Coexistence of anti-RNA polymerase III and anti-U1RNP antibodies in patients with systemic lupus erythematosus: two cases without features of scleroderma.

Anti-RNA polymerase III (RNAP III) antibodies are highly specific for scleroderma (SSc) and associated with diffuse SSc and renal crisis. Coexistence of anti-RNAP III and other SSc autoantibodies is rarely documented. We report three cases with coexisting anti-RNAP III and anti-U1RNP. Autoantibodies in 3829 sera from rheumatology clinics were screened by immunoprecipitation. Anti-RNAP III-positive sera were also examined by immunofluorescence and anti-RNAP III ELISA. In total, 35 anti-RNAP III-positive sera were identified by immunoprecipitation, in which three had coexisting anti-U1RNP. All three were anti-RNAP III ELISA positive. Two had anti-RNAP I dominant (vs. RNAP III) reactivity and showed strong nucleolar staining. A case with anti-U1/U2RNP (U2RNP dominant) had systemic lupus erythematosus (SLE)-SSc overlap syndrome; however, the remaining two cases had SLE without signs of SSc. All three cases of anti-RNAP III + U1RNP fulfilled ACR SLE criteria but none in the group with anti-RNAP III alone (p = 0.0002). In contrast, only one case in the former group had sclerodermatous skin changes and Raynaud's phenomenon, vs. 92% with scleroderma in the latter (p < 0.05). Although anti-RNAP III is highly specific for SSc, cases with coexisting anti-U1RNP are not so uncommon among anti-RNAP III positives (8%, 3/35) and may be SLE without features of SSc.

Murine lupus susceptibility locus Sle1a requires the expression of two sub-loci to induce inflammatory T cells.

The NZM2410-derived Sle1a lupus susceptibility locus induces activated autoreactive CD4(+) T cells and reduces the number and function of Foxp3(+) regulatory T cells (Tregs). In this study, we first showed that Sle1a contributes to autoimmunity by increasing antinuclear antibody production when expressed on either NZB or NZW heterozygous genomes, and by enhancing the chronic graft versus host disease response indicating an expansion of the autoreactive B-cell pool. Screening two non-overlapping recombinants, the Sle1a.1 and Sle1a.2 intervals that cover the entire Sle1a locus, revealed that both Sle1a.1 and Sle1a.2 were necessary for the full Sle1a phenotype. Sle1a.1, and to a lesser extent Sle1a.2, significantly affected CD4(+) T-cell activation as well as Treg differentiation and function. Sle1a.2 also increased the production of autoreactive B cells. As the Sle1a.1 and Sle1a.2 intervals contain only 1 and 15 known genes, respectively, this study considerably reduces the number of candidate genes responsible for the production of autoreactive T cells. These results also show that the Sle1 locus is an excellent model for the genetic architecture of lupus, in which a major obligate phenotype results from the coexpression of multiple genetic variants with individual weak effects.

High sensitivity C-reactive protein in systemic lupus erythematosus: relation to disease activity, clinical presentation and implications for cardiovascular risk.

Measurement of high sensitivity C-reactive protein (hs-CRP), has been used in the assessment of disease activity in numerous rheumatic conditions including systemic lupus erythematosus (SLE). However, the utility of hs-CRP measurement in patients with lupus is uncertain. This study examined if hs-CRP can be used to assess disease activity, severity and cardiovascular risk in SLE. Serum samples from 601 visits of 213 SLE patients and 134 controls were analysed for hs-CRP by nephelometry. Detailed demographic data were obtained from all subjects and medication history and key laboratory parameters were collected. Disease activity was assessed using the SLEDAI. High sensitivity CRP was not associated with disease activity (SLEDAI), number of ACR SLE criteria or presence of any particular organ involvement. hs-CRP levels were significantly correlated with standard cardiovascular risk factors including body weight (P = 0.0002), hypertension (P = 0.001), and apolipoprotein A-I (P < 0.0001). Interestingly an inverse correlation was seen between hs-CRP levels and antimalarial use (P = 0.0018). Our results suggest that measurement of hs-CRP, though not valuable as marker of disease activity in SLE may be of some use in the assessment of cardiovascular risk. We speculate that antimalarials may help to reduce cardiovascular risk in patients with SLE.

Aberrant expression of the NF-kappaB and IkappaB proteins in B cells from viable motheaten mice.

