Cell Biology
COPI and COPII Vesicle Transport
The two protein coats that shuttle cargo between the ER and Golgi — Sar1, Sec23/24/13/31, ARF1, and KDEL retrieval
COPI and COPII vesicles are the two protein-coated carriers that shuttle cargo between the endoplasmic reticulum and the Golgi apparatus. COPII coats bud anterograde carriers that leave the ER, built by the small GTPase Sar1 plus the Sec23/Sec24 inner adaptor and the Sec13/Sec31 outer cage; COPI coats bud retrograde carriers under the GTPase ARF1 that retrieve escaped ER-resident proteins bearing the KDEL signal. The COPII coat was reconstituted in vitro from yeast by Randy Schekman's laboratory — whose sec mutant screen won a share of the 2013 Nobel Prize — while James Rothman's group defined COPI biochemistry from mammalian Golgi. Roughly one-third of the eukaryotic proteome, from insulin to procollagen, transits through these coats on its way out of the cell.
- DirectionCOPII = ER→Golgi, COPI = Golgi→ER
- COPII GTPaseSar1 (GEF: Sec12)
- COPI GTPaseARF1 (coatomer coat)
- Retrieval signalsKDEL (soluble), KKXX (membrane)
- Vesicle size~60–90 nm
- Nobel PrizeSchekman & Rothman, 2013
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 COPI and COPII transport matters
- It is the front door of the secretory pathway. Roughly one-third of all proteins a eukaryotic cell makes are destined for membranes, lysosomes, or secretion, and essentially every one of them leaves the ER inside a COPII vesicle. Insulin, antibodies, collagen, digestive enzymes, and every cell-surface receptor begins its journey this way.
- It enforces quality control. Only correctly folded, correctly assembled proteins expose their export signals and get packaged by Sec24. Misfolded proteins are held back by chaperones such as BiP and either refolded or routed to ER-associated degradation — so COPII export is a licensing checkpoint, not a conveyor belt.
- It keeps the ER from emptying itself. Bulk anterograde flow constantly sweeps ER-resident chaperones and enzymes forward. COPI retrograde retrieval, guided by the KDEL receptor and di-lysine motifs, drags them back. Without it, the ER would lose its own folding machinery within a few cell divisions.
- Coat mutations cause named human diseases. SEC23A mutations cause cranio-lenticulo-sutural dysplasia; SAR1B mutations cause chylomicron retention disease (fat malabsorption); SEC23B mutations cause congenital dyserythropoietic anemia type II; COPA mutations cause an autoimmune lung-and-joint syndrome. Each disease is a specific step of this pathway breaking.
- Big cargo needs special machinery. Procollagen leaves the ER as a rigid rod 300 to 400 nm long — far too big for a standard 90 nm vesicle. The receptor TANGO1/MIA3 at the ER exit site builds enlarged COPII carriers, which is why connective-tissue cells depend so heavily on this pathway.
- It is a drug and toxin target. Brefeldin A, a fungal metabolite, locks ARF1 in its inactive GDP form, blocks COPI budding, and causes the Golgi to collapse back into the ER within minutes — a textbook demonstration that the Golgi is maintained by continuous retrograde traffic.
- It defines organelle identity. The steady balance of anterograde COPII output and retrograde COPI retrieval is what keeps the ER, ERGIC, and Golgi as distinct compartments with distinct protein compositions despite constant membrane exchange between them.
Common misconceptions
- COPII vesicles fuse directly with the Golgi. In most animal cells they do not fuse straight onto a Golgi cisterna. They first deliver cargo to the ER-Golgi intermediate compartment (ERGIC, also called VTCs), a sorting station where anterograde cargo is separated from retrograde cargo before onward transport. The Golgi is reached only after this intermediate step.
