HLA Class I Histocompatibility Antigen, Alpha Chain E
Historically, the gene cluster on chromosome 6 that harbors the human leukocyte antigen (HLA) system was termed major histocompatibility complex (MHC), because their discovery was linked to experiments regarding tissue transplantations. It became later apparent that the HLA system is comprised of several molecule classes of which the HLA class I and HLA class II molecules are surface-expressed glycoproteins that present peptides of self or pathogenic origin to different T cell subtypes. This procedure termed MHC restriction sheds new light on T cell–mediated immunity and also elucidates the obstacle of organ rejection. The extensive influence of this discovery was later appreciated by the Nobel Prize dedicated to Peter Doherty and Rolf Zinkernagel for their work on MHC restriction. This peptide ligand-mediated interaction of HLA class I molecules (HLA-A, -B, or -C) with CD8+ T cells preceded a complex process within the cell providing selected peptides from the cytosolic compartment. During pathologic conditions, such as viral infections, peptides of pathogenic origin can be presented by HLA molecules, thus enabling the immune system to surveil the health status of the cell. In 1987 (Bjorkman et al. 1987), the first crystal structure of an HLA molecule was solved and highlighted the impact of distinct amino acids (AA) within the peptide binding region (PBR) that are in contact with AAs of the bound peptide. In order to present an abundance of peptide-antigens to T cells, alterations in the PBR through AA exchanges (mismatches) are crucial because these changes in conformation of this overall peptide-HLA structure would directly affect thymic T-cell selection and activation of T-cells. Therefore, to achieve maximum diversity of the presentable antigen repertoire, HLA genes are among the most polymorphic of the human genes. These polymorphisms are mostly located within Exon 2 and 3, which encode the PBR. In 1988 (Koller et al. 1988), a novel gene in the HLA gene cluster between the HLA-A and HLA-C loci was identified and designated as HLA-E. In contrast to the classical HLA class I molecules HLA-A, -B, or -C (HLA class Ia), HLA-E exhibits only a few polymorphisms, accounting for nine coding allelic variants to date as published in the IMGT/HLA database (http://www.ebi.ac.uk/ipd/imgt/hla/). Subsequently, HLA-E was grouped in the nonclassical HLA class I category (class Ib) together with the less-polymorphic variants HLA-F and HLA-G.
Nonclassical HLA molecules differ not only in their limited polymorphisms from class Ia molecules, but also in their tissue distribution while exhibiting specific functions. HLA-F is not well characterized, but it is thought to be mainly expressed intracellularly, rarely reaching the cell surface (Lepin et al. 2000). Its expression seems to be restricted to the bladder, skin, and liver as well as immune cells; however, receptor interactions with HLA-F are not well understood. HLA-F potentially interacts with ILT-2 and ILT-4 (Lepin et al. 2000), although the function of this interaction remains elusive. The expression of HLA-G is even more confined to specific immune privileged tissues such as the placenta or the eye. Here, HLA-G interacts with KIR2DL4 present on NK cells as well as ILT-2 and ILT-4 present on a wide array of immune cells such as T cells, B cells, monocytes, and macrophages. In the placenta, HLA-G primarily acts as an immune modulator, locally inhibiting the immune response in order to prevent the rejection of the fetus. However, based on their function, peptide restriction and selection of HLA-F and HLA-G appear to be a minor factor in adaptive immune responses.
HLA-E, by contrast, is an intermediate molecule, undergoing peptide restriction and pathogen presentation for adaptive immune recognition but also an immune tolerance mediating molecule in innate immunity. HLA-E is expressed in virtually all nucleated cells, however, to varying degrees. First conclusions to its differential function were reached via sequencing of the presented peptide repertoire. HLA-E presents nonameric peptides derived from the signal sequence of other HLA class I molecules (i.e., leader peptides). The strict peptide motif in combination with the conserved nature of the HLA-E heavy chain could be correlated to its conserved function to primarily interact with natural killer (NK) cells. By primary presentation of leader peptides and the interaction with NK cells, HLA-E enables the innate immune system to indirectly monitor the expression of HLA class I molecules and constitutes a barrier for pathogens specifically targeting HLA class Ia molecules as part of their immune evasion strategy against T cell recognition (Braud et al. 1997). However, newer research indicates that HLA-E is not confined to the presentation of such leader peptides but is also able to present peptides of extraordinary length with unrestricted anchor motifs (Celik et al. 2015).
