Cell Biology
Lipid Rafts
Cholesterol- and sphingolipid-rich membrane microdomains — signaling platforms, caveolae, GPI-anchored proteins
Lipid rafts are cholesterol- and sphingolipid-rich ordered microdomains that float within the more fluid bulk of the plasma membrane, concentrating specific proteins into transient functional platforms. Enriched in cholesterol and saturated-tail sphingolipids, rafts adopt a tightly packed liquid-ordered (Lo) phase distinct from the surrounding liquid-disordered (Ld) membrane, so they selectively recruit GPI-anchored proteins and lipid-modified signaling molecules while excluding others. Typically 10 to 200 nm across and dynamic on a millisecond-to-second timescale, rafts coalesce on demand to platform signal transduction, form flask-shaped caveolae, and drive raft-mediated endocytosis. The concept was named by Kai Simons and Elina Ikonen in a 1997 Nature paper, building on Simons and van Meer's 1988 lipid-sorting model.
- Raft size~10–200 nm, dynamic
- Defining lipidscholesterol + sphingolipids
- Physical phaseliquid-ordered (Lo)
- Caveolae50–100 nm, caveolin-1 coat
- CoinedSimons & Ikonen 1997
- Raft-bustermethyl-β-cyclodextrin
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 lipid rafts matter
- They break the "random sea" picture of the membrane. The 1972 Singer–Nicolson fluid mosaic model treated the bilayer as a homogeneous 2D solvent in which proteins diffuse freely. Rafts add lateral heterogeneity: the plasma membrane is a patchwork of ordered and disordered domains that sorts proteins in the plane of the membrane, not just across it.
- They platform immune signaling. On T cells, engaging the T-cell receptor drives raft coalescence at the immunological synapse, concentrating the Src-family kinase Lck and the phosphorylated adaptor LAT to nucleate the activation cascade. Rafts similarly organize B-cell receptor, FcεRI (mast-cell allergy), and innate immune receptor signaling.
- Caveolae buffer mechanical stress. Muscle and endothelial cells pack 50-to-100-nm caveolae by the thousands; when the membrane is stretched, caveolae flatten out and release stored membrane area, protecting the cell from rupture. Loss of caveolin-3 causes limb-girdle muscular dystrophy and rippling muscle disease.
- Pathogens hijack the raft door. Simian Virus 40, cholera toxin (via ganglioside GM1), and several other toxins and viruses enter cells through cholesterol-rich domains and caveolae, bypassing the clathrin route the cell uses for most receptor uptake.
- Cholesterol handling is a raft problem. Because rafts depend on cholesterol, disorders of cholesterol trafficking (Niemann–Pick type C) and drugs that deplete cholesterol perturb raft-dependent signaling. Amyloid precursor protein processing that generates Alzheimer's Aβ peptide is enriched in rafts, tying membrane order to neurodegeneration.
- They explain a blood disease. Paroxysmal nocturnal hemoglobinuria arises from a somatic PIGA mutation that blocks GPI-anchor synthesis. Red cells then lose the raft-resident complement inhibitors CD55 and CD59 and are lysed by the patient's own complement — a direct clinical consequence of raft biology.
How lipid rafts work
The physics starts with cholesterol and sphingolipids. Sphingolipids (sphingomyelin, glycosphingolipids, gangliosides) carry long, mostly saturated fatty-acid tails that pack straight and tight. Cholesterol, a rigid four-ring sterol, slots its flat face against those saturated tails and fills the gaps, condensing the lipids into a liquid-ordered (Lo) phase: the acyl chains are extended and closely packed, so the membrane there is slightly thicker (by roughly 0.5–1 nm) and less fluid than the surrounding liquid-disordered (Ld) phase, which is rich in kinked, unsaturated phospholipid tails. Crucially the Lo phase is still a liquid — lipids diffuse laterally — so a raft is an ordered patch, not a solid.
Proteins are sorted by that order. A protein partitions into a raft if its membrane-facing surface matches the ordered environment. The strongest raft affinities come from lipid modifications: a GPI anchor in the outer leaflet (folate receptor, CD59, Thy-1, prion protein PrP), and palmitoylation or myristoylation of inner-leaflet proteins such as the Src-family kinases Lck and Fyn and the alpha subunits of heterotrimeric G proteins. Long, saturated transmembrane helices also favor the raft; proteins with bulky, kinked, or unsaturated environments are excluded. The two leaflets are coupled, so an outer-leaflet raft can register a partner assembly on the inner leaflet.
