Thymic lymphoproliferative disease after successful correction of CD40 ligand deficiency by gene transfer in mice.

Inherited deficiency of the CD40 ligand (X-linked hyper-IgM syndrome) is characterized by failure of immunoglobulin isotype switching and severe defects of cell-mediated immunity. To test the potential for gene transfer therapy to correct this disorder, we transduced murine bone marrow or thymic cells with a retroviral vector containing the cDNA for the murine CD40 ligand (CD40L) and injected them into CD40L-/- mice. Even low-level, constitutive expression of the transgene stimulated humoral and cellular immune functions in these mice. With extended follow-up, however, 12 of 19 treated mice developed T-lymphoproliferative disorders, ranging from polyclonal increases of lymphoblasts to overt monoclonal T-lymphoblastic lymphomas that involved multiple organs. Our findings show that constitutive (rather than tightly regulated), low-level expression of CD40L can produce abnormal proliferative responses in developing T lymphocytes, apparently through aberrant interaction between CD40L+ and TCRalphabeta+CD40+ thymocytes. Current methods of gene therapy may prove inappropriate for disorders involving highly regulated genes in essential positions in proliferative cascades.

T cells that encounter virus in the complete absence of a particular H-2 antigen are nonresponsive when stimulated again in the context of that H-2 antigen.

Immunologically naive BALB/c (H-2d) and C57BL/6J (B6) (H-2b) T-cell populations can, after filtration to remove alloreactive precursor lymphocytes, be induced to respond to vaccinia virus presented in the context of H-2Kk when stimulated in an appropriate recipient. Exposure to vaccinia virus 6 wk previously completely abrogated the capacity of BALB/c T cells to interact with H-2Kk-vaccinia virus. This is also true for negatively selected B6 thoracic duct lymphocytes taken at 14 or 18 d, but not at 6 wk after immunization: the discrepancy is thought to reflect the progressive emergence of new T cells in the latter group. No evidence could be found for the operation of suppression, and the results are considered to indicate that T cells that interact with virus in the absence of the relevant H-2 antigen are tolerized. Whereas stimulation to effector function is H-2 restricted, induction of immune paralysis may be unrestricted. The capacity of T-cell populations to respond to virus presented in the context of allogeneic H-2 determinants thus depends upon previous antigenic experience.

H-2 compatability requirement for T-cell-mediated lysis of target cells infected with lymphocytic choriomeningitis virus. Different cytotoxic T-cell specificities are associated with structures coded for in H-2K or H-2D;.

Use of syngeneic, allogeneic, F1, AND H-2 recombinatn mice has shown that animals injected with lymphocytic choriomeningitis (LCM) virus generate T cells which are cytotoxic for H-2K or H-2D compatible, but not H-2 different, virus-infected target cells. Three separate lines of evidence are presented which indicate that these immune T cells are sensitized to "altered-self," the self antigens involved being coded for in the H-2K or H-2d regions. Firstly, cytotoxic activity associated with mutuality at H-2D iy, lysis mediated by immune T cells from F1 or H-2 recombinant mice is specifically inhibited only by presence of unlabeled, virus-infected cells that are H-2 compatible with the targets. Thirdly, LCM-immune F1 and H-2 recombinant T cells inoculated into irradiated, virus-infected recipients proliferate only to kill target cells that are H-2 compatible with both the donor and the recipient. All of these experiments establish that there is a dissociation of T-cell activities between parental haplotypes in F1 mice, and between H-2K and H-2D in recombinants. It would thus seem that there are at least two specificities of tlcm-immune T cells in homozygotes, associated with either H-2K or H-2D, and four specificities in F1 hybrids. The significance of these findings, with respect both to gene duplication and to the marked polymorphism in the H-2 system, is discussed.

H-2 compatibility is required for T-cell-mediated lysis of target cells infected with lymphocytic choriomeningitis virus.

Maximal cell-mediated lysis of targets infected with lymphocytic choriomeningitis virus occurs only within a H-2 compatible system. Syngeneic immune spleen cells are at least 100 times as effective as are allogeneic lymphocytes. Reciprocal restriction of cytotoxic T-cell activity has been shown to operative between H-2k, H-2d, and H-2b. Experiments with cogenic mice have localized the effect to the H-2 gene complex. Furthermore, the observation that lymphocytes from H-2a mice cause high specific 51Cr release from either H-2d virus-infected cells, indicates that identity at either the K or the D end of the H-2 gene complex is sufficient for this lytic interaction.

Patterns of virus-immune T-cell responsiveness. Comparison of (H-2k X H-2b) leads to H-2b radiation chimeras and negatively selected H-2b lymphocytes.

