Structure of Semi-Permeable Membranes

While corresponding with a friend today about the phenomenon of cell suicide (apoptosis) I remembered that I had recently written about a personal revelation in cell membranes. And that since then I had completely forgotten about it. After rereading it and doing some touch editing and clarification I decided it was definitely blog-worthy, I don't know why I didn't post it before. So this was originally written on November 7th 2005:

I just got out of my physics research seminar, which is basically a one hour pass course where we listen to various professors explain about their research and tour their labs. Today Professor Ritchie, a new professor, lectured about his work in biophysics. Since he is so new, he doesn't have a web domain set up with Purdue yet, I asked him about it and he said he's "working on it", it'll probably be a while. Anyway, if you're interested in his research here is his contact page at Purdue.

So what he said in class was pretty amazing. He was talking about his research in cell membranes, specifically the lipid bilayer. A single lipid/phosphate head molecule (I'll just call it a lipid) moves around the general surface. They move like plastic ducks on water; staying within the general 2-D plane of the membrane, but the individual lipids and proteins float around each other randomly.

The problem with that is that if this were entirely accurate, a single lipid should move randomly at a certain speed. After tagging a single lipid with a photoluminescer and watching it's progress, they find it moves at about a fifth of the speed expected. Also, in such a model, the cell wouldn't be able to efficiently localize information. He gave the example of nerve cells that have definite outgoing and ingoing regions, if the lipids and proteins move freely/randomly about one another, then the cell could not have the specified regions of receptor proteins (they would be mixed randomly).

So, what he did was to use an incredibly high frame/s camera. (digression: this thing is wicked awesome! He showed us a demo of the camera he used in which the camera recorded a balloon filled with water being punctured by an exacto. As the exacto bit through the balloon, the balloon quickly snapped away and crumpled as would be expected, but the water almost perfectly held it's form, due to the high elasticity of water. Even the air bubble kept it's shape, and the camera was able to record this! Link) So, using this new camera he set the same experiment with photoluminescer tagged lipid and watched what happened (I presume he was careful not to change variables from the original experiment). The lipid moved randomly in a certain perimeter, then shifted a bit to move randomly in another perimeter, then shifted again to a neighboring perimeter. What they found was that the speed of the lipid within a perimeter was that which they had expected for the duck/pond model of a membrane, in which the lipid could float freely anywhere, and the speed that it took the lipid to move from one sector to the next was what they had recorded to cause the whole problem in the first place (1/5 expected speed). Apparently there is some sort of pattern of boundaries on the membrane which causes freely moving bodies to get stuck, but eventually make through, it holds them up.

They used a nifty little method of anchoring one of the more tightly bound proteins to a gold colloid and literally dragged the protein around the membrane in order to sound out the boundaries. If they hadn't of used a protein that binds tightly with gold, then the gold would have had a higher chance of breaking off when meeting resistance, which was the whole point, to measure resistance.

After mapping the boundaries they found that they align with the actin microfilaments of the cytoskeleton!

All through my biological studies in high school I had been confused as to what the cytoskeleton actually does, what function it performs.

This is what professor Ritchie says, the cytoskeletal filaments are right below the membrane and proteins are anchored to the structure. These proteins stick up inside (sometimes through) the membrane and cause what he calls "fences". A lipid is free to move within any of these little areas, but it takes some time for it to get through/around the fence. I didn't ask if all proteins are anchored to the actin, but knowing generally how a cell works, I'd say that the proteins which are advantageous to be localized are anchored and the ones that need to move aren't. Which explains how a nerve cell can partition one branch for ingoing and another for outgoing proteins.