Cell Biology
Aquaporins (Water Channels)
Membrane channels that rush water through — while blocking every ion and even the proton
Aquaporins are integral membrane channel proteins that conduct water across the lipid bilayer at up to 3 billion molecules per second per channel, yet exclude protons and every ion. Each hourglass-shaped pore threads water in single file through a constriction only about 2.8 angstroms wide, and a pair of asparagine–proline–alanine (NPA) motifs at the pore center forces each passing water to flip its orientation — breaking the hydrogen-bonded chain a proton would need to hop along. Peter Agre isolated the first channel (CHIP28, now AQP1) from red-cell and kidney membranes in 1992 and won the 2003 Nobel Prize in Chemistry. Humans carry thirteen isoforms (AQP0–AQP12) that set water balance in the kidney, eye lens, brain, red cells, and secretory glands — and AQP2, controlled by the hormone vasopressin, is the switch that decides how concentrated your urine becomes.
- Throughput~3 billion H₂O/s per channel
- Pore width~2.8 Å (single-file)
- Selectivity motifdual NPA + ar/R constriction
- DiscoveredAgre, CHIP28, 1992
- NobelChemistry 2003 (Agre)
- Human isoforms13 (AQP0–AQP12)
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 aquaporins matter
- They solve a hidden paradox of life. The lipid bilayer is a decent water barrier, but red cells and kidney tubules move water tens of times faster than raw diffusion allows. For decades physiologists insisted membranes must contain dedicated water pores; aquaporins are those pores, and they conduct water at a rate approaching the diffusion limit while still refusing to leak charge.
- They set your urine concentration. The kidney filters about 180 liters of plasma per day, and you excrete only around 1.5 liters. Almost all of that reabsorbed water crosses through aquaporins — AQP1 in the proximal tubule and descending limb (constitutive) and AQP2 in the collecting duct (hormone-controlled). Regulating AQP2 lets a mammal shift from dilute to maximally concentrated urine and back within an hour.
- They keep the lens transparent. AQP0 (major intrinsic protein, MIP) is the single most abundant protein in eye-lens fiber cells. It both conducts water and physically tethers neighboring fibers into a tight, low-scatter lattice. Mutations in AQP0 cause hereditary congenital cataracts because water regulation and cell packing both fail.
- They govern brain water and edema. AQP4 studs the astrocyte end-feet wrapping brain capillaries. It drives water into and out of the brain, so it dominates the swelling of cytotoxic edema after stroke and the clearance in vasogenic edema. AQP4 is also the autoantigen attacked by antibodies in neuromyelitis optica (Devic disease).
- They dry you out — or don't. AQP5 powers fluid secretion in salivary, lacrimal, and airway glands. Its loss or autoimmune targeting (as in Sjögren syndrome) produces dry mouth and dry eyes. A related subfamily, the aquaglyceroporins (AQP3, AQP7, AQP9, AQP10), also passes glycerol, linking these channels to skin hydration, fat mobilization, and metabolism.
- They are ancient and universal. Aquaporins appear across bacteria, archaea, plants, fungi, and animals. Plants alone encode dozens (the maize genome has more than thirty), and plant PIP channels open and close to manage water during drought — one of the few aquaporins that is truly gated rather than simply trafficked.
Common misconceptions
- "Aquaporins pump water." They do not. They are passive channels that conduct water only down its osmotic gradient and in whichever direction that gradient points. No ATP is hydrolyzed and no per-molecule gating cycle occurs. The energy that ultimately moves the water comes from solute pumps like the sodium–potassium pump elsewhere in the tissue.
- "Water just diffuses across membranes, so channels are unnecessary." Water does slowly cross a bare bilayer, but far too slowly for red cells, kidney, or lens. Aquaporins raise a membrane's osmotic water permeability (Pf) 10- to 100-fold. Where you need bulk water flux — reabsorbing 179 of 180 filtered liters — diffusion alone cannot keep up.
- "Protons are tiny, so a water channel must leak them." Protons don't travel as free particles through water; they relay along a hydrogen-bonded chain by the Grotthuss mechanism. The dual NPA motifs and the two half-helix dipoles at the pore center reorient each passing water and break that chain, so there is no wire for the proton to hop along even though the water flows freely.
- "All aquaporins only pass water." The aquaglyceroporin branch (AQP3, AQP7, AQP9, AQP10, and bacterial GlpF) has a wider, less charged pore that also conducts glycerol and other small uncharged solutes. Some aquaporins additionally pass gases (CO₂, NH₃), hydrogen peroxide, or urea, depending on isoform.