Negatively selected H-2K(b)D(b) TDL can be induced to respond strongly to vaccinia virus presented in the context of both H-2K(k) and H-2D(b) when stimulated in irradiated H-2K(k)D(b) recipients. Addition of excess (H- 2K(k)D(b) x H-2K(b)D(b))F1 TDL, which are low responders to H-2D(b)-vaccinia virus, does not obviously suppress the reactivity pattern of the H-2K(b)D(b) T cells. However, lymphocytes from chimeras made by reconstituting H- 2K(b)D(b) mice with (H-2K(k)D(k) x H-2K(b)D(b))F(l) bone marrow cells make little, if any, cytotoxic T-cell response to vaccinia virus when sensitized in H-2K(k)D(b) recipients. We have thus documented one instance where the responder phenotype of T ceils from an F(l) {arrow} parent chimera is not equivalent to that associated with the H-2 type of the parental thymus. Lymphocytes from both the chimera and the H-2K(b)D(b) parent (after negative selection) are tolerant to the H-2K(k) and I-A(k) alloantigens encountered in the recipient, but the chimera T cells are also defective in their response to a neoantigen (vaccinia virus) presented in the context of H-2K(k) which the parental T cells invariably recognize. It is thus possible that at least part of the phenomenology associated with the F(l) {arrow} parent radiation chimeras reflects deletion of repertoire in the context of H-2 antigens present during thymocyte ontogeny on other than radiation-resistant thymic epithelium.

Vaccinia-specific cytotoxic T-cell responses in the context of H-2 antigens not encountered in thymus may reflect aberrant recognition of a virus-H-2 complex.

BALB/c (H-2Kd-Dd) spleen and lymph node populations were specifically depleted of alloreactive potential by filtration through H-2 different, irradiated recipients. These negatively selected T cells were then stimulated with vaccinia virus in mice expressing the foreign H-2 determinants encountered previously in the filter environment. Strong virus-immune cytotoxic T-cell responses were seen in the context of H-2Kk and H-2Ks, but not 2H-2Kb. The T cells generated were not cross-reactive for the H-2Kk and H-2Kd alleles, and responsiveness was independent of concurrent presence of effector populations operating at H-2D. These findings are consisent with the idea that recognition is mediated via a complex receptor, part of which is specific for virus and part for self H-2. The capacity to interact with allogeneic, virus-infected cells may then reflect aberrant recognition of a virus-H-2-antigen complex by this single, large binding site. For instance, the T cell which would normally recognize H-2Kd-virus x, or H-2Dd-minor histocompatibility antigen Z, may now show specificity for H-2Kk-vaccinia virus. Implications for both the selective role of the thymus and for mechanisms of tolerance are discussed.

T-cell populations specifically depleted of alloreactive potential cannot be induced to lyse H-2-different virus-infected target cells.

Mouse lymphocyte populations of one parental H-2 type (A) were specificially depleted of alloreactive potential by filtration through irradiated A X B F1 recipients, and thoracic duct cells were then stimulated with virus in an A X B F1 environment. Experiments using T cells that had previously been exposed to influenza virus in the context of A established that cross-priming for recognition of viral components expressed on H-2-different (B) target cells does not occur. Furthermore, immunologically naive T cells stimulated with vaccinia virus, subsequent to negative selection for reactivity to B, could not be shown to interact with virus-infected cells of type B. Either there is no significant T-cell repertoire for recognition of virus associated with an H-2 determinant not encountered during ontogeny, or such T cells are also alloreactive and are removed during filtration.

Suppression of cell-mediated immunity by street rabies virus.

Mice lethally infected with street rabies virus failed to develop cytotoxic T cells specific for rabies virus-infected target cells, whereas high levels of cell-mediated cytotoxicity (CMC) were generated after nonfatal infection with the attenuated high egg passage (HEP) or ERA rabies virus strains. Furthermore concurrent infection with street, but not with HEP, rabies virus suppresses development of a primary (but not a secondary) CMC response specific for influenza virus. No cross-reactivity is found between effector T-cell populations from mice immunized with HEP or with influenza virus. It thus appears that street rabies virus, which is not known to replicate in the cells of immune system, induces some general defect in the primary CMC lymphocyte response, though restimulation of memory T-cell populations is unimpaired and there is no defect in antibody formation. Development of fatal rabies may reflect the operation of this selective immunosuppressive mechanism.

Induction of neonatal tolerance to H-2k in B6 mice does not allow the emergence of T cells specific for H-2k plus vaccinia virus.

