Allosuppressor- and allohelper-T cells in acute and chronic graft-vs.-host (GVH) disease. III. Different Lyt subsets of donor T cells induce different pathological syndromes.

Previous work from this laboratory has led to the hypothesis that the stimulatory pathological symptoms of chronic graft-vs.-host disease (GVHD) are caused by alloreactive donor T helper (TH) cells, whereas the suppressive pathological symptoms of acute GVHD are caused by alloreactive T suppressor (TS) cells of the donor. In the present paper we analyzed the Lyt phenotypes of B10 donor T cells required for the induction of either acute or chronic GVHD in H-2-different (B10 X DBA/2)F1 recipients. First, nonirradiated F1 mice were used as the recipients. We found that unseparated B10 T cells induced only a moderate formation of systemic lupus erythematosus (SLE)-like autoantibodies, but a high percentage of lethal GVHD (LGVHD). In contrast, Lyt-1+2- donor T cells were unable to induce LGVHD in these recipients; these cells were capable, however, of inducing a vigorous formation of SLE-like autoantibodies and the formation of severe immune-complex glomerulonephritis. Lyt-1-2+ T cells were incapable of inducing either acute or chronic GVHD. In another experiment, the sensitivity and accuracy of the GVH system were increased by using irradiated F1 mice as recipients and by comparing donor-cell inocula that contained similar numbers of T lymphocytes. In addition, donor-cell inocula were used that had been tested for their allohelper and allosuppressor effects on F1 B cells in vitro. In the irradiated F1 recipients, too, unseparated donor T cells were superior to T cell subsets in inducing LGVHD; Lyt-1-2+ donor cells were completely and Lyt-1+2- donor cells were almost incapable of doing so. In contrast, Lyt-1+2- T cells, but neither unseparated T cells nor Lyt-1-2+ T cells, were capable of inducing a vigorous formation of SLE-like auto-antibodies. We conclude that the stimulatory pathological symptoms of chronic GVHD are caused by Lyt-1+2- allohelper T cells. In contrast, the development of the suppressive pathological symptoms of acute GVHD appears to involve alloreactive Lyt-1+2+ T suppressor cells.

The autoantigen-binding B cell repertoires of normal and of chronically graft-versus-host-diseased mice.

A quantitative analysis of the frequencies of autoantibody-producing B cells in GVHD and in normal mice has been undertaken by generating collections of hybridomas of activated B cells. These hybridomas secreted sufficient quantities of Ig to allow binding analyses on a panel of autoantigens. B cells have been activated in a variety of ways. In vivo they were activated by injection of alloreactive T cells of one parent, leading to GVHD by a foreign antigen, sheep erythrocytes, in a secondary response, or by the polyclonal activator LPS. B cells from an experimentally unstimulated animal were used for an analysis of the normal background. In vitro B cells were activated by alloreactive T cells or by LPS. The frequencies of hybridomas and, therefore, of activated B cells producing autoantibodies to DNA or to kidney were not significantly different in mice activated by a graft-vs.-host T cell response as compared with B cell populations activated by any of the other procedures. They were found to compose 7.1-17.1% of the total repertoire of activated B cells. Moreover, the frequencies of autoantibody-producing activated B cells does not change with time after induction of the graft-vs.-host reaction. The pattern and frequencies of autoantigen-binding specificities to cytoskeleton, smooth muscle, nuclei, mitochondria, and DNA were not found to be different in any of the groups of hybridomas. The single notable exception, found in GVHD mice, were hybridomas producing autoantibodies to kidney proximal tubular brush border. These results allow the conclusion that autoantigen-binding B cells exist in an activated state in GVHD mice, as well as in mice activated by a foreign antigen or by a polyclonal activator, in B cell populations activated in vitro either by alloreactive T cells or by a polyclonal activator, and even in the background of experimentally unstimulated animals. T cell-mediated graft-vs.-host activation, in large part, does not lead to a selective expansion of autoantigen-binding B cells. The main difference between the graft-vs.-host-activated B cell repertoire and all others is that approximately 90% of teh autoantibodies were of the IgG class, whereas al autoantibodies found in the other groups were IgM.

