MHC Class I and Class II

Why MHC matters

• Foundation of all T cell recognition. Unlike B cells, which see free antigen via BCR, T cells only see peptide-MHC. Every adaptive cellular response — viral clearance, tumor surveillance, transplant rejection, autoimmunity — runs through MHC presentation. About 10^4-10^5 MHC molecules per cell, each holding one peptide.
• HLA matching is life-or-death for transplants. A 10/10 HLA-A/B/C/DR/DQ-matched bone marrow transplant has 60-70% 5-year survival; a 7/10 mismatched transplant drops to 30-40% from graft-versus-host disease. The 1990 first successful sibling donor stem cell transplant (Thomas, Nobel 1990) hinged on MHC matching.
• HLA polymorphism is unmatched in the genome. Over 30,000 alleles known across the HLA loci as of 2024, with most variation concentrated in the peptide-binding groove. Selection pressure: each MHC variant binds a different peptide repertoire, and a polymorphic population resists pathogens that any single haplotype cannot.
• HLA association with disease is enormous. HLA-B27 carriers have ~80x higher ankylosing spondylitis risk; HLA-B57:01 predicts abacavir hypersensitivity (FDA pre-treatment screen since 2008); HLA-DQ2/DQ8 underlies essentially all celiac disease. Forty-plus diseases have strong HLA associations.
• Cancer immunotherapy depends on MHC-I. Tumors that downregulate MHC-I escape CD8+ killing — a major mechanism of checkpoint inhibitor resistance. Roughly 40-90% of metastatic tumors show MHC-I loss. Restoring MHC-I expression with IFN-γ or epigenetic drugs is an active therapeutic axis.
• Vaccine design starts with MHC binding. Predicting which peptides bind a patient's HLA alleles (NetMHCpan, MHCflurry) is the entry point for personalized cancer vaccines. The 8-10 aa MHC-I groove and the 13-25 aa MHC-II groove are computationally tractable thanks to ~3 million peptide-HLA training datapoints.
• Pathogens evolved MHC-evasion arsenals. HIV Nef downregulates MHC-I from infected T cells; HSV ICP47 blocks TAP; HCMV encodes seven separate MHC-I antagonists. The depth of viral countermeasures testifies to how much fitness MHC-I surveillance imposes.

Common misconceptions

• MHC-I shows only viral or foreign peptides. No — MHC-I displays a representative sample of EVERYTHING in the cytosol, ~99% self peptides plus any pathogen-derived peptides. T cell selection in the thymus prunes self-reactive clones so that only foreign-presenting cells trigger killing.
• MHC-II is for engulfed pathogens only. Some intracellular peptides reach MHC-II via autophagy. About 20-30% of MHC-II-bound peptides come from cytosolic or nuclear proteins, not endosomal cargo.
• Cross-presentation is rare and unimportant. It is largely restricted to cDC1 dendritic cells, but it is essential for priming CD8+ T cells against viruses that don't infect DCs and against tumor antigens. Without cross-presentation, anti-tumor CTL responses fail.
• One TCR sees one peptide. TCRs are degenerate — a single TCR can productively engage 10^4 to 10^6 different peptide-MHC combinations, almost always at very low affinity. High-affinity peptides trigger; low-affinity ones don't. The repertoire's size (~10^7-10^8 unique TCRs in an adult) plus degeneracy gives near-universal pathogen coverage.
• HLA = MHC. HLA is the human nomenclature for MHC. In mice it's H-2; in chickens B; cattle BoLA. They are orthologous loci in the same chromosomal region, just named differently.
• Loading is irreversible. Both classes have peptide editors. Tapasin in the MHC-I peptide-loading complex preferentially retains high-affinity peptides; HLA-DM in MIIC catalyzes CLIP-to-peptide exchange and edits MHC-II for stable binders. Loading is iterative — peptides come off and back on until a high-affinity one wins.