In viable motheaten mice, a mutation in the gene encoding the phosphatase, SHP1, causes severe immunodeficiency and autoimmunity. A defective phosphatase may result in modified phosphorylation of proteins involved in gene regulation. Since the NFkappaB/IkappaB proteins are regulated through phosphorylation, we wished to understand if the expression of these proteins was altered by the SHP1 defect. Splenic B cells from viable motheaten mice were isolated and assessed for purity by flow cytometry. Levels of each protein in isolated B cells were examined by Western blot analyses. Measurement of RNA levels for each protein was assessed by semi-quantitative RT-PCR. Western blots revealed that, in me(v) whole cell lysates, there were reduced levels of RelA and RelB proteins and increased levels of p50 and c-Rel. Furthermore, we analyzed the protein levels of IkappaBalpha and found that, in me(v), this inhibitor was significantly reduced, while the level of another member of the IkappaB family, IkappaBbeta, was not. To determine if these findings in me(v) were secondary to the autoimmune process, we evaluated NF-kappaB/IkappaB expression in the BXSB murine model of autoimmunity. Unlike me(v), B cells from BXSB/Yaa mice had NF-kappaB complexes composed of the RelA submit, and IkappaBalpha was readily detected. In addition, RNA for the RelA and IkappaBalpha proteins in me(v) and control littermates was detected by RT-PCR, indicating that the reduced amounts of these proteins was not exclusively due to transcriptional defects. We conclude that the differences in NF-kappaB/IkappaB proteins that we have described in me(v) are likely a consequences of the SHP1 defect and could contribute to the clinical disorder that characterizes me(v) mice.

Genetic dissection of SLE pathogenesis: adoptive transfer of Sle1 mediates the loss of tolerance by bone marrow-derived B cells.

Sle1 is a potent autoimmune susceptibility locus on chromosome 1 originally identified in a genome scan of testcross progeny between the systemic lupus erythematosus-prone NZM2410 strain and C57BL/6. We subsequently produced B6.NZMc1, a congenic strain carrying the NZM2410-derived Sle1 genomic interval on the B6 background and demonstrated that Sle1 mediated the loss of tolerance to chromatin in both the B and T cell compartments. In this communication, we show by adoptive transfer experiments that the autoimmune phenotypes of Sle1 are completely reconstituted in B6 radiation chimeras receiving B6.NZMc1 bone marrow but not by the reciprocal reconstitution, demonstrating that Sle1 is functionally expressed in B cells. In additional experiments, cotransfer of mixtures of bone marrow derived from B6.NZMc1 and nonautoimmune congenic B6 mice carrying allelic T and B cell markers showed that only B cells derived from B6.NZMc1 bone marrow produced anti-chromatin autoantibodies. In contrast, increased expression of CD69 was equivalent in CD4+ T cells derived from either B6.NZMc1 or congenic B6 bone marrow, suggesting that either T cell population could be activated subsequent to loss of tolerance in the B cell compartment. These findings indicate that the expression of Sle1 in B cells is essential for the development of autoimmunity.

Resistance to HgCl2-induced autoimmunity in haplotype-heterozygous mice is an intrinsic property of B cells.

Exposure to low doses of mercury chloride induces autoantibodies to the nucleolar protein fibrillarin in H-2s, but not in H-2b, mice. Surprisingly, F1 crosses between resistant and sensitive haplotypes are resistant. Previously, we have shown that the resistance in these F1 mice was due to coexpression of the resistant class II allele. Using adoptive transfer techniques we have examined several mechanisms by which the resistant haplotype could be down-regulating the antifibrillarin response in F1 (s/b) mice. Similar to other autoimmune models, mercury-induced autoimmunity requires cognate MHC-restricted T cell help. The absence of autoantibody production in F1 mice was not due to a difference in thymic education or to the absence of antifibrillarin-specific T cell help. These results suggest that the resistance is due to an intrinsic property of the haplotype-heterozygous B cells.

Functional consequences of the SHP-1 defect in motheaten viable mice: role of NF-kappa B.

