Expression of class I and class II MHC antigens in neuromuscular diseases

Journal of the Neurological Sciences, 1989, 89:213-226

213

Elsevier JNS 03116

Expression of class I and class II MHC antigens in neuromuscular diseases Rhoda M. McDouall*, Michael J. Dunn** and Victor Dubowitz Jerry Lewis Muscle Research Centre, Department of Paediatrics and Neonatal Medicine, Royal Postgraduate Medical School, London W12 ONN (U.K.)

(Received 7 September, 1988) (Accepted 11 October, 1988)

SUMMARY The distribution of HLA class I and class II antigens has been investigated in cryostat sections of a series of 200 skeletal muscle biopsy specimens from patients with various neuromuscular disorders. Normal muscle fibres expressed no detectable class I antigens, whereas muscle fibres of patients with inflammatory myopathies and Duchenne (DMD) and Becker (BMD) muscular dystrophy showed consistently strong expression. In other neuromuscular diseases expression of class I antigens was more variable. No expression of class I antigens was observed on muscle fibres in samples from fetuses "at risk" for D M D and BMD or from female carders of these disorders. The immunocytochemical assessment of HLA class I antigen expression was confLrmed by a quantitative radioimmunoassay which demonstrated a 3-fold increase in the level of expression in muscle samples from patients with D M D and juvenile dermatomyositis. Class II antigen expression was never observed on muscle fibres in biopsies from normal individuals or any of the neuromuscular disorders. However, these antigens were expressed by endothelial cells present in these samples. Muscle specimens from fetuses and early in postnatal life showed very limited expression of class II antigens. They were expressed at a reduced level by about 3 months of age, but strong expression of class II antigens was not observed until about 1 year of age. The mechanism of induction of class I antigen expression in diseased muscle is not known. The appearance of class I antigens on diseased muscle may make the affected tissue a target for cytotoxic

* Present address: Cardiothoracic Unit, HarefieldHospital, Harefield, Middlesex,U.K. ** Present address: Department of Cardiothoracic Surgery, Cardiothoraeie Institute, Dovehouse

Street, London SW3 6LY, U.K. Correspondence to: Dr. M.J. Dunn, Department of Cardiothoracic Surgery,Cardiothoracic Unit, Dovehouse Street, London SW3 6LY, U.K. 0022-510X/89/$03.50 © 1989Elsevier Science Publishers B.V. (BiomedicalDivision)

214 T cells and may thus have a role in muscle fibre damage in inflammatory myopathies and the X-linked muscular dystrophies.

Key words: 3u-Microglobulin; Dermatomyositis; HLA class I antigens; HLA class II antigens; Inflammatory myopathy; Muscular dystrophy; Neuromuscular diseases; Skeletal muscle

