Search

We designed a trap system to isolate different amino acid sequences which could target proteins to the cell surface via GPI anchor transfer. This selection procedure is based on the insertion of various sequences which regenerate a functional GPI anchor signal sequence and therefore provoke re-expression at the surface of a reporter molecule. Using this trap for cell surface targeting sequences, we could show the importance of the defined elements essential for GPI anchor addition. Such a system could be used for an exhaustive analysis of the carboxyl terminus structural requirements for GPI membrane anchoring.

The attachment of glycolipid anchors to the Thy-1 glycoprotein during biosynthesis was followed by the change of detergent-binding properties of biosynthetically labelled Thy-1 precursors upon phospholipase C treatment in the murine thymoma lines BW5147 and S1A. In S1A, 80% of the Thy-1 molecules were phospholipase-C-sensitive after a 2 min pulse with [35S]methionine, indicating that these molecules were already anchored via a glycolipid tail. In BW5147, 47% of the Thy-1 molecules had phospholipase-C-sensitive anchors attached after a 1.5 min labelling and, with longer pulses, this percentage rose to 76%. Tunicamycin did not block the addition of glycolipid anchors, and glycolipid attachment also occurred at 21 degrees C. The findings suggest that the attachment of glycolipid anchors occurs in the rough endoplasmic reticulum.

A rabbit antiserum to the mouse CD8 antigen (Lyt-2/3) was obtained through the use of a synthetic peptide corresponding to the N-terminal segment of the 37 kDa subunit of the CD8 molecular complex. This anti-peptide antiserum detected specifically the membrane antigen from detergent extracts of surface-labeled mouse thymocytes. The efficiency of the immunoprecipitation increased significantly upon treatment of the cell lysates with 0.1% SDS at 100 degrees C before immunopurification. Finally both the 37 kDa and the 32 kDa polypeptides expressing the Lyt-2 antigen were revealed by immunoblotting.

A monoclonal antibody (mAb), 3A35, produced against mouse macrophages (M phi) was found to react against certain activated T cells. This mAb, a rat IgM, resulted from a cell fusion between a mouse plasmacytoma and rat lymphocytes immunized against mouse M phi. It bound more avidly to activated than to resident M phi. It did not react against B cells and resting T lymphocytes but recognized certain dividing T cells like EL4 lymphoma, concanavalin A-activated and interleukin 2-expanded spleen cells, and helper T cell hybridomas. By contrast, other T lymphocyte-derived cell lines such as YAC-1 and CTLL2 were unreactive. No clear relationship was found between the binding of 3A35 to cells and the expression of L3T4 and Lyt-2 antigens. The specific stimulation of T cell clones with antigen rapidly induced a strong reactivity with 3A35 mAb which declined thereafter to a low (helper clones) or non-reactivity (cytotoxic clones) after 10 days of culture. Immunoprecipitation experiments, performed with M phi derived from bone marrow cell cultures, surface iodinated with 125I or metabolically labeled with [35S]methionine, showed that 3A35 bound to a 200-kDa molecule, shifting to 175 kDa under reducing conditions. In peritoneal M phi activated in vivo, in addition to the 175-kDa band, new bands migrating at 140, 120 and 85 kDa were identified by 3A35 and could be absorbed on a commercial anti-T200 mAb bound to Sepharose beads. After strengthening the cell binding of 3A35 to EL4 lymphoma cells by a cross-linking agent, only a 85-kDa molecule was immunoprecipitated. Thus, 3A35 identifies a new epitope of the T200 molecule family which is expressed on M phi and activated T cells.

Recent evidence shows that the mature Thy-1 surface glycoprotein lacks the C-terminal amino acids 113 to 143 predicted from the cDNA sequence and is anchored in the plasma membrane by a complex, phosphatidylinositol-containing glycolipid attached to the alpha-carboxyl group of amino acid 112. Here we studied the biosynthesis of Thy-1 in two previously described and two newly isolated Thy-1-deficient mutant cell lines. Somatic cell hybridization indicated that their mutations affected some processing step rather than the Thy-1 structural gene. The Thy-1 made by mutants of classes C, F, and H bound detergent but, in contrast to wild-type Thy-1, their detergent-binding moieties could not be removed by phospholipase C. In addition, tryptophan, which only occurs in position 124, was incorporated into Thy-1 of these mutants but not of wild-type cells. Last, the Thy-1 of wild-type but not mutant cells could be radiolabeled with [3H]palmitic acid. Together, these findings strongly suggest that mutants of classes C, F, and H accumulate a biosynthetic intermediate of Thy-1 which retains at least part of the hydrophobic C-terminal peptide. The Thy-1 of these mutants remained endoglycosidase H sensitive, suggesting that it accumulated in the rough endoplasmic reticulum or the Cis-Golgi. A different Thy-1 intermediate was found in a class B mutant cell line: the Thy-1 of this mutant was 2 kilodaltons smaller than the Thy-1 of other cell lines, did not bind detergent, and was rapidly secreted via a normal secretory pathway.

