Surface markers of T cells causing lethal graft-vs-host disease to class I vs class II H-2 differences

Information was sought on the phenotype of lymphoid cells causing lethal graft-vs-host disease (GVHD) in irradiated mice expressing whole or partial H-2 differences. In all strain combinations tested, pretreating donor lymph node (LN) cells with anti-Thy-1 monoclonal antibodies (MAb) plus complement (C) abolished mortality. With GVHD directed to class I H-2 differences, pretreating LN cells with anti-Lyt-2 MAb prevented mortality, whereas MAb specific for Ly-1 or L3T4 cell surface determinants caused severe mortality. These data imply that lethal GVHD directed to class I H-2 differences is mediated by L3T4-, Lyt-2+ cells; this subset of T cells was shown previously to control GVHD directed to multiple minor histocompatibility antigens, i.e., antigens seen in the context of self-class I molecules. With whole H-2 differences, GVHD appeared to be controlled largely but not exclusively by L3T4+, Lyt-2-T cells. This T cell subset was also the predominant cause of GVHD directed to class II differences. With class II incompatibilities, depleting donor cells of L3T4+ T cells, either by pretreatment with anti-L3T4 MAb + C or by fluorescence activated cell sorter selection, greatly reduced but did not completely abolish GVHD. These data might imply that L3T4-, Lyt-2+ cells have some capacity to elicit anti-class II GVHD. A more likely possibility, however, is that the residual GVHD to class II differences observed with Lyt-2+-enriched cells reflected minor contamination with L3T4+ cells.     

Synergistic effect of murine cytomegalovirus on the induction of acute graft-vs-host disease involving MHC class I differences only. Analysis of in vitro T cell function

The development of acute graft-vs-host disease (GVHD) is a common outcome after the injection of fully MHC disparate parental T cells into unirradiated F1 mice. Murine cytomegalovirus (MCMV) infection has been previously shown to augment the development of acute GVHD in the parent-into-F1 (P----F1) model, such that 10-fold fewer parental cells are required. In the present study, we have investigated the effect of MCMV infection on the induction of non-lethal GVHD that occurs in P----F1 combinations involving MHC class I only or class II only differences. Using P----F1 combinations involving either an H-2K only difference or an H-2D only difference, MCMV infection of F1 mice 3 days before the injection of parental spleen cells led to a profound T cell immunodeficiency that strongly resembled that observed in acute GVHD. Further studies examining the H-2K disparate P----F1 combination, C57Bl/6---- (C57Bl/6xB6.C-H-2bm1) F1 and combined MCMV infection showed that the immunodeficiency is characterized by a profound loss of in vitro Th cell production of IL-2 and an intrinsic defect in T effector function as shown by an inability of rIL-2 to restore defective CTL responses. Additional experiments in these mice revealed the presence of suppressor cells as well as significant parent-anti-F1 CTL activity possibly accounting for the suppressor effect. This pattern of immunodeficiency was not seen after the administration of either MCMV or MHC class I disparate parental cells alone. MCMV infection did not detectably alter the immunodeficiency observed in a P----F1 combination involving a MHC class II difference only. These results indicate that MCMV infection can alter the pattern of GVHD in the setting of an MHC class I disparity, but not in the setting of class II disparity, such that it resembles acute GVHD. These results may have relevance to the human transplant setting where intercurrent CMV infection has been associated with an adverse clinical outcome.     

L3T4-positive T cells participate in the induction of graft-vs-host disease in response to minor histocompatibility antigens

The phenotype of T cells that initiate graft-vs-host disease (GVHD) in response to minor histocompatibility antigens (minor HA) was determined in three H-2 compatible strain combinations by using negative selection with monoclonal antibodies to Lyt-2 and L3T4 antigens to test the hypothesis that Lyt-2-positive T cells alone initiate GVHD. The phenotype of T cells required to initiate GVHD was different in each of the three strain combinations studied. Both Lyt-2+ and L3T4+ LP spleen cells were necessary to cause lethal GVHD in C57BL/6 recipients. In the reciprocal transplant, Lyt-2+, but not L3T4+ C57BL/6 spleen cells were sufficient to initiate GVHD in LP recipients. In contrast, L3T4+, but not Lyt-2+ B10.D2 spleen cells were found to initiate GVHD in BALB/c recipients. The optimal response to minor HA requires both Lyt-2+ and L3T4+ T cells because a mixture of the two subsets of spleen cells resulted in a more severe form of GVHD than either subset alone in all three strain combinations studied. This study demonstrates that L3T4+ cells participate in the initiation of GVHD in response to minor HA. The dominant T cell subset that initiates GVHD varies with the specific strain combination tested. The specific minor HA expressed in the transplant recipient, the H-2 type, and possibly non-major histocompatibility complex immune response genes of the donor strain appear to determine the phenotype of the initiator T cells.     

