key: cord-0919071-14do9vp0 authors: Horwitz, Marc S.; Sarvetnick, Nora title: Viruses, host responses, and autoimmunity date: 2006-04-28 journal: Immunol Rev DOI: 10.1111/j.1600-065x.1999.tb01319.x sha: a1e257d220cff66c3430c759ba5322b57df13a57 doc_id: 919071 cord_uid: 14do9vp0 Summary: Conceptually, the initiation of autoimmune disease can be described as a three‐stage process involving both genetic and environmental influences. This process begins with the development of an autoimmune cellular repertoire, followed by activation of these autoreactive cells in response to a localized target and, finally, the immune system's failure to regulate these self reactive constituents. Viruses have long been associated with inciting autoimmune disorders. Two mechanisms have been proposed to explain how a viral infection can overcome immunological tolerance to self components and initiate an organ specific autoreactive process, these mechanisms arc molecular mimicry and bystander activation. Both pathways, as discussed here, could play pivotal roles in the development of autoimmunity without necessarily excluding each other. Transgene technology has allowed us and others to examine more closely the roles of these mechanisms in mice and to dissect the requirements for initiating disease. These results demonstrate that bystander activation is the must likely explanation fur disease development. Additional evidence suggests a further role for viruses in the reactivation and chronicity of autoimmune diseases. In this scenario, a second invasion by a previously infecting virus may restimulate already existing autoreactive lymphocytes and thereby contribute to the diversity of the immune response. Summary: Conceptually, the initiation of autoimmune disease (.an be described as a three stage process involving botb genetic and envininniental infhieiKes. This process begins with the development oi an aiitoiiiiiniuie cellular repertoire, followi'd by aelivatioii ai these diiU>reactive ceils m response to a localized target and, finally, the immune S) stem's laihLrc to regulate these self reactive constittieiits. Viruses have long been associ ated with inciting aLiuiiinmuiic disorders. Two mechanism^ have been proposed to explain how a viral infection can inerLome inimimologLcai tolerance to self eomponciits and initiate an organ specific autoreactive process, these mechanisms arc molecular mimicr) and bystander acti tiem. Both pathways, as discussed here, could play pivotal roles in the development of autoimmiinity withotit necessarily excluding each other. Transgcne tt'chnohigy has allowed us duil olliers to examine more clo.sely tilt' roles lA these mt't.haiijsnis in mice and lo di.s.seti the recjtiirements for iriitiaiing disea.se. Tliese results deiitoiistrjie ihat liystander .ictivatinn is tbe must likely explanation fur disease deveiopineiit. .-Yddititinal evidtnue suggests a further role for viruses in the reactivation aiidehrunicity ui'auiuiniinune diseases. In this scenario, a second invasion by a previoti^ly infeeting virus may restimiilate already existing aiUoreactive lymphoc)tes. and thereby contribute iu (he diversity ot llie iniiiiune resjionse. T!ie organ-specific autoimmune diseases insvilin-dependent diabetes me!liuts (!DDM) and multiple sclerosis (MS) are characterized hy c!iroiiic inf!ammation, tissue destruction, and loss of funciion lo die pancreas or centra! nervous system {CNS), respectively, without the obviovis presence of a pathogen. T!ie challenge of working bac!i ards to identify and predict common denominators invo!ved in initiating STICIT complex chronic diseases of humans is immense. A great dea! o! research has gone into generating anitnal tiujdcls tliat resemh!e soine of the c!inica! manifestations of these two complex diseases. Organ-specifk autohntnuiiil.y is tlioug!it to result fr(.iiTi individua!s' !oss oi tolerance to se!f-antigens. Appropriate tests can identify autoreactive T cells and antibodies preceding t!ie onset of c!inica! disease in, for example, IDDM (1-4). Then, as organ-.speLilic autoimmune disease progresses, an immune response directed at a sing!e aiiligen of that organ targets particular cell types for destruction. Rare!y do additiona! aberran-24! Sarvetnick -Vinjsei, lost respionses. and autoimmunity cics in the immune response occur outside of rhe affected tissue. In IDDM, pancreatic p cells and their auiigens are the targets of destruction; in MS, the myelin sheath produced by oligodendrocytcs ol the CNS is affected, and in rhettmatoid arthritis, collagen is attacked as a self-antigcii present iu the syuovial membranes of joints. Many autoiminune diseases are mtiltifactorial with associations t(3 both genetic and