Monoclonal antibodies to RT7 and LCA antigens in the rat: cell distribution and segregation analysis

Fluorescein-conjugated monoclonal antibodies (mAb) have been used to examine the cellular distribution and genetic relationship of the RT7 and the leukocyte common antigen (L-CA) in the rat. We demonstrate here that the RT7.1 alloantigen, recognized by the mAb BC84, is present on B lymphocytes as well as T lymphocytes, although at markedly different apparent concentrations; T cells binding approximately 4 times more BC84 than do B cells. In contrast, the polymorphic L-CA recognized by the NDS58 mAb and the non-polymorphic L-CA recognized by the OX1 mAb appear to be expressed in equal concentrations on T and B cells and differ in their tissue distribution from the RT7.1 alloantigen. We have also tested a new mAb termed 8G6.1, that is reactive with RT7b rat strains and has a cell, tissue and strain distribution profile resembling that of a polymorphic L-CA. Segregation analysis of the RT7 and polymorphic L-CA antigenic systems using a single backcross model demonstrate that the alloantigens recognized by BC84 and NDS58 cosegregate and that both are allelic with respect to 8G6.1. The L-CA antigenic determinant defined by OX1 was non-polymorphic and thus was not allelic with any of the antigenic systems tested. These results suggest that there is a close genetic relationship between the expression of RT7 and L-CA in the rat.     

Evidence from capping experiments for independence of the RT7 alloantigen and the leucocyte common antigen in the rat

Experiments reported here demonstrate that the RT7 alloantigen and the L-C antigen are separate and distinct structures on the surface of rat lymphocytes. The distribution of the antigens in different rat strains, including the mutant WF/fz, clearly establish the RT7 antigenic system as a polymorphic diallelic system, whereas the recognized L-C antigenic determinant is monomorphic and present on the lymphocytes of all rat strains tested. These data were obtained using monoclonal antibodies to the antigens in indirect immunofluorescence experiments. The two antigens were shown to redistribute (cap) independently of one another on the surface of rat thymocytes. Cells that had been exposed to anti-L-C antibody and FITC-conjugated anti-Ig followed by anti-RT7.1 antibody and RITC-conjugated anti-Ig demonstrated FITC caps and RITC rings.     

Diabetes-prone BB rats express the RT6 alloantigen on intestinal intraepithelial lymphocytes.

The RT6 alloantigen of the rat is expressed on most peripheral T cells but not on thymocytes and thus represents a marker for postthymic T lymphocyte maturation in this species. Diabetes-prone (DP) BB rats exhibit a genetically determined T cell lymphopenia associated with a deficiency of RT6+ T cells. In this study we have analyzed the expression of RT6 on lymph node (LN) cells and intestinal intraepithelial lymphocytes (IEL) in two DP BB strains (BB/OK and BB/Mol) and two control strains (non-lymphopenic BB/PhiK and LEW) by flow cytometry. In the DP BB rats the number of LN T cells was substantially reduced (less than 25% TcR2+ cells) and completely lacked RT6 expression. The IEL population was also reduced in number and in marked contrast to normal rats consisted predominantly of CD4+ cells. The majority of IEL, however clearly expressed RT6. Treatment with a phosphatidylinositol (PI)-specific phospholipase C markedly reduced the RT6 density showing that PI-mediated anchoring of RT6 in the cell membrane also applies to IEL of DP BB rats. The results demonstrate that the DP BB strains possess a functional RT6 gene and are also able to generate the PI anchor. The defect in RT6 expression is thus unlikely to be the primary cause of the T cell lymphopenia.     

SHORT COMMUNICATION: MOUSE MONOCLONAL ANTIBODY AGAINST Lyt - 1.1 ALLOANTIGEN *

A cloned hybridoma, producing an IgM kappa antibody, has been derived from a fusion between NSI myeloma cells and spleen cells from BALB/c mice immune to CBA lymphocytes. Its specificity for Lyt-1.1 alloantigen has been established from (1) the lymphoid tissue distribution restricted to T cells and (2) the reactivity pattern on a panel of thymocytes from standard and Lyt congenic mouse strains.     

