Chapter 6: Antigen Presentation to T Lymphocytes PDF

213 Antigen Presentation to T Lymphocytes Vertebrate adaptive immune cells possess two types of antigen receptors: the immunoglobulins that serve as antigen receptors on B cells, and the T-cell receptors. While immunoglobulins can recognize native antigens, T cells recognize only antigens that are...

213 Antigen Presentation to T Lymphocytes Vertebrate adaptive immune cells possess two types of antigen receptors: the immunoglobulins that serve as antigen receptors on B cells, and the T-cell receptors. While immunoglobulins can recognize native antigens, T cells recognize only antigens that are displayed by MHC complexes on cell surfaces. The conventional α:β T cells recognize antigens as peptide:MHC complexes (see Section 4-13). The peptides recognized by α:β T cells can be derived from the normal turnover of self proteins, from intracellular pathogens, such as viruses, or from products of pathogens taken up from the extracellular fluid. Various tolerance mechanisms normally prevent self peptides from initiating an immune response; when these mechanisms fail, self peptides can become the target of autoimmune responses, as discussed in Chapter 15. Other classes of T cells, such as MAIT cells and γ:δ T cells (see Sections 4-18 and 4-20), recognize different types of surface molecules whose expression may indicate infection or cellular stress. The first part of this chapter describes the cellular pathways used by various types of cells to generate peptide:MHC complexes recognized by α:β T cells. This process participates in adaptive immunity in at least two different ways. In somatic cells, peptide:MHC complexes can signal the presence of an intracellular pathogen for elimination by armed effector T cells. In dendritic cells, which may not themselves be infected, peptide:MHC complexes serve to activate antigen-specific effector T cells. We will also introduce mechanisms by which certain pathogens defeat adaptive immunity by blocking the production of peptide:MHC complexes. The second part of this chapter focuses on the MHC class I and II genes and their tremendous variability. The MHC molecules are encoded within a large cluster of genes that were first identified by their powerful effects on the immune response to transplanted tissues and were therefore called the major histocompatibility complex (MHC). There are several different MHC molecules in each class, and each of their genes is highly polymorphic, with many variants present in the population. MHC polymorphism has a profound effect on antigen recognition by T cells, and the combination of multiple genes and polymorphism greatly extends the range of peptides that can be presented to T cells in each individual and in populations as a whole, thus enabling individuals to respond to the wide range of potential pathogens they will encounter. The MHC also contains genes other than those for the MHC molecules; some of these genes are involved in the processing of antigens to produce peptide:MHC complexes. The last part of the chapter discusses the ligands for unconventional classes of T cells. We will examine a group of proteins similar to MHC class I molecules that have limited polymorphism, some encoded within the MHC and others encoded outside the MHC. These so-called nonclassical MHC class I proteins serve various functions, some acting as ligands for γ:δ T-cell receptors and MAIT cells, or as ligands for NKG2D expressed by T cells and NK cells. In addition, we will introduce a special subset of α:β T cells known as invariant NKT cells that recognize microbial lipid antigens presented by these proteins. ERRNVPHGLFRVRUJ 6 IN THIS CHAPTER The generation of α:β T-cell receptor ligands. The major histocompatibility complex and its function. Generation of ligands for unconventional T-cell subsets. 214 Chapter 6: Antigen Presentation to T Lymphocytes The generation of α:β T-cell receptor ligands. The protective function of T cells depends on their recognition of cells harboring intracellular pathogens or that have internalized their products. As we saw in Chapter 4, the ligand recognized by an α:β T-cell receptor is a peptide bound to an MHC molecule and displayed on a cell surface. The generation of peptides from native proteins is commonly referred to as antigen processing, while peptide display at the cell surface by the MHC molecule is referred to as antigen presentation. We have already described the structure of MHC mole­ cules and seen how they bind peptide antigens in a cleft, or groove, on their outer surface (see Sections 4-13 to 4-16). We will now look at how peptides are generated from the proteins derived from pathogens and how they are loaded onto MHC class I or MHC class II molecules. 6-1 Antigen presentation functions both in arming effector T cells and in triggering their effector functions to attack pathogeninfected cells. The processing and presentation of pathogen-derived antigens has two distinct purposes: inducing the development of armed effector T cells, and triggering the effector functions of these armed cells at sites of infection. MHC class I molecules bind peptides that are recognized by CD8 T cells, and MHC class II molecules bind peptides that are recognized by CD4 T cells, a pattern of recognition determined by specific binding of the CD8 or CD4 molecules to the respective MHC molecules (see Section 4-18). The importance of this specificity of recognition lies in the different distributions of MHC class I and class II molecules on cells throughout the body. Nearly all somatic cells (except red blood cells) express MHC class I molecules. Consequently, the CD8 T cell is primarily responsible for pathogen surveillance and cytolysis of somatic cells. Also called cytotoxic T cells, their function is to kill the cells they recognize. CD8 T cells are therefore an important mechanism in eliminating sources of new viral particles and bacteria that live only in the cytosol, and thus freeing the host from infection. By contrast, MHC class II molecules are expressed primarily only on cells of the immune system, and particularly by dendritic cells, macrophages, and B cells. Thymic cortical epithelial cells and activated, but not naive, T cells can express MHC class II molecules, which can also be induced on many cells in response to the cytokine IFN-γ. Thus, CD4 T cells can recognize their cognate antigens during their development in the thymus, on a limited set of ‘professional’ antigen-presenting cells, and on other somatic cells under specific inflammatory conditions. Effector CD4 T cells comprise several subsets with different activities that help eliminate the pathogens. Importantly, naive CD8 and CD4 T cells can become armed effector cells only after encountering their cognate antigen once it has been processed and presented by activated dendritic cells. In considering antigen processing, it is important to distinguish between the various cellular compartments from which antigens can be derived (Fig. 6.1). These compartments, which are separated by membranes, include the cytosol and the various vesicular compartments involved in endocytosis and secretion. Peptides derived from the cytosol are transported into the endoplasmic reticulum and directly loaded onto newly synthesized MHC class I molecules on the same cell for recognition by T cells, as we will discuss below in greater detail. Because viruses and some bacteria replicate