Cell Biology
Clathrin-Coated Vesicles
Triskelions self-assemble into hexagonal/pentagonal lattices that pinch ~100 nm vesicles from membranes
Clathrin-coated vesicles are ~100 nm membrane carriers built from a self-assembling lattice of three-legged triskelions. Each triskelion = three heavy chains (~190 kDa each) + three light chains (~25 kDa each); the lattice has exactly 12 pentagons (an Euler-theorem requirement for a closed cage) plus however many hexagons its size needs — ~36 triskelions for a small minicoat, more for a ~100-nm vesicle. Cargo is selected by adaptor complexes (AP-2 at the plasma membrane, AP-1 at the TGN) that recognize sorting motifs (YxxΦ, NPxY, dileucine) on cargo tails. Dynamin polymerizes a helical collar around the neck and severs the vesicle within seconds. Discovered by Barbara Pearse in 1975 from pig brain.
- Vesicle diameter80–120 nm
- Triskelion3 heavy chains (~190 kDa) + 3 light chains
- Lattice geometryalways 12 pentagons + size-dependent hexagons; ~36 triskelions in a minicoat
- Formation time30–60 s plasma membrane
- AdaptorsAP-2 (PM); AP-1, GGAs (TGN)
- DiscoveredBarbara Pearse 1975
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
Why clathrin-coated vesicles matter
- Selective uptake from outside the cell. Clathrin-mediated endocytosis is the cell's main route for nutrient capture and receptor downregulation. A single human cell uses up to ~2000 clathrin-coated pits per minute under stimulated conditions — recycling roughly an area equal to the entire plasma membrane every 30 to 60 minutes.
- Cholesterol uptake via the LDL receptor. Joseph Goldstein and Michael Brown's discovery (1976 onward) that LDL receptors cluster in coated pits to internalize cholesterol — and that mutations preventing this clustering cause familial hypercholesterolemia — won the 1985 Nobel Prize. The FDNPVY motif on the receptor's cytoplasmic tail is recognized by the adaptor ARH (or Dab2), recruiting clathrin and pulling the LDL into the cell.
- Iron acquisition. Transferrin loaded with two ferric ions binds the transferrin receptor; the YTRF motif on the receptor's tail recruits AP-2 and clathrin. The vesicle delivers transferrin to early endosomes, where pH-driven iron release frees the receptor for recycling. Each receptor cycles every 10 to 20 minutes; an erythroblast takes up enough iron via this loop to assemble ~400 million hemoglobin molecules.
- Receptor signaling termination. EGFR, after ligand binding and autophosphorylation, is monoubiquitinated by Cbl. Eps15 and epsin recognize the ubiquitin and recruit AP-2, internalizing the receptor in clathrin-coated vesicles for lysosomal degradation. Loss of this shut-off mechanism — Cbl mutations, or Cbl-binding-defective EGFR — drives oncogenic hyperproliferation.
- Synaptic vesicle recycling. Neurons release ~150 to 200 synaptic vesicles per active zone per stimulus train and recycle them within seconds via clathrin-mediated endocytosis. Dynamin-1 is required; the Drosophila shibire mutant — a temperature-sensitive dynamin allele — paralyzes flies at 32°C as synaptic vesicles deplete.
- Viral entry routes. Influenza A, hepatitis C, and many flaviviruses enter cells through clathrin-coated pits. The virus binds a surface receptor (sialic acid for flu) that already cycles via clathrin; the virus exploits the host machinery. Dynasore, a small-molecule dynamin inhibitor, blocks entry of these viruses in cell culture.
- Foundation for vesicle biology. The COPI, COPII, and retromer coats discovered later all use the same logic — adaptor recognition of cargo plus self-assembling polyhedral cage. Clathrin was the prototype that established "coat protein recruitment + curvature + scission" as the universal vesicle-budding paradigm.
Common misconceptions
- Clathrin alone bends the membrane. Clathrin lattices have intrinsic curvature, but real coated pits also depend on adaptor proteins (AP-2, epsin, CALM) that insert amphipathic helices into the outer leaflet wedging the lipid bilayer, on BAR-domain proteins (endophilin, amphiphysin) at the neck, and on actin-driven force in mammalian cells. Pure clathrin in vitro can curve membranes only weakly.
- Triskelions are rigid building blocks. The clathrin heavy-chain leg has flexible hinges between proximal, distal, and ankle segments, and the foot region (β-propeller) at the N-terminus rotates relative to the leg. That flexibility is required to switch hexagons to pentagons during curvature transitions and for the post-assembly disassembly by Hsc70/auxilin.