environmental factors. Although strong evidence links some autoimmune diseases vvidi inherited genes, including those ofthe major histo-conipatibiliLy complex (MHC) (5, 6), theparciLipation of enviroiinieiital influences (7, 8) has been well documeuted, Iu fact, viral infections often precede both the onset ofdiabetes (9-14) aud relapses t>f MS (15-17), In experimental animals, virtises can actually cause diabetes (14, 18, 19) as well as demyelination like that iu MS (20, 21), Understanding the etiology of autoimmune diseases thoroughly enough to account for the complexity of such multifactorial processes is difficult. To simplify aud group the many components associated with the progression of tiiese diseases, we can separate their onset into three sequential steps. First, a repertoire of immuue cells with the capacity for autoreactivity is estabhshed. Appropriate MHC and T-cell receptor (TCR) alleles must be available both to present and to recognize self-antigens, thereby generating au autoreactive response. For example, susceptibility to diabetes is inherited by biimaus via MHC allek^s like HLA DQ8 or by noii-obese diabetic (NOD) mice through I-A g7, but the presence of these alleles alone does not lead to disease. Therefore, an cuvironmeutal event like infection is apt lo be a contributory factor. Second, pocencially autoreactive T cells must be activatt^d. Target tissue inflammation, possibly through viral infection, is the most likely mediator for T-cell activation. An initial, localiztid iufection would not only serve to release sequestered antigens that stimtilate autoreactive cells, but would also create a high local concentracion of cytokines aild chemokines that would further attract immune cells to the organ. Once an infection is neutralized (i,e, viral clearance), a chronic respouse to the target tissue by these newly activated autoreactive lympliocytes could remain. Third, a faihire of the Immune system to counter-regulate the autoreactive response would result in further chronicity. Without proper regulation, an excess of activated lymphocytes would remain after che infeccion had dissipaced, rcsulcing iu a breakdown of colerance to specific self-antigens. The lack of effective regulation could be gcuecically determined so that tbe individual's immune response is otU of control. A similarly plausible scenario is that a sufficiently strong iuflammatory resp(jnse could activate an equally strong auCoreacCive T-cell populaCion for which normal lymphocyte counter-regulation would be itisufTicient, Inflammation as an autoimmune process Infccticnis and iuflammatory processes can elicit cytokiue and/or chemokine release, immune cell infiltration, and tissue destruction, as well as release and presentation of both foreign and self-an tig ens. Consequently, self-reactive lymphocytes arc activated. Normal immune regulation mitiimizes the impact of tiiese autoreactive T cells by limiting their migratory capability and hfespan after infection. Logically, chen, defects in this regulatory process lead to autoimmuue disease as is the case in the lpr/lpr mouse, which develops a lupus-like syndrotiie (22). These mice lack functional Fas/FasL iutcractious that regulate targeted cell death via apoptosis as a mechanism of cotitrolling and reducing lymphocyte numbers after a pachogen iufection. Without this mechanism, autoreactive lymphocytes generated as a consequence of inflammation persist Co react with self-tissue. Chronic infection such as that observed in HIV can also promote autoreactivity (23), Similarly, prolonged inflammation iu a specific locale can cause the autoreactive respouse to spread from one antigen to another (epitope spreading) as has been observed in experimental autoimmnne encephalitis (EAE) (24) and Theiler's virus infection ofthe CNS (25-27), Local increases in specific cytokines during viral infection are most likely a major element in controlling the activation of self-reactivc lymphocytes and/or loss of tolerance t(5 self-antigen. The expression of both Thl and Th2 cytokiues like interferon (IFN)-y, ttuiior necrosis factor (TNF)-a, iuterleukin (IL)-4, IL-10, and transforming growth factor (TGF)-p in the pancreata of transgenic mice incite autoimmune diabetes (28-33) but, conversely, IL-4, IL-6 and TGF-p stippress spontaneous diahetes in NOD mice (34-36), Similarly, transgenic expression of IFN-yand TNF-a iu the CNS (37, 38) leads to the dtivckDpmcut of demyeliuacing disease and, in the case of IFN-y, enhances disease after EAE induction (39), Therefore, increased local cycokine levels that tnimic and Induce organ-specific inflammation can result in tissue damage and release of tissuespecific self-antigetis that xiltimately activate