key: cord-0010212-p4okc6hh
authors: Bao, Shisan; King, Nicholas J. C.; Dos Remedios, Cristobal G.
title: Flavivirus induces MHC antigen on human myoblasts: A model of autoimmune myositis?
date: 2004-10-13
journal: Muscle Nerve
DOI: 10.1002/mus.880151109
sha: 4659936747ffeab537dec592fdaa042308eb5437
doc_id: 10212
cord_uid: p4okc6hh

Infection of human embryonic myoblasts by West Nile virus (WNV), a flavivirus, caused significant upregulation of class I and II MHC expression as determined by flow cytometry. After 48 hours at a multiplicity of infection of 5 pfu/cell, a sixfold increase in MHC class I expression was induced from initially low levels of expression. In contrast, MHC class II was induced de novo to five times the control fluorescence level. At least 70% of the cells were infected as determined using fluorescence microscopy and anti‐WNV antibody labeling. Myoblasts were > 90% pure as shown by anti–Leu‐19 labeling. MHC class I (but not class II) was increased threefold after exposure to virus‐inactivated supernatant from 48‐hour–infected cells, indicating the presence of factor(s) contributing to the MHC class I increase. These findings may be important in establishing a link between viral infection of human cells and induction of inflammatory autoimmune disease. We discuss the possibility of using WNV as an in vivo model. © 1992 John Wiley & Sons, Inc.

Major histocompatibility complex antigens (MHC class I and 11) are transmembrane glycoproteins. They bind viral antigen for recognition by cytotoxic (CD8+) T cells through MHC class I, and for presentation to T-helper (CD4+) cells through MHC class 11. The ensuing cellular immune response consists of the controlled, specific eradication of virus-infected cells,41 followed by suppression of this response. However, in certain individuals, autoimmune disease is thought to supervene.

Several different organs may be involved simultaneously in autoimmune attack, but in some diseases only a relatively narrow range of tissues may be affected. In the latter category, diseases such as polymyositis have been extensively investigated. Much work has uncovered evidence of humoral involvement in poly~nyositis,~~~ 1832 but during the last decade it has become increasingly clear that cellular immunity may play a significant part in disease pathogenesis. Thus, increases in MHC class 1' ' and 113* on muscle tissue have been found, as well as local infiltrations of CD4+ and CD8' T cells.',*,' Significantly, antecedent virus infection is also thought to be an important factor in the precipitation of p o l y m y o~i t i s .~~~~'~~~~~~~ Most viruses reduce cell surface MHC ex ression, either specifically or nonspecifically, 10,zK34 although some, for example, neurotropic coronaviruses, have been found to increase MHC expression on oligodendrocytes, a s t r~c y t e s "~~~~~~ and brain endothe-Viruses which induce MHC expression on normally MHC-negative cells would presumably be candidate etiologic agents in autoimmune polymyositis as well as other autoimmune diseases.

Recently, it has been found that flaviviruses increase MHC class I on fibroblastsz6 and class I1 on astrocytesz9 with little interference with cell metabolism for the first 24-48 hours. West Nile virus (WNV) is an arborvirus belonging to the family Flauiuiridae. Transmitted by mosquitoes, this hu-man pathogen causes significant morbidity and mortality throughout the world.

We report here that WNV infection of cultured human embryonic myoblasts increases MHC class I expression and induces de novo expression of MHC class I1 on these cells. We suggest that inappropriate expression and presentation of class I and 11 MHC by normally negative myoblasts/myotubes, particularly in the context of virus infection, could augment the pathogenesis of autoimmune disease. This virus may thus be useful as a tool to create a laboratory model for autoimmune muscle disease.

