Immune responses to LCM virus infection in vivo and in vitro

Immune responses to LCM virus infection in vivo and in vitro Mechanisms of immune-mediated disease* G. A. COLE 1 & E. D. JOHNSON 2 Both recovery and death of mice following acute infections with lymphocytic chorio- meningitis virus appear to be mediated by a population of virus-specific thymus-derived (T) effector lymphocytes that possess lytic activity in vitro against virus-infected syngeneic fibroblasts. Whether recovery or death occurs is determined by the balance between two interdependent factors: (1) the extent of virus-induced modifications in the surfaces of cells comprising " target " tissues, and (2) the efficiency of the immune inductive process leading to the generation of effector T lymphocytes that recognize and destroy these modified cells. INTRODUCTION Since its original isolation more than 30 years ago, lymphocytic choriomeningitis (LCM) virus has re- mained both an enigma and a delight to the numer- ous experimentalists who have studied the diversity of responses elicited by this agent in the mouse, its natural host, as well as in other animal species. After long remaining a singular oddity, divorced from other major families of animal viruses by its appar- ent lack of serological or morphological relatives, LCM virus is now recognized as the prototype of the newly emerged arenavirus group. Our present knowledge of the immunopathological phenomena that occur during naturally acquired or experimental LCM virus infections of the mouse can serve as a background against which the pathogenesis of other arenavirus infections of man can be compared. GENERAL FEATURES OF MURINE INFECrIONS An appreciation of the biology of LCM virus came first from the pioneering studies of Traub (1-3) followed by those of Rowe (4) and Hotchin (5). Collectively, these investigators defined most of the experimental conditions under which the administra- tion of virus to mice could lead to a life-long carrier * From the Division of Infectious Disease, Department of Epidemiology, School of Hygiene and Public Health, The Johns Hopkins University, Baltimore, MD 21205, USA. Professor. 'Predoctoral fellow. state, an abortive immunizing infection, or an acute- ly fatal disease. The work of Rowe and his colleagues (4, 6) unequivocally established that fatal cerebral infec- tion, or its visceral analogue, was immune-mediated and it also clarified the differences in tissue tropisms that exist between different virus strains. Thus, the nature of the acute disease produced in adult mice by viscerotropic strains depends on the dose of virus and the route by which it is given. The pathological hallmarks of infection with low doses range from destructive lesions of the lung, liver, and kidney following peripheral inoculation, to choriomeningitis following intracerebral inoculation (7). Recovery from visceral infection is accompanied by a solid immunity to a second, normally lethal, virus chal- lenge. Infection with large doses, however, may lead to the induction of a persistent carrier state (6, 8), the immunological aspects of which have been given little attention. On the other hand, small or large doses of brain- passaged or " neurotropic " strains replicate ineffi- ciently in adult mouse parenchymal tissues and produce an abortive immunizing infection when administered extraneurally, yet they uniformly kill when given intracerebrally. The inoculation of newborn mice, by any route, with all common strains ofLCM virus usually results in the survival of a high proportion of animals who subsequently develop permanent carrier infections. At any time during their lifespan, high titres of virus 3372 - 465 - BULL. WORLD HEALTH ORGAN., Vol. 52, 1975 G. A. COLE & E. D. JOHNSON are readily demonstrable in blood and many paren- chymal tissues of carrier mice, and the infection is perpetuated by congenital transmission to their progeny (3, 7, 8). For many years, the LCM virus carrier state was considered to be the paradigm of specific immunol- ogical tolerance since, by conventional means, no virus-specific antibodies could be detected in carrier mouse sera. This notion was dispelled, however, when Oldstone & Dixon (9, 10) using specific fluoresceinated antisera to stain sections of kidneys from carrier mice, were able to demonstrate the presence of glomerular deposits containing viral antigens, IgG, and complement. In addition, IgG eluted from carrier mouse kid- neys contained virus-specific complement-fixing anti- body. These observations established that LCM virus carrier infections are accompanied by a humoral response directed against one or more virus-specific determinants resulting eventually in chronic immune complex disease. In additional studies (11), Oldstone and his associates suggested that the severity of immune-mediated disease may be determined by genes that map within the H-2 gene complex. In retrospect, these findings appear to provide the basis