Consequences of cyclophosphamide treatment in murine lymphocytic choriomeningitis: evidence for cytotoxic T cell replication in vivo

The pas de deux of viruses and CD8 T cells

The vertebrate immune system has evolved to deal with parasitic life forms, so it is hardly surprising that experiments with pathogens have proved illuminating for immunology. Those of us who have worked for years with infectious processes are acutely conscious that we do little more than probe the vast 'experiment of nature.' There is no place for doctrinaire rigidity in this extraordinarily complex area of biology. Though we have obviously tried to do rational experiments, much of the novelty that has been brought in to immunology from the analysis of the virus-specific host response has, in a very real sense, been the product of what we now call 'discovery science'. The following relates some of the research that I was involved in personally and attempts to put it in the context of both the history of the field and the events of the time. Virus infections, particularly HIV, pose enormous challenges for the future. It is generally helpful to know a little of what went before.

Analysis of virus-specific CD4(+) t cells during long-term gammaherpesvirus infection

Major histocompatibility complex class II-mediated antigen presentation after intranasal infection with murine gammaherpesvirus 68 differs in mediastinal lymph nodes and spleen. Evidence that virus-specific CD4(+) T cells were being stimulated was found as late as 6 to 8 months after infection, and cells specific for the viral gp150(67-83) and ORF11(168-180) peptides were maintained as a fairly stable proportion of the total response.

In vivo proliferation of naïve and memory influenza-specific CD8(+) T cells

The virus-specific CD8(+) T cell response has been analyzed through the development, effector, and recovery phases of primary and secondary influenza pneumonia. Apparently, most, if not all, memory T cells expressing clonotypic receptors that bind a tetrameric complex of influenza nucleoprotein (NP)(366-374) peptide+H-2D(b) (NPP) are induced to divide during the course of this localized respiratory infection. The replicative phase of the recall response ends about the time that virus can no longer be recovered from the lung, whereas some primary CD8(+)NPP(+) T cells may proliferate for a few more days. The greatly expanded population of CD8(+)NPP(+) memory T cells in the lymphoid tissue of secondarily challenged mice declines progressively in mean prevalence over the ensuing 100 days, despite the fact that at least some of these lymphocytes continue to cycle. The recall of cell-mediated immunity thus is characterized by massive proliferation of the antigen-specific CD8(+) set, whereas the extent of lymphocyte turnover in the absence of cognate peptide is variable, at a low level, and can be influenced by intercurrent infection.

T-cell responsiveness to LCMV segregates as a single locus in crosses between BALB/cA and C.B-17 mice. Evidence for regulation by a gene outside the Igh region

The course of systemic infection with lymphocytic choriomeningitis virus (LCMV) was studied in BALB/cA and C.B-17 mouse strains differing in the immunoglobulin heavy chain region (Igh). Susceptibility to intracerebral infection and the ability to clear the virus differed significantly between these presumably congenic strains, suggesting that a gene in the Igh region might influence the course of this infection. A difference in virus spread prior to appearance of the immune response could not explain the observed differences. On the other hand, the differences in course of infection correlated with a difference in virus-specific T-cell responsiveness measured in terms of virus-specific cytotoxicity in vitro and delayed-type hypersensitivity in vivo. Analysis of F1, BC1 and F2 progeny showed that differential T-cell responsiveness was influenced by a single gene or gene complex; however, no linkage was found between this locus and the Igh-C region. Taken together, these results indicate that an additional, and previously unknown, genetic difference exists between these two mouse strains, and that the involved locus carries a gene which significantly affects T-cell responsiveness to LCMV.

Dissection of an inflammatory process induced by CD8+ T cells

A massive delayed type hypersensitivity (DTH) reaction occurs in the cerebrospinal fluid (CSF) of mice with lymphocytic choriomeningitis (LCM). In this article, Peter Doherty and colleagues analyze this reaction together with the population dynamics of the regional lymph node to give a comprehensive picture of the events underlying this CD8+ T-cell-mediated immunopathological disease. Their findings are of general relevance to the understanding of inflammation.

