Multiple sclerosis: doubling down on MHC

Trends in Genetics OPEN ACCESS Opinion Multiple sclerosis: doubling down on MHC Roland Martin, 1,* Mireia Sospedra, 1 Thomas Eiermann, 2 and Tomas Olsson 3 Human leukocyte antigen (HLA)-encoded surface molecules present antigenic peptides to T lymphocytes and play a key role in adaptive immune responses. Besides their physiological role of defending the host against infectious pathogens, specific alleles serve as genetic risk factors for autoimmune diseases. For multiple sclerosis (MS), an autoimmune disease that affects the brain and spinal cord, an association with the HLA-DR15 haplotype was described in the early 1970s. This short opinion piece discusses the difficulties of disentangling the details of this association and recent observations about the functional involvement of not only one, but also the second gene of the HLA-DR15 haplotype. This information is not only important for understanding the pathomechanism of MS, but also for antigen-specific therapies. Highlights Genes of the human leukocyte antigen (HLA)-DR15 haplotype show a strong association with disease risk for multiple sclerosis (MS). Current genotyping tools failed to pinpoint the second HLA-DR allele, DRB5*01:01, besides the well-known DRB1*15:01 gene, despite their close to 100% linkage disequilibrium (LD). Functional studies illustrate the complex mechanisms of how both DR15 alleles shape a CD4+ T cell repertoire that is selected and maintained by peptides from the HLA-DR molecules but reacts more strongly to peptides from both MS-associated infectious organisms and autoantigens. Introduction MHC, called HLA in humans, describes a set of surface proteins that play central roles in adaptive immune recognition by T lymphocytes. The genes coding for HLA proteins are located on a 3.6 Mb stretch on the short arm of chromosome 6 (6p21) (Figure 1A, Key figure). In order to understand the difficulties in deciphering the involvement of different HLA molecules in autoimmune diseases (AIDs), we briefly summarize the genetic organization of the HLA locus, which genes code for which proteins, and how these form membrane proteins that are involved in immune recognition. The multiple roles of disease-associated HLA-DR molecules include serving as antigen-presenting structure and as a source of antigen. The cooperativity between two HLA-DR alleles of the same haplotype likely applies to other HLA-disease associations as well and should be examined. There are three HLA class I proteins, HLA-A, -B, and -C, encoded by their respective genes. HLA class I proteins consist of three domains (α1, α2, and α3) and form membrane heterodimers together with the invariant β2 microglobulin (Figure 1B). Their membrane-distal α1 and α2 domains shape a groove, to which antigenic peptides bind and are recognized by T cell receptors (TCRs) of CD8+ T lymphocytes (Figure 1B). HLA class I binding grooves are closed at either end and bind peptides with a length of nine or ten amino acids. All nucleated cells and platelets express HLA class I molecules, perhaps with the exception of intact neurons, the cornea, sperm, and some trophoblast cells in immune privileged sitesi [1,2]. A second set of HLA genes codes for the HLA class II molecules, HLA-DR, -DQ, and -DP. HLA class II molecules consist of two different proteins, α and β chains, which pair to build the respective HLA class II molecules. The α1 and β1 domains of the α and β chains in HLA class II molecules shape a groove, which binds peptides recognized by TCRs of CD4+ T lymphocytes (Figure 1B). Different from HLA class I molecules, the antigen-binding groove is open at either end and HLA class II molecules can present peptides of up to 25 amino acids. However, the number of amino acids inside the binding groove is also nine to ten. HLA class II molecules are only expressed on the cell surface of immune cells that are involved in antigen presentation, so-called antigenpresenting cells (APCs) [3]. In inflammatory conditions they can also be expressed on other cell types, for example, thyroid epithelial cells [4] or astrocytes [5] and oligodendrocytes in the brain [6]. HLA-DQ and -DP molecules consist of polymorphic α and β chains encoded by HLA-DQA1, -DPA1, -DQB1, and -DPB1 genes. HLA-DR molecules are formed by a nonpolymorphic α 784 1 Neuroimmunology and Multiple Sclerosis Research, Neurology Clinic, Frauenklinikstrasse 26, 8091 Zurich, University Hospital Zurich, University Zurich, Switzerland 2 Institute of Pathology, University Medical Center Hamburg-Eppendorf, Hamburg 20251, Germany 3 Neuroimmunology Unit, Department of Clinical Neuroscience, Karolinska Institutet, 17176 Stockholm, Sweden *Correspondence: [email protected] (R. Martin). Trends in Genetics, September 2021, Vol. 37, No. 9 https://doi.org/10.1016/j.tig.2021.04.012 © 2021 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Trends in Genetics OPEN ACCESS Key figure Glossary (A) Schematic view of the human leukocyte antigen (HLA) region on chromosome 6p21 Diversity: the human leukocyte antigen (HLA) complex on chromosome 6p21 encompasses approximately 400 genes, among them the highly polymorphic HLA class I genes -A, -B, and -C and the HLA class II genes DRB1, -DRB3, -DRB4, -DRB5, -DPA, DPB, - DQA, and -DQB, which play important roles in adaptive immune function. Besides genes coding for olfactory receptors and natural killer cell receptors, HLA class I and II genes are among the most diverse of the human genome. Genome-wide association study (GWAS): study that employs sets of genetic markers spanning the entire genome, usually in a large number of individuals with or without a trait, for example, a specific disease, to search for an association between specific genetic markers and the trait. Linkage disequilibrium (LD): the occurrence of genes of two or more loci that are nonrandomly associated in a given population. SNP: difference of a single nucleotide between members of a given species that occurs with a frequency of at least 1%. (A) Class II DP DQ 400 kb B1 Class III A1 DR B 50 kb B1 A1 Class I C 100 kb B1 B3/4/5 A A 1,270 kb 1,070 kb 6:29,942,554 6:33,076,042 (B) CD8 binds the α3 domain of HLA class I CD4 binds the β2 domain of HLA DR α/β APC APC (C) DR B1 HLA class I β α3 α1 α2 α β Peptide HLA class II TCR α β CD8+ T cell Peptide CD4 TCR A α2 β2 α1 β1 CD8 B5 α DR2a DR2b β CD4+ T cell B cell Trends in Genetics Figure 1. In the DR15 haplotype, but also in many other HLA-DR haplotypes, the genes coding for two HLA-DR beta chain proteins lie in close proximity. In the DR15 haplotype DRB1*15:01 and DRB5*01:01 are in close to perfect linkage disequilibrium (i.e. always occur together). (B) Trimolecular complex of a HLA class I molecule, associated beta 2 microglobulin with embedded peptide, the coreceptor molecule CD8 and the T cell receptor (TCR) alpha/beta heterodimer, which contacts the HLA class I/peptide molecule (left). Trimolecular complex of a HLA class II molecule, that is HLA-DR alpha and -beta with embedded peptide, the coreceptor molecule CD4, and the TCR alpha/beta heterodimer, which contacts the HLA-class II/peptide molecule (right). Note that the alpha 1 domain of HLA-DR alpha and the beta 1 domain of HLA-DR beta form the peptide binding groove. (C) The DR beta chains, together with the DR alpha chain, are codominantly expressed and form two HLA-DR alpha/beta heterodimers (i.e., DRA*/DRB5*01:01 = DR2a and DRA* and DRB1*15:01 = DR2b). The functional heterodimers appear on the surface of antigen-presenting cells, for example, a B cell (Art credit: Katie Vicari). chain (DRα), encoded by the HLA-DRA gene, and a highly polymorphic β chain (DRβ) that can be encoded by four different genes, HLA-DRB1, -DRB3, -DRB4, and -DRB5. The HLA complex belongs to the most polymorphic