Update on the genetics of the idiopathic inflammatory myopathies

A number of lines of investigation suggest that, as is likely the case for other autoimmune diseases, the idiopathic inflammatory myopathies (IIM) develop as a result of specific environmental exposures in genetically susceptible individuals. Current data imply that multiple genes are involved in the etiology of these complex disorders. Targeted gene studies and whole genome approaches have begun to identify several genetic risk factors for autoimmune diseases, but the rarity and heterogeneity of the IIM have limited our knowledge of their associated genes. Current findings suggest that human leukocyte antigen (HLA) genes on chromosome 6, particularly HLA DRB1*0301 and the linked allele DQA1*0501, have the strongest associations with all clinical forms of IIM in white patients. Different HLA alleles, however, may confer risk or protection for myositis in distinct ethnic, serologic, and environmental exposure groups. Non-HLA genetic risk factors, which have been documented for other autoimmune diseases, are now being identified for the IIM. These include polymorphic genes encoding immunoglobulin heavy chains (defined by serologic markers known as Gm allotypes), cytokines and their receptors, and certain proteins that accumulate in the myocyte vacuoles of inclusion body myositis patients. Selected allelic polymorphisms of interleukin-1 receptor antagonist variable number tandem repeats and genes for tumor necrosis factor alpha and interleukin-1 alpha also have recently been associated with IIM. The pathogenic bases for the differences among the many clinically, pathologically and immunologically defined syndromes known as the IIM will be elucidated through a better understanding of the multiple genes that define risks for their development, as well as through investigations of genegene and gene-environment interactions. Curr Opin Rheumatol 2000,


Update on the genetics of the idiopathic inflammatory myopathies
Ejaz A. Shamim, MD, MS, Lisa G. Rider, MD, and Frederick W. Miller, MD, PhD A number of lines of investigation suggest that, as is likely the case for other autoimmune diseases, the idiopathic inflammatory myopathies (IIM) develop as a result of specific environmental exposures in genetically susceptible individuals. Current data imply that multiple genes are involved in the etiology of these complex disorders. Targeted gene studies and whole genome approaches have begun to identify several genetic risk factors for autoimmune diseases, but the rarity and heterogeneity of the IIM have limited our knowledge of their associated genes. Current findings suggest that human leukocyte antigen (HLA) genes on chromosome 6, particularly HLA DRB1*0301 and the linked allele DQA1*0501, have the strongest associations with all clinical forms of IIM in white patients. Different HLA alleles, however, may confer risk or protection for myositis in distinct ethnic, serologic, and environmental exposure groups. Non-HLA genetic risk factors, which have been documented for other autoimmune diseases, are now being identified for the IIM. These include polymorphic genes encoding immunoglobulin heavy chains (defined by serologic markers known as Gm allotypes), cytokines and their receptors, and certain proteins that accumulate in the myocyte vacuoles of inclusion body myositis patients. Selected allelic polymorphisms of interleukin-1 receptor antagonist variable number tandem repeats and genes for tumor necrosis factor alpha and interleukin-1 alpha also have recently been associated with IIM. The pathogenic bases for the differences among the many clinically, pathologically and immunologically defined syndromes known as the IIM will be elucidated through a better understanding of the multiple genes that define risks for their development, as well as through investigations of genegene and gene-environment interactions. The idiopathic inflammatory myopathies (IIM) are a heterogeneous group of rare systemic connective tissue diseases that are defined by the presence of acquired muscle weakness and chronic inflammation of unknown cause. Polymyositis (PM), dermatomyositis (DM), and inclusion body myositis (IBM) are the most common clinical forms of these diseases and they are likely increasing in prevalence [1]. The frequency of the clinical and serologic subsets, as well as the risk factors for developing IIM, may differ among distinct ethnic groups in geographically isolated parts of the world [2,3•,4,5].
The pathogeneses of the IIM remain unclear, although evidence from family, cohort, and candidate gene studies suggest a genetic component and that many different genes are likely necessary for the development of these diseases [6•]. Current findings imply, however, that genetic factors themselves are not adequate for the development of disease. Rather, in genetically predisposed individuals, disease probably develops only after exposure to critical environmental triggers [7,8]. In this article, we summarize the current understanding of the genetics of the IIM. The genetics of the familial noninflammatory forms of the inclusion body myopathies, which are not considered IIM, will not be covered here, but have been recently reviewed elsewhere [9].

Approaches to the study of genetic risk factors for autoimmune disease
Several approaches have been used to assess the role of genetics in the development of autoimmune diseases and some of these have been applied to the IIM (Table 1) [6•]. The oldest is the simple description of the same autoimmune disease occurring in multiple members of a single family. These multicase families with autoimmunity suggest a genetic etiology and/or shared environmental factors in the pathogenesis of these diseases. In recent decades, familial associations of several autoimmune diseases, including IIM, have been reported. To date, a total of 34 families in which two or more members have an IIM, including DM, PM, IBM, orbital myositis, or amyopathic myositis, have been published (Table 1).