H-2 homology requirements for secondary cell-mediated lympholysis and miced lymphocyte reactions to TNP-modified syngeneic lymphocytes.

Secondary cell-mediated lympholysis (CML) and mixed lymphocyte reactions (MLR) were generated in a tissue culture system against trinitrophenyl (TNP)-modified murine syngeneic spleen cells. H-2 homology between primary and secondary TNP-modified stimulating cells was required in order to restimulate in the secondary CML. Strong proliferative responses (MLR) were detected only in the secondary cultures, for which H-2 homology was also required between TNP-modified primary and secondary immunogens. Intra-H-2 mapping for the secondary MLR indicated that the relevant regions of homology were I, D, and K and/or I-A. Homology throughout the entire major histocompatibility complex or at K plus I-A gave stronger MLR than did cultures in which there was homology between the primary and secondary phases at I or D only.     

Cyclosporine blocks the induction of class I and class II MHC products in mouse kidney by graft-vs-host disease

The ability of cyclosporine to prevent the increase in Ia and H-2K expression that occurs in mice with graft-vs-host disease (GVHD) was examined by means of absorption, indirect immunofluorescent staining (IIF), and indirect immunoperoxidase staining (IIP). Acute GVHD was induced in irradiated C3H/HeJ mice (H-2k) by injections of bone marrow and spleen cells from C57BL/6J mice (H-2b). Ten days after induction of acute GVHD, the spleens of mice not receiving cyclosporine expressed only donor Ia, reflecting their reconstitution by donor cells. The kidneys of such mice had a 10-fold increase in host Ia and H-2K expression, as previously reported. Treatment with cyclosporine reduced the amount of donor Ia and H-2K in spleens, and prevented the enhanced expression of recipient Ia and H-2K in kidneys in a dose-dependent manner. IIF or IIP staining showed that the principal change was in kidney tubules, where the induction of Ia and H-2K expression was greatly diminished. Cyclosporine administered to normal mice did not alter Ia expression except at high doses, at which it decreased Ia expression in kidneys and in spleens. The results suggest that prevention of enhanced MHC product expression could be part of the immunosuppressive actions of cyclosporine.     

Tissue graft rejection in mice

The influence ofH-2 subregions on graft survival in a liver slice-tokidney bed grafting system has been investigated.H-2K-region andH-2IA-region donor-recipient differences, either individually or in concert, cause acute graft rejection.H-2D-region donor-recipient differences cause chronic immunological reaction as evaluated by histological criteria. Grafts across this barrier may ultimately be rejected or may survive indefinitely. Several possible explanations for the variation in survival are proposed. The remaining knownH-2 regions (IB, IC, S, andG) all appear to cause immunological reactivity in a recipient animal which differs from the liver tissue donor at any of these regions. However, only anIC-region difference may ultimately cause complete graft destruction following an extended chronic immunological course. Grafts across background histocompatibility barriers of several genetic types show rejection patterns equivalent to those seen acrossK andIA barriers. These patterns are unchanged, whether or not the donor and recipient are congenic forH-2 alleles. DifferentH-2 allelic donor-recipient differences do, however, show different times of survival, indicating variation in strength or number of donor antigens or differences in recipient immune response.     