In this paper, we describe the activity of conjugates constituted of monoclonal antibodies chemically coupled to ricin and the ricin A-chain-like toxin gelonin. Several antibody-toxin conjugates (immunotoxins) were assayed against a Thy 1.2+ and Lyt 2.2+ T-cell line. Anti-Thy 1.2-gelonin and anti-Lyt 2.2-ricin immunotoxins (IT) had a good cytotoxic activity against target cells, whereas anti-Lyt 2.2-gelonin showed weak or no cytotoxicity. An anti-Lyt 2.2-gelonin IT displayed detectable cytotoxicity only when cells were pretreated with ammonium chloride. IT binding activity was very similar for anti-Thy 1.2 and anti-Lyt 2.2 IT, and surface antigen density was of the same order of magnitude for Thy 1.2 and Lyt 2.2 antigens. Our results suggest that intrinsic properties of the surface target antigen are able to influence the IT cytotoxic efficacy.

The biosynthesis and the maturation of Thy-1 antigen of mouse thymocytes have been studied by using a xenogeneic rabbit anti-mouse Thy-1 antibody. The earliest form of Thy-1 detected after a 5-min pulse with [35S]methionine and [35S]cysteine had an apparent m.w. of 26,500. During chase, this band converted to a molecular ratio (Mr) = 25,000 polypeptide, probably derived from the latter by trimming of glucose or mannose residues from the three high-mannose glycan units of Thy-1. Mature Thy-1 molecules were detected at the cell surface after a 15-min chase. At least one of the three N-linked oligosaccharide units was shown to be in the high mannose form at the cell surface, as indicated by its susceptibility to endo-beta-N-acetylglucosaminidase H digestion. Treatment of the early and late forms of Thy-1 antigen with endo-beta-N-acetylglucosaminidase F generated a single polypeptide of Mr = 13,500. The same precursor was obtained when cells were labeled in the presence of tunicamycin. This indicates the absence of O-linked glycan in the mature cell surface antigen. Finally, the resistance of Thy-1 antigen to trypsin digestion when associated with membranes confirmed that this molecule has no cytoplasmically oriented portion.

Glycosyl-inositolphospholipid (GPL) anchoring structures are incorporated into GPL-anchored proteins immediately posttranslationally in the rough endoplasmic reticulum, but the biochemical and cellular constituents involved in this "glypiation" process are unknown. To establish whether glypiation could be achieved in vitro, mRNAs generated by transcription of cDNAs encoding two GPL-anchored proteins, murine Thy-1 antigen and human decay-accelerating factor (DAF), and a conventionally anchored control protein, polymeric-immunoglobulin receptor (IgR), were translated in a rabbit reticulocyte lysate. Upon addition of dog pancreatic rough microsomes, nascent polypeptides generated from the three mRNAs translocated into vesicles. Dispersal of the vesicles with Triton X-114 detergent and incubation of the hydrophobic phase with phosphatidylinositol-specific phospholipases C and D, enzymes specific for GPL-anchor structures, released Thy-1 and DAF but not IgR protein into the aqueous phase. The selective incorporation of phospholipase-sensitive anchoring moieties into Thy-1 and DAF but not IgR translation products during in vitro translocation indicates that rough microsomes are able to support and regulate glypiation.

The biochemical features of a membrane antigen detected by a mouse monoclonal antibody (A1) raised against the murine thymoma cell line EL4 are described. This reagent detected a novel disulfide-linked 90,000 mol. wt dimeric membrane glycoprotein composed of two chains of approx 45,000 mol. wt. Endo-beta-N-acetylglucosaminidase F digestion generated a single 28,000 polypeptide, thus suggesting that the A1 molecule is a homodimer. No structural homology between the A1 molecule and the human T 90/44 protein (9.3 antigen) could be revealed by peptide mapping analysis. In view of the fact that three polypeptides of mol. wts 28,000-30,000, 21,000 and 15,000 respectively co-precipitated with the A1 antigen, the possible relationship of the A1 molecular complex to other known T-cell surface antigens including the antigen receptor is discussed.