Thymocytes and spleen cells from C57BL/6 mice (H-2b) neonatally tolerized to H-2k alloantigens do not generate an anti-vaccinia response restricted to H-2Kk when adoptively transferred to appropriate irradiated hosts. This is in sharp contrast to the case for negatively selected C57BL/6 spleen cells acutely depleted of alloreactivity. No evidence for suppression was found in cell mixture experiments. We have shown elsewhere that our neonatally tolerized animals have a centrally induced delection-type tolerance in the absence of obvious suppression.2 We now suggest that in the neonatally tolerized mouse, chronic, central delection of anti-H-2k clones during early T cell ontogeny eliminates the major source of cells able to give rise, via somatic mutation and expansion, to anti-H-2Kk + vaccinia specific cytotoxic T lymphocyte precursors (CTL-P) in the adult. A similar mechanism may operate in the (k + b) leads to b chimera; however, the presence of H-2kxb accessory and presenting cells may permit the eventual generation (via cross-stimulation) of an H-2k-restricted vaccinia-specific repertoire. This would account for our observation of such "aberrant recognition" CTL-P emerging in the spleens of older (k x b) leads to b chimeras.

The control of specificity of cytotoxic T lymphocytes by the major histocompatibility complex (AG-B) in rats and identification of a new alloantigen system showing no AG-B restriction.

The regulatory influence of the rat major histocompatibility complex (MHC) (Ag-B complex) on the specificity of cytotoxic T lymphocytes was investigated. It was shown that the effector cells were specific for the original Ag-B phenotype in rat systems in which the responder and stimulator cell populations were unquestionably MHC identical but expressed different minor alloantigens of viral antigens. However, combined in vivo immunization and restimulation in culture of lymphocytes from rat strains previously thought to be MHC compatible resulted in the generation of cytotoxic T lymphocytes which effectively lyse not only target cells from the specific stimulating strains but also, to varying degrees, target cells from third party strains regardless of their Ag-B haplotypes. Genetic analysis indicates that expression of these cytotoxic T-cell-defined ("CT") antigens, found on both T and B lymphocytes, detectable thus far only with cytotoxic lymphocytes, is controlled by a single locus which segregates in backcross populations with the rat MHC. Discrepancies between the nature of CT antigens of the rat Ag-B and I-region specificities of the mouse H-2 are discussed.

Activation of cytokine genes in T cells during primary and secondary murine influenza pneumonia.

The patterns of cytokine mRNA expression in mice with primary or secondary influenza pneumonia have been assessed by in situ hybridization analysis of cells from both the mediastinal lymph node (MLN) and the virus-infected lung. Evidence of substantial transcriptional activity was found in all lymphocyte subsets recovered from both anatomical sites. The kinetics of cytokine mRNA expression after primary infection with an H3N2 virus were in accord with the idea that the initial response occurs in regional lymphoid tissue, with the effector T cells later moving to the lung. This temporal separation was much less apparent for the more rapid secondary response resulting from challenge of H3N2-primed mice with an H1N1 virus. Among the T cell receptor alpha/beta+ subsets, transcripts for interferon (IFN) gamma and tumor necrosis factor beta were most commonly found in the CD8+ population whereas mRNA for interleukin (IL) 4 and IL-10 was much more prevalent in CD4+ T cells. The gamma/delta T cells expressed mRNA for all cytokines tested, with IL-2, IL-4, and IFN-gamma predominating among those recovered from the inflammatory exudate. At particular time points, especially early in the MLN and late in the infected lung, the frequency of mRNA+ lymphocytes was much higher than would be expected from current understanding of the prevalence of virus-specific precursors and effectors. If this response is typical, induction of cytokine gene expression for T cells that are not responding directly to the invading pathogen may be a prominent feature of acute virus infections.

Clearance of influenza virus respiratory infection in mice lacking class I major histocompatibility complex-restricted CD8+ T cells.

Transgenic mice homozygous for a beta 2-microglobulin (beta 2-m) gene disruption and normal mice that had been treated with a CD8-specific mAb were infected intranasally with an H3N2 influenza A virus. Both groups of CD8T cell-deficient mice eliminated the virus from the infected respiratory tract. Potent CTL activity was detected in lung lavage populations taken from mice with intact CD8+ T cell function, with minimal levels of cytotoxicity being found for inflammatory cells obtained from the antibody-treated and beta 2-m mutant mice. We therefore conclude that cells infected with an influenza A virus can be cleared from the respiratory tract of mice lacking both functional class I major histocompatibility complex (MHC) glycoproteins and class I MHC-restricted, CD8+ effector T cells.

Generation of both cross-reactive and virus-specific T-cell populations after immunization with serologically distinct influenza A viruses.