regions in the human genome [7] and all HLA class I and II genes, except HLA-DRA, are highly polymorphic. To illustrate this complexity, so far, more than 2400 alleles are known for the HLA-DRB1 gene alone, resulting in an enormous diversityii (see Glossary) [8]. In fact, it is expected that no two individuals with completely matching HLA class I and II types can be found in the human population, except for monozygotic twins. Considering all HLA-A, -B, and -DRB1 alleles that can currently be typed by molecular Trends in Genetics, September 2021, Vol. 37, No. 9 785 Trends in Genetics OPEN ACCESS methods, the estimated number of possible genotypes is 9 × 1025, although the real number will be considerably lower due to the linkage disequilibrium (LD) [9]. Certain combinations of HLA genes, and in particular HLA-DRB genes, due to their close proximity, are usually inherited together and referred to as haplotype. Most HLA-DR haplotypes express a HLA-DRB1 gene, which makes up a heterodimer with HLA-DRα, and a second HLA-DRB gene, -DRB3, -DRB4, or -DRB5, which forms a second membrane heterodimer with HLA-DRα (Figure 1A,C). Exceptions are HLA-DR1, -8, and -10 haplotypes, which encode only the HLADRB1 gene (Table 1). Seminal discoveries mark the increasing understanding of MHC/HLA molecules. They were first identified by P. Gorer in inbred mice in 1936 [10]. Subsequently, Dausset and Bercy demonstrated their equivalent in humans [11] and Benacerraf and Snell their involvement in immune responses, organ rejection, and uniqueness with respect to histocompatibility in transplantation [12,13]. Zinkernagel and Doherty showed that T cell responses are restricted by self-MHC (i.e., T cells recognize antigen together with autologous MHC molecules) [14]. Snell, Dausset, and Benacerraf received the Nobel Prize in 1980 and Zinkernagel and Doherty were awarded the Table 1. Discovery and evolution of the HLA-D region and its correlation with DR antigens and DRB genes 1st β-chain 2nd β-chain HTCa DRB1*01:01 - - DRB1*15:01 DRw51 DRB5*01:01 [16] DRB5*01:02 [17] DRB5*02:02 [79] Cellular split DR1 – Dw1 DR2 DR15 Dw2 PGF Dw12 BGE DRB1*15:02 DR16 Dw'AZH' AZH DR17 Dw3 – DRB1*04:01 Dw10 DRB1*04:02 Dw13 DRB1*04:03 Dw15 DR11 DR13 DR14 DRw52b DRB3*01:01 DRw53 DRB4*01:01 [80] Dw5 DRB3*01:01 [81] DRB1*04:05 DRB1*11:01 DRw52 DRB1*12:01 Dw18 HHKB DRB1*13:01 DRw52 Dw18 CB6B DRB1*13:01 DRB3*02:01 Dw19 WT47 DRB1*13:02 DRB3*03:01 Dw'HAG' HAG DRB1*13:03 DRB3*01:01 [24] Dw9 TEM DRB1*14:01 DRB3*02:01 [82] Dw16 AMALA DRB1*14:02 DRB3*01:01 DR7 – Dw7 DRB1*07:01 DRw53 DR8 – Dw8 DRB1*08:01 DRw52b DR9 – – DRB1*09:01 DRw53 DR10 – – DRB1*10:01 – Abbreviation: HTC, homozygous typing cell. DRw52 epitope encoded by the DRB1 gene in DR8. 786 Refs DRB1*04:04 KT3 DR12 DR6 DRB gene DRB1*03:02 Dw4 Dw14 DR5 DRB1*16:01 DRB1*03:01 DR18 DR4 b antigen Serological split DR3 a DRB gene antigen Trends in Genetics, September 2021, Vol. 37, No. 9 – – [34] Trends in Genetics OPEN ACCESS same in 1996, thus underscoring the importance of MHC/HLA molecules for immune function and physiology in general. Even before the genes for individual HLA class I and II alleles had been identified, the involvement of HLA not only in host defense, but also their association with AID became clear, for example, in MS, rheumatoid arthritis, type 1 diabetes, and ankylosing spondylitis [8,15]. For the sake of brevity, other class I and II molecules such as HLA-E, -G, or HLA-DM, -DO, which are not directly involved in antigen presentation, are not discussed here [8,15]. Genetic studies of an association between HLA molecules and MS Following observations from organ transplantation and the landmark discoveries discussed earlier, knowledge about the genetic organization of the HLA complex has rapidly grown and more and more individual alleles are still being discovered. In fact, new alleles are being added each year. The HLA gene nomenclature evolved in parallel and has been adapted several times during the past more than five decades by consensus among experts and guidance by the World Health Organizationiii. Even for immunologists, the complexity is sometimes bewildering and not easy to grasp. To give one example, the MS-associated HLA-DR15 haplotype was formerly termed HLA-DR2 (Table 1) [16,17]. Assignment of the DR2 type was based on serological typing techniques and it was already known at that time, by cellular typing with primed lymphocytes, that individuals who were typed by sera from multiparous women as DR2+ could be subdivided further into subtypes, which were denoted with HLA-Dw (w standing for workshop or preliminary denotation) types (e.g., HLA-DR2 Dw2). Later, it was found that there were multiple DRB1 genes, which were responsible for the serotyping group HLA-DR2, and these were subsequently termed HLA-DRB1*15- and -DRB1*16 alleles [16–18]. In contrast to HLA-DR4 and -DR6 groups, no single gene could be identified for the different HLA-Dw types within the HLA-DR2 group and it was realized that the cellular typing result might be due to differences in a second, tightly linked HLA-DRB5 gene, which was tentatively assigned the term HLA-DRw51, based on serotyping (Table 1). At that time, it also became clear that large pieces of 6p21 encompassing multiple HLA genes are usually inherited together and referred to as haplotype. A genetic method, restriction length polymorphisms (RFLPs), which had been used prior to typing by sequencespecific oligonucleotides, and DNA sequencing, were used to identify such haplotypes [19]. The HLA class II haplotype designation is now based on the HLA-DRB1 gene of the individual (e.g., HLA-DRB1*15:01+ individuals are considered HLA-DR15 haplotype carriers). In the HLA-DR15 and -DR16 haplotypes, the second HLA-DRB gene was termed HLA-DRB5. In HLA-DR3, -DR11, -DR12, -DR13, and -DR14 haplotype carriers, the second HLA-DRB gene is termed HLA-DRB3, and in HLA-DR4, -DR7, and -DR9 haplotype carriers, HLA-DRB4. As already mentioned, more than 2400 HLA-DRB1 alleles are knownii. Also, it has to be noted that the difference in the number of genes between the ‘standard tool’ of immunologists, that is, inbred rodent strains such as C57/BL6 or SJL mice, which express two MHC class I and one MHC class II allele, and outbred heterozygous humans, which express up to six HLA class I and eight HLA class II molecules on their two chromosomes, is enormousiii [7]. This difference explains why findings about antigen specificity and MHC/HLA restriction can only be extrapolated to a limited degree from rodents to humans, or there is no equivalent data in rodents at all. The observations of functional involvement of two HLA class II alleles in MS, that will be discussed later, could therefore not have been examined in experimental rodent models. Regarding the association of certain HLA class II alleles with certain organ-specific AIDs, the complexity extends beyond single HLA alleles, because HLA class II haplotypes are often in tight LD and this is particularly true for certain HLA-DRB1 and the respective HLA-DRB3, -4, and -5 alleles, for example, in the MS-associated HLA-DR15 haplotype [20]. Combinations of Trends in Genetics, September 2021, Vol. 37, No. 9 787 Trends in Genetics OPEN ACCESS specific HLA-DR and -DQ alleles and even HLA class I alleles are frequently found in populations of distinct ethnicity [7]. These are referred to as extended or complex haplotypes. HLA-A3, -B7, -DR15, and -DQw6, the genes of which are associated with MS, is one example in individuals with Caucasoid background. It is assumed that the evolutionary pressure of infections and mounting efficient or poor cell-mediated immune responses in the context of certain HLA allele combinations is responsible for the over- or under-representation of certain haplotypes in different ethnicities