- "Aquaporins are gated open and shut like ion channels." Most mammalian aquaporins are regulated by trafficking — inserting or removing the whole channel from the membrane (AQP2 is the textbook case) — not by opening a gate. True conformational gating exists mainly in plant PIP channels (closed by dephosphorylation and low pH during drought) and yeast AQY1.
- "AQP1 is the kidney's water channel, so it makes urine concentrated." AQP1 is constitutive and unregulated; it handles the bulk isosmotic reabsorption in the proximal tubule and thin descending limb. The hormone-tunable step that actually sets final urine concentration is AQP2 in the collecting duct, under vasopressin control.
How aquaporins work
An aquaporin monomer folds into six transmembrane alpha-helices connected by five loops, plus two short half-helices (from loops B and E) that dip into the membrane from opposite sides and meet in the middle. This gives each subunit an hourglass shape: wide vestibules at both membrane surfaces funnel toward a single narrow channel about 20 angstroms long. Four monomers assemble into a homotetramer, but — unusually — each of the four monomers has its own independent water pore, so the tetramer carries four channels (a central fifth pore in the tetramer may conduct gases or ions in some isoforms).
Two features enforce selectivity, positioned in series along the pore. Near the extracellular mouth sits the aromatic/arginine (ar/R) constriction, the narrowest point at roughly 2.8 angstroms — just above the diameter of a water molecule. A conserved arginine there presents a fixed positive charge that electrostatically repels cations (including the proton's carrier, the hydronium ion), while the sheer tightness excludes any hydrated ion, which cannot shed its water shell to fit. Deeper in, at the pore center, the two half-helices point their positive N-terminal dipoles inward, and their tips carry the tandem asparagine–proline–alanine (NPA) motifs. The two asparagine side chains donate hydrogen bonds to the oxygen of the water molecule passing through.
The elegant part is what this does to proton transport. In bulk water, a proton hops rapidly along a continuous hydrogen-bonded chain — the Grotthuss (or "proton wire") mechanism — without any single molecule moving far. Inside the aquaporin, water crosses in single file, one molecule at a time. At the NPA junction the reversed helical dipoles and the asparagine hydrogen bonds force each water to flip its hydrogens away from the pore axis, so the water dipoles above and below the center point in opposite directions. That reorientation breaks the continuous donor–acceptor chain: there is no unbroken proton wire, so a proton cannot relay through even though water molecules keep flowing. The net result is a channel that conducts water at up to ~3 billion molecules per second while blocking H⁺ and every ion.
Because the transport is purely passive, the direction and rate depend only on the osmotic gradient across the membrane. Cells therefore regulate aquaporins not by gating individual molecules but by controlling how many channels sit in the membrane. The canonical example is AQP2 in the kidney collecting duct: vasopressin (ADH) binds the V2 receptor, a Gs-coupled GPCR, raising cyclic AMP and activating protein kinase A, which phosphorylates AQP2 at Ser256. Phosphorylated AQP2, held in subapical storage vesicles, is trafficked to and fused into the apical membrane, opening a high-conductance path for water to leave the tubule and concentrate the urine. When ADH falls, the channels are re-internalized by endocytosis.