Specificity of cytotoxic T-cell function was investigated for a range of different influenza viruses. T cells from mice immunized with A or B strain influenza viruses, or with vaccinia virus, showed reciprocal exclusion of cytotoxicity. Extensive cross-reactivity was, however, found for lymphocyte populations from mice infected with a variety of serologically distinct influenza A viruses, though serum antibodies did not cross-react when tested in a radioimmunoassay using comparable target cells as immunoadsorbents. This apparent lack of T-cell specificity was recognized for immune spleen cells generated after intraperitoneal inoculation of high titers of virus, and for mediastinal lymph node populations from mice with pneumonia due to infection with much less virus. The phenomenon could not be explained on the basis of exposure to the chicken host component, which is common to A and B strain viruses. However, not all of the virus-immune T-cell clones are cross-reactive. Competitive-inhibition experiments indicate that a considerable proportion of the lymphocyte response is restricted to the immunizing virus. Even so, the less specific component is significant. Also, exposure to one type A virus was found to prime for an enhanced cell-mediated immunity response after challenge with a second, serologically different A strain virus.

Antibody to influenza virus matrix protein detects a common antigen on the surface of cells infected with type A influenza viruses.

Antisera to the type-specific internal influenza virus matrix (M) protein of a type A influenza virus were produced in goats. In the presence of complement, anti-M serum was cytotoxic for target cells which were infected with a variety of serologically distinct type A influenza viruses, but did not react with type B influenza virus-infected cells. Absorption experiments indicated that anti-M serum detected a common antigen(s) on the surface of type A-infected cells. This serological cross-reactivity parallels the cross-reactivity observed for the cytotoxic T-cell response to type A viruses.

Inhibition of autoimmune neuropathological process by treatment with an iron-chelating agent.

Lewis rats that are primed with guinea pig spinal cord homogenized in complete Freund's adjuvant (GPSCH-CFA) develop overt symptoms of experimental allergic encephalomyelitis (EAE). Treatment with the iron-chelating agent, desferrioxamine B mesylate (DFOM), at various times before the onset of EAE, dramatically suppressed both the severity and duration of disease. When DFOM was administered to rats soon after the development of neurological signs, a rapid recovery occurred, though mild, transient symptoms could be seen approximately 1 wk after withdrawal of the drug. Treatment with DFOM was always accompanied by a diminution of T cell responsiveness on the part of the delayed-type hypersensitivity/helper subset and, on histological examination, an absence of inflammatory cells from lesions. Iron is believed to influence both the migration and function of immune effector cells. It can also act as a catalyst in the formation of free radicals, which are highly toxic agents causing tissue damage in sites of inflammation. The mechanisms underlying the effect of DFOM on the severity of EAE, and the possible implications for treatment of multiple sclerosis are discussed.

Progressive loss of CD8+ T cell-mediated control of a gamma-herpesvirus in the absence of CD4+ T cells.

A unique experimental model has been developed for dissecting the integrity of CD8+ T cell-mediated immunity to a persistent gammaherpesvirus under conditions of CD4+ T cell deficiency. Respiratory challenge of major histocompatibility complex class II -/- and +/+ C57BL/6J mice with the murine gammaherpesvirus 68 (MHV-68) leads to productive infection of both lung and adrenal epithelial cells. Virus titers peak within 5-10 d, and are no longer detected after day 15. Persistent, latent infection is established concurrently in splenic and lymph node B cells, with higher numbers of MHV-68+ lymphocytes being found in all lymphoid sites analyzed from the +/+ mice concurrent with the massive, but transient splenomegaly that occurred only in this group. From day 17, however, the numbers of infected B lymphocytes were consistently higher in the -/- group, while the frequency of this population diminished progressively in the +/+ controls. Infectious MHV-68 was again detected in the respiratory tract and the adrenals of the -/- (but not the +/+) mice from day 22 after infection. The titers in these sites rose progressively, with the majority of the -/- mice dying between days 120 and 133. Even so, some CD8+ effectors were still functioning as late as 100 d after infection. Depletion of CD8+ T cells at this stage led to higher virus titers in the -/- lung, and to the development of wasting in some of the -/- mice. Elimination of the CD8+ T cells from the +/+ group (day 80) increased the numbers of MHV-68+ cells in the spleen, but did not reactivate the infection in the respiratory tract. The results are consistent with the interpretation that CD8+ T cell-mediated control of this persistent gammaherpesvirus is progressively lost in the absence of the CD4+ T cell subset. This parallels what may be happening in AIDS patients who develop Kaposi's sarcoma and various Epstein Barr virus associated disease processes.

Involvement of H-2L gene products in virus-immune T-cell recognition. Evidence for an H-2L-restricted T-cell response.

