Significant advances in mass-spectroscopy (MS) have made it possible to investigate the cellular immunopeptidome, a large collection of MHC-associated epitopes presented on the surface of healthy, stressed and infected cells. These approaches have hitherto allowed the unambiguous identification of large cohorts of epitope sequences that are restricted to specific MHC class I and II molecules, enhancing our understanding of the quantities, qualities and origins of these peptide populations. Most importantly these analyses provide essential information about the immunopeptidome in responses to pathogens, autoimmunity and cancer, and will hopefully allow for future tailored individual therapies. Protein post-translational modifications (PTM) play a key role in cellular functions, and are essential for both maintaining cellular homeostasis and increasing the diversity of the proteome. A significant proportion of proteins is post-translationally modified, and thus a deeper understanding of the importance of PTM epitopes in immunopeptidomes is essential for a thorough and stringent understanding of these peptide populations. The aim of the present review is to provide a structural insight into the impact of PTM peptides on stability of MHC/peptide complexes, and how these may alter/modulate immune responses.
The immune system makes use of leukocytes to scan all tissues of the organism for infected and/or stressed cells, followed by their potential elimination. To make this possible, Nature has invented a mechanism based on cell surface presentation of intracellularly processed peptides bound to major histocompatibility complexes (MHC). The two classes of MHC molecules, class I (MHC-I) and class II (MHC-II) bind repertoires of endogenously processed peptide antigens, termed immunopeptidomes, that are displayed on the cell surface enabling recognition by CD8+ cytotoxic T cell lymphocytes (CTL), CD4+ T helper cells and NK cells. The overall 3D structures of the extracellular regions of MHC-I and MHC-II molecules are similar, comprising a peptide-binding domain (PBD) and two immunoglobulin-like (Ig) domains that separate the PBD from the membrane. The peculiar structure of the PBD resembles a cradle with its base formed by eight β-strands and its sides formed by two broken α-helices. The processed peptide is presented on the base of this cradle and is partially entwined between the two α-helices. T cell and most NK cell receptors bind on the top of the pMHC, interacting almost always with both the MHC chains and the presented peptides.
This review will mainly focus on structural descriptions of the effects of post-translational modifications (PTM) and chemical modifications in MHC-restricted peptides. Thus, although we provide here below a short description of MHC-I and MHC-II antigen processing and presentation, we refer the readers to other more comprehensive and detailed reviews of these pathways. Proteolysis is the main step in intracellular production of peptides that will be loaded onto MHC molecules. Most MHC-I epitopes usually originate from proteasomal degradation of cytosolic proteins. Intracellular proteins (including pathogen-associated products) are labeled with ubiquitin and degraded by cytosolic proteasome and/or immunoproteasome. The products are thereafter transferred by the transporter associated with antigen processing (TAP) transmembrane proteins into the endoplasmic reticulum (ER), where an array of peptides is selected and tested for binding to nascent MHC-I by the peptide loading complex (PLC), that includes tapasin, calreticulin and ERp57. Finally, the N-termini of peptides are trimmed by ERAP aminopeptidases before and after loading into MHC-I, and the formed pMHC complexes are thereafter transported via the Golgi system to the cell surface where they are exposed to immune cells. It should be noted that besides the conventional MHC-I antigen processing and presentation pathway described here above, in which peptides result from the degradation of elderly proteins (retirees), several studies have demonstrated that a substantial fraction of peptides can be generated significantly more rapidly, and independently from ubiquitylation or proteasome cleavage, including non-canonical translation of defective ribosomal products (DriPs) or short-lived proteins (SLiPs).
