Antigen Processing and Presentation: MHC-I, MHC-II, Cross-Priming

Every T-cell response begins with a protein being taken apart. Before a CD8 or CD4 T cell can see anything, an antigen has to be degraded into peptides of the right length, delivered to the right compartment, edited against a competing pool of self peptides and displayed on a class I or class II molecule that survives long enough to reach the surface. Three semi-independent routes do this — the proteasomal class I pathway, the endosomal class II pathway, and cross-presentation — and the differences between them decide whether an immunogen produces cytotoxic immunity, helper immunity, or tolerance.

The class I route: from a mistranslated protein to a loaded groove

The class I pathway samples what a cell is making right now. Most of its peptide supply is not slowly turned-over cytoplasm but ubiquitin-tagged, E3 ligase-marked protein retrotranslocated by p97/VCP and, above all, DRiPs — defective ribosomal products destroyed within minutes of synthesis, which is why a virus becomes visible to CD8 T cells long before viral protein accumulates. The 26S proteasome cuts and TPP2 trims what it releases. IFN-γ then swaps in PSMB8 (β5i), PSMB9 (β1i) and PSMB10 (β2i) plus the PA28 (PSME1) cap, shifting cleavage towards the hydrophobic C-termini class I prefers — so an inflamed cell displays a qualitatively different repertoire, not simply more of it.

Transport is the bottleneck. TAP1/TAP2 pumps 8–16mers into the ER, where the peptide-loading complex assembles: tapasin bridges TAP to a nascent MHC-I heavy chain already folded with calnexin and paired with β₂m, with ERp57 and calreticulin completing the module. Tapasin is a proofreader rather than a chaperone: it holds the groove open and biases loading towards peptides with slow off-rates, while ERAP1 and ERAP2 trim N-termini to the 8–10mer optimum. Only a correctly loaded complex is released for the ER → Golgi → surface route, and NLRC5 sets how much of this machinery exists at all.

The class II route: invariant chain, the MIIC and H2-DM

Class II expression is largely the output of one transactivator. CIITA does not bind DNA itself but is recruited to the SXY module of the MHC-II genes, and its promoter usage explains why professional presenters are constitutively positive while other cells need IFN-γ. Newly made αβ dimers are born in an ER full of class I peptide, so CD74, the invariant chain, trimerises with them and plugs the groove with its CLIP segment. Its dileucine motif carries the address too, diverting the complex into the endocytic system where antigen taken up through DEC-205, DC-SIGN, Langerin or CD206 is already being degraded.

As Rab5 compartments convert to Rab7 ones and the v-ATPase acidifies the lumen, cathepsin S, cathepsin L, cathepsin D and legumain activate, GILT (IFI30) reduces disulfides to expose buried epitopes, and Ii is stripped back to CLIP. H2-DM then catalyses exchange until a peptide with a slow enough off-rate arrives — selecting for kinetic stability, not sequence. Loaded MHC-II:peptide leaves the LAMP1⁺ MIIC by tubular exocytosis. On an immature dendritic cell MARCH1 ubiquitinates the β-chain tail and sends it back for degradation; maturation switches MARCH1 off, and that single change, not new synthesis, produces most of the class II surge.

Cross-presentation: the cDC1 exception

Anti-viral and anti-tumour CD8 responses usually depend on a dendritic cell presenting antigen it never made. That capacity sits in one lineage: BATF3- and IRF8-dependent cDC1s, expanded by Flt3L and identified by XCR1, CLEC9A (DNGR-1) and, in tissue, CD103 (M290). CLEC9A binds F-actin exposed only once a cell has lost membrane integrity, so dead-cell cargo is routed preferentially into cross-presentation, and HMGB1 with ecto-calreticulin flags it as immunogenic. Uptake itself is promiscuous — CD36, LOX-1, αvβ5, HSP70/GRP94 complexes and FcγRIII/II (2.4G2) or FcγRI on immune complexes all feed the same phagosome.

The phagosome is what is unusual. Sec22b/SNAP23-dependent fusion imports ERGIC membrane carrying TAP, tapasin and Sec61, while NOX2 with p47-phox generates ROS that consume protons and hold the lumen alkaline. Restrained proteolysis is the point: cargo is deliberately under-digested so intact protein survives export to the cytosol, proteasomal cleavage and re-import through TAP. A parallel vacuolar route skips the cytosol entirely, using cathepsin S and IRAP in Rab11⁺ recycling endosomes, with Rab34 delaying maturation. Proteasome and TAP inhibitors abolish the first route and spare the second — the cleanest way to tell which one a vaccine formulation actually uses.