HLA-E and Peptide Presentation
HLA-E is expressed on the cell surface as a heterotrimeric complex consisting of the HLA-E heavy chain (hc), the noncovalently associated β2microglobulin (β2m) and a bound peptide. The structural accessibility of its immunogenic areas is therefore similar to HLA class Ia molecules. The membrane-anchored HLA-E hc consists of three extracellular domains (α1–3), whereby the α1- and α2-domain form the peptide binding region (PBR) and the Ig-like α3 domain acts as the primary anchoring point for β2m as well as acting as a CD8 binding site. In concordance with its structural features, peptide loading also occurs in a similar manner to HLA class Ia molecules (Braud et al. 1998b), utilizing the peptide loading complex (PLC) in the endoplasmic reticulum (ER). Synthesized cotranslationally at the ER membrane, the HLA-E hc is associated with β2m as well as with the chaperone calreticulin and the thiol reductase ERp57 in the ER lumen. Through the subsequent interaction with tapasin (TPN), the PLC is brought into close proximity of TAP (transporter associated with antigen processing). By peptide translocation from the cytosol to the ER via TAP, peptide loading can be facilitated by the PLC in the ER lumen. Once assembled, the stable trimeric HLA-E complex is released and subsequently transported to the cell surface via the trans-Golgi network. The nature of the bound peptide is predefined by the biochemical properties of the PBR that is layered over a β-sheet and framed by two α-helices, building a total of six specificity pockets (pocket A-F) for peptide binding (Saper et al. 1991). Canonical ligands for HLA-E are derived from the conserved leader sequences of HLA class I molecules composed of nine AAs in length (Braud et al. 1997). Resolving the structure of HLA-E*01:01 bound to the nonameric leader peptide of HLA-B8, VMAPRTVLL, binding properties similar to HLA-A2 are observed whereby peptide positions p2 and pΩ are anchored in pocket B and pocket F, respectively. The peptide positions p4, p5, and p8 are exposed to the solvent, whereas p6 and p7 are accommodated by pockets C and E, respectively (PDB: 1MHE). As it is known by now, peptides of extraordinary length (>8–10 AA) can be presented by HLA class I molecules and HLA-E is no exception; however, the function of such noncanonical peptides and their binding properties remains to be elucidated. As no structures are available it is not entirely clear how these peptides are presented, however modeling studies suggest that bulging of the peptide could provide solvent accessible AAs (Kraemer et al. 2015).
HLA-E Receptor Ligand Interactions
Of the described coding HLA-E variants, two alleles are maintained in diverse populations at nearly equal frequencies: HLA-E*01:01 and HLA-E*01:03 (http://www.allelefrequencies.net/). A possible advantage for heterozygous individuals based on this balancing selection suggests that there is an underlying functional difference between both alleles. However, the only difference between HLA-E*01:01 and HLA-E*01:03 is a single AA at position 107 located in the α2 domain, where Arginine (Arg) is substituted by Glycine (Gly). This single polymorphism affects its immune function fundamentally (Kraemer et al. 2015). In contrast to most substitutions in HLA class Ia molecules that are located inside the PBR directly influencing the features of bound peptides, the Arg107Gly substitution in HLA-E is located outside the PBR in a loop connecting two β-strands. The AA exchange does not lead to any profound structural difference when bound to the same leader peptide, however, in comparison to HLA-E*01:01 (Arg107), the Gly107 polymorphism leads to a higher thermal stability of HLA-E*01:03. By extension, this also impacts peptide binding affinities: Utilizing the same leader peptide, HLA-E*01:03 is stabilized by lower peptide concentrations in comparison to HLA-E*01:01 (Strong et al. 2003). These features eventually result in lower levels of surface expression for HLA-E*01:01 thus impacting the half-life of the molecule on the cell surface. These biochemical features combined with the appearance of nearly equal frequencies throughout populations implicate that both alleles are possibly able to execute functions that differ from their role in immune surveillance of HLA class I molecules. Indeed, peptide binding studies with random peptide libraries (Stevens et al. 2001) were able to proof that HLA-E is also capable of binding peptides not derived from leader sequences of other HLA class I molecules, but also peptides derived from stress signals or pathogens. Subsequent research showed that in the absence of HLA class I molecules the presented peptide repertoire shifts between the polymorphic variants HLA-E*01:01 and HLA-E*01:03 (Celik et al. 2015).
