key: cord-021069-v9f9874x authors: Morrison, Lynda A.; Fields, Bernard N. title: Viral pathogenesis and central nervous system infection date: 2004-11-23 journal: nan DOI: 10.1016/1044-5765(91)90002-6 sha: doc_id: 21069 cord_uid: v9f9874x Both host defense and viral genetic factors influence the development of viral infection and disease. Due to the presence of the blood-brain barrier, infection of the central nervous system creates additional complexities in interactions between a virus and its host. Stages in viral pathogenesis defined as (1) virus entry, (2) spread, (3) tropism, (4) virulence and injury to the host, and (5) the outcome of infection are discussed for viral infections in general and those aspects unique to infections of the central nervous system. Information about neuronal physiology and function has also been revealed through studying virus infection. An increased understanding of viral pathogenetic mechanisms and host response to infection raises interesting possibilities for vaccine development and for basic studies in neurology and neurobiology. Seminars in THE NEUROSCIENCES, Vol 3, 1991 : pp 8 3 -91 Viral pathogenesis and central nervous system infection Lynda A . Morrison and Bernard N. Fields Both host defense and viral genetic factors influence the development of viral infection and disease. Due to the presence of the blood-brain barrier, infection of the central nervous system creates additional complexities in interactions between a virus and its host. Stages in viral pathogenesis defined as (1) virus entry, (2) spread, (3) tropism, (4) virulence and injury to the host, and (5) the outcome of infection are discussed for viral infections in general and those aspects unique to infections of the central nervous system . Information about neuronal physiology and function has also been revealed through studying virus infection . An increased understanding of viral pathogenetic mechanisms and host response to infection raises interesting possibilities for vaccine development and for basic studies in neurology and neurobiology . Key words : virus / infection / pathogenesis / central nervous system VIRAL PATHOGENESIS has been defined as the process by which a virus causes disease in a susceptible host .' Several stages in pathogenesis can be identified that represent unifying themes in this process . All viruses must successfully penetrate the host's barriers to entry, undergo primary replication and then spread to their ultimate target tissue . Any virus able to cause disseminated infection has the potential to infect the CNS . Viruses infecting the CNS can either circumvent the blood-brain barrier by entry into peripheral nerve endings and axonal transport into the CNS, or they can invade from the blood stream by infection of or transport across the vascular endothelium in the CNS, in regions of greater permeability, or by passive transport in infected monocytes . But with all potentially neurotropic viruses successful invasion of the CNS is rare, limited by host defenses, the mechanics of the blood-brain From the Department of Microbiology and Molecular Genetics, and the Shipley Institute of Medicine, Harvard Medical School, 25 activity and are thought to gain entry through small immune host defenses . breaks in the integrity of the mucosa, or possibly by uptake at the intact mucosal surface . Infection of the conjunctiva, presumably by direct inoculation from contaminated fingers, can occur with enterovirus 70, HSV and some coxsackie viruses . 3 Regardless of the route of entry, viruses must evade local phagocytic cells that make an important contribution to early host defense . Alveolar macrophages and histiocytes of the reticuloendothelial system efficiently phagocytose and eliminate invading viruses, but some (lymphocytic choriomeningitis virus, cytomegalovirus and the lentiviruses visna and HIV) escape destruction by productively infecting these cells . 1,2,5 Elements of the host immune system, immunoglobulin A in mucous secretions and intraepithelial lymphocytes in the intestine, must also Spread Tissue invasion A period of replication in epithelial or subepithelial tissue at the site of inoculation typically occurs, except when virus is directly inoculated into the blood . Whether a virus infection will remain localized at the site of entry or become disseminated is influenced by the temperature of the epithelial surface (e .g . rhinovirus), the propensity of a virus to bud from the apical (influenza) or basolateral (HSV) surface of epithelial cells, 3 Following replication at the site of inoculation, virus enters the subepithelial lymphatic capillaries and is carried into the blood stream 3 where it must survive encounters with phagocytic cells and elements of the immune system . Primary viremia permits the rapid dissemination either of free virus or of bloodcell associated virus to other susceptible tissues :' successful penetration of the vascular endothelium and replication in these organs amplifies the infection and augments viremia . Viremia must be of a magnitude and duration adequate to sustain infection and, for some viruses, to permit an assault on the blood-brain barrier . Replication of some arboviruses in cerebral capillary endothelium precedes viral invasion of the neural parenchyma, 6 and replication in or passive transport across the endothelium has been postulated for several other viruses, including enteroviruses, arboviruses and retroviruses .6' 7 Because of its fenestrated endothelium and sparse basement membrane, the choroid plexus is also a target for replication or passive transport of mumps and arboviruses ; 1,3 through it, virus enters the cerebral spinal fluid to infect ependymal cells lining the ventricles and subseqently the underlying brain tissue . 