To define the functional consequences of the src-homology domain-1 protein (SHP-1) defect, we examined cytokine production and NF-kappa B activity in motheaten viable (Mev) mice. We found elevated levels of interleukin-6 (IL-6), interleukin-10 (IL-10), tumor necrosis factor (TNF), and interferon-gamma (IFN-gamma) in Mev mice sera and cultured B and T cells compared to littermate control adult mice. The levels of interleukin-2 (IL-2) detected in Mev sera and activated Mev T cells were decreased, but IL-2 receptor expression was increased. We then evaluated the activity of NF-kappa B and found that this protein is highly expressed in Mev B and T cells. To determine if NF-kappa B had a role in causing the elevated levels of cytokines in Mev mice, we treated activated Mev T cells with an NF-kappa B decoy and found that cell culture treatment with the decoy resulted in significant reduction of the secretion of IL-6, GM-CSF, and TNF, but not IFN-gamma. Therefore, our data show that Mev mice secrete elevated levels of inflammatory cytokines, which can be mediators in the development of the Mev clinical disorder, and that NF-kappa B has an important role in this process, impacting upon the regulation of the immune response.

Novel immunoregulatory B cell pathways revealed by lpr-+ mixed chimeras.

lpr, a murine mutation of the Fas apoptosis receptor, causes lymphadenopathy and autoantibody production, with lymphadenopathy primarily due to a population of CD4-CD8-B220+ T cells. Previous in vivo experiments, in which lpr and normal bone marrow cells were coinfused into lpr hosts, have demonstrated that only T cells of lpr origin accumulated abnormally and only B cells of lpr origin produced autoantibodies. Moreover, in these chimeras, B cells of normal origin were unable to respond to conventional, T cell-dependent exogenous Ag. To address the role of lpr B cells in regulation of lpr autoimmunity, we have prepared lpr-+ mixed chimeras and selectively eliminated lpr B cells using allele-specific, mAb treatment, thus allowing normal B cells to develop in an environment with lpr T cells. From these data, we arrived at four major conclusions: 1) Compared with control-treated chimeric mice, lpr B cell-depleted mice had greatly reduced total lymph node cell counts; 2) the T cells were derived equally from normal and lpr donors, and the percentage of lpr-derived CD4-CD8- T cells was greatly reduced; 3) despite the presence of the remaining lpr T cells, no autoantibodies were produced by the normal derived B cells; and 4) lpr T cells without lpr B cells were unable to prevent a normal B cell response to conventional Ag. These data demonstrate that B cells can play a critical and expansive regulatory role, not only for T cells, but for other B cells as well.

Use of phosphorothioate-modified oligodeoxynucleotides to inhibit NF-kappaB expression and lymphocyte function.

NF-kappaB is a potential target for immunosuppressive therapy. Two methods were evaluated to inhibit NF-kappaB: the antisense (AS) approach in which single-stranded oligodeoxynucleotides (ODNs) bind the mRNA for the RelA subunit of NF-kappaB and the transcription factor decoy (TFD) approach in which double-stranded ODNs bind the NF-kappaB protein. AS and TFD inhibited NF-kappaB binding and decreased total IgG and anti-dsDNA antibody production in splenocytes from the BXSB/Yaa autoimmune mouse strain. TNF-alpha expression was reduced by AS and TFD, as were the levels of IL-2. But AS effects did not last beyond 24 h, whereas TFD inhibited cytokine production after 72 h. AS had no effect upon IL-6, while the TFD reduced the secretion of IL-6. Therefore, the suppression of immune response mediators by AS or TFD, through inhibition of NF-kappaB, is substantial. These inhibitors can serve as novel choices for therapy in the treatment of autoimmune disorders.

Prevention of experimental allergic encephalomyelitis by an antibody to CD45RB.

CD45 is involved in the regulation of lymphocyte activation, and it has been demonstrated that ligation of CD45 induces apoptosis of T and B lymphocytes. Recently anti-CD45RB antibody therapy was shown to block acute allograft rejection in a mouse model of transplantation. Therefore, we wanted to examine the effects of anti-CD45RB antibody treatment on the course of an autoimmune disorder, experimental allergic encephalomyelitis (EAE), a Th1-mediated process. Mice immunized with myelin basic protein and treated with anti-CD45RB antibody did not develop EAE. Histologically, there was no evidence of lymphocytic infiltrates in the central nervous system. T cell proliferation and TNF-alpha production were significantly decreased in anti-CD45RB-treated mice. Furthermore, there was a significant reduction in the production of other Th1 cytokines including interferon-gamma and IL-2, but not IL-4 or IL-6. However, levels of a number of adhesion markers or markers of activation such as VLA-4 and LFA-1 on T cells were no different in treated versus control animals. Thus, anti-CD45RB can prevent EAE and appears to do so by altering T cell proliferation and cytokine production.

