key: cord-0039800-3lunx9tt authors: Sedgwick, Jonathon D.; Dörries, Rüdiger title: The immune system response to viral infection of the CNS date: 2004-11-23 journal: nan DOI: 10.1016/1044-5765(91)90003-7 sha: 6d52b0772b48b0a804a6c48619c4869f6184f65e doc_id: 39800 cord_uid: 3lunx9tt In immunological terms the CNS is at a severe disadvantage in its ability to respond to infection by virus. First, both glial and neuronal cells normally do not express molecules of the major histocompatibility complex. Second, the most efficient cells for stimulating an immune response (leucocyte dendritic cells) are not present in the healthy CNS; and third, there is no specialized lymphatic drainage from the CNS to lymph nodes to enable the immune system to be quickly informed of the presence of an infection. Nevertheless, the immune system apparently copes with the vast majority of viral infections in the CNS. This is clearly evidenced by the reactivation of latent CNS viral infections in some immunosuppressed patients and the dramatic increase in the seventy of CNS disease in young or otherwise immunologically incompetent experimental animals infected with neurotropic viruses. The routes by which the CNS and the immune system may communicate and the varied ways in which an immune response may affect the outcome of a viral infection of the CNS are discussed. seminars in THE NEUROSCIENCES, Vo13, 1991 : pp 93-100 The immune system response to viral infection of the CNS jonathon D. Sedgwick In immunological terms the CNS is at a severe disadvantage in its ability to respond to infection by virus. First, both glial and neuronal cells normally do not express molecules of the major histocompatibility complex. Second, the most efficient cells for stimulating an immune response (leucocyte dendritic cells) are not present in the healthy CNS; and third, there is no specialized lymphatic drainagefrom the CNS to lymph nodes to enable the immune system to be quickly informed of the presence of an infection. Nevertheless, the immune system apparently copes with the vast majority of viral infections in the CNS. This is clearly evidenced by the reactivation of latent CNS viral infections in some immunosuppressed patients and the dramatic increase in the severity of CNS disease in young or otherwise immunologically incompetent experimental animals infected with neurotropic viruses. The routes by which the CNS and the immune system may communicate and the varied ways in which an immune response may affect the outcome of a viral infection of the CNS are discussed. Key words: central nervous system / viral encephalomyelitis I antigen presentation I immunity I lymphocytes I viral persistence THE IMMUNE SYSTEM has evolved in higher vertebrates into a complex, multi-factorial surveillance mechanism designed to limit both the spread of infectious organisms as well as the outgrowth of neoplastic cells. There appear to be few, if any, body compartments that are not patrolled by the immune system. The central nervous system (CNS) is, however, compared to most other sites, normally lacking in the essential elements and conditions known to lead to a rapid and affective immunological response. The paucity of immune elements in the CNS and the consequent lack of immune interactions that occur obviously has important consequences for the way in which viral infections of the CNS are handled. Recent experimental evidence indicates that the 'immunological privilege' of the CNS is by no means absolute and that there are some forms of communication between the immune and central nervous systems. Indeed, the immune surveillance of the CNS appears to be sufficient in most cases to control or eliminate most viral infections. Here we briefly describe those components of the immune system which are fundamental for effective defence against viral infection and illustrate why the normal CNS is considered to be so immunologically deficient. The possible mechanisms that enable the immune system to respond to a CNS viral infection are discussed and brief examples given of the way in which the course of infection may be altered by the action of the immune system. The major histocompatibility complex and T and B lymphocytes Studies in the 1970s and subsequently (reviewed in ref 1) established the nature of the specific recognition requirements necessary for the immune system to respond to a given antigen and, in particular, for cytotoxic T lymphocytes to recognize and destroy virus-infected cells ( Figure 1 ). Most T lymphocytes can be subdivided into two non-overlapping populations expressing distinct surface molecules known as CD4 and CD8. The former subset predominantly exhibit helper/inducer function, whereas the latter is the most active in terms of cytotoxic capacity.2 Both subsets also express a T cell-specific receptor which recognizes peptide antigen (for example, derived from a virus present in an infected cell), only when it is bound to and displayed on the cell surface in the context of a major histocompatibility complex ( CD4 + T cell subset also provides help for antigenspecific B lymphocytes resulting, ultimately, in their differentiation into antibody-secreting plasma cells . Unprimed CD4+ (T) lymphocytes are activated following