Major Histocompatibility Complex (MHC) Class I and MHC Class II Proteins: Conformational Plasticity in Antigen Presentation

Abstract

Antigen presentation by major histocompatibility complex (MHC) proteins is essential for adaptive immunity. Prior to presentation, peptides need to be generated from proteins that are either produced by the cell's own translational machinery or that are funneled into the endo-lysosomal vesicular system. The prolonged interaction between a T cell receptor and specific pMHC complexes, after an extensive search process in secondary lymphatic organs, eventually triggers T cells to proliferate and to mount a specific cellular immune response. Once processed, the peptide repertoire presented by MHC proteins largely depends on structural features of the binding groove of each particular MHC allelic variant. Additionally, two peptide editors-tapasin for class I and HLA-DM for class II-contribute to the shaping of the presented peptidome by favoring the binding of high-affinity antigens. Although there is a vast amount of biochemical and structural information, the mechanism of the catalyzed peptide exchange for MHC class I and class II proteins still remains controversial, and it is not well understood why certain MHC allelic variants are more susceptible to peptide editing than others. Recent studies predict a high impact of protein intermediate states on MHC allele-specific peptide presentation, which implies a profound influence of MHC dynamics on the phenomenon of immunodominance and the development of autoimmune diseases. Here, we review the recent literature that describe MHC class I and II dynamics from a theoretical and experimental point of view and we highlight the similarities between MHC class I and class II dynamics despite the distinct functions they fulfill in adaptive immunity.

Keywords: HLA; HLA-DM; adaptive immunity; antigen presentation; major histocompatibility complex; peptide exchange; protein dynamics; tapasin.

Figure 1

Structural characteristics of major histocompatibility complex (MHC) class I and MHC class II proteins and their compartment-dependent loading with processed peptides. (A) Domain topology of a pMHC class I and pMHC class II complex. (B) Structure of HLA-A68 in complex with an HIV-derived peptide (PDB: 4HWZ, left) and HLA-DR1 in complex with a hemagglutinin-derived peptide (1DLH, right). Indicated are the supposed interaction sites of MHC class I with tapasin and of MHC class II with DM as dashed gray lines. The peptide is shown in yellow with its N and C-terminus marked and relevant pockets are labeled green (C) Simplified illustration of MHC class I (left) and II (right) processing and peptide-editing pathways. CLIP, class II-associated invariant chain peptide; Caln., calnexin; Calr., calreticulin; ER, endoplasmic reticulum; PLC, peptide loading complex.

Figure 2

Global B-factor analysis of X-ray crystal structures of MHC class I and MHC class II. Shown is the variance of the normalized residual B factor values of CA atoms (A) derived from 297 human pMHC class I structures is plotted as blue to red spectrum on a HLA-A*0201/peptide complex (PDB: 5HHN) and (B) from 41 human pMHC class II structures is plotted on a DR1/peptide complex (PDB: 4X5W).

Figure 3

Conformational rearrangements upon DM binding and structural variations in type 1 diabetes-susceptible DQ complexes. (A) Structural rearrangement in the α1-S4 strand and 3₁₀-helical region seen in DR1 when bound to DM (limon cartoon) compared to DR1 unbound DM (red). (B) DM-induced rearrangements in the P1-pocket and the surrounding helical segments. PDBs used in (A,B) 1DLH and 4FQX. (C) Overlay of DQ2/ag (PDB: 1S9V), DQ6/hyp_1–13 (PDB:1UVQ) and DQ8/InsB_9–23 (PDB: 1JK8) showing the structural variations of the 3₁₀ helix and the P1-proximal β1-helix. Interdomain communication as exemplarily indicated by the hydrogen bond between αR52 and βE86/βT89 in the DQ8 allele variant is thought to increase the stability of these regions and was previously discussed to be linked to a lowered DM-susceptibility (62, 63). ag, αI-gliadin; hyp, hypocretin peptide 1–13; InsB, insulin B chain 9–23. (D) Structural alignment of DR1/CLIP (PDB: 3QXA) and DR1-αF54C/CLIP (PDB: 3QXD), a mutant that shows an altered conformation in the 3₁₀ helix and an increased DM susceptibility.

Figure 4

Thermodynamic model for peptide exchange of major histocompatibility complex (MHC) class I. Peptide–MHC class I (pMHCI) complexes can follow two mechanistic pathways for peptide exchange starting from pMHCI ground state (state 1). In the tapasin-catalyzed pathway, tapasin modulates conformational changes in the α2-1 helix (red) of the F pocket region (pink) and the α3 domain (not shown) that accelerate the kinetics of peptide dissociation (state 2) and the loading of a high-affinity peptide (3). More intermediates states (between state 1 and state 3) need to be identified by computational studies and/or NMR and X-ray crystallography. In the non-catalyzed pathway, the peptide dissociates from the sub-optimally-loaded intermediate state (state 1′). The resulting empty MHC molecule shows subtype-dependent dynamics (especially at the F pocket region, pink) and thus can exist in a stable peptide-receptive form (state 2′) or in an unstable form (state 2″) that is chaperoned by tapasin for peptide binding. The structures used in states 1 and state 3 were modified from PDB: 1UXS (shown in white). The models used in states 1′, 2, 2′, and 2″ represent suggested states by computational and experimental studies (shown in limon).

Figure 5

Thermodynamic model for peptide editing of major histocompatibility complex class II. pMHCII complexes can follow two mechanistic pathways for peptide exchange. The DM-catalyzed route requires multi-step transitions starting from the pMHCII ground state (1). This includes initially an out-flip movement of αW43 or the destabilization of β80–93 region via spontaneous conformational sampling of rare conformations (state 2′). DM would preferably select for conformations that are sampled on longer timescales and which show both, an out-flip movement of αW43 and a destabilized β80–93 region. Binding of DM to the energetically excited intermediate (which shows in part features of the DM-bound state) would then induce further rearrangements in the 3₁₀-helical region (state 3′) and thereby accelerate peptide-release. Binding of peptides which can displace the stabilizing interactions complete the peptide exchange process (state 4). Spontaneous (non-catalyzed) peptide exchange depends on the intrinsic stability of the pMHCII complex and does not rely on the sampling of rare conformations (state 2). Binding of a new peptide would likely require dissociation of the bound peptide, leading to the empty state (state 3) which rapidly converts into the non-receptive state (state 3b) but can also be chaperone by DM (state 3c) in order to allow for high-affinity peptide binding (state 4). Structures used in state 1, 3′, and 4 were derived from PDB: 4QXA, 4FQX, and 1DLH, respectively. Cartoons shown in 2, 2′, 3, and 3b were derived from molecular dynamic simulations (46, 91).