in the cytosol or in the contiguous nuclear compartment, peptides from their components can be loaded onto MHC class I molecules by this process (Fig. 6.2, first upper panel). ERRNVPHGLFRVRUJ The generation of α:β T-cell receptor ligands. Fig. 6.1 There are two categories of major intracellular compartments, separated by membranes. One compartment is the cytosol, which communicates with the nucleus via pores in the nuclear membrane. The other is the vesicular system, which comprises the endoplasmic reticulum, Golgi apparatus, endosomes, lysosomes, and other intracellular vesicles. The vesicular system can be thought of as being continuous with the extracellular fluid. Secretory vesicles bud off from the endoplasmic reticulum and are transported via fusion with Golgi membranes to move vesicular contents out of the cell. Extracellular material is taken up by endocytosis or phagocytosis into endosomes or phagosomes, respectively. The fusion of incoming and outgoing vesicles is important both for pathogen destruction in cells such as neutrophils and for antigen presentation. Autophagosomes surround components in the cytosol and deliver them to lysosomes in a process known as autophagy. Golgi secretory vesicle apparatus nucleus endoplasmic reticulum endosome This pathway of recognition is sometimes referred to as direct presentation, and can identify both somatic and immune cells that are infected by a pathogen. Certain pathogenic bacteria and protozoan parasites survive ingestion by macrophages and are able to replicate inside the intracellular vesicles of the endosomal–lysosomal system (Fig. 6.2, second panel). Other pathogenic bacteria proliferate outside cells, and can be internalized, along with their toxic products, by phagocytosis, receptor-mediated endocytosis, or macro­ pinocytosis into endosomes and lysosomes, where they are broken down by digestive enzymes. For example, receptor-mediated endocytosis by B cells can efficiently internalize extracellular antigens through B-cell receptors (Fig. 6.2, third panel). Virus particles and parasite antigens in extracellular fluids can also be taken up by these routes and degraded, and their peptides presented to T cells. cytosol lysosome autophagosome Immunobiology | chapter 6 | 06_001 Murphy et al | Ninth edition © Garland Science design by blink studio limited Some pathogens may infect somatic cells but not directly infect phagocytes such as dendritic cells. In this case, dendritic cells must acquire antigens from exogenous sources in order to process and present antigens to T cells. For example, to eliminate a virus that infects only epithelial cells, activation of CD8 T cells will require that dendritic cells load MHC class I molecules with peptides derived from viral proteins taken up from virally infected cells. This exogenous pathway of loading MHC class I molecules is called crosspresentation, and is carried out very efficiently by some specialized types of dendritic cells (Fig. 6.3). The activation of naive T cells by this pathway is called cross-priming. Degraded in Peptides bind to Presented to Effect on presenting cell Cytosolic pathogens Intravesicular pathogens Extracellular pathogens and toxins any cell macrophage B cell Cytosol Endocytic vesicles (low pH) Endocytic vesicles (low pH) MHC class I MHC class II MHC class II Effector CD8 T cells Effector CD4 T cells Effector CD4 T cells Cell death Activation to kill intravesicular bacteria and parasites Activation of B cells to secrete Ig to eliminate extracellular bacteria/toxins Immunobiology | chapter 6 | 06_002 Murphy et al | Ninth edition © Garland Science design by blink studio limited ERRNVPHGLFRVRUJ Fig. 6.2 Cells become targets of T-cell recognition by acquiring antigens from either the cytosolic or the vesicular compartments. Top, first panel: viruses and some bacteria replicate in the cytosolic compartment. Their antigens are presented by MHC class I molecules to activate killing by cytotoxic CD8 T cells. Second panel: other bacteria and some parasites are taken up into endosomes, usually by specialized phagocytic cells such as macrophages. Here they are killed and degraded, or in some cases are able to survive and proliferate within the vesicle. Their antigens are presented by MHC class II molecules to activate cytokine production by CD4 T cells. Third panel: proteins derived from extracellular pathogens may bind to cell-surface receptors and enter the vesicular system by endocytosis, illustrated here for antigens bound by the surface immunoglobulin of B cells. These antigens are presented by MHC class II molecules to CD4 helper T cells, which can then stimulate the B cells to produce antibody. 215 216 Chapter 6: Antigen Presentation to T Lymphocytes Cross-presentation of exogenous antigens by MHC class I molecules by dendritic cells MHC class I phagolysosome antigens ER Immunobiology | chapter 6 | 06_013 Murphy et al | Ninth edition © Garland Science design by blink studio limited Presentation of cellular antigens by MHC class II molecules self antigens autophagosome MHC class II CLIP MIIC Fig. 6.4 Autophagy Immunobiology | chapter 6 pathways | 06_111 Murphy et al | Ninth edition can deliver cytosolic antigens Garland Science design byby blinkMHC studio limited © for presentation class II molecules. In the process of autophagy, portions of the cytoplasm are taken into autophagosomes, specialized vesicles that are fused with endocytic vesicles and eventually with lysosomes, where the contents are catabolized. Some of the resulting peptides of this process can be bound to MHC class II molecules and presented on the cell surface. In dendritic cells and macrophages, this can occur in the absence of activation, so that immature dendritic cells may express self peptides in a tolerogenic context, rather than inducing T-cell responses to self antigens. Fig. 6.3 Cross-presentation of extracellular antigens on MHC class I molecules by dendritic cells. Certain subsets of dendritic cells are efficient in capturing exogenous proteins and loading peptides derived from them onto MHC class I molecules. There is evidence that several cellular pathways may be involved. One route may involve the translocation of ingested proteins from the phagolysosome into the cytosol for degradation by the proteasome, with the resultant peptides then passing through TAP (see Section 6-3) into the endoplasmic reticulum, where they load onto MHC class I molecules in the usual way. Another route may involve direct transport of antigens from the phagolysosome into a vesicular loading compartment—without passage through the cytosol—where peptides are allowed to be bound to mature MHC class I molecules. For loading peptides onto MHC class II molecules, dendritic cells, macro­ phages, and B cells are able to capture exogenous proteins via endocytic vesicles and through specific cell-surface receptors. For B cells, this process of antigen capture can include the B-cell receptor. The peptides that are derived from these proteins are loaded onto MHC class II molecules in specially modified endocytic compartments in these antigen-presenting cells, which we will discuss in more detail later. In dendritic cells, this pathway operates to activate naive CD4 T cells to become effector T cells. Macrophages take up particulate material by phagocytosis and so mainly present pathogen-derived peptides on MHC class II molecules. In macrophages, such antigen presentation may be used to