Slippery lipids

So, as you might have seen reading this website, the main molecule that we work on is a membrane lipid: PI(4,5)P2 or PIP2. This molecule does a lot in cells; proteins of all shapes and sizes stick to it, either pinning them to the membrane, or switching them into an active conformation. These proteins control everything that happens at the cell surface, from controlling cell shape and attachment to its neighbors, through transmitting information to the inside of the cell, to controlling the flux of materials into, out of and across the membrane.

These proteins all organize into separate regions of the membrane. A theory cell biologists have had is that PIP2 might be separated in the membrane into these separate regions, reserved to activate specific sub sets of proteins. It follows that we would need to understand what controls this separation in order to understand how specific membrane-associated functions fail in disease. But to start with, we’d need to see how easily the lipids can be separated like this.

To that end, we used fluorescent lipid biosensors – for our favorite lipid, PIP2 – and for several others. There is a catch to these experiments: the biosensors bind to the lipid’s head that sticks out of the membrane surface, which is the same place that the proteins do. So our experiments only follow the “unbound” lipids – we can’t see the lipids that already bind proteins. But don’t worry – biosensors and proteins are continuously popping on and off lipids every second, so the mix is jumbled up every couple of seconds or so.

To detect the motion of these lipid biosensors, we use a very expensive camera on a microscope to look at the membrane underneath the cell. With a few optical tricks, we are able to see individual fluorescent molecules. This is how they look in a living cell (and the movie is real time):

Wow, they jiggle about pretty fast, huh? In fact, we use an algorithm to track the molecules, and it turns out to be a little weirder that that. The molecules spend roughly two-thirds of their time moving quickly, and one third moving much more slowly. Sometimes you can resolve “slow” and “fast” molecules, but mostly, the molecules swap between slow and fast, as you can see in this example:

Spoiler: we can still only really guess at why the molecules do this. But we wondered whether specific, PIP2-binding protein domains in the membrane could trap the PIP2. Maybe we see slower motion, or more “slow” particles at these domains? Here are some examples of the domains:

To our great surprise, we actually saw no difference in PIP2 mobility for most of these domains – the biosensor-bound lipids slip into, out of and through these domains with absolutely no impediment.

However, there was one really convincing exception: the septin cytoskeleton. We saw slower mobility (diffusion or “D” in technical terms), and more “slow” particles around regions of the membrane with lots of septin filaments:

Why does septin do this? Well, septin filaments lie across the membrane in contact with the lipids, so they could be a literal barrier to diffusion.

We could test this using a really freaky but cool trick: if you disassemble the cellular F-actin skeleton with a poison, the septin filaments rearrange into micrometer-sized rings – kind of tiny donuts. We could see the lipid biosensors bouncing around the outside of these, but almost never getting inside:

So, septin filaments are a barrier to PIP2 diffusion (so too are spectrin filaments). But most PIP2 binding proteins organize into domains that do not and cannot really corral the lipids when not directly sticking to them.

So, our conclusions are that PIP2 (and other lipids) move very rapidly in the membrane, and most proteins are unable to constrain the lipids beyond the individual lipids they may be holding onto at that time. This is an important finding, because it tells us that the lipid cannot really be used to organize the proteins – it has to happen the other way around.

Therefore, the mystery of how PIP2 is able to control so many functions of the cell membrane in parallel, and regulate them separately, remains unsolved. Maybe you can help us figure out why?

In the mean time, you can read the full article for free at journal of cell biology.