INTRODUCTION The major histocompatibility antigens are a family of antigens encoded by a complex of genes, located in humans on chromosome 6 (Jongsma et al. 1973), called the major histocompatibility complex (MHC). The gene products of the MHC are divided into class I and class II antigens comprising a set of cell-surface glycoproteins. In man, class I MHC glycoproteins are encoded by 3 separate genetic loci, called HLA-A, HLA-B and HLA-C. Each of these loci encodes a single, highly polymorphic, glycosylated polypeptide chain (Mr = 45 kDa) which is associated non-covalently through its extracellular portion with fl2-microglobulin. The latter is a non-polymorphic, non-glycosylated polypeptide (Mr = 12 kDa) encoded by a gene on chromosome 15 (Goodfellow et al. 1975). Class II MHC glycoproteins are encoded by the HLA-D locus, and are subdivided into DP, DQ and DR. They are composed of two noncovalently associated polypeptides, an 0t-chain (Mr = 33kDa) and a fl-chain ( M r = 28 kDa), both of which are transmembrane in orientation and are glycosylated. As in the case of the class I MHC antigens, class II antigens exhibit a very high degree of allelic polymorphism. Class I antigens have a wide tissue distribution and their expression at the surface of target cells is essential for the recognition of foreign (e.g. viral) antigen by cytotoxic T cells. Indeed, until recently it was believed on the basis of cytotoxic assays of single-cell suspensions and absorption assays of allo- and hetero-antisera that all nucleated mammalian cells express class I M H C antigens (Berah et al. 1970). However, the use of monoclonal antibodies in immunocytochemical techniques has enabled detailed characterisation of tissue antigens in situ. From such studies it is now clear that class I antigens are not expressed in certain tissues, including both skeletal and cardiac muscle (Daar et al. 1984a; Appleyard et al. 1985; Isenberg et al. 1986). In an initial study we investigated the expression of class I antigens in skeletal muscle biopsy samples from patients with a variety of neuromuscular disorders (Appleyard et al. 1985). We found that normal muscle fibres and those from patients with congenital muscular dystrophy expressed little or no class I antigens, whereas muscle fibres of patients with inflammatory myopathies or X-linked muscular dystrophies showed consistently strong expression. In other neuromuscular diseases class I antigen expression was more variable. In the present investigation we have extended our study of class I MHC antigen

215 expression in skeletal muscle to encompass a greater diversity of neuromuscular disorders and a larger number of patients. We have also attempted to obtain a quantitative estimate of class I antigen expression using a radioimmunoassay. This analysis has been complemented by a parallel study of the expression of flz-microglobulin at the surface of normal and diseased skeletal muscle. Class II MHC antigens generally have a narrow tissue distribution, their expression at the surface of antigen presenting cells being essential for the recognition of foreign antigen by helper T cells. However, it is now clear that class II antigens can be expressed at the surface of a variety of cells and tissues (Daar et al. 1984b; Germain and Quill 1986; Pujol-Borrell et al. 1987). We have, therefore, investigated the distribution of class II MHC antigens in normal skeletal muscle and in various neuromuscular diseases.

MATERIALS AND METHODS

Muscle specimens Specimens of muscle tissue were obtained by needle biopsy of the quadriceps from patients attending the paediatric muscle clinic at Hammersmith Hospital. Control specimens were those muscle biopsy samples that showed no evidence of neuromuscular disease. Samples of muscle from fetuses "at risk" for DMD were obtained through Guy's Hospital. Normal fetal muscle was obtained from fetuses aborted for reasons not related to neuromuscular disease. Diagnosis of the various neuromuscular disorders was established according to accepted clinical, genetic and morphological criteria (Dubowitz 1985). The muscle samples used in this study are detailed in Table 1.

Immunocytochemistry Muscle samples were rapidly frozen in Arcton 12 cooled with liquid nitrogen. Transverse cryostat sections (6/am) were mounted on glass coverslips, air-dried, fixed for 10 rain in acetone, washed in phosphate-buffered saline (PBS) and incubated for 30 rain in a humid atmosphere with the appropriate primary mouse monoclonal antibody. The antibodies used were: (1)W6/32 (gift of Professor H. Festenstein, The London Hospital), specific for a human non-polymorphic determinant common to all class I antigens (Barnstaple et al. 1978), (2) anti-fl2-microglobulin (Becton Dickinson), specific for fl2-microglobulin which is associated with MHC class I antigens, (3) L227/CA2 (gift of Professor H. Festenstein, The London Hospital), specific for human monomorphic class II determinants, and (4) EN6 (gift of Professor C. Spry, St. George's Hospital), specific for an antigen on human endothelial ceils. The sections were then washed in PBS and incubated for 30 rain with affinity purified, biotinylated goat antibody against mouse immunoglobulins (TAGO) diluted in normal human serum (North London Blood Transfusion Service) to minimise non-specific binding of immunoglobulin. After incubation, the sections were washed in PBS and incubated for 60 rain with avidin-biotin-peroxidase complex (Vector Laboratories), prepared by incubating avidin D with biotinylated horseradish peroxidase for 30 rain before use. Control specimens were treated in the same way with the omission of the primary

