We demonstrate that transfer of OVA-specific DO11 CD4+ T cells into mice that lack T and B cells and produce secreted OVA as an endogenous self-protein results in a severe systemic autoimmune reaction with skin inflammation, wasting, and death. The transferred DO11 T cells undergo massive expansion and produce IL-2 and IFN-γ abundantly. Transfer of DO11 cells into OVA-expressing animals in which T cells are absent but B cells are present, leads to mild disease with no death. In this situation, the DO11 cells undergo similar expansion but show poor Th1 differentiation. This regulatory effect of B cells correlates with profound TCR down-regulation. If T cells are present, the DO11 cells fail to expand independent of B cells. These results suggest that both endogenous T and B lymphocytes control T cell tolerance induction and pathogenicity, but at different stages of an anti-self response. Although endogenous T cells prevent expansion and maintain homeostasis, endogenous B cells limit subsequent effector responses of autoreactive CD4+ T cells.
Afundamental feature of the immune system is tolerance to self-Ags. T cell tolerance is maintained by deletion of immature lymphocytes that recognize self-Ags in the generative lymphoid organs, and by anergy, death, and suppression of self-reactive lymphocytes that enter peripheral tissues (1, 2, 3, 4, 5). A question that has not been answered is the nature of the APCs that are important in the induction and maintenance of peripheral tolerance. Candidates include immature dendritic cells and B lymphocytes (6, 7). However, their relative contributions and mechanisms of action are largely matters of speculation (7, 8). It has also been suspected that endogenous polyclonal (i.e., not self-Ag-specific) lymphocytes normally control the activation of self-reactive T cells (9). Clinical immunologists have known that up to 20–25% of patients with various immunodeficiencies develop rheumatoid arthritis and other diseases caused by T cell-mediated autoimmunity (10, 11), and lymphopenia has been described to be a major contributor to autoimmunity in murine models (12) and in graft-vs-host disease (13).
To define the contributions of various lymphocyte subsets to the development of tolerance, we have generated a model in which naive Ag-specific DO11.10 (DO11) CD4+ T cells recognize OVA as a systemically expressed endogenous Ag in the absence of inflammatory stimuli, either in the presence or in the absence of endogenous polyclonal lymphocytes. We show that endogenous CD4+ T lymphocytes and B lymphocytes are both able to control a severe systemic pathologic reaction, however at different levels. These cellular interactions may be essential for strategies to induce tolerance in settings of graft-vs-host disease and autoimmunity.
Materials and Methods
All experimental mice were used at 6–12 wk of age. All mice were age and sex matched (±2 wk). Transgenic mice expressing the DO11.10 TCR (DO11), specific for the chicken OVA peptide (OVA323–339) in the context of the MHC class II molecule I-Ad, were obtained from Dr. K. Murphy (Washington University, St. Louis, MO). All experiments were performed using DO11 cells that had been crossed either onto a Rag1−/− or Rag2−/− background. Soluble OVA transgenic mice (sOVA Tg)4 on a BALB/c background, expressing a soluble form of OVA in the serum under control of the metallothionein promoter I, and PCR typing for OVA expression have been described (our unpublished data and Ref. 14). sOVA Tg mice were bred onto a Rag2−/−, a TCRα−/− (kindly provided by Dr. R. Locksley, University of California, San Francisco (UCSF)), a JHD−/− (kindly provided by Dr. D. Umetsu, Harvard University, Boston, MA), or crossed onto a JHD−/− × TCRα−/− background.
All mice were bred and maintained in our pathogen-free facility in accordance with the guidelines of the Laboratory Animal Resource Center of UCSF. All experiments were conducted with the approval of the Committee on Animal Research of UCSF.