3,5 Migrating infected monocytes, lymphocytes or leukocytes may be an important vehicle for virus transport into the CNS, especially in lentivirus infection (see Zink and Narayan, this issue,$ and refs 9,10) . Neural spread from the periphery to the CNS can be considered as two processes, penetration and propagation . Penetration probably occurs at nerve terminals, either at the site of initial entry into and replication in the host or at a secondary site of replication in extraneural tissues to which virus is disseminated by the blood . Electron microscopic observations have documented accumulation of neuroinvasive rabies virus particles in infected myocytes at neuromuscular junctions and neuromuscular 85 spindles, followed by entry into motor or sensory nerve terminals . 3 Receptors used by viruses may be concentrated at nerve terminals and synapses, facilitating uptake and possibly transport of virus (see below, Tropism) . Once within nerves, propagation is mediated by axonal transport, replication and transsynaptic transfer of virus . HSV, rabies, poliovirus and reovirus are known to spread along nerves by the system of fast axonal transport ." Transport has been demonstrated both towards and away from the cell body in somatic motor, somatic sensory and autonomic nerves, under various experimental conditions . Bidirectional transport is important in the establishment of, and reactivation from, latency seen with HSV and varicella zoster . 12,13 Viruses may have a predilection for a subset of nerve types, as evidenced by the more rapid uptake of reoviruses in motor than sensory fibers innervating the footpad 14 and by the restriction of rabies mutants to transport by sensory but not autonomic paths . 15 In contrast to the fast axonal transport of most viruses, the scrapie agent has been postulated to use the system of slow axonal transport . 3,6,16 Disruption of the neurofilament cytoskeleton that supports slow axonal transport has been observed in scrapie-infected neurons and has been implicated in generation of the spongiform changes that characterize this degenerative neurological disease (see Hope, this issue, 17 and refs 16, 18) . Viral replication occurs once virus particles reach the neuronal cell body. Newly synthesized virion components and virus particles then accumulate at post-synaptic sites, probably reflecting directed transport of virus components analogous to directed budding in polarized epithelial cells (see Entry into the host, above ; ref 11) . Both rabies and vesicular stomatitis virus accumulate at synaptic terminals and beneath the nodes of Ranvier . When antiviral antibody is added to infected neuronal cultures, vesicular stomatitis virus buds preferentially from the synapses, indicating that exocytosis by neurons may be facilitated at this point .3 Neurons thus transfer virus at synaptic junctions in the periphery . In the CNS, the high density of neurons and glia permits many viruses to spread from cell to cell with less fidelity . The site of virus inoculation can influence the ultimate distribution of lesions within the CNS, especially for viruses that spread along neural routes . Neurons can transport viruses great distances intraaxonally both to and within the CNS, and the final destinations of a virus will be determined in part by the particular synaptic connections made with individual neurons whose processes took up virus in the periphery . 19 HSV gaining entry through the genitourinary tract infects and becomes latent in sacral ganglia, whereas HSV entering in the oral mucosa becomes latent in trigeminal sensory neurons . Though cases of strictly blood-borne or neural spread of viruses to the nervous system have been documented, a combination of the two routes may be used by certain viruses, or under certain conditions . For example, flaviviruses can spread through the blood to exposed neurons of the olfactory bulb which acetylcholine receptor, localized on muscle cells in the region of neuromuscular junctions, has been identified as a receptor for rabies virus . Phosphatidyl serine may function as an alternative receptor for rhabdoviruses, 5,26 indicating that lipid molecules (and carbohydrate residues) as well as proteins can serve as neurotropic virus receptors . Other molecules acting as receptors for neurotropic viruses are listed in Table 2 . Specific interactions with receptors may regulate binding of a particular virus at epithelial surfaces and vascular endothelium as well as nerve terminals, and may explain to some degree tissue specificity and differential susceptibility to infection . Numerous viral cell-attachment proteins have also been identified ( gp350, gp220 heparin sulfate proteoglycana gB, gC, basic fibroblast growth gD/gH1 factor receptorb 87 selected with antibody against the E2 envelope protein revealed that a single amino-acid substitution in the E2 molecule determines loss of tropism for neurons without change in tropism for oligodendrocytes . 