Function comes from coalescence. Resting rafts are small (10–200 nm) and short-lived, so their partner proteins are dispersed and signaling is off. A trigger — ligand binding, receptor cross-linking, antibody clustering — makes individual rafts fuse into a larger, more stable platform. That raises the local concentration of receptors, kinases, and adaptors, excludes inhibitory phosphatases, and lets the cascade fire fast and specifically. The T-cell receptor is the textbook example: engagement clusters rafts at the immunological synapse, bringing Lck and phosphorylated LAT together to ignite downstream signaling.
Caveolae are the stable, visible form of a raft. Caveolin-1 (or the muscle isoform caveolin-3) is an integral membrane protein that inserts a hairpin into the inner leaflet, binds cholesterol, and oligomerizes; peripheral cavin proteins (cavin-1/PTRF is required) coat the cytoplasmic face and sculpt the membrane into a 50-to-100-nm flask-shaped invagination. To internalize, a caveola recruits the large GTPase dynamin-2, which constricts and severs the neck, releasing a caveolar vesicle — the core of raft-mediated (caveolar) endocytosis, a clathrin-independent pathway used for transcytosis of albumin across endothelium and for the entry of SV40 and cholera toxin.
Lipid rafts vs bulk membrane vs caveolae
| Feature | Lipid raft (planar) | Bulk membrane (non-raft) | Caveola |
|---|---|---|---|
| Physical phase | Liquid-ordered (Lo) | Liquid-disordered (Ld) | Liquid-ordered (Lo) |
| Lipid enrichment | Cholesterol + sphingolipids | Unsaturated phospholipids | Cholesterol + sphingolipids |
| Shape | Flat patch | Flat | 50–100 nm flask invagination |
| Defining protein coat | None (lipid-defined) | None | Caveolin-1/-3 + cavins |
| Size / lifetime | ~10–200 nm, transient | Continuous phase | ~50–100 nm, stable |
| Detergent resistance | Resistant (DRM) | Solubilized | Resistant (DRM) |
| Marker proteins | GPI-anchored, Src kinases, flotillin | Transferrin receptor | Caveolin-1, cavin-1 |
| Main role | Signaling platform | Bulk lipid solvent | Endocytosis, transcytosis, tension buffer |
Raft/caveolar vs clathrin-mediated endocytosis
| Property | Raft/caveolar endocytosis | Clathrin-mediated endocytosis |
|---|---|---|
| Membrane domain | Cholesterol-rich ordered raft / caveola | Clathrin-coated pit (non-raft) |
| Coat | Caveolin-1 + cavins (or coatless) | Clathrin triskelions + AP-2 adaptors |
| Cholesterol dependence | Absolute — blocked by cholesterol depletion | Largely cholesterol-independent |
| Scission GTPase | Dynamin-2 | Dynamin-1/2 |
| Typical cargo | GPI proteins, GM1, albumin, SV40, cholera toxin | Transferrin, LDL, EGFR, most receptors |
| Vesicle size | ~50–100 nm | ~100–150 nm |
| Throughput / speed | Lower, slower | High, fast, constitutive |
| Destination | Early endosome / caveosome; some transcytose | Early endosome → sorting/recycling/degradation |
Common misconceptions
- Rafts are big, stable structures you can see under a microscope. Native rafts are nanoscale (10–200 nm) and short-lived, well below the diffraction limit; only caveolae and ligand-induced coalesced platforms approach visible scale. The consensus (2006 Keystone) definition explicitly calls them small and dynamic.
- A detergent-resistant membrane (DRM) equals a raft. Cold Triton X-100 extraction enriches ordered lipids, but the DRM is a biochemical artifact of the extraction, not a snapshot of a pre-existing domain. DRMs are useful evidence but do not prove a raft existed in the living membrane — this ambiguity fueled the decades-long controversy.