Production of a rat T cell hybridoma that stably expresses the T cell differentiation marker RT6.2

The rat T cell alloantigen RT6.2 shows a slow rate of synthesis in isolated T cells which hampers studies on the metabolism of RT6.2 in these cells (1). In order to facilitate further molecular and functional characterization of this molecule we have established a T-T hybridoma cell line which stably expresses RT6.2. Only 1 of 102 T cell hybridomas obtained upon fusion of the rat thymoma C58NT with DA rat lymph node cells expressed this antigen. This clone--EpD3--initially showed a relatively slow rate of cell division and a highly unstable pattern of expression of RT6.2. Thus, the relative number of RT6.2 bearing cells consistently decreased during cultivation of EpD3 and of EpD3-derived subclones. Cell separation studies suggest that the switch in RT6.2 phenotype of EpD3 cultures was due to the appearance of variant cells of RT6.2- phenotype with a higher proliferative capacity and was not induced by factors in the culture medium. By repeated panning and subcloning procedures to select for RT6.2 expressing cells, a subline of EpD3--EpD3/87--was obtained which stably expresses high levels of RT6.2. Metabolic labeling studies of RT6.2 in EpD3 show that it is synthesized very efficiently in these cells and support our previous suggestion that RT6.2 does not bear any classic oligosaccharide side chains. Moreover, RT6.2 can be released almost completely from EpD3 cells by phosphatidylinositol-specific phospholipase C (PI-PLC).     

Dendritic cells internalise and re-present conformationally intact soluble MHC class I alloantigen for generation of alloantibody

Following organ transplantation soluble MHC class I is released from the graft and may contribute to alloimmunity. We determined in a well-established rat model whether DC are able to internalise soluble MHC class I alloantigen and then re-present intact alloantigen to B cells and T cells for generation of an alloantibody or CD8 T cell response. PVG.RT1(u) BM-derived DC internalised (via an active process) and retained intact a recombinant soluble form of RT1-A(a) (sRT1-A(a)). When PVG.RT1(u) rats were immunised with sRT1-A(a)-pulsed syngeneic DC, they developed a strong anti-sRT1-A(a) alloantibody response and showed accelerated rejection of RT1-A(a)-disparate PVG.R8 heart grafts. Alloantibody production and accelerated heart graft rejection were both dependent on immunisation with viable sRT1-A(a)-pulsed DC. The alloantibody response to sRT1-A(a)-pulsed DC was directed exclusively against conformational epitopes expressed by sRT1-A(a) and not epitopes expressed, for example, by non-conformational sRT1-A(a) heavy chain. Immunisation with sRT1-A(a)-pulsed syngeneic DC did not stimulate a CD8 T cell response. Our findings suggest a novel alloantigen recognition pathway whereby soluble MHC class I alloantigen released from an allograft may be taken up by recipient DC and presented in an intact unprocessed form to B cells for the generation of an alloantibody response.     

RT7-Defined Alloantigens in Rats are Part of the Leucocyte Common Antigen Family

Haemopoietic cells carry a variety of cell-surface molecules, some of which are known to have allotypic variation. In rats, the RT7 alloantigenic system has been well documented using alloantisera. We have produced the first mouse hybridomn cell line secreting an antibody. H1S41, which binds to leucocytes of rat strains carrying the RT7.2 but not the RT7.1 determinant. An lgG2b isotype switch variant (HIS4l.2b) of the original HIS41 (IgG1 isotype) was also made. HIS41 showed a clear and discrete binding in immunofluorescenl and histological experiments and has already been used in several studies on haemopoietic cell turnover and differentiation employing PVG rats congenic for RT7. The present study addresses the question of whether the RT7 gene products are members of the L-CA family, which has been a matter of controversy over the last decade. When using HIS41 for the analysis of tissue distribution and molecular weight of RT7 gene products, a strong similarity was evident with the data reported for the L-CA detected by MRC OX-1 and MRC OX-30. These two MoAb have been reported to bind to all members of the L-CA family. All haemopoietic cells, excluding erythrocytes and the more mature stages of erythropoiesis, stained with HIS41. The molecular weights of HIS41 binding molecules on thymocytes and peripheral T cells were comparable to the L-CA precipitated by MRC OX-1. Capping and sequential immunoprecipitation studies indicated that HIS41 and MRC OX-30- binding molecules were identical. MRC OX-1. however, appeared to bind only a subset of these molecules. Thus, our study confirms the identity of RT7.2 gene products and L-CA. lt also revealed a difference between MRC OX-1 and MRC OX-30 not noticed previously.     