6,26 Little is known about the murine coronavirus receptor protein on neurons or its similarity to the receptor on oligodendrocytes . X-ray crystallographic analysis of the poliovirus virion has provided a detailed molecular view of a ligand in virus-receptor molecular interactions . The analysis has revealed a series of peaks in the VP1 protein surrounded by a broad valley composed of VPI and VP3 that forms the receptor binding pocket (see Almond, this issue 29 ; ref 13) . Other members of the picornavirus family are known or postulated to have cell-attachment proteins of similar topology . Transcriptional regulatory elements are an important post-receptor-binding determinant of viral tropism . 5,30 An enhancer element confers tissuespecific expression for JC papovavirus in human oligodendrocytes . Similarly, promoter elements of some neurotropic murine retroviruses determine their species-specificity . On the other hand, host cells that lack enzymatic activities required for maturational cleavage of viral proteins or completion of the virus replicative cycle (as observed for poliovirus) will not be productively infected, thereby limiting tropism .5,25 A number of host factors modulate viral tropism and virulence, influencing viral pathogenetic potential and ultimately the potential for CNS infection . 3,31 Many neurotropic viruses more readily infect the young. This has been shown to variously result from the immaturity of the immune response, reduced capacity to produce interferon, increased susceptibility to infection of some cell types, and age-specific nature and distribution of receptor proteins2 '3 (see Coyle, this issue 32) . Genetic determinants of disease susceptibility have been found for infection of mice with strains of most neurotropic viruses, in at least one case of coronavirus reflecting lack of a gene encoding a virus receptor protein . Determinants have been linked to histocompatibility genes that regulate immune responses only in chronic diseases of immunopathological etiology . Last, sex of the host and nutritional status can influence resistance to infection . The selective vulnerability of different species or one species at different ages, coupled with viral determinants of neurotropism, may explain differences in clinical manifestations of CNS disease . The efficiency with which a virus can multiply and extend infection after it has invaded the nervous system, its neurovirulence, determines in part the extent of injury suffered by the host . The effectiveness of the host response and the resulting immunopathology also play a role . Although each virus has evolved its own genetic determinants of virulence, some generalizations can be drawn . Outer capsid and envelope glycoproteins are frequently implicated as virulence factors because changes in their amino-acid sequence usually weaken their infectivity (attenuation), although this may be a subtle reflection of altered tropism . Genetic reassortants between virulent and avirulent strains of bunyaviruses and reoviruses have also been used to identify determinants of neurovirulence by segregation of phenotype with a particular viral gene . 3 Genetic recombination between strains of poliovirus and of HSV have similarly been used to identify genes critical for neurovirulence . 5 More than one region of a viral protein identified as a virulence factor may be important3 and, when neurotropism and neurovirulence are determined by the same protein, changes in that protein can have multiple and dramatic effects on the resulting infection . In studies of Theiler's murine encephalomyelitis virus, a model of human demyelinating disease (see Nash, this issue 33), variants selected with L .A . Morrison and B .N. Fields antibody to the VP1 protein possess tissue-specific alterations in extent and duration of infection and in capacity to induce demyelination . 13 But not all changes in viral genes are attenuating : mutations causing reversion to neurovirulence have been identified in poliovirus isolates from the feces of recipients of the attenuated trivalent oral vaccine . 34 Mutations in genes encoding proteins other than those of the capsid or envelope can also result in attenuation, for example, sequences encoding a retroviral core protein or in the 5' noncoding region of poliovirus RNA . 34 Transacting transcriptional activators may play a role in determining virulence, exemplified by a mutant of HIV with a defective tatIII gene that shows reduced replication and cytopathic effect .3,6 Importantly, neurovirulence may be under multigenic control and determinants of virulence may differ depending on the host species infected . 