Abs and flow cytometry
Splenocytes were stained with the clonotypic Ab KJ1-26 (Caltag Laboratories), anti-CD4 (GK1.5, H129.19, and RM4-5), anti-CD3 (2C11). All Abs were obtained from BD Pharmingen unless otherwise stated. Abs were used as FITC, PE, PE-Cy7, PE-Texas Red, allophycocyanin, or PerCP conjugates. Fc-block (anti-CD16/CD32) was added before staining. Flow cytometric analyses were done on a FACSCalibur with CellQuest Software (both BD Biosciences). For intracellular cytokine stains of IL-2, IFN-γ or, with appropriate isotype controls, DO11 T cells recovered from splenocytes of transfer recipients were restimulated on mitomycin C-treated BALB/c splenocytes for 14 h in the presence of 1 μg/ml OVA peptide. Brefeldin A (Epicenter) was added (10 μg/ml) for the last 2 h of stimulation.
Cell preparations, purifications, and adoptive transfer
CD4+ cells for adoptive transfer were purified from spleen and lymph nodes using Dynabeads according to the manufacturer’s protocol (Dynal). CD4+ purified DO11 Rag−/− cells were adoptively transferred into the indicated recipients by tail vein injection. BALB/c recipients were used as naive controls.
Histology and immunohistochemistry
Tissues were fixed in 10% neutral buffered formalin and embedded in paraffin. Five-micrometer sections were cut and stained with H&E. For immunohistochemistry, tissues were immersed in OCT (TissueTek; Miles), flash frozen, cut into 5-μm sections, and stained with rat anti-CD4-biotin (GK1.5; BD Pharmingen) or biotinylated KJ1-26 and subsequently with streptavidin-HRP (BD Pharmingen). Visualization was done with 3,3′-diaminobenzidine (Sigma-Aldrich).
Results and Discussion
Lack of endogenous lymphocytes prevents tolerance to systemic Ag
We have previously shown that naive OVA-specific DO11 T cells, when transferred into sOVA Tg mice that express OVA under the control of the metallothionein promoter, similar to other models of T cell tolerance, undergo an initial proliferation, rapidly expand, and then contract and become functionally unresponsive (14, 15, 16, 17, 18, 19). This model of systemic T cell tolerance allows one to analyze the stimuli and conditions that may cause or abrogate tolerance induction. We chose to study the effects of selective lymphocyte deficiencies, based on the hypothesis that endogenous lymphocytes serve to maintain self-tolerance, and therefore, elimination of endogenous lymphocytes should lead to a loss of tolerance and an autoimmune response.
To test this idea, DO11 Rag−/− cells were transferred into sOVA Tg mice that had been crossed on a Rag−/− background lacking T and B lymphocytes. As previously reported,5 the sOVA Tg Rag−/− transfer recipients develop a wasting syndrome, and in different experiments, ∼50% of the mice die within 2 wk (Fig. 1, A and B). No weight loss or lethality is seen in either sOVA Tg wild-type (WT) recipients (lymphocyte-sufficient) or in non-Tg Rag−/− recipients that lack the self-Ag (Fig. 1,A, and data not shown). All of the mice show severe ruffling of fur, and swelling of eyes, ears, and paws. Pathologic studies reveal extensive cellular infiltrates in s.c. tissues, most of them staining positive for CD4 (Fig. 1, C and D) and KJ1-26 (data not shown). No infiltration is seen when DO11 Rag−/− cells are transferred into WT sOVA Tg mice. Thus, the severe immune reaction requires both recognition of systemic Ag and the absence of endogenous lymphocytes. Importantly, the recipient mice were not immunized in any way, indicating that recognition of the circulating Ag alone is a sufficient stimulus for the pathologic reaction, and no additional activation is needed to trigger this reaction.
Absence of endogenous T lymphocytes leads to expansion of Ag-reactive T cells but not their differentiation into effector cells
To examine the role of individual polyclonal lymphocyte subsets in the control of autoimmunity, we analyzed reactions of self-reactive DO11 cells to the systemic Ag in the absence of these subsets. This approach was used because Rag−/− mice could not be reconstituted with mature lymphocytes, particularly B cells, due to abnormalities in their lymphoid architecture. We initially postulated that the severe systemic reaction and Th1 development resulted from a deficiency of regulatory T cells. However, surprisingly, if sOVA Tg TCRα−/− mice, which only lack αβ T cells, are used as recipients for DO11 Rag−/− cells, i.e., if B cells are present, the clinical disease is less severe with less weight loss, and all animals survive (Fig. 1, A and B). This correlates with mild cellular infiltrates in the skin, which are much reduced compared with sOVA Tg Rag−/− recipients (Fig. 1, C and D). Thus, the absence of αβ T cells alone (which includes regulatory T cells) is not sufficient to allow for a severe pathologic reaction against the systemic Ag.