216 monoclonal antibody step. After washing in PBS, sites of antibody binding were visualized by incubation with 0.05 ~ (w/v) diaminobenzidene tetrahydrochloride and 0.01 ~o (w/v) hydrogen peroxide in PBS for 5 min. Sections were counterstalned with haematoxylin, dehydrated, and mounted in XAM neutral medium. Quantitation of class I MHC antigen expression Class I expression in muscle tissue was quantitated by an indirect radioimmunoassay using radioiodinated second antibody. Cryostat tissue sections (6 #m) were cut onto plastic coverslips (GelBond FMC), air-dried, fixed in acetone for 10 min, air-dried and washed with PBS. For each muscle sample, duplicate sections were incubated with W6/32 primary antibody for 30 min, while duplicate control sections were incubated with PB S in the absence of primary antibody. The sections were washed in PBS and then incubated with 5 #Ci sheep anti-mouse immunoglobulin ~25I-labelled F(ab')2 fragment (Amersham International) diluted in human serum for 30 min. Sections were washed thoroughly with PBS and air-dried. The excess plastic backing around the sections was cut off, and the tissue on its plastic backing counted in a gamma-counter. Serial cryostat sections on glass coverslips were stained with W6/32 using our standard avidin-biotin-peroxidase procedure. The stained sections were photographed and the photographs digitised (Reichert MOP Videoplan) to assess the total area of muscle fibres and connective tissue in each section. The results are expressed as cpm 1251 bound to the sections per mm 2 of muscle and connective tissue.

RESULTS Class I MHC antigen expression The expression of class I MHC antigens by normal muscle and in neuromuscular diseases is summarised in Table 1. Class I antigens as assessed by the binding of antibody W6/32 were not expressed by fibres in normal muscle, whereas they were expressed by blood vessels and some interstitial cells (Fig. la). In contrast, samples from patients with X-linked dystrophies (DMD, BMD, EDMD) (Fig. lb, c) and inflammatory myopathies (JDM, DM, PM, inclusion body myositis) (Fig. ld, e) showed very strong expression of class I antigens. There was intense binding of antibody at the sarcolemma of all muscle fibres within these specimens and it was not limited to focal areas of necrosis, areas of cellular infiltrate, or to regions of perifascicular atrophy. A population of muscle fibres in these biopsies showed strong internal staining with antibody W6/32. This was often observed in areas of perifascicular atrophy in inflammatory myopathies (Fig. le). Some of the fibres with internal staining were vacuolated and necrotic, while others corresponded with the small basophilic fibres generally regarded as regenerating fibres (Fig. le). Samples from patients with other forms of muscular dystrophy (LGD, CMD, FSH) showed more variable expression of class I antigens (Fig. lg, h, i), with the degree of antibody binding to muscle fibres ranging from very slight in some biopsies to moderately strong in other cases (Table 1).

217 TABLE 1 CLINICAL DETAILS OF PATIENTS AND VISUAL ASSESSMENT OF CLASS I ANTIGEN EXPRESSION Assessed by the binding of monoclonal antibody W6/32, scored on the scale: - , none detected; - / +, trace detected; and +, weakly expressed, to + + +, strongly expressed. Ages of patients expressed as: wg (weeks gestation), h (hours), d (days), m (months), y (years). Diagnosis

n

Age

HLA class I expression

Normal Duchenne muscular dystrophy (DMD) Becker muscular dystrophy (BMD) Emery-Dreifuss muscular dystrophy (EDMD) Limb girdle dystrophy (LGD) Congenital muscular dystrophy (CMD) Facioscapulohumeral muscular dystrophy (FSH) Spinal muscular atrophy (SMA) Juvenile dermatomyositis (JDM) Dermatomyositis (DM) Polymyositis (PM) Inclusion body myositis Normal fetus "At risk" DMD fetus Definite DMD carrier Possible DMD carrier Definite BMD carrier Possible BMD carrier