Other approaches used to investigate the genetics of complex disorders include studies of disease concordance in monozygotic and dizygotic twins. Although the rarity of the IIM has prohibited such studies from being performed, twin studies in other autoimmune diseases have shown concordance rates to be higher in monozygotic than in dizygotic twins, but still less than 50% in most cases [6•]. These findings suggest important roles for both genetic and environmental factors in the pathogenesis of many autoimmune diseases. Genetic investigations of familial autoimmune diseases, in which multiple cases of a particular autoimmune disease are present within a single family, have led to interesting findings. These include: (1) in many diseases, the same genetic risk factors are often present in familial as in sporadic cases [10], although in some disorders, unique risk factors have been identified for the familial forms [11]; (2) the same genotypes can predispose to many different autoimmune disease phenotypes as well as to other diseases that are not thought to be autoimmune [12]; and (3) in some cases, serologic and genetic screening techniques can accurately predict the likelihood of familial disease [13].
In addition to the observation that the same autoimmune disease can be found within a given family, a tendency for multiple different autoimmune diseases to occur with increased frequency among first and second degree relatives of a proband with a given autoimmune disease has also been described [14,15]. These reports imply that there are common genes predisposing to many forms of autoimmunity, as has been demonstrated by meta-analyses of whole genome scanning studies [16]. In addition to these shared autoimmunity predisposing genes, however, other genes and/or environmental exposures likely determine which specific autoimmune disease an individual will develop [6•,16].
Candidate gene studies, using cohort comparisons between affected patients and racially-and geographically-matched healthy controls to assess potential disease-associated genes based upon suspected pathogenic mechanisms, have shown that the major histocompatibility complex (MHC) region on chromosome 6 has the strongest association with most immune-mediated diseases. In some cases, these risk factors have also been confirmed by family-based methodologies, such as transmission disequilibrium testing, in which unaffected parents are used as controls [17]. Through candidate gene approaches in cohort-based studies, other polymorphic genetic loci, including genes encoding cytokines and their receptors, T-cell receptors, immunoglobulins, Fc receptors, and autoantigens have been identified as risk factors for various autoimmune diseases, but their statistical associations with disease are weaker than those of the MHC [18,19]. The data overall, however, suggest that many of these other genetic loci are involved as secondary risk factors [16,20••,21•]. Candidate gene approaches, which constitute the primary data that have been used to identify immunogenetic risk factors for the IIM, are detailed below.
The most recent genetic approaches have involved whole genome scans in multiple-case families with single autoimmune diseases (  Tables 3,4) Non-MHC genes beginning to be identified (see Table 5) No studies to date Examples in other autoimmune diseases IDDM: Mennonite kindred [64] RA: Four generations of women in one family with RA [65] IDDM: 53% concordance in MZ twins vs 11% in DZ twins [66] RA: 15% concordance in MZ twins vs 4% in DZ twins [67] MS: DR2 is a risk factor for familial and sporadic MS [68] IDDM: HLA DQB1*0302-DR1*0401/DQB1*0201-DRB1*03 genotype is thought to be the major genetic risk factor in Europe [69] SLE: Non-MHC genes, including IL-10, have been identified [70] Preliminary genome scans performed for SLE, RA, JRA, and AS [71]  . Of interest, in most of the familial IIM patients, the same clinical group was shared among all members of a given family who developed myositis. The sporadic IIM group had a higher frequency of myositisspecific autoantibodies (MSAs) compared with familial IIM, and after adjustment for the clinical associations with the MSAs, no significant clinical differences were observed between the familial and sporadic IIM groups. HLA DRB1*0301 was a risk factor for familial IIM, as is the case for sporadic IIM; however, homozygosity of DQA1 was an additional unique risk factor for familial forms of myositis [28].
[29] assessed histories for autoimmune diseases in family members of juvenile rheumatoid arthritis (JRA), juvenile DM patients with sporadic disease, and geographically matched healthy controls. They noted that pedigrees of children with JRA had a significantly higher frequency of rheumatoid arthritis and pernicious anemia than controls, but this was not observed in the juvenile DM families.
The first study to assess the frequency of all autoimmune diseases in families of sporadic IIM patients was published in 1998 [14]. This prospective case-controlled investigation demonstrated that the prevalence of autoimmune disease was significantly higher (21.9%) in the 151 first-degree relatives of 21 IIM patients compared with that in 143 first degree relatives of 21 age-, gender-and race-matched probands without evidence of an autoimmune disease (4.9%) [14]. As expected, these pedigrees also demonstrated a female predominance for autoimmunity, an increase in frequency of autoimmune diseases with increasing age, and, through genetic modeling, that autoimmune diseases in these families likely have a polygenic mode of inheritance.