Monoclonal antibodies reactive with the mouse interleukin 5 receptor.

The rat mAbs R52.120 and R52.625 inhibit the action of IL-5 on both IL-5-sensitive cell lines and freshly isolated splenic B lymphocytes. Neither antibody inhibits the proliferative cell responses promoted by IL-2, IL-3, or IL-4. Purified R52.120+ lymphoid spleen cells contain 15-20-fold higher numbers of B lymphocytes responding to IL-5 in the form of maturation into antibody-producing cells. By immunofluorescence staining and flow fluorocytometry, the R52.120 and R52.625 antibodies bound to all 12 IL-5-sensitive cell lines tested. Both antibodies react with 2-4% cells in the spleen, 5% lymphoid cells, and 10-15% myeloid cells in the bone marrow, and 10-14% in the peritoneum of C57BL/6, DBA/2, and BALB/c adult mice. No positive cells for either antibody were detected in the thymus and lymph nodes of these mice. Both R52.120 and R52.625 antibodies specifically inhibit the binding of radiolabeled IL-5 to its receptor. Finally, R52.120 and R52.625 antibodies precipitate from 35S-methionine-labeled IL-5-R+ cell lysates three proteins with Mr 46,000, 130,000, and 140,000. Taken together from these results, we conclude that the R52.120 and R52.625 mAbs recognize epitopes on the IL-5-R complex very close or identical to the IL-5 binding sites.

Allosuppressor and allohelper T cells in acute and chronic graft-vs.-host disease. II. F1 recipients carrying mutations at H-2K and/or I-A.

By induction of a graft-vs.-host reaction (GVHR) in nonirradiated H-2-different F1 mice, one can induce stimulatory pathological symptoms, such as lymphadenopathy and hypergammaglobulinemia, combined with the production of autoantibodies characteristic of systemic lupus erythematosus (SLE). Alternatively, the GVHR can lead to the suppressive pathological symptoms, such as pancytopenia and hypogammaglobulinemia, characteristic of acute GVH disease (GVHD). Whether stimulatory or suppressive symptoms are induced by a GVHR depends, in our view (2-4), on the functional subset of donor T cells activated in the F1 host. The purpose of the present study was to investigate whether class I and/or class II H-2 alloantigens can selectively trigger, out of a pool of unselected donor T cells, those subpopulations of T cells responsible for the stimulatory and suppressive GVH symptoms, respectively. For the induction of the GVHR, 10(8) lymphoid cells from C57BL/6 (B6) donors were injected into three kinds of F1 hybrid mice, which had been bred from H-2 mutant strains on a B6 background. Whereas the I-A-disparate (B6 X bm12)F1 recipients exclusively developed stimulatory GVH symptoms, including SLE-like autoantibodies and immune complex glomerulonephritis, the K locus-disparate (B6 X bm1)F1 recipients showed neither clearly stimulatory nor clearly suppressive GVH symptoms. In marked contrast, the (bm1 X bm12)F1 recipients, which differ from the B6 donor strain by mutations at both K and I-A locus, initially developed stimulatory GVH symptoms, but rapidly thereafter showed the suppressive pathological symptoms of acute GVHD and died. Moreover, spleen cells obtained from (B6 X bm12)F1 mice injected with B6 donor cells helped the primary anti-sheep erythrocyte (SRBC) response of normal (B6 X bm12)F1 spleen cells in vitro, whereas spleen cells (bm1 X bm12)F1 mice injected with B6 donor cells strongly suppressed the primary anti-SRBC response of normal (bm1 X bm12)F1 spleen cells. Spleen cells from the K locus-disparate (B6 X bm1)F1 recipients also suppressed the primary anti-SRBC of normal (B6 X bm1)F1 spleen cells; this suppression, however, was weak when compared with the suppression induced by spleen cells from GVH (bm1 X bm12)F1 mice. Taken together, these findings indicate that a small class II (I-A) antigenic difference suffices to trigger the alloreactive donor T helper cells causing SLE-like GVHD. In contrast, both class I (H-2K) and class II (I-A) differences are required to trigger the subsets of donor T cells responsible for acute GVHD. It appears that alloreactive donor T helper cells induce the alloreactive T suppressor cells, which then act as the suppressor effector cells causing the pancytopenia of acute GVHD. These findings may help to understand the variability of GVH-like diseases caused by a given etiologic agent, their cellular pathogenesis, and association with certain HLA loci.