Binding of monoclonal antibodies and T cell effector function in vivo

The capacity of adoptively transferred CD8+ effector T cells to induce meningitis in immunosuppressed, or unsuppressed, recipients infected with lymphocytic choriomeningitis virus (LCMV) may be diminished by prior incubation of the lymphocytes with IgM monoclonal antibodies (MAbs) specific for CD8 or Thy1.2. The same is true, though to a lesser extent, for the further proliferation of donor T cells in the spleens of the immunosuppressed mice. This inhibition of cell mediated immunity can be overcome, at least for the unsuppressed recipients, by increasing the numbers of cells that are transferred, even though exposure to Mab+ complement abrogates all cytotoxic T cell activity in vitro. The LCM model thus provides a quantitative system for assessing the consequences of MAb binding for T cell trafficking and effector function in vivo.

Persistence of the irradiated host component in thymocyte populations from bone marrow radiation chimeras infected with lymphocytic choriomeningitis virus

The thymus of chimeras made using T cell-depleted donor bone marrow from Thy1.1+ mice and 950 rad Thy 1.2+ recipients is dominated initially by cells expressing the Thy 1.2+ phenotype of the irradiated host. The thymocyte population recovered at 2 weeks after reconstitution comprises 80% Thy 1.2+ cells (host), the remainder being Thy 1.1+ (donor). This situation is normally reversed within a further week, with the host Ty 1.2+ (donor). This situation is normally reversed within a further week, with the host Thy 1.2+ thymocytes being present at a frequency of less than 5% from Week 4. Infection with lymphocytic choriomeningitis virus (LCMV) at 1 week after reconstitution with bone marrow causes a profound and persistent drop in the total number of thymocytes. The decline is equivalent for all categories of donor-derived thymocytes defined by two-color flow microfluorometric analysis for CD4 and CD8. However, there is a partial compensation by the retention of cells originating from the Thy 1.2+ host, which constitute 30-40% of the total thymocyte pool as late as 8 weeks after administration of bone marrow in the LCMV-infected chimeras. These radiation-resistant precursors give rise to CD4-8-, CD4-8+, CD4+8-, and CD4+8+ thymocytes, with the latter category being present at increased frequency. The potential skewing of the mature T cell repertoire as a consequence of persistent virus infection is discussed.

Consequences of exposure to ionizing radiation for effector T cell function in vivo

The adoptive transfer of actuely primed and memory virus-immune CD8+ T cells causes enhanced meningitis in both cyclophosphamide (Cy) suppressed, and unsuppressed, recipients infected with lymphocytic choriomeningitis virus (LCMV). The severity of meningitis is assessed by counting cells in cerebrospinal fluid (CSF) obtained from the cisterna magna, which allows measurement of significant inflammatory process ranging from 3 to more than 300 times the background number of cells found in mice injected with virus alone. Exposure of the donor immune population to ionizing radiation prior to transfer has shown that activated T cells from mice primed 7 or 8 days previously with virus may still promote a low level of meningitis in unsuppressed recipients following as much as 800 rads, while this effect is lost totally in Cy-suppressed mice at 600 rads. Memory T cells are more susceptible and show no evidence of in vivo effector function in either recipient population subsequent to 400 rads, a dose level which also greatly reduces the efficacy of acutely-primed T cells. The results are interpreted as indicating that heavily irradiated cells that are already fully functional show evidence of primary localization to the CNS and a limited capacity to cause pathology. Secondary localization, and events that require further proliferation of the T cells in vivo, are greatly inhibited by irradiation.

Contributions of host and donor T cells to the inflammatory process in murine lymphocytic choriomeningitis

The severe inflammation characteristic of the infection of adult mice with murine lymphocytic choriomeningitis virus (LCMV) is induced earlier in unsuppressed, virus-infected recipients by the adoptive transfer of class I MHC-compatible, CD4- CD8+ LCMV-immune spleen cell populations. The time to onset of fatal LCM may also be slightly diminished, though not to the extent that would be expected from the enhanced kinetics of the extravasation of cells into cerebrospinal fluid. The development of symptoms is thus not solely related to the magnitude of the inflammatory process. The majority of the T lymphocytes in the inflammatory exudate are of host origin and have the size characteristics of resting cells, while the minority population of donor T cells show more of a lymphoblast morphology. The findings are consistent with the idea that relatively few CD8+ virus-immune effectors trigger an inflammatory process which consists largely of secondarily recruited host T cells and monocyte/macrophages.

