The CD45 tyrosine phosphatase regulates Campath-1H (CD52)-induced TCR-dependent signal transduction in human T cells.

Campath-1H, a humanized mAb undergoing clinical trials for treatment of leukemia, transplantation and autoimmune diseases, produces substantial lymphocyte depletion in vivo. The antibody binds to CD52, a highly glycosylated molecule attached to the membrane by a glycosylphosphatidylinositol anchor. Cross-linked Campath-1H is known to activate T cells in vitro. We have investigated the molecular basis for these effects by comparing the protein tyrosine phosphorylation signals induced by Campath-1H and the CD3 mAb OKT3 in primary T cells, and in CD45(+)TCR(+), CD45(-)TCR(+) and CD45(+)TCR(-) Jurkat subclones transfected with CD52. Our results show that Campath-1H triggers similar tyrosine phosphorylation events as OKT3 in both primary T cells and in the CD45(+)TCR(+) Jurkat sub-clone, albeit at quantitatively lower levels. However, no phospholipase C gamma 1 activation nor calcium signals were detected in response to CD52 ligation. The CD52-mediated induction of protein tyrosine phosphorylation was absolutely dependent upon the expression of both the TCR and the CD45 phosphotyrosine phosphatase at the cell surface. Cross-linking of Campath-1H was essential for signal transduction in all cells investigated. Fluorescence resonance energy transfer was used to demonstrate CD52 homo-association at the cell surface in Jurkat T cells in a TCR- and CD45-independent manner, and CD52-TCR association in CD45(+)TCR(+) cells. We propose a model to explain the activating effects of Campath-1H in which CD52 mAb cross-linking causes the trapping of TCR polypeptides within molecular complexes at the cell surface, thereby inducing signals via the TCR by a process which depends on the CD45-mediated regulation of the p56(lck) and p59(fyn) tyrosine kinases.

Aberrant TCR-mediated signaling in CD45-null thymocytes involves dysfunctional regulation of Lck, Fyn, TCR-zeta, and ZAP-70.

CD45 is a transmembrane phosphotyrosine phosphatase expressed on all nucleated hemopoietic cells. Targeting of CD45 exon 9 has generated a mouse line completely lacking CD45 expression (CD45-null) in which there are severe abnormalities in T cell development. Defects in TCR-mediated signals underlying these abnormalities have now been investigated using CD45-null T cells. No T cell proliferation was detected in response to a CD3 mAb. In thymocytes the p56(lck) and p59(fyn) tyrosine kinases were hyperphosphorylated, and p56(lck) was in its inactive conformation. Both basal and TCR-stimulated tyrosine phosphorylation of TCR-zeta and CD3-epsilon were much reduced, and TCR stimulation induced an abnormal p18 phosphoisomer of TCR-zeta previously noted in T cells stimulated by altered peptide ligands. These defects were associated with the failure of ZAP-70 kinase recruitment to the TCR-zeta chain. TCR coupling to the tyrosine phosphorylation of several proteins, including HS1 and p120(cbl), was also much reduced. However, TCR-induced signaling was not ablated, and significant inositol phosphate and calcium signals were observed in CD45-null thymocytes. Our molecular analysis suggests that the threshold for TCR signal transduction is greatly increased in CD45-null T cells, thus explaining the profound defects in thymic development.

Acceptance of skin grafts between mice bearing different allelic forms of beta 2-microglobulin.

Single amino acid disparities in MHC class I molecules can elicit transplantation responses. Since beta 2 microglobulin (beta 2m) is noncovalently associated with class I antigens on the cell membrane we investigated whether the single amino acid polymorphism at position 85 (Asp-->Ala) in the mouse beta 2m molecule can cause skin graft rejection. A B2mb transgene was introduced into CBA(B2ma) mice which subsequently expressed both forms of beta 2m. Skin from these CBA beta 2mb transgenic mice was not rejected by the parental CBA strain. Previous studies showed that cytotoxic T lymphocyte (CTL) responses directed against beta 2mb use H2Kb as a restriction element. We therefore produced mice expressing H2Kb and H2Ab as well as beta 2mb by crossing CBA.beta 2mb mice with either CBA.Kb (CBK) transgenic mice or C3H.SW mice and used these as skin graft donors for beta 2mb negative littermates. In both cases rejection of transgenic skin only occurred when mice had received both a beta 2mb graft and an H2-disparate allograft lying adjacent in the same site. Introduction of the male specific antigen, H-Y, as a helper determinant did not result in rejection of beta 2mb skin. Neither did two CTL determinants (P91A and beta 2mb) on the same graft complement one another to elicit a transplantation response. Prior immunisation with tissues expressing the beta 2m disparity alone did not generate in vivo or in vitro beta 2mb-specific CTL responses, suggesting that this single amino acid difference is not sufficient to elicit a CTL or helper T cell response.

Characterization of human IgG-binding xenoantigens expressed by porcine aortic endothelial cells.

Xenoreactive antibodies (XAb) play a major role in the rejection of xenografts. In this study, human IgG XAb that bind to xenoantigens expressed by porcine aortic endothelial cells (PAEC) were characterized, together with their corresponding xenoantigens. Using an ELISA with both fixed and unfixed confluent monolayers of PAEC, XAb of both IgG and IgM classes in pooled and individual normal human serum were identified. The binding of these IgG XAb to the endothelium is mediated by F(ab')2 and the only detectable subclasses that bind to the endothelium are IgG1 and IgG2. On the basis of direct binding experiments, inhibition and antibody adsorption studies, and enzymatic digestions, it is shown that only a minor component of the XAb binding is directed against galactose in an alpha 1,3 linkage with galactose on PAEC surfaces. There is some cross-reactivity with antigens expressed on porcine lymphocytes, but not porcine red blood cells. Histological examination of sections of porcine aortae, snap-frozen and stained using immunoperoxidase techniques, confirmed interaction with the vascular endothelium. Labeling of the PAEC with 125I, followed by cell lysis and immunoprecipitation under reducing conditions, showed binding of IgG XAb to several components on the endothelial cell surface, the most prominent of which have apparent molecular masses of 75 kDa, 110 kDa, 180 kDa, and 210 kDa. The 110-kDa component and the 180-kDa component were sensitive to digestion with endoglycosidase F, which suggests the participation of N-linked carbohydrate structures. These studies demonstrate that human IgG XAb recognize multiple determinants expressed by PAEC, a minor population of which contain alpha 1,3-linked galactose residues. Cross-reactive determinants are expressed on porcine lymphocytes but not porcine red blood cells.

A cDNA clone encoding the mouse Qa-1a histocompatibility antigen and proposed structure of the putative peptide binding site.

We have isolated a cDNA clone which encodes the Qa-1a histocompatibility Ag from a library prepared from Con A-activated B10.BR mouse spleen cells. The clone encodes a protein of 322 amino acids with three potential N-glycosylation sites. The coding sequence shows strongest similarity with that of the T23d gene of DBA/2 mice which encodes the Qa-1b molecule. Molecular modeling of the putative peptide-combining site indicates most of the differences between Qa-1a and Qa-1b are located peripheral to the binding cleft, with only two amino acid substitutions, at positions 9 and 24, which might affect peptide binding. Many features of the Qa-1 binding cleft are also conserved in the rat RTBM.1 and in human HLA-E molecules. This suggests that all of these molecules may associate with structurally similar peptides.