Class II haplotype differentially regulates immune response in HgCl2-treated mice.

One of the most striking features of exposure to low doses of mercury in mice is the high-titer haplotype-linked anti-nucleolar (ANoA) autoantibody response. Mice of H-2(s) haplotype have been high responders, while H-2(b) mice have been low. This pattern has been attributed to the class II molecule itself, but the poor response of F1 crosses between high and low responders raised the possibility that the anti-fibrillarin specificity was actually due to a closely linked dominant negative gene. To test the role of class II explicitly, F1 crosses between congenic B6.SJL (H-2(s)) and C57BL/6 (H-2(b)) mice with a targeted deletion of I-AbAbeta were generated, creating mice heterozygous for all MHC loci, but expressing only I-As. In comparison with B6.SJL, no diminution of ANoA titers was found, proving that I-As itself was responsible for susceptibility and I-Ab for downregulation. Unlike I-A, expression of the I-E class II molecule could not downregulate the response in an otherwise susceptible mouse. These results suggest a complicated role for class II in the regulation of a novel, environmentally induced autoimmune response.

B cell genotype determines the fine specificity of autoantibody in lpr mice.

Anti-Sm Abs are specific markers of human systemic lupus erythematosus (SLE) and of murine models of this disease. In humans, anti-Sm Abs are mostly IgG1, and in MRL/lpr mice, IgG2a; both are T-dependent isotypes. Other lpr strains, such as B6/lpr, do not produce anti-Sm Ab spontaneously. The present study was aimed at identifying the cellular expression of background genes responsible for generation of the anti-Sm Ab response in MRL/lpr mice. We used double chimeric mice made by transferring MRL/lpr and B6/lpr bone marrows into irradiated allotype heterozygous F1 mice. Five mo after reconstitution, FACS analysis of lymph node (LN) and spleen cells revealed that both MRL/lpr and B6/lpr cells coexisted in roughly equal numbers. Ab produced by each donor could be distinguished by allotype-specific assays. IgG2a anti-Sm was made only by MRL-derived B cells despite the presence of T cells that might potentially provide help to the B6/lpr B cells. The frequency of anti-Sm Ab-producing individuals was similar to that of unmanipulated MRL/lpr mice (about 25%). IgG2a anti-chromatin and total IgG2a was mostly dominated by the MRL-derived B cells. B6-derived B cells produced more rheumatoid factor (RF) against their own IgG2b(b), while RF against IgG2a was dominated by MRL-derived B cells. This suggests that the control of the production of particular autoantibody specificities, such as anti-Sm, is determined by the expression of MRL or B6 background genes in B cells.

Inhibition of the p50 (NKkappaB1) subunit of NF-kappaB by phosphorothioate-modified antisense oligodeoxynucleotides reduces NF-kappaB expression and immunoglobulin synthesis in murine B cells.

NF-kappaB is a regulatory protein of immune response genes and a candidate for targeting in immunosuppressive therapy. NF-kappaB proteins are formed from components of which p50 (NFkappaB1) is a subunit. By targeting p50 gene expression with specific antisense 3' phosphorothioate-oligodeoxynucleotides (3' PS-ODNs), an effect upon NF-kappaB regulation and immunoglobulin synthesis in murine B cells was achieved. A 49% decrease in p50 protein was induced by treatment of WEHI 231 B cells with p50 antisense 3' PS-ODNs and not by control 3' PS-ODNs. p50 antisense specifically reduced the expression of NF-kappaB by 51%, but not the transcription factor, Oct-1. In the BXSB murine model of autoimmunity, p50 antisense inhibited NF-kappaB expression and total IgM and IgG synthesis, but, more importantly, dsDNA antibodies were reduced 90%. These results validate the use of p50 antisense to reduce NF-kappaB expression and, by downregulating the immune response, has application in the treatment of autoimmune disorders.

Hyperproliferation of BXSB B cells is linked to the Yaa allele.