- The KDEL receptor keeps ER proteins from ever leaving the ER. It does not prevent escape — it retrieves. BiP and protein disulfide isomerase are constantly swept into the Golgi by bulk flow; the KDEL receptor captures them in the cis-Golgi and returns them by COPI. Retrieval, not retention, is the mechanism, which is why KDEL proteins are found transiently in the Golgi at all.
- COPI is the coat for endocytosis. No — the plasma-membrane and trans-Golgi endocytic coat is clathrin, working with adaptor complexes such as AP-2. COPI and COPII are the non-clathrin coats dedicated to the early secretory pathway between ER and Golgi. Confusing them is one of the most common exam errors.
- The coat both selects cargo and provides the driving force in one protein. The jobs are split across layers. In COPII, the inner Sec23/Sec24 layer selects cargo and hydrolyzes the GTPase, while the outer Sec13/Sec31 cage provides the structural scaffold that bends the membrane. Assembly is sequential, not a single event.
- Higher numbers mean later discovery. COPI was actually characterized biochemically before COPII, so the numbers do not follow the direction of transport or the order of discovery. COPII simply moves cargo "out" of the ER and COPI brings it back "in," and the numbering is a historical accident that trips up nearly every student.
- GTP hydrolysis is only for uncoating. It is a timer with two jobs. On COPII, Sec23 (the GAP) and Sec31 accelerate Sar1's GTP hydrolysis, which both times coat release and acts as a proofreading step — coats that assemble before cargo is loaded tend to hydrolyze and fall off before they can bud, biasing the system toward loaded vesicles.
How COPI and COPII transport works
Anterograde export begins at discrete patches of ER membrane called ER exit sites, marked by the scaffolding protein Sec16. The trigger is nucleotide exchange on the small GTPase Sar1: the ER-anchored guanine-nucleotide exchange factor Sec12 swaps GDP for GTP on Sar1. GTP loading exposes an N-terminal amphipathic helix on Sar1 that inserts into the outer leaflet of the ER bilayer, anchoring the nascent coat and beginning to bend the membrane. Membrane-bound Sar1-GTP then recruits the Sec23/Sec24 heterodimer — the inner coat layer. Sec24 is the cargo adaptor: it carries several independent binding sites that recognize export signals on cargo, including the di-acidic DxE motif (found on the model cargo VSV-G glycoprotein), di-hydrophobic FF or LxxLE motifs, and other short signals. Mammals express four Sec24 paralogs (SEC24A–D) with overlapping cargo repertoires, which is how a single coat can gather thousands of different proteins.
With cargo concentrated by the inner layer, the Sec13/Sec31 heterotetramer polymerizes into a self-assembling cuboctahedral outer cage that scaffolds the growing bud and drives further membrane curvature. The completed prebudding complex sculpts a vesicle roughly 60 to 90 nm across, pinches it off, and — critically — carries the trigger for its own disassembly. Sec23 is the GTPase-activating protein for Sar1, and Sec31 stimulates that activity about tenfold once the full coat is assembled. When Sar1 hydrolyzes GTP to GDP, its membrane-anchoring helix retracts, Sar1 releases, and the coat sheds. The naked vesicle then fuses with the ERGIC using the Rab1 GTPase and the SNARE proteins (syntaxin-5, membrin, rBet1, Sec22b), delivering its anterograde cargo.
Retrograde transport runs the machinery in reverse under a different GTPase. ARF1, activated by a Golgi GEF (of the GBF1/BIG family) and carrying an N-terminal myristoyl anchor, embeds in the Golgi or ERGIC membrane and recruits the seven-subunit coatomer (α-, β-, β'-, γ-, δ-, ε-, ζ-COP) en bloc. Coatomer selects retrograde cargo by two routes: it directly binds the cytosolic di-lysine KKXX or KXKX motif on ER-resident membrane proteins, and it accepts soluble cargo handed off by the KDEL receptor. The KDEL receptor grips the C-terminal KDEL tail of escaped chaperones like BiP in the mildly acidic Golgi (around pH 6.2) and releases it in the near-neutral ER (around pH 7.2), so the receptor shuttles loaded to the ER and empty back to the Golgi. As with COPII, GTP hydrolysis — triggered by ArfGAP proteins — retracts ARF1's anchor, sheds coatomer, and readies the vesicle to fuse back with the ER.