Precursor B cell receptor-dependent B cell proliferation and differentiation does not require the bone marrow or fetal liver environment.

The capacity of precursor B (pre-B) I cells from fetal liver and bone marrow to proliferate and differentiate into surface immunoglobulin-positive immature B cells in vitro was analyzed. Both fetal liver- and bone marrow-derived progenitors do so in a pre-B cell receptor (pre-BCR)-dependent manner in tissue culture medium alone, without addition of other cells or cytokines. Approximately 20% of the initial pre-B I cells enter more than one division. Analyses at the single-cell level show that approximately 15% divide two to five times. Coculture of pre-B I cells with stromal cells did not enhance proliferation or differentiation, whereas the presence of interleukin 7, especially in combination with stromal cells, resulted mainly in the expansion of pre-B I cells and prevented their further differentiation. Thus, the environment of fetal liver or bone marrow is not required for the pre-BCR to exert its function, which is to select and expand cells that have undergone an inframe V(H)-D(H)J(H) rearrangement that produces a pre-BCR-compatible muH chain. It appears unlikely that a ligand for the pre-BCR drives this pre-B cell proliferation.

Development of autoimmune disease in SCID mice populated with long-term "in vitro" proliferating (NZB x NZW)F1 pre-B cells.

Pre-B cell lines proliferating for several months on stromal cells in the presence of interleukin 7 (IL-7) were established from fetal liver of (NZB x NZW)F1 mice. They express the B lineage-specific markers PB76, B220, and VpreB, but do not express surface immunoglobulin (sIg). Upon removal of IL-7 from the culture, they differentiate to sIg+ B cells that can then be stimulated by lipopolysaccharide to become IgM-secreting cells. Transfer of these pre-B cell lines into SCID mice leads to hypergammaglobulinemia of IgM (600-900 micrograms/ml), IgG2a (1-3 mg/ml), and IgG3 (300-500 micrograms/ml) for the next 3-5 mo. The spleen appears populated with (NZB x NZW)F1-derived pre-B cells, few B cells, and many IgM and/or IgG-producing plasma cells. In contrast, SCID mice populated with pre-B cell lines of normal (C57BL/6 x DBA/2)F1 mouse fetal liver develop normal levels of serum IgM (approximately 100-300 micrograms/ml), almost no detectable levels of IgG, and no plasma cell hyperplasia. The (NZB x NZW)F1 pre-B cell-populated SCID mice contain elevated serum titers of IgG antinuclear autoantibodies, but no retroviral gp70-specific nor erythrocyte-specific autoantibodies. Up to 20% of the SCID mice develop proteinuria as a consequence of IgG deposits in the kidney glomeruli during a 7-mo period of observation. All signs of autoimmune disease seen in these mice are independent of the sex of the SCID host. This experimental system provides a distinction between the disease-determining (NZB x NZW)F1 genes, which are expressed in the B lymphocyte lineage and cause the development of the disease, from those expressed in other cell lineages which only modulate its progression.

Capacity of genetically different T lymphocytes to induce lethal graft-versus-host disease correlates with their capacity to generate suppression but not with their capacity to generate anti-F1 killer cells. A non-H-2 locus determines the inability to induce lethal graft-versus-host disease.