Immunogenetic analysis of cellular interactions governing the recruitment of T lymphocytes and monocytes in lymphocytic choriomeningitis virus-induced immunopathology

The Lyt2+ class I major histocompatibility complex (MHC)-restricted virus-immune T cells that induce murine lymphocytic choriomeningitis (LCM) are targeted onto radiation-resistant cells in the central nervous system of virus-infected mice. The use of appropriate bone marrow radiation chimeras as LCM virus-infected, (immunosuppressed recipients for immune T-cell transfer has established that, though bone marrow-derived cells can stimulate virus-specific cytotoxic T lymphocytes (CTL) in spleen, they do not reconstitute the barrier to T-cell recruitment from blood to cerebrospinal fluid. This is true for chimeras made up to 8 months previously, even though the inflammatory monocytes and macrophages in such chimeras are all of donor bone marrow origin. Radiation-resistant cells in the spleens of these chimeras are also still able to further stimulate virus-immune CTL. There is no requirement for H-2 compatibility between virus-immune T lymphocytes and secondarily recruited monocytes, or T cells of an inappropriate specificity. The key event in LCM immunopathology may thus be localization of T cells to the antigen-presenting endothelium in brain, leading to the secretion of mediators that promote the nonspecific recruitment of monocytes and other T cells.

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.

The acute inflammatory process in murine lymphocytic choriomeningitis is dependent on Lyt-2+ immune T cells

Virus-immune spleen cells induce fatal immunopathology following adoptive transfer into adult C57B1/6J mice that have been infected with lymphocytic choriomeningitis virus (LCMV) and immunosuppressed with cyclophosphamide. This is accompanied by the development of potent cytotoxic T-lymphocyte (CTL) activity of donor origin in the recipient spleen. Both the capacity to trigger the acute meningitis observed at 72 hr and to generate CTL effectors in lymphoid tissue are completely abrogated by the removal of Lyt-2+ cells from the donor population. However a lower level of inflammatory process in the central nervous system may emerge, in the absence of significant CTL function in recipient spleen, by 5 days after transfer of the Lyt-2-depleted cell population. Treatment of the transferred cells with antibody to the L3T4 marker does not reduce either the severity of inflammation or the level of CTL effector function in the recipient. Thus Lyt-2+ cells are required for the acute, fatal immunopathology characteristic of LCM, but it is not clear that in a more chronic situation, they are the sole effectors capable of triggering inflammatory process in this disease.

Anti-asialo GM1 eliminates both inflammatory process and cytotoxic T-cell function in the lymphocytic choriomeningitis adoptive transfer model

The induction of severe inflammatory process and fatal neurological disease by transfer of lymphocytic choriomeningitis virus (LCMV)-immune T cells into cyclophosphamide (Cy)-suppressed LCMV-infected mice is greatly inhibited by treatment of these recipients with antibody to the asialo GM1 ganglioside (anti-ASGM1). Examination of cytotoxic activity in lymphoid tissue of the Cy-suppressed recipients at 72 hr after cell transfer revealed that anti-ASGM1 treatment prevented the development of the cytotoxic T lymphocyte (CTL) response, even though the dose of antibody used did not significantly decrease CTL generation in unsuppressed mice. Abrogation of CTL activity was also observed following antibody treatment of NK-deficient (bg/bg) Cy-suppressed recipients, indicating that the anti-ASGM1 was unlikely to be operating via removal of NK cells that are in some way involved in the development of the CTL response. The possibility that anti-ASGM1 may act directly on T cells should be considered in all protocols involving the use of this reagent in immunosuppressed mice.

Expression of Pgp-1 (or Ly24) by subpopulations of mouse thymocytes and activated peripheral T lymphocytes

The expression of Pgp-1 (Ly24) by subpopulations of thymocytes was investigated and a subpopulation of Lyt-2-/L3T4-/J11d- thymocytes was identified which contained significant numbers (80%) of Pgp-1+ cells. Among freshly isolated lymph node T cells but not cortisone-resistant thymocytes, Pgp-1 expression was heterogeneous. Stimulation of T lymphocytes with either concanavalin A or the combination of phorbol myristate acetate plus calcium ionophore resulted in increased Pgp-1 expression which was found to be regulated independently of DNA synthesis and interleukin 2 receptor expression. T cells in the cerebrospinal fluid exudate of mice infected with lymphocytic choriomeningitis virus were also found to be Pgp-1+.