and geographic areas. Other important factors that contributed to haplotype selection are ancestral population composition and founder effects. As noted earlier, an association between HLA class I, particularly HLA-B*03:01, and class II molecules, namely HLA-DR15 and MS, has been known for almost five decades [21,22]. There are few AIDs for which the HLA association is even tighter than for MS. Narcolepsy and the association with HLA-DQw6, the heterodimer of HLA-DQA*01:02 and -DQB*06:02, which is also part of the extended HLA-DR15 haplotype, is the best example [23,24]. More than 90% of patients with narcolepsy are positive for HLA-DQw6. The association between MS and HLA-DR15 is not as strong but has been confirmed by numerous studies during the past decades. The HLA-DR15 haplotype contributes by far the most to the genetic risk of Caucasoid MS patients, with estimates reaching up to 60% [25]. Until 2007, when the first large genome-wide association study (GWAS) was published [26], the only genes that had been confirmed by multiple studies assessing disease association with candidate genes had been HLA-DR15, -DQw6 [27,28], and the HLA class I alleles A3 and B7 [29,30]. Since then, the International Multiple Sclerosis Genetics Consortium (IMSGC) has conducted multiple GWAS studies, with steadily increasing patient numbers; the last one with 47 000 patients and 68 000 controls [31]. The latter reported a P value for the association between MS and the HLA-DRB1*15:01 allele of approximately 10–1900. One of the GWAS studies focused only on HLA associations with MS [32] and confirmed the very strong association between the HLA-DRB1*15:01 allele and MS [32]. Besides the previously known HLA class II alleles HLA-DRB1*13:03, -03:01, -08:01, -DQA1*01:02/DQB1*06:02, and -DQB1*03:02, which confer risk with different effect sizes and less than HLA-DRB1*15:01 and a few protective HLA class I alleles, particularly HLA-A*02:01, it described interactions between pairs of class II alleles HLA-DQA1*01:01-DRB1*15:01 and -DQB1*03:01-DQB1*03:02 [27,28,32]. Further, the second most important genes affecting MS are also within the HLA complex and either confer risk (HLA-A*03 and -B*07) [22,30] or protection (HLA-A*02 and -B*44) [32], a phenomenon that has remained enigmatic functionally. However, recent evidence suggest that the influence is not through antigen presentation, but possibly due to actions on the type I interferon system [33]. Interestingly in the context of HLA associations with MS, it was already described in 1991 in the HLA typing field that the alleles for the two HLA-DRB genes that are found in the HLA-DR15 haplotype, DRB1*15:01 and DRB5*01:01, are in very strong, almost 100% LD in Caucasoid individuals [20,34], the ethnic group with the highest prevalence of MS. This strong LD also holds for MS patients and was first reported in 1995 [35]. Despite this long-known LD from HLA typing data in the context of organ transplantation, the Moutsianas et al. GWAS study [32], which specifically addressed HLA associations with MS, did not mention the second allele of the HLA-DR15 haplotype DRB5*01:01, nor any of the other alleles of the second HLA-DR locus. The reason for not addressing this point, despite its importance for understanding which and how many HLA class II alleles are functionally involved in MS, was that the SNP-based genotyping arrays allowed the assignment of HLA-DRB1 genes, but SNPs were not sufficiently tightly spaced for assigning the second HLA-DRB gene in the respective haplotype. Probably even more important, the near perfect LD between HLA-DRB1*15:01 and -DRB5*01:01 does not permit imputation of HLADRB5* genotypes with sufficient statistical power, despite the large cohorts of recent studies. 788 Trends in Genetics, September 2021, Vol. 37, No. 9 Trends in Genetics OPEN ACCESS This led to the curious fact that almost no attention had been paid to the HLA-DRB5*01:01 allele in the rapidly advancing genetics field in MS during the past 15 years. Functional studies show that both DR15 alleles are associated with MS Support for the importance of both HLA-DR15 alleles came primarily from functional studies. When it was realized that there is a second HLA-DRB gene in the HLA-DR15 haplotype, the membrane heterodimer consisting of HLA-DRα and the HLA-DRβ chain of the HLADRB5*01:01 allele was termed DR2a and the heterodimer of HLA-DRα with the HLA-DRβ chain of HLA-DRB1*15:01 was termed DR2b (Figure 1C). In order to address their importance, two groups examined the HLA class II restriction of autoreactive CD4+ T cells specific for the myelin protein, myelin basic protein (MBP), by using transfectants expressing either DR2a or DR2b and demonstrated that distinct MBP peptides were recognized in the context of either DR2a or DR2b [36,37]. Interestingly at that time, DR2a appeared dominant when cytotoxicity was used as a functional readout for CD4+ MBP-specific T cell lines (TCLs) [38]. The relevance of MBP 83-99 as an immunodominant myelin peptide in MS has been shown by numerous studies [36,37,39–42]. Seminal findings are the crossreactivity of MBP 83–99-specific T cell clones (TCC) with peptides of the MS-associated environmental risk factor, Epstein Barr virus (EBV) [43], the proinflammatory phenotype of these cells [44], and that humanized mice expressing either DR2b and a MBP 83-99-specific TCRα/β [45], but also those expressing DR2a and another MBP 83–99-specific TCR [46], develop spontaneous experimental autoimmune encephalomyelitis (EAE), an animal model showing parallels to MS. Further interesting observations (summarized in Table 2) were that the cytotoxic mechanisms of MBP-specific TCCs that are restricted by DR2a or DR2b differ [44,47]. DR2a-restricted MBP-specific TCC employ perforin-mediated damage, while DR2b-restricted, MBP-specific TCCs lyse target cells by Fas/Fas-ligand-mediated interactions [44,47]. When the sequences of peptides that had been eluted from HLA-DR2a and -DR2b were aligned, it became clear that the amino acids that are responsible for binding to these molecules are similar in distinct pockets of the HLA class II binding grooves, but with a shift of three amino acids [48,49] (Figure 2). The predictions of DR2a- and DR2b-binding from immunopeptidome analyses were confirmed by X-ray crystallographic studies [50–52] and indicated that not only the MBP 83–99 peptide, but also other peptides with certain characteristics, would bind to both alleles. These data suggested an as yet unrecognized ‘cooperativity’ between the tightly linked HLA-DR15 alleles. Important to note, HLA-DR15 is not only associated with MS, but also other AIDs, autoimmune uveitis [53], and Goodpasture syndrome [54], with infectious diseases such as John Cunningham virus (JCV) infection of the brain [55], and with EBV [56]. Due to the increasing understanding of T cell recognition, several studies added important further data to this theme. Lang et al. described that a DR2b-restricted, MBP 83–99-specific TCR [43] not only recognized the MBP peptide, but, surprisingly, also an EBV peptide, although presented by DR2a [57]. In the context of DR2a, DR2b, and MS, this was the first report to show that a CD4+ TCC and its TCR were not only crossreactive to two peptides from an MS-relevant autoantigen and an environmental pathogen, but also crossrestricted, that is, capable of recognizing different peptides together with two MS-associated HLA-DR alleles. Subsequent studies demonstrated that crossreactivity and crossrestriction extend beyond this special case. Several TCC that had been isolated from an MS patient’s cerebrospinal fluid (CSF) during acute relapse recognized multiple different peptides in the context of DR2a and DR2b, some even with HLA-DQw6, the HLA-DQ α/β combination of