for an explanation of the appearance of a runting syndrome in certain strains of mice with long- standing carrier infections and termed " late dis- ease" by Hotchin (5). the virus-induced immune response components generated during acute LCM virus infection is responsible for the elicitation of, and recovery from, disease. REQUIREMENT FOR T LYMPHOCYTES As first shown by studies from our laboratory (12, 13), splenic lymphocytes obtained from mice im- munized by intraperitoneal injection with Armstrong virus, when adoptively transferred to syngeneic CY- induced virus carrier BALB/c mice, regularly pro- duced an acute lethal central nervous system (CNS) disease, which, clinically and histologically, was typical of classical LCM. The administration of serum with high virus-specific complement-fixing activity was without effect. Furthermore, both CNS disease and virus clearance mediated by exogenous immune lymphoid cells required the participation of T lymphocytes (14). Depletion of this population with anti-C serum and complement completely abro- gated the ability of the transferred cells to elicit LCM, although their capacity to produce virus- specific complement-fixing antibody was retained. These findings, together with the fact that neutraliz- ing antibody is not detected during acute infections, constitute a strong argument against any significant role of antibody in mediating acute disease. CELLULAR IMMUNE REACTIVITY IN VITRO THE EFFECT OF IMMUNOSUPPRESSION The most convincing evidence for the immuno- pathological basis of acute LCM comes from the dramatic effect of immunosuppression on the course of the disease. Since Rowe's observation that irradia- tion prior to infection prevented or delayed the onset of disease (4) a number of investigators have shown that other immunosuppressive procedures are also effective to a greater or lesser degree (see review by Cole et al., 12). One of the most efficient ways to convert a potentially lethal cerebral infection to a carrier state is by the use of cyclophosphamide (CY). A single immunosuppressive dose of this drug given to adult mice within 2-3 days after intracerebral inoculation of the neurotropic Armstrong strain of LCM virus prevents the development of immune-mediated choriomeningitis and results in the development of a chronic asymptomatic carrier infection (12). This simple procedure, combined with adoptive immuni- zation, provided the means of determining which of The characteristic ability of LCM virus to estab- lish a persistent infection in cultured mouse fibro- blasts without accompanying cytopathology pro- vides a readily obtainable population of " target " cells against which the lytic (effector) activity of lymphoid cells from infected or immune mice can be measured. When 61Cr-labelled infected target cells are co-cultivated overnight with syngeneic lymphoid cells, any released radioactive label can be quanti- tated and directly correlated with the effector activity of the lymphoid population. Using this procedure, we have shown that effector lymphocytes appear in the spleens of adult mice about 4 days after an immunizing infection with the Armstrong strain of LCM virus (15). The response reaches a peak after 8-10 days and then declines rapidly, becoming undetectable by 30 days. Effector activity is not displayed against uninfected targets; it is abolished by treatment with anti-C serum and complement and is, therefore, a function of virus- specific T lymphocytes. 466 IMMUNE RESPONSES TO LCM VIRUS INFECTION The generation of the effector T cell response is heralded by marked splenic lymphoproliferation, as measured by the incorporation of radioactively label- led DNA precursors in addition to direct cell counts. A sharp decline in both effector and lymphoproli- ferative responses is seen at about the time infectious virus in the spleen becomes undetectable (15). THE PROTECTIVE ROLE OF EFFECTOR T LYMPHOCYTES The temporal relationship between the kinetics of the effector cell response and the disappearance of infectious virus from the spleen suggested that re- covery from infection was mediated by effector T lymphocytes. To test this supposition, two experi- mental approaches were used. First, non-immune BALC/c mice were given intra- venous adoptive transfers of 108 viable spleen cells taken from syngeneic immunized donors, either at the peak of their effector T cell response (day 8) or after its disappearance (day 30); 24 hours later, they were challenged with a normally lethal, intracerebral dose of LCM virus. As controls, groups of animals that had not been adoptively immunized, or that had received an intraperitoneal dose of virus 24 hours previously, were similarly challenged. Only those animals that received spleen cells con- taining effector activity were protected against intra- cerebral challenge, although some exhibited clinical evidence of a transient, mild CNS disease. Treatment of these donated cells with anti-O serum