23 52 15 3 4 12 2 22 23 2 5 3 4 12 2 13 1 2

ld-14y 6m-13y 3-16y 6-36y 7-12y 7h-5m 43-51y 17d-14y 3-15y 20-30y 11-75y 7-56y 8-20wg 12-20wg 30-44y 16-45y 33y 34y

+++ +++ ++ +/+ + +/+ + + -/+ +++ +++ +++ ++ - /+ -

M o s t cases o f S M A showed no expression of class I antigens by muscle fibres (Fig. If), but in a few cases weak expression on both atrophic and reinnervated muscle fibres was observed (Table 1). Class I antigen expression by muscle fibres was not generally observed in biopsies from carriers o f D M D (Fig. 2a) and B M D (Table 1), although in one such biopsy weak antibody binding to some muscle fibres was observed (Table 1). N o class I antigen expression by muscle fibres was observed in samples from normal fetuses (Fig. 3a) and fetuses "at risk" for D M D (Fig. 3b). Control sections of muscle biopsies processed by the same immunocytochemical procedure, but with the primary antibody step omitted, showed no staining associated with either muscle fibre membranes, blood vessels or interstitial cells (Fig. 2d).

Quantitation of class I MHC antigen expression The results o f the quantitative analysis of class I antigen expression in sections of muscle biopsy specimens is shown in Table 2. These data confirm the subjective, immunocytochemical assessment o f class I antigen expression, with J D M and D M D samples showing a significant, approximately 3-fold increase in the level o f expression. In contrast, there was no significant increase in class I antigen expression by samples o f muscle from patients with SMA.

218

Fig. 1. Class I MHC antigen expression by normal muscle and in various neuromuscular diseases using immunoperoxidase staining with antibody W6/32. a, normal muscle; b, DMD; c, BMD; d, JDM; e, JDM showing perifascieular atrophy; f, SMA; g, LGD; h, CMD; i, FSH. x 260.

Fig. 2. MHC antigen expression by a muscle biopsy from a definite carrier of DMD. a, class I (W6/32); b, class I (/~2-microglobulin); c, class II (L227/CA2); d, control section of muscle with no primary antibody. × 260.

219

Fig. 3. MHC antigen expression by muscle from normal fetuses (a,c,e) and fetuses "at risk" for DMD (b,d,f). a,b, class I (W6/32); c,d, class II (L227/CA2); e,f, endothelial cell marker (EN6). x 260.

TABLE 2 QUANTITATION OF CLASS I MHC ANTIGEN EXPRESSION BY INDIRECT RADIOIMMUNOASSAY USING W6/32 AND 125I-LABELLED SECOND ANTIBODY Diagnosis

n

1251 cpm- ram-2 ( + SEM) (muscle + connective tissue)

Normal JDM DMD SMA

12 9 10 5

320 ( + 40) 1059" (+ 105) 857* (+ 127) 420** ( + 75)

* P < 0.001.

** not significant.

220

Expression of ~2-microglobulin The distribution of class I M H C antigens in normal and diseased muscle tissue was also assessed using a monoclonal antibody specific for fl2-microglobulin. The results paralleled those observed using the class I specific antibody, W6/32 (Fig. 4). No fl2-microglobulin was detectable on fibres in normal muscle, whereas it was detectable in blood vessels and the majority of interstitial cells (Fig. 4a). Strong expression of fl2-microglobulin was observed on all muscle fibres in the X-linked dystrophies (Fig. 4b, c) and inflammatory myopathies (Fig. 4d). Results in patients with other dystrophies and neuropathies were again more variable. No flz-microglobulin was detectable on muscle fibres in biopsies from carriers of D M D (Fig. 2b) and BMD or in muscle samples from normal and "at risk" D M D fetuses (not shown).

