Mouse hepatitis virus is cleared from the central nervous systems of mice lacking perforin-mediated cytolysis.

Perforin-deficient [perforin (-/-)] mice were infected with two strains of JHM virus (JHMV) to analyze the role of perforin-mediated cytotoxicity in acute lethal and subacute central nervous system (CNS) infections. During both acute and subacute infections, the overall mortality of the perforin (-/-) mice was not different from that of the controls. Perforin (-/-) mice survived longer than the controls, consistent with reduced morbidity. Both strains of virus were cleared from the perforin (-/-) mice as in the controls; however, the rate of clearance was delayed in the perforin (-/-) mice, indicating that perforin-mediated cytolysis is involved in viral clearance. The absence of perforin-mediated cytolysis did not prevent encephalomyelitis or extensive demyelination. Cells undergoing apoptosis were detected in the CNS of both the perforin (-/-) and control groups, indicating that perforin is not essential for programmed cell death. Neutralizing antibodies were not detected in either group of mice until day 9 postinfection, when the majority of the virus had been cleared. These data further confirm the importance of cell-mediated cytotoxicity and suggest that additional components of the immune response contribute to the clearance of JHMV from the CNS.     

Cutting edge: tumor rejection mediated by NKG2D receptor-ligand interaction is dependent upon perforin

We have investigated the primary immunity generated in vivo by MHC class I-deficient and -competent tumor cell lines that expressed the NKG2D ligand retinoic acid early inducible-1 (Rae-1) beta. Rae-1beta expression on class I-deficient RMA-S lymphoma cells enhanced primary NK cell-mediated tumor rejection in vivo, whereas RMA-Rae-1beta tumor cells were rejected by a combination of NK cells and CD8(+) T cells. Rae-1beta expression stimulated NK cell cytotoxicity and IFN-gamma secretion in vitro, but not proliferation. Surprisingly, only NK cell perforin-mediated cytotoxicity, but not production of IFN-gamma, was critical for the rejection of Rae-1beta-expressing tumor cells in vivo. This distinct requirement for perforin activity contrasts with the NK cell-mediated rejection of MHC class I-deficient RMA-S tumor cells expressing other activating ligands such as CD70 and CD80. Thus, these results indicated that NKG2D acted as a natural cytotoxicity receptor to stimulate perforin-mediated elimination of ligand-expressing tumor cells.     

CD8(+) T cells secreting type 2 lymphokines are defective in protection against viral infection

Effector T cells secreting type 1 and/or type 2 lymphokines (Tc1, Tc0, Tc2) were generated in vitro from CD8(+) T cells of mice with a transgenic TCR recognizing lymphocytic choriomeningitis virus (LCMV) glycoprotein to compare their effector function in vitro and in vivo. Tc1, Tc2, and Tc0 showed similar Fas- and perforin-mediated cytotoxicity in vitro. Upon adoptive transfer, Tc2 and Tc0 effectors were less efficient than Tc1 at controlling LCMV or recombinant vaccinia virus expressing the LCMV glycoprotein in vivo. Tc2 and Tc0 had decreased surface VLA-4 density and deficient activation-induced LFA-1/ICAM-1-dependent homotypic adhesion in vitro. Therefore, the reduced antiviral activity in vivo of Tc2 and Tc0 compared with Tc1 is not due to reduced cytotoxic activity or IFN-gamma secretion but may be explained by defective homing to the target organ due to decreased expression and/or lower activity of adhesion molecules.     

Multifunctional CD4 cells expressing gamma interferon and perforin mediate protection against lethal influenza virus infection