Differential effect of hybrid resistance on the localization of virus-immune effector T cells to spleen and brain

The hybrid resistance (Hr) effect operates in the lymphocytic choriomeningitis (LCM) in vivo transfer model to inhibit both the level of cytotoxicity T lymphocyte (CTL) generation in spleen and the induction of inflammation in cerebrospinal fluid (CSF). The effect is seen when LCM virus-immune T cells that are homozygous for H-2Db are injected into virus-infected, immunosuppressed recipients that are heterozygous for this allele, or into radiation chimeras that express an appropriate F1 phenotype. Evidence that Hr to T-cell transfer is cell-dose-dependent and tends to diminish with age was found in both chimeric and normal F1 mice. Inhibition of the capacity of injected T cells to cause meningitis is a more sensitive measure of Hr than is the further stimulation of CTL effectors in recipient lymphoid tissue. The injection of large numbers of H-2b virus-immune T cells into (H-2k X H-2bF1----H-2k) virus-infected recipients did not induce any cellular extravasation into CSF, though potent H-2b-restricted CTL effectors were generated in recipient spleen. Evidence of minimal inflammatory process was found in one experiment where these chimeras were given a comparable dose of (H-2b X H-2d)F1 immune spleen cells. Development of this T-cell-mediated immunopathological process depends essentially on the expression of the appropriate H-2 restriction element on radiation-resistant host cells which, in this case, presumably constitute part of the physiological barrier between blood and CSF.

Role of the major histocompatibility complex in targeting effector T cells into a site of virus infection

Bone marrow radiation chimeras have been used as virus-infected, cyclophosphamide-suppressed recipients to analyze the major histocompatibility complex (MHC) restriction constraints on the adoptive transfer of lymphocytic choriomeningitis by immune T cells. The basic protocol employed [(A X B)F1----(A X C)F1] chimeras, where A, B, C are different MHC haplotypes, and the establishment of appropriate chimerism was measured by the capacity of, for instance, the transferred A, T cells to generate MHC-restricted, virus-specific cytotoxic T lymphocytes (CTL) in the spleen of the [(A X B)F1----(B X C)F1] recipients. The experiments show quite clearly that maximal inflammatory process is only induced when donor bone marrow, irradiated recipient and transferred T cells share at least one MHC haplotype. Compatibility of the T cells and the radiation-resistant phenotype alone in, for instance, transfer of C immune lymphocytes into [(A X B)F1----(B X C)F1] recipients produced little inflammation if the chimeras had been established for at least 10 weeks. These results are compatible with a model which proposes that the transferred T cells first replicate in MHC-compatible lymphoid tissue and are then targeted onto the appropriate MHC plus virus expressed on cells in the blood-cerebrospinal fluid (CSF) barrier. An alternative postulate, that all that is required for the development of inflammatory process is MHC-restricted T cell replication in lymphoid tissue, with subsequent nonspecific localization to the central nervous system (CNS), was explored by transferring B, immune T cells into [(A X B)F1----A] recipients. The finding was that H-2b effectors did not cause any meningitis in [H-2kXb F1----H-2k] recipients, though potent H-2b-restricted CTL were generated in the spleens of comparable chimeras. However, in the H-2k immune----[H-2kXbF1----H-2b] transfer there was evidence of moderate inflammation that was about 9 times less severe than that caused by the H-2b effectors in comparable recipients. This indicates that, for maximal inflammatory process to occur, the T cells must encounter MHC-compatible, virus-infected cells in the CNS, with the effect being absolute in one strain combination and partial in another. Possible mechanisms underlying this divergence are discussed.

Natural killer cells contribute to inflammation but do not appear to be essential for the induction of clinical lymphocytic choriomeningitis

The inflammatory exudate found in cerebrospinal fluid (CSF) of mice 6 days after intracerebral infection with lymphocytic choriomeningitis virus (LCMV) contains substantial populations of both cytotoxic T lymphocytes (CTL) and natural killer (NK) cells. Removal of NK cell activity by in vivo treatment with antibody to the asialo GM1 ganglioside and studies with NK-deficient bg/bg mice did not clearly determine whether NK cells contribute in any way to the development of clinical LCM. However, the LCM disease process induced in cyclophosphamide-suppressed, LCMV-infected recipients by the adoptive transfer of LCMV-immune spleen cells occurs in the absence of NK cell effector function in spleen, lymph nodes, or CSF of the recipients, though potent CTL populations are present in all of these sites. In this situation, NK cells are apparently not required for the induction of neurological symptoms that are indistinguishable from those of classical LCM.