COPI vs COPII vs clathrin coats
| Feature | COPII | COPI | Clathrin |
|---|---|---|---|
| Direction | Anterograde (ER → ERGIC/Golgi) | Retrograde (Golgi/ERGIC → ER) | Endocytic / TGN sorting |
| Small GTPase | Sar1 | ARF1 | ARF1/ARF6 (for adaptors) |
| GEF | Sec12 | GBF1 / BIG1/2 | — |
| Coat subunits | Sec23/24 + Sec13/31 | Coatomer (7 subunits, α–ζ COP) | Clathrin triskelia + AP adaptors |
| Cargo signal read | DxE, FF/LxxLE (via Sec24) | KKXX/KXKX; KDEL via KDELR | YXXΦ, [DE]XXXL[LI] (via AP) |
| Main site | ER exit sites (Sec16) | cis-Golgi, ERGIC | Plasma membrane, TGN, endosomes |
| Uncoating trigger | Sar1 GTP hydrolysis (Sec23 GAP) | ARF1 GTP hydrolysis (ArfGAP) | Hsc70 + auxilin, PIP loss |
| Classic inhibitor / tool | Sar1 GTP-locked mutant (H79G) | Brefeldin A (blocks ARF1 GEF) | Dynasore, chlorpromazine |
Anterograde COPII vs retrograde COPI in detail
| Property | COPII (anterograde) | COPI (retrograde) |
|---|---|---|
| Buds from | ER exit sites | cis-Golgi and ERGIC |
| Delivers to | ERGIC, then cis-Golgi | ER (and within-Golgi retrograde) |
| Cargo carried | Newly folded secretory / membrane proteins | Escaped ER residents; recycled machinery |
| Cargo receptor | Sec24 (four paralogs in mammals) | Coatomer + KDEL receptor |
| Signal decoded | DxE, di-hydrophobic export motifs | KDEL (soluble); KKXX/KXKX (membrane) |
| Membrane curvature | Sar1 helix + Sec13/31 outer cage | ARF1 myristoyl helix + coatomer |
| GAP for the GTPase | Sec23 (boosted by Sec31) | ArfGAP1/2/3 |
| Big-cargo adaptation | TANGO1/MIA3 builds enlarged carriers for procollagen | Not applicable — recycling small carriers |
| Human disease example | CLSD (SEC23A); chylomicron retention (SAR1B); CDA-II (SEC23B) | COPA syndrome (COPA) |
Famous experiments and history
- Schekman's yeast sec screen (1979–1980). Peter Novick and Randy Schekman isolated temperature-sensitive Saccharomyces cerevisiae mutants that stopped secreting and bloated with internal membranes, ordering the secretory pathway genetically. The screen eventually named more than twenty sec genes; sec12, sec13, sec16, sec23, sec24, and sec31 encode the COPII machinery, and SAR1 its GTPase.
- In vitro reconstitution of COPII (1994). Charles Barlowe, Lelio Orci, Randy Schekman, and colleagues showed that purified Sar1, Sec23/24, and Sec13/31 plus GTP were sufficient to bud synthetic COPII vesicles from ER-derived membranes — proving the gene list was complete and defining the minimal budding machine (Cell 77: 895–907).
- Rothman's coatomer and ARF biochemistry. James Rothman, Lelio Orci, and co-workers purified the COPI coatomer and ARF1 from mammalian Golgi membranes in the late 1980s and early 1990s, reconstituted non-clathrin-coated budding, and established that GTP hydrolysis governs coat cycling — the biochemical complement to Schekman's genetics.
