Programmed death-1 is required for systemic self-tolerance in newly generated T cells during the establishment of immune homeostasis.

Lymphopenia driven T cell activation is associated with autoimmunity. That lymphopenia does not always lead to autoimmunity suggests that control mechanisms may exist. We assessed the importance of the co-inhibitory receptor programmed death-1 (PD-1) in the control of lymphopenia-driven autoimmunity in newly generated T cells vs. established peripheral T cells and in thymic selection. PD-1 was not required for negative selection in the thymus or for maintenance of self tolerance following transfer of established PD-1⁻/⁻ peripheral T cells to a lymphopenic host. In contrast, PD-1 was essential for systemic self tolerance in newly generated T cells under lymphopenic conditions, as PD-1⁻/⁻ recent thymic emigrants (RTE), generated after transfer of PD-1⁻/⁻ hematopoietic stem cell (HSC) precursors or thymocytes into lymphopenic adult Rag⁻/⁻ recipients, induced a rapidly lethal multi-organ inflammatory disease. Disease could be blocked by using lymph node deficient recipients, indicating that lymphopenia driven PD-1⁻/⁻ T cell activation required access to sufficient lymph node stroma. These data suggested that PD-1⁻/⁻ mice themselves might be substantially protected from autoimmunity because their T cell repertoire is first generated early in life, a period naturally deficient in lymph node stroma. Consistent with this idea, neonatal Rag⁻/⁻ recipients of PD-1⁻/⁻ HSC were resistant to disease. Thus, a critical role of PD-1 resides in the control of RTE in lymphopenia. The data suggest that PD-1 and a paucity of lymphoid stroma cooperate to control autoimmunity in newly generated T cells. Clinical therapies for autoimmune disease employing lymphoablation and hematopoietic stem cell transplantation will need to take into account functional polymorphisms in the PD-1 pathway, if the treatment is to ameliorate rather than exacerbate autoimmunity.

Developmentally regulated changes in glucosidase II association with, and carbohydrate content of, the protein tyrosine phosphatase CD45.

Glucosidase II (GII) stably interacts with the external domain of CD45 in a carbohydrate-dependent manner. We have found that the association occurs in immature cells, but is significantly reduced in mature T cells. Using mannose-binding protein (MBP), in both FACS analysis and pull-down assays, we find that MBP can specifically recognize cell surface CD45 from immature, but not mature T cells. Analysis of thymocytes reveals increased MBP binding and GII association with CD45 in double-positive thymocytes compared with either double-negative or single-positive thymocytes. As well, the same pool of CD45 recognized by MBP can also associate with GII. Initial analysis of the basis of the interaction between CD45 and MBP suggests MBP binds two different glycoforms of CD45 based on the differential competition with glucose. Finally, inhibition of GII activity in cells that do not normally express MBP ligands results in significant increases in cell surface MBP ligands, including CD45. Taken together, these data suggest that the glucose content of the cell surface CD45 changes as thymocytes undergo maturation to mature T cells, and may be regulated by GII interactions. Such changes in the cell surface carbohydrate on CD45 may affect the development of thymocytes, perhaps via binding of CD45 on thymocytes to lectins on stromal cells.

Specific isoforms of the resident endoplasmic reticulum protein glucosidase II associate with the CD45 protein-tyrosine phosphatase via a lectin-like interaction.

We have previously demonstrated that CD45 physically associates with the endoplasmic reticulum processing enzyme glucosidase II (GII). GII consists of the catalytic alpha-chain and an associated beta-chain. To gain insight into the basis of the association between CD45 and GII, we examined the biochemical requirements for the interaction. We show that the alpha-subunit is essential for the interaction. Interestingly, only a higher molecular weight form of GIIalpha is capable of associating with CD45 in a competitive situation where multiple GIIalpha isoforms are expressed. Further, transfection studies demonstrate that only isoforms containing the alternatively spliced sequence Box A1 are capable of binding CD45, although all isoforms are catalytically active. The interaction between CD45 and GII is dependent on the active site of GII, is mediated through the carbohydrate on CD45, and can be inhibited with mannose. Taken together, these results suggest that GIIalpha acts as a lectin and binds to CD45 in an exon-dependent manner. This lectin activity of GII may be a novel mechanism for the regulation of CD45 biology and play a role in immune function, possibly by regulating CD45 glycosylation.