- Rafts and caveolae are the same thing. Caveolae are one morphological subtype of raft, defined by the caveolin/cavin coat and flask shape. Most cells' rafts are flat and coatless; cells that lack caveolin-1 (and thus have no caveolae) still have flat rafts.
- Cholesterol is bad, so depleting it is harmless. Cholesterol is structurally essential to the ordered phase. Extracting it with methyl-β-cyclodextrin dissolves rafts and shuts down raft-dependent signaling and endocytosis — a research tool, but a reminder that some membrane cholesterol is indispensable.
- GPI-anchored proteins can't signal because they have no cytoplasmic tail. They have no transmembrane or cytoplasmic domain, yet cross-linking a GPI-anchored protein transmits signals inward — apparently through raft coalescence engaging inner-leaflet Src-family kinases, a striking case of outside-in signaling with no direct protein bridge.
- Rafts only exist in the plasma membrane. The lipid-sorting logic operates through the secretory pathway; sphingolipid and cholesterol enrichment increases from the ER toward the plasma membrane, and raft-like sorting in the trans-Golgi network helps route GPI-anchored proteins to the cell surface.
Famous experiments and history
- Simons & van Meer (1988). Gerrit van Meer and Kai Simons proposed that glycosphingolipids cluster with cholesterol in the trans-Golgi network to sort apical membrane proteins in polarized epithelial cells — the intellectual seed of the raft concept, before the word "raft" existed.
- Simons & Ikonen (1997). Kai Simons and Elina Ikonen's Nature review "Functional rafts in cell membranes" named and framed lipid rafts as cholesterol/sphingolipid platforms for sorting and signaling. It became one of the most-cited papers in membrane biology and launched the field.
- Brown & Rose DRM isolation (1992). Deborah Brown and John Rose showed that GPI-anchored proteins resist cold Triton X-100 solubilization and float to a low density on sucrose gradients, giving the field its workhorse biochemical assay — and, later, its central controversy over whether DRMs reflect real domains.
- Model-membrane phase separation. Giant unilamellar vesicles made from cholesterol plus a saturated and an unsaturated lipid visibly separate into micron-scale liquid-ordered and liquid-disordered domains, imaged by fluorescence microscopy (e.g., work from the Feigenson, Baumgart, and Keller labs). This proved the underlying physical chemistry of coexisting liquid phases is real.
- Caveolae and caveolin. Caveolae were described by electron microscopy in the 1950s (Palade; Yamada, 1955), but caveolin-1 was identified as their coat protein only in the 1990s. Cavin-1/PTRF was later shown to be required for caveola formation; its loss produces lipodystrophy and muscular dystrophy in humans, cementing caveolae as bona fide functional organelles.
- Cholera toxin and SV40 entry. Tracking cholera toxin B-subunit (which binds raft ganglioside GM1) and Simian Virus 40 showed both concentrate in caveolae/rafts and enter cells by a cholesterol-dependent, dynamin-2-requiring, clathrin-independent route — the defining functional demonstration of raft-mediated endocytosis.
Frequently asked questions
What is a lipid raft?
A lipid raft is a small, dynamic region of the cell membrane that is enriched in cholesterol and sphingolipids and packed more tightly than the surrounding bilayer. Because the sphingolipids carry long, saturated fatty-acid tails and cholesterol slots between them, the raft adopts a liquid-ordered (Lo) phase: the lipids are still mobile but their tails are extended and closely packed, making the raft slightly thicker and less fluid than the disordered (Ld) membrane around it. This ordered patch selectively recruits certain proteins — GPI-anchored proteins, palmitoylated and myristoylated signaling molecules, and specific receptors — while excluding others, so the raft acts as a sorting platform. Rafts are typically 10 to 200 nm across and transient, coalescing into larger functional platforms when a receptor is triggered. The term was coined by Kai Simons and Elina Ikonen in 1997.
What is the difference between lipid rafts and caveolae?