Alloantigen pretreatment induces a down-regulation of the in vivo subsequent development of cytotoxic T lymphocytes directed against linked alloantigens

In the present study, we describe a new regulatory system that influences the in vivo development of cytotoxic T lymphocytes (CTL) and that could be related to epitopic suppression. Epitopic suppression has been previously shown to occur when carrier-primed mice are subsequently immunized with a "new" epitope coupled to the priming carrier. The suppression specifically inhibited the antibody response to the "new" epitope without affecting the secondary antibody response to the carrier. In this report, using a carrier/hapten-carrier type of immunization protocol, we have demonstrated that a similar regulatory system could also affect the induction of CTL directed against allogeneic cells. Priming mice with an alloantigen 1 (carrier) inhibits the induction of alloantigen 2 (hapten)-specific cytotoxic responses when the alloantigen 2 is presented in association with the alloantigen 1 on an F1 stimulator cell (hapten-carrier conjugate). This has been demonstrated by the specific decrease of anti-H-2b or anti-H-2d CTL responses generated in C3H/He mice (H-2k) previously primed with, respectively, H-2d or H-2b spleen cells before immunization with F1 (H-2d x b) spleen cells. This suppression of the CTL responses against the second immunizing alloantigen is associated with a strong CTL response against the first priming alloantigen. The induction of the suppression is dependent on the dose of H-2d spleen cells administered before immunization with F1 spleen cells and is not related to antigen elimination since a strong suppression of the CTL response against H-2b antigens is shown following immunization with a mixture of F1 cells and H-2b-bearing cells of H-2d-primed animals.     

Monoclonal antibodies to the murine Ly-2.1 cell surface antigen.

Six monoclonal anti-Ly-2.1 antibodies are described that arose from a single fusion using the spleen from a 129/ReJ mouse immunized with CBA/H thymus cells. The specificity was determined from the strain distribution, and testing of congenic strains where all six monoclonal antibodies detected the Ly-2.1 alloantigen. Furthermore, the antibodies fall into two groups: Group I (four antibodies) detected the expected number of Ly-2+ thymocytes and peripheral T cells, whereas Group II (two antibodies) detected 10% fewer peripheral T cells. Functional studies demonstrated that the Ly-2.1 determinant detected by a Group I antibody (49-31.1) was presented on cytotoxic T cells (Ly-1+2+), on concanavalin A (Con A)-induced suppressor T cells (Ly-1-2+), but absent from helper T cells (Ly-1+2-). One antibody (49-11.1) precipitated a 68,000-75,000 mol.wt glycoprotein, which on reduction yielded 30,000 and 35,000 mol.wt moieties. Thus, by functional, genetic and biochemical criteria these antibodies detected the Ly-2.1 specificity. In addition, a monoclonal, noncytotoxic, IgGl, anti-Thy-1.2 antibody is described.     

Surface IgG-mediated mixed lymphocyte culture and cell-mediated lympholysis responses by immunocompetent thymus cells

Radiolabeled preparations of murine surface immunoglobulin have been shown to bind selectively to immunocompetent responder thymus cells. Such bound immunoglobulin (especially IgG) facilitate the mixed lymphocyte culture and cell-mediated lympholysis responses between immunocompetent thymus-cell subsets. Papain digest of IgG also promote alloantigen recognition between immunocompetent thymocytes; most of the activity residues in the Fc piece. The binding of IgG per se does not activate responder thymocytes. Alloantigen-specific MLC and CML responses in vitro appear to require the participation of alloantigen receptors on the responder cell, the expression of alloantigens by the stimulator thymus cells and the binding of B-cell immunoglobulin by the immunocompetent responder thymocytes. When any of these processes are interrupted an alloantigen-specific response is not generated in vitro.     

Monoclonal antibody definition of the DRB3 allele, HLA-Dw25

Monoclonal antibodies are powerful tools for analyzing HLA antigen polymorphism. We have investigated the serological and biochemical nature of the DRw52-related antigen defined by the monoclonal antibody NDS10. A detailed analysis of the population distribution of NDS10 reactivity revealed that the epitope was present on a subpopulation of DRw52 positive cells. A distinct pattern of reactivity was found within DR3 individuals: all of the B18,DR3 cells were NDS10 positive, whereas the A1,B8,DR3 cells were negative. All of the DR5(w11) cells and two of three DRw12 cells reacted with NDS10. NDS10 reactivity with DRw6 was not restricted to either of the serologically defined subtypes; three of 17 DRw13 and nine of 10 DRw14 cells were NDS10 positive. NDS10 was unreactive with all of the DRw8 cells tested. Two-dimensional gel analyses revealed that the NDS10 molecule precipitated from DR3, DR5(w11) and DRw6(w14) cell lines had an identical beta chain profile. These data indicate that NDS10 recognises the Dw25 allele of the DRw52 complex.     

