Modification of emerging repertoires by immunosuppression in immunodeficient mice results in autoimmunity

Expansion and clonal deletion of peripheral T cells induced by bacterial superantigen is independent of the interleukin-2 pathway

Injection of the bacterial superantigen Staphylococcus aureus enterotoxin B (SEB) into mice provokes a rapid expansion and subsequent contraction of the pool of SEB-reactive T cells bearing T cell receptor (TcR) V beta 8 gene products. Given that interleukin 2 (IL-2) stimulates proliferation, abolishes anergy, and counteracts apoptotic cell death in T cells in vitro, we tested whether the IL-2 synthesis inhibitor cyclosporin A (CsA) or a vaccinia virus recombinant releasing high amounts of human IL-2 modulate SEB responses in vivo. Surprisingly, neither IL-2 nor CsA were able to change the in vivo kinetics and magnitude of SEB-induced expansion, unresponsiveness to SEB, and peripheral clonal deletion of T cells expressing products of the SEB-reactive TcR V beta 8 gene family. In accord with these in vivo observations, IL-2 is incapable of reversing "anergy" and apoptotic cell death of V beta 8+ SEB-reactive T cells isolated from SEB-primed mice in vitro. Accordingly, upon SEB injection V beta 8+ T cells expand rapidly, without expressing IL-2 receptor (IL-2R)alpha chains in vivo, although SEB induces IL-2R alpha in vitro. Altogether, these results indicate that the IL-2/IL-2R-mediated pathway is not involved in T cell repertoire modulation by bacterial superantigens. Moreover, the data suggest that unresponsiveness of V beta 8+ T cells from SEB-primed mice is not a reversible process, but involves an unreversible commitment to programmed cell death. Absence or presence of IL-2 responsiveness could be a hallmark to distinguish truly reversible anergy and peripheral clonal deletion.

Autoantibody formation after bone marrow transplantation: a comparison between autologous and allogeneic grafts

Various autoantibodies were screened in 11 allogeneic, 14 autologous bone marrow transplanted patients and 11 healthy controls. Except anti-mitochondria antibodies, autoantibodies were found with similar frequencies after allogeneic and autologous transplantation. Moreover, 50% of autoantibodies were already present before transplantation. These results suggest that factors other than allogeneic reaction could be responsible for autoantibody formation after bone marrow transplantation, in particular, pre-graft conditioning treatment and/or viral and bacterial infection. Nevertheless, the number of patients with 2 or more autoantibodies was significantly higher after allogeneic than after autologous grafting (respectively, 82% and 36.7%; p less than 0.05). Forty percent of allografted but none of autografted patients were still positive at 12 months post graft. So, allogeneic reaction could play a role in intensity of humoral response and its maintenance for a long period.

Cyclosporine-induced autoimmunity and immune hyperreactivity

Cyclosporine (CS) is a potent immunosuppressive agent which under some circumstances paradoxically augments DTH responses, aggravates some autoimmune diseases, and induces specific forms of autoimmunity. The enhancement of DTH and other immune responses is closely related to the timing of CS administration relative to immunization. CS inhibits IL-2 production (and several other lymphokines) at a pretranscriptional level, but does not usually prevent the antigen-specific priming of T cells, such that T cells may be poised to respond as soon as CS is withdrawn. Thus, accelerated GVHD and allograft rejection may occur after withdrawal of CS. CS has been shown to aggravate and/or induce relapse in several autoimmune diseases including collagen-induced arthritis, EAE, autoimmune thyroiditis, uveitis in SDA chickens, and an autoimmune form of myocarditis in mice. CS may enhance immune responses by inactivating suppressor cells, by altering Th1/Th2 antagonism (e.g., CS promotes a protective Th1-type response in BALB/c mice infected with Leishmania major), or by promoting T cell activation through a CS-resistant IL-2-independent T cell activation/differentiation pathway. At least three forms of CS-induced autoimmunity have been described: Syngeneic or autologous GVHD which occurs in CS-treated syngeneic or autologous bone marrow transplant recipients after CS is withdrawn in rats, mice, and humans; a systemic autoimmune disease with polyarthritis and glomerulonephritis which occurs in irradiated CBA/N mice treated with CS; and organ-specific autoimmune diseases which occur in mice treated with CS during the neonatal period. The precise mechanisms by which CS induces these autoimmune diseases are not clear, however, CS affects immune tolerance at three levels. CS induces thymic medullary involution with loss of medullary Ia+ cells, and appears to at least partially block the transition from double positive (CD4+CD8+) to single positive (mature type) thymocytes. In syngeneic bone marrow chimeras, CS appears to inhibit the intrathymic deletion of clones with relatively low affinity, but not those with high affinity, to self antigens. CS appears to inhibit the action of suppressor T cells which normally maintain an innate form of resistance to autoimmunity. Finally, CS has been shown to prevent the development of T cell clonal anergy. There is redundancy in immune tolerance mechanisms, i.e., clonal deletion, clonal anergy, and suppressor cells can each maintain tolerance to similar antigens, such that it is likely that CS must cripple more than one tolerance mechanism for autoimmunity to occur.