- The coat stays on the vesicle until fusion. No — within seconds of scission, the chaperone Hsc70 is recruited by the J-domain protein auxilin (or GAK in non-neuronal cells) and ATP-dependently pries the triskelions off the vesicle. The naked vesicle is then competent for downstream fusion. Coats trapped on vesicles by Hsc70 inhibition cannot fuse.
- One adaptor per pathway. A coated pit at the plasma membrane uses AP-2 plus a constellation of accessory adaptors and CLASP (clathrin-associated sorting protein) factors: ARH, Dab2, NUMB, β-arrestin, epsin, eps15, CALM, HIP1R, and others. Each recognizes a distinct subset of cargo. AP-2 alone is necessary but not sufficient for selective cargo capture.
- Pentagons sit only at the "corners" in a soccer-ball pattern. True for the textbook truncated-icosahedron cage (12 pentagons + 20 hexagons, 60 triskelions), but real coated vesicles vary widely in size; the 12-pentagon rule still holds (Euler), but the pentagons can sit anywhere in the lattice, not just at symmetric positions.
- All endocytosis goes through clathrin. Caveolin-coated pits, CLIC/GEEC tubules, fast endophilin-mediated endocytosis (FEME), and macropinocytosis are clathrin-independent routes. Clathrin handles a large share but not all internalization. Dynamin is required by both clathrin-dependent and many independent routes.
How clathrin-coated vesicle formation works
Initiation begins with FCHo1/2 and Eps15 sensing the local PI(4,5)P2 enrichment on the inner plasma membrane leaflet. AP-2 is recruited and undergoes a conformational opening — its μ2 subunit is phosphorylated by AAK1 at Thr156 and the heterotetramer pivots to expose its cargo-binding sites. AP-2's α and μ2 subunits engage the lipid; its β2 subunit binds the clathrin-box motif (LLNLD) on the heavy chain via the β-propeller. Cargo enters the pit when its cytoplasmic tail's sorting motif — YxxΦ on transferrin receptor, NPxY on LDL receptor (recognized by ARH or Dab2), dileucine [DE]xxxL[LI] on CD3 chains, monoubiquitin on EGFR (recognized by epsin) — docks into the corresponding pocket on AP-2 or an alternative adaptor.
Triskelions then polymerize on the cytoplasmic face. Each triskelion contributes one leg per vertex, so a hexagonal lattice has three legs interdigitating at each vertex. The lattice nucleates at the rim of the recruited adaptor patch and propagates outward and inward, gradually adding pentagons that introduce curvature. By Euler's formula, any closed cage whose faces are pentagons and hexagons (with three meeting at every vertex) has exactly 12 pentagons; the triskelion count is then 20 + 2×(number of hexagons). The smallest possible cage is a pure dodecahedron — 12 pentagons, no hexagons, 20 triskelions; the smallest barrel-shaped cage seen on real vesicles has 12 pentagons and 8 hexagons (36 triskelions); a ~100-nm coated vesicle has 12 pentagons and several dozen hexagons. The lattice deepens the pit from a flat patch to a hemispherical bud over 30 to 60 seconds, with actin polymerization (via N-WASP, Arp2/3, and HIP1R) providing additional inward force in mammalian cells.
Scission requires dynamin. The 100-kDa GTPase polymerizes into a helical collar around the ~15-nm neck of the deeply invaginated bud — typically 1 to 2 helical turns containing 13 to 14 dynamin dimers per turn. BAR-domain proteins endophilin and amphiphysin pre-recruit dynamin via their SH3 domains binding dynamin's PRD. Cooperative GTP hydrolysis induces a conformational change that simultaneously constricts and twists the helix, severing the lipid bilayer in seconds. Immediately post-scission, auxilin (or GAK) binds the released vesicle's coat and recruits the chaperone Hsc70, which uses ATP hydrolysis to disassemble the triskelion lattice within seconds, releasing soluble triskelions for re-use and exposing the naked vesicle for downstream tethering and SNARE-mediated fusion with the early endosome.