When comparing, in a murine model, the kind of graft-versus-host (GVH) disease (GVHD) induced by the donor strain DBA/2 on the one hand and several H-2-congenic resistant B10 donor strains on the other, we found that strain DBA/2 was a universal nonkilling GVH donor for H-2-incompatible nonirradiated F1 hybrid recipients. In this respect, DBA/2 T cells differed from those of the H-2-identical donor strain B10.D2 as well as those of other b10 donor strains. The inability of strain DBA/2 to kill by GVH reaction was not limited to certain H-2 incompatibilities in the F1 recipients, but was nonspecific. The inability to kill is determined by a dominant locus not linked to H-2. DBA/2 T cells were also incapable of inducing the severe suppression of hematocrit values, bone marrow erythropoiesis, thymic cell proliferation, and splenic IgG production in the F1 recipients that was observed after the injection of T cell from the B10 strains. However, DBA/2 T cells, in contrast with those of the B10 donor strains, were vigorous stimulators of IgG production in H-2-incompatible F1 hybrid recipients. Surprisingly, strain DBA/2 as well as the B10 donor strains had good capacity to generate anti-F1 TK cells. Taken together, these findings raise the possibility that lethal GVHD disease is not caused, or not caused exclusively, by donor killer T cells, but by those donor T cells that directly or indirectly induce a suppression of cell proliferation in certain vital organs of the recipient.

Intrinsic B cell defects in NZB and NZW mice contribute to systemic lupus erythematosus in (NZB x NZW)F1 mice.

We have previously shown that long-term in vitro proliferating fetal liver pre-B cell lines derived from autoimmune-prone (NZB x NZW)F1 (BW) mice, but not normal (B6 x DBA2)F1 mice, can differentiate in severe combined immunodeficient (SCID) mice to produce elevated levels of serum immunoglobulin (Ig) M and IgG, and high titers of antinuclear antibodies The contribution of parental NZB and NZW strains to B cell abnormalities of BW hybrid mice was investigated here by preparing pre-B cells and transferring them into immunodeficient SCID- and RAG-2-targeted mice. We show that transfer of NZB pre-B cells led to a marked IgM hypergammaglobulinemia and to the production of limited amounts of IgG2a. On the other hand, the transfer of NZW pre-B cell lines led to moderately elevated IgM levels and marked hypergammaglobulinemia of IgG2a. High IgM and low IgG anti-DNA titers are found in the recipients of NZB pre-B cells, whereas those receiving NZW pre-B cells contained lower levels of IgM and high titers of IgG anti-DNA. In marked contrast, essentially identical titers of antibodies directed against a non-self-antigen, DNP, are found in all group of pre-B cell recipients. Thus, B-lineage cells of both NZB and NZW parental strains manifest abnormalities associated with the development of this lupus-like disease. Therefore, the present study strongly suggests a complex inheritance of B cell abnormalities in autoimmune-prone (NZB x NZW)F1 mice and emphasizes the critical importance of intrinsic B cell defects in the development of murine systemic lupus erythematosus.

Allosuppressor and allohelper T cells in acute and chronic graft-vs-host disease. I. Alloreactive suppressor cells rather than killer T cells appear to be the decisive effector cells in lethal graft-vs.-host disease.

Splenic T cells from B10 donors were injected into irradiated (B10 x DBA/2)F1 mice. Either 5 or 6 d later, activated donor T cells were recovered from the spleens of these primary F1 (1 degree F1) recipients and transferred to groups of nonirradiated syngeneic F1 (2 degrees F1) recipients. Whereas day-5-activated parental T cells induced the characteristic symptoms of acute graft-vs.-host disease (GVHD) and eventually lethal GVHD, day-6-activated B10 T cells failed to induce acute GVHD but induced symptoms of chronic GVHD. Interestingly, the inability of day-6-activated T cells to induce lethal GVHD could not be ascribed to a lack in anti-F1 T killer cells. The combined results of functional studies indicated that day-6 cells were enriched for alloreactive helper T cells, whereas day-5 cells were enriched for alloreactive suppressor cells. Hence, our findings indicate that acute GVHD and lethal GVHD are caused by alloreactive donor T suppressor but not T killer cells, and that symptoms of chronic GVHD are caused by alloreactive donor T helper cells.

Diseases caused by reactions of T lymphocytes towards incompatible structures of the major histocompatibility complex. VI. Autoantibodies characteristic of systemic lupus erythematosus induced by abnormal T-B cell cooperation across I-E.