Tissue distribution and quantitation of la-like antigens in the rat

The tissue distribution of a major histocompatibility complex (MHC)-linked alloantigen in the rat was studied using antiserum raised by alloimmunization in a MHC-congenic pair of rat strains. HO (Ag-B5) rats were immunized with HO.B2 (Ag-B2) skin grafts, and the tissue distribution of antigens recognized by the resultant antiserum was determined by cellular radioimmunoassay using different HO.B2 tissues as targets or immunoabsorbent. Linkage to the rat MHC was formally established by backcross analysis. The results obtained indicated that the tissue distribution of antigens recognized by this antiserum was closely similar to that reported for the distribution of Ia antigens in the mouse. Thus, T cells and thymocytes bound, respectively, 2% and 10% as much antibody as peripheral B cells while no binding to erythrocytes was detectable. Quantitative immunoabsorption studies revealed that thymocytes and kidney tissue had, respectively, 20% and 2% of the absorbing capacity of spleen cells. However, both tissues were capable of completely absorbing out the anti-B cell activity showing that they both possessed all the Ia-like specificities detectable by this antiserum on peripheral B cells. A variety of procedures aimed at reducing the passenger leukocyte content of kidneys had only a slight effect on the absorbing capacity. The cellular radioimmunoassay results indicated that on B cells, Ia-like molecules are a major membrane component, present in amounts comparable with surface immunoglobulin molecules. Studies at the single cell level, using a fluorescence-activated cell sorter, indicated that the expression of Ia-like alloantigens by rat thymocytes was very heterogeneous and that large thymocytes were the most heavily labeled. No Ia-like alloantigens were detected on alloantigen-activated T cells. The findings are compared with published data on the mouse.     

A human thymocyte antigen defined by a hybrid myeloma monoclonal antibody.

Spleen cells from a BALB/c mouse that had been immunized with human thymocytes were fused with the myeloma line P3-NS 1/1 Ag 4.1. One of the resulting hybrid clones (NA 1/34) secreted an antibody that was highly specific for human thymocytes. Eighty-five % of thymocytes expressed the antigen designated HTA1. There were an estimated 15 x 10(4) molecules of HTA 1 per cell, and it is therefore a major surface molecule. The expression of this antigen on thymocytes appears to be reciprocal to HLA, as recognized by another monoclonal antibody W6/32. Immunoprecipitated material from [125I]-labeled thymocyte membranes was analyzed by polyacrylamide gel electrophoresis in sodium dodecyl sulfate which disclosed a single component of 45,000 molecular weight.     

Thymic epithelial defects and predisposition to autoimmune disease in BB rats.

We report an association between thymic epithelial defects and predisposition to autoimmunity. Diabetes-prone (DP) BB rats develop spontaneous hyperglycemia and are deficient in T cell subsets expressing the RT6 alloantigen. Diabetes resistant (DR) BB rats become diabetic if depleted of RT6+ T cells. The inciting immune system defects are unknown. We made the following observations: 1) Regions of thymic cortex and medulla devoid of thymic epithelium exist in DP-BB, DR-BB, and Lewis rats, all of which are susceptible to autoimmune disorders. Such defects were absent in eight normal rat strains. 2) Thymic epithelial defects are absent at birth, but present in BB rats at 4 weeks of age. 3) The genetic predisposition to thymic epithelial defects is an autosomal dominant trait. 4) The observation of thymic defects in (DP x WF)F1 rats led to the prediction that such animals, which never develop spontaneous autoimmunity, might be susceptible to its induction. Following depletion of RT6+ T cells we observed diabetes in 91%, and thyroiditis in 43%, of treated F1 animals (n = 23). Pancreatic insulitis was uniformly present. Because thymic epithelium participates in the positive and negative selection of developing thymocytes, we propose that thymic epithelial defects may play an important role in the predisposition of BB rats to autoimmunity.     

Ontogeny and immunohistochemical localization of thymus-dependent and thymus-independent RT6+ cells in the rat.