and comple- ment eliminated their protective effect. The results of similar experiments published elsewhere (15) showed that histological choriomeningitis was attenuated in protected animals, and the maximal virus levels reached in their brains were approximately 10% of those that normally occur in control animals devel- oping fatal LCM. The second approach was to employ essentially the same experimental protocol, except that the reci- pients of the adoptive cell transfer were 8-week-old (C3H x C5,B1)F1 mice that had been thymectomized at 4 weeks of age, lethally irradiated, and then given a syngeneic bone marrow (ATxBM) graft. This procedure results in a severe depletion of T lympho- cytes and renders mice incapable of mounting a cellular immune response so that they invariably become chronic carriers after LCM virus inocula- tion (15). When ATxBM mice were intracerebrally chal- lenged 24 hours after adoptive immunization with effector T lymphocytes, infection was aborted (15). These animals developed no overt disease and when killed 2-3 weeks after virus challenge, no viral antigen was detectable in their neural membranes by immunofluorescent staining. However, histological- ly, all presented with a mild choriomeningitis, indicat- ing that the transferred effector cells were able to mediate some degree of pathology, apparently with- out the cooperative benefit of recipient T lympho- cytes. In contrast, the transfer of non-immune spleen cells to ATxBM mice restored their capability to respond immunologically to virus challenge and they all developed clinical LCM between 8 and 10 days after infection (15). Thus, it seems clear that protection by effector T lymphocytes against the development of either a normally fatal CNS disease or a chronic carrier state is accompanied by a limited choriomeningitis, which results from immune-mediated elimination of virus and/or infected cells after infection is established. MEDIATION OF ACUTE LCM BY EFFECTOR T LYMPHOCYTES Previous studies (14) have established that T lym- phocytes are required for the production of acute LCM by adoptive immunization of CY-induced carrier mice. To determine whether CNS disease is caused specifically by T lymphocytes with effector activity, groups of CY-induced carrier mice were given, intravenously, 108 syngeneic splenic lympho- cytes obtained from donors either 8 or 30 days after immunization. Although 90% of all recipients devel- oped fatal LCM, the group receiving lymphocytes with effector activity died 4 days after adoptive immunization, whereas carriers given immune lym- phocytes devoid of effectors died after 8 days. The increased survival time of the latter group indicated that a certain period of time was required for a population of effector T lymphocytes to be generated in vivo from a donated precursor cell pool, pre- sumably by proliferation mediated by (viral) antigen and/or differentiation. By comparison, the relatively rapid production of fatal LCM by donated lymphoid cells containing an already established effector pop- ulation is strong evidence that this population actu- ally mediates immunopathological CNS disease. DISCUSSION On the basis of the collective observations de- scribed above and elsewhere (8, 15), it can be concluded that the functional activity ofLCM virus- 5 467' G. A. COLE & E. D. JOHNSON induced effector T lymphocytes as expressed in vitro, namely, their ability to recognize and destroy in- fected target cells, is the same in the intact animal. Furthermore, the net result of acute infection with LCM virus is undoubtedly determined by the kinetic relationship between two interdependent variables. The first is the extent of infection in relevant target tissues, while the second is the efficiency with which the infectious process can lead to the generation of a cellular immune response. The various outcomes of murine infection can be explained in terms of these variables. Classical LCM is the result of widespread im- mune-mediated damage to the meninges, choroid plexus, and ependyma. These membranes are heavily infected when the peak of the cellular immune reactivity is reached (12) and they present an exten- sive target for effector T lymphocytes. The fatal convulsive CNS disease that ensues is probably secondary to the severe choriomeningitis. During primary abortive infections, the balance between infective and immune induction is shifted in favour of the host, either because of a decreased susceptibility of extraneural tissues to virus or the greater ability of these tissues to sustain a consider- able degree of immune-mediated injury without fatal consequences. There is no evidence that the initial site of primary infection has any significant effect on the kinetics of immune induction. The resistance of primed mice to reinfection can be explained by the presence of an established pool of lymphoid precursor cells from which effector T lym- phocytes can be rapidly generated (15). A