How presentation actually works

MHC-I biogenesis starts in the ER. The heavy chain (44 kDa, encoded at HLA-A, B, or C) folds with β2-microglobulin (12 kDa, separate chromosome 15 gene) and joins the peptide-loading complex: TAP1/TAP2 transporter, tapasin chaperone, ERp57 oxidoreductase, calreticulin lectin. Cytosolic proteasomes (the 26S complex with 20S core; immunoproteasome β1i/2i/5i subunits induced by IFN-γ) cleave proteins into peptides 3-22 aa long. TAP pumps peptides into the ER lumen with a preference for 8-12 aa lengths and hydrophobic or basic C-termini. Tapasin holds nascent MHC-I in a peptide-receptive state and edits the peptide pool until a high-affinity 8-10 aa fragment locks in. Loaded MHC-I exits via Golgi to the cell surface in 30-60 minutes. Surface half-life is hours; degradation is by retroendocytosis and lysosomal turnover.

MHC-II takes a different path. α/β heterodimer assembles in the ER but the groove is plugged by the invariant chain (Ii), which blocks any ER peptide loading. The Ii sorting motif targets the complex through the Golgi to MIIC, a late endosomal compartment with the proton pump V-ATPase keeping pH ~5. Cathepsin S (in DCs and B cells) and cathepsin L (in thymic epithelium) digest Ii down to a residual CLIP fragment occupying the groove. HLA-DM, a non-classical MHC-II molecule, catalyzes CLIP exchange for high-affinity endosomal peptides and acts as a peptide editor. Because the MHC-II groove is open at both ends, peptide length is unconstrained at 13-25 aa with conserved registration. Loaded MHC-II exports to the surface and is recycled through endosomes for re-loading.

The peptide-MHC molecule on the surface is what the TCR sees. TCRαβ CDR3 loops contact the peptide directly while CDR1 and CDR2 contact MHC residues. CD8 binds MHC-I α3 domain (so CD8+ T cells only recognize MHC-I); CD4 binds MHC-II β2 domain (so CD4+ T cells only recognize MHC-II). A productive TCR-pMHC engagement plus CD28-B7 costimulation triggers naive T cell activation; without costimulation it triggers anergy or deletion. MHC restriction means the TCR co-evolved with self MHC during thymic selection — most TCR clones are dual-specific for foreign peptide on self MHC.

MHC Class I vs Class II

Feature	MHC Class I	MHC Class II
Peptide source	Cytosolic — proteasome → TAP → ER	Endosomal — engulfed material → cathepsins
Peptide length	8-10 aa (groove closed at both ends)	13-25 aa (groove open, peptide protrudes)
Presented to	CD8+ cytotoxic T cells	CD4+ helper T cells
Structure	Heavy chain + β2-microglobulin	α + β chain heterodimer
Expressing cells	All nucleated cells (not RBCs)	Professional APCs: DCs, macrophages, B cells; thymic epithelium
Human loci	HLA-A, HLA-B, HLA-C	HLA-DR, HLA-DQ, HLA-DP
Loading machinery	TAP, tapasin, ERp57, calreticulin	Invariant chain, HLA-DM, cathepsins
Effector outcome	Direct cytotoxic killing of infected/cancer cells	Help: B cell class switch, macrophage activation, CTL priming

Famous experiments and case studies

• Doherty and Zinkernagel 1974. Working with LCMV-infected mice at John Curtin School in Canberra, they showed that immune CTLs from one mouse strain killed virus-infected target cells only from the same H-2 strain — discovering MHC restriction. Their 1996 Nobel changed how the field thought about T cell recognition forever.
• Bjorkman and Wiley 1987 crystal structure. Pamela Bjorkman and Don Wiley solved HLA-A2's crystal structure showing the peptide-binding groove with electron density of unidentified peptides — the first molecular view of antigen presentation. The image of the α1/α2 helices clamping a peptide became iconic.
• Townsend 1986 endogenous processing. Alain Townsend showed that cytosolic synthesis of influenza nucleoprotein was sufficient to generate MHC-I-presented peptides — the discovery that MHC-I samples internal proteins, decoupling presentation from the secretory pathway.
• HIV elite controllers and HLA-B57. About 1 in 300 HIV-infected individuals controls viremia without therapy — strongly enriched for HLA-B57:01 and B27, which present highly conserved Gag epitopes. Mutations escaping these alleles cripple viral fitness, demonstrating the depth of CD8+ MHC-I selection pressure.
• Bare lymphocyte syndrome. Patients with mutations in CIITA, RFX5, RFXAP, or RFXANK lose MHC-II expression, producing severe combined immunodeficiency with recurrent infections from infancy. Less than 100 cases reported worldwide; bone marrow transplantation is the only curative therapy.