Treatment. Human embryonic myoblasts, isolated from 14-17 week normal human embryos, were prepared as previously d e~c r i b e d ,~ and cultured for l week in Ham's F10 medium (CSL #733 2049) containing 10% fetal calf serum (HF1 O/FCS) under standard culture conditions (SCC). NH and MRC guidelines were strictly adhered to throughout. Myoblasts were then seeded into 50-mm plastic Petri dishes at 5 x lo5 cells/ dish. The cells were infected 24 hours later with a multiplicity of infection (moi) of 5 plaque-forming units (pfu) of' WNV/cell for 48 hours. The virus concentration and the time of infection was titrated to give maximum response for both class I and I1 MHC induction with minimum cytotoxicity (Fig. 1) . Infected cells were incubated under SCC for 48 hours. Control cells were mock-infected using HFIO/FCS only. A control for MHC induction consisted of myoblasts incubated with IFN-y (100 U/mL) for 48 hours. This was titrated for maximal MHC induction (Fig. 2) .

Supernatants from WNV-infected cultures were irradiated with UV light at 1600 pW/cm2 for 12 minutes. This destroys WNV without interfering with the activity of I F N s .~~,~' Supernatants were added to fresh myoblast cultures for 48 hours under SCC. Supernatants from mock-infected cultures were divided into two. Fresh WNV at 5 pfu/cell was added to half, the other was left untreated. Both were irradiated as above, and added to fresh cultures as controls. 

Previously, using quantitation by fluorescence microscopy, we reported low MHC class I expression and no class I1 expression on cultured normal human embryonic myoblast~.~ Here we provide a more accurate quantitation of MHC expression using FCM. Figure 3A illustrates the distribution of MHC class I expression on normal human myoblasts (b), compared with that of the isotype control antibody (anti-GFAP) (a). The difference in fluorescence is about tenfold. Thus, this level is not as Figure 3B shows the distribution of MHC class I1 expression on normal human myoblasts (b), compared with that of the isotype control antibody (a). In agreement with previous finding^,^ this demonstrates that there is no detectable MHC class I1 expression on these cells.

low as was previously thought. 7,22,31

MHC class 1 and I1 expression was measured by FCM on 48-hour WNV-infected (5 pfukell), mock-infected, and IFN-y-treated myoblasts. Figure 3C shows a sixfold increase in MHC class I expression by infected myoblasts (b) compared with mock-infected myoblasts (a). MHC class I expression on IFN-y-treated myoblasts (c) was almost 10-fold higher than on mock-infected cells. Figure  3D illustrates the de novo induction of MHC class I1 expression after WNV infection (b). The fluorescence intensity is about fivefold greater than mock-infected myoblasts (a), while the IFN-ytreated group (c) showed a 16-fold increase in fluorescence. This demonstrates that the capacity for MHC class I1 induction on myoblasts is significantly greater than previously reported. This may in part reflect differences in technique3 andlor tissue origin.lg

To investigate whether soluble factor(s) secreted by infected myoblasts could upregulate MHC expression, UVirradiated supernatants from WNV-or mockinfected cultures were added to fresh myoblasts for 48 hours. Figure 3E shows that MHC class I expression was increased threefold (c), while the mock-infected control groups, either with (b) or without (a) UV-irradiated WNV, were unaffected. Thus, there was no apparent influence of irradiation breakdown products on MHC class I expression. The IFN-y-treated group (d) shows an 1 1-fold higher expression than mock-infected m yoblasts. Figure 3F demonstrates that there is no increase in MHC class I1 expression in the presence of irradiated supernatant (c), mock-infected with (b) and without (a) UV-irradiated WNV. The IFN-y-treated group (d) shows a fluorescence intensity greater than 20-fold that of non-UVirradiated WNV (a).

The purity of myoblast cultures is shown in Figure 3G . More than 90% of cells were positively labeled by the myogenic marker, antihuman Leu-19 MAb (b), compared with an isotype antibody control (a).