Clathrin-mediated vs caveolar vs macropinocytosis vs FEME
| Property | Clathrin-mediated | Caveolar | Macropinocytosis | FEME (fast endophilin) |
|---|---|---|---|---|
| Coat protein | Clathrin triskelions + AP-2 | Caveolin-1 (Cav1) | None | Endophilin (BAR domain) |
| Vesicle diameter | 80–120 nm | 50–80 nm flask-shaped | 0.2–10 µm | 50–80 nm tubular |
| Cargo selectivity | Sorting motifs (YxxΦ, NPxY) | Lipid-raft cargo (GPI-anchored) | Bulk fluid + receptors | Specific GPCRs, RTKs (interleukin-2) |
| Dynamin required | Yes | Yes | Variable; CDC42/Rac driven | Yes |
| Actin involvement | Yes (mammalian) | Limited | Yes (essential, protrusions) | Yes |
| Time scale | 30–60 s | Minutes (slower) | Seconds to minutes | ~10 s (fastest) |
| Canonical cargo | Transferrin, LDL, EGFR, GPCRs | SV40, albumin, GPI-anchored proteins | Antigens, fluid, growth factors | IL-2R, β1-AR upon activation |
AP-1 vs AP-2 vs AP-3 vs AP-4 vs AP-5 vs GGAs
| Adaptor | Location | Subunits | Cargo motifs recognized | Disease association |
|---|---|---|---|---|
| AP-1 | TGN, recycling endosomes | γ, β1, μ1, σ1 | YxxΦ, dileucine | MEDNIK syndrome (AP1S1) |
| AP-2 | Plasma membrane | α, β2, μ2, σ2 | YxxΦ, dileucine; cargo via PI(4,5)P2 | Transferrin uptake required |
| AP-3 | Late endosomes, lysosomes | δ, β3, μ3, σ3 | Dileucine; YxxΦ (μ3A) | Hermansky-Pudlak type 2 (β3A) |
| AP-4 | TGN (no clathrin coat) | ε, β4, μ4, σ4 | YxxΦ; FxxY (autophagy receptors) | SPG47/50/51/52 hereditary spastic paraplegia |
| AP-5 | Late endosomes, lysosomes | ζ, β5, μ5, σ5 | Less defined | SPG48 hereditary spastic paraplegia |
| GGAs (1–3) | TGN | Monomeric (single chain) | DXXLL acidic dileucine; ubiquitin | Mannose-6-phosphate receptor sorting |
Famous experiments
- Pearse purifies clathrin (1975). Barbara Pearse at the MRC LMB in Cambridge published the isolation of coated vesicles from pig brain and identified the major polypeptide as a single ~180-kDa chain in PNAS 73: 1255–1259. She named it clathrin from Latin clathratus (latticed). The paper transformed vesicle biology — what had been opaque "coats" in EM became a defined molecular system.
- Brown & Goldstein on the LDL receptor (1976 onward). Demonstrated that fibroblasts from familial hypercholesterolemia patients fail to cluster their LDL receptors in coated pits; the receptor's tail had a mutation in what would later be identified as the FDNPVY motif. The receptor-mediated endocytosis paradigm — and the recognition that coated pits are the cellular cholesterol gate — won the 1985 Nobel Prize.
- Schmid & Drubin TIRF live imaging (early 2000s). Total internal reflection fluorescence microscopy of fluorescently tagged clathrin (mainly via Tomas Kirchhausen's lab plus Drubin and others) recorded the formation lifetime of individual coated pits in living cells: ~30 to 60 seconds in mammalian cells, with characteristic recruitment kinetics for AP-2, then dynamin spike at scission, then disappearance of clathrin signal upon Hsc70 disassembly.
- shibire and dynamin scission. The temperature-sensitive Drosophila shibire mutation paralyzes flies at 32°C. Electron microscopy at restrictive temperature revealed deeply invaginated coated pits arrested with characteristic dynamin collars at necks — the visual proof that dynamin's GTPase activity is required for membrane scission. Mol cloning in 1991 identified the gene as a Drosophila dynamin.
- Cryo-EM structures of clathrin lattices (Fotin et al. 2004, Morris et al. 2019). Sub-nanometer cryo-EM maps from Tomas Kirchhausen and Stephen Harrison resolved the triskelion-leg interdigitation at vertices, the heavy-chain proximal-distal-ankle hinges, and the position of the light chains. The 2004 Nature paper at 7.9 Å was the first near-atomic view of how the lattice assembles at vertices.
Frequently asked questions
What is a clathrin triskelion?
A triskelion is the elementary subunit of the clathrin coat — a three-legged pinwheel structure named for the Triskelion symbol. It is a hexamer of three clathrin heavy chains (~190 kDa each, encoded by CLTC) and three clathrin light chains (CLTA or CLTB, ~25 kDa each). The heavy chains assemble at their C-terminal hubs and project as long curved legs (~50 nm each) terminating in N-terminal β-propeller domains that bind adaptor proteins. The legs of adjacent triskelions interdigitate at vertices to form the lattice. That lattice always contains exactly 12 pentagons — Euler's theorem requires it for any closed cage of pentagons and hexagons — plus however many hexagons its size calls for: the number of triskelions equals 20 + twice the number of hexagons, so a small ~70 nm "minicoat" (12 pentagons + 8 hexagons) uses 36 triskelions, while a ~100 nm coated vesicle uses several dozen more.