similar degree of resistance can be achieved in nonimmune mice by providing them with exogenously generated effector lymphoid cells. The establishment of the LCM virus carrier state in either the neonate or the immunosuppressed adult occurs under similar conditions, namely, the un- impeded progression of infection to all tissues in the absence of a wholly functional immune system. With the eventual acquisition (or re-acquisition) of im- mune competence, the high levels of free and cell- associated viral antigen(s) already present probably serve as a paralytic stimulus that prevents the devel- opment of virus-specific effector T lymphocyte pre- cursors but not, however, the development of com- plement-fixing antibody. The requirements for induction of the humoral response of carrier mice have yet to be defined. Recent studies by Lehmann-Grube and his associates (unpublished) indicate that its specificity is for a subviral component found within infected cells but not expressed on their surfaces. The implication of this finding is that virus-specific antigens involved in immune complex disease are different from those recognized by effector T lymphocytes. The provoca- tive and conceptually important findings of Zinker- nagel & Doherty (16) favour this view since they show that the specificity of LCM virus-induced effector T lymphocytes is directed toward surface H- 2 antigens modified by virus as a result of infection and not simply toward viral determinants alone. The influx of monocytes and macrophages into sites of developing LCM virus-induced immuno- pathology has been well-documented by sequential light microscopy (13, 17) and by electron microscopy (M. del Cerro & A. A. Monjan, unpublished observa- tions). Although macrophages are obviously respon- sible for eliminating virus-infected cells destroyed in situ during this process, it is not clear if, or how, they inactivate infectious, cell-associated virus, since this interaction has been little studied. One possibility, for which there is no direct evidence, is that the virus- induced modifications in the surface of permissive cells which render them susceptible to lysis by effector T lymphocytes occur prior to the develop- ment of infectious viral progeny. If such were the case, then no special virucidal property would be required of the macrophage. Alternatively, macro- phages may acquire this property as a result of becoming activated at foci of inflammation (18). ACKNOWLEDGEMENTS This work was supported by USPHS grants NS 11286, RR 05445, RCDA NS 46242 (G.A.C.) and a Brown-Hazen Grant from the Research Corporation. 468 IMMUNE RESPONSES TO LCM VIRUS INFECTION 469 RFSUMIt REPONSES IMMUNITAIRES IN VIVO ET IN VITRO A L INFECTION PAR LE VIRUS DE LA CML: MiCANISMES DE LA MALADIE A MtDIATION IMMUNITAIRE I1 est possible d'etablir un rapport entre la fonction lytique (fonction d'effecteur) observee in vitro dans une population de lymphocytes T produite dans la rate de souris atteintes d'infection aigue par le virus de la chorio- meningite lymphocytaire et la m8me fonction in vivo qui aboutit soit i la guerison, soit a la mort. L'issue est determinee par l'6quilibre entre deux facteurs interdepen- dants, a savoir l'etendue des modifications provoqu6es par le virus dans les surfaces cellulaires de tissus 4 cibles )) et l'efficacite du processus immunitaire aboutissant A la production de lymphocytes T effecteurs qui reconnaissent et detruisent les cellules alterees. Seuls les lymphocytes T qui lysent specifiquement les cultures de fibroblastes infectees par le virus peuvent etre utilises 1) soit pour conf6rer aux souris adultes synge- niques non-immunes ou d6munies de lymphocytes T une protection contre, respectivement, la chorio-meningite mortelle ou 1'6tat chronique de porteur de virus, qui fait inevitablement suite a l'6preuve virale intracerebrale chez les animaux temoins non trait6s, 2) soit pour provoquer une forme rapidement accl6r16e de chorio-m6ningite lymphocytaire chez des souris devenues porteurs par l'effet de la cyclophosphamide. La production de lymphocytes T effecteurs parait etre la voie finale commune tant pour le d6veloppement de la maladie a induction immunitaire que pour 1'6limina- tion de l'infection. REFERENCES 1. TRAUB, E. An epidemic in a mouse colony due to the virus of acute lymphocytic choriomeningitis. J. exp. Med., 63: 533-546 (1936). 2. TRAUB, E. Persistence of lymphocytic choriomen- ingitis virus in immune animals and its relation to immunity. J. exp. Med., 63: 847-861 (1936). 3. TRAUB, E. The epidemiology of lymphocytic chorio- meningitis in white mice. J. exp. Med., 64: 183-200 (1936). 4. ROWE, W. P. Studies on pathogenesis and immunity in lymphocytic choriomeningitis infection in the mouse. Bethesda, Naval Medical Research Institute, 1954 (Research report NM 005 048.14.01). 5. HOTCHIN, J. Virus, cell surface and self: Lymphocytic choriomeningitis of mice. Amer. J. clin. Path., 56: 333-349 (1971). 