autoreaccive lymphocytes. Likewise, a viral infection can disturb ihe balance of immtuie regulation and lead to organ-specific autoimmune disease. For instance, in recent work with herpes simplL;x virus type 1 (HSV-1), which iudtices herpes stromal kcraticis (HSK), an autoimmuue disease of the eye, this virus was found to induce disease without the presence of an autoreactive T-cell population (40). The virus's presence in the eye is enough to stimulate cytokine production and, in turn, iuflammatiou. However, this inflammatory response is largely non-specific, comprised of transgeuic T cells speciftc to ovalbumin, an antigen not present iti either the virus or the iufccted mouse. The and most importantly, the seeming inability to discover infectious agents responsible for the disease, The concept that a pathogen can initiate autoimmune disease by activating lymphocytes or antibodies with the capacity to recognize cross-reacting viral and host determinants is termed molecular mimicry (44) (45) (46) (47) (48) (49) , To date, molecular mimicry has been assigned a presumptive role in the pathogenesis of several htiman diseases, including IDDM (3), ankylosing spoiidylitis Co drive chc autoimmuuicy. Unlike bystander activation described below, molecular mimicry between epitopes common to the virus aud host may restilt in an activated imnuinity whose purpose is to clear virus. Immunologicd Revievri 169/1999 I loiwitz & Saivetnck • Viruses, host -esponses, and ^l but results in an attack on cells presenting similar self-epitopes. In this inaiiiicT, a virus would induce responsiveness to a selfepitope systemically and be cleared before the onset of clinical disease. The role of molecular mimicry may be limited to the activation of the auCoreaccive respouse aud chis, as we will describe later, may not be sufficienc to initiate clinical disease. Additional environmental influences may be required to recruit the autoreactive respouse to the target tissue and induce disease. The concept of bystander damage and activation following a viral infection requires destruction of specific tissue, release of sequestered antigen and increased local immune iuflammatiou. Lymphocytes would be recruited to the tissue and those reactive to the released, sequestered self-antigen would in turn be restimulated in the inflammatory response. Thus, aut(jreactivc lymphocytes would gain access to the target tissue withotit being directly involved in the initial vira! insult or reactive to viral antigens. Successive targeted viral infections over a lifetime would ftilfiU the requirement ft)r both the generation and activation of autoinunune lymphocytes and their targeted recruitment. Tht^ role t)f virus iu this mechanism is uot only Co select the tissue, but also to indtice a strong inflammatory response. As we stated previously, recent data support the existence of molecular mimicry and bystander activation as two probable mechanisms of virus-indticed autoimmune disease. Although bystander activation requires tissue-specific damage, molecular mimicry allows for systemic infection to break tolerance Co tissue-specific antigens. By either mechanism, viruses can initiate an autoinunune process but are uot rcquirt^d at the clinical onset of disease. So, even though the mechanisms of molecular mimicry and bystander activation are somewhat different and dependent (wi chc Cropism of the virus, they are certainly not mutually exclusive, Experimental models to test both these mechanisms have been established in transgenic mice. Among the first of these models was one in wliich mice manipulated to express the viral gene products of lyniphocytit chorioineuingitis virus (LCMV) in their pancreatic [i cells did not develop spoutaneous diabetic di.sease. However, after LCMV infection of the mice, viruses were cleared by a vigorous cellular immune response tbat subsequently led to an attack on p cells and resulted in diabetes (73, 74) , This sequence of events was not limited to the pancreas, because other transgenic mice were eventually generated whose CNS specifically expressed LCMV and, after infection by that virus, developed a disease similar to MS (75). In neither case did che investigators find any evidence of infectious LCMV at the onset of clinical disease. However, virus-specific memory T lymphocytes and antibodies were found. Thus, in these experimental systems, any virus with molecular identity to an infected host's "self aiitigcu cotild initiate disease. An important emphasis is that these experimental models reflect molecular identity rather chan true molecular mimicry. Efforts to discriminate between mimicry and identity have been hindered by the inability to generate mutations in LCMV in