Lymphocyte-mediated cytotoxicity in vitro and in vivo: mechanisms and significance.
Explore the source record for details and available documents.
Biomedical subjects
Publications and source records attributed to H Hengartner.
Explore the source record for details and available documents.
The role of innate, alpha/beta interferon (IFN-alpha/beta)-dependent protection versus specific antibody-mediated protection against vesicular stomatitis virus (VSV) was evaluated in IFN-alpha/beta receptor-deficient mice (IFN-alpha/beta R0/0 mice). VSV is a close relative to rabies virus that causes neurological disease in mice. In contrast to normal mice, IFN-alpha/beta R0/0 mice were highly susceptible to infection with VSV because of ubiquitous high viral replication. Adoptive transfer experiments showed that neutralizing antibodies against the glycoprotein of VSV (VSV-G) protected these mice efficiently against systemic infection and against peripheral subcutaneous infection but protected only to a limited degree against intranasal infection with VSV. In contrast, VSV-specific T cells or antibodies specific for the nucleoprotein of VSV (VSV-N) were unable to protect IFN-alpha/beta R0/0 mice against VSV. These results demonstrate that mice are extremely sensitive to VSV if IFN-alpha/beta is not functional and that under these conditions, neutralizing antibody responses mediate efficient protection, but apparently only against extraneuronal infection.
Antiviral immune responses of mice lacking interleukin-2 (IL-2) or IL-4 or both IL-2 and IL-4 (IL-2/4) were compared by using different viruses. Primary cytotoxic T-lymphocyte (CTL) responses against lymphocytic choriomeningitis virus (LCMV) were only moderately reduced in mice lacking IL-2 and were normal in mice lacking IL-4. Mice deficient in both interleukins exhibited variable and more strongly reduced but nevertheless in vivo protective LCMV-specific CTL responses. Similar results were obtained with vaccinia virus. Upon virus-specific restimulation in vitro, spleen cells from IL-2- and IL-2/4-deficient mice failed to generate CTL responses against virus-infected target cells, whereas the response of mice deficient in only IL-4 was comparable to that of control mice. The addition of IL-2 during in vitro restimulation completely restored the responses of both IL-2 and IL-2/4-deficient mice. T-helper-cell-independent immunoglobulin M and T-helper-cell-dependent immunoglobulin G antibody responses against vesicular stomatitis virus glycoprotein were within normal ranges for the various mutant mice. After LCMV infection, specific antibody responses against LCMV nucleoprotein were reduced four- to eightfold. These results show that mice lacking IL-2/4 have an overall tendency to exhibit more severely reduced CTL responses than IL-2- or IL-4-deficient mice. Nevertheless, and surprisingly, in vivo protective immune responses were mounted in the absence of IL-2/4, suggesting that besides a minor contribution from IL-4, other interleukins compensate in vivo for the lack of IL-2 in IL-2-deficient mice.
Expression of self antigen on lymphohemopoietic cells and in the thymus has been shown to cause tolerance by negative selection. To investigate the role of self antigen expression levels on the induction of tolerance, we generated transgenic mice expressing the lymphocytic choriomeningitis virus (LCMV) glycoprotein (GP) driven by the H-2Kb promoter. Two mouse lines differing in transgene expression levels were obtained and evaluated for the induction of tolerance to LCMV-GP. LCMV-GP high (GPhi) expressing animals thymically deleted self-reactive thymocytes. Low expressors (GPlo) partially deleted self-reactive mature T cells in the periphery in the absence of any obvious signs of negative selection in the thymus. Functionally, the LCMV-GP-specific cytotoxic T cell (CTL) response was absent in GPhi mice, whereas GPlo mice produced diminished LCMV-GP-specific CTL responses. Therefore limiting levels of expression of self antigen influence efficiency of negative selection, enabling potentially self-reactive T cells to escape from tolerance induction.
This study investigated whether or not the localization of persisting Ag influenced the localization of memory B and/or of Ab-forming cells (AFC). As a model Ag, we used vesicular stomatitis virus, which does not measurably replicate extraneuronally in adult mice after peripheral infection. Our results show that memory B cells and AFC are induced at the site where Ag is present; induced memory B cells then recirculate throughout the lymphoid system independently of Ag localization. In contrast, AFC are induced only at the sites of Ag persistence. These triggered Ab producing cells do not recirculate through the lymphoid tissue but migrate to the bone marrow.