Evidence that HLA alleles are risk factors for idiopathic inflammatory myopathy
Although hereditary aspects of myositis were recognized half a century ago, only in the last two decades have laboratory approaches been applied in an attempt to identify the specific genes responsible for the development of the IIM (Tables 3,4). Of the many polymorphic genetic loci that may be risk factors for myositis and other autoimmune diseases, the HLA region on chromo-      Table 5. Non-HLA polymorphic genetic marker associations with the idiopathic inflammatory myopathies recognition particle autoantibodies have an increased frequency of HLA DR5 and DQA1*0301 compared with normal controls [33]. White IIM subjects with anti-PM-Scl and anti-Ku autoantibodies have higher frequencies of HLA DRB1*0301 and DQA1*0501 alleles compared with controls (Table 4). Genetic studies of other IIM serologic groups have not yet been reported.
Though gene-environment interactions are inherently difficult to study in rare diseases, certain HLA class II alleles may be risk factors for IIM associated with particular environmental exposures. Furthermore, the genetic risk factors for the development of myositis after certain exposures may differ among distinct racial groups, as may be the case with D-penicillamine [37]. In white patients from the US and in patients from the Indian subcontinent, DR2 appears to be associated with development of myositis after D-penicillamine exposure [37,38]. In the later group, DQw1 also appears to be a risk factor. In white subjects in Australia, however, HLA DR4, B18, and B35 were associated with the development of myositis after Dpenicillamine exposure. A small retrospective comparison study noted that white women who developed myositis after silicone breast implantation had a higher frequency of HLA DQA1*0102, compared with racematched controls or white IIM patients without implants [28]. These studies, though limited in power by their size, suggest the importance of both genetic and environmental components in the pathogenesis of myositis in some individuals.

Evidence that non-HLA alleles are risk factors for idiopathic inflammatory myositis
Non-HLA loci, which have been defined as likely risk factors for other autoimmune diseases [16,18,39], may also play a role in the pathogeneses of the IIM [3•,39-41] (Table 5). Risk factors for other autoimmune diseases include genes encoding cytokines and their receptors, immunoglobulins, and loci containing a large number of genes whose functions remain unclear [3•,39]. To date, few such loci have been studied in the IIM, and the published associations are preliminary and require confirmation. The likelihood that the non-HLA genes will have much weaker associations with disease than the HLA genes suggests that very large populations of patients and race-matched controls will be needed to identify most of the polygenic risk factors for IIM.
Gm allotypes are serologic markers on IgG heavy chains encoded on chromosome 14 and thought to be associated with the regulation of immunoglobulin responses to certain exposures. They have been defined as markers for a number of autoimmune, infectious and malignant diseases [41]. Gm 21 has recently been shown to be a protective factor for IIM in Koreans, but not in white patients in the US [3•].
The gene encoding the interleukin-1 receptor antagonist (IL-1RN), located on chromosome 2, has five allelic polymorphisms (A1-A5) of an 86 base pair variable number tandem repeat in intron 2 which represent two to six copies of the repeat sequence [42,43]. The specific allele IL-1RN A1, which is defined by four copies of the variable number tandem repeat, has been shown to be a risk factor for juvenile IIM in white patients (OR = 2.5) [44••]. This allele is not increased in black patients, but IL-1RN A3, representing five repeats, is a possible risk factor in this ethnic group [44••]. In contrast to studies of other autoimmune diseases, the IL-1RN A2 has not been identified as a severity factor for juvenile IIM [44••]. A second polymorphic gene on chromosome 2, IL-1α 1,2 (defined by a single G to T substitution in exon 5), has also been identified as a possible risk factor in juvenile IIM in white patients [45]. In contrast, no alleles of the polymorphic IL-1β gene have been identified as risk or protective factors in white juvenile IIM patients.
Tumor necrosis factor alpha (TNF-α), encoded within the MHC, plays a number of important roles in the regulation of normal and pathologic immune responses. The promoter region of TNF-α has two polymorphic sites at nucleotide positions -238 and -308 (referred to as TNF1 and TNF2 at each site). At both nucleotide positions, TNF2 is distinguished from TNF1 by a G to A substitution. Thus, the two alleles at position -238 are named TNF1 (-238G) and TNF2 (-238A), and the two alleles at -308 are referred to as TNF1 (-308G) and TNF2 (-308A). The TNF2 (-238A) allele appears to be a possible risk factor for juvenile IIM in white patients, although this finding may be due to its linkage with DRB1*0301 [45]. The other polymorphism of the TNF-α gene, resulting from a G to A substitution at position -308, appears to be a risk factor for juvenile DM in white patients [45]. Homozygosity of TNF (-308A) confers a high risk of death or severe neurologic involvement in patients with cerebral malaria [46]. This allele may also be a risk factor for systemic lupus erythematosus and several other autoimmune diseases [43,47]. Furthermore, in patients with juvenile DM, TNF2 (-308A), which is associated with higher levels of TNF-α after in vitro stimulation of peripheral lymphocytes, is increased in white patients with calcinosis, a chronic disease course, and capillary occlusion on biopsy, suggesting the -308 TNF-α allele may be a severity factor [45,48,49•].