CD4 effectors generated in vitro can promote survival against a highly pathogenic influenza virus via an antibody-independent mechanism involving class II-restricted, perforin-mediated cytotoxicity. However, it is not known whether CD4 cells activated during influenza virus infection can acquire cytolytic activity that contributes to protection against lethal challenge. CD4 cells isolated from the lungs of infected mice were able to confer protection against a lethal dose of H1N1 influenza virus A/Puerto Rico 8/34 (PR8). Infection of BALB/c mice with PR8 induced a multifunctional CD4 population with proliferative capacity and ability to secrete interleukin-2 (IL-2) and tumor necrosis factor alpha (TNF-α) in the draining lymph node (DLN) and gamma interferon (IFN-γ) and IL-10 in the lung. IFN-γ-deficient CD4 cells produced larger amounts of IL-17 and similar levels of TNF-α, IL-10, and IL-2 compared to wild-type (WT) CD4 cells. Both WT and IFN-γ(-/-) CD4 cells exhibit influenza virus-specific cytotoxicity; however, IFN-γ-deficient CD4 cells did not promote recovery after lethal infection as effectively as WT CD4 cells. PR8 infection induced a population of cytolytic CD4 effectors that resided in the lung but not the DLN. These cells expressed granzyme B (GrB) and required perforin to lyse peptide-pulsed targets. Lethally infected mice given influenza virus-specific CD4 cells deficient in perforin showed greater weight loss and a slower time to recovery than mice given WT influenza virus-specific CD4 cells. Taken together, these data strengthen the concept that CD4 T cell effectors are broadly multifunctional with direct roles in promoting protection against lethal influenza virus infection.     

Critical Role of Perforin-dependent CD8+ T Cell Immunity for Rapid Protective Vaccination in a Murine Model for Human Smallpox

Vaccination is highly effective in preventing various infectious diseases, whereas the constant threat of new emerging pathogens necessitates the development of innovative vaccination principles that also confer rapid protection in a case of emergency. Although increasing evidence points to T cell immunity playing a critical role in vaccination against viral diseases, vaccine efficacy is mostly associated with the induction of antibody responses. Here we analyze the immunological mechanism(s) of rapidly protective vaccinia virus immunization using mousepox as surrogate model for human smallpox. We found that fast protection against lethal systemic poxvirus disease solely depended on CD4 and CD8 T cell responses induced by vaccination with highly attenuated modified vaccinia virus Ankara (MVA) or conventional vaccinia virus. Of note, CD4 T cells were critically required to allow for MVA induced CD8 T cell expansion and perforin-mediated cytotoxicity was a key mechanism of MVA induced protection. In contrast, selected components of the innate immune system and B cell-mediated responses were fully dispensable for prevention of fatal disease by immunization given two days before challenge. In conclusion, our data clearly demonstrate that perforin-dependent CD8 T cell immunity plays a key role in MVA conferred short term protection against lethal mousepox. Rapid induction of T cell immunity might serve as a new paradigm for treatments that need to fit into a scenario of protective emergency vaccination.     

Essential roles of perforin in antigen-specific cytotoxicity mediated by human CD4+ T lymphocytes: analysis using the combination of hereditary perforin-deficient effector cells and Fas-deficient target cells

Although the cytotoxic mechanisms of murine CTLs have been investigated extensively using various mutant and knockout mice, those of human CTLs, especially CD4+ CTLs, are still obscure. To clarify the roles of perforin in Ag-specific cytotoxicity mediated by human CD4+ CTLs, alloantigen-specific and HSV-specific human CD4+ T lymphocyte bulk lines and clones were established from a patient with hereditary perforin deficiency and her healthy father, and their cytotoxic activities were investigated. Alloantigen-specific CD4+ T lymphocytes expressing perforin exerted cytotoxicity against Fas-negative as well as Fas-positive allogeneic B lymphoblastoid cell lines established from members of a family with hereditary Fas deficiency. Perforin-deficient, but not perforin-expressing, CD4+ T lymphocytes failed to show strong cytotoxicity against HSV-infected autologous B lymphoblastoid cells. Perforin-deficient CD4+ T lymphocytes could exert relatively low level cytotoxicity against allogeneic IFN-gamma-treated keratinocytes. Although cytotoxicity mediated by perforin-expressing CD4+ CTLs was almost completely inhibited by concanamycin A, a potent inhibitor of the perforin-mediated cytotoxic pathway, cytotoxicity against IFN-gamma-treated keratinocytes mediated by perforin-deficient CD4+ T lymphocytes was inhibited only partially by concanamycin A, but was inhibited significantly by antagonistic anti-Fas Ab and anti-Fas ligand Ab. The combination of perforin-deficient effector T lymphocytes and Fas-negative target cells used in the present study provides a novel experimental system for studying the detailed mechanisms of human CTL-mediated cytotoxicity. The present data demonstrate that perforin-negative CD4+ CTLs can exert cytotoxicity against Fas-sensitive target cells; however, perforin plays essential roles in Ag-specific cytotoxicity mediated by human CD4+ as well as CD8+ CTLs.     