Two molecular mechanisms of T cell-mediated cytotoxicity, one perforin-based, the other Fas-based, have been demonstrated. To determine the extent of their contribution to T cell-mediated cytotoxicity, a range of effector cells from normal control or perforin-deficient mice were tested against a panel of target cells with various levels of Fas expression. All cytotoxicity observed was due to either of these mechanisms, and no third mechanism was detected. Thus, the perforin- and Fas-based mechanisms may account for all T cell-mediated cytotoxicity in short-term in vitro assays.
The influence of the murine MHC gene complex H-2 class I alleles or of the genetic background and the role of virus variants on establishment of a persistent infection with lymphocytic choriomeningitis virus (LCMV) was analyzed in immunocompetent mice of H-2d haplotype by evaluation of cytotoxic T cell responses. Susceptibility to establishment of a virus carrier state increased in various H-2d mice with lower CTL response and slower CTL kinetics. Low responder BALB/c-H-2dm2, lacking H-2Ld molecules to present the major T cell epitope amino acids 118-126 of the LCMV nucleoprotein or DBA/2 mice possessing relatively few CD8+ T cells, were more susceptible than high responder BALB/c mice expressing H-2Ld. Additional critical factors were LCMV isolate and dose of infection. The rapidly replicating LCMV-DOCILE and Cl 13 Armstrong induced viral persistence readily, whereas the slowly replicating parental virus strains WE or Armstrong did not. The presented findings illustrate a model of MHC-linked or MHC-unlinked susceptibility for virus persistence and may help to explain pathogeneses of chronic virus infections in humans that are often associated with slowly progressing immunopathology.
Perforin-deficient mice have been generated by homologous recombination to determine whether the effects of CD8+ cytolytic T cells and natural killer cells are mediated by pore formation involving perforin. These mice are viable and fertile and have normal numbers of CD8+ T cells and natural killer cells which do not lyse virus-infected or allogeneic fibroblasts or natural killer target cells in vitro. The mice fail to clear lymphocytic choriomeningitis virus and they eliminate fibrosarcoma tumour cells with reduced efficiency. Perforin is therefore a key effector molecule for T-cell- and natural killer-cell-mediated cytolysis.
The size of the Ab repertoire has been estimated to comprise theoretically somewhere between > 10(10) and approximately 10(4) specificities, dependent on the criteria used. In an attempt to estimate the anti-viral protective Ab repertoire of the mouse the B cell and Ab-forming cell (AFC) frequencies and protective neutralizing Ab levels during the course of an infection with vesicular stomatitis virus (VSV) were analyzed. Determination of AFC frequencies, limiting dilution assays, and adoptive transfer experiments to SCID mice revealed that during the acute phase (day 8) of the immune response, more than 50% of all IgG2a-producing AFCs were specific for VSV, most of them recognizing the neutralizing determinant. In a later phase (days 21 or 50), 10 to 20 times fewer VSV-specific AFCs were present, corresponding to a frequency of approximately 1:10(4) spleen cells. Finally, in a protection assay in SCID mice, adoptively transferred protective Ab concentrations were found to be approximately 1 to 10 micrograms Ab/ml mouse serum. Because during the memory phase of the anti-VSV response usually 10(4) AFC/mouse are engaged to maintain a high level of memory IgG against the neutralizing determinant on VSV and if one assumes a total number of about 10(6) AFCs/mouse, these data suggest a rather limited neutralizing anti-viral protective memory-AFC repertoire of 10(2) to 10(4) different specificities.
T lymphocytes have been implicated in the onset of many autoimmune diseases; however, the mechanisms underlying T-cell activation toward self antigens are poorly understood. To study whether T-lymphocyte costimulation can overcome the immunologic unresponsiveness observed in an in vivo model, we have created transgenic mice expressing the costimulatory mouse molecule B7-1, a ligand for the CD28 receptor, on pancreatic beta cells. We now report that triple-transgenic mice expressing both B7-1 and a viral glycoprotein on their beta cells, along with T cells expressing the viral-glycoprotein-specific transgenic T-cell receptor, all develop insulitis (lymphocytic infiltration of the pancreatic islets) and diabetes. In striking contrast, the T cells in double-transgenic mice expressing the same viral glycoprotein (but no B7) on their pancreatic beta cells and the transgenic T-cell receptor on their T cells, reported earlier, remain indifferent to the glycoprotein-expressing beta cells. In fact, all three transgenes are required to initiate immune-mediated destruction of the beta cells. Mice expressing any of the transgenes alone, or any two in combination, maintain normal islet architecture and never spontaneously develop insulitis or diabetes. Our results show that aberrant B7 expression on peripheral tissues may play an important role in the activation of self-reactive T cells and further suggest that abnormal expression of costimulatory receptors may be involved in various T-cell-mediated autoimmune diseases.