interaction with leucocyte dendritic cells (DC) which are a specialized and highly efficient type of antigen-presenting cell (APC) . DC may provide factors (indicated by the small arrow) which assist in this process . T lymphocytes use a specific T cell receptor (TCR) to recognize peptide forms of (viral) antigen which are presented to the T cell in the context of MHC class II molecules (C1 II) on the DC . DC are not normally present in the CNS and so this step probably first occurs outside the CNS . The activated, primed CD4+ T helper lymphocyte (Th) may now interact with a variety of APC, which could include macrophages, astrocytes, microglia and vascular endothelial cells . B lymphocytes can bind virus and also act as APC for CD4+ Th lymphocytes . The Th cells produce factors (indicated by small arrows) that allow B lymphocytes to differentiate into antibody-secreting plasma cells . CD8+ T lymphocytes (CTL) see virusinfected cells through viral peptides presented in the context of MHC class I molecules (C1 I) . Th lymphocytes also provide factors which enable further differentiation of the CD8 + T cells into active cytotoxic T lymphocytes (CTL) . Nevertheless, after viral infection, leucocytes, predominantly T lymphocytes and blood monocytes, may cross the blood-brain barrier and enter the parenchyma of the CNS. The appearance of these cells is usually accompanied by the upregulation of MHC expression on resident CNS cells and this is illustrated in Figure 3 . Although there are relatively few infdtrating cells (in Figure 3b only infiltrating cells and not the CNS are stained), substantial MHC class I and II induction on the CNS itself results (Figure 3a,d) . The MHC induction is probably due to secretion of cytokines such as interferon y6 by the infiltrating T lymphocytes. CNS antigen presentation, lymphatic drainage and the blood-brain barrier In tissues other than the CNS (for example the skin), a specialized system exists to transport antigen through afferent lymphatics to local lymph nodes where full activation of an appropriate immune response can occur. The efficiency of this system is at least partly related to the presence of dendritic antigen presenting cells in the skin, as in many other organs, that pick up antigen and transport it to the lymph nodes (reviewed in ref 7). The CNS is also deficient in this respect. First, no specialized dendritic antigen-presenting cells (see Figure 1 ) are present in the CNS. 7 Macrophage-like antigen-presenting cells are present 8)g but most of these are not constitutively MHC class II positive. Moreover, there is also no evidence that these cells, or indeed any other resident CNS cells with potential to act as antigen-presenting cells, such as astrocytes,1° can pick up antigen and pass to lymph nodes to allow sensitization of the immune system to occur. Second, the CNS has no specialized lymphatic drainage. The importance of this mode of antigen transfer to lymph Figure 4 for details) . However, none of these routes are likely to be as efficient in this task as are the dedicated afferent lymphatics present in other organs and tissues . Passage of effector leucocytes into the CNS and antigen presentation Once the immune system is sensitized there does not appear to be anything preventing the initiating of an inflammatory response within the CNS, despite its many immunological deficiencies . But exactly how leucocytes circulating in the blood know that they should cross the brain vascular endothelium and enter the CNS is not definitely known . It seems likely that there is a requirement for some sort of nonspecific adhesion to occur between circulating leucocytes and the brain vascular endothelium, as a first step in the process of extravasation of these cells from the blood into the parenchyma of the CNS . A complex array of molecules has now been Alternatively, antigen may traverse the olfactory nerves, leaving the brain by the cribriform plate, where it is collected and drained by afferent lymphatics into the cervical lymph nodes ." There is some evidence that antigen may be able to passage from brain to blood through the vascular endothelium . 15 Antigen recognition by T cells could then occur on the endothelial cells themselves . There is, theoretically, no requirement for transport of whole molecules by the latter pathway as T lymphocytes see only peptide antigens, which may more readily be processed and transported through the endothelial cells for presentation on the luminal surface of the vasculature . The next step is unclear . It is possible that foreign antigen in the CNS is degraded and presented to the adhering T cells as peptides in the context of MHC molecules on the luminal face of the vascular endothelial cells . This could impart a signal to the T cells to initiate the process of extravasation or, alternatively, the T cells may directly damage the endothelial cells 17 allowing the rapid passage of many leucocytes into the CNS . The latter process certainly occurs during lymphocytic choriomeningitis virus infection in mice . 