Skip to lesson
OncoGuideeducationDiana’s wiki
THE EDUCATION LIBRARY

Antigen-processing machinery

In one sentence

Antigen-processing machinery generates protein fragments and loads them onto display molecules for T-cell inspection.

The intuition

A display case needs a supply chain. Producing a protein is only the first step; fragments must be generated, loaded and transported. This is an analogy for linked biochemical processes. There is no single factory switch that every antigen uses.

How it works

In a common class-I route, proteasomes cut intracellular proteins. The transporter associated with antigen processing (TAP) moves fragments into the endoplasmic reticulum, a cellular processing compartment. Aminopeptidases can trim fragments further.

A class-I human leukocyte antigen (HLA) heavy chain assembles with beta-2 microglobulin (Beta-2 microglobulin (B2M)) and a peptide. Supporting proteins assist loading and quality control. Suitable complexes then reach the surface, where conventional CD8 T cells may recognize them.

Class-II presentation uses a different route, often involving proteins taken into endosomal compartments. Cross-presentation also allows externally acquired material to reach class I. Individual targets can have different dependencies; the common class-I chain is a teaching model, not an exhaustive map.

Intracellular protein Fragments generated Transport and trimming Peptide loaded with class-I heavy chain and B2M Surface display Receptor recognition is a separate gate

The common class-I pathway has shared bottlenecks; expression alone does not prove display.

Why it matters in cancer

A tumor may retain an antigen’s sequence but lose useful presentation. More candidate peptides cannot bypass a shared broken step. RNA measurements of machinery describe expression, rather than assembled surface complexes.

Worked example

A hypothetical cell makes an altered protein but lacks functional TAP. A TAP-dependent peptide may no longer reach class I efficiently. Measuring the altered transcript would miss this bottleneck. A target-specific presentation experiment would address a different question.

Common confusions

  • Binding prediction does not test the whole processing pathway.
  • Class I and class II do not use identical machinery.
  • Low expression, gene loss and an experimentally demonstrated functional defect are different findings.

Sources and scope

Source check: October 8, 2026. This is a mechanism explanation, not a treatment recommendation. Expert and learner review remain pending.

Used in