indicate the presence of a pathogen within its vesicular compartment. Effector CD4 T cells, on recognizing antigen, produce cytokines that can activate the macrophage to destroy the pathogen. Some intravesicular pathogens have adapted to resist intracellular killing, and the macrophages in which they live require these cytokines to kill the pathogen: this is one of the roles of the TH1 subset of CD4 T cells. Other CD4 T cell subsets have roles in regulating other aspects of the immune response, and some CD4 T cells even have cytotoxic activity. In B cells, antigen presentation may serve to recruit help from CD4 T cells that recognize the same protein antigen as the B cell. By efficiently endocytosing a specific antigen via their surface immunoglobulin and presenting the antigen-derived peptides on MHC class II molecules, B cells can activate CD4 T cells that will in turn serve as helper T cells for the production of antibodies against that antigen. Beyond the presentation of exogenous proteins, MHC class II molecules can also be loaded with peptides derived from cytosolic proteins by a ubiquitous pathway of autophagy, in which cytoplasmic proteins are delivered into the endocytic system for degradation in lysosomes (Fig. 6.4). This pathway can serve in the presentation of self-cytosolic proteins for the induction of tolerance to self antigens, and also as a means for presenting antigens from pathogens, such as herpes simplex virus, that have accessed the cell’s cytosol. 6-2 Peptides are generated from ubiquitinated proteins in the cytosol by the proteasome. Proteins in cells are continually being degraded and replaced with newly synthesized proteins. Much cytosolic protein degradation is carried out by a large, multicatalytic protease complex called the proteasome (Fig. 6.5). A typical proteasome is composed of one 20S catalytic core and two 19S regulatory caps, one at each end; both the core and the caps are multisubunit complexes of proteins. The 20S core is a large cylindrical complex of some 28 subunits, arranged in four stacked rings of seven subunits each around a hollow core. The two outer rings are composed of seven distinct α subunits and are noncatalytic. The two inner rings of the 20S proteasome core are composed of seven distinct β subunits. The constitutively expressed proteolytic subunits are β1, β2, and β5, which form the catalytic chamber. The 19S regulator is composed of a base containing nine subunits that binds directly to the α ring of the 20S ERRNVPHGLFRVRUJ The generation of α:β T-cell receptor ligands. core particle and a lid that has up to 10 different subunits. The association of the 20S core with a 19S cap requires ATP as well as the ATPase activity of many of the caps’ subunits. One of the 19S caps binds and delivers proteins into the proteasome, while the other keeps them from exiting prematurely. Proteins in the cytosol are tagged for degradation via the ubiquitin–proteasome system (UPS). This begins with the attachment of a chain of several ubiquitin molecules to the target protein, a process called ubiquitination. First, a lysine residue on the targeted protein is chemically linked to the glycine at the carboxy terminus of one ubiquitin molecule. Ubiquitin chains are then formed by linking the lysine at residue 48 (K48) of the first ubiquitin to the carboxy-terminal glycine of a second ubiquitin, and so on until at least 4 ubiquitin molecules are bound. This K48-linked type of ubiquitin chain is recognized by the 19S cap of the proteasome, which then unfolds the tagged protein so that it can be introduced into the proteasome’s catalytic core. There the protein chain is degraded with a general lack of sequence specificity into short peptides, which are subsequently released into the cytosol. The general degradative functions of the proteasome have been co-opted for antigen presentation, so that MHC molecules have evolved to work with the peptides that the proteasome can produce. Various lines of evidence implicate the proteasome in the production of peptide ligands for MHC class I molecules. Experimentally tagging proteins with ubiquitin results in more efficient presentation of their peptides by MHC class I molecules, and inhibitors of the proteolytic activity of the proteasome inhibit antigen presentation by MHC class I molecules. Whether the proteasome is the only cytosolic protease capable of generating peptides for transport into the endoplasmic reticulum is not known. The constitutive β1, β2, and β5 subunits of the catalytic chamber are sometimes replaced by three alternative catalytic subunits that are induced by interferons. These induced subunits are called β1i (or LMP2), β2i (or MECL-1), and β5i (or LMP7). Both β1i and β5i are encoded by the PSMB9 and PSMB8 genes, which are located in the MHC locus, whereas β2i is encoded by PSMB10 outside the MHC locus. Thus, the proteasome can exist both as both a constitutive proteasome present in all cells and as the immunoproteasome, which is present in cells stimulated with interferons. MHC class I proteins are also induced by interferons. The replacement of the β subunits by their interferoninducible counterparts alters the enzymatic specificity of the proteasome such that there is increased cleavage of polypeptides after hydrophobic residues, and decreased cleavage after acidic residues. This produces peptides with carboxy-terminal residues that are preferred anchor residues for binding to most MHC class I molecules (see Chapter 4) and are also the preferred structures for transport by TAP. Another substitution for a β subunit in the catalytic chamber has been found to occur in cells in the thymus. Epithelial cells of the thymic cortex (cTECs) express a unique β subunit, called β5t, that is encoded by PSMB11. In cTECs, β5t becomes a component of the proteasome in association with β1i and β2i, and this specialized type of proteasome is called the thymoproteasome. Mice lacking expression of β5t have reduced numbers of CD8 T cells, indicating that the peptide:MHC complexes produced by the thymoproteasome are important in CD8 T-cell development in the thymus. Interferon-γ (IFN-γ) can further increase the production of antigenic peptides by inducing expression of the PA28 proteasome-activator complex that binds to the proteasome. PA28 is a six- or seven-membered ring composed of two proteins, PA28α and PA28β, both of which are induced by IFN-γ. A PA28 ring, which can bind to either end of the 20S proteasome core in place of the 19S regulatory cap, acts to increase the rate at which peptides are released (Fig. 6.6). In addition to simply providing more peptides, the increased rate of ERRNVPHGLFRVRUJ One 20S core combines with two 19S regulatory caps to form a proteasome in the cytosol 19S 20S 19S α β βα Polyubiquitinated proteins are bound by the 19S cap and degraded within the catalytic core, releasing peptides into the cytosol protein ubiquitin peptide fragments Immunobiology | chapter 6 | 06_100 Fig. 6.5 Cytosolic proteins are degraded Murphy et al | Ninth edition by the Science ubiquitin–proteasome design by blink studio limited system © Garland into short peptides. The proteasome is composed of a 20S catalytic core, which consists of four multisubunit rings (see text), and two 19S regulatory caps on either end. Proteins (orange) that are targeted become covalently tagged with K48-linked polyubiquitin chains (yellow) through the actions of various E3 ligases. The 19S regulatory cap recognizes polyubiquitin and draws the tagged protein inside the catalytic chamber; there, the protein is degraded, giving rise to small peptide fragments that are released back into the cytoplasm. 