Target antigens determine graft-versus-host disease phenotype

Chronic graft-vs-host disease (cGVHD) is an increasingly frequent complication of allogeneic stem cell transplantation. Phenotypically, cGVHD differs from patient to patient; in particular, a subset of patients develops extensive cutaneous fibrosis. Similarly, graft-vs-host disease (GVHD) is distinct in inbred murine donor:recipient pairings, indicating a genetic component to disease phenotype. The B10.D2 -->BALB/c (H-2d) strain pairing uniquely recapitulates key pathologic features of fibrotic human cutaneous cGVHD. To distinguish whether this genetic component is due to differences in genes that modulate immune responses or to the specific Ags targeted, we asked whether skin-dominant cGVHD also develops in the B10 -->BALB.B (H-2b) and B10.BR -->BALB.K (H-2k) MHC-congenic pairings. Because each MHC haplotype presents different peptides and selects different T cell repertoires, GVHD in each donor:recipient pair undoubtedly targets different Ags. We found that, in contrast to BALB/c recipients, BALB.B mice never manifested skin disease while BALB.K mice developed a modified form of skin disease. Instead, BALB.B and BALB.K recipients developed systemic GVHD which was absent in BALB/c mice. Moreover, in (B10 x B10.D2)F1 -->(BALB.B x BALB/c)F1 H-2b/d transplants, recipients developed both cutaneous and systemic disease. Thus, the selection of immunodominant Ags determines the target and character of GVHD, providing insight into the genetic basis for different forms of GVHD.     

A comparison of the neonatal tolerance-inducing capacities of H-2 class I and class II antigens

We have investigated the abilities of murine major histocompatibility complex-encoded antigens to induce in vitro hyporeactivity of T lymphocytes when these antigens are injected neonatally. Class I molecules, presented on F1 donor cells having an H-2 K or D region difference from recipients, can readily induce tolerogen-specific cytotoxic T cell hyporeactivity; as few as 1 X 10(6) neonatally injected donor cells suffice. In contrast, class II molecules, presented on F1 donor cells having an H-2 I region difference from recipients, can induce tolerogen-specific helper T cell hyporeactivity only when at least 1 X 10(7) neonatally injected donor cells are used, and then only in some of these recipients. Results from another in vitro assay system, taken in conjunction with these data, indicate that the molecular class of the tolerizing disparity, rather than the effector function of the responding cell type assayed, may be the most important factor in controlling the ease with which neonatally induced alloantigen tolerance can be achieved. In each type of tolerance described here, the hyporeactivity seen is antigen specific, in its induction and its expression; the implications of this fact for considerations of possible mechanisms of tolerance maintenance are discussed.     

Cellular and antigenic requirements for production of mixed leukocyte reaction suppressor factor

When murine spleen cells, alloantigen-sensitized previously in vivo, are incubated with spleen cells bearing the sensitizing alloantigens, a supernatant factor is produced that inhibits 3H-thymidine incorporation by responding lymphocytes in the mixed leukocyte reaction. This study evaluates the cellular and antigenic requirements during restimulation for elaboration of this suppressor factor, MLR-TsF. BALB/c spleen cells, sensitized to C57BL/6 (B6) alloantigens in vivo, produced MLR-TsF when cultured with B6 spleen cells in vitro, despite depletion of Sephadex G-10-adherent cells from factor-producing cells, stimulator cells, or from both populations. T cells were not required within the stimulating population, but a requirement for viable stimulator cells was demonstrated when heat-killed or glutaraldehyde-fixed stimulator cells failed to induce MLR-TsF production. The alloantigenic requirements for MLR-TsF production were addressed by 2 approaches. Treatment of stimulator cells with appropriate anti-I region antisera and complement did not affect MLR-TsF production, demonstrating that an absolute requirement for cells expressing I region determinants did not exist. However, spleen cells primed against entire H-2 haplotype differences produced significant quantities of MLR-TsF when they were restimulated with spleen cells homologous to the priming cells in only the I region, in the K and D regions, or in the D region alone. The additive nature of subregion-specific restimulation suggests that distinct subpopulations of K, I, and D region-specific MLR-Ts comprise the MLR-Ts population primed to entire H-2 haplotype differences.     

Activation or suppression of bactericidal activity of macrophages during a graft-versus-host reaction against I-A and I-J-region differences, respectively

Systemic graft-versus-host reactions (GVHR) were induced in F1 heterozygous mice by injecting 10(8) parental lymphocytes. The Anti-Thy 1.2-sensitive, T-cell mediated activation of macrophages was assessed by their increased capacity to destroy a facultative intracellular bacterium Listeria monocytogenes. The difference in MHC regions causing a GVHR that induced high levels of macrophage activation mapped to I-A. In contrast, differences at K or D, in any of the other H-2 subregions or in the non-H-2 background, including Mls alone or in combination, did not induce a GVHR leading to macrophage activation, unless these differences were combined with a difference at I-A. The numbers of parental cells needed to activate macrophages via a GVHR caused by I-A vs. non-I-A differences, varied at least 30- to 100-fold. When parental cells were injected into F1 offspring of parents differing at I-J, growth of Listeria was enhanced significantly; this negative effect on macrophages was not seen when parental combinations differing at I-A alone were compared with those differing at I-A plus I-J or I-J plus other H-2 regions.     