CD4 T cells contribute to virus control and pathology following central nervous system infection with neurotropic mouse hepatitis virus

Replication of the neurotropic mouse hepatitis virus strain JHM (JHMV) is controlled primarily by CD8(+) T-cell effectors utilizing gamma interferon (IFN-gamma) and perforin-mediated cytotoxicity. CD4(+) T cells provide an auxiliary function(s) for CD8(+) T-cell survival; however, their direct contribution to control of virus replication and pathology is unclear. To examine a direct role of CD4(+) T cells in viral clearance and pathology, pathogenesis was compared in mice deficient in both perforin and IFN-gamma that were selectively reconstituted for these functions via transfer of virus-specific memory CD4(+) T cells. CD4(+) T cells from immunized wild-type, perforin-deficient, and IFN-gamma-deficient donors all initially reduced virus replication. However, prolonged viral control by IFN-gamma-competent donors suggested that IFN-gamma is important for sustained virus control. Local release of IFN-gamma was evident by up-regulation of class II molecules on microglia in recipients of IFN-gamma producing CD4(+) T cells. CD4(+) T-cell-mediated antiviral activity correlated with diminished clinical symptoms, pathology, and demyelination. Both wild-type donor CD90.1 and recipient CD90.2 CD4(+) T cells trafficked into the central nervous system (CNS) parenchyma and localized to infected white matter, correlating with decreased numbers of virus-infected oligodendrocytes in the CNS. These data support a direct, if limited, antiviral role for CD4(+) T cells early during acute JHMV encephalomyelitis. Although the antiviral effector mechanism is initially independent of IFN-gamma secretion, sustained control of CNS virus replication by CD4(+) T cells requires IFN-gamma.     

Apoptotic neurodegeneration induced by influenza A virus infection in the mouse brain

Neurodegeneration in the brain induced by the WSN strain of influenza A virus was investigated after stereotaxic introduction into the olfactory bulb of C57BL/6 mice. Immunohistochemistry detected WSN virus-infected neurons in the anterior olfactory nucleus as early as day 3 postinfection. Thereafter, they became shrunken and showed loss of neurite-immunolabeling and chromatin condensation. Infected neurons died by day 12, degenerating into multiple small granular bodies. The terminal deoxynucleotidyltransferase-mediated dUTP nick end-labeling method demonstrated DNA fragmentation in infected neurons at day 7 and also in such granular bodies at day 12. In perforin-deficient mice, the appearance of virally induced apoptotic neurodegeneration was delayed and virus infection continued for a longer period of 35 days postinfection. These findings indicate that perforin-mediated neuroapoptosis appears significant in exterminating the intracellular pathogen at an early stage of infection.     

Intimate cell conjugate formation and exchange of membrane lipids precede apoptosis induction in target cells during antibody-dependent, granulocyte-mediated cytotoxicity

Ab-dependent polymorphonuclear granulocyte (PMN)-mediated cytotoxicity may play an important role in the control of malignant diseases. However, little is known as to which particular pathways are used for the killing of malignant cells by PMN. The production of reactive oxygen intermediates (ROI) has been observed to occur during Ab-dependent, cell-mediated cytotoxicity (ADCC). However, PMN from a patient with chronic granulomatous disease demonstrated strong ADCC against malignant lymphoma cells. Furthermore, the inhibition of ROI production in PMN from healthy donors had no significant effect on ADCC. Therefore, ROI production by the NADPH oxidase of PMN does not appear to be mandatory for PMN-mediated ADCC. Recent data suggest a role for perforins in PMN-mediated cytotoxicity. However, in our assays concanamycin A, an inhibitor of perforin-mediated ADCC by mononuclear cells, had no inhibitory effect on PMN-mediated ADCC. Using electron microscopy we observed that PMN and their target cells intimately interact with the formation of interdigitating membrane protrusions. During PMN and target cell contact there was a mutual exchange of fluorescent membrane lipid dyes that was strongly increased in the presence of tumor-targeting Abs. This observation may be closely related to the recently described process of trogocytosis by lymphocytes. The presence of transient PMN-tumor cell aggregates and the accumulation of PMN with tumor cell-derived membrane lipids and vice versa were associated with effective ADCC as measured by chromium-release or apoptosis induction.     