3 CNS injury resulting from viral infections Viral infections of the nervous system can have very severe consequences because the neurons vital to host function and survival cannot be replaced once injured or destroyed ; the effect of even minor damage may be catastrophic .2 Furthermore, once infection in the CNS has begun, the reduced permeability and dense structure of the blood-brain barrier which often impedes the initial spread of virus to the CNS may reduce the capacity of the host to resolve infection and limit injury. 9 Injury at a cellular level may be directly due to virus infection, occurring when cytopathic effect is significant enough to damage irreparably the plasma membrane (lysis) or to arrest metabolic activity by inhibition of protein, RNA or DNA synthesis, e .g . poliovirus, vesicular stomatitis virus and herpesvirus, respectively . 35,36 Viruses such as rabies may alter nerve impulse conductance without histologically perturbing infected neurons . 19 Accumulations of virus capsids or even individual viral proteins may in themselves be toxic . The sequence homology between HIV gp120 envelope protein and certain essential neurotransmitters and neuropeptides has evoked the hypothesis that competitive binding of gp120 may block normal neuronal function (see Tillman and Wigdahl, this issue 37). gp120 also mediates another potential form of direct injury by virus, syncytium formation . Indirect injury to cells has been postulated to occur by release of toxic`factors' from neighboring cells damaged by infection . 36 Systemic injury occurs by both direct and indirect mechanisms as well . The consequences of infection vary with the location and function of tissue injured by the virus : for example, motor neuron destruction in poliomyelitis results in paralysis ; demyelinating reactions to virus infection cause incoordination ; and virus infections of cells in the developing nervous system produce a variety of congenital abnormalities and neurological diseases (see Coyle, this issue, 32 and refs 19, 38) . Indirect mechanisms of viral injury to the host CNS include the actions of interferons, 39-41 and the virus-specific cellular and humoral immune response (see Sedgwick and Dörries, this issue, 42 and refs 2, 40, 43) . Paradoxically, the elaborate system of host response to virus infection that may be protective outside the CNS can be destructive when generated within this normally isolated compartment, 1,43 where cell lysis and complement activation may injure the host while helping to clear virus . Virus replication in macrophages or immunocompetent T or B cells can also induce immune disregulation . 41,43 Last, autoimmune reactions may be provoked by viral infections, resulting in indirect injury to the host even without continued presence of virus. 41,44 Autoimmunity can be triggered in genetically susceptible individuals by non-specific inflammatory stimuli that alter immune regulation or by cross-reactivity between the viral inducing agent and normal host proteins (molecular mimicry) . 44,45 Exposure of normally sequestered antigen due to virus-induced cytolysis and destruction or dysfunction of suppressor T cells have also been postulated as mechanisms in the development of autoimmune reactions . 41 Autoimmune demyelination reactions, initiated in the CNS against nervous system antigens, are especially prevalent . 46 Several outcomes are possible in viral infections of the CNS that do not acutely result in death . Virus may be effectively cleared from the CNS, may remain latent in the nervous system and cause recurrent disease, or may produce persistent disease . Clearance is thought to be highly dependent on the immune response, though limitation of viral replication and the type of disease produced can be influenced by nonspecific factors . Failure to clear varicella zoster, cytomegalovirus, adenovirus and 89 measles infections in natural cell-mediated immunodeficiency states suggest that T cell responses may be more important than antibody in eventual clearance of many infections . ) Successful clearance may still leave the host impaired, depending on the nature and severity of the injury sustained .9" 10 A chronic autoimmune reaction may continue even after virus clearance . When latency is established by some viruses, the viral genome remains within the nervous system after acute infection ceases . No infectious virus is produced until reactivation occurs and the absence of viral protein products transcribed during latency may permit the infected cell to escape immune surveillance . Capacity for latent infection in the nervous system is characteristic of the herpesviruses, HSV and varicella zoster (see Stevens, this issue, 47 and refs 12,13) . Persistent viral infections, in contrast to latency, are characterized by continued production and shedding of virions from infected cells . They are especially prevalent in CNS tissues, 9 possibly because these contain non-dividing cells, are isolated behind the blood-brain barrier, have low expression of major histocompatibility complex proteins and may permit increased generation of defective viruses . 