217 218 Chapter 6: Antigen Presentation to T Lymphocytes Fig. 6.6 The PA28 proteasome activator binds to either end of the proteasome. Panel a: in this side view cross-section, the heptamer rings of the PA28 proteasome activator (yellow) interact with the α subunits (pink) at either end of the core proteasome (the β subunits that make up the catalytic cavity of the core are in blue). Within this region is the α-annulus (green), a narrow ringlike opening that is normally blocked by other parts of the α subunits (shown in red). Panel b: a close-up view from the top, looking into the α-annulus without PA28 bound. Panel c: with the same perspective, the binding of PA28 to the proteasome changes the conformation of the α subunits, moving those parts of the molecule that block the α-annulus, and opening the end of the cylinder. For simplicity, PA28 is not shown. Structures courtesy of F. Whitby. PA28 α β catalytic chamber β b α PA28 a c Immunobiology | chapter 6 | 06_004 flow allows antigenic peptides to escape additional processing that Murphy et al | Ninthpotentially edition might destroy their antigenicity. © Garland Science design by blink studio limited Translation of self or pathogen-derived mRNAs in the cytoplasm generates not only properly folded proteins but also a significant quantity—possibly up to 30%—of peptides and proteins that are known as defective ribosomal products (DRiPs). These include peptides translated from introns in improperly spliced mRNAs, translations of frameshifts, and improperly folded proteins, which are tagged by ubiquitin for rapid degradation by the proteasome. This seemingly wasteful process provides another source of peptides and ensures that both self proteins and proteins derived from pathogens generate abundant peptide substrates for eventual presentation by MHC class I proteins. 6-3 Peptides from the cytosol are transported by TAP into the endoplasmic reticulum and further processed before binding to MHC class I molecules. The polypeptide chains of proteins destined for the cell surface, such as the two chains of MHC molecules, are translocated during synthesis into the lumen of the endoplasmic reticulum, where two chains fold correctly and assemble with each other. This means that the peptide-binding site of the MHC class I molecule is formed in the lumen of the endoplasmic reticulum and is never exposed to the cytosol. The antigen fragments that bind to MHC class I molecules, however, are typically derived from proteins made in the cytosol. This raises the question, How are these peptides able to bind to MHC class I molecules and be delivered to the cell surface? The answer was aided by analysis of mutant cells that had a defect in antigen presentation by MHC class I molecules. These cells expressed far fewer MHC ERRNVPHGLFRVRUJ The generation of α:β T-cell receptor ligands. class I proteins than normal on their surface despite normal synthesis of these molecules in the cytoplasm. This defect could be corrected by adding synthetic peptides to the culture medium, suggesting that the supply of peptides to the MHC class I molecules in the endoplasmic reticulum might be the limiting factor. Analysis of the DNA of the mutant cells identified the problem responsible for this phenotype to be in genes for members of the ATP-binding cassette (ABC) family of proteins; the ABC proteins mediate the ATP-dependent transport of ions, sugars, amino acids, and peptides across membranes. Missing from the mutant cells were two ABC proteins, called transporters associated with antigen processing-1 and -2 (TAP1 and TAP2), that are normally associated with the endoplasmic reticulum membrane. Transfection of the mutant cells with the missing genes restored the presentation of peptides by the cell’s MHC class I molecules. The two TAP proteins form a heterodimer in the membrane (Fig. 6.7), and mutations in either TAP gene can prevent antigen presentation by MHC class I molecules. The genes TAP1 and TAP2 are located in the MHC locus (see Section 6-10), near the PSMB9 and PSMB8 genes, and their basal level of expression is further enhanced by interferons produced in response to viral infection, similar to MHC class I and β1, β2, and β5 subunits of the proteasome. This induction results in increased delivery of cytosolic peptides into the endoplasmic reticulum. Microsomal vesicles from non-mutant cells can mimic the endoplasmic reticulum in assays in vitro, by internalizing peptides that then bind to MHC class I molecules present in the microsome lumen. In contrast, vesicles from TAP1or TAP2-deficient cells do not take up peptides. Peptide transport into normal microsomes requires ATP hydrolysis, confirming that the TAP1:TAP2 complex is an ATP-dependent peptide transporter. The TAP complex has limited specificity for the peptides it will transport, transporting peptides of between 8 and 16 amino acids in length and preferring peptides that have hydrophobic or basic residues at the carboxy terminus—the precise features of peptides that bind MHC class I molecules (see Section 4-15). The TAP complex has a bias against proline in the first three amino-terminal residues, but lacks in any true peptide-sequence specificity. The discovery of TAP explained how viral peptides from proteins synthesized in the cytosol gain access to the lumen of the endoplasmic reticulum and are bound by MHC class I molecules. Peptides produced in the cytosol are protected from complete degradation by cellular chaperones such as the TCP-1 ring complex (TRiC), but many of these peptides are longer than can be bound by MHC class I molecules. Evidence indicates that the carboxy terminus of peptide antigens is produced by cleavage in the proteasome. However, the amino terminus of peptides that are too long to bind MHC class I molecules can be trimmed by an enzyme called the endoplasmic reticulum aminopeptidase associated with antigen processing (ERAAP). Like other components of the antigen-processing pathway, expression of ERAAP is increased by IFN-γ stimulation. Mice lacking the enzyme ERAAP have an altered repertoire of peptides loaded onto MHC class I molecules. Although the loading of some peptides is not affected by the absence of ERAAP, other peptides fail to load normally, and many unstable and immunogenic peptides not normally present are found bound to MHC molecules on the cell surface. This causes cells from ERAAP-deficient mice to be immunogenic for T cells from wild-type mice, demonstrating that ERAAP is an important editor of the normal peptide:MHC repertoire. 