Expression of class H MHC antigens Class II M H C antigen expression as measured by the binding of the monoclonal antibody, L227/CA2, was never observed on muscle fibres in biopsies from normal individuals (Fig. 5a), any of the neuromuscular diseases examined (Fig. 5b-f), and carriers of D M D (Fig. 2c) or BMD. However, these antigens were expressed by endothelial cells and the majority of the interstitial cells present in these biopsy samples. In the case of fetal muscle samples, there appeared to be very limited expression of class II M H C antigens (Fig. 3c, d), even by endothelial cells. We investigated this phenomenon in more detail using a monoclonal antibody, EN6, specific for an antigen

Fig. 4. Expressionof~z-microglobulinby normal muscleand in various neuromuscular disorders, a, normal muscle; b, DMD; c, BMD; d, JDM. × 260.

221

Fig. 5. Class II MHC antigen expression by normal muscle and in various neuromusculardiseases as assessed using antibody L227/CA2.a, normal muscle; b, DMD; c, BMD; d, JDM; e, SMA; f, CMD. x 260. expressed by endothelial cells. This antibody was bound very strongly by endothelial cells in muscle samples from both normal and "at risk" DMD fetuses (Fig. 3e, f). Muscle biopsies in early postnatal life also showed strong binding of antibody EN6 to endothelial cells (Fig. 6b) with little corresponding expression of class II M H C antigens (Fig. 6a). By about 3 months of age class II M H C antigens were expressed by blood vessels present in the muscle biopsy specimens (Fig. 6c), but at a reduced level compared with binding of EN6 (Fig. 6d). Strong expression of class II antigens was not observed until about 1 year of age (Fig. 6e), at which time binding of EN6 remained strong (Fig. 6f).

222

Fig. 6. Class II MHC antigen expression by muscle during the first year of life using antibody L227/CA2 (a,c,e) compared with the expression of the endothelial cell marker EN6 (b,d,f). a,b, 14 days; c,d, 3 months; e,f, I year. x 260.

DISCUSSION

The results of the present study have confLrmed our previous finding that normal human skeletal muscle fibres do not express class I M H C antigens as assessed by the binding of antibodies either to the 45 kDa glycosylated glycoprotein (W6/32) or to fl2-microglobulin. These findings are also consistent with those of other reports that class I antigens are not expressed by either skeletal (Rowe et al. 1983; Daar et al. 1984a; Isenberg et al. 1986) or cardiac (Rose et al. 1986) muscle. In addition, we have also demonstrated that class II MHC antigens are not expressed by normal muscle fibres. However, we cannot exclude the possibility that either (or both) of these antigens are expressed below the limit of sensitivity of the immunocytochemical procedure we have used.