Aquaporins vs pumps, transporters, and the bare bilayer
| Feature | Aquaporin (water channel) | Na⁺/K⁺ pump | Facilitated transporter (e.g. GLUT) | Bare lipid bilayer |
|---|---|---|---|---|
| Energy source | None (passive, osmotic) | ATP hydrolysis | None (down gradient) | None |
| Directionality | Bidirectional, follows gradient | Against gradient (uphill) | Down gradient | Down gradient |
| Per-cycle conformational change | No — open pore | Yes, large | Yes (alternating access) | N/A |
| Throughput | ~10⁹–3×10⁹ molecules/s | ~10²–10³ ions/s | ~10³–10⁴ molecules/s | slow, unassisted |
| What it moves | Water (± glycerol, gases) | 3 Na⁺ out / 2 K⁺ in | Specific solute | Anything, slowly |
| Selectivity trick | NPA reorientation + ar/R constriction | Ion-coordination sites | Binding-site geometry | Solubility-diffusion |
| Regulation | Trafficking / abundance (AQP2) | Kinase, hormones | Trafficking (GLUT4) | Membrane composition |
Human aquaporin isoforms and where they act
| Isoform | Main location | Also conducts | Physiology / disease link |
|---|---|---|---|
| AQP0 (MIP) | Eye lens fiber cells | Water (low Pf); adhesion role | Most abundant lens protein; mutations → congenital cataract |
| AQP1 (CHIP28) | Red cells, proximal tubule, thin descending limb, choroid plexus, endothelium | CO₂, NO (debated) | The founding water channel; constitutive bulk reabsorption |
| AQP2 | Collecting-duct apical membrane | Water only | Vasopressin-regulated; mutations → nephrogenic diabetes insipidus |
| AQP3 | Collecting-duct basolateral, skin, gut | Glycerol, urea (aquaglyceroporin) | Water exit to blood; skin hydration |
| AQP4 | Astrocyte end-feet, ependyma, collecting-duct basolateral | Water only | Brain water & edema; autoantigen in neuromyelitis optica |
| AQP5 | Salivary, lacrimal, airway secretory cells | Water; CO₂ | Gland fluid secretion; targeted in Sjögren syndrome |
| AQP7 | Adipose tissue, testis, kidney | Glycerol (aquaglyceroporin) | Fat metabolism, glycerol release during fasting |
Famous experiments and history
- The accidental discovery (1988–1992). Peter Agre's lab at Johns Hopkins was purifying the Rh blood-group antigen from red-cell membranes when a mysterious 28-kilodalton protein kept co-purifying. They named it CHIP28 (channel-forming integral protein), sequenced it, and realized it was also abundant in kidney tubules — exactly where physiologists had long predicted water pores must exist.
- The oocyte-swelling proof (Preston, Carroll, Guggino, Agre, Science 1992). They injected CHIP28 messenger RNA into Xenopus oocytes and dropped the cells into hypotonic saline. CHIP28 oocytes swelled and burst within minutes; water-injected controls barely changed. Mercuric chloride, which reacts with the pore-lining cysteine Cys189, abolished the permeability, and a Cys189→Ser mutant became mercury-insensitive — nailing the pore location. The protein was renamed aquaporin-1.
- The atomic structure (2000–2001). Electron and X-ray crystallography of AQP1 and the bacterial glycerol channel GlpF revealed the hourglass fold, the tetramer with four independent pores, and the NPA motifs at the pore center. Molecular-dynamics simulations (Tajkhorshid, Schulten, and colleagues, 2002) then showed water molecules single-filing through and flipping their orientation at the NPA junction — visualizing the proton-exclusion mechanism directly.
- The 2003 Nobel Prize in Chemistry. Agre received half the prize "for the discovery of water channels," sharing it with Roderick MacKinnon, honored for structural and mechanistic studies of ion channels. It was one of the fastest Nobel recognitions relative to a discovery in modern chemistry — barely a decade from oocyte experiment to prize.
- AQP2 and human disease. Positional cloning showed that congenital nephrogenic diabetes insipidus is caused by loss-of-function mutations in either the V2 vasopressin receptor (X-linked, ~90% of cases) or AQP2 itself (autosomal). Affected patients cannot insert working water channels into the collecting duct and pass many liters of dilute urine per day — direct clinical confirmation that AQP2 trafficking is the switch for water reabsorption.
- AQP4 and neuromyelitis optica. In 2004–2005 the NMO-IgG autoantibody, which distinguishes neuromyelitis optica from multiple sclerosis, was identified as targeting AQP4 on astrocyte end-feet. This turned a channel discovery into a diagnostic test and a disease mechanism, and made AQP4 a therapeutic target.
Frequently asked questions
How do aquaporins let water through but block ions and protons?
Two filters act in series. First, size and charge: an aromatic/arginine (ar/R) constriction near the extracellular mouth narrows the pore to roughly 2.8 angstroms — barely wider than a water molecule (about 2.8 angstroms across) — and a fixed positive charge from a conserved arginine electrostatically repels cations while the constriction is too tight for a hydrated ion to pass. Second, and more subtly, the channel blocks protons even though bare protons are tiny. Protons normally travel through water by the Grotthuss mechanism, hopping along a continuous hydrogen-bonded chain of water molecules. At the center of the aquaporin pore two half-helices point their positive N-terminal dipoles at the passing water, and the tandem asparagine–proline–alanine (NPA) motifs donate hydrogen bonds to the water oxygen. This forces each water molecule to reorient — flipping its hydrogens away from the pore axis — so the single-file water chain is broken and cannot relay a proton. Water still slips through at up to 3 billion molecules per second; the proton simply has no wire to hop along.
How fast do aquaporins move water?