18 It is possible that no antigen-specific recognition at the vascular endothelium is required at all . There is accumulating evidence that the CNS is routinely patrolled by the immune system, although the discussion above illustrates that this function must be performed by relatively few T cells at any one time, given their scarcity in the healthy CNS . Data are emerging that pre-activated but not resting T lymphocytes may preferentially pass from the blood into the CNS . 19 In this way, the immune system through its random surveillance of the CNS may become aware of the presence of foreign antigen . Which cell type(s) in the CNS present antigen to the extravasated T lymphocytes is not known but perivascular macrophages8 or astrocytes 1° are likely candidates as both are intimately associated with the CNS vasculature . One point that keeps emerging in this discussion is that immune reactions may proceed in the CNS but only after the response has been initiated outside the CNS . Once specific T cells have interacted with CNS antigen, a relatively rapid accumulation of other, non-specific inflammatory cells may follow . Subsequently, the immune response can develop to some extent within the CNS as evidenced by high levels of, for example, specific anti-viral immunoglobulin in the cerebrospinal fluid and the restricted (oligoclonal) nature of this antibody . 20 There is also evidence that expansion of populations of T cells with restricted specificity may occur in the CNS . 21 How the immune response may influence the outcome of viral infection Here we outline briefly examples of CNS viral infections to illustrate the variable ways in which an immune response may affect the outcome of the infection . The protective nature of an effective immune response is illustrated by the course of CNS viral infection in immunologically immature or immunosuppressed hosts . Infection of immunologically incompetent neonatal rats or mice with the neurotropic coronavirus JHM causes acute lethal encephalitis . 22, 23 In animals which are nursed by pre-immunized mothers the onset of neurological symptoms is delayed, indicating protection by transfer of maternal antibodies . 24,25 In the human neonate, disseminated infections involving the CNS may occur and are usually accompanied by high mortality or severe consequences for the survivors 26 (see also Coyle this issue 27) . Immunosuppression during experimental CNS infection in mice by corona virus JHM and Theiler's virus is associated with enlarged areas of histopathological change, spread of virus and increased incidence of fatal encephalitis . 23,28 In man, immunosuppressive treatment during organ transplantation may induce reactivation of chronic persistent virus infection resulting in, for example, lytic infection of oligodendrocytes and death, as seen during progressive multifocal leucoencephalopathy caused by polyomavirus JC .26 Moreover, the aquired immunodeficiency syndrome (AIDS) is frequently complicated by opportunistic viral infections of the CNS including JC-, herpes and cytomegalovirus . 29 In contrast, the immunocompetent host controls viral CNS infections by protective humoral and cellular effector mechanisms . Data from experimental corona virus and Theiler's virus infection in mice suggests that antibodies limit viral spread by neutralizing extracellular virus 25,28 whereas action of CD8+ cytotoxic T lymphocytes is required for viral clearance from the brain . 28,30 These events usually parallel the recovery of the infected host from acute neurological disease or even prevent clinical manifestation of CNS infection . The elimination of virus in the CNS by cellular immune-mediated mechanisms may also contribute to the severity of the disease . Direct evidence for the contribution of local virus-specific immunity j D . Sedgwick and R . to CNS disease is seen following intracerebral infection of immunosuppressed mice by lymphocytic choriomeningitis virus . A persistent infection is established in these animals which remain healthy until they receive CD8 + lymphocytes specific for the virus. These cells, in performing their task of viral clearance, ultimately induce so much damage to the CNS vasculature and other local tissue that the animals die . 18 A rare and immunologically less understood complication is postinfectious encephalomylitis which follows recovery from peripheral infection with some exanthema viruses . 26 Strong inflammation in the CNS accompanied by widespread primary demyelination in the absence of viral particles in brain and no detectable virus-specific immune response in the cerebrospinal fluid suggests an autoimmunological destruction of the CNS as a consequence of peripheral infection . Failure to eliminate virus from the CNS Failure of complete virus clearance from the brain by a competent immune system may end in persistent or latent infection of the CNS (see also Stevens, this issue31 ) . In this context, viruses with strong tropism for neurons are of particular interest, because neurons are extremely resistant to induction of MHC class I and so viruses infecting neurons may escape elimination by MHC class I-restricted CD8 + cytotoxic T lymphocytes . Lack of sufficient CD8 + cytotoxic T lymphocyte activity in combination with a strong local antiviral antibody response may help to maintain the persistent state of the infection . Selection pressure put on the virus by these antibodies results in the generation of nonneutralizable virus variants, as seen in visna virus infection of sheep 32 (see also Zink and Narayan, this issue33) . Alternatively, the viral strategy of replication is modified . RNA viruses, like measles, can shut off viral protein synthesis in the presence of high titers of neutralizing antibodies . However, the genetic information of the virus is still present in the target cell and synthesis of viral proteins can start again after the antibody titer has dropped beyond a certain threshold. 