- Brefeldin A and the ARF1 switch. Brefeldin A was shown to trap the ARF1–GEF complex in an abortive GDP-bound state, blocking COPI assembly and causing the Golgi to redistribute into the ER within minutes. It became the definitive tool proving that Golgi identity depends on continuous COPI retrograde traffic.
- The 2013 Nobel Prize. Randy Schekman, James Rothman, and Thomas Sudhof shared the Nobel Prize in Physiology or Medicine "for their discoveries of machinery regulating vesicle traffic, a major transport system in our cells" — Schekman for the genetics of secretion, Rothman for the biochemistry of coats and fusion, and Sudhof for the timing of neurotransmitter release.
Frequently asked questions
What is the difference between COPI and COPII vesicles?
The two coats move cargo in opposite directions between the same two compartments. COPII carriers bud from the endoplasmic reticulum at ER exit sites and travel anterograde — forward, toward the Golgi — carrying newly synthesized secretory and membrane proteins. Their assembly is nucleated by the small GTPase Sar1 and built from Sec23/Sec24 (the inner cargo-selecting layer) and Sec13/Sec31 (the outer structural cage). COPI carriers bud from the Golgi and the ER-Golgi intermediate compartment and travel retrograde — backward, toward the ER — retrieving escaped ER-resident proteins and recycling trafficking machinery. COPI assembly is nucleated by the GTPase ARF1 and built from the seven-subunit coatomer complex. A useful mnemonic: COPII goes 'out' from the ER (roughly two arrows out), COPI comes back 'in' toward the ER. Both are non-clathrin coats, distinct from the clathrin coats that operate at the plasma membrane and trans-Golgi network.
How does the COPII coat assemble on the ER membrane?
Assembly is a sequential, GTP-driven program at ER exit sites. First, the ER-anchored guanine-nucleotide exchange factor Sec12 loads GTP onto the small GTPase Sar1. GTP-Sar1 exposes an amphipathic N-terminal helix that inserts into the outer leaflet of the ER membrane, anchoring the coat and beginning to curve the bilayer. Membrane-bound Sar1-GTP then recruits the Sec23/Sec24 heterodimer; Sec23 is the GTPase-activating protein (GAP) for Sar1, and Sec24 is the cargo adaptor that reads export signals such as the di-acidic DxE motif and hydrophobic patches on transmembrane cargo. Finally the Sec13/Sec31 heterotetramer polymerizes into a cuboctahedral outer cage that scaffolds the growing bud and further deforms the membrane. The whole prebudding complex concentrates cargo and sculpts a 60 to 90 nm vesicle.
What is the KDEL signal and how does retrograde retrieval work?
KDEL (lysine-aspartate-glutamate-leucine) is a four-residue C-terminal signal on soluble ER-resident proteins such as the chaperone BiP/GRP78 and protein disulfide isomerase. These proteins are constantly swept forward into the Golgi by bulk flow. The KDEL receptor (KDELR/ERD2) in the cis-Golgi binds the KDEL tail, and the binding is pH-sensitive: the receptor grips KDEL in the mildly acidic Golgi (around pH 6.2) and releases it in the near-neutral ER lumen (around pH 7.2). Ligand-bound KDELR is packaged into ARF1-driven COPI vesicles and carried retrograde back to the ER, where the higher pH releases the cargo and the empty receptor recycles forward again. Membrane ER-resident proteins are retrieved differently: they carry a cytosolic C-terminal KKXX or KXKX (di-lysine) motif that COPI coatomer binds directly. This retrieval system corrects the leakiness of bulk anterograde flow and keeps the ER stocked with its own proteins.
How do the coats know which cargo to take?