In an effort to derive monoclonal antibodies (mAb) which can induce production of macrophage-activating factor (MAF) by cloned murine cytolytic T lymphocyte (CTL) lines, we have fused spleen cells from a rat immunized with a CTL clone with the nonsecreting mouse myeloma X63-Ag8.653. Three mAb (designated I-22, III-5 and V-8) were found to stimulate MAF production by the immunizing CTL clone and (with a single exception) two other unrelated CTL clones. However, none of these mAb inhibited the cytolytic activity of the clones. Immunoprecipitation studies indicated that the three mAb reacted primarily with a 25-30-kDa protein which could not be distinguished from that precipitated by either a reference anti-Thy-1.2 mAb or a polyclonal rabbit anti-Thy-1 antiserum. Moreover, competition binding experiments demonstrated that the three mAb competed with each other and with the reference anti-Thy-1.2 mAb. Flow cytofluorometric analysis of the strain distribution of the molecules defined by the mAb revealed that two of the antibodies (I-22 and III-5) were directed against nonpolymorphic determinants of Thy-1, whereas V-8 mAb reacted only with Thy-1.2+ lymphocytes. One of the mAb (III-5) was also able to stimulate proliferation and interleukin 2 secretion by normal splenic T cells. Since mAb directed against a number of other surface structures on CTL clones did not stimulate MAF production, it thus appears that Thy-1 (or molecules associated with Thy-1) may play a functional role in T lymphocyte triggering.

Recent evidence shows that mature Thy-1 glycoprotein lacks amino acids 113-143 predicted from the cDNA sequence and is anchored to the plasma membrane by a phosphatidylinositol-containing glycolipid attached to amino acid 112. Previously characterized Thy-1-deficient mutant lymphoma lines of complementation classes A and E were analysed. They make detergent binding Thy-1 precursors but, in contrast to wild-type, the detergent binding moiety cannot be removed by phospholipase C. Moreover, tryptophan which only occurs at position 124 is incorporated into mutant but not parental Thy-1. This suggests that the mutants make a Thy-1 precursor of 143 amino acids but fail to replace its C-terminal end by a glycolipid anchor.

In the present study, we examined the biosynthesis and the maturation of two distinct membrane glycoproteins detected by two monoclonal antibodies (RL388 and RL119), which were selected on the basis of their ability to stimulate the production of interleukin 2 by a subline of the murine EL4 thymoma. RL388 detected a disulfide-linked heterodimer complex (Mr = 130,000) composed of a glycosylated heavy (Mr = 86,000) and a nonglycosylated light (Mr = 39,000) subunit. The unglycosylated precursor of the heavy chain was a polypeptide of Mr = 57,500, which was converted upon maturation into a Mr = 73,000 core-glycosylated intermediate, and then into the Mr = 86,000 surface-expressed molecule. Partial endo-H digestion of the core-glycosylated form suggested the presence of four N-linked glycan units. The antibody reacted with a protein determinant expressed on the mature form as well as the unglycosylated precursor of the heavy chain. Moreover, both subunits assembled rapidly during biosynthesis, and the glycosylation of the heavy chain was not required for this association. Taken together, these data suggest that the antigen detected by RL388 may be the murine homologue of the human 4F2 antigen. The antigen identified by RL119 was a surface glycoprotein of Mr = 55,000 with three to five N-linked glycan units. The unglycosylated precursor polypeptide was of Mr = 29,000. The fully core-glycosylated form of Mr = 41,000, which was detected after inhibition of glucosidase I with 1-deoxynojirimycin, was converted into a Mr = 39,000 intermediate, and upon further trimming, into a Mr = 36,000 endo-H-sensitive form. The latter could be detected for chase periods of over several hours, thus suggesting a low rate of intracellular processing. The wide cellular distribution of the molecules identified by RL388 and RL119 and their preferential expression on the surface of growing cells suggests that they may be associated with cell activation events.

The Pgp-1 glycoprotein was identified on a minor (27%) subset of peripheral Lyt-2+ or L3T4+ T cells. In contrast, mature medullary-type thymocytes (Lyt-2+ L3T4-, Lyt-2- L3T4+) were nearly devoid of cells expressing detectable surface Pgp-1. The appearance of peripheral Pgp-1- T cells was found to be thymus dependent, as demonstrated by the diminished proportion of Pgp-1- T cells after thymectomy and their virtual absence in athymic nude mice. The subsequent acquisition of surface Pgp-1 was found to be a stable differentiation event occurring concomitantly with primary antigenic stimulation; selected Pgp-1- mature T cells from thymus or periphery acquired constitutive expression of Pgp-1 after stimulation in vitro with alloantigen or mitogens. These observations were extended by studies in vivo showing that immunization with various antigens augmented the percentage of Pgp-1+ spleen cells within the Lyt-2+ subset. Furthermore, the frequencies of antigen-specific CTLp, after immunization by any of three different antigens tested, were greatly enriched in the Pgp-1+ compared with the Pgp-1- subpopulations. Peritoneal exudate Lyt-2+ cells, after a localized allograft rejection, demonstrated a particularly prominent Pgp-1+ subpopulation (78%) that contained virtually all the allospecific cytolytic activity. A model consistent with all of these data proposes that mature thymocytes lacking surface Pgp-1 upon emigration to the periphery acquire its expression at the time of primary antigenic stimulation. Hence, expression of Pgp-1 among peripheral T cells is an important differentiation marker for identifying antigen-stimulated memory T cells.