34 DNA viruses like polyoma virus JC reveal host cell dependent rearrangements of genomic control regions regulating replication . 35 In conjunction with constant monitoring of infected cell foci by the patrolling immune system, this probably results in subclinical infection over decades . A lethal outcome is, however, still possible during severe immunosuppressive states (AIDS, leukemia, transplantation) . Once immune sensitization outside the CNS has occurred, the immune system may then respond to antigen present in the CNS . Without. prior sensitization, the ability to respond is much diminished . 36 Information briefly presented here provides possible mechanisms by which immune surveillance of the CNS may be achieved but further work is required before the relative importance of thesè communication' routes is defined . The clear examples of reactivation of latent CNS viral infections when individuals are immunosuppressed illustrates that the immune system is generally effective in limiting the potentially damaging effects of uncontrolled viral spread in the CNS . How this is achieved is not understood, given that in such latent viral infections, many or most of the elements described above which are considered essential for the effective functioning of anti-viral immunity are usually absent . A substantial amount of work is required before these and other allied aspects of the immune-pathogenesis of CNS viral infections are clarified . Antigen recognition by class I-restricted T lymphocytes The role of CD4 in T cell-activation : accessory molecule or co-receptor? T cell reactivity to MHC molecules : immunity versus tolerance Interferon--y induces the expression of H-2 and la antigens on brain cells Functional analysis of T lymphocyte subsets in antiviral host defence Macroglia : neural cells responsive to lymphokines and growth factors Control of the immune response at the level of antigen-presenting cells : a comparison of the function of dendritic cells and B lymphocytes Perivascular microglial cells of the CNS are bone marrow-derived and present antigen in vivo Immunobiology of microglial cells Astrocytes present myelin basic protein to encephalitogenic T cell lines Effector mechanisms in allograft rejection The Blood Brain Barrier in Health and Disease Role of cervical lymph nodes in the systemic humoral immune response to human serum albumin microinfused into rat cerebrospinal fluid Ultracytochemical distribution of myelin basic protein after injection into the cerebrospinal fluid . Evidence for transport through the blood-brain barrier and binding to the luminal surface of cerebral veins Adhesion receptors of the immune system Antigen-specific damage to brain vascular endothelial cells mediated by encephalitogenic and non-encephalitogenic CD4' T cell lines in vitro Dissection of an inflammatory process induced by CD8' T cells T cell entry into the rat central nervous system Intrathecal synthesis of virus-specific oligoclonal IgG, IgA and IgM antibodies in a case of varicella zoster meningoencephalitis Limited heterogeneity of rearranged T cell receptor Va transcripts in brains of multiple sclerosis patients Comparative analysis of coronavirus JHM-induced demyelinating encephalomyelitis in Lewis and Brown Norway rats Pathogenesis of demyelination induced by a mouse hepatitis virus (JHM virus) Coronavirus JHM-induced demyelinating encephalomyelitis in rats : influence of immunity on the course of disease Spread of a neurotropic murine coronavirus into the CNS via the trigeminal and olfactory nerves Viral Infections of the Nervous System Viral infections of the developing nervous system Differential pathogenesis between mouse strains resistant and susceptible to Theiler's virus-induced demeyelination Central nervous system dysfunction in acquired immunodeficiency syndrome Effective clearance of mouse hepatitis virus from the central nervous system requires both CD4' and CD8' T cells Herpes simplex virus : neuroinvasiveness, neurovirulence and latency Immune selection of virus variants Narayan 0 (1991) The neuropathogenesis of visna virus infection in sheep Effect of measles antibodies on a measles SSPE virus persistently infected C6 rat glioma cell line DNA rearrangements in organspecific variants of polyomavirus JC strain GS The fate of allogeneic and xenogeneic neuronal tissue transplanted into the third ventricle of rodents Thanks are extended to Drs Heiner Körner and Helmut Wege for providing viral stocks as well as advice and assistance with the infection of rats, and to Hanne Weinand for the preparation and staining of histological material . This work was supported by grants from the Bundesministerium für Forschung und Technologie (Grant No 01 Ki 8839) and The Hertie Foundation .