Inside the thymus, Mls antigen is exclusively presented by B lymphocytes

The ability to stimulate an Mls-1 mixed lymphocyte reaction (MLR) is predominantly expressed by low density B lymphocytes in the spleen and peritoneal cavity of normal adult mice, and is absent in splenic B cells 1 month after lethal irradiation and reconstitution from autologous bone marrow. Coreconstitution of these mice with normal syngeneic peritoneal cells restores the stimulatory potential of splenic B cells, but sorted CD5+ or CD5- IgM+ lymphocytes from peritoneum are equally good stimulators, suggesting that functional Mls-1 expression may require long life spans and selection. Bone-marrow-reconstituted DBA/2 mice that fail to express Mls-1 antigens in the periphery nevertheless maintain T-cell receptor V beta 6 and 8.1 deletions among the newly formed T cells. These findings led us to directly investigate the Mls stimulatory ability of purified antigen-presenting cell populations inside the thymus. We report here that thymic B lymphocytes seem to represent the only intrathymic cell population able to stimulate Mls-1 MLR.

The relationship between connectivity and tolerance as revealed by computer simulation of the immune network: some lessons for an understanding of autoimmunity

According to a classical, antigen-driven view of the immune system, autoimmunity is due to the presence of self-reactive lymphocyte clones which have not been eliminated. However, computer simulations of the immune network show that the greater the degree of connectivity of a clone, the greater its degree of tolerance to chronic antigenic stimulation. This tolerance does not correspond to an absence of response on the part of the system as a whole. On the contrary, stimulation by a 'tolerogenic antigen' results in widespread modification and overall activation of the whole network. This suggests that on an autopoietic network view of the immune system, autoimmunity arises not because of the presence of self-reactive clones, which is completely normal, but because such clones are inadequately connected to the network. This amounts to a complete reversal in perspective, whose significance for the clinical treatment of autoimmunity and the future of immunology is discussed.

Recognition of idiotypic structures in the thymus

During the passage through the thymus, T cells are selected which recognize self major histocompatibility complex (MHC) antigens with low avidity. Whether the T-cell repertoire for recognition of altered self is also built up intrathymically or in the periphery, and whether it is determined exclusively by external antigens or is shaped by the internal environment is still a matter of debate. This question was addressed by analysing the responsiveness of thymocytes during post-natal development towards a nominal antigen [trinitrophenyl (TNP)] and an anti-TNP monoclonal antibody (Sp6), which carries a recurrent idiotype. During the first weeks of life, in vitro cultures of thymocytes proliferated strongly in the absence of nominal antigen. Proliferation rates were not increased by the addition of nominal antigen [TNP-ovalbumin (OA)], but a significant increase was noted in the presence of Sp6, thymocytes recognizing the processed immunoglobulin. After in vivo stimulation with TNP conjugates, 'antigen-specific' clones could also be detected in the thymus, the frequency of clones proliferating in response to Sp6 being further augmented. With increasing age, the proliferative capacity of thymocytes from unstimulated and antigenically stimulated mice decreased significantly. Responsiveness of spleen cells (SC) differed in some respects. The response towards Sp6 decreased with age, while antigen-specific clones were detected at increasing frequencies during post-natal development. Furthermore, after antigenic stimulation, the frequency solely of antigen-specific, but not of Sp6-specific clones was increased. Thus, it appears that the T-cell repertoire is shaped already during the intrathymic passage, being influenced primarily by the B-cell repertoire and modulated further by external antigen.