HLA-E in Pathophysiological Conditions
Downregulation of HLA class I molecules is not only an immune evasion mechanism for pathogens but also part of the repertoire of many endogenous malignancies, in part to prevent recognition through CD8+ T cells. However, presentation of HLA is needed to sustain the self-signal that inhibits NK cell–mediated cytotoxicity. Oftentimes, effects of HLA class I downregulation are not as severe for HLA-E and thus provide an advantage for survival due to its inhibitory function regarding NK cell–mediated lysis. Novel insights from cancer research suggest that differential expression of HLA-E in various malignancies could contribute to disease progression. Cancer cells in the nonsmall cell lung cancer (NSCLC), for instance, that are HLA class Ia negative and HLA-E positive are associated with a worse outcome for survival (Talebian Yazdi et al. 2016). In acute myeloid leukemia (AML), HLA-E expression on AML blasts exerted an inhibiting effect on immature CD94/NKG2A NK cells. Inhibition of the early NK cell response post HSCT (hematopoietic stem cell transplantation) is likely to interfere with the graft vs leukemia effect, necessary for clearance of remaining malignant cells (Nguyen et al. 2005). The role of soluble HLA molecules in disease progression and its potential use as a biomarker have come into focus in recent years as well. In particular, soluble HLA-E was found to be significantly increased in melanoma (Allard et al. 2011) and neuroblastoma patients (Morandi et al. 2013), potentially being involved in the antitumor immune response. However, HLA-E expression being just one factor in a complex disease entity, different roles of HLA-E and soluble HLA-E as well as its contributing factors remain elusive.
Interaction of HLA-E with CD8+ T cells, of which autoreactive T cells are known to contribute to autoimmunity, also provides a link to autoimmune diseases. Extensive studies on Qa-1, the mouse homologue of HLA-E, showed the importance of nonclassical HLA molecules for regulation of autoreactive T cells. In diabetes type 1, it was found that an immune reaction against HLA-E in T1D patients was facilitated by defective CD8+ T cell recognition of HLA-E presenting a peptide derived from Hsp60, resulting in failure of self-tolerance (Jiang et al. 2010). In Multiple Sclerosis (MS), HLA-E upregulation was found in the white matter of the CNS and the endothelial cells, colocalizing with CD8+ T cells (Durrenberger et al. 2012). MS is a multifactorial autoimmune disease, however, HLA-E expression might be altered in stressed tissues such as MS lesions, leading to a subset of regulatory CD8+ T cells.
HLA-E fulfills a special role in the relationship between the HLA system and the immune system. Typically, HLA molecules are known to be part of the most polymorphic genes in the human genome; however, in contrast to that HLA-E exhibits only a few polymorphisms leading to only two alleles that are effectively present in populations worldwide. The occurrence of such balanced HLA-E expression appears to go hand in hand with its function to primarily interact with NK cells of the innate immune system, where peptide presentation needs to be confined to a very strict binding motif due to the conserved nature of the interacting receptor. Physiologically, HLA-E presents a nonameric peptide derived from the leader sequence of other HLA molecules. Because such leader peptides are vastly similar throughout different HLA alleles, it becomes possible for NK cells to surveil individual cells for their health status. This mechanism acts as a countermeasure against pathogens that evolved to target the HLA system as part of their immune evasion strategy. However, as an HLA class I molecule, HLA-E also constitutes a ligand for CD8+ T cells and newer research indicates that HLA-E is not confined to the presentation of leader peptides. It is quite possible that during special occurrences, HLA-E is able to present peptides of extraordinary length and diverse intracellular origins, including pathogens.