DQA1*01:02/DQB1*06:02, which is tightly linked with the two HLA-DR alleles in the MS-associated HLA-DR15 haplotype [58,59]. These findings indicated that crossreactivity and also crossrestriction of antigen-specific T cells are phenomena that occur more frequently than realized before. One conclusion from these findings was that crossreactivity and crossrestriction of these TCC may be due to the fact that they are autoreactive and therefore not as stringent in recognizing distinct peptide/HLA-DR combinations, possibly due Trends in Genetics, September 2021, Vol. 37, No. 9 789 Trends in Genetics OPEN ACCESS Table 2. Functional evidence supporting that both DRB1*15:01 (DR2b) and DRB5*01:01 (DR2a) are involved in the pathogenesis of MS MBP-specific T cell lines recognize epitopes in the context of either DR2a or DR2b [36,37]. The immunodominant epitope MBP 83–99 binds to and is recognized in the context of both alleles [36,37]. DR2a-restricted, MBP-specific T cell clones are either noncytotoxic or kill via perforin (DR2a-restricted TCCs), DR2b-restricted TCCs via less efficient Fas/Fas-L-mediated mechanisms [44,47]. DR2a- and DR2b-restricted myelin-specific and generally autoreactive T cells express proinflammatory phenotypes [83]. Both an MBP 83–99-specific TCR restricted by DR2b [45] and another restricted by DR2a [46] lead to spontaneous EAE in humanized transgenic mouse models coexpressing the TCR and the HLA-DR molecule. The peptide binding motifs of DR2a and DR2b show similarities in HLA anchor positions. The peptide repertoires therefore are also partly overlapping (e.g., for MBP 83–99) [48,49,75]. Vaccination with an altered peptide ligand of MBP 83–99 led to disease exacerbation via crossreactivity with the unmodified MBP 83–99 peptide [39]. DR2a- and DR2b-restricted autoreactive T cell clones can be activated by mature dendritic cells in the absence of exogenous antigen [72]. An encephalitogenic, MBP 83–99-specific TCR derived from an MS patient recognizes MBP 83–99 in the context of DR2b and EBV 627–641 in the context of DR2a [57]. Clonally expanded T cell clones from the CSF of an MS patient in relapse recognize multiple viral and self-peptides in the context of DR2a and DR2b [58,59]. Disease susceptibility in humanized mice expressing DR2b and an MBP-specific TCR is reduced when DR2a is also expressed as transgene, indicating that DR2a can epistatically modify the effects of DR2b [77]. DRB1*15:01 and DRB5*01:01 are the two genes with highest differential expression in normal appearing gray matter of the brains of MS patients [66]. Spontaneous activation and autoproliferation are increased in DR15+ individuals and supported by self-peptides from both DR2a and DR2b [73]. DR2a- and DR2b-derived HLA-DR-SPs are presented by the two alleles and can be recognized by T cells with low avidity, DR2a and DR2b serve as antigen and antigen-presenting molecule [75]. The immunodominant RASGRP2 peptide 78–87 is recognized by high avidity T cell clones in the context of both DR2b and DR2a. These clones also recognize HLA-DR-derived self-peptides and foreign peptides from the environmental ‘risk factors’ EBV and Akkermansia [75]. to low antigen avidity. However, crossrestriction had also been observed with TCC specific for Mycobacterium tuberculosis [60], herpes simplex virus 2 [61], and HIV [62] and we reported TCC that had been isolated from the brain of a patient with progressive multifocal leukoencephalopathy due to brain infection with JCV [63]. In the latter case, multiple TCC with specificity for different peptides of the JCV major capsid protein VP1 were either restricted by DR2a, DR2b, both alleles, or even both together with HLA-DQw6 [63]. Interestingly, crossrestriction of these clones was related to higher antigen avidity, which is biologically meaningful, in that a TCC can use more than one HLA class II restriction element for antigen recognition in an organ like the brain with very limited HLA class II expressioni. Finally, a recent study demonstrated antibody crossreactivity against a novel MS autoantigen, anoctamin-2, and an EBV peptide, particularly in HLA-DR15+ individuals, and positivity for these antibodies is associated with a strongly increased risk in a very large patient cohort, demonstrating that molecular mimicry between these antigens indeed is pathogenetically important [64]. It will be interesting to examine if both CD4+ T cells restricted by either DR2a and/or DR2b play a role in this process as well. One possibility of how a specific HLA class II molecule could be involved in an AID is by ectopic or aberrant expression on tissue-specific cells and not only on professional APCs. This could lead to presentation of a tissue-derived autoantigen to autoreactive CD4+ T cells within the organ, which in turn could start an inflammatory response and cell damage of the HLA class II-expressing cells 790 Trends in Genetics, September 2021, Vol. 37, No. 9 Trends in Genetics OPEN ACCESS MBP 83-99 ENPVVHFFKNIVTPRTP ENPVVHFFKNIVTPRTP F Y L M Q V I M R K L V I F Y I I L V M F Figure 2. The immunodominant myelin basic protein (MBP) 83–99 peptide is used as an example to illustrate how one peptide can be presented by both DR15 alleles DR2a and DR2b. The preferred anchor amino acids that mediate peptide binding are shown underneath the MBP 83–99 sequence. As can be seen, DR2a prefers aromatic (large) amino acids in the first binding pocket, aliphatic amino acids in the middle pocket, and a positively charged arginine or lysine in the third pocket, while it is aliphatic, aromatic, and aliphatic again in the DR2b binding groove. In two of three main DRA*/DRB5*01:01 DRA*/DRB1*15:01 HLA-binding anchor amino acids, the DR2a DR2b two DR15 molecules are similar and, as shown above, MBP 83–99 contains Trends in Genetics such amino acids in the correct spacing. The conclusion is that the peptides that bind to the two DR 15 alleles share certain similarities (Art credit: Katie Vicari). in the tissue. That this is probably relevant, at least for some AIDs, was first demonstrated for thyroid epithelial cells in the thyroid gland affected by autoimmune inflammation [4]. Aberrant HLA class II expression can even be found in the brain and on oligodendrocytes [6], one of the major central nervous system (CNS)-specific glial cell types, which forms myelin, under inflammatory conditions. Furthermore, studies of DR2a and DR2b expression by different immune cells, but also in the thymus and the brain of MS patients, demonstrated that both alleles are always coexpressed, although at varied levels in different cell types [65]. A recent study of gene expression in normal-appearing cortical gray matter in MS patients surprisingly demonstrated that the most strongly upregulated genes are DR2b and DR2a, even in the absence of overt signs of inflammation or damage [66]. Most likely, microglial cells are responsible for these changes. At present, it is not clear what causes the upregulation of HLA-DR15 expression in seemingly unaffected brain tissue, but it could mean that aberrant HLA-DR15 expression due to as yet unknown factors functions as a local vulnerability factor and contributes to starting brain inflammation. Injury of the facial nerve with subsequent inflammation of its nucleus inside the brain [67], traumata, such as concussion of the head during adolescence [68], and death by ablation of oligodendrocytes [69] can be inciting insults. The potential importance of the latter observations (i.e., HLA class II expression by microglia in cortical gray matter) as well as the role of MHC class II expression by injured/dying oligodendrocytes for T cell infiltration and inflammation should be addressed in more detail in the future, for example, by conditionally ablating MHC class II in these cell types. While speculative