vitro. Nevertheless, in subsequent experiments widi these transgenic mice, challenges with recombinant vaccinia viruses encoding tiie well-described cytotoxic T-lyniphocyte epitopes of LCMV did not indtice diabetes, Althotigli these vaccinia virus infecticDus did indeed generate significant numbers of anti-self (viral) specific lymphocytes (1/6,000), a threshold of 1/1,500 cyiotoxic T cells appears to be required to induce clinical disease (49, (76) (77) (78) , Thus, even in these idealized circumstances of molecular identity, not enough autoreactive lymphocytes are generated to induce disease; therefore, molecular mimicry on its own is probably inadequate to induce autoimmune disease, A second model that reflects chronic inflammatory disease similar to that observed following aii environmental insult, such as a virus, is tissue-specific IFN-y expression, p-cell spe- Memory T iymphocytes are activated more easily and produce higher levels of cytokines than naive T lymphocytes. Repeated virai infections from distinct viruses with no known cross-reactivity liave been shown to cross-activate memory lymphocytes from cariicr infectious (92, 93) More specifically, this activation of CD8' T cells hy IFN was induced by IL-1 5 and appeared to utilize the IL-2 receptor (5 that is found on CD8-T celis but not CD4+ T cells (106), Tiierefore, this CD8"-seiected reactivation makes sense and is appropriate considering that, after virai intrusion, the cyt(jiytic arm of the immune response is generaiiy the most critical in clearing the infectiou. Therefore non-specific cross-activation may participate only when CDS' T cells and their epitopes are involved in disease development. However, in the coxsackie-induced IDDM of BDG2,5 cransgeuic mice (89), in HSV-1-induced HSK (40), and iu pachogen-induced EAE (103), chc accivacion of pathogenic CD4' lymphocytes was essential for disease inducciou, and targeC tissue breakdown vi'as required. In systems dependent on CD8"^ lymphocytes, snch as the LCMV crausgenic models previously described, systemic cross-activation may be the pathology in dengue virus infections F3iabctes melhtus and aiUoininuinity in patients witb ihc congenital rubella syndrome RevInfectDis Sequence fiomology between certain viral proteins and proieins rdated io encephalomyelitis and neuritis Amino acid homology between the eucephaliiogenic site ofmydin basic protein and virtis: a mechanism for autoimniuniiy Virus-induced autoinimunity through molecular mimicry 10-2 1 7, 47. I-lall R, Molecular mimicry Molecular mimicry and autoimnuine disease Molctular mimicry and ininiunf mediated diseases Molecular mimicry and ankylosing spondylitis: possible role of a novel sequence in pullulanase o('Klf^s•rfll(l / ti-GMI igG aniibodies and Campylobactei bacteria in Guiilain-Barre syndrome Niho Y HLA DRB4 0101-restricted immunodominant T cell autoepitope of pyruvaie dehydrcigenase complex in primary biliary cirrhosis: evidence i:if molecular mimicry in human autoinunune diseases 8M83S 1845 Molecular mimicry in T cell-mediated autoimmunity: viral peptides activate htunan T cell clones specific for myehn basic protein OueUet F, \lyT.lin basic protein and Iniman coronavirus 229E cross-reactive T cdls in iTiuhipIc sclerosis Oligodendrocyte specific expression and autoantigenicity of transaldolase in multipk"-sclerosis j >'xp Med Enteroviruses and myocarditis: viral palhogencsis through replication, cytokine induction, and immunopathogenicity Coxsackievirus B-3 myocarditis. Acute and chronic forms of the disease caused by dilTcrent iinmunopatliogenic Detection of coxsackie B virus RN.A sequences m whole biood samples from adult patients at the onset of type I diabetes mellitus A prospective sEud) ol the role ofcoxsackie B and olher enierovirus inleciions in the pathogenesis of TDDM Indications that maternal coxsackie B virus infection during pregnancy is a risk factor for childhood-onset Diabetologia Several different enterovirus serotypes can be associated with prediabetic antoimmune episodes and onset of overt IDDVL Childhood DiabeCes in Finland (DiMe) Study Group Antibodies to glntamic acid detarboxylase and P2-C peptides in sera from coxsackie virus B4-infected mice and ID15M patients GAD 65 epitopes in insulin-dcpendeiU diabcies melhtus Spontaneous loss of T cell lolerance to ghitamic acid decarboxylase in muriup insulin-dependent diabeies Immune response to glutamic acid decarboxylase correlates with insulitis m non-obese diabetic mite Administering giuCamic acid decarboxylase to MOn mice prevcnis diabetes T cell cross reactivity between coxsackievirus and giutamate decarboxylase is associated with a murine diabeies susteptibilit) alJele Responses of NOD congenic mice to a