Granzyme B activity in target cells detects attack by cytotoxic lymphocytes

Lymphocyte-mediated cytotoxicity via granule exocytosis operates by the perforin-mediated transfer of granzymes from CTLs and NK cells into target cells where caspase activation and other death pathways are triggered. Granzyme B (GzB) is a major cytotoxic effector in this pathway, and its fate in target cells has been studied by several groups using immunodetection. In this study, we have used a newly developed cell-permeable fluorogenic GzB substrate to measure this protease activity in three different living targets following contact with cytotoxic effectors. Although no GzB activity is measurable in CTL or NK92 effector cells, this activity rapidly becomes detectable throughout the target cytoplasm after effector-target engagement. We have combined the GzB substrate with a second fluorogenic substrate selective for caspase 3 to allow both flow cytometry and fluorescence confocal microscopy studies of cytotoxicity. With both effectors, caspase 3 activity appears subsequent to that of GzB inside all three targets. Overexpression of Bcl-2 in target cells has minimal effects on lysis, NK- or CTL-delivered GzB activity, or activation of target caspase 3. Detection of target GzB activity followed by caspase 3 activation provides a unique readout of a potentially lethal injury delivered by cytotoxic lymphocytes.     

Lymphocyte-mediated cytotoxicity to autologous cells infected with measles virus. II. Specificity of the cytotoxic reaction and characterization of the effector cells involved.

The mechanism of lymphocyte-mediated cytotoxicity to cells infected with measles virus was investigated. Cytotoxicity was measured in a direct assay, immediately after the isolation of lymphocytes from human peripheral blood; mononuclear leukocytes, infected with measles virus in vitro, served as autologous target cells. Virus-specific cytotoxicity required the presence of both IgG antibodies against measles virus and of effector lymphocytes. The effector lymphocytes had Fc receptors and were mainly present in a fraction of non-T lymphocytes. Monocytes were not cytotoxic but rather inhibitory. These results indicate that lysis of virus-infected cells in this direct assay is due to antibody-dependent cellular cytotoxicity (ADCC), caused by K cells. Control experiments showed that the virus-infected target cells were sensitive to incubation with human serum or IgG, resulting in a nonspecific increase of ⁵¹Cr release; however, this did not affect the results of K-cell cytotoxicity. Maximal virus-specific lysis by ADCC did not reach the level obtained by complement-dependent cytotoxicity. Possible explanations for this difference are discussed.     

EBV latent membrane proteins (LMPs) 1 and 2 as immunotherapeutic targets: LMP-specific CD4+ cytotoxic T cell recognition of EBV-transformed B cell lines

The EBV-latent membrane proteins (LMPs) 1 and 2 are among only three viral proteins expressed in EBV-associated Hodgkin's lymphoma and nasopharyngeal carcinoma. Since these tumors are HLA class I and class II-positive, the LMPs could serve as both CD8+ and CD4+ T cell targets. In contrast to CD8 responses, very little is known about CD4 responses to LMPs. In this study, we describe CD4+ T cell clones defining four LMP1- and three LMP2-derived peptide epitopes and their restricting alleles. All clones produced Th1-like cytokines in response to peptide and most killed peptide-loaded target cells by perforin-mediated lysis. Although clones to different epitopes showed different functional avidities in peptide titration assays, avidity per se was a poor predictor of the ability to recognize naturally infected B lymphoblastoid cell lines (LCLs) expressing LMPs at physiologic levels. Some epitopes, particularly within LMP1, consistently mediated strong LCL recognition detectable in cytokine release, cytotoxicity, and outgrowth inhibition assays. Using cyclosporin A to selectively block cytokine release, we found that CD4+ T cell cytotoxicity is the key effector of LCL outgrowth control. We therefore infer that cytotoxic CD4+ T cells to a subset of LMP epitopes could have therapeutic potential against LMP-expressing tumors.     