HLA DRB1*0301 appears to be an important risk factor for sporadic inflammatory forms of IBM, but not for the hereditary noninflammatory forms of the inclusion body myopathies [50]. Nonetheless, recent findings imply that non-HLA genes may also play a role in the pathogenesis of IBM, the most common form of IIM in the elderly. IBM is characterized on muscle biopsy by intracellular inclusions containing many components. The prion protein has been found to be one constituent of these intracellular inclusions and is encoded by a polymorphic gene, which results in a protein containing either a methionine or valine at codon 129 [51]. An unconfirmed study has recently suggested that homozygosity at codon 129 (encoding methionine) of the prion gene may be a predisposing factor for IBM [52••]. This is of interest because overexpression of the prion gene in transgenic mice leads to a pathologic state similar to IBM [53]. An additional protein component of the intracellular rimmed vacuoles of IBM is apolipoprotein E (apoE), which is also found in the brain plaques of patients with Alzheimer disease [54]. The major known function of apoE is the transport of lipoproteins in the blood. In humans, 3 major isoforms are found: E2, E3 and E4. The single amino acid differences between the apolipoprotein isoforms confer differing affinities for apoE receptors as well as for lipoprotein subtypes. The frequency of one variant of the apoE gene, the APO E ε4 allele, is increased in patients with IBM compared with the general population. This unconfirmed result in a small population suggests that the APO E ε4 allele may be a predisposing factor for the development of IBM [55].

Future approaches for the investigation of genetic risk factors in the idiopathic inflammatory myopathies
Many challenges lie ahead in the further definition of genetic risk factors for the IIMs. A common difficulty faced by all researchers assessing the genetic basis of autoimmune diseases is the identification of genes of relatively small effect size, against a background of substantial genetic and environmental variation, using populations often of inadequate size. In addition, specific challenges in identifying genetic risk factors for the myositis syndromes include the rarity and heterogeneity of the IIM, the apparent differences in genetics among the various clinical, serologic, and ethnic groups, and the likely gene-gene interactions that define these diseases. It may be that certain alleles of many polymorphic, physically unlinked genes have been evolutionarily selected to work together as functional units to optimize immune responses to infections and other environmental exposures. If this is the case, then it is possible that some combinations of these polymorphisms may result in abnormal immune responses in certain environments, which could lead to autoimmune disease [56].
As is the case for many other autoimmune diseases, gene-environment interactions also seem to play an important role in the initiation of the IIM. Thus, multidisciplinary approaches which explore the interaction of genetic and environmental factors by integrating a variety of study designs and take into account the advantages and disadvantages of each, should be considered in future studies [8,44,57,58]. Because it is likely that the heterogeneity of the myositis syndromes reflects varied pathogenic mechanisms in different individuals, it may also be useful to study genotypes in groups of patients classified in new ways. These approaches could include assessing novel IIM phenotypes defined by (1) unique clusters of specific signs, symptoms, and laboratory abnormalities; (2) common patterns of target tissue gene expression using cDNA microarray technology; or (3) different responses to immunomodulating agents.
A major limitation has been the relatively small numbers of subjects within each of the subgroups available for investigation. To address this problem, national and international collaborative efforts have been initiated, and need to be extended, to develop serum and DNA repositories on large populations of well-characterized subjects. This will also allow expansion of these studies to other clinical, serologic and environmental exposure subgroups of myositis that have not yet been studied.
In addition to collaborative efforts to increase population sizes, new technologies and epidemiologic approaches should enhance these efforts [17, [59][60][61][62][63]. The human genome project, and its resultant high resolution mapping of MHC and non-MHC regions, may allow an understanding of the true genetic risk factors, rather than those that may only be linked to the true risks. The expanding fields of gene chip and microarray technologies also hold special promise in understanding the pathogeneses of complex disease traits like myositis, inasmuch as the interaction of many different polymorphic gene products can be rapidly assessed in large populations using these methodologies.
In summary, our understanding of the genetics of IIM is in an early stage of development. Many of the current findings in this field are preliminary observations, limited by small samples and weak effect sizes, and should not be considered definitive. The early indications of heterogeneity of genetic risk and protective factors among different serologic and ethnic groups, as well as possible geneenvironment interactions, suggest that enhanced understanding in this area should have important diagnostic, therapeutic and preventative implications for improving outcomes of myositis patients in the future.