Cargo selection is signal-based, not random. On the anterograde side, Sec24 is the primary cargo receptor of the COPII coat and has multiple independent binding sites that recognize distinct export motifs: the di-acidic DxE motif (as on VSV-G glycoprotein), the di-hydrophobic FF or LxxLE motifs, and the YNNSNPF-type signals. Mammals express four Sec24 paralogs (SEC24A to D) with overlapping but non-identical cargo preferences, which is how one coat can select thousands of different cargoes. Large or bulky cargo such as procollagen (a rigid 300 to 400 nm rod that cannot fit a standard 90 nm vesicle) needs the accessory protein TANGO1/MIA3 at the ER exit site to build an enlarged carrier. On the retrograde side, COPI coatomer directly reads di-lysine KKXX or KXKX motifs on membrane proteins, and the KDEL receptor hands soluble KDEL-bearing proteins to the same coat. Bulk-flow cargo without a signal is taken less efficiently, at roughly the background concentration of the ER lumen.
Who discovered the COP vesicle coats?
The genetics came first. In the late 1970s and 1980s Randy Schekman's laboratory at Berkeley ran a screen for temperature-sensitive yeast mutants that accumulated secretory proteins internally, isolating dozens of sec genes that mapped the secretory pathway; sec12, sec13, sec16, sec23, sec24, and sec31 turned out to encode the COPII machinery, and Sar1 was identified as its GTPase. Schekman's group then reconstituted COPII budding from purified components in vitro in the early 1990s, proving those genes were sufficient to make a vesicle. In parallel, James Rothman's laboratory purified the COPI coatomer and its GTPase ARF1 biochemically from mammalian Golgi membranes, defining coat assembly and the role of GTP hydrolysis in budding. Schekman and Rothman shared the 2013 Nobel Prize in Physiology or Medicine with Thomas Sudhof for discoveries of the machinery regulating vesicle traffic. Confusingly for students, COPI was characterized biochemically before COPII, which is why the numbering does not follow the direction of transport.
How does a COPII vesicle shed its coat after budding?
Uncoating is timed by GTP hydrolysis on the small GTPase. Sar1 is a GTPase, but on its own it hydrolyzes GTP very slowly; the Sec23 subunit of the inner coat is its dedicated GTPase-activating protein, and Sec31 of the outer cage further stimulates that activity roughly tenfold when the full coat is assembled. So the completed, cargo-loaded coat contains the very trigger for its own disassembly — a built-in timer. When Sar1 hydrolyzes GTP to GDP, its membrane-anchoring amphipathic helix retracts, Sar1 leaves the membrane, and the coat loses its grip and falls apart. The naked vesicle can then fuse with its target, the ER-Golgi intermediate compartment, using Rab1 and the SNARE proteins. COPI uncoating works the same way in principle: ARF1's GAP (ArfGAP1/2/3) triggers GTP hydrolysis, ARF1's myristoylated helix retracts, and coatomer releases. The timer design means a coat that assembles too fast, before cargo is loaded, tends to fall off before it can bud — a proofreading effect.
What happens to human health when COPI or COPII transport fails?
Because one-third of the proteome transits this route, coat mutations cause distinct human diseases. Mutations in SEC23A cause cranio-lenticulo-sutural dysplasia (Boyadjiev-Jabs syndrome), a skeletal and facial disorder driven by impaired procollagen export from chondrocytes and osteoblasts. Mutations in SAR1B cause chylomicron retention disease (Anderson disease), in which intestinal cells cannot bud the large COPII carriers needed to export dietary-fat-laden chylomicrons, producing fat malabsorption and failure to thrive. SEC23B mutations cause congenital dyserythropoietic anemia type II (CDA-II), where defective ER export disrupts red-cell membrane glycoprotein maturation. On the COPI side, mutations in COPA cause COPA syndrome, an autoimmune interstitial lung and joint disease driven by defective retrograde retrieval and consequent ER stress. Several pathogens exploit the pathway too: Brefeldin A, a fungal toxin, freezes ARF1 in its GDP state, collapses the Golgi back into the ER, and is a classic laboratory tool for probing secretion.