One likely explanation for the less severe disease in the sOVA Tg TCRα−/− mice is that there is less expansion than in the sOVA Tg Rag−/− mice, simply because there is less available space (9). To address this question, the transferred DO11 T cells were followed for cell numbers over time. As we had observed previously, DO11 cells that encounter the soluble Ag in WT Tg recipients undergo cycling with limited expansion (3- to 4-fold) followed by deletion and development of tolerance, as indicated by reduced IL-2 and complete lack of IFN-γ production upon restimulation (Fig. 2) (14). In contrast, in the sOVA Tg Rag−/− recipients, the DO11 Rag−/− cells undergo massive expansion (Fig. 2,A). Remarkably, when DO11 cells are transferred into sOVA Tg TCRα−/− recipients, the T cells undergo a similar expansion compared with sOVA Tg Rag−/− recipients (Fig. 2 A). Thus, T cells in the host control the proliferation of Ag-recognizing lymphocytes and function to maintain homeostasis.
Another possibility to explain the less severe reaction in the sOVA Tg TCRα−/− recipients, compared with the Rag−/− recipients, could be that effector differentiation is reduced. To address this question, the cytokine profiles of the DO11 cells were examined over time by intracellular staining. DO11 Rag−/− cells that have encountered their Ag for 5 or 10 days in sOVA Tg Rag−/− recipients produce abundant amounts of IL-2 and IFN-γ, correlating with their severe clinical disease (Fig. 2, B and C). The effector differentiation is associated with up-regulation of the activation marker CD25 and down-regulation of CD62L (data not shown). Strikingly, DO11 cells transferred into the sOVA Tg TCRα−/− recipients produce less IL-2 and greatly reduced IFN-γ compared with the sOVA Tg Rag−/− recipients (Fig. 2, B and C). This indicates that despite similar expansion of Ag-specific T cells, effector differentiation is reduced in the presence of B cells, presumably accounting for the less severe disease in the TCRα−/− sOVA Tg recipients.
B cells prevent effector differentiation and autoimmunity only in the absence of endogenous T cells
Because the sOVA Tg TCRα−/− recipients (lacking all αβ T cells) develop less disease than sOVA Tg Rag−/− recipients and no lethality, it is likely that B cells (which are normal in these mice) play a role in preventing the severe reaction of the DO11 T cells. The next set of experiments was designed to examine the relative contributions of endogenous T and B lymphocytes to the induction of tolerance. To address this question, we crossed sOVA Tg mice on a JHD−/− (B cell-deficient) background or on a TCRα−/− × JHD−/− double knockout (T and B cell-deficient) background and used these mice as recipients for adoptively transferred DO11 Rag−/− cells. Fig. 3,A shows that sOVA Tg JHD−/− recipients show no clinical disease, similar to sOVA Tg WT mice. In contrast, sOVA Tg TCRα−/− × JHD−/− recipients develop severe systemic disease and show massive T cell infiltrates in the skin, recapitulating the histologic picture seen in sOVA Tg Rag−/− mice (Fig. 3, B and C).
Assays of cell numbers and functional responses showed that the transferred DO11 cells in the B cell-deficient recipients (sOVA Tg JHD−/−) behave like in the sOVA Tg WT recipients, showing only limited expansion followed by contraction (Fig. 4,A). DO11 cells recovered on day 5 after transfer are functionally tolerant by IL-2 and IFN-γ production (Fig. 4, B and C). However, if DO11 T cells are transferred into sOVA Tg TCRα−/− × JHD−/− recipients, the T cells undergo massive expansion (Fig. 4,A) and develop effector function, just as in sOVA Tg Rag−/− recipients (Fig. 4, B and C).