Cell-mediated immunity in herpes simplex virus-infected mice: H-2 mapping of the delayed-type hypersensitivity response and the antiviral T cell response

An adoptive transfer system was used to investigate the H-2 restriction of delayed-type hypersensitivity (DTH) to herpes simplex virus. A successful DTH transfer was achieved when donor and recipient were compatible at the I-A region, with K and D region compatibility unnecessary. However, the rapid clearance of infectious virus from the inoculation site was found only when the donor and recipients were compatible at H-2K (and presumably D) and I-A regions.     

Keratinocytes synthesize Ia antigen in acute cutaneous graft-vs-host disease

Keratinocytes express la antigen (Ia) during cutaneous graft-vs-host disease (GVHD); it is, however, unclear whether this Ia is adsorbed from alloactivated donor lymphocytes or from Ia-bearing host Langerhans cells (LC), or whether it is actively synthesized by host keratinocytes. We therefore sought to determine the origin of keratinocyte Ia in a murine model of GVHD. Lethally irradiated C3H/He (H-2k) mice developed characteristic histopathologic changes of acute cutaneous GVHD 7 days after injection of BALB/c (H-2d) bone marrow and spleen cells, and expressed keratinocyte Ia of host (Iak) but not donor (Iad) origin in immunofluorescence studies. To determine whether the Ia was synthesized by keratinocytes or adsorbed from host LC, we investigated GVHD that was induced in chimeric mice. Parental strain A mice were made chimeric by lethal irradiation and reconstitution with (A X B)F1 bone marrow cells as follows: (BALB/c X C3H/He)F1 (H-2d,k) leads to C3H/He (H-2k), B6C3F1 (H-2b,k) leads to C57BL/6 (H-2b), and B6C3F1 (H-2b,k) leads to C3H/He (H-2k). After 3 mo, the LC in the skin of these chimeric mice were mainly of F1 haplotype. The chimeric mice were again lethally irradiated and injected with marrow and spleen cells from a third strain of mouse (C57BL/6, H-2b or BALB/c, H-2d) histoincompatible with both F1 parental strains. In the ensuing GVHD, the chimeric recipients only expressed keratinocyte Ia syngeneic to the original haplotype of the animal (strain A), despite the fact that the majority of their LC were derived from F1 marrow and expressed Ia of both F1 parental strain haplotypes (strains A and B). Together, these findings indicate that keratinocyte Ia in GVHD is synthesized by keratinocytes and is not derived from donor lymphocytes or adsorbed from host LC.     

The generation of cytolytic activity in mixed leukocyte cultures: cortisone-resistant thymocytes are deficient in helper T lymphocytes

Cortisone-resistant (CR) thymocytes did not generate cytolytic activity toward H-2 K or D alloantigen unless they were also stimulated by H-2 I or non-H-2 alloantigens, even though spleen cells generated brisk cytolytic activity toward H-2 K or D alone. Backstimulation by stimulating strain T lymphocytes accounted for neither the response of spleen cells toward H-2 K or D alloantigen nor the response of CR thymocytes to a full set of alloantigens. In addition, lack of non-T accessory cells did not account for the CR thymocyte pattern of reactivity. Rather, CR thymocytes appeared to be relatively deficient in helper T lymphocytes (HTL). CR thymocytes generated specific cytolytic activity toward H-2 D alloantigen when T cell growth factors (TCGF) or cloned alloreactive helper T lymphocytes were added to culture. CR thymocytes contained fewer HTL precursors detected at limit dilution than spleen cells did. Thus spleen cells generated cytolytic activity toward class I alloantigens alone, but under the same culture conditions CR thymocytes had to be stimulated by both class I and class II alloantigens. Class II alloantigens may be required to stimulate cytolytic activity only under culture conditions in which class I-reactive HTL are not sufficient to provide a minimal threshold of help.     