at the moment, these data could reignite the debate of whether MS develops from outside-in (i.e., from changes of the peripheral immune systems into the brain) or the reverse, inside-out. However, as in most of the earlier observations, both DR15 alleles appear to be involved. Besides aberrant expression in the target tissue, there are other ways how a disease-related HLA class II molecule could contribute to autoimmunity. The most obvious function is via preferential presentation of self-peptides (SPs) that stimulate autoreactive CD4+ T cells and foster their migration to the brain. Even before this activation step can occur in the peripheral immune system with naïve or memory T cells, the spectrum of SPs that is presented to developing T cells in the thymus could select a T cell repertoire that is prone to show autoreactivity against a specific Trends in Genetics, September 2021, Vol. 37, No. 9 791 Trends in Genetics OPEN ACCESS self-tissue. If one wanted to separate the steps that lead to organ-specific autoimmunity, they are (Figure 3): (i) The shaping of a T cell repertoire in the thymus by SPs by disease-associated HLA class II molecules and subsequent egress of such T cells into the peripheral immune system (Figure 3A). Only T cells that recognize SPs, which are presented by thymic APCs, with low to intermediate avidity are positively selected to become naïve T cells in the peripheral immune system. Conversely, T cells recognizing SPs with high avidity, which are potentially dangerous and could more easily cause autoimmune reactions, are destroyed (negative selection). (ii) Peripheral blood naïve CD4+ T cells are maintained for long periods of time, potentially life-long, by homeostatic proliferation and the two most important signals are certain cytokines, such as interleukin (IL)-7, IL-2, and IL-15, but again the low avidity interactions of the TCRs of naïve T cells with SPs presented in the context of autologous HLA class II molecules [70] (Figure 3B). (A) Thymus APC Naïve T cells DRα α2 β2 DRβ Memory T cells (B) α1 β1 Peptide TCR α (C) Homeostasis (D) Activation Re-activation HLA-peptide IL-7, IL-2, IL-15 β Foreign antigen Self-antigen e.g. RASGRP2 Infection e.g. with EBV, Akkermansia, influenza CD4+ T cell Antigen concentration μM nM–fM TCR avidity TCR-autoantigens/foreign peptides Antigen concentration μM TCR avidity TCR-self peptides/autoantigens Death by neglect Positive selection Negative selection Foreign antigens, viruses Homeostatic proliferation, maintenance Full activation; protective immune responses Trends in Genetics Figure 3. Scenario of how autoreactive T cells may be selected, maintained, expanded, and lead to tissue damage in multiple sclerosis (MS). (A) Positive selection of CD4+ T lymphocytes occurs in the thymus, where self-peptides (SPs) are presented by human leukocyte antigen (HLA)-class II molecules (top). Only T cells that recognize these self-peptides including HLA-DR-SPs at relatively high concentration, that is, with low to intermediate functional avidity (the blue zone in the bottom part of Figure 1A), are positively selected and released into the peripheral immune system as mature naïve CD4+ T cells (B). Low avidity recognition of HLA-DR with SP complexes (blue zone of the graph) in combination with certain cytokines assures homeostatic maintenance of these naïve T cells (B). Activation by peptides from foreign agents such as Epstein-Barr virus (EBV), Akkermansia, influenza, and others can lead to full activation and conversion to memory T cells (C). The latter peptides are recognized at much lower concentration or with higher functional avidity (see red zone in graph). Repeated reactivation of potentially autoreactive T cells by foreign agents, but also by autoantigens or self-peptides that are pathogenetically relevant and recognized with high avidity, leads to migration of autoreactive T cells to the brain in the case of MS and contributes to tissue damage (D) (Art credit: Katie Vicari). Abbreviations: APC, antigen-presenting cell; RASGRP2, RAS guanyl-releasing protein 2; TCR, T cell receptor. 792 Trends in Genetics, September 2021, Vol. 37, No. 9 Trends in Genetics OPEN ACCESS (iii) When naïve T cells recognize an antigen with higher avidity (e.g., a viral peptide) they get fully activated and differentiate into either effector or central memory T cells. This activation step involves crossreactivity, that is, recognition of a foreign peptide with higher avidity, but again presented by self-HLA class II (Figure 3C). Memory cells (e.g., after mumps or tetanus vaccination) are again maintained by homeostatic proliferation involving HLA-DR/SPs and cytokines over prolonged periods of time. During activation, particularly if it occurs repetitively, the activation threshold decreases and the sensitivity to activation increases. Different to antibodies, which undergo affinity maturation by sequence changes, the TCRs that T cells employ for antigen recognition do not change; however, the signaling machinery of the respective T cell adapts and hence functional antigen avidity increases as well. (iv) The final step, activation of T cells and homing to the target organ, the brain (Figure 3D), is least understood. The acquisition of receptors and molecules that are necessary for adhering to the blood–brain barrier, crossing it, and moving in the CNS tissue or CSF are one important aspect [71]. The recognition of target autoantigen and finally the induction of an inflammatory response within the brain can then lead to MS. Recent studies examined the involvement of DR2a and DR2b in this cascade and the most important findings will briefly be summarized. Following earlier observations that mature dendritic cells can activate myelin-specific TCC in the absence of exogenous antigen [72] Mohme et al. demonstrated that peripheral blood CD4+ T cells from HLA-DR15+ MS patients show increased spontaneous proliferation, also referred to as autoproliferation, without an antigenic stimulus and that DR2a- and DR2b-derived SPs are probably involved in this process [73]. These data suggested that homeostatic maintenance mechanisms and proliferation of T cells are easily activated and that the MS-associated HLA-DR15 haplotype participates in this phenomenon. The increased autoproliferation involves not only autoreactive CD4+ T cells, but also proinflammatory B cells and cognate interaction between the TCRs and CD4 molecules on the T cell side and HLA-DR15 or generally HLA-DR and likely SPs on the B cell side [74]. The latter study identified a novel target autoantigen, RAS guanyl-releasing protein 2 (RASGRP2), which is expressed by activated B cells and cortical neurons in the brain [74], and furthermore disclosed that the autoproliferating CD4+ T cells are enriched for cells that are also found in inflammatory brain lesions. At this stage, it was not yet clear how the two HLA-DR15 alleles and DR-derived SPs (HLA-DR-SPs) might contribute, although Mohme et al. already provided hints that they participated [73]. This gap has recently been filled by Wang et al. after studying the immunopeptidomes, that is, peptides naturally presented by HLA-DR of B cells and specifically by the two HLA-DR15 molecules DR2a and DR2b [75] (Figure 4A). Surprisingly, a large fraction of peptides that are presented by the two alleles are derived from the HLA-DR15 molecules themselves, which means the MS-associated HLA-DR molecules not only serve as antigenpresenting molecules, but also as a source of antigens. HLA-DR-SPs eluted from DR2a are primarily derived from the HLA-DRβ chain of DRB1*15:01, while DR2b immunopeptidomes are enriched for peptides from the nonpolymorphic HLA-DRα chain (Figure 4B) [75]. Both polymorphic (i.e., from either HLA-DRB1*15:01 or -DRB5*01:01) and nonpolymorphic (i.e., from HLA-DRA1* or a nonpolymorphic sequence of HLA-DRB1*/-DRB5*) have been found (Figure 4B). While the relative abundance of these HLA-DR-SPs was lower on thymic APCs compared with peripheral B cells, their presence in the T cell selection compartment indicates their involvement in positive selection of the T cell repertoire in HLA-DR15+ individuals [see point (i) in earlier list, Figure 3A]. Further, HLA-DR-SPs could weakly activate peripheral blood memory CD4+ T cells, suggesting a role in peripheral homeostasis/maintenance [list point (ii), Figure 3B,C]. Using an unbiased antigen discovery approach, the authors further showed that peptides from two environmental agents, EBV and the gut microbiota Akkermansia muciniphila, which have been associated with MS [76], can fully activate HLA-DR-SP-specific T cells [list point (iii), Figure 3C]. The very same Trends in Genetics, September 2021, Vol. 37, No. 9 793 Trends in Genetics OPEN ACCESS (A) Analysis of HLA-DR-SPs Primary B cells Sequencing by tandem mass spectrometry Relative intensity Surface HLA-DR/peptide complexes, peptide elution Peptides 100 y7 y5 50 y1 0 ADSGEGDFLAEGGGVR b -H2O y6 * 4 y8 b1* * b2* b3 -H2O a1* a * 2 200 400 600 y9 y11 y10 y12 y13 800 1,000 1,200 m/z (B) Source of HLA-DR-SPs presented by DR2a and DR2b 70 DR2a DR2b Epitope sources α1 β1 α1 β1 α2 β2 α2 β2 85 N- -C 72 86 184 199 57 70 72 86 184 199 N- -C N- Antigen-presenting molecules -C DR2a DR2b (C) Summary of contribution of DR2a and DR2b to MS APC APC Peptides DR2b α2 β2 α2 β2 α1 β1 α β DR α/β α1 β1 EBV DR2a Akkermansia TCR α β TCC14 RASGRP2 TCR TCC14 Trends in Genetics Figure 4. Experimental strategy pursued by Wang et al. [75], to demonstrate the complex interactions of DR2a and DR2b and how they can contribute to multiple sclerosis (MS). Human leukocyte antigen (HLA)-DR2a and -DR2b molecules with bound peptides were specifically precipitated from the surface of primary B cells (A), bound self-peptides (SPs) released, and were sequenced by tandem mass spectrometry. These experiments demonstrated that a large fraction of the two HLADR15 molecules on B cells are loaded with SPs derived from these molecules themselves (B). The specific HLA-DR alpha and -DR beta peptides and their position in the sequence of these molecules are shown on the right. Two of the five HLA-DR SPs are polymorphic and specific for HLA-DRB1*15:01 (i.e., 57–70 and 72–86), one for HLADRB5*01:01 (i.e., 57–70), and two (i.e., HLADRB1*/-B5* 184–190 and HLA-DRα 70–85) are shared by many HLA-DRB1*/-B5* alleles or part of the nonpolymorphic HLADRα chain. (C) Summarizes that the two MS-associated HLA-DR15 alleles serve as: (i) antigen-presenting molecules, (ii) as source of peptides, and (iii) autoreactive CD4+ T cells and their single T cell receptor (TCR) molecules can recognize HLA-DR-SPs and peptides derived from disease-associated pathogens [Epstein-Barr virus (EBV), Akkermansia] and pathogenic SPs in the context of both DR15 molecules, that is, these TCRs are both crossreactive (they respond to multiple peptides) and crossrestricted (they can recognize the same peptides in the context of both DR molecules) (Art credit: Katie Vicari). Abbreviations: APC, antigen-presenting cell; RASGRP2, RAS guanyl-releasing protein 2. 794 Trends in Genetics, September 2021, Vol. 37, No. 9 Trends in Genetics OPEN ACCESS T cells and TCRs that can recognize HLA-DR-SPs with low avidity and EBV and Akkermansia peptides with high avidity, also responded strongly to RASGRP2, the autoantigen expressed by proinflammatory B cells and neurons in the brain [74,75]. Whether activation by the EBV and Akkermansia peptides and/or RASGRP2 are a prerequisite for and involved in list point (iv) [i.e., homing to brain lesions (Figure 3D)] is currently not clear, but is suggested by the earlier findings [74]. As a final surprising observation, HLA-DR2a and -DR2b not only serve as presenting molecules and antigen source, but CD4+ T cells and their TCRs with crossreactivity for HLA-DR-SPs, for EBV and Akkermansia peptides, and for RASGRP2 can ‘see’ these antigens on both MS-associated HLA-DR15 molecules; that is, they are also crossrestricted (Figure 4C and Table 2). With respect to the relative importance of DR2a versus DR2b, all the earlier data indicate that both play a functional role in MS pathogenesis. However, one study argues for epistatic protective influences of DR2a to balance out the MS risk of DR2b [77]. These data have been generated in a humanized mouse model expressing an MBP 83–99-specific TCR and DR2b [45,77]. Introduction of DR2a as second transgenic HLA-DR allele led to lower EAE incidence. The authors argue that the tight LD of the two HLA-DR15 alleles may be due to DR2a balancing out the MS risk of DR2b. Since MS is a relatively ‘young’ disease, does not substantially shorten the lifespan of affected individuals, and therefore also does not compromise producing progeny, it should not be MS that exerts evolutionary pressure and causes the tight LD of DR2a and DR2b, but more likely fitness to cope with infectious pathogens. Whether EBV, Akkermansia, or other microorganisms are involved in this process is not clear. In the mouse model, we do not know the relative abundance of HLA-DR-SPs in the immunopeptidomes, whether the expression levels of the two alleles are comparable with humans and whether B cells are involved, and how the artificial situation with a highly abundant single TCR may relate to the reduced disease incidence. Evidence for a stronger contribution of DR2b or the DRB1*15:01 allele comes from HLA typing of African American MS patients [78]. MS was found in patients expressing a rare haplotype carrying HLA-DRB1*15:03 and no HLA-DRB5 allele. These individuals were more likely to develop progressive MS, indicating an attenuating effect of HLA-DRB5. When weighing all the available data, they indicate that both HLA-DR15 alleles are involved, either alone or jointly, in the pathogenesis of MS. Concluding remarks These data demonstrate the remarkable complexity of how two MS-associated and very tightly linked HLA class II molecules and peptides derived from these can contribute to disease. Whether similar interactions between two HLA class II or even class I alleles [30,33] in the same or different haplotype play a role in other AIDs will be interesting to examine (see Outstanding questions). With respect to studying the association of one or several HLA molecules with a disease, one should not only rely on GWAS studies, at least not when certain HLA alleles are not identifiable by SNP-based typing platforms with insufficient coverage in that region and particularly not when it is well-known from conventional HLA typing that two alleles, like HLADRB1*15:01 and -DRB5*01:01 in MS, always occur together and functional studies suggest involvement of both. In MS, both HLA-DR15 alleles should be considered as pathogenetic factors and also for the development of antigen-specific tolerance approaches. Outstanding questions When searching for disease-associated HLA alleles, which methodologies could be developed that allow typing for HLA alleles which would be missed because the spacing of SNPs on conventional genotyping platforms does not allow the assignment of specific alleles? Should functional data and/or knowledge about frequent HLA haplotypes be incorporated in such algorithms? Following the identification of certain HLA genes that are associated with defined conditions, for example, autoimmune or infectious diseases, understanding how they contribute at the functional level is paramount. How can one best address this at the level of individual HLA molecules and at the same time assure that complex interactions are not missed? Could small molecules or peptidic compounds that ‘fill’ the binding