These results make two important points. First, the burst of expansion, which is limited by the presence of endogenous T cells, is critical for the breakdown of tolerance because tolerance induction is not affected in sOVA Tg JHD−/− mice. Second, the DO11 transfer into sOVA Tg TCRα−/− × JHD−/− recipients formally proves that B cells are the key players in the prevention of effector differentiation and not γδ T cells or other T cell populations that do not express the TCRα chain.
TCR down-modulation on self-Ag-specific T cells in the presence of B lymphocytes
Phenotypic analyses of DO11 cells that recognized the systemic Ag in the presence of endogenous B and/or T lymphocytes revealed that, in the presence of B lymphocytes (i.e., in TCRα−/− sOVA Tg recipients), the levels of the TCR and CD3, both surface and intracellular, are profoundly reduced, compared with Rag−/− and TCRα−/− × JHD−/− recipients (which lack both T and B cells) (Fig. 5 and data not shown). No TCR down-regulation is seen in either Rag−/− recipients or in TCRα−/− recipients that do not express sOVA (data not shown) demonstrating that TCR down-regulation in the presence of B cells depends on self-Ag recognition. TCR down-regulation has been associated with T cell Ag recognition and activation. However, the kinetics and the fact that no TCR down-regulation occurs in Rag−/− and TCRα−/− × JHD−/− recipients, despite strong effector differentiation, suggests that TCR down-regulation is mediated primarily by B cells and may serve as a mechanism of tolerance maintenance.
By studying the consequences of deficiency of a particular host lymphocyte population, we have tried to define the role of that population in the induction of T cell tolerance. These results demonstrate that normal, polyclonal T cells and B cells control different and perhaps sequential steps in the development of tolerance in naive T cells that encounter a systemic Ag. Endogenous T cells control the initial activation and proliferation of T cells that specifically recognize a self-Ag. The expansion that occurs in the absence of host T cells is critical for the breakdown of tolerance, because it does not occur in sOVA Tg WT or JHD−/− recipients, and in these recipients the transferred T cells become tolerant. Endogenous B cells, in contrast, do not prevent the accumulation of the potentially pathogenic T cells but limit the subsequent differentiation of these autoreactive lymphocytes into effector cells. Thus, B cells would be the second line in preventing autoimmunity once the initial control point at the stage of the T cell expansion has been passed.
The tolerogenic potential of Ag-presenting B cells has been suggested in previous studies (7, 20), but the underlying mechanisms remain obscure. For instance, B cells may provide a tolerogenic stimulus to T cells in a MHC-dependent fashion, as has been suggested for immature dendritic cells, or by an Ag-independent effect (21, 22). In our model, B cells may function to induce self-tolerance by several ways. First, because B cells are abundant and may be presenting self-Ag chronically, they may lead to down-modulation of the Ag receptor on self-reactive T cells or other proximal molecules in the TCR signaling pathway, e.g., through the E3-ubiquitin-ligase pathway (23, 24). The fact that both surface and intracellular TCR levels are down-regulated in sOVA Tg TCRα−/− recipients suggests that both internalization and degradation occur. B cells may also inhibit T cell differentiation by producing inhibitory cytokines such as IL-10 and TGFβ (25, 26).
Our data are the first evidence that different classes of endogenous lymphocytes may regulate different phases of a T cell response to a systemic self-Ag. It suggests that the relative contribution of T and B cells to immunity or tolerance is influenced by timing and the local environment at the site of Ag recognition. These results have important clinical implications in settings where one tries to induce tolerance in the presence of an empty host, e.g., in bone marrow transplantation or certain autoimmune diseases.
We are indebted to C. Benitez for mouse typing We thank the members of the Abbas lab and Bluestone lab for helpful discussion.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by National Institutes of Health Grants RO1 AI64677 and PO1 AI35297 (to A.K.A.) and Deutsche Forschungsgemeinschaft Grants LO 808/1-1 and KN 533/1-1 (to J.L. and B.K.).
Abbreviations used in this paper: sOVA Tg, soluble OVA transgenic; WT, wild type.
B. Knoechel, J. Lohr, E. Kahn, and A. K. Abbas. Sequential development of IL-2-dependent effector and regulatory T cells in response to endogenous systemic antigen. Submitted for publication.