Systemic lupus erythematosus is characterized by polyclonal B cell activation, the production of autoantibodies, and often by renal disease. Previous studies demonstrated that unfractionated B cells from several strains of mice with lupus hyperproliferate in culture when stimulated with lipopolysaccharide (LPS) or anti-IgM. We wished to further examine proliferation of resting B cells from the BXSB mouse model of lupus and mice with the Yaa allele, when activated with a number of stimuli. Our work demonstrates that: (1) resting B cells from mice containing the Yaa allele hyperproliferated compared to that seen with B cells from mice lacking the Yaa allele, (2) this hyperproliferation occurred whether cells were stimulated with phorbol myristate acetate/ionomycin, LPS, anti-IgM, or CD40L cross-linking, (3) this hyperproliferation is specific to B and not T cells. Taken together these data suggest that one mechanism by which the Yaa allele contributes to the accelerated onset of lupus in BXSB male mice is through its influence on B cell activation.

Cellular interactions in the lpr and gld models of systemic autoimmunity.

The lpr and gld murine models have been important contributors to our understanding of systemic autoimmune diseases. Mice homozygous for either of these autosomal recessive genes develop a phenotypically identical disease characterized by the accumulation of CD4-CD8- T-cells and the production of a wide spectrum of autoantibodies. The lpr (lymphoproliferation) mutation encodes a defective Fas apoptosis receptor gene. More recently, gld (generalized lymphadenopathy) has been shown to be a point mutation in the Fas ligand gene. Despite the molecular characterization of these mutations, the exact mechanism by which tolerance is lost is still unknown, although in vivo cell transfer studies have provided clues. Chimera studies, in which normal and lpr bone marrow were co-infused into lpr mice, demonstrated not only that the normal Fas receptor is functionally expressed in both T- and B-cells, but that the Fas mutation is required in both for full expression of the lpr phenotype. Conversely, in analogous experiments with gld mice, co-infusion of normal and gld bone marrow largely prevented the development of autoantibodies. Sporadic autoantibody titers were seen in some mice, but were derived from both donors. The effects on T-cells were subtly different: The CD4-CD8- T-cells were also greatly reduced in number, but all were of gld origin. These data indicate that the gld defect is extrinsic to B-cells but only partially extrinsic to T-cells, and suggest that Fas ligand in T-cells may have an autocrine and paracrine function.

Phenotypic abnormalities of splenic and bone marrow B cells in lpr and gld mice.

Mice homozygous for the mutant Fas gene lpr develop generalized lymphoproliferation and produce autoantibodies resembling those found in human SLE. We have previously shown that these autoantibodies are produced by B2 cells rather than B1 cells and that the autoantibody- producing B cells are intrinsically abnormal. We investigated further the lpr B cell with a large panel of antibodies to B-cell surface markers to identify phenotypic abnormalities. B cells from spleen and bone marrow of age-matched congenic mice differing only at the lpr locus were examined by flow cytometry. Two consistent phenotypic differences were identified. First, spleen cells from older lpr mice had an increase in the number and percentage of IgM+ B cells expressing low levels of CD23. Second, lpr bone marrow had decreased numbers of B220hiIgM+-syndecan-1+CD23+ B cells. All other markers tested, except the previously identified modest increase of Ia on lpr spleen cells, showed no consistent differences. B cells from gld mice showed the same phenotypic abnormalities as those from lpr. Compared to T cells, the relative paucity of cell surface marker differences between lpr and +/+ B cells suggests that B cells may have fewer regulatory mechanisms to silence autoreactive specificities. The phenotypic differences identified may provide clues to the mechanism of autoantibody production in lpr mice, while the overwhelming phenotypic similarity between lpr and +/+ B cells suggests that the major abnormality of lpr B cells may lie in their specificity, that is, in their inability to delete autoreactive subsets.

The Yaa gene-mediated acceleration of murine lupus: Yaa- T cells from non-autoimmune mice collaborate with Yaa+ B cells to produce lupus autoantibodies in vivo.

The BXSB Y chromosome-linked mutant gene, Yaa, promotes autoimmune responses in mice predisposed to a lupus-like autoimmune disease. We have previously shown that a cognate interaction of T cells with B cells expressing the Yaa gene appears to be responsible for the accelerated production of autoantibodies. To investigate whether T cells that provide help for autoantibody production by Yaa+ B cells need to express the Yaa gene, we have made radiation bone marrow chimeras containing two sets of T and B cells from mice with or without the Yaa gene and differing by the Thy-1 and Igh allotypes. We then determined autoantibody production following the selective elimination of T cells of Yaa+ origin by treating mice with allele-specific anti-Thy-1 monoclonal antibody. Our results demonstrated that the selective production of autoantibodies by Yaa+ B cells in Yaa(+)-Yaa- double bone marrow chimeras can be mediated as efficiently by T cells from non-autoimmune mice lacking the Yaa gene as by T cells from autoimmune mice bearing the Yaa gene. This indicates that T cells from non-autoimmune Yaa- mice are capable of providing help for autoimmune responses by collaborating with Yaa+ B cells. These data thus strongly suggest that the Yaa gene defect is not functionally expressed in T cells, but only in B cells, and contrast with parallel experiments in the lpr model, in which defects of the Fas antigen in both T and B cells are crucial for the lpr gene-mediated promotion of autoantibody production.