RT6 is a cell surface alloantigen that identifies a regulatory subset of peripheral T cells in the rat. Diabetes-prone BB rats are deficient in peripheral RT6+ T cells and develop spontaneous autoimmune insulin-dependent diabetes mellitus. Diabetes-resistant BB rats have normal numbers of RT6+ T cells, and insulin-dependent diabetes mellitus can be induced in these animals by in vivo depletion of peripheral RT6+ cells. Athymic rats are also severely deficient in peripheral RT6+ T cells. Although very different with respect to the peripheral RT6+ cell compartment, normal, athymic, and diabetes-prone BB rats all generate RT6+ intestinal epithelial lymphocytes (IELs). The goal of these studies was to analyze the ontogeny of RT6+ IELs and peripheral lymphoid cells by in situ immunohistochemistry. We observed the following. 1) RT6+ IELs appear before alpha(beta) T-cell-receptor- expressing IELs in diabetes-prone BB, diabetes-resistant BB, and athymic WAG rats. 2) In vivo depletion of peripheral RT6+ T cells in diabetes-resistant BB rats using a cytotoxic monoclonal antibody is not accompanied by depletion of RT6+ IELs. 3) A population of RT6+ T-cell-receptor-negative IELs is present in normal, euthymic diabetes-resistant BB rats, constitutes a larger percentage of the euthymic but lymphopenic diabetes-prone BB rat IEL population, and is the predominant IEL phenotype in athymic WAG rats. These results suggest that RT6+ cells are composed of both thymus-dependent and thymus-independent cell subsets that have different developmental characteristics and may differ in function.     

Four distinct antigen systems on human thymus and T cells defined by monoclonal antibodies: immunohistological and immunochemical studies.

Human thymus and T cell antigens were identified by using four distinct monoclonal antibodies (MoAb), designated 2D5, 5B3, 7A5 and 9D4. 2D5 antibody reacted with most human thymocytes and a few peripheral lymphocytes as well as with a subpopulation (20%) of bone marrow cells, and precipitated a 45K molecular weight (mol. wt.) component from 125I-labelled thymus cell lysate. 7A5 antibody also reacted with the majority (80%) of thymocytes but neither with peripheral lymphocytes nor with bone marrow cells. The antigen detected by 7A5 was a glycoprotein consisting of 48K and 12K mol. wt. components, which were non-covalently associated with each other. 5B3 reacted with virtually all of human thymus and T cells but not with the majority of B cells and bone marrow cells. This reagent precipitated a 72K mol. wt. glycoprotein from thymus and T cells. An additional 65K mol. wt. glycoprotein was precipitated by 5B3 together with the 72K mol. wt. component, but with poor reproducibility. 9D4 antibody, on the other hand, reacted with a 200K mol. wt. component from thymus and T cells as well as 220K and 210K components from the non-T cell fraction of tonsil lymphocytes. Whereas antigens detected by 2D5 and 7A5 appeared to be highly expressed on cortical thymocytes, the antigen defined by 5B3 occurred much more abundantly on medullary thymocytes and peripheral T cells than on cortical thymocytes. Based on the data described above, it is suggested that 7A5, 5B3 and 9D4 MoAb recognize human homologues of mouse TL, Ly-1 and Ly-5 antigens, respectively, whereas 2D5 antibody seems to resemble OKT10, as described by others.     

Simultaneous detection of two different cell surface antigens by electron microscopy by means of multivalent hybrid antibody with double specificity

Multivalent hybrid antibody (MHA) complexes with dual specificity were prepared by combining two antibodies of different specificities, one against ferritin (Fer) the other against horseradish peroxidase (HRP), with protein A of Staphylococcus aaureus (SpA).Electron microscopy of mouse spleen lymphocytes and thymocytes (previously coated with mouse IgG anti-Thy-1 antibody) treated with IgG anti-Fer/SpA/IgG anti-mouse Ig complex and Fer gave better resolution and higher accuracy than previously obtained with IgG anti-HRP/SpA/IgG anti-mouse Ig complex and HRP (Mandache et al., 1980). Surface Thy-1 alloantigen and Fc receptor (charged with human IgG) treated with a mixture of IgG anti-Fer/SpA/IgG anti-Thy-1 and IgG anti-HRP/SpA/IgG anti-human Fab could be simultaneously detected on the thymocyte surface by either light or electron microscopy using Fer and HRP. The concomitant visualization of Thy-1 alloantigen (with Fer) and FcR (with HRP) on mouse thymocyte clearly shows that their distribution is largely independent and that the amount of Thy-1 antigen is greater.These results show that electron microscopy with a mixture of MHA is a useful technique for simultaneous location of two antigenic markers on the cell surface.     