Infection of myoblasts on coverslips was detected using anti-WNV antibody, labeled with SAMIg-FITC and fluorescence microscopy. Infection was detectable in at least 70% of myoblasts in all experiments as a bright specific perinuclear fluorescence (Fig. 4A ) compared with uninfected controls (Fig. 4B) . DISCUSSION We report here that infection of human myoblasts by WNV significantly upregulates MHC class I and induces de novo class I1 expression. This was shown to be due, at least in part, to virusinduced secreted but unidentified soluble fact o r~, '~,~' since the virus-inactivated supernatants from WNV-infected myoblast cultures were able to induce increased MHC class I expression in fresh myoblast cultures. This may also be due to a direct virus effect.26 In contrast, MHC class I1 on myoblasts may be induced by the virus directly, as there was no induction of MHC class I1 in response to the virus-inactivated supernatant. The mechanism of this apparent direct induction by the active WNV is presently unknown. We think it may relate to the way some protein product(s) of viral transcription interact with the host cell. Investigation into possible mechanisms is presently being undertaken. Interestingly, and by contrast, UV-inactivated nonreplicating neurotropic coronavirus particles appear to be able to induce class 11 MHC in a small subpopulation of astrocytes in ~i t r o .~' This suggests different mechanisms of induction of class I1 by these two viruses and, if true, would clearly have different implications for the respective antiviral immune responses.

As the concentration of MHC is directly related to the efficiency of both induction and execution of the cellular immune r e~p o n s e , '~.~~ increases in MHC expression due to IFN-y or virus (including WNV) are accompanied by increased lysis of target cells b virus-immune and alloimmune Tc cells. 5212,2432J In autoimmune muscle disease, long speculated to have a viral etiology, increased MHC class 1" and class Il3* expression on muscle fiber membranes is frequently found in biopsies, and is associated with an influx of CD8+ and CD4+ T Evidence suggests that CD8+ cells may mediate lysis of muscle cells." However, although there is an association in inflammatory muscle disease with several different viruses (e.g., retroviruses,8*"8 Coxsackie virus, l 7 pic o r n a v i r u~, '~ and infl~enza'~), direct viral involve-ment in the pathogenesis has been difficult to prove.

Hypothetically, viral infection resulting in a virus-specific cellular immune response would induce the release of IFN-y by T cells35 and thus induce an increase in MHC expression on local uninfected muscle cells.20 In susceptible individuals, this could facilitate crossreactive killing of uninfected muscle cells by antiviral T cells. Moreover, the induction of aberrant MHC expression might break self-tolerance and elicit an MHCrestricted antimuscle T-cell r e s~o n s e .~ Subsequent IFN-y induction of MHC by such autoimmune T cells would act as a positive feedback in this scenario.

Most viruses, however, in reducing cell surface MHC expression, would preferentially encoura e the formation of high-affinity antiviral Tc cells. These undoubtedly kill more efficiently if there is increased MHC expression on target cells that are infected.5724933 Nevertheless, the high affinity of these Tc cells for virus-infected target cells would not usually result in crossreactive killing of uninfected cells despite the increased MHC expression of the latter. In theory, viruses which increase MHC expression in infected cells in vivo would induce an antiviral cellular immune response consisting mostly of low-affinity Tc cells with a high degree of crossreactivit for MHC alone. This occurs with WNV in vitro.2 In addition, a tissue-specific, MHC-restricted response might simultaneously be induced in susceptible individuals in vivo. Such viruses would thus be candidate etiologic agents in autoimmune polymyositis and other autoimmune disease.

For reasons of hosthirus survival, virusinduced autoimmunity cannot be a common event, although the reasons for this are not well understood. Nevertheless, under certain conditions, particularly in geneticall susceptible individuals (e.g., HLA-B8 carriers *), it seems likely that autoimmune muscle disease could develop subsequent to such a viral infection. Ideally, we would like to be able to demonstrate a virus-specific and allospecific Tc-cell response in the human myoblast system reported here. However, although practicable in the murine system, there are severe ethical problems associated with doing this in the human embryonic system.

In conclusion, these findings suggest the use of WNV infection as a possible model for autoimmune muscle disease development in mice. Our preliminary findings suggest that murine embryonic myoblasts infected with WNV behave in the 1657 s 7 .

1276 WNV Increases Myoblast MHC same way as human myoblasts. It seems reasonable to suggest that investigation of such an animal model would shed useful light on events related to virus-induced autoimmune muscle disease pathogenesis in humans.

