Superhero aquaporin

How does water get into and out of our cells?

Water is an amazing molecule. Common, but complicated. Clear, essential, and taken for granted, but also chemically very weird. Water is more prominent as one goes out in the solar system. The Oort cloud is a vast, cold swirl of ice and dust, from which Earth is thought to have obtained most of its water by bombardment, after the original light gasses were driven off in the earliest hot phase of planet formation.

We are 60% water. Biology originated in water, and all biochemical reactions take place in water as the solvent. One of the first steps to the origin of life may have been controlling water by forming oily membranes to separate inside from outside, creating primordial cells.

Biological membranes are very slightly permeable to water, and less so as one adds cholesterol. By the time one gets to all existing cells in the evolutionary story, there is a need for special proteins that conduct water through the cell membrane at a high rate. These are all passive pores, (not pumps, which would require extra energy), though they can be regulated in various ways to turn on and off. Our kidneys are full of them, with at least six different genes encoding water pores, called aquaporins.

Cells typically regulate their water balance not by pumping water itself around, but by manipulating ion concentrations (of sodium, potassium, and chloride, principally) within. They can build internal water pressure (turgor pressure) by pumping an excess of an ion inward, and waiting for diffusion to attempt to equalize the concentrations inside and outside by sending water after the ions into the cell.

But here is where the aquaporins come in. In bare membranes, this diffusion process is way too slow for most cases of regulation, so cells need a passive, specific, and high-volume way to let water in and out, faster than the membranes themselves allow. And the toughest part of this is making such a channel specific to water alone. Making a hole in a membrane is not difficult. Many pathogens do just that to kill cells. An aquaporin needs to conduct H2O only, not salts, not hydrogen ions (protons), and not OH- ions, either. It is a job for a very intricately structured protein.

A recent paper described a very high-resolution X-ray structure of an aquaporin from yeast cells, taken as a model for all kinds of aquaporins. Typical protein structures come in at around 2Å resolution, which is enough to follow a protein backbone around and get a good idea of the overall protein structure. This one is 0.88Å, a remarkable achievement for membrane proteins, which tend to be very difficult to crystalize and solve (prior work). 

The quality of this structure is so high that the researchers can tell the difference between an electron cloud shared (conjugated) between the carbon on amino acid Gln137 and its oxygen (red), and an electron cloud more evenly shared elsewhere in the structure, between the carbon on Glu51 and its two oxygens. X-ray crystallography provides only electron densities, from which the identities of the associated atoms and molecules must be deduced.

The overall structure determined by this group, in cut-away cartoon view, with water molecules as red dots. SF is the "selectivity filter", and NPA is another well-studied area, with a "asparagine-proline-alanine" protein sequence.

This resolution allowed the researchers to see individual water molecules snaking their way through the central channel (at least very roughly, there was still some motion that blured them out) in the low-temperature frozen crystals, something very rare in protein structural studies. Because yes, as you would guess, aquaporin channel is a big protein structure that exists entirely to create a narrow channel running through its center.

Youtube provides a computer simulation run by other researchers using another model aquaporin. Brownian motion of the peripheral waters dominates the video, while the waters inside the channel bounce around more slowly. But quickly enough to pass at a rate of billions per second! The channel shows distinct charged areas, blue (+) and red (-), which organize the slightly dipolar water molecules as they slip through. One of the water molecules is highlighted in yellow so you can track it. The simulation shows two particularly narrow regions, upper, near the large blue blob, and also a lower one, where the waters have to squeeze through.

"A single human aquaporin-1 channel facilitates water transport at a rate of roughly 3 billion water molecules per second."

Fine.. but anything could run through here, at least anything smaller than water, which being a three-atom molecule, means that most single-atom ions would be smaller. What makes this channel selective for only water? Here we turn from structure to electrostatics. The channel works by funnelling each water molecule through a channel that is not only narrow, but also lined in very specfic ways to test that each molecule passing is actually water.

Example of a site where one water (#4) is extensively coordinated to tell by shape and charge that it really is water.

"MD [computer simulation / molecular dynamics] snapshots illustrate how this geometry achieves exceptional water selectivity, because all four H-bond donor and acceptor interactions are filled as water moves through the SF [selectivity filter] (C). The presence of four closely spaced water-selective sites optimizes the aquaporin SF’s ability to discriminate water from other small molecules. Hydroxide ions, in particular, suffer a geometric penalty, because they cannot simultaneously donate H bonds to the backbone hydroxyl of Ala221 and to Nε of His212. Conversely, all H-bond interactions are distorted from ideal water geometry, and this avoids binding water too tightly, such that efficient transport is compromised."

A larger view of one of these "selectivity filters" is shown below, where side chains from the protein (in green and blue) reach out to touch the water molecules, shown smaller than their real effective size so that we can see what is going on. Pairs of waters are shown in progressive positions in the channel. The researchers conclude that the pair move in lock-step from one pair of positions to the next pair, all of which show up in the structure, but only two of which can be occupied at a time.

Pas de deux of two waters through the selectivity filter, which coordinates them closely with slight positive charges from Arg227 touching the negative water dipole on the oxygen, while the slight negative charges on Gly220 and Ala221 touch the positive dipoles of the hydrogens.

Which brings up the most sensitive issue, which is that water needs to be sensed very carefully on all sides to bring the right shape through, and kept from bringing along H+ protons, or coming through in OH- form- but not be held so tightly that it can't get through at all. So all these touches are done like in a car wash, on the fly as the water molecules come through one by one.

It is great to see basic scientists digging deeper into the basic mechanisms of life, unearthing knowledge whose application may be very far off, but whose beauty and elegance is a permanent and worthy achievement in itself.

• Advanced guide to protein crystallography.
• Free the whales!
• Goldman, doing god's work corrupting markets and stealing from the rest of us.
• Doesn't "guest" labor sound so nice?
• A theologian responds to the Martin case: time to be an atheist!
• What's wrong with Detroit? Right and left wing prisms.
• The answer for social security ... expand it.
• Is the economic ideology tide turning?
• North Carolina takes up the supreme court's challenge to prevent voting by the "wrong" people.
• Foxes running the climate change hen house.
• Searle on souls.