6-4 Newly synthesized MHC class I molecules are retained in the endoplasmic reticulum until they bind a peptide. Binding a peptide is an important step in the assembly of a stable MHC class I molecule. When the supply of peptides into the endoplasmic reticulum is disrupted, as in TAP-mutant cells, newly synthesized MHC class I molecules ERRNVPHGLFRVRUJ Schematic diagram of TAP lumen of ER TAP1 TAP2 ER membrane hydrophobic transmembrane domain cytosol ATP-binding cassette (ABC) domain a b Immunobiology | chapter 6 | 06_003 Fig. and TAP2 form a peptide Murphy6.7 et alTAP1 | Ninth edition transporter in the endoplasmic Garland Science design by blink studio limited © reticulum membrane. Upper panel: TAP1 and TAP2 are individual polypeptide chains, each with one hydrophobic and one ATPbinding domain. The two chains assemble into a heterodimer to form a four-domain transporter typical of the ATP-binding cassette (ABC) family. The hydrophobic transmembrane domains have multiple transmembrane regions (not shown here). The ATP-binding domains lie within the cytosol, whereas the hydrophobic domains project through the membrane into the lumen of the endoplasmic reticulum (ER) to form a channel through which peptides can pass. Lower panel: electron microscopic reconstruction of the structure of the TAP1:TAP2 heterodimer. Panel a shows the surface of the TAP transporter as seen from the lumen of the ER, looking down onto the top of the transmembrane domains, while panel b shows a lateral view of the TAP heterodimer in the plane of the membrane. The ATP-binding domains form two lobes beneath the transmembrane domains; the bottom edges of these lobes are just visible at the back of the lateral view. TAP structures courtesy of G. Velarde. 219 220 Chapter 6: Antigen Presentation to T Lymphocytes are held in the endoplasmic reticulum in a partly folded state. This explains why the rare human patients who have been identified with immunodeficiency due to defects in TAP1 and TAP2 have few MHC class I molecules on their cell surfaces, a condition known as MHC class I deficiency. The folding and assembly of a complete MHC class I molecule (see Fig. 4.19) depends on the association of the MHC class I α chain first with β2-microglobulin and then with peptide, and this process involves a number of accessory proteins with chaperone-like functions. Only after peptide has bound is the MHC class I molecule released from the endoplasmic reticulum and transported to the cell surface. MHC Class I Deficiency Newly synthesized MHC class I α chains that enter the endoplasmic reticulum membranes bind to calnexin, a general-purpose chaperone protein that retains the MHC class I molecule in a partly folded state (Fig. 6.8). Calnexin also associates with partly folded T-cell receptors, immunoglobulins, and MHC class II molecules, and so has a central role in the assembly of many immunological as well as non-immunological proteins. When β2-microglobulin binds to the α chain, the partly folded MHC class I α:β2-microglobulin heterodimer dissociates from calnexin and binds to an assembly of proteins called the MHC class I peptide-loading complex (PLC). One component of Partly folded MHC class I α chains bind to calnexin until β2-microglobulin binds calreticulin MHC class I ER calnexin MHC class I α:β2m complex is released from calnexin, binds a complex of chaperone proteins (calreticulin, ERp57) and binds to TAP via tapasin Cytosolic proteins and defective ribosomal products (DRiPs) are degraded to peptide fragments by the proteasome. TAP delivers peptides to the ER A peptide binds the MHC class I molecule and completes its folding. The MHC class I molecule is released from the TAP complex and exported to the cell membrane ERp57 tapasin β2m TAP ERAAP normal proteins (>70%) cytosol peptide fragments DRiPs (<30%) ribosome proteasome ubiquitinated protein nucleus Immunobiology | chapter 6 | 06_005 Fig. 6.8 MHC class I molecules do not leave the endoplasmic Murphy et al | Ninth edition reticulum they bind Garland Scienceunless design by blink studio limitedpeptides. © Newly synthesized MHC class I α chains assemble in the endoplasmic reticulum (ER) with the membrane-bound protein calnexin. When this complex binds β2-microglobulin (β2m), the MHC class I α:β2m dimer dissociates from calnexin, and the partly folded MHC class I molecule then binds to the TAP-associated protein tapasin. Two MHC:tapasin complexes may bind with the TAP dimer at the same time. The chaperone molecules ERp57, which forms a heterodimer with tapasin, and calreticulin also bind to form the MHC class I peptide-loading complex. The MHC class I molecule is retained within the ER until released by the binding of a peptide, which completes the folding of the MHC molecule. Even in the absence of infection, there is a continual flow of peptides from the cytosol into the ER. Defective ribosomal products (DRiPs) and proteins marked for destruction by K48-linked polyubiquitin (yellow triangles) are degraded in the cytoplasm by the proteasome to generate peptides that are transported into the lumen of the endoplasmic reticulum by TAP. Some of these peptides will bind to MHC class I molecules. The aminopeptidase ERAAP trims the peptides at their amino termini, allowing peptides that are too long to bind to MHC class I molecules and thereby increasing the repertoire of potential peptides for presentation. Once a peptide has bound to the MHC molecule, the peptide:MHC complex leaves the endoplasmic reticulum and is transported through the Golgi apparatus and finally to the cell surface. ERRNVPHGLFRVRUJ The generation of α:β T-cell receptor ligands. the PLC—calreticulin—is similar to calnexin and probably also has a general chaperone function, like calnexin. A second component of the complex is the TAP-associated protein tapasin, encoded by a gene within the MHC. Tapasin forms a bridge between MHC class I molecules and TAP, allowing the partly folded α:β2-microglobulin heterodimer to await the transport of a suitable peptide from the cytosol. A third component of this complex is the chaperone ERp57, a thiol oxidoreductase that may have a role in breaking and re-forming the disulfide bond in the MHC class I α2 domain during peptide loading (Fig. 6.9). ERp57 forms a stable disulfide-linked heterodimer with tapasin. Tapasin seems to be a component of the PLC that is specific to antigen processing, while calnexin, ERp57, and calreticulin bind various other glycoproteins assembling in the endoplasmic reticulum and seem to be part of the cell’s general quality control machinery. TAP itself is the final component of the PLC, and it delivers peptides to the partially folded MHC class I molecule. MOVIE 6.1 The PLC maintains the MHC class I molecule in a state that is receptive to peptide binding and mediates the exchange of low-affinity peptides bound to the MHC molecule for peptides of higher affinity, a process called peptide editing. The ERp57:tapasin heterodimer functions in editing peptides binding to MHC class I. Cells lacking calreticulin or tapasin show defects in the assembly of MHC class I molecules, and those molecules that reach the cell surface are bound to suboptimal, low-affinity peptides. The binding of a peptide to the partly folded MHC class I molecule releases it from the PLC, and the peptide:MHC complex leaves the endoplasmic reticulum and is transported to the cell surface. Most of the peptides transported by TAP will not bind to the MHC molecules and are rapidly cleared out of the endoplasmic reticulum; these appear to be transported back into the cytosol by Sec61, an ATP-dependent transport complex distinct from TAP. As mentioned above, the MHC class I molecule must bind a peptide in order to be released from the PLC. In cells lacking functional TAP genes, the MHC class I molecules fail to exit the endoplasmic reticulum, and so must be degraded instead. Since the ubiquitin–proteasome system is located in the cytosol, these terminally misfolded MHC molecules must somehow be transported back into the cytoplasm for degradation. This is achieved by a system of quality control pathways called endoplasmic reticulum-associated