Identification of a CD4 homologue in the cat

Monoclonal antibody, Fel 7, produced against cat T cells, was found to react with a single-chain glycoprotein of Mr 65,000 present on a majority of the thymocytes, 40% of lymph node cells, 15% of splenocytes and 25% of blood mononuclear cells. Using a previously reported antibody that recognizes the feline CD8 antigen, approximately 65% of cat thymocytes were shown to express both the Fel 7 and fCD8 antigens, while 14% and 6% expressed either the Fel 7 or the fCD8 determinant respectively. The Fel 7 and fCD8 antigens were expressed by mutually-exclusive subpopulations of peripheral T cells, and not by B cells, macrophages or other types of blood cells. Expression of the Fel 7 antigen was down-regulated and the molecule was phosphorylated when T cells were stimulated with phorbol ester, while the expression of the fCD8 antigen was unaffected by this treatment. The addition of soluble Fel 7 antibodies efficiently blocked Con A-induced proliferation of T cells in a dose-dependent manner. The data suggest that the mAb Fel 7 identifies a feline CD4 homologue, providing an important reagent for the study of normal and abnormal T cell development in cats.     

Cytokine Production and Responsiveness of Fetal T-Cell Receptor Vγ3 Thymocytes

The aim of this study was to examine the cytokine production and cytokine responsiveness of the first T-cell receptor (TcR) positive cells that appear in the murine fetal thymus, namely TcR V gamma 3 cells. It is shown that IL-2-cultured fetal TcR V gamma 3 thymocytes were capable of producing IL-3, GM-CSF, TNF-alpha and IFN-gamma upon TcR triggering. IL-2, IL-4, IL-5 and IL-6 could not be detected. With regard to cytokine responsiveness, TcR V gamma 3 cells proliferated to a high extent when high concentrations of rIL-2 were added. rIL-4 or rIL-7 alone, but not rIL-1 alone, were capable of inducing a modest proliferation of TcR V gamma 3 thymocytes. When combined with low concentrations of IL-2, a synergistic effect could be observed with IL-1, IL-4 or IL-7. It is shown that the synergistic effect of IL-2 with IL-4 was mainly due to induction of IL-2 receptor expression. The synergistic effect of IL-2 and IL-7 on the proliferation of TcR V gamma 3 cells could only be partially inhibited by anti-IL-2 receptor MoAb, and this antibody had no effect on the IL-2 + IL-1 cultures. These observations can explain the extensive proliferation of TcR V gamma 3 thymocytes during fetal life and they indicate that TcR V gamma 3 thymocytes have the potential to play a functional role during fetal thymus development.     

Deletion of alloantigen-reactive thymocytes as a mechanism of adult tolerance induction following intrathymic antigen administration

Direct injection of foreign antigen into the adult thymus is a potent route of antigen delivery for the induction of tolerance in vivo. In this report, we demonstrate that tolerance to C57BL/10 (H2^b/BL10) alloantigens can be induced in CBA/Ca (H2^k/CBA) mice by intrathymic (IT) administration of BL10 spleen leukocytes coincident with transient peripheral immunomodulation of CD4⁺ T cells using a depleting anti-CD4 monoclonal antibody. T cell receptor (TCR) transgenic mice (BM3.6; H2^k) expressing a CD8-independent TCR specific for H2K^b were used as recipients to facilitate investigation of the mechanisms responsible for tolerance induction by allowing visualization of events in the thymus following IT injection. IT administration of 5 × 10⁷ BL10 spleen leukocytes and concomitant transient peripheral T cell depletion in BM3.6 mice resulted in a substantial H2K^b-specific deletion of transgenic-TCR⁺ (tg-TCR) thymocytes which was dependent on the level of tg-TCR expression. IT deletion and the failure to export CD8⁺ T cells to the peripheral lymphoid organs correlated with the induction of tolerance to H2K^b; TCR transgenic mice that had received IT injection of BL10 splenocytes and peripheral T cell depletion accepted a H2K^b+ cardiac allograft indefinitely. Analysis of tolerant BM3.6 mice revealed that there were low numbers of CD8⁺ T cells in the periphery giving rise to a substantially reduced reactivity in vitro despite the fact that no donor cells or IT deletion were observed in the thymi of the majority of tolerant mice. These results demonstrate for the first time that IT injection of foreign alloantigen into an adult thymus results in the deletion of thymocytes expressing a TCR specific for the injected alloantigen and suggest that this is an important mechanism of tolerance induction following IT injection of alloantigen in vivo. Furthermore, analysis of tolerant TCR-transgenic mice suggests that IT deletion is not required for the maintenance of tolerance, and that peripheral mechanisms enforce continued hyporesponsiveness to H2K^b following transplantation.