223 In contrast to normal muscle, we found that class I antigens are expressed strongly and consistently by muscle fibres in biopsy samples of patients with inflammatory myopathies (JDM, PM, inclusion body myositis) and X-linked muscular dystrophies (DMD, BMD, EDMD). In these cases it was clear that class I antigens were present at the sarcolemma of all muscle fibres within the muscle biopsy. This finding is at variance with a recent report (Karpati et al. 1988), which generally corroborated our previous findings (Appleyard et al. 1985), but where it was found that class I antigen expression was associated specifically with regenerating fibres in dystrophic muscle and with areas of inflammatory cell infdtrate, muscle cell damage and perifascicular atrophy in cases ofinfiammatory myopathy. The basis for this discrepancy is unclear at present, but is should be pointed out that the antibodies we have used to assess class I antigen expression (W6/32 and anti-I~2-microglobulin) are different from those used by Karpati et al. (1988). The causes and consequences of the increased expression of class I MHC antigens which we have observed in certain neuromuscular disorders have not been elucidated. Class I antigen expression is a prerequisite for recognition and lysis by cytotoxic T lymphocytes of virus-infected cells (McMichael et al. 1977) and cells bearing aUoantigens (Zinkernagel and Doherty 1979). In this context, it is of interest that a recent report (Bowles et al. 1987) documented the presence ofcoxsackie B viral genomic RNA in a proportion of muscle biopsy samples from patients with polymyositis and dermatomyositis, suggesting a viral aetiology. However, no viral genomic sequences were detected in DMD muscle biopsy samples, suggesting that this mechanism may not be the general cause of class I antigen expression in diseased muscle. There is clear evidence that class I antigen expression can be induced in response to cell-mediated allograft rejection (Nagafuchi et al. 1985; Suitters et al. 1987) and that interferons can induce the expression of both class I (Fellous et al. 1979; Wallach et al. 1982; Lampson and Fisher 1984; Wong et al. 1984; Hunt and Wood 1986) and class II (Virelizier et al. 1984; Wong et al. 1984; Pujol-Borrell et al. 1987). These findings suggest that class I antigen expression by diseased muscle may be a consequence of the release of interferons by activated lymphocytes or macrophages present in cellular infdtrates which are commonly present in biopsy samples from patients with inflammatory myopathies (Rowe et al. 1981, 1983; Arahata and Engel 1984; Giorno et al. 1984) and are sometimes observed in muscle samples from patients with DMD (Arahata and Engel 1984). In support of this hypothesis, it has recently been reported (Isenberg et al. 1986) that, in biopsy samples from patients with polymyositis, ~-,/~- and y-interferons could be detected immunocytochemically at the periphery of muscle fibres. However, we have been unable to reproduce this finding in muscle biopsies from patients with juvenile dermatomyositis or adult forms of inflammatory myopathy (McDouall et al. 1988, and unpublished results) using monoclonal antibodies which bind to y-interferon either obtained commercially (Boehringer Mannheim) or as used in the study of Isenberg et al. (1986) (gift of Drs. D. Novick and D. Isenberg). In any case, it is unlikely that this mechanism is generally responsible for class I antigen expression in diseased muscle since Isenberg et al. (1986) detected little or no interferon deposits in dystrophic muscle. In certain tissues such as endocrine cells, which are normally class II negative,

224 induction of class II expression in response to ?-interferon or other factors can result in autoimmunity (Pujol-BorreU et al. 1987). However, we have never observed class II expression by muscle fibres in any of the neuromuscular diseases we have investigated, including the inflammatory myopathies which may have an autoimmune aetiology. These antigens are expressed by endothelial cells and some of the interstitial cells present in the biopsy samples. Interestingly, we find that class II antigens are not expressed at detectable levels by endothelial cells in fetal muscle, although such cells do express the specific endothelial cell surface antigen detected by the monoclonal antibody, EN6. Class II antigens are similarly not expressed in blood vessels early in postnatal life, become detectable at about 3 months of age, but are not expressed at a normal level until about 1 year of age. Further investigations are required to establish the mechanism of induction of class I antigen expression in muscle from patients with inflammatory myopathies and X-linked muscular dystrophies. It seems unlikely that class I expression is simply a non-specific consequence of muscle degeneration since muscle from patients with other dystrophic diseases (LGD, CMD, FSH) showing severe pathological changes did not strongly and consistently express class I antigens. The appearance of class I antigens on diseased muscle may make the affected tissue a target for cytotoxic T cells and thus may precede muscle fibre destruction. Abnormal expression of class I antigens is certainly an early event in dystrophic muscle since we have detected strong expression of these antigens in three 6-month-old patients with DMD, although we have not observed class I expression by muscle fibres from "at risk" DMD fetuses. Interestingly, it has recently been shown (Allison et al. 1988) that a class I histocompatibility gene, H-2K b, linked to the rat insulin promoter, was overexpressed in pancreatic fl cells of transgeneic mice. The mice, whether syngeneic or allogeneic to the transgenes, developed insulin-dependent diabetes without the involvement ofT lymphocytes. These results suggest that the upregulation of MHC gene products can be sufficient per se to induce cell damage. Abnormal class I antigen expression cannot be a direct result of the primary genetic defect in DMD since the protein product of the DMD gene, termed "dystrophin" has recently been identified (Hoffman et al. 1987). Nevertheless, our findings suggest a role for abnormal expression of class I antigens in muscle fibre damage in inflammatory myopathies and in some other neuromuscular disorders, particularly the X-linked muscular dystrophies.