Similarly to MHC-I epitopes, a thorough understanding of the molecular processes underlying MHC-II antigen processing and presentation, combined with the discovery of e.g. MHC-II-restricted cancer- or other disease-associated epitopes, enhances significantly our capacity to design novel and/or improve already existing CD8+ and CD4+ T cell-based therapies through specific modifications In contrast to MHC-I antigen processing, most MHC-II-bound peptides are derived from the endosomal proteolytic machinery. It has been previously thought that a majority of proteins from which MHC-II-restricted peptides are derived are extracellular or host membrane proteins. Indeed, cells endocytose extracellular material as well as membrane proteins, and degrade these after fusion of vesicles with lysosomes. However, it should be noted that mass-spectrometry (MS) studies of MHC-II immunopeptidomes revealed that 25–55% of presented peptides are derived from cytosolic or nuclear proteins, demonstrating that MHC-II antigen processing and presentation covers all kinds of epitopes, similarly to MHC-I molecules. The main mechanism providing these intracellular proteins into lysosomes is autophagy, in which protein aggregates or even whole organelles are enclosed by a double membrane forming autophagosomes that fuse with lysosomes, ensuring MHC-II presentation of intracellular protein-derived peptides.
The size of MHC-I and MHC-II binding grooves allows both these molecules to bind 8–10 amino acid (aa) long epitopes in an extended conformation. Most often, peptide binding to different MHC alleles requires the use of two to three peptide anchor residues, with their side chains fitting snuggly within pockets of the MHC peptide binding cleft. There are however specific differences in how these two MHC classes bind peptides. The peptide-binding cleft is closed at both termini by large heavy chain residues that are conserved among most known MHC-I alleles, and that form hydrogen bonds with peptide termini. These structural restrictions explain why the MHC-I immunopeptidome is significantly enriched with mostly 8–12 aa-long peptides. However, it should be noted that significantly longer peptides are also always present in the analyzed immunopeptidomes, although in smaller amounts. The closed ends of MHC-I binding clefts do not entirely prevent the extension of peptides at both termini, or only at one end, a feature that seems to prevail more in specific MHC-I alleles compared to others. In contrast to MHC-I, MHC-II peptide-binding grooves are open at both ends, allowing presented epitopes to extend out of the cleft. Accordingly, the MHC-II immunopeptidome is mainly composed of 10–25 aa-long peptides.
The significantly enhanced quality and analytical stringency in sample preparation, MS methods and bioinformatics allow a thorough investigation of the cellular immunopeptidome, revealing that the abundance of each identified peptide ranges from 1 to approximately 10,000 copies per given cell, but also that immunopeptidomes are plastic and molded by cell-intrinsic and -extrinsic factors. Over the last 15 years, MS methods have allowed the identification and quantification of MHC-restricted peptide populations in a large ensemble of cell lines and tissues. The results of most of these studies are presented in databases such as SysteMHC, a catalog of MHC-I epitopes generated by MS, standardized MHC allele-specific peptide spectral libraries and links to other proteomics and immunology databases, or caAtlas which combines 43 immunopeptidomes containing information from 311 cancer samples from nine different kinds of cancer and over 700 non-cancerous samples. These MS-based workflows, combined with more reliable MHC ligand prediction algorithms, were particularly helpful to better understand the molecular mechanisms that process cellular and/or phagocyted proteins into epitopes, and identify disease-associated peptides. Besides numbers and sequence identities, these MS studies also provide novel insights into the protein origin of identified peptides, and quantification of peptide amounts that may represent one single protein. For example, more than 12,000 unique peptides were identified in a B-lymphoblastoid cell line, originating from 5,603 proteins located in the ER, nucleus and cytoplasm. The identified immunopeptidome corresponded to more than half of the proteome, composed of about 10,000 proteins. While half of the source proteins were represented by only one single peptide, the remaining proteins were represented by multiple unique peptides, each one binding to a different allele. Similarly, a study of 19 different tissues demonstrated the tissue specificity of the immunopeptidome. Among the 28,448 high-confidence H-2D b/H-2K b-associated peptides, about 30% were detected in only one particular tissue, and a very small amount (0.2% for H-2D b and 1.9% for H-2K b) was shared across all 19 studied tissues, emphasizing that the immunopeptidome may be a partial replica of the cellular peptidome.