6. ROWE, W. P. ET AL. Protective effect of neonatal thymectomy on mouse LCM infection. Proc. Soc. exp. Biol. (N.Y.), 114: 248-251 (1963). 7. COLE, G. A. & Nathanson, N. Lymphocytic chorio- meningitis. Pathogenesis. Prog. Med. Virol., 18: 95- 110 (1974). 8. HOTCHIN, J. The biology of lymphocytic chorio- meningitis infection. Virus-induced immune disease. Cold Spr. Harb. Symp. quant. Biol., 27: 479499 (1962). 9. OLDSTONE, M. B. A. and DIXON, F. J. Pathogenesis of chronic disease associated with persistent lym- phocytic choriomeningitis viral infection. I. Rela- tionship of antibody production to disease in neo- natally infected mice. J. exp. Med., 129: 483-505 (1969). 10. OLDSTONE, M. B. A. & DIXON, F. J. Immune complex disease in chronic viral infections. J. exp. Med., 134: 32s-40s (1971). 11. OLDSTONE, M. B. A. ET AL. Histocompatability- linked genetic control of disease susceptibility. Mu- rine lymphocytic choriomeningitis infection. J. exp. Med., 137: 1201-1212 (1973). 12. COLE, G. A. ET AL. Lymphocytic choriomeningitis virus. Pathogenesis of acute central nervous system disease. Fed. Proc., 30: 1831-1841 (1971). 13. GILDEN, D. H. ET AL. Immunopathogenesis of acute central nervous system disease produced by lympho- cytic choriomeningitis virus. II. Adoptive immuniza- tion of virus carriers. J. exp. Med., 135: 874-889 (1972). 14. COLE, G. A. ET AL. Requirement for e-bearing cells in lymphocytic choriomeningitis virus-induced central nervous system disease. Nature (Lond.), 238: 335- 337 (1972). 15. JOHNSON, E. D. & COLE, G. A. Functional hetero- geneity of lymphocytic choriomeningitis virus-spe- cific T lymphocytes. I. Identification of effector and memory subsets. J. exp. Med., 141: 866-881 (1975). 16. ZINKERNAGEL, R. M. & DOHERTY, P. C. Immuno- logical surveillance against altered self components by sensitized T lymphocytes in lymphocytic chorio- meningitis. Nature (Lond.), 251: 547-548 (1974). 17. MONJAN, A. A. ET AL. Pathogenesis of cerebellar hypoplasia produced by lymphocytic choriomeningi- tis virus infection of neonatal rats. I. Evolution of disease following infection at 4 days of age. J. Neuropath., 32: 110-124 (1973). 18. BLANDEN, R. V. & MIms, C. A. Macrophage activa- tion in mice infected with ectromelia or lymphocytic choriomeningitis viruses. Aust. J. exp. Biol. med. Sci., 51: 393-398 (1973). G. A. COLE & E. D. JOHNSON DISCUSSION OLDSTONE: In the experimental allergic encephalomyelitis model, if you transfer sensitized lymph node cells or effector T cells direct into the target organ (the brain), the time necessary for immunopathological lesions to appear is the same as if you transferred the cells intra- venously. Do you know whether in the LCM adoptive transfer system it makes any difference if the effector cells are introduced direct into the CNS instead of being given intravenously? COLE: We have talked about doing this experiment for years, but have never done so. OLDSTONE: Is there any evidence that when you give an effector T cell, it goes direct to the lesion, or do you think that the T cell produces something that recruits other cells to the lesion? COLE: This is a difficult question to answer because the only semidirect evidence comes from the work of Doherty who has done cisternal taps on mice inoculated with appropriate effector cells. He suggests that, using the right parent-F1 combination, effector cells go direct to the target. OLDSTONE: Have you tried radioactive labelling of the cells that you transfer? COLE: We are doing that at present. Mims: In these experiments you have transferred sen- sitized spleen cells to your cyclophosphamide-induced carriers and produced neurological disease. With LCM, however, there is a difficulty that nobody has yet thought out. Many years ago, Dr Volkert injected sensitized cells into congenital carriers in an infected colony of mice and produced, at most, only the faintest histological evidence of lesions and no neurological disease. As far as I know, nobody has interpreted this in terms of the targets, the surfaces of infected cells offered in those animals, or other immunological inhibitors. Have you any comments on that? COLE: I really should agree and acknowledge that we are following in the footsteps of Dr Volkert who did similar studies 10 years ago or more but who, as you mentioned, was using a congenital carrier mouse. Carrier mice are quite different because the virus load in the cyclophosphamide-induced carrier mouse is largely con- fined to the CNS. In fact, if the cyclophosphamide- induced carriers are kept for 50 or 60 days, one has great difficulty in killing them with CNS disease. This seems to be related to the fact that there is a gradual centripetal extension of virus to all tissues; the mice then behave biologically very much like congenital car- riers. So I think that we are dealing with an antigen gradient and we should not expect sensitized cells to " swim upstream " towards a target that has no attraction for them. 470