H-2-restricted graft-versus-host reaction: Foreign determinants and restriction elements

Nylon-wool-purified T cells from radiation chimeras cause a lethal graft-versus-host reaction (GVHR) in irradiated, bone-marrow-protected recipients only if the recipient shares a restriction element with the T-cell donor and also expresses antigens foreign to the donor. Class I molecules (H-2K and H-2D) can act as restriction elements, but restriction to class II molecules could not be demonstrated. However, class II molecules as well as H-2K and some non-H-2 determinants could serve as foreign antigens.     

Flagellin, a TLR5 agonist, reduces graft-versus-host disease in allogeneic hematopoietic stem cell transplantation recipients while enhancing antiviral immunity

Graft-versus-host disease (GVHD) is a major cause of morbidity and mortality in patients treated with allogeneic hematopoietic stem cell transplantation (HSCT). Posttransplant immunosuppressive drugs incompletely control GVHD and increase susceptibility to opportunistic infections. In this study, we used flagellin, a TLR5 agonist protein (～50 kDa) extracted from bacterial flagella, as a novel experimental treatment strategy to reduce both acute and chronic GVHD in allogeneic HSCT recipients. On the basis of the radioprotective effects of flagellin, we hypothesized that flagellin could ameliorate GVHD in lethally irradiated murine models of allogeneic HSCT. Two doses of highly purified flagellin (administered 3 h before irradiation and 24 h after HSCT) reduced GVHD and led to better survival in both H-2(b) → CB6F1 and H-2(K) → B6 allogeneic HSCT models while preserving >99% donor T cell chimerism. Flagellin treatment preserved long-term posttransplant immune reconstitution characterized by more donor thymic-derived CD4(+)CD25(+)Foxp3(+) regulatory T cells and significantly enhanced antiviral immunity after murine CMV infection. The proliferation index and activation status of donor spleen-derived T cells and serum concentration of proinflammatory cytokines in flagellin-treated recipients were reduced significantly within 4 d posttransplant compared with those of the PBS-treated control recipients. Allogeneic transplantation of radiation chimeras previously engrafted with TLR5 knockout hematopoietic cells showed that interactions between flagellin and TLR5 expressed on both donor hematopoietic and host nonhematopoietic cells were required to reduce GVHD. Thus, the peritransplant administration of flagellin is a novel therapeutic approach to control GVHD while preserving posttransplant donor immunity.     

Evaluation of in vitro cytotoxic T lymphocyte assays as a predictive test for the occureence of graft vs host disease

The potential value of in vitro cytotoxic T lymphocyte (CTL) assays for predicting the occurrence of graft vs host disease (GVHD) following allogeneic bone marrow transplantation was evaluated in 12 mouse donor-host combinations associated with various degrees of GVHD. These donor-host combinations were selected after evaluation of GVHD triggered by minor histocompatibility antigens (MiHA) in 24 allogeneic strain combinations derived from six strains of H-2 ^b mice. Recipients (n=475), previously submitted to total body irradiation (9.5 Gy), were transplanted with 10⁷ bone marrow cells along with 5 x 10⁷ spleen cells. While lethal GVHD was observed in half of the strain combinations, it was possible to select 12 donor-host combinations characterized by severe, mild, or absent GVHD. When levels of anti-host CTL activity were assessed following in vivo priming and in vitro boosting, strong CTL-mediated cytotoxicity was observed in all combinations wheteer they developed GVHD or not. CTL frequency measured by limiting dilution analysis (LDA) ranged from 1/16880-1/306. The Spearman rank test revealed no positive correlation between GVHD intensity and donor anti-host CTL activity assayed either in bulk culture experiments or in LDA conditions. These results indicate that MiHA capable of triggering potent CTL responses in vitro do not necessarily initiate GVHD, and that in vitro measurement of donor CTL activity against host-type Con A blasts is not a predictive assay for anti-MiHA GVHD. However, the possibility to recruit CTL populations targeting host MiHA expressed specifically on hematopoietic cells suggests a novel therapeutic strategy for the cure of hematopoietic malignancies. Indeed, transplantation of donor hematopoietic stem cells supplemented with T cells aimed at MiHA specifically expressed by host hematopoietic cells, could possibly potentiate the desirable graft vs leukemia effect without increasing the risk of GVHD.     