A synthetic peptide corresponding to an immunodominant epitope of lymphocytic choriomeningitis virus glycoprotein (LCMV GP) was used to prime or to tolerize CD8+ T cells in vivo, dependent on mode of immunization. Peptide-specific tolerance was then examined in transgenic mice expressing LCMV GP in the beta islet cells of the pancreas; these mice develop CD8+ T-cell-mediated diabetes within 8-14 days after LCMV infection. Specific peptide-induced tolerance prevented autoimmune destruction of beta islet cells and diabetes in this transgenic mouse model.
Successful interaction of the T cell receptor (TCR) with major histocompatibility complex (MHC) molecules during thymic selection down-regulates the expression of the recombination activating genes (RAG)-1 and -2 in cortical thymocytes and thereby prevents further endogenous TCR alpha-chain gene rearrangements (Borgulya, P., Kishi, H., Uematsu, Y. and von Boehmer, H., Cell. 1992. 69: 529-537; Brändle, D., Müller, C., Rülicke, T., Hengartner, H. and Pircher, H., Proc. Natl. Acad. Sci. USA 1992. 89: 9529-9533). To address the question whether down-regulation of RAG-1 activity represents an irreversible process we have blocked TCR-MHC interactions of thymocytes with thymic stromal cells. Firstly, transgenic (Tg) mice expressing a virus-specific MHC class I (H-2Db)-restricted TCR were injected with anti-Db or anti-CD8 monoclonal antibodies and RAG-1 expression was examined by in situ hybridization on thymus sections. The results show that cortical thymocytes up-regulated RAG-1 expression within 24 h after antibody administration. Secondly, immature thymocytes from TCR Tg mice were released from the thymic microenvironment and cultured in vitro for 14 h in single-cell suspension. The amount of RAG-1 mRNA was increased sixfold in cultured cells when compared to freshly isolated thymocytes. In addition, we show that immature thymocytes from TCR transgenic mice bearing non-selective MHC molecules (H-2d) down-regulated RAG-1 expression after antigen-induced TCR engagement. Cytofluorometric analysis further revealed that surface expression of CD69 on immature thymocytes inversely correlated with RAG-1 expression during positive and negative selection processes.
We investigated the mechanism leading to an IgG autoantibody response in two transgenic mouse lines expressing the glycoprotein of vesicular stomatitis virus (VSV-G). Previous experiments have shown that these animals do not mount a transgene-specific IgG response upon stimulation with purified VSV-G or infection with recombinant vaccinia virus expressing VSV-G. However, infection of VSV-G transgenic animals with wild-type vesicular stomatitis virus, serotype Indiana, readily induced VSV-G-specific, neutralizing IgG autoantibodies. We have tested whether this labile state of tolerance reflected differential availability of VSV-G-specific T help. For this, we immunized transgenic mice with the self-antigen VSV-G covalently coupled to sperm-whale myoglobulin (VSV-G-SWM), to provide new T helper epitopes that are linked to the B cell epitope; co-injected uncoupled VSV-G and SWM served as control. High titers of VSV-G specific IgG autoantibodies were detected in serum of mice immunized with VSV-G-SWM but not after co-injection of uncoupled VSV-G and SWM. Transgenic animals depleted of CD4+ T cells prior to injection of VSV-G-SWM failed to mount an IgG response. Priming of transgenic mice with the foreign carrier did not accelerate the IgG autoantibody response to VSV-G-SWM, suggesting that B cells were limiting the rate of the response. Thus, self-reactive B cells could be triggered to produce IgG, if they received linked T help specific for a foreign carrier determinant provided either by a classical carrier determinant or a virus.
We have characterized a prominent (15-20%) thymocyte population expressing CD4 at a high and CD8 at a low level (CD4+8lo) in mice transgenic for a T cell receptor (TCR) restricted by major histocompatibility complex (MHC) class I molecules. The results demonstrate that the CD4+8lo population is an intermediate stage between immature CD4+8+ and end-stage CD4+8- thymocytes and that the survival of these cells crucially depends on the successful interaction of the transgenic TCR with self MHC class I molecules. In addition we demonstrate that the avidity of the interaction between TCR and self MHC class I molecules determines whether CD4+8lo thymocytes are found in significant numbers in this transgenic model. Our findings support a selective and multi-step model of T cell differentiation in the thymus.