By induction of a suitable graft-vs-host reaction (GVHR) in H-2-different F1 mice, one can induce the production of autoantibodies characteristic of systemic lupus erythematosus (SLE). The purpose of the present study was to define the intra-H-2 differences in the F1 recipients that are capable of triggering this process. A GVHR was induced in [B10.A(2R) x B10.A(4R)]F1 mice by injecting 10(8) lymphocytes from either parental strain. Whereas the donor B10.A(4R) induced a massive formation of autoantibodies to thymocytes, erythrocytes, nuclear antigens, and double-stranded DNA, the donor B10.A(2R) failed to do so. The intra-H-2 genetics of these two parent leads to F1 combinations are such that the observed autoantibody formation after the injection of B10.A(4R) T cells must have been triggered exclusively by the incompatible I-Ek subregion of the [B10.A(2R) x B10.A(4R)]F1 recipients. Because I-E appears to be the murine analogue of HLA-D/DR, this finding is of interest with respect to the increased frequency of certain HLA-DR alleles in SLE patients, as discussed.

Activated murine B lymphocytes and dendritic cells produce a novel CC chemokine which acts selectively on activated T cells.

Genes were isolated using the suppression subtractive hybridization method by stimulation of pro/pre B cells with anti-CD40 and interleukin (IL)-4 to mature S mu-Sepsilon-switched cells. One of the strongly upregulated genes encodes a novel murine CC chemokine we have named ABCD-1. The ABCD-1 gene has three exons separated by 1. 2- and 2.7-kb introns. It gives rise to a 2.2-kb transcript containing an open reading frame of 276 nucleotides. Two polyadenylation sites are used, giving rise to cDNAs with either 1550 or 1850 bp of 3' untranslated regions. The open reading frame encodes a 24 amino acid-long leader peptide and a 68 amino acid-long mature protein with a predicted molecular mass of 7.8 kD. ABCD-1 mRNA is found in highest quantities in activated splenic B lymphocytes and dendritic cells. Little chemokine mRNA is present in lung, in unstimulated splenic cells, in thymocytes, and in lymph node cells. No ABCD-1 mRNA is detected in bone marrow, liver, kidney, or brain, in peritoneal exudate cells as well as in the majority of all unstimulated B lineage cells tested. It is also undetectable in Concanavalin A-activated/IL-2-restimulated splenic T cells, and in bone marrow-derived IL-2-induced natural killer cells and IL-3-activated macrophages. Recombinant ABCD-1 revealed a concentration-dependent and specific migration of activated splenic T lymphoblasts in chemotaxis assays. FACS(R) analyses of migrated cells showed no preferential difference in migration of CD4(+) versus CD8(+) T cell blasts. Murine as well as human T cells responded to ABCD-1. Freshly isolated cells from bone marrow, thymus, spleen, and lymph node, IL-2-activated NK cells, and LPS-stimulated splenic cells, all did not show any chemotactic response. Thus, ABCD-1 is the first chemokine produced in large amounts by activated B cells and acting selectively on activated T lymphocytes. Therefore, ABCD-1 is expected to play an important role in the collaboration of dendritic cells and B lymphocytes with T cells in immune responses.

Selection events operating at various stages in B cell development.

B cells have to progress through various checkpoints during their process of development. The three transcription factors E2A, EBF (early B cell factor) and Pax5 play essential roles in B cell commitment checkpoints. The various forms of the BCR and their downstream signaling molecules, which are expressed at different stages of B cell development, act as critical checkpoint guards allowing (positive selection) or preventing (negative selection) developmental progression. The recent advances on the molecular mechanisms operating at these various checkpoints are here summarized and discussed.

Lineage commitment in lymphopoiesis.