protein degradation (ERAD). ERAD comprises several general cellular pathways that involve the recognition and delivery of misfolded proteins to a retrotranslocation complex that unfolds and translocates the proteins across the membrane of the endoplasmic reticulum and into the cytosol. The proteins are ubiquitinated during this process and so are targeted to the ubiquitin– proteasome system (UPS) for eventual degradation. We shall not delve deeply into the details of ERAD here, since these pathways are not unique to MHC Side view of the calreticulin, tapasin, ERp57, and MHC chaperone complex P domain calreticulin ERp57 tapasin a Fig. 6.9 The MHC class I peptide-loading complex includes the chaperones calreticulin, ERp57, and tapasin. This model shows a side (a) and top view (b) of the peptide-loading complex (PLC) oriented as it extends from the luminal surface of the endoplasmic reticulum. The newly synthesized MHC class I and β2-microglobulin are shown as yellow ribbons, with the α helices of the MHC peptide-binding groove clearly identifiable. The MHC and tapasin (cyan) would be tethered to the membrane of the endoplasmic reticulum by carboxy-terminal extensions not shown here. Tapasin and ERp57 (green) form a heterodimer linked by a disulfide bond, and tapasin makes contacts with the MHC molecule that stabilize the empty conformation of the peptide-binding groove; they function in editing peptides binding to the MHC class I molecule. Calreticulin (orange), like the calnexin it replaces (see Fig. 6.8), binds to the monoglucosylated N-linked glycan at asparagine 86 of the immature MHC molecule. The long, flexible P domain of calreticulin extends around the top of the peptide-binding groove of the MHC molecule to make contact with ERp57. The transmembrane region of tapasin (not shown) associates the PLC with TAP (see Fig. 6.8), bringing the empty MHC molecules into proximity with peptides arriving into the endoplasmic reticulum from the cytosol. Structure based on PDB file provided by Karin Reinisch and Peter Cresswell. Top view of chaperone complex tapasin ERp57 MHC b calreticulin Immunobiology | chapter 6 | 06_006 Murphy et al | Ninth edition ERRNVPHGLFRVRUJ © Garland Science design by blink studio limited 221 222 Chapter 6: Antigen Presentation to T Lymphocytes class I assembly or antigen processing. However, we will see in Chapter 13 how many viral pathogens co-opt the ERAD pathways to block assembly of MHC class I molecules as a way to evade recognition by CD8 T cells. In uninfected cells, peptides derived from self proteins fill the peptide-binding groove of the mature MHC class I molecules and are carried to the cell surface. In normal cells, MHC class I molecules are retained in the endoplasmic reticulum for some time, which suggests that they are present in excess of peptide. This is important for the immunological function of MHC class I molecules, which must be immediately available to transport viral peptides to the cell surface if the cell becomes infected. 6-5 Dendritic cells use cross-presentation to present exogenous proteins on MHC class I molecules to prime CD8 T cells. The pathway described above explains how proteins synthesized in the cytosol can generate peptides that become displayed as complexes with MHC class I molecules on the cell surface. This pathway is sufficient to ensure detection and destruction of pathogen-infected cells by cytotoxic T cells. But how do these cytotoxic T cells first become activated? Our explanation so far would require that dendritic cells become infected as well, so that they express the peptide:MHC class I complex needed to activate naive CD8 T cells. But many viruses exhibit a restricted tropism for different cells types, and not all viruses will infect dendritic cells. This creates the chance that antigens from such pathogens might never be displayed by dendritic cells, and that cytotoxic T cells that recognize them might not be activated. As it turns out, certain dendritic cells are able to generate peptide:MHC class I complexes from peptides that were not generated within their own cytosol. Peptides from extracellular sources—such as viruses, bacteria, and phagocytosed dying cells infected with cytosolic pathogens—can be presented on MHC class I molecules on the surface of these dendritic cells by the process of cross-presentation. Long before its role in priming T-cell responses to viruses was appreciated, cross-presentation was observed in studies of minor histocompatibility antigens. These are non-MHC gene products that can elicit strong responses between mice of different genetic backgrounds. When spleen cells from B10 mice of MHC type H-2b were injected into BALB mice of MHC type H-2b×d (which express both b and d MHC types), BALB mice generated cytotoxic T cells reactive against minor antigens of the B10 background. Some of these cytotoxic T cells recognized minor antigens presented by the H-2b B10 cells used for immunization, as one might expect from direct priming of T cells by the B10 antigen-presenting cells. But other cytotoxic T cells recognized minor B10 antigens only when presented by cells of the H-2d MHC type. This meant that these CD8 T cells had been activated in vivo by recognizing the minor B10 antigens presented by the BALB host’s own H-2d molecules. In other words, the minor histocompatibility antigens must have become transferred from the original immunizing B10 cells to the BALB host’s dendritic cells and processed for MHC class I presentation. We now know that cross-presentation by MHC class I molecules occurs not only for antigens on tissue or cell grafts, as in the original experiment described above, but also for viral and bacterial antigens. It appears that the capacity for cross-presentation is not equally distributed across all antigen-presenting cells. While still an area of active study, it seems that cross-presentation is most efficiently performed by certain subsets of dendritic cells that are present in both humans and mice. Dendritic cell subsets are not identified by the same markers in humans and mice, but in both species, one strongly cross-presenting dendritic cell subset requires the transcription factor BATF3 for its development, and these cells uniquely express the chemokine receptor XCR1. In lymphoid tissues such as the spleen, this ERRNVPHGLFRVRUJ The generation of α:β T-cell receptor ligands. lineage of dendritic cells expresses the CD8α molecule on the cell surface, and migratory dendritic cells in lymph nodes capable of cross-presentation are identified by their expression of the αE integrin (CD103). Mice lacking a functional BATF3 gene lack these types of dendritic cells and are also unable to generate normal CD8 T-cell responses to many viruses, including herpes simplex virus. The biochemical mechanisms enabling cross-presentation are still unclear, and there may be several different pathways at work. It is not clear whether all proteins captured by phagocytic receptors and taken into endosomes need to be transported into the cytosol and degraded by the proteasome in order to be cross-presented. Some evidence supports a direct pathway in which the PLC is transported from the endoplasmic reticulum to the endosomal compartments, allowing exogenous antigens to be loaded onto newly synthesized MHC class I molecules in phagosomes (see Fig. 6.3). Another pathway of cross-presentation by dendritic cells may involve an interferon-γ-induced GTPase known as IRGM3 (short for immune-related GTPase family M protein 3). IRGM3 interacts with adipose differentiation related protein (ADRP) in the endoplasmic reticulum and regulates the generation of neutral lipid storage organelles called lipid bodies, which are thought to originate from ER membranes. Dendritic cells from mice lacking IRGM3 are selectively deficient in cross-presentation of antigens to CD8 T cells, but have a normal process for presenting antigens on MHC class II molecules. The relationship between this and other pathways remains an area of active research. 