grooves of the MS-associated HLADR15 alleles be developed into a new class of therapies? How do the complex functional interactions between the two MSassociated HLA-DR15 alleles, which are outlined here, inform similar studies in other autoimmune diseases, for which associations with single or several HLA class II alleles are well known? Examples are rheumatoid arthritis, type I diabetes, and Goodpasture syndrome. Even in diseases that bear strong association with a single HLA class II molecule (e.g., narcolepsy and the heterodimer of DQA1*01:02/ DQB*06:02), other alleles from the extended HLA-DR15 haplotype, such as the two MS-associated HLA-DR15 alleles, could play a role. Acknowledgments The authors were supported by European Research Grant Advanced Grant ERC-2013- ADG 340733 to R.M., the Clinical Research Priority Program Precision-MS (CRPPMS) of the University of Zurich (UZH) to R.M. and M.S., and T.O. has received grant support from the Swedish Research council, The Swedish Brain Foundation, and Knut and Alice Wallenberg foundation. We would like to thank Lars Fugger, Weatherall Institute of Molecular Medicine, Oxford University, Oxford, UK, Trends in Genetics, September 2021, Vol. 37, No. 9 795 Trends in Genetics OPEN ACCESS for discussion and insightful advice, and Karolin Léger, Neuroimmunology and MS Research, University Zurich, for her help in editing the manuscript. Declaration of interests T.O. has received honoraria for advisory boards/lectures and unrestricted MS research grants from Biogen, Novartis, Sanofi, Merck, and Roche. R.M. received unrestricted MS research grants from Biogen, Novartis, Roche, and Third Rock Ventures, personal compensation for lecture or advisory board functions from Biogen, Merck, Novartis, Roche, Sanofi Aventis, Teva, CellProtect, Neuway, and Third Rock Ventures. He is a co-founder and co-owner of Cellerys, a startup company of the University of Zurich. He is co-inventor and patent holder on patents related to antigen-specific tolerization, treatment/vaccination of PML (together with M.S.), and the use of daclizumab as a treatment for multiple sclerosis. T.E. has no interests to declare. Resources i www.proteinatlas.org ii www.ebi.ac.uk/ipd/imgt/hla/; http://allelefrequencies.net/hla.asp iii http://hla.alleles.org/nomenclature/index.html References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Hewitt, E.W. (2003) The MHC class I antigen presentation pathway: strategies for viral immune evasion. Immunology 110, 163–169 Boegel, S. et al. (2018) HLA and proteasome expression body map. BMC Med. Genet. 11, 36 Ting, J.P. and Trowsdale, J. (2002) Genetic control of MHC class II expression. Cell 109, S21–S33 Pujol-Borrell, R. et al. (1983) Lectin-induced expression of DR antigen on human cultured follicular thyroid cells. Nature 304, 71–73 Fontana, A. et al. (1984) Astrocytes present myelin basic protein to encephalitogenic T-cell lines. Nature 307, 273–276 Falcao, A.M. et al. (2018) Disease-specific oligodendrocyte lineage cells arise in multiple sclerosis. Nat. Med. 24, 1837–1844 Leen, G. et al. (2021) The HLA diversity of the Anthony Nolan register. HLA 97, 15–29 Dendrou, C.A. et al. (2018) HLA variation and disease. Nat. Rev. Immunol. 18, 325–339 Kröger, N. (2015) Allogene Stammzelltherapie - Grundlagen, Indikationen und Perspektiven (4th edn), UNI-MED Science Gorer, P.A. (1936) The detection of antigenic differences in mouse erytlirocytes by the employment of immune sera. Brit. J. Exp. Pathol. 17, 42–50 Dausset, J. and Brecy, H. (1957) Identical nature of the leucocyte antigens detectable in monozygotic twins by means of immune iso-leuco-agglutinins. Nature 180, 1430 Snell, G.D. and Higgins, G.F. (1951) Alleles at the histocompatibility-2 locus in the mouse as determined by tumor transplantation. Genetics 36, 306–310 Martin, W.J. et al. (1970) Histocompatibility type and immune responsiveness in random bred Hartley strain guinea pigs. J. Exp. Med. 132, 1259–1266 Zinkernagel, R.M. and Doherty, P.C. (1975) H-2 compatability requirement for T-cell-mediated lysis of target cells infected with lymphocytic choriomeningitis virus. Different cytotoxic T-cell specificities are associated with structures coded for in H-2K or H-2D. J. Exp. Med. 141, 1427–1436 Nepom, G.T. and Erlich, H. (1991) MHC class-II molecules and autoimmunity. Annu. Rev. Immunol. 9, 493–525 Lee, B.S. et al. (1987) HLA-DR2 subtypesform an additional supertypic family of DR beta alleles. Proc. Natl. Acad. Sci. U. S. A. 84, 4591–4595 Kawai, J. et al. (1989) Analysis of gene structure and antigen determinants of DR2 antigens using DR gene transfer into mouse L cells. J. Immunol. 142, 312–317 Batchelor, J.R. et al. (1989) Antigen Society #22 report (DR2). In Immunology of HLA (Dupont, B., ed.), pp. 240–242, Springer-Verlag Segall, M. et al. (1986) DNA restriction fragment length polymorphisms characteristic for Dw subtypes of DR2. Hum. Immunol. 15, 336–343 796 Trends in Genetics, September 2021, Vol. 37, No. 9 20. Tiercy, J.M. et al. (1991) Oligonucleotide typing analysis for the linkage disequilibrium between the polymorphic DRB1 and DRB5 loci in DR2 haplotypes. Tissue Antigens 37, 161–164 21. Jersild, C. et al. (1973) Histocompatibility-linked immune-response determinants in multiple sclerosis. Transplant. Proc. 5, 1791–1796 22. Bertrams, J. et al. (1972) HL-A antigens and multiple sclerosis. Tissue Antigens 2, 405–408 23. Hillert, J. and Olerup, O. (1993) Multiple sclerosis is associated with genes within or close to the HLA-DR-DQ subregion on a normal DR15,DQ6,Dw2 haplotype. Neurology 43, 163–168 24. Tiercy, J.M. et al. (1990) A new HLA-DRB1 allele within the DRw52 supertypic specificity (DRw13-DwHAG): sequencing and direct identification by oligonucleotide typing. Eur. J. Immunol. 20, 237–241 25. Oksenberg, J.R. et al. (2008) The genetics of multiple sclerosis: SNPs to pathways to pathogenesis. Nat. Rev. Genet. 9, 516–526 26. International Multiple Sclerosis Genetics Consortium et al. (2007) Risk alleles for multiple sclerosis identified by a genome-wide study. N. Engl. J. Med. 357, 851–862 27. Dyment, D.A. et al. (2005) Complex interactions among MHC haplotypes in multiple sclerosis: susceptibility and resistance. Hum. Mol. Genet. 14, 2019–2026 28. Lincoln, M.R. et al. (2009) Epistasis among HLA-DRB1, HLADQA1, and HLA-DQB1 loci determines multiple sclerosis susceptibility. Proc. Natl. Acad. Sci. U. S. A. 106, 7542–7547 29. Bertrams, H.J. and Kuwert, E.K. (1976) Association of histocompatibility haplotype HLA-A3-B7 with multiple sclerosis. J. Immunol. 117, 1906–1912 30. Harbo, H.F. et al. (2004) Genes in the HLA class I region may contribute to the HLA class II-associated genetic susceptibility to multiple sclerosis. Tissue Antigens 63, 237–247 31. International Multiple Sclerosis Genetics Consortium (2019) Multiple sclerosis genomic map implicates peripheral immune cells and microglia in susceptibility. Science 365, eaav7188 32. Moutsianas, L. et al. (2015) Class II HLA interactions modulate genetic risk for multiple sclerosis. Nat. Genet. 47, 1107–1113 33. Lundtoft, C. et al. (2020) Function of multiple sclerosis-protective HLA class I alleles revealed by genome-wide protein-quantitative trait loci mapping of interferon signalling. PLoS Genet. 16, e1009199 34. Ballas, M. et al. (1990) Mapping of an HLA-DRw52-associated determinant on DR beta 1 molecules. Tissue Antigens 36, 187–193 35. Fogdell, A. et al. (1995) The multiple sclerosis- and narcolepsy-associated HLA class II haplotype includes the DRB5*0101 allele. Tissue Antigens 46, 333–336 36. Martin, R. et al. (1990) Fine specificity and HLA restriction of myelin basic protein-specific cytotoxic T cell lines from multiple sclerosis patients and healthy individuals. J. Immunol. 