19 Immune responses are usually generated and maintained but are ineffectual in clearing virus .48 As a result, chronic immune reactions may contribute much of the tissue injury .43 Possible mechanisms of establishment and maintenance of persistent infection are listed in Table 3 . Persistent infection takes many forms . In lymphocytic choriomeningitis virus infection of newborn mice, chronic production of virus occurs without serious detriment to cells . 43 A more thorough understanding of these stages in pathogenesis of viral CNS infections may engender new or more effective measures for prevention of disease . Development of more efficacious vaccines is a primary objective but requirements for protection from individual viruses differ . 40,50 Knowledge of pathogenetic stages of each virus may reveal points of greatest vulnerability and will help to determine requirements for stimulating immunity with a minimum of immunopathological side effects . Ideally, infection by any virus with potential to invade the CNS should be stopped before entering the nervous system . Characterization of neurotropic virus infections also offers interesting possibilities for studying structure and function of the nervous system . Basic neurological and neuroscience research has begun to benefit from using the strongly neurotropic rabies, HSV and pseudorabies viruses for tracing neural pathways . 51 More speculative is the use of attenuated neurotropic viruses for delivering foreign proteins into regions of the CNS defined through studies of viral pathogenesis . Foreign genes can be packaged by many viruses and may be selectively expressed, permitting intracerebral production of pharmacological agents or neurotransmitters . Thus studies of viral pathogenesis of the CNS may offer tools for the future as well as an understanding of disease states . Viral Infections of the Nervous System Viral Pathogenesis and Immunology Pathogenesis of viral infections Host defense mechanisms at mucosal surfaces Pathogenesis of neurotropic viral infections Molecular and genetic aspects of the pathogenesis of viral infections of the central nervous system Vascular endothelium in immunology and infectious disease The neuropathogenesis of visna virus infection in sheep Narayan 0 (1988) The neurobiology of human immunodeficiency virus infections The human immunodeficiency virus : infectivity and mechanisms of pathogenesis Implications of axoplasmic transport for the spread of virus infections in the nervous system The use of molecular techniques in studying viral pathogenesis in the nervous system Penetration of the nervous system of suckling mice by mammalian reoviruses Pathways of the early propagation of virulent and avirulent rabies strains from the eye to the brain Slow infections, in Infections of the Nervous System The key role of the nerve membrane protein AP in scrapie-like diseases Fields Viral pathogenesis and CNS infection Hypothesis : interference with axonal transport of neurofilament as a common pathogenetic mechanism in certain diseases of the central nervous system Persistence of RNA viruses in the central nervous system Enteroviruses, in Handbook of Poliovirus type 1 enters the human host through intestinal M cells Intestinal M cells : a pathway for entry of reovirus into the host Isolation of cellular receptors for viruses Viral receptors : expression, regulation and relationship to infectivity Host and viral factors that influence viral neurotropism . I . Viral cell attachment proteins and target cell receptors Fibroblast growth factor receptor is a portal of cellular entry for herpes simplex virus Pathogenesis of reovirus infections of the central nervous system Enhancers and eukaryotic gene transcription Host and viral factors that influence viral neurotropism . II . Viral genes, host genes, site of entry and route of spread of virus Viral infections of the developing nervous system Virological and pathological processes involved in Theiler's virus infection of the central nervous system The attenuation of poliovirus neurovirulence Impact of virus infection on host cell protein synthesis Virus-host-cell interactions Neuropathogenesis of human immunodeficiency virus infection The viral infections of the developing nervous system : an overview, in Virus Infections and the Developing Nervous System The pharmacology and toxicology of the interferons : an overview Immunological aspects of the prevention of viral diseases Interactions of viruses with the immune system The immune system response to viral infection of the CNS Viruses perturb lymphocyte functions : selected principles characterizing virusinduced immunosuppression Autoimmune diseases : the failure of self tolerance Molecular mimicry and autoimmune disease Virus-induced autoimmune demyelinating disease of the central nervous system Herpes simplex virus : neuroinvasiveness, neurovirulence and latency Viral persistence Pathogenetic aspects of persistent measles virus infections in brain tissue Modern Approaches to Vaccines Viruses as transneuronal tracers The authors are grateful to Dr K Tyler for much helpful advice and discussion, and for reviewing this manuscript . Work on the manuscript was supported by Public Health Service grants A108207-02 and program project 5 P50 NS16998 from the National Institutes of Health .