6-6 Peptide:MHC class II complexes are generated in acidified endocytic vesicles from proteins obtained through endocytosis, phagocytosis, and autophagy. The immunological function of MHC class II molecules is to bind peptides generated in the intracellular vesicles of dendritic cells, macrophages, and B cells, and to present these peptides to CD4 T cells. The purpose for this pathway is different for each cell type. Dendritic cells primarily are concerned with activating CD4 T cells, while macrophages and B cells are concerned with receiving various forms of help from these CD4 T cells. For example, the intracellular vesicles of macrophages are the sites of replication for several types of pathogens, including the protozoan parasite Leishmania and the mycobacteria that cause leprosy and tuberculosis. Because these pathogens reside in membrane-enclosed vesicles, the proteins of these pathogens are not usually accessible to proteasomes in the cytosol. Instead, after activation of the macrophage, the pathogens are degraded by activated intravesicular proteases into peptide fragments that can bind to MHC class II molecules, which pass through this compartment on their way from the endoplasmic reticulum to the cell surface. Like all membrane proteins, MHC class II molecules are first delivered into the endoplasmic reticulum membrane, and are then transported onward as part of membrane-enclosed vesicles that bud off the endoplasmic reticulum and are directed to intracellular vesicles containing internalized antigens. Complexes of peptides and MHC class II molecules are formed there and are then delivered to the cell surface, where they can be recognized by CD4 T cells. Antigen processing for MHC class II molecules begins when extracellular pathogens and proteins are internalized into endocytic vesicles (Fig. 6.10). Proteins that bind to surface immunoglobulin on B cells and are internalized by receptor-mediated endocytosis are processed by this pathway. Larger particulate materials, such as fragments of dead cells, are internalized by phagocytosis, particularly by macrophages and dendritic cells. Soluble proteins, such as secreted toxins, are taken up by macropinocytosis. Proteins that enter cells through endocytosis are delivered to endosomes, which become ERRNVPHGLFRVRUJ 223 224 Chapter 6: Antigen Presentation to T Lymphocytes Antigen is taken up from the extracellular space into intracellular vesicles In early endosomes of neutral pH, endosomal proteases are inactive Acidification of vesicles activates proteases to degrade antigen into peptide fragments Vesicles containing peptides fuse with vesicles containing MHC class II molecules extracellular space cytosol Immunobiology | chapter 6 | 06_009 Murphy et al | Ninth edition Garland Science design by blink studiobind limited to MHC © Fig. 6.10 Peptides that class II molecules are generated in acidified endocytic vesicles. In the case illustrated here, extracellular foreign antigens, such as bacteria or bacterial antigens, have been taken up by an antigen-presenting cell such as a macrophage or an immature dendritic cell. In other cases, the source of the peptide antigen may be bacteria or parasites that have invaded the cell to replicate in intracellular vesicles. In both cases the antigen-processing pathway is the same. The pH of the endosomes containing the engulfed pathogens decreases progressively, activating proteases within the vesicles to degrade the engulfed material. At some point on their pathway to the cell surface, newly synthesized MHC class II molecules pass through such acidified vesicles and bind peptide fragments of the antigen, transporting the peptides to the cell surface. increasingly acidic as they progress into the interior of the cell, eventually fusing with lysosomes. The endosomes and lysosomes contain proteases, known as acid proteases, that are activated at low pH and eventually degrade the protein antigens contained in the vesicles. Drugs such as chloroquine that raise the pH of endosomes, making them less acidic, inhibit the presentation of intravesicular antigens, suggesting that acid proteases are responsible for processing internalized antigen. These proteases include the cysteine proteases—so called because they use a cysteine in their catalytic site—known as cathepsins B, D, S, and L, of which L is the most active. Antigen processing can be mimicked to some extent by the digestion of proteins with these enzymes in vitro at acid pH. Cathepsins S and L may be the predominant proteases in the processing of vesicular antigens; mice that lack cathepsin B or cathepsin D process antigens normally, whereas mice with no cathepsin S show some deficiencies, including in cross-presentation. Asparagine endopeptidase (AEP), a cysteine protease cleaving after asparagines, is important for processing some antigens, such as the tetanus toxin antigen for MHC class II presentation, but is not required in all cases where antigens contain asparagine residues near their relevant epitopes. It is likely that the overall repertoire of peptides produced within the vesicular pathway reflects the activities of the many proteases present in endosomes and lysosomes. Disulfide bonds, particularly intramolecular disulfide bonds, help in the denaturation process and facilitate proteolysis in endosomes. The enzyme IFN-γ-induced lysosomal thiol reductase (GILT) is present in endosomes and functions by breaking and re-forming disulfide bonds in the antigenprocessing pathway. The various endosomal proteases act in a largely redundant and nonspecific manner to digest regions of the polypeptide that have become accessible to proteolysis by denaturation and previous steps of degradation. The peptides generated vary in sequence and abundance throughout the endocytic pathway, so that MHC class II molecules can bind and present many different peptides from these compartments. A significant number of the self-peptides bound to MHC class II molecules arise from common proteins that are cytosolic in location, such as actin and ubiquitin. The most likely way in which cytosolic proteins are processed for MHC class II presentation is by the natural process of protein turn­over known as autophagy, in which damaged organelles and cytosolic proteins are delivered to lysosomes for degradation. Here their peptides could encounter MHC class II molecules present in the lysosome membranes, and the resulting peptide:MHC class II complex could be transported to the cell surface via endolysosomal tubules (see Fig. 6.4). Autophagy is constitutive, but it is increased by cellular stresses such as starvation, when the cell catabolizes intracellular proteins to obtain energy. In microautophagy, cytosol is ERRNVPHGLFRVRUJ The generation of α:β T-cell receptor ligands. continuously internalized into the vesicular system by lysosomal invaginations, whereas in macroautophagy, which is induced by starvation, a doublemembraned autophagosome engulfs cytosol and fuses with lysosomes. A third autophagic pathway uses the heat-shock cognate protein 70 (Hsc70) and the lysosome-associated membrane protein-2 (LAMP-2) to transport cytosolic proteins to lysosomes. Autophagy has been shown to be involved in the processing of the Epstein–Barr virus nuclear antigen 1 (EBNA-1) for presentation on MHC class II molecules. Such presentation enables cytotoxic CD4 T cells to recognize and kill B cells infected with Epstein–Barr virus. 