Disparate cytotoxic activity of nickel-specific CD8+ and CD4+ T cell subsets against keratinocytes

Allergic contact dermatitis (ACD) is the result of an exaggerated immune reaction to haptens mediated by skin-homing T cells, but the effector mechanisms responsible for the tissue damage are poorly understood. Here we studied the capacity of distinct subsets of hapten-specific T cells to induce apoptosis in autologous keratinocytes. Skin- and blood-derived nickel-specific CD8+ T cytotoxic 1 (Tc1) and Tc2 clones as well as CD4+ Th1 and Th2 expressed the cutaneous lymphocyte-associated Ag and exhibited strong MHC-restricted cytotoxicity against nickel-coupled B lymphoblasts, as detected by the [3H]TdR release assay. Both Tc1 and Tc2 clones, but not CD4+ T cells, displayed a significant cytotoxic activity against resting nickel-modified keratinocytes. Following IFN-gamma treatment, keratinocytes expressed MHC class II and ICAM-1 and became susceptible to Th1-mediated, but not Th2-mediated, cytotoxicity. The molecules of the two major cytotoxic pathways, Fas ligand (FasL) and perforin, were expressed by Tc1, Tc2, and Th1 cells, whereas Th2 cells expressed only FasL. Experiments performed in the presence of specific inhibitors of the perforin (concanamycin A) and FasL (brefeldin A) pathway indicated that perforin-mediated killing dominated in Tc1 and Tc2, and FasL-mediated cytotoxicity prevailed in Th2 clones, with a more heterogeneous behavior in the case of Th1 cells. Finally, perforin mRNA was expressed in ACD lesional skin, as assessed by RT-PCR analysis. In aggregate, our results indicate that keratinocytes can be target of multiple hapten-specific CTL responses, that may have distinct roles in the epidermal injury during ACD.     

Perforin-mediated effector function within the central nervous system requires IFN-gamma-mediated MHC up-regulation

CD8(+) T cells infiltrating the CNS control infection by the neurotropic JHM strain of mouse hepatitis virus. Differential susceptibility of infected cell types to clearance by perforin or IFN-gamma uncovered distinct, nonredundant roles for these antiviral mechanisms. To separately evaluate each effector function specifically in the context of CD8(+) T cells, pathogenesis was analyzed in mice deficient in both perforin and IFN-gamma (PKO/GKO) or selectively reconstituted for each function by transfer of CD8(+) T cells. Untreated PKO/GKO mice were unable to control the infection and died of lethal encephalomyelitis within 16 days, despite substantially higher CD8(+) T cell accumulation in the CNS compared with controls. Uncontrolled infection was associated with limited MHC class I up-regulation and an absence of class II expression on microglia, coinciding with decreased CD4(+) T cells in CNS infiltrates. CD8(+) T cells from perforin-deficient and wild-type donors reduced virus replication in PKO/GKO recipients. By contrast, IFN-gamma-deficient donor CD8(+) T cells did not affect virus replication. The inability of perforin-mediated mechanisms to control virus in the absence of IFN-gamma coincided with reduced class I expression. These data not only confirm direct antiviral activity of IFN-gamma within the CNS but also demonstrate IFN-gamma-dependent MHC surface expression to guarantee local T cell effector function in tissues inherently low in MHC expression. The data further imply that IFN-gamma plays a crucial role in pathogenesis by regulating the balance between virus replication in oligodendrocytes, CD8(+) T cell effector function, and demyelination.     

Perforin and gamma interferon-mediated control of coronavirus central nervous system infection by CD8 T cells in the absence of CD4 T cells

Infection of the central nervous system (CNS) with the neurotropic JHM strain of mouse hepatitis virus produces acute and chronic demyelination. The contributions of perforin-mediated cytolysis and gamma interferon (IFN-gamma) secretion by CD8(+) T cells to the control of infection and the induction of demyelination were examined by adoptive transfer into infected SCID recipients. Untreated SCID mice exhibited uncontrolled virus replication in all CNS cell types but had little or no demyelination. Memory CD8(+) T cells from syngeneic wild-type (wt), perforin-deficient, or IFN-gamma-deficient (GKO) donors all trafficked into the infected CNS in the absence of CD4(+) T cells and localized to similar areas. Although CD8(+) T cells from all three donors suppressed virus replication in the CNS, GKO CD8(+) T cells expressed the least antiviral activity. A distinct viral antigen distribution in specific CNS cell types revealed different mechanisms of viral control. While wt CD8(+) T cells inhibited virus replication in all CNS cell types, cytolytic activity in the absence of IFN-gamma suppressed the infection of astrocytes, but not oligodendroglia. In contrast, cells that secreted IFN-gamma but lacked cytolytic activity inhibited replication in oligodendroglia, but not astrocytes. Demyelination was most severe following viral control by wt CD8(+) T cells but was independent of macrophage infiltration. These data demonstrate the effective control of virus replication by CD8(+) T cells in the absence of CD4(+) T cells and support the necessity for the expression of distinct effector mechanisms in the control of viral replication in distinct CNS glial cell types.     