The H-2L locus is closely linked to H-2D and codes for antigenic specificities present on a 45,000 mol wt glycoprotein that is distinct from the molecule which bears the D region private specificity. It was found that BALB/c-H-2db mice, which lack detectable cell-surface H-2L gene products, were able to generate influenza- and vaccinia-immune cytotoxic T cells which lyse D region-compatible target cells, although they have been reported to be incapable of making a similar response to ectromelia virus (7). Thus, the lack of H-2L antigenic specificities does not produce a general loss of responsiveness for other viruses even when a highly cross-reactive pox virus (vaccinia) was studied. Antisera-blocking experiments utilizing sera specific for either L or D molecules indicated that BALB/c mice generate influenza virus-immune cytotoxic T-cell subsets which independently recognize H-2L and H-2D gene products in association with viral antigens. These results are the first indication that products of the H-2L locus can operate analogously to H-2K/D gene products in virus-immune T-cell recognition.

Cytotoxic T-cell responses in mice infected with influenza and vaccinia viruses vary in magnitude with H-2 genotype.

Secondary effector T-cell populations generated by cross-priming with heterologous influenza A viruses operate only in H-2K or H-2D compatible situations, when assayed on SV40-transformed target cells infected with a range of influenza A viruses. The H2-Kb allele is associated with a total failure in the generation of influenza-immune cytotoxic T cells, though this is not seen for the primary response to vaccinia virus. In both influenza and vaccinia development of effector T cells operating at H-2Db is greatly depressed in B10.A(2R) (kkkddb) and B10.A(4R) (kkbbbb), but not in B10 (bbbbbb), mice. However, there is no defect in viral antigen expression at either H-2Kk or H-2Db in B10.A(2R) target cells. This apparently reflects some inadequacy in the stimulator environment, as (A/J X B6) F1 T cells can be induced to respond at H-2Db when exposed to vaccinia virus in an irradiated B6 but not in a B10.A(4R) recipient. The present report, together with the accompanying paper by Zinkernagel and colleagues, records the first rigorous demonstration of both a nonresponder situation and a probable Ir-gene effect for conventional infectious viruses. Possible implications for the evolution of H-2 polymorphism and mechanisms of Ir gene function are discussed.

H-2 compatibility requirement for virus-specific T-cell-mediated cytolysis. Evaluation of the role of H-2I region and non-H-2 genes in regulating immune response.

Lymphocytic choriomeningitis virus (LCMV) and ectromelia virus-specific T-cell-mediated cytotoxicity was assayed in various strain combinations using as targets peritoneal macrophages which have been shown to express Ia antigens. Virus-specific cytotoxicity was found only in H-2K- or D-region compatible combinations. I-region compatibility was not necessary nor alone sufficient for lysis. Six different I-region specificities had no obvious effect on the capacity to generate in vivo specific cytotoxicity (expressed in vitro) associated with Dd. Low LCMV-specific cytotoxic activity generated in DBA/2 mice was caused by the non-H-2 genetic background. This trait was inversely related to the infectious virus dose and recessive. Non-H-2 genes, possibly involved in controlling initial spread and multiplication of virus, seem to be, at least in the examples tested, more important in determining virus-specific cytotoxic T-cell activity in spleens than are Ir genes coded in H-2.

Phenotypic analysis of the inflammatory exudate in murine lymphocytic choriomeningitis.

The massive inflammation of the cerebrospinal fluid (CSF) which occurs in adult mice injected with lymphocytic choriomeningitis virus (LCMV) has been analyzed by flow microfluorometry (FMF). The great majority of the T cells detected by direct examination of freshly obtained CSF were found to be Lyt-2+, with an almost total absence of L3T4+ lymphocytes. The Lyt-2/L3T4 ratio of lymphocytes in blood was within normal limits. Predominance of the Lyt-2+ subset was confirmed by culturing the CSF cells after mitogenic stimulation. In addition, the T lymphocytes in CSF of cyclophosphamide-suppressed, virus-infected recipients that had been injected 4 d previously with LCMV-immune spleen cells were almost entirely donor Lyt-2+ cells, while the nonlymphoid elements were exclusively of host origin. However this pattern of donor and host T cell distribution was reversed when the LCMV-infected recipients were not immunosuppressed. The frequency of LCMV-specific CTL precursors in CSF taken immediately before the development of symptoms was as low as 1:3,000 cells. Thus most of the T lymphocytes extravasating into the CSF of mice with LCM are passive participants recruited as a consequence of the function of relatively few LCMV-specific effector T cells. The dominance of the Lyt-2+ T cell subset in the CSF of mice with LCM is intriguing.