6-7 The invariant chain directs newly synthesized MHC class II molecules to acidified intracellular vesicles. The biosynthetic pathway for MHC class II molecules begins with their translocation into the endoplasmic reticulum. Here, it is important to prevent them from prematurely binding to peptides transported into the endoplasmic reticulum lumen or to the cell’s own newly synthesized polypeptides. The endoplasmic reticulum is full of unfolded and partly folded polypeptide chains, and so a general mechanism is needed to prevent these from binding in the open-ended peptide-binding groove of the MHC class II molecule. Premature peptide binding is prevented by the assembly of newly synthesized MHC class II molecules with a membrane protein known as the MHC class II-associated invariant chain (Ii, CD74). Ii is a type II membrane glyco­ protein; its amino terminus resides in the cytosol and its transmembrane region spans the membrane of the endoplasmic reticulum (Fig. 6.11). The remainder of Ii and its carboxy terminus reside within the endoplasmic reticulum. Ii has a unique cylindrical domain that mediates formation of stable Ii trimers. Near this domain, Ii contains a peptide sequence, the class II-associated invariant chain peptide (CLIP), with which each Ii subunit of the trimer binds noncovalently to an MHC class II α:β heterodimer. Each Ii subunit binds to an MHC class II molecule with CLIP lying within the peptidebinding groove, thus blocking the groove and preventing the binding of either peptides or partly folded proteins. The binding site of an MHC class II molecule is open relative to the binding site of an MHC class I molecule. C terminus CLIP trimerization domain Invariant chain (Ii) binds in the groove of MHC class II molecule ER Ii is cleaved initially to leave a fragment bound to the class II molecule and to the membrane Further cleavage leaves a short peptide fragment, CLIP, bound to the class II molecule Ii CLIP N terminus transmembrane domain LIP10 cytosol Fig. 6.11 The invariant chain is cleaved to leave a peptide Immunobiology | chapter 6 | 06_010 Murphy et al | Ninth edition fragment, CLIP, bound to the MHC class II molecule. © Garland Science design by blink studio limited A model of the trimeric invariant chain bound to MHC class II α:β heterodimers is shown on the left. The CLIP portion is shown in purple, the rest of the invariant chain is shown in green, and the MHC class II molecules are shown in yellow (model, and leftmost of the three panels). In the endoplasmic reticulum, the invariant chain (Ii) binds to MHC class II molecules with the CLIP section of its polypeptide chain lying along the peptide-binding groove. After transport into an acidified vesicle, Ii is cleaved, initially just at one side of the MHC class II molecule (center panel), first by non-cysteine proteases to give a remaining portion of Ii known as the leupeptin-induced peptide LIP22 (not shown), and then by cysteine protease to the LIP10 fragment shown. LIP10 retains the transmembrane and cytoplasmic segments that contain the signals that target Ii:MHC class II complexes tothe endosomal pathway. Subsequent cleavage (right panel) of LIP10 leaves only a short peptide still bound by the class II molecule; this peptide is the CLIP fragment. Model structure courtesy of P. Cresswell. ERRNVPHGLFRVRUJ 225 226 Chapter 6: Antigen Presentation to T Lymphocytes This allows MHC class II molecules to more easily allow the CLIP region of Ii to pass through their binding sites. While this complex is being assembled in the endoplasmic reticulum, its component parts are associated with calnexin. Only when a nine-chain complex—three Ii chains, three α chains, and three β chains—has been assembled is the complex released from calnexin for transport out of the endoplasmic reticulum. As part of the nine-chain complex, the MHC class II molecules cannot bind peptides or unfolded proteins, so that peptides present in the endoplasmic reticulum are not usually presented by MHC class II molecules. There is evidence that in the absence of Ii many MHC class II molecules are retained in the endoplasmic reticulum as complexes with misfolded proteins. Trafficking of membrane proteins is controlled by cytosolic sorting tags. In this regard, Ii has a second function, which is to target delivery of the MHC class II molecules to a low-pH endosomal compartment where peptide loading can occur. The complex of MHC class II α:β heterodimers with Ii trimers is retained for 2–4 hours in this compartment (see Fig. 6.11). During this time, the Ii molecule undergoes an initial cleavage by acid proteases to remove the trimerization domain, generating a truncated 22-kDa fragment of Ii called LIP22. This is further cleaved by cysteine proteases into a 10-kDa fragment called LIP10, which remains bound to the MHC class II molecule and retains it within the proteolytic compartment. A subsequent cleavage of LIP10 releases the MHC class II molecule from the membrane-associated Ii, leaving the CLIP fragment bound to the MHC molecule. This cleavage is carried out by cathepsin S in most MHC class II-positive cells but by cathepsin L in thymic epithelial cells. Being associated with CLIP, the MHC class II molecules cannot yet bind other peptides. However, since CLIP does not carry the Ii-encoded signals that retain the complex in the endocytic compartment, the MHC–CLIP complex is now free to escape to the cell surface. MIIC G Immunobiology | chapter 6 | 06_011 Fig. 6.12 MHC class II molecules Murphy et al | Ninth edition are loaded with peptide in a late endosomal compartment called the MIIC. MHC class II molecules are transported from the Golgi apparatus (labeled G in this electron micrograph of an ultrathin section of a B cell) to the cell surface via intracellular vesicles called the MHC class II compartment (MIIC). These have a complex morphology, showing internal vesicles and sheets of membrane. Antibodies labeled with different-sized gold particles identify the presence of both MHC class II molecules (visible as small dark spots) and the invariant chain (large dark spots) in the Golgi, whereas only MHC class II molecules are detectable in the MIIC. This compartment is thought to be a late endosome, an acidified compartment of the endocytic system (pH 4.5–5) in which the invariant chain is cleaved and peptide loading occurs. Photograph (×135,000) courtesy of H.J. Geuze. © Garland Science design by blink studio limited To allow another peptide to bind to the MHC class II molecule, CLIP must either dissociate or be displaced. Newly synthesized MHC class II molecules are brought toward the cell surface in vesicles, most of which at some point fuse with incoming endosomes. However, some MHC class II:Ii complexes may first be transported to the cell surface and reinternalized into endosomes. In either case, MHC class II:Ii complexes enter the endosomal pathway, where they