How does the lattice bend the membrane?
Pure clathrin lattices are slightly curved by design — the triskelion legs angle out of the plane at ~30°, so flat assemblies are unstable and tend to puckers spontaneously. To go from flat to curved, the lattice must convert hexagons (which tile a plane) into pentagons (which curve a surface, as in a soccer ball). This conversion happens through cooperative rearrangement: clathrin coats can grow as predominantly flat hexagonal arrays and then remodel, or assemble curved from the start — both pathways are observed by live-cell imaging. Membrane curvature is also driven by adaptor proteins (AP-2, epsin, CALM) that insert amphipathic helices into the outer leaflet, by the BAR-domain proteins endophilin and amphiphysin at the neck, and by actin polymerization providing the final pulling force in mammalian cells.
What does AP-2 do?
AP-2 is the heterotetrameric adaptor complex (subunits α, β2, μ2, σ2) that bridges clathrin to the plasma membrane and to cargo. Its core binds simultaneously to PI(4,5)P2 lipid (via α and μ2 subunits), to clathrin's terminal β-propeller (via the β2 hinge containing the clathrin-box motif LLNLD), and to cargo cytoplasmic tails carrying sorting signals: tyrosine-based YxxΦ motifs (Φ = bulky hydrophobic) bind the μ2 subunit; dileucine [DE]xxxL[LI] motifs bind the α/σ2 interface. The transferrin receptor uses YTRF (a YxxΦ); the LDL receptor uses FDNPVY (a noncanonical NPxY recognized by ARH or Dab2 — an AP-2-related but distinct route). AP-2 also undergoes a phosphorylation switch: AAK1 phosphorylates μ2 at Thr156 to lock the open conformation that binds cargo and PI(4,5)P2.
How does dynamin pinch off the vesicle?
Dynamin is a 100-kDa GTPase that polymerizes into a helical collar around the ~15-nm neck of an invaginating coated pit. Each helical turn contains roughly 13 to 14 dynamin dimers contacting the membrane via their PH domains. Cooperative GTP hydrolysis induces a conformational squeeze that constricts the helix and twists, severing the membrane within seconds. In dynamin-deficient mice or in shibire flies (a temperature-sensitive Drosophila dynamin mutant) at restrictive temperature, coated pits arrest at deeply invaginated stages with characteristic collared necks — the diagnostic 'shibire phenotype' first imaged by EM in the 1970s. Three dynamin paralogs exist in mammals: dynamin-1 (neuronal synaptic vesicles), dynamin-2 (ubiquitous), dynamin-3 (testis, brain). Dynamin-2 mutations cause Charcot-Marie-Tooth disease and centronuclear myopathy.
What cargo travels in clathrin-coated vesicles?
Iron uptake (transferrin–transferrin receptor complex), cholesterol uptake (LDL bound to LDL receptor), epidermal growth factor receptor (EGFR) signaling termination, GPCR desensitization (β-arrestin recruits clathrin), and synaptic vesicle recycling (in neurons, dynamin-1 recycles vesicles within seconds of fusion) are the canonical examples. Influenza A virus, HIV, hepatitis C virus, and SARS-CoV-2 (in some entry routes) hijack clathrin-mediated endocytosis. Each cargo carries a sorting signal: transferrin receptor's YTRF tyrosine motif, LDL receptor's FDNPVY (NPxY family), EGFR's monoubiquitin tag, β-arrestin's AP-2-binding motif. Cargo selection is the cell's main filter on what enters.
How was clathrin discovered?
Barbara Pearse, working at the MRC Laboratory of Molecular Biology in Cambridge, identified clathrin in 1975 by purifying coated vesicles from pig brain and identifying the major coat protein as a single ~180-kDa polypeptide. She named it clathrin from the Latin clathratus (latticed), reflecting the polyhedral lattice she had imaged by negative-stain EM. The triskelion structure was visualized by Ernst Ungewickell and Daniel Branton in 1981 by rotary shadowing of dissociated coats. Frances Brodsky, then at UCSF, dissected the heavy/light chain composition and the regulation by light-chain phosphorylation in the 1980s. Tomas Kirchhausen at Harvard contributed the structural and assembly biophysics over decades. Live-cell TIRF imaging by Tom Kirchhausen, Sandra Schmid, and David Drubin revealed the ~30 to 60 second formation kinetics in real time.