Induction of CD8+, class I-restricted T cells by non-infectious, exogenous antigens has been documented for model protein antigens such as ovalbumin and for major histocompatibility complex restricted short peptides in viral and tumor systems. However, the protective capacity of cytotoxic T cells induced by conventional proteins has not been tested in vivo so far. We, therefore, evaluated the induction of protective cytotoxic T cells against three different full-length recombinant viral proteins derived from a baculovirus expression system, i.e. the glycoprotein and nucleoprotein of lymphocytic choriomeningitis virus (LCMV) and the nucleoprotein of vesicular stomatitis virus (VSV). These viral proteins induced cytotoxic T cells in a T helper cell-independent fashion which lysed infected target cells in vitro and protected mice from viral replication, immunopathological disease and growth of a tumor expressing the same antigen as a tumor antigen. These results are surprising, since it had been shown earlier for completely inactivated nonreplicating viral vaccines and again here for beta-propiolactone-inactivated VSV or UV-light inactivated LCMV that nonreplicating viral vaccines were incapable of inducing protective cytotoxic T cells. Our data show that immunization of mice with as little as 10 micrograms of non-infectious viral proteins triggered long-lasting CD8+ T cell-mediated antiviral immunity. It was found that the protein alone was only weakly able to induce cytotoxic T cells, and that association with cellular debris functioned as an adjuvant. These findings may be relevant for our understanding of the phenomenon of cross-priming and have obvious implications for vaccine strategies.
Antigens normally induce an immunoglobulin (Ig)G response which stays at an elevated level for several weeks or months, constituting an important part of the immunological memory. This study investigated factors influencing the level of neutralizing IgG titers against a virus and shows that within the range tested it was independent of the number of initially available and potentially responding T helper and B cells, but was regulated by the amount of specific IgG-immune complexes forming depots of persisting antigen. These findings support the notion that the efficiency of vaccines in inducing long-lasting protective IgG is regulated predominantly by the amount of persisting (and presumably follicular dendritic cell-associated) antigen-antibody complexes.
T cell tolerance is usually established by clonal deletion of self-specific T cells in the thymus, or some times, in the periphery. Alternatively, tolerance may also be achieved by induction of clonal T cell unresponsiveness by a poorly understood mechanism called "anergy". We found that transgenic mice expressing a soluble form of vesicular stomatitis virus (VSV) glycoprotein (G) predominantly in liver and kidney exhibited normal B cell responses. VSV-G-specific T help-independent neutralizing IgM responses were within normal ranges, but no T help-dependent neutralizing IgG antibodies were generated upon immunization with recombinant VSV-G protein and recombinant vaccinia virus expressing VSV-G. This demonstrated absence of B cell tolerance but presence of T helper cell unresponsiveness. After adoptive transfer of transgenic spleen cells into thymectomized immuno-incompetent hosts, the unresponsive T helper cells regained function and switched the neutralizing IgM response to IgG, comparably to control T helper cells, within 7 days. Conversely, when naive non-transgenic spleen cells were transferred into transgenic mice, VSV-G-specific T helper cells became unresponsive within 3-4 days. These results suggest that VSV-G-specific T helper cells are rendered unresponsive within a few days in the VSV-G transgenic host also outside of the thymus and that this unresponsiveness was reversed by transfer into antigen-free recipients.
Growth of Listeria monocytogenes is mainly controlled by macrophages, which are activated by specific T cells. A potential role of CD8+ T cells by direct lysis of infected cells was investigated in perforin-deficient mice generated by homologous recombination. The absence of perforin-mediated cytotoxicity resulted in delayed clearance of Listeria from the spleen but not the liver after primary infection, overall susceptibility to Listeria however was not increased. Protection against a secondary infection was drastically impaired in perforin-deficient mice. Adoptive transfer of immune spleen cells to recipients revealed that anti-Listeria protection by CD8+ T cells from perforin-deficient versus normal mice was about 10-fold reduced in livers and about 100-fold reduced in the spleen of recipients. CD4+ T cells from immune control and perforin-deficient mice conferred comparable protection. These results indicate that the protective effect of CD8+ T cells against an intracellular bacterium mainly evident in secondary infection is mediated by a perforin-dependent pathway, presumably cytotoxicity, and less by other direct or indirect effector mechanisms.