The (Cyto)Skeleton in the Closet
It’s a bit weird to think of prokaryotes as having cytoskeletons, since the most common dogma for undergraduate teaching tends to be that in bacteria, things just diffuse. Bacteria are small enough, the professor says, that they don’t need active transport systems that rely on the long polymers of the cytoskeleton. After all, an E. coli bacterium is only 1 micrometer long, and with the diffusion constant of an aqueous protein being around 0.1 μm2 / ms, the time for proteins to diffuse from one side of the E. coli to the other is only a couple milliseconds. Really, (so the theory goes) is it surprising that bacteria don’t have a cytoskeleton? (The theory has been extended to analysis of purely physical methods of bacterial segregation, due to entropic considerations. Suckjoon Jun and Bela Mulder have a nice article in PNAS on this. An older paper describes a model of the bacterial nucleoid as a phase of supercoiled DNA.)
But in the last ten years or so, bacteria have been found to have many proteins that are very similar to eukaryotic cytoskeletal proteins, as well as a few that don’t have eukaryotic homologs at all. Mindblowing. I’ve been reading this review paper by Zemer Gitai to catch up. So, really, half the stuff they teach about bacteria to high schoolers and college students is wrong.
Though in retrospect, it’s not all that surprising that bacteria have cytoskeletons. How else would bacteria divide? Something has to squeeze the wall closed in between the two daughter cells, and it’s probably going to be some sort of cytoskeletal protein. It turns out that a tubulin homolog, FtsZ, forms a ring at the division site. In many mammalian and plant cells, we know that actin forms a ring and squeezes shut to cut cell apart (imagine a garroting wire forming a loop, and you have the right, though gruesome, idea). We don’t know exactly what FtsZ does, but it may have a similar role in bacteria. And bacteria also have lots of interesting shapes, which have to be maintained somehow.
One weird thing is how FtsZ is controlled by bacteria. Different bacterial species seem to use really different ways to control where FtsZ assembles. C. crescentus seems to use the most obvious solution, which is to have a protein that stops FtsZ assembly — MipZ — gather at the ends of the bacterium, so that the split away from the ends, right in the middle. B. subtilis uses a similar mechanism, with a different protein, MinC.
E. Coli, though, uses a really weird system, which is to have MinC form at one end, recruit another protein (MinE) which disassembles it, and then form again at the other end. So, essentially you have MinC turning on and off on either end of the bacterium, which forces FtsZ to form right in the middle. But why go through all that trouble? The oscillator also doesn’t seem particularly stable, the way it’s conceived right now, so there’s definitely something more to this oscillatory behavior than what’s known right now.
Anyway, I’m really curious as to the future discoveries on the bacterial cytoskeleton. I’m not only interested in the biochemistry, but also on the evolutionary side of things. Did eukaryotic cytoskeletal proteins evolve from this prokaryotic stuff? Harold Erickson seems to think so, even with the large sequence differences between the prokaryotic and eukaryotic versions. This could lead to interesting insights on eukaryotic cell shape, by simply studying bacterial shape.
And bacterial cytoskeletal proteins might just be different enough to make interesting broad spectrum antibiotics in the future.