Postnatal B cell development: influence of trinitrobenzenesulfonic acid treatment during pregnancy

Prenatal treatment with a reactive hapten may be well suited for analyzing the establishment of self tolerance because the hapten binds ubiquitously to proteins and cells and persists for a long period in the developing organism. Based on this consideration, pregnant BALB/c mice were treated with 2,4,6-trinitrobenzenesulfonic acid (TNBS), searching for differences in 2,4,6-trinitrophenyl (TNP) responsiveness in their offspring as compared to litters of untreated mice. The frequency of TNP-specific T-independent B cells of litters from TNBS-treated mothers was very low at birth and remained below 10% of controls until the age of 42 days. On the contrary, in 8-day-old prenatally TNBS-treated litters, the frequency of TNP-specific T-dependent B cells was higher than in controls. Expansion of TNP-specific B cells after antigenic stimulation of control mice started at the age of 3-4 weeks and expansion rates increased with age, while in prenatally TNBS-treated mice, significant expansion rates were seen at the age of 2 weeks only. Yet, after restimulation with TNP-lipopolysaccharide or with a TNP-anti-TNP conjugate, but not after restimulation with TNP-ovalbumin, similar numbers of plaque-forming cells (PFC) were observed with spleen cells of prenatally untreated and TNBS-treated mice, the latter revealing an exceptional predominance of IgG PFC. Thus, TNP-specific B cells were not deleted, but prenatal TNBS treatment resulted in an altered composition of TNP-specific B cell subpopulations, their regulation differing qualitatively from the one observed in prenatally untreated mice.

T cell dependence of the "natural" autoreactive B cell activation in the spleen of normal mice

A high frequency of clonal precursor B cells producing lytic antibodies to syngeneic erythrocytes treated with bromelain (BMRC) is revealed in normal mouse spleen cells by lipopolysaccharide-driven limiting dilution analysis. All such specificities are recovered as activated blasts after density gradient fractionation, the "small lymphocyte" pool being depleted of anti-BMRC reactivities. In contrast, the spleens of athymic (nude) mice contain undetectably low frequencies of these specificities in either lymphocyte compartment. Transfer of relatively low numbers of normal syngeneic splenic T lymphocytes to adult nude mice restores the high frequency of anti-BMRC clonal precursor B cells, again in the activated, but not in the resting spleen cell fractions. Large total T cells are more than tenfold better than resting T cells in reconstitution potential, as are enriched CD4+ as compared to CD8+ cells which are practically devoid of activity in this respect. These results apply exclusively to B cells at a differentiative stage that allows for extensive clonal expansion, since there is a marked difference between the frequency of clonal precursors determined by limiting dilution analysis and the frequency of Ig-secreting plaque-forming cells of the same specificity, and induced by the same mitogen in short-term cultures. The implications of these findings for the physiology of autoreactivity and repertoire selection in the compartment of perinatal B cells are discussed.

B cell participation in the recursive selection of T cell repertoires

Normal BALB/c mice produce 2,4,6-trinitrophenyl (TNP)-I-Ad specific T helper (Th) cells expressing a receptor heterodimer which share with anti-TNP antibodies an idiotope defined by the F6(51) anti-idiotypic antibody. Expression of this Th idiotype is controlled by major histocompatibility complex and immunoglobulin heavy chain-linked genes and results from antibody-dependent selection of T cell repertoires (Martinez-A. et al., Eur. J. Immunol. 1986. 16: 417). We now present evidence for the recursive nature of T----B cell repertoire selection and suggest that perinatal B cells, present in adult peritoneal cavity, operate in the early phases of this process. Thus, the Th idiotype is absent in BALB/c mice which are either suppressed from birth with anti-mu antibodies, or reconstituted with autologous bone marrow after lethal irradiation as adults. Supplementation of bone marrow reconstitution with syngeneic Thy-1-, Ly-1+ peritoneal B cells, however, selects Th cell repertoires that are undistinguishable from normal mice as to expression of the F6(51) clonotype. This effect is lost after depletion of Ly-1+ cells in the reconstituting Thy-1- peritoneal cell population. Interestingly, large in vivo "naturally" activated Ly-1- splenic B cells can also reconstitute Th idiotype expression if they are isolated from normal, but not from athymic, nude donors. However, transfer of normal large splenic T cells to adult nude mice "educates" the splenic "large B cell" compartment in these animals such that they acquire the ability to recursively select, upon transfer to bone marrow reconstituted recipients, the Th clonotype.(ABSTRACT TRUNCATED AT 250 WORDS)