The mechanisms controlling the commitment of hematopoietic progenitor cells to the lymphoid lineages are still mostly unknown. Recent findings indicate that the earliest phase of B cell development may proceed in two steps. At the onset of B-lymphopoiesis, the transcription factors E2A and EBF coordinately activate the B-cell-specific gene expression program. Subsequently, Pax5 appears to repress the promiscuous transcription of lineage-inappropriate genes and thus commits progenitor cells to the B-lymphoid pathway by suppressing alternative cell fates. B-lineage commitment by Pax5 seems to occur in a stochastic manner in the bone marrow, as indicated by the random activation of only one of the two Pax5 alleles in early pro-B cells. In contrast, loss- and gain-of-function analyses have implicated the Notch1 receptor in the specification of the T cell fate, which may thus be controlled by instructive signals in the thymus.

Targeted inactivation of mouse RAD52 reduces homologous recombination but not resistance to ionizing radiation.

The RAD52 epistasis group is required for recombinational repair of double-strand breaks (DSBs) and shows strong evolutionary conservation. In Saccharomyces cerevisiae, RAD52 is one of the key members in this pathway. Strains with mutations in this gene show strong hypersensitivity to DNA-damaging agents and defects in recombination. Inactivation of the mouse homologue of RAD52 in embryonic stem (ES) cells resulted in a reduced frequency of homologous recombination. Unlike the yeast Scrad52 mutant, MmRAD52(-/-) ES cells were not hypersensitive to agents that induce DSBs. MmRAD52 null mutant mice showed no abnormalities in viability, fertility, and the immune system. These results show that, as in S. cerevisiae, MmRAD52 is involved in recombination, although the repair of DNA damage is not affected upon inactivation, indicating that MmRAD52 may be involved in certain types of DSB repair processes and not in others. The effect of inactivating MmRAD52 suggests the presence of genes functionally related to MmRAD52, which can partly compensate for the absence of MmRad52 protein.

Autoreactive B-cell repertoire in mice with chronic graft versus host disease.

A quantitative analysis of the frequencies of autoantibody producing B-cells has been undertaken by producing and analyzing random hybridoma collections generated in fusions with activated B-cells. Activated B-cells were derived from mice injected with LPS and SRBC and normal mice. They were compared to those derived from mice undergoing chronic GVHD. The frequencies of successful fusion events correlate well with the number of activated B-cells used in the fusions, so that it is reasonable to conclude that the hybridoma collections reflect the activated B-cell repertoires in the different animals. The frequencies of hybridomas producing autoantibodies as well as their specificities for self-antigens, were not significantly different between the different collections of hybridomas. Moreover, no difference in VH gene family expression was found in the different collections of autoantibody producing hybridomas. So, the activated autoreactive B-cell repertoires in GVHF1 mice and in normal mice is similar. In contrast to the normal activated autoreactive B-cell repertoires, which make predominantly IgM antibodies, the GVH-activated autoreactive B-cells make predominantly antibodies of the IgG class. Therefore, we conclude that T-cell mediated graft versus host activation does not generally lead to selective expansion of autoreactive B-cells, but appears to play a crucial role in the switch from IgM to IgG production.

Pax5 determines the identity of B cells from the beginning to the end of B-lymphopoiesis.

Despite being one of the most intensively studied cell types, the molecular basis of B cell specification is largely unknown. The Pax5 gene encoding the transcription factor BSAP is required for progression of B-lymphopoiesis beyond the pro-B cell stage. Pax5-deficient pro-B cells are, however, not yet committed to the B-lymphoid lineage, but instead have a broad lymphomyeloid developmental potential. Pax5 appears to mediate B-lineage commitment by repressing the transcription of non-B-lymphoid genes and by simultaneously activating the expression of B-lineage-specific genes. Pax5 thus functions both as a transcriptional repressor and activator, depending on its interactions with corepressors of the Groucho protein family or with positive regulators such as the TATA-binding protein. Once committed to the B-lineage, B cells require Pax5 function to maintain their B-lymphoid identity throughout B cell development.

Age-dependent changes in B lymphocyte development in man and mouse.

In recent years, detailed analyses of B cell development in both humans and mice have revealed similar subsets of precursors along the same pathway of differentiation. From these studies it also became clear that both species undergo age related changes in this B lymphocyte development program. In this review we summarize these findings and discuss, potential mechanisms underlying these age related changes, and possible causative correlations between these changes and age related B cell abnormalities.