ACKNOWLEDGEMENTS We thank Dr. Marlene L. Rose for helpful discussions during the course of this work. We are grateful to Professor H. Festenstein, Professor C. Spry, Dr. D. Novick and Dr. D. Isenberg for their generous gifts ofmonoclonal antibodies used in this study. We thank Mrs. C. Trand for preparing the manuscript and Mrs. K. Davidson for preparing the photographs. Financial support of the Muscular Dystrophy Group of Great Britain is acknowledged.

225 REFERENCES Allison, J., I.L. Campbell, G. Morahan, T.E. MandeU, L.C. Harrison and J. F. A. P. Miller (1988) Diabetes in transgeneic mice resulting from over-expression &class I histoeompatibility molecules in pancreatic //cells. Nature, 339: 529-533. Appleyard, S.T., M. J. Dunn, V. Dubowitz and M.L. Rose (1985) Increased expression of HLA ABC class I antigens by muscle fibres in Duchenne muscular dystrophy, inflammatory myopathy and other neuromuscular disorders. Lancet, i: 361-363. Arahata, K. and A.G. Engel (1984) 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 fibres invaded by T cells. Ann. Neurol., 16: 193-208. Barnstaple, C.J., W.F. Bodmer, G. Brown, G. Galfre, C. Milstein, A.F. Williams and A. Zeigler (1978) Production of monoclonal antibodies to group A erythroeytes, HLA and other human cell surface antigens - new tools for genetic analysis. Cell, 14: 9-20. Berah, M., J. Hers and J. Dausset (1970) A study of the HLA antigens in human organs. Transplantation, 9: 185-192. Bowles, N.E., C.A. Sewry, V. Dubowitz and L.C. Archard (1987) Dermatomyositis, polymyositis, and Coxsackie-B-virus infection. Lancet, i: 1004-1007. Daar, A.S., S.V. Fuggle, J.W. Fabre, A. Ting and P.J. Morris (1984a) The detailed distribution of HLA-A, B, C antigens in normal human organs. Transplantation, 38: 287-292. Daar, A. S., S.V. Fuggie, J.W. Fabre, A. Ting and P.J. Morris (1984b) The detailed distribution of MHC class II antigens in normal human organs. Transplantation, 38: 293-298. Dubowitz, V. (1978) Muscle Disorders in Childhood, Saunders, London. Fellous, M., M. Kamoun, I. Gresser and R. Bone (1979) Enhanced expression of HLA antigens and fl2-microglobulin on interferon-treated lymphoid cells. Eur. J. Immunol., 9: 446-449. Germain, R.N. and H. Quill (1986) Unexpected expression of a unique mixed-isotype class II MHC molecule by transfected L-cells. Nature, 320: 72-75. Giorno, R., M.T. Barden, P.F. Kohler and S.P. Ringel (1984) Immunohistochemical characterization of the mononuclear cells infiltrating muscle of patients with inflammatory and noninflammatory myopathies. Clin. Irnmunol. Immunopathol., 30: 405-412. Goodfellow, P.N., E.A. Jones, V. van Heyningen, E. Solomon, M. Bobrow, V. Miggiano and W. F. Bodmer (1975) The/~-microglobulin gene is on chromosome 15 and not in the HLA region. Nature, 254: 267-269. Hoffman, E.P., R.H. Brown and L.M. Kunkel (1987) Dystrophin: the protein product of the Duchenne muscular dystrophy locus. Cell, 51: 919-928. Hunt, J.S. and G.W. Wood (1986) Interferon-)' induces class I HLA and fl2-microglobulin expression by human amnion cells. J. Immunol., 136: 364-367. Isenberg, D.A., D. Rowe, M. Shearer, D. Novick and P.C.L. Beverley (1986) Localization of interferons and interleukin 2 in polymyositis and muscular dystrophy. Clin. Exp. lmmunol., 63: 450-458. Jongsma, A., H. Someren, A. Westerwald, A. Hagemeijer and P. Pearson (1973) Loacalisation of genes on human chromosomes by studies of human-Chinese hamster somatic cell hybrids. Humangenetik, 20: 195-200. Karpati, G., Y. Pouliot and S. Carpenter (1988) Expression of immunoreactive major histocompatibility complex products in human skeletal muscles. Ann. Neurol., 23: 64-72. Lampson, L.A. and C.A. Fisher (1984)Weak HLA and fl2-microglobulin expression of neuronal cell lines can be modulated by interferon. Proc. Natl. Acad. Sci. USA, 81: 6476-6480. McDouall, R.M., M.J. Dunn and V. Dubowitz (1988) The cellular infiltrate and expression of MHC antigens in the skeletal muscle of patients with juvenile dermatomyositis. Neuropath. Appl. Neurobiol., in press. McMichael, A.J., A. Ting, H. F. Zweevink and B. A. Askonas (1977) HLA restriction of cell mediated lysis by influenza virus infected human cells. Nature, 270: 524-526. Nagafuchi, Y., K. E. F. Hobbs, H.C. Thomas and P.J. Scheuer (1985) Expression of beta-2-microglobulin on hepatocytes after liver transplantation. Lancet, i: 551-554. Pujol-Borrell, R., I. Todd, M. Doshi, G.F. Bottazzo, R. Sutton, D. Gray, G.R. Adolf and M. Feldmann (1987) HLA class II induction in human islet cells by interferon-), plus tumour necrosis factor or lymphotoxin. Nature, 326: 304-306. Rose, M.L., M.I. Coles, R.J. Griffin, A. Pomerance and M.H. Yacoub (1986) Expression of class I and class II major histocompatibility antigens in normal and transplanted human heart. Transplantation, 41 : 776-780.