145, 540–548 Trends in Genetics 37. Pette, M. et al. (1990) Myelin autoreactivity in multiple sclerosis: recognition of myelin basic protein in the context of HLA-DR2 products by T lymphocytes of multiple-sclerosis patients and healthy donors. Proc. Natl. Acad. Sci. U. S. A. 87, 7968–7972 38. Jaraquemada, D. et al. (1990) HLA-DR2a is the dominant restriction molecule for the cytotoxic T cell response to myelin basic protein in DR2Dw2 individuals. J. Immunol. 145, 2880–2885 39. Bielekova, B. et al. (2000) Encephalitogenic potential of the myelin basic protein peptide (amino acids 83-99) in multiple sclerosis: results of a phase II clinical trial with an altered peptide ligand. Nat. Med. 6, 1167–1175 40. Olsson, T. et al. (1990) Autoreactive T lymphocytes in multiple sclerosis determined by antigen-induced secretion of interferongamma. J. Clin. Invest. 86, 981–985 41. Ota, K. et al. (1990) T-cell recognition of an immunodominant myelin basic protein epitope in multiple sclerosis. Nature 346, 183–187 42. Richert, J.R. et al. (1989) Human cytotoxic T-cell recognition of a synthetic peptide of myelin basic protein. Ann. Neurol. 26, 342–346 43. Wucherpfennig, K.W. and Strominger, J.L. (1995) Molecular mimicry in T cell-mediated autoimmunity: viral peptides activate human T cell clones specific for myelin basic protein. Cell 80, 695–705 44. Vergelli, M. et al. (1997) Human autoreactive CD4+ T cell clones use perforin- or Fas/Fas ligand-mediated pathways for target cell lysis. J. Immunol. 158, 2756–2761 45. Madsen, L.S. et al. (1999) A humanized model for multiple sclerosis using HLA-DR2 and a human T-cell receptor. Nat. Genet. 23, 343–347 46. Quandt, J.A. et al. (2012) Myelin basic protein-specific TCR/ HLA-DRB5*01:01 transgenic mice support the etiologic role of DRB5*01:01 in multiple sclerosis. J. Immunol. 189, 2897–2908 47. Vergelli, M. et al. (1997) T cell response to myelin basic protein in the context of the multiple sclerosis-associated HLA-DR15 haplotype: peptide binding, immunodominance and effector functions of T cells. J. Neuroimmunol. 77, 195–203 48. Vogt, A.B. et al. (1994) Ligand motifs of HLA-DRB5*0101 and DRB1*1501 molecules delineated from self-peptides. J. Immunol. 153, 1665–1673 49. Scholz, E.M. et al. (2017) Human leukocyte antigen (HLA)DRB1*15:01 and HLA-DRB5*01:01 present complementary peptide repertoires. Front. Immunol. 8, 984 50. Li, Y. et al. (2005) Structure of a human autoimmune TCR bound to a myelin basic protein self-peptide and a multiple sclerosis-associated MHC class II molecule. EMBO J. 24, 2968–2979 51. Li, Y. et al. (2000) Structural basis for the binding of an immunodominant peptide from myelin basic protein in different registers by two HLA-DR2 proteins. J. Mol. Biol. 304, 177–188 52. Smith, K.J. et al. (1998) Crystal structure of HLA-DR2 (DRA*0101, DRB1*1501) complexed with a peptide from human myelin basic protein. J. Exp. Med. 188, 1511–1520 53. Tang, W.M. et al. (1997) The association of HLA-DR15 and intermediate uveitis. Am J. Ophthalmol. 123, 70–75 54. Ooi, J.D. et al. (2017) Dominant protection from HLA-linked autoimmunity by antigen-specific regulatory T cells. Nature 545, 243–247 55. Sundqvist, E. et al. (2014) JC polyomavirus infection is strongly controlled by human leucocyte antigen class II variants. PLoS Pathog. 10, e1004084 56. Zdimerova, H. et al. (2020) Attenuated immune control of Epstein-Barr virus in humanized mice is associated with the multiple sclerosis risk factor HLA-DR15. Eur. J. Immunol. 51, 64–75 57. Lang, H.L. et al. (2002) A functional and structural basis for TCR cross-reactivity in multiple sclerosis. Nat. Immunol. 3, 940–943 58. Sospedra, M. et al. (2005) Recognition of conserved amino acid motifs of common viruses and its role in autoimmunity. PLoS Pathog. 1, e41 59. Sospedra, M. et al. (2006) Redundancy in antigen-presenting function of the HLA-DR and -DQ molecules in the multiple OPEN ACCESS 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. sclerosis-associated HLA-DR2 haplotype. J. Immunol. 176, 1951–1961 Caccamo, N. et al. (2003) Cytokine profile, HLA restriction and TCR sequence analysis of human CD4+ T clones specific for an immunodominant epitope of Mycobacterium tuberculosis 16-kDa protein. Clin. Exp. Immunol. 133, 260–266 Doherty, D.G. et al. (1998) Structural basis of specificity and degeneracy of T cell recognition: pluriallelic restriction of T cell responses to a peptide antigen involves both specific and promiscuous interactions between the T cell receptor, peptide, and HLA-DR. J. Immunol. 161, 3527–3535 Galperin, M. et al. (2018) CD4(+) T cell-mediated HLA class II cross-restriction in HIV controllers. Sci. Immunol. 3, eaat0687 Yousef, S. et al. (2012) TCR bias and HLA cross-restriction are strategies of human brain-infiltrating JC virus-specific CD4+ T cells during viral infection. J. Immunol. 189, 3618–3630 Tengvall, K. et al. (2019) Molecular mimicry between Anoctamin 2 and Epstein-Barr virus nuclear antigen 1 associates with multiple sclerosis risk. Proc. Natl. Acad. Sci. U. S. A. 116, 16955–16960 Prat, E. et al. (2005) HLA-DRB5*0101 and -DRB1*1501 expression in the multiple sclerosis-associated HLA-DR15 haplotype. J. Neuroimmunol. 167, 108–119 Enz, L.S. et al. (2020) Increased HLA-DR expression and cortical demyelination in MS links with HLA-DR15. Neurol. Neuroimmunol. Neuroinflamm. 7, e656 Maehlen, J. et al. (1989) Local enhancement of major histocompatibility complex (MHC) class I and II expression and cell infiltration in experimental allergic encephalomyelitis around axotomized motor neurons. J. Neuroimmunol. 23, 125–132 Montgomery, S. et al. (2017) Concussion in adolescence and risk of multiple sclerosis. Ann. Neurol. 82, 554–561 Traka, M. et al. (2016) Oligodendrocyte death results in immunemediated CNS demyelination. Nat. Neurosci. 19, 65–74 Surh, C.D. and Sprent, J. (2005) Regulation of mature T cell homeostasis. Semin. Immunol. 17, 183–191 Ransohoff, R.M. et al. (2003) Three or more routes for leukocyte migration into the central nervous system. Nat. Rev. Immunol. 3, 569–581 Kondo, T. et al. (2001) Dendritic cells signal T cells in the absence of exogenous antigen. Nat. Immunol. 2, 932–938 Mohme, M. et al. (2013) HLA-DR15-derived self-peptides are involved in increased autologous T cell proliferation in multiple sclerosis. Brain 136, 1783–1798 Jelcic, I. et al. (2018) Memory B cells activate brain-homing, autoreactive CD4(+) T cells in multiple sclerosis. Cell 175, 85–100 Wang, J. et al. (2020) HLA-DR15 molecules jointly shape an autoreactive T cell repertoire in multiple sclerosis. Cell 183, 1264–1281 Cekanaviciute, E. et al. (2017) Gut bacteria from multiple sclerosis patients modulate human T cells and exacerbate symptoms in mouse models. Proc. Natl. Acad. Sci. U. S. A. 114, 10713–10718 Gregersen, J.W. et al. (2006) Functional epistasis on a common MHC haplotype associated with multiple sclerosis. Nature 443, 574–577 Caillier, S.J. et al. (2008) Uncoupling the roles of HLA-DRB1 and HLA-DRB5 genes in multiple sclerosis. J. Immunol. 181, 5473–5480 Wu, S. et al. (1986) Polymorphism of human Ia antigens generated by reciprocal intergenic exchange between two DR beta loci. Nature 324, 676–679 Gregersen, P.K. et al. (1986) Molecular diversity of HLA-DR4 haplotypes. Proc. Natl. Acad. Sci. U. S. A. 83, 2642–2646 Tiercy, J.M. et al. (1989) DNA typing of DRw6 subtypes: correlation with DRB1 and DRB3 allelic sequences by hybridization with oligonucleotide probes. Hum. Immunol. 24, 1–14 Obata, F. et al. (1990) Sequence analysis and HLA-DR genotyping of a novel HLA-DRw14 allele. Immunogenetics 32, 313–320 Hemmer, B. et al. (1996) Cytokine phenotype of human autoreactive T cell clones specific for the immunodominant myelin basic protein peptide (83-99). J. Neurosci. Res. 45, 852–862 Trends in Genetics, September 2021, Vol. 37, No. 9 797