Across the Cell Membrane

Two Layers, Under Five Nanometres

The body of a cell membrane is two layers of lipid molecules. Each has a head that likes water and two oily tails; in water the tails hide inwards and the heads face out, and the molecules settle into two layers by themselves.

A pure phospholipid bilayer is about 4.5 nm thick with its heads, and its oily core about 2.7 nm. Membranes from rat liver cells measure 3.6⁠–⁠4.3 nm from the phosphates on one side to those on the other.

A square micrometre of membrane holds a few million lipids. They jostle constantly within their own layer, but crossing to the other is slow: without an enzyme, half of them take 6.5⁠–⁠15 hours to flip.

On the plinth is a patch 16 nm wide, enlarged about 32 million times (computed).

Pure phospholipid: core about 2.7 nm, about 4.5 nm with heads · cell membranes: phosphate to phosphate 3.6⁠–⁠4.3 nm

What Gets Through on Its Own

The barrier is the oily core. Small uncharged molecules such as oxygen and carbon dioxide barely notice it; water is polar but seeps through slowly; urea and glycerol are a few hundred times slower again.

Ions hardly get through at all: across an artificial bilayer, sodium and potassium have permeabilities of about 5 × 10⁻¹⁴ cm/s, some ten orders of magnitude below water. Carrying a charge into oil costs too much. Glucose is large and polar, and does not get through either.

For ions and sugars to move in and out, a cell needs proteins in its membrane.

Artificial bilayers, cm/s: O₂ 23⁠–⁠125 · CO₂ 0.35 · water about 0.003 · urea about 0.000004 · Na⁺ about 5 × 10⁻¹⁴

Water in Single File

The water channel aquaporin-⁠1 is abundant in red blood cells and kidney tubules. It is four identical subunits, each with a narrow pore of its own; the model is cut open along one of them.

At its narrowest the pore is about 2.8 Å, barely wider than a water molecule; along some 20 Å in the middle, waters can only pass one behind another. Two NPA motifs sit at the middle, where the field turns each water round; the electrostatic barrier there is what mainly keeps protons out, so the pore passes water but not H⁺.

From its measured unit permeability, each subunit can pass up to about 3 billion water molecules a second. How many actually flow depends on the osmotic difference: with the two sides equal, the water shuffles back and forth with no net flow.

AQP1 (PDB 7UZE) · narrowest about 2.8 Å · waters in single file (schematic)

An Egg That Burst

In 1988 researchers purified a 28 kDa protein from red cell membranes: 120,⁠000⁠–⁠160,⁠000 copies on each cell, and no one knew what it did.

In 1992 they injected its RNA into eggs of the African clawed frog, waited 3 days for the protein to appear in the egg membrane, and moved the eggs into a hypotonic solution. Within 30 seconds the eggs began to swell; after 3 minutes they burst. Eggs injected with water alone barely changed. A mercury salt blocked the effect, reversibly.

It was the first water channel to be recognised, later named AQP1, and Peter Agre shared the 2003 Nobel Prize in Chemistry for it. From above, the four subunits make a square, each with its own pore; the central axis is not a water path.

120,⁠000⁠–⁠160,⁠000 per red cell · eggs swell in 30 s and burst in 3 min · four subunits, four pores

A Pocket That Turns

Glucose enters red cells through GLUT1. GLUT1 is not an open pore: a sugar enters a pocket in its middle from one side, the protein changes shape, the pocket opens to the other side, and the sugar leaves.

It moves sugar only down the concentration difference and uses no energy: facilitated diffusion. One GLUT1 carries about 1,⁠200 glucose molecules a second; an open potassium channel passes hundreds of millions of ions a second.

The model’s GLUT1 comes from an inward-⁠open crystal (PDB 6THA); the sugar ring in its pocket is the glucoside of a detergent used to grow the crystal, and the outward-⁠open shape is known from its close relative GLUT3. Moving things against the difference takes energy, which is a pump’s job: the sodium⁠–⁠potassium pump moves 3 sodium ions out and 2 potassium ions in for each ATP (see “Nerve Impulse”).

GLUT1 (PDB 6THA, inward-⁠open) · about 1,⁠200 glucose a second · the sugar’s path is a schematic

Which Way the Water Goes

When one side of a membrane holds more dissolved particles that cannot cross it, water flows, on balance, towards that side: osmosis. What counts is the number of particles, not what they are: sodium chloride splits into a sodium ion and a chloride ion, so each counts twice.

Holding the flow back takes extra pressure on the side with more particles: the osmotic pressure. In a dilute solution π = cRT. With the measured osmotic coefficients of sodium chloride, 0.9 % salt at 25 °C gives about 0.71 MPa and 3 % about 2.3 MPa, some 23 atmospheres (computed).