ghitamic acid decarboxylase-derived peptide Polygenic control ol autoimmune fhabeies in nonobese diabetic mice Ablation of "tolerance" and induciif::>n of diabetes by virus infection in viral antigen transgenie mice Virus infection triggers insulin-dependent diabetes meliitus in a transgenic model: role of anti-self (virns) immune response CeU Viral infection of transgenic mice expressing a viral protein in oligodendrocytes leads to chronic central nervous system autoimmune disease How virus induces a rapid or slow unsei insuiin-dependent diabetes melliuis in a transgenie model Focal expression oi interleukin-2 does not break unresponsiveness to "self (viral) antigen expressed in |5 cells but enhances developmeni of aiiiominuine disease (diabetes) after initiation of an anti self immune response Using iransgenie mouse models to dissect ihe pailiogencsis of virus-induced autoimmune disorders ofthe isli^Ls o( I.angerJians and tfie cencral nervous system liiterfcron-y: mechanism of action and Cherapeutic poceutiaL Biocliem Pharmacol loiv/iLz & Sd'veL'iick • Vi-uses, ncsL responses, and autommjnity 80 Activated microglia mediate neuronal cell injury via a uitrtc oxide mecbanisni Cytokine release from micrtiglia: differential inhibition by jieiitoxilylline and dexamelhasoue. IlnleciDis 1 992; I FITocts of interferon-y on primary cultures ol human brain micixivessd endotheliai cells Detailed in vivo analysis of ijitcrleron-y induced MHC expressionin the CNS: astrocytes l.iil io express Mi IC dass land II molecules Genes in the MHC that may alTect antigen processing A molecular model oI'MIlC class-100 Interferon-mcdiated induction of class I MHC products in liuman netironal cell lines: analysis of I IT A and Cytokines up-regulate la expression in organotypic cultures of central 102. nervous system tissue Coxsackie virus-103. induced diabetes; initiation by bystander damage and not molecular mimicry Cype I interferon in vivo Recruitment and proliferation of CD8" T cells in respiratory virus infections Cross-reactivities in memory cytocoxic T lymphocyte recognition of heterolngous viruses High frequency of crossreactive cytotoxic T lymphocytes elicited during the virus-induced polyclonal cycotoxic T lymphocyCe response Counting aniigen specific CD8 T cells a reevaluation of I 08 bystander activation during viral infection Triggers of auioimniune disease in a murine TCR iransgeiiit model lor muliiple sclerosis. I 10 Bystander activation of eytotoxic T cells: studies on the mechanism 11 I and evaluation of in vivo significance in a transgeiiK mouse model Molecnlar mimicry by herpes simplex virus-type 1: auCoimmune disease 11 3. alcer viral infection Ancigeuic I 14 mimicry between measles virus and human T lymphocytes Molecular mimicry: frequency of reactivity of monodonal 1 1 5. aiitiviral antibodies iviili normal tissues Transgenic mice that express a myelin basic protcin-specillc T cell recepior develop spouianeous autoimmunity, 1 I 6 Cell Inducers of interferon and hose resistance, II. Multistranded polynudeotide complexes Pocenc and selective stimulacion of memoryphejiotype CD8+ T cells in vim by IL-1 5 How virus induces a rapid (H slow onset insulin-dependeut diabetes mellitus in a transgenic model Producl lOii of interleukin 10 by islet cells accelerates iinmuue-iiiediaied destruction of beta cells in uoiiobese diabetic mice IL-1 0 impacts autoimmune diabetes via a CD8" T cell pathuay circumventing the requircTiient lor CD4" T and B lymphocytes Frequent expression ot interleukin-10 by Epsiem-Barr virusharboring ttuiior cells of Hodgkm's disease Kinetics of cytokine expression during primary human immunodeficiency virus type I infeciion Differential expression of inflammatory cytokines parallels progression of central nervous system pathology in two dhiitally distinct models of multiple sclerosis Responses to influenza vaccination in difl'ereni T-cell subsets; a comparison of healthy young and older adults Regulatory T cells in the control ofautoimniunity: ihc. esseniial role of transforming growth factor P and inCerfeukin 4 in lhe prevention of anloimmiuie thyroiditis in rats by peripheral CD4+ CD45RC-cells and CD4+ CD8-thymocytes Hon/^it7 & Saivetnick • Viruses, host responses A role lor transforming and virus-stimulated airway smooth niusde Oidstone MBA, Sarvetnick N, Sensitizaiion to growth facior-P 1 in the inereased cells produce IL-1 1 and other IL-6-type self (virns) antigen by in iitu expression of pnenmouiiis in murine allogeneic bone cycokines, murine interferon-y marrow transplant recipients with grafi-AmJPhysiol Myasthenia gravis-like syndrome simplex virus type I iniection, 1 in chronic hepatitis C, Induced by expression of interferon y in the Blood