226 Rowe, D., D.A. Isenberg, J. McDougalt and P.C.L Beverley (1981) Characterization of polymyositis infiltrates using monoclonal antibodies to human leucocyte antigens. Clin. Exp. Immunol., 45: 290-298. Rowe, D., D.A. Isenberg and P. C. L. Beverley (1983) Monoelonal antibodies to human leucocyte antigens in polymyositis and muscular dystrophy. Clin. Exp. Immunol., 54: 327-336. Suitters, A., M. L. Rose, A. Higgins and M. H. Yaeoub (1987) MHC antigen expression in sequential biopsies from cardiac transplant patients - correlation with rejection. Clin. Exp. Immunol., 69: 575-583. Virelzier, J.-L., N. Perez, F. Arenzana-Seisdedos and R. Devos (1984) Pure interferon- 7 enhances class II antigens on human monocyte cell lines. Eur. J. Immunol., 14: 106-108. Wallach, D., M. Fellous and M. Revel (1982) Preferential effect of ~,-interferon on the synthesis of HLA antigens and their mRNAs in human cells. Nature, 299: 833-836. Wong, G. H. W., P.F. Bartlett, I. Clark-Lewis, F. Battye and J.W. Schrader (1984) Inducible expression of H-2 and Ia antigens on brain ceils. Nature, 310: 688-691. Zinkernagel, R.M. and P.C. Doherty (1979) MHC cytotoxic T cells: studies on the biological role of polymorphic major transplantation antigens determining T cell restriction-specificity, function and responsiveness. Adv. Immunol., 27: 51-177.