Caveolae are a specialized, morphologically visible subtype of lipid raft. Ordinary (planar or flat) rafts are flat patches of liquid-ordered membrane that are too small and dynamic to see directly and are detected biochemically or by super-resolution imaging. Caveolae, by contrast, are stable 50-to-100-nm flask-shaped invaginations of the plasma membrane that are coated on their cytoplasmic face by the integral membrane protein caveolin-1 (with caveolin-2 or -3) and stabilized by peripheral cavin proteins (cavin-1/PTRF is essential). Both share the cholesterol- and sphingolipid-rich, detergent-resistant character, but only caveolae have a fixed shape and a defining protein coat. Caveolae are abundant in adipocytes, endothelial cells, and muscle, where they buffer membrane tension and mediate transcytosis and specialized endocytosis; flat rafts are ubiquitous and mostly organize signaling. Cells lacking caveolin-1 form no caveolae but still have flat rafts.
How do lipid rafts organize cell signaling?
Rafts act as pre-assembly platforms that raise the local concentration of partner molecules so signaling can fire fast and specifically. Many receptors and their downstream effectors carry lipid modifications — a GPI anchor, palmitoylation, or myristoylation — that preferentially partition them into the ordered raft phase, while inhibitory phosphatases and antagonists are often excluded. When a ligand binds, individual rafts coalesce into a larger platform, bringing receptors, non-receptor kinases such as the Src-family members Lck and Fyn, adaptor proteins like LAT, and G proteins into contact. The clearest case is the T-cell receptor: engagement drives raft clustering at the immunological synapse, concentrating Lck and phosphorylated LAT to nucleate the signaling cascade. Extracting membrane cholesterol with methyl-beta-cyclodextrin dissolves the ordered phase and blocks many of these pathways, which is the classic experimental proof of raft dependence.
What are GPI-anchored proteins and why do they cluster in rafts?
GPI-anchored proteins are cell-surface proteins tethered to the membrane not by a transmembrane helix but by a glycosylphosphatidylinositol (GPI) anchor — a glycolipid, attached post-translationally in the ER, whose two lipid tails insert into the outer leaflet. Because those tails are typically long and saturated, GPI-anchored proteins such as the folate receptor, alkaline placental phosphatase, CD59, Thy-1, and the prion protein PrP partition strongly into the liquid-ordered raft phase and are classic raft markers. Their clustering matters functionally: cross-linking a GPI-anchored protein can transmit signals across the bilayer even though the protein has no cytoplasmic domain, apparently by raft coalescence engaging inner-leaflet Src-family kinases. Defective GPI-anchor synthesis causes paroxysmal nocturnal hemoglobinuria, in which red cells lose the raft-resident complement inhibitors CD55 and CD59 and are destroyed by complement.
What is raft-mediated (caveolar) endocytosis?
Raft-mediated endocytosis is a clathrin-independent uptake pathway in which cargo concentrated in cholesterol-rich ordered domains is internalized, most visibly through caveolae. A caveola pinches off from the plasma membrane in a process requiring the large GTPase dynamin-2, which severs the neck, and forms a caveolar vesicle that can deliver cargo to early endosomes or the caveosome/late-endosomal compartments. This route is exploited by several pathogens and toxins: Simian Virus 40 (SV40) and cholera toxin (which binds raft ganglioside GM1) both enter cells largely via caveolae. Endothelial caveolae also carry out transcytosis, ferrying albumin and other macromolecules across the blood vessel wall. The pathway is cholesterol-dependent, so it is blocked by cholesterol-extracting agents and by inhibitors of dynamin, and it is generally slower and lower-throughput than clathrin-mediated endocytosis.
Are lipid rafts real, and why were they controversial?
Lipid rafts are now widely accepted, but for roughly two decades their existence was contested because the early evidence relied on detergent-resistant membranes (DRMs) — the fraction that survives cold Triton X-100 extraction. Critics, notably Kai Simons's frequent debate partner, argued that cold detergent could artificially create or expand ordered domains, so a DRM did not prove a raft existed in the living membrane. Two developments settled the question. First, model membranes (giant unilamellar vesicles) made from cholesterol plus saturated and unsaturated lipids visibly phase-separate into micron-scale liquid-ordered and liquid-disordered domains, proving the physical chemistry is real. Second, super-resolution microscopy, single-molecule tracking, and FRET in living cells detected nanoscale (10 to 200 nm), short-lived ordered assemblies. A 2006 Keystone Symposium redefined rafts as small, dynamic, cholesterol- and sphingolipid-enriched domains that can coalesce, which is the consensus definition used today.