Cell death-associated translocation of plasma membrane components induced by CTL

In the very early stages of target cell apoptosis induced by CTL, we found that fluorescence of labeling probes of the target plasma membrane, such as N-(3-triethylammoniumpropyl)-4-(p-dibutylaminostyryl)pyridin ium dibromide (FM1-43), was translocated into intracellular membrane structures including nuclear envelope and mitochondria. This translocation was associated with the execution of CTL-mediated killing, because neither the CTL-target conjugation alone nor the binding of noncytotoxic Th2 clone with target cell was sufficient to provoke the process. Although FM1-43 translocation was observed in perforin-mediated cytotoxicity, examinations with several other dyes failed to detect the evidence for membrane damages that may cause influx of the dye. Moreover, the translocation was also observed in Fas-dependent apoptosis. These data indicate that the translocation precedes the damage of plasma membrane and intracellular organella in the course of apoptotic cell death and may represent the existence of a membrane trafficking that mediates the translocation of plasma membrane components in the early onset of apoptotic cell death.     

Resistance of cytolytic lymphocytes to perforin-mediated killing. Inhibition of perforin binding activity by surface membrane proteins

The mechanism whereby cytolytic lymphocytes protect themselves from killing mediated by their own cytotoxic protein, perforin, was studied. By using a competition assay, we demonstrated that the resistance of cells to perforin-mediated cytolysis is inversely correlated with their ability to absorb perforin, with tumor cells and noncytotoxic lymphocytes that are susceptible to perforin-mediated lysis being able to absorb perforin from the supernatant much better than CTL. The evidence implies that there is molecule on cytolytic lymphocytes that interferes with perforin-binding activity, resulting in the inability of perforin to lyse these cells. The molecule is most likely a surface protein or complex of proteins because its activity decreases after CTL treatment with the proteolytic enzymes trypsin and papain, and the activity can be recovered by incubation of the treated CTL cells at 37 degrees C for 6 h. The recovery can be blocked by emetine, cycloheximide, and actinomycin D, inhibitors of protein and RNA/DNA synthesis. The protein contains carbohydrate groups that play an important role in the function of the protein, as indicated by the fact that inhibition of glycosylation by tunicamycin and cleavage of sialic acid from the protein with neuraminidase result in a significant increase of perforin binding to CTL. Cross-linkage of CTL membrane proteins with glutaraldehyde and formaldehyde and blockage of the functional domains of the protein with an antiserum against CTL also inhibit the activity of this protein. Temperature-dependence studies that allow for a dissociation of the binding and pore-forming stages of perforin-mediated hemolysis suggest that the protective protein interferes at the perforin-binding stage.     

Perforin-deficient CD8+ T cells mediate fatal lymphocytic choriomeningitis despite impaired cytokine production

Intracerebral (i.c.) infection with lymphocytic choriomeningitis virus (LCMV) is one of the most studied models for virus-induced immunopathology, and based on results from perforin-deficient mice, it is currently assumed that fatal disease directly reflects perforin-mediated cell lysis. However, recent studies have revealed additional functional defects within the effector T cells of LCMV-infected perforin-deficient mice, raising the possibility that perforin may not be directly involved in mediating lethal disease. For this reason, we decided to reevaluate the role of perforin in determining the outcome of i.c. infection with LCMV. We confirmed that the expansion of virus-specific CD8(+) T cells is unimpaired in perforin-deficient mice. However, despite the fact that the virus-specific CD8(+) effector T cells in perforin-deficient mice are broadly impaired in their effector function, these mice invariably succumb to i.c. infection with LCMV strain Armstrong, although a few days later than matched wild-type mice. Upon further investigation, we found that this delay correlates with the delayed recruitment of inflammatory cells to the central nervous system (CNS). However, CD8(+) effector T cells were not kept from the CNS by sequestering in infected extraneural organ sites such as liver or lungs. Thus, the observed dysfunctionality regarding the production of proinflammatory mediators probably results in the delayed recruitment of effector cells to the CNS, and this appears to be the main explanation for the delayed onset of fatal disease in perforin-deficient mice. However, once accumulated in the CNS, virus-specific CD8(+) T cells can induce fatal CNS pathology despite the absence of perforin-mediated lysis and reduced capacity to produce several key cytokines.