The abnormal lpr double-negative T cell fails to proliferate in vivo.

Mice homozygous for the autosomal recessive gene lpr develop marked lymphadenopathy and a systemic autoimmune disease resembling human systemic lupus erythematosus. The enlarged nodes are dominated by T cells with an unusual surface phenotype: dull Thy-1+, dull CD3+, CD4-, CD8-, B220+ (double-negative T cells or DNTs). Despite their massive accumulation in vivo, these cells fail to proliferate in response to conventional T-cell mitogens in vitro. The identification of the lpr mutation as a defect in the Fas apoptosis receptor gene suggests that DNT accumulation may result from abnormal persistence rather than overproliferation. To test in vivo whether DNTs persist abnormally or have a capacity to differentiate into single-positive T cells, we have performed cell transfer experiments between congenic strains of lpr and +/+ mice differentially marked by expression of the Ly-1 or Thy-1 alleles. Although transferred lpr lymph node cells were mostly DNTs at the time of injection, most recovered cells of donor origin were single positive, particularly CD8+, at all time points after transfer. Furthermore, transfer of purified DNTs resulted in recovery of relatively few cells of donor origin. Transfer of lpr T cells enriched for CD8 expression confirmed the preferential survival of this subset. Thus, DNTs are a surprisingly transient population and have little capacity for transformation to single positives. This would suggest that DNTs are constantly being renewed, perhaps from CD4+ and CD8+ precursors.

In vivo depletion of Thy-1-positive cells originating from normal bone marrow abrogates the suppression of gld disease in normal-gld mixed bone marrow chimeras.

Mice homozygous for gld develop an autoimmune syndrome characterized by hypergammaglobulinemia, massive accumulation of abnormal T cells and the production of autoantibodies. Previous studies in our laboratory have shown that reconstitution of lethally irradiated B6/gld recipients with a mixture of normal and gld bone marrow (BM) suppresses the gld-induced syndrome. In this report we extend this observation by demonstrating that the depletion of normal Thy-1+ cells, but not normal B cells, restores gld disease in mixed BM chimeras congenic for Thy-1 and IgH alleles. These results strongly suggest that normal T cells suppress the development of gld-related abnormalities. It is probable that the mechanism by which normal Thy-1+ cells mediate the suppression is Fas ligand dependent.

Co-infusion of normal bone marrow partially corrects the gld T-cell defect. Evidence for an intrinsic and extrinsic role for Fas ligand.

Ipr and gld mice develop systemic autoimmune diseases with nearly indistinguishable manifestations, including the accumulation of massive numbers of CD4-CD8- T lymphocytes. In vivo chimera experiments have shown that the Ipr mutation is functionally expressed in both T and B cells. When lethally irradiated Ipr mice were given a combination of normal and Ipr bone marrow, only Ipr-derived B cells produced autoantibodies and only Ipr-derived T cells hyperproliferated. In contrast, analogous experiments with gld mice showed that the co-infusion of normal bone marrow greatly reduced autoantibody production. These results indicated that the gld B cell defect was extrinsic to those cells producing autoantibodies, in agreement with the recent molecular data showing that the normal gene products of the Ipr and gld loci form an interacting receptor-ligand pair. In the present study, we have extended our functional studies with gld mice using T cell-marked congenic donors. Lymphadenopathy was reduced three- to fourfold in gld mice given a combination of congenic normal and gld bone marrow compared with mice given gld bone marrow alone, and the absolute number of CD4-CD8- T cells was reduced by a factor of 7. Surprisingly, the residual CD4-CD8- T cells present in the mixed chimeras were derived entirely from the gld donor marrow. This suggests that the gld mutation results in both an extrinsic and intrinsic defect in T cells.