In 1877 Pfeffer first measured osmotic pressure, with a membrane that let water through but not sugar. Van ’t Hoff found that a dilute solution’s osmotic pressure equals the pressure of the same number of molecules as a gas, and in 1901 won the first Nobel Prize in Chemistry.

More solute above (schematic) · 0.9 % NaCl about 0.71 MPa · 3 % about 2.3 MPa (25 °C, computed)

Why Drips Are 0.9 %

Blood plasma runs at about 275⁠–⁠295 mOsm/kg. Saline of 0.9 % holds 9 g of sodium chloride per litre: 308 mOsm/L if each splits into two particles, about 286 mOsm/kg as measured, close to plasma. That is why it is called isotonic and used in drips. It holds only sodium and chloride, so it is not the same as plasma.

Red blood cells have no wall. In weaker salt, water flows in and they swell into spheres: at about 0.45⁠–⁠0.53 % the most fragile begin to leak haemoglobin, and by about 0.27⁠–⁠0.33 % nearly all have (the figures shift with temperature). The haemoglobin mostly leaks out through a stretched membrane; the membrane need not break apart.

In strong salt, red cells lose water and shrink into spiky little balls.

0.9 % NaCl: 308 mOsm/L nominal, about 286 mOsm/kg measured · plasma 275⁠–⁠295 mOsm/kg

A Cell with a Wall

A plant cell has a wall outside its membrane: a cross-⁠linked mesh of pectin, cellulose and other polymers, stiff and full of pores that let water and salt through. What chooses what goes in and out is the membrane inside it.

The outer skin of a red onion scale is a single layer of long cells; onion epidermal cells that have been measured are about 0.4 mm long and 50⁠–⁠120 µm wide. The purple is anthocyanin, dissolved in the sap of the vacuole. In a mature plant cell the vacuole often fills over nine tenths of the cell, and the cytoplasm is a thin layer pressed to the membrane.

In fresh water the cell takes up water and presses its membrane against the wall: turgor, about 0.3⁠–⁠0.7 MPa in onion skin, depending on how it is measured. The wall holds against it, which is why a plant cell in fresh water does not burst. The cells on the plinth are enlarged about 160 times (computed).

Red onion epidermis · turgor about 0.3⁠–⁠0.7 MPa · vacuole often over nine tenths of the cell

The Purple Pulls Back

Drop strong salt water at one edge of the cover slip and draw the water off at the other with blotting paper, and the salt water flows past the cells. With more particles outside than in the sap, water leaves through the membrane: the vacuole shrinks, the purple looks deeper, and the protoplast pulls away from the wall. This is plasmolysis.

The wall hardly moves, and what fills the space between wall and protoplast is the salt water from outside, not air. Where the membrane leaves the wall, fine threads often still tie it on: Hechtian strands.

In strong salt the protoplast can be seen leaving the wall within tens of seconds, and settles in minutes to half an hour, faster in stronger solutions. In 1877 de Vries named the effect plasmolysis, and later used it to find the strength of solution that matches the sap.

3 % NaCl: the protoplast shrinks to about 41 % (this page’s model) · salt water between wall and protoplast

Back in Fresh Water

Back in fresh water, with fewer particles outside, water flows back into the cell and the protoplast swells against the wall again. Tobacco cells finish plasmolysing in 3⁠–⁠5 minutes and recover their shape 10⁠–⁠20 minutes after going back into their medium.

Only a living cell can go back and forth like this. Living cells survive more than a day plasmolysed; a cell whose membrane has failed cannot hold its sap, so it no longer shrinks and swells. Dried too fast and then flooded, the membrane may not have enough area and the protoplast bursts, which is why researchers bring cells back through a weaker solution first.

Salting cabbage works the same way: water flows out of the leaves into the brine, salt slowly moves in, and the leaves lose water, soften and shrink.

Tobacco cells: plasmolysed in 3⁠–⁠5 min, recovered in 10⁠–⁠20 min · only living cells recover

A Patch of Membrane, Taken Apart

The model comes apart in five: the outer leaflet, the inner leaflet, aquaporin-⁠1, the glucose carrier GLUT1, and the file of water in the pore. The outer leaflet lifts off, and the two proteins and the water file come forward.

Both proteins’ surfaces are computed from Protein Data Bank atoms and set in the membrane at the depth the OPM database gives. Rest the pointer on a part and it is lifted out on its own and named.

Outer leaflet · AQP1 · GLUT1 · water file · inner leaflet