A single AQP1 channel conducts on the order of 3 billion water molecules per second (roughly 10^9 to 3×10^9 s^-1) when a physiological osmotic gradient is applied — fast enough that the transport is essentially diffusion-limited at the pore mouth. Because water crosses in single file through a narrow one-molecule-wide channel, the flux is enormous per unit area yet still passive: there is no ATP consumption and no conformational gating cycle per molecule, unlike a pump or a symporter. Classic evidence came from expressing AQP1 in Xenopus oocytes: injected oocytes placed in hypotonic saline swelled and burst within minutes, while control oocytes barely changed volume, and mercury (which reacts with a pore-lining cysteine, Cys189) abolished the water permeability. The osmotic water permeability coefficient (Pf) of an AQP1-loaded membrane can exceed that of a bare lipid bilayer by 10- to 100-fold.
What is the role of AQP2 and vasopressin in the kidney?
Aquaporin-2 (AQP2) is the vasopressin-regulated water channel of the renal collecting duct, and it is the switch that decides how concentrated your urine becomes. When the body is dehydrated, the posterior pituitary releases antidiuretic hormone (ADH, arginine vasopressin). ADH binds the V2 receptor on the basolateral membrane of principal cells, a Gs-coupled GPCR that raises cyclic AMP, activates protein kinase A, and triggers phosphorylation of AQP2 (notably at Ser256). Phosphorylated AQP2, stored in subapical vesicles, traffics to and fuses with the apical membrane, inserting water channels that let water flow out of the tubular fluid down the osmotic gradient built by the medullary countercurrent system — concentrating the urine. AQP3 and AQP4 on the basolateral side provide the exit route back to blood. When ADH falls, AQP2 is internalized by endocytosis and urine stays dilute. Loss-of-function mutations in AQP2 or the V2 receptor cause nephrogenic diabetes insipidus, in which patients pass many liters of dilute urine per day and cannot concentrate it even when severely dehydrated.
Who discovered aquaporins and when?
Peter Agre and colleagues at Johns Hopkins identified the first water channel in 1991–1992. They had actually stumbled on the protein — a 28-kilodalton band, then called CHIP28 (channel-forming integral protein) — while purifying the Rh blood-group antigen from red-cell membranes, and noticed it was also abundant in kidney tubules. In a decisive 1992 experiment (Preston, Carroll, Guggino, and Agre, Science), they expressed CHIP28 in Xenopus oocytes; the cells swelled and lysed in hypotonic medium, proving the protein was a water pore. The channel was renamed aquaporin-1. Physiologists had argued for decades that membranes must contain dedicated water pores to explain the high water permeability of red cells and kidney, but the molecule had eluded everyone. Agre received the 2003 Nobel Prize in Chemistry, sharing it with Roderick MacKinnon, who solved the structural basis of ion channels.
Where are aquaporins found in the human body?
Humans express thirteen aquaporin isoforms (AQP0 through AQP12), each with a tissue-specific distribution. AQP1 lines red blood cells, the kidney proximal tubule and descending thin limb, choroid plexus, and vascular endothelium. AQP0 (MIP) is the most abundant protein in the eye lens fiber cells, where it both conducts water and glues adjacent cells together — mutations cause hereditary cataracts. AQP2, AQP3, and AQP4 handle collecting-duct water reabsorption. AQP4 is the dominant channel in the brain, concentrated on astrocyte end-feet at the blood–brain barrier, where it governs cerebral water flux and edema and is the autoantigen in neuromyelitis optica. AQP5 drives fluid secretion in salivary, lacrimal, and airway glands — its loss produces dry mouth and dry eyes. A subgroup called aquaglyceroporins (AQP3, AQP7, AQP9, AQP10) also conducts glycerol and small solutes, linking water channels to fat metabolism and skin hydration.
Do aquaporins pump water or is transport passive?
Aquaporins are passive channels, not pumps. They conduct water strictly down its own concentration (osmotic) gradient and can move it in either direction depending on which side is more concentrated; they never build a gradient and never hydrolyze ATP. Water still needs a driving force, which is supplied by solute transporters and pumps elsewhere. In the kidney, for example, the sodium–potassium pump and secondary active solute transport in the proximal tubule and the countercurrent multiplier in the loop of Henle create the osmotic gradient; aquaporins then simply provide a low-resistance path for water to follow that gradient rapidly. This is why regulation happens at the level of channel abundance and localization — inserting or removing AQP2 from the membrane — rather than by gating each molecule. Some aquaporins are gated by pH or phosphorylation (plant PIP channels close during drought, for instance), but the transport itself is always downhill and energy-free.