Recombinant haplotype H-2 in mice of the congenic resistant strain B10.D1(R108)/Y

Recombinant H-2 haplotype of mouse strain B10.D1(R108)/Y (symbol R108) obtained in experiments with skin grafting in the course of developing the CR B10.D1/Y strain (strain DBA/LacY--the donor of H-2q) was studied. Strains with recombinant H-2 haplotypes a, h2, g1, i3, i5, i7, m, y1 were used. Alleles of different H-2 (K, I, D) regions were determined according to the presence or absence of genetic complementation in the F1 test with skin grafts. R108 recombinant was studied by serological methods with panel of anti-H-2 sera. Anti-H-2Kb (H-2.33) and anti-H-2Dq (H-2.30) monospecific antisera were used in microcytotoxicity test and in absorption experiments in vitro. It was concluded that crossing over between H-2b and H-2q chromosomes, which led to formation of recombinant H-2 haplotype of R108 mice, occurred at I region, between IA and IC subregions. The H-2 complex of R108 line has KbIAbIJ?IE?ICqSqDq alleles. bq1 symbol was proposed for the H-2 haplotype of B10.D1(R108)/Y strain.     

Changes in Ia expression in mouse kidney during acute graft-vs-host disease

We induced graft-vs-host disease (GVHD) in mice to determine whether immunologic stimuli could alter renal Ia expression. Two strain combinations were used: B6.C-H-2bm12 into C57BL/6, an I-A mutation difference, and A.SW into A.TL, differing in the I and D regions of H-2. By day 10 after allogeneic reconstitution of lethally irradiated recipients with bone marrow and spleen cells, the recipients had developed acute GVHD, as measured by their spleen to body weight ratio. Histologic examination revealed focal interstitial infiltrates of mononuclear cells in the kidneys. The expression of host Ia in these kidneys was increased up to 10-fold, as measured by absorption, and indirect immunofluorescence indicated that certain renal tubule cells had become strongly positive, suggesting that these were the principal sites of the increase in Ia expression. Similar increases were not observed in donor Ia. Tubule cells may have become Ia positive by passive uptake, or more probably, by the increase of Ia biosynthesis in cells that usually synthesize little or no Ia. Lethal irradiation without reconstitution tended to decrease renal Ia expression, as assessed by absorption and immunofluorescence. The results indicate that renal Ia expression, particularly in renal tubules, can be altered by changes in the immune system, raising the possibility of a role for such altered Ia expression in autoimmune or alloimmune responses involving the kidney.     

Synergistic interactions between lymph node and thymus cells in response to antigenic differences limited to selected regions of the H-2 complex.

Thymus and lymph nodes from the A.TL recombinant line were utilized as sources of responding cells in MLR (mixed lymphocyte response) assays to MHC-determined (major histocompatibility complex) antigenic differences. Cells from both sources were stimulated to proliferate by antigenic determinants controlled by the H-2K region alone, H-2D region and the H-2I-H-2S regions. Nylon-fiber-adherent splenic cells from each of the stimulating cell strains stimulated T-cell-dependent responses. Synergistic interactions between A.TL thymus and lymph node cells were initiated by antigenic products limited to single H-2 regions. Antigenic differences determined within the H-2I region were not required for synergistic responses to H-2K-controlled products or for the generation of cytotoxic killer cells to H-2D-associated antigens. The H-2I-region-associated products also were very effective in stimulating T-cell synergy. These data demonstrate that the two responsive T-cell subpopulations can both be stimulated by alloantigens coded within a single known H-2 region.     

The effect of H-21J disparity on induction of allotransplantation tolerance by Lentil Lectin

The effect of an H-21J disparity on skin graft survival was studied in 18 mouse donor-recipient strain combinations, in which the recipients were treated with an efficient immunosuppressant, lentil lectin (LCA). The simultaneous I-J disparity essentially had no (or a slightly adverse) effect on the graft survival times in strain combinations differing at the K and I-A loci or in the entire H-2 complex. In two strain combinations incompatible at the D locus, the simultaneous I-J disparity promoted graft survival. The disparity at the I-J locus therefore seems to have only a marginal effect on the survival of allografts in most of the LCA-treated recipients, but it may promote graft survival in some animals. A similar tolerance-promoting effect was also observed with D disparity.