Monoclonal antibody analysis of mononuclear cells in myopathies. I. Quantitation of subsets according to diagnosis and sites of accumulation and demonstration and counts of muscle fibers invaded by T cells

Monoclonal antibody analysis of mononuclear cells in myopathies. 111. lmmunoelectron microscopy aspects of cell-mediated muscle fiber injury

Elevated MHC class I and I1 antigens in cultured human embryonic myoblasts following stimulation with ?-interferon. Immunol Cell B i d Blanden RV, Hodgkin PD, Hill A, Sinickas VG, Mullbacher A: Quantitative considerations of T-cell activation and self tolerance

Interferon enhances the susceptibility of virus-infected fibroblasts to cytotoxic T cells

Circulating antibodies in polymyositis and other muscle-wasting disorders

The detailed distribution of HLA-A, B, C antigens in normal human organs

Polymyositis associated with AIDS retrovirus

Neuromuscular diseases associated with human immunodeficiency virus infection

Expression of histocompatibility antigens H-ZK, -D, and -L is reduced in adenovirus-12-transformed mouse cells and is restored by interferon-y

Monoclonal antibody analysis of mononuclear cells in myopathies. 11: Phenotypes of autoinvasive cells in polymyositis and inclusion body myositis

Retrovirus-induced changes in major histocompatibility complex expression influence susceptibility to lysis by cytotoxic lymphocytes

Class I1 transplantation antigens: Distribution in tissues and involvement in disease

Acute fulminant myoglobinuric polymyositis with picornavirus-like crystals

Isolation of influenza virus from muscle in myoglobinuric polymyositis

Cytotoxic T cell activation by class I protein on cell-size artificial membranes: Antigen density and Lyt-213 function

Coxsackievirus aggregates in MUSCLE & NERVE

HLA-D related (DR) antigens in various kinds of myositis

Localization of interferons and interleukin-2 in polymyositis and muscular dystrophy

Differential modulation of MHC class I antigen expression on mouse brain endothelial cells by MHV-4 infection

WNV-induced modulation of class I major histocompatibility antigens: implications for T c cell/target cell interactions, T lymphocyte response to flavivirus infections

Relationship between surface H-2 concentration, size of different target cells, and lysis by cytotoxic T cells

Interferon independent increases in class I major histocompatibility complex antigen expression following flavivirus infection

Induction of class I major histocompatibility complex antigen expression by West Nile virus on y interferon-refractory early murine trophoblast cells

Altered serological and cellular reactivity to H-2 antigens after target cell infection with vaccinia virus

Flavivirus infection up-regulates the expression of class I and class I1 major histocompatibility antigens on and enhances T cell recognition of astrocytes in vztro

Viral particles induce la antigen expression on astrocytes

Circulating autoantibody against human myoglobin in polymyositis

HLA-DR expression, T lymphocyte phenotypes, OKMl and OKT9 reactive cells in inflammatory myopathy

Quantitative differences in the expression of parentally-derived H-2 antigen expression on lymphomyeloid cells in semi-allogeneic irradiation chimeras

Adenoviruses of subgenera B, C, D, and E modulate cell-surface expression of major histocompatibility complex class I antigens

Release of discrete interferons by cytotoxic T lymphocytes in response to immune and non-immune stimuli

Lymphocyte antigen Leu-19 as a molecular marker of regeneration in human skeletal muscle

Clonal analysis of cytolytic T lymphocyte-mediated lysis of target cells with inducible antigen expression: correlation between antigen density and requirement for Lyt-2/3 function

Human immunodeficiency virus-associated myopathy: analysis of 1 1 patients

Coronavirus infection induces H-2 antigen expression on oligodendrocytes and astrocytes

Silberberg DH: Induction of glial cell MHC antigen expression in neurotropic coronavirus infections

Restriction of in nitro T cell mediated cytotoxicity in lymphocytic choriomeningitis within a syngeneic or semiallogeneic system