NK cells as they lacked surface NK1.1 staining. The percentage of CD4⁺CD25⁺FoxP3⁺ regulatory T cells in GPI⁻ cells was only 10% of that in GPI⁺ cells in peripheral blood in both untransplanted and transplanted animals, indicating that the expanded T cell population in the GPI⁻ cell fraction contained few cells with immunosuppressive property. We further investigated T cell clonality by usage of T cell receptor beta variable region (Vbeta); approximately 5-6 Vbeta subfamilies were over represented in the GPI⁻ CD8⁺ T cells. In particular, Vbeta 5.1/5.2 was prominent in GPI⁻CD8⁺ T cells, constituting 22-23 ± 5% GPI⁻ T cells from untransplanted and transplanted animals; a significant increase in comparison to 8-9.1 ± 0.5% Vbeta 5.1/5.2 clonal representation in GPI⁺CD8⁺ T cells. Our results are consistent with an antigen-driven T cell response in the GPI⁻ lymphocyte population, independent of pancytopenia. Functionally, GPI⁻CD8 T cells showed no response to lectin stimulation as measured by gamma interferon production, but they were capable of effecting target cell apoptosis when co-incubated with minor-H antigen mismatched BM cells in vitro. Our data agrees with observations in humans, in which an immune process driven by a restricted set of (unknown) antigens appears active in the pathogenesis of PNH (Gargiulo et al., Blood 2007). We conclude that deletion of PigA gene in hematopoietic cells, independent of frank hematopoietic failure, leads to enrichment of lymphocytes, especially CD8 T cells, in the GPI⁻ cell fraction that have an inactive and naive phenotype. These expanded, clonally-restricted, T cells may provide an initial pool of immune effectors, which in the proper immune activated environment, contribute to bone marrow failure in PNH.