Licensing the presenter: pattern recognition and interferon

A dendritic cell that has processed antigen but received no danger signal is tolerogenic, not immunogenic. TLR4 sensing LPS and TLR9 sensing CpG DNA signal through MyD88 to NF-κB, while TLR3 responding to poly(I:C) runs the TRIF arm to IRF3 and type I interferon. The maturation programme follows: CD80, CD86, CD40, CD70 and CD83 rise, macropinocytosis falls, and CCR7 appears so the cell can reach the draining node. Processing capacity and migratory capacity are deliberately anti-correlated — a mature DC carries a fixed peptide cargo rather than continuing to sample.

Interferon then rebuilds the machinery. IFN-α and IFN-β act through IFNAR1, IFN-γ through IFNGR1, with JAK1 and JAK2 converging on STAT1 and IRF1. IRF1 ties the diagram together: immunoproteasome subunits, PA28, TAP, NLRC5 for class I, CIITA for class II and CXCL9 to recruit effectors all move at once, which is why IFN-γ blockade gives such a broad phenotype. The same signal installs its brakes — SOCS1 terminates JAK–STAT within hours, IL-10 suppresses class II and co-stimulation together, and STAT1 induces PD-L1. CD1d runs alongside as a separate, lipid-presenting axis.

The output: priming, effector function and tolerance

Peptide-MHC is only signal one. MHC-I:peptide — H-2Kd or H-2Kb in the common haplotypes — engages the TCR on a CD8 T cell with CD8 (Ly-2) as co-receptor, while MHC-II:peptide engages a CD4 (GK1.5) T cell. Lck delivered by the co-receptor phosphorylates the CD3ζ ITAMs, ZAP-70 docks and phosphorylates LAT, and the LAT signalosome converts occupancy into calcium flux and transcription. Signal two comes from the maturation programme: CD80 and CD86 engaging CD28, and CD70 engaging CD27.

Help is transitive. A CD4 T cell recognising class II delivers CD40L to CD40 on the same dendritic cell, licensing it to prime CD8 T cells; IL-12 then commits the response to the Th1/CTL programme and IL-15 supports the memory transition, with perforin, granzyme B and FasL as the effector output. Remove signal two and the identical contact produces anergy: CTLA-4 (9D9) outcompetes CD28, PD-1 (29F.1A12) meeting PD-L1 (10F.9G2) recruits SHP-2 to the TCR module, and IL-10 pushes towards Treg tolerance. Tumours work both ends — losing β₂m, TAP or NLRC5 hides the target while PD-L1 disarms the T cells that still see it.

Key targets and matching reagents

Studying antigen processing and presentation in vivo

This pathway is unusually well served by functional-grade antibodies, because so many of its decisive nodes are surface receptors or secreted cytokines. Experiments generally fall into three groups.

1. Targeting and depleting the presenting cell

Anti-CD11c, anti-CD103 (M290) and anti-CD209b identify or deplete dendritic-cell subsets, and antigen conjugated to DEC-205 or Langerin is the classical way to force cargo into a defined processing route. Always block with anti-CD16/CD32 (2.4G2) before staining, since FcγR is both a real cross-presentation receptor here and a major source of artefact. Quantify the compartment with Flt3L, XCR1, CLEC9A, BATF3 and IRF8 readouts.

2. Blocking presentation and co-stimulation

Anti-MHC-II (M5/114) and anti-H-2Kb (Y-3) or anti-H-2Kd block the presenting surface directly; anti-CD4 (GK1.5), anti-CD8a (Ly-2) and anti-CD3 (145-2C11) address the responding compartment. To separate signal one from signal two, combine these with anti-CD80, anti-CD86, anti-CD28, anti-CD40 and anti-CD70 — blocking CD40 removes licensing without touching peptide display, which is the cleanest way to show that a CD8 defect is help-dependent.

3. Manipulating the cytokine environment

Anti-IFN-γ, anti-IFNAR1 and anti-IFN-β collapse the transactivation arm, and anti-IL-10 releases its main brake; anti-CXCL9 (MIG, 2F5.5) separates recruitment from priming. Pair blockade with checkpoint reagents — anti-PD-1 (29F.1A12), anti-PD-L1 (10F.9G2) and anti-CTLA-4 (9D9) — and read out with ELISAs for IL-12p40, IL-15, granzyme B, perforin and STAT1 to turn a blocking experiment into a quantitative one.

Featured products for this pathway

Frequently asked questions

For research use only. Not for use in diagnostic or therapeutic procedures.