Blob-ology

The advent of electron microscopy - in particular single particle cryoEM - has revolutionised our understanding of these fascinating conformational changes in the molecular chaperones and other proteins. With EM it has been possible to trace the pathway of these changes, initiated by the binding and hydrolysis of ATP, which then drives the allosteric cycle observed in GroEL.

Before the development of EM, scientists worked in the field of 'blobology' - "everything just looked like blobs", Saibil says. "But even at the blob level, GroEL looked quite interesting.

"Ten years ago we made a series of little movies that we still show of the conformational changes in the GroEL system at 25 angstrom resolution. You could just see the shapes of proteins in their separated domains or subunits - you could see the symmetry. But nowadays we are getting resolutions of seven or eight angstrom and you can see the alpha helical densities. We are able to see the internal features of the protein in quite a lot of detail.

"People are developing this for other things like viruses. With GroEL now, people are using it as a test object and getting very good resolution on it by electron microscopy. So it's now possible to look at protein complexes in quite a lot of detail by these methods. And that's all happened at the same time as we have been working on the GroEL story."

In part due to her interest in the conformational changes observed in the molecular chaperones, Saibil has also ventured into the world of bacterial pore-forming toxins, a group that includes the anthrax toxin. Saibil has been studying pneumolysin, a pore-forming toxin secreted by Streptococcus pneumoniae that is part of the cholesterol binding family of bacterial toxins. They are characterised by the formation of large rings on the surface of cholesterol-containing membranes through which the toxin enters the cell or allows its contents to leak out.

"It is the same idea as the big conformational changes in protein complexes," she says. "Pore-forming toxins have to go through a huge change because they start as small, individual, soluble proteins - monomers - and then they assemble into huge rings that jump into membranes. Proteins are not supposed to do that - they are supposed to be soluble or membrane proteins. Normally they stay one or the other. But pore-forming toxins are fascinating for the reason that they go through huge changes and it is mysterious as to how a protein can do both of those things."

Protein misfolding

As well as looking at how molecular chaperones assist protein folding, there is the obverse - what happens when it all goes horribly wrong and protein aggregates form into amyloid fibrils. While it is not known exactly how they do it, amyloid fibrils are thought to be the toxic agents responsible for nerve cell death and loss of brain function in a number of diseases, including Alzheimer's and CJD.

Saibil often wishes she hadn't gone into this area as misfolded or denatured proteins are extremely difficult to work with. "Misfolded proteins look pretty disgusting and you can't recognise much," she says. "But there is a final destination point for a lot of these aggregates, which is amyloid fibrils."

From a structural perspective, amyloids are interesting because the proteins, whatever their starting structure was, are converted into a beta sheet structure. When proteins fold they make two regular, secondary structures. The polypeptide chain either curls into a spring shape (an alpha helix) or what Saibil describes as a Venetian blind, in which each slat is hooked onto the next, known as the beta sheet structure.

The spring-shaped type has hydrogen bonds along its length, folding it into a coil, and the beta sheet has them folding into a flat shape. "For some reason that secondary structure ends up in these fibrils and it doesn't matter what the starting structure of the protein was - all of them seem to end up in what is called a cross-beta fold," she says. "The beta sheets are perpendicular to the long axis of the fibre. So it is like very long ribbons of Venetian blinds.

"We have been looking at the structure of these fibrils, which is quite difficult. We haven't yet got very far but we hope it will eventually help us understand this mysterious conversion of the protein that has become denatured or messed up in some way. If we can understand the conformation of the fibril better we might be able to work backwards to see what the conversion is."

Saibil last visited Australia a decade ago, when she visited for the 23rd Lorne Protein conference. For the 33rd conference, she is an invited speaker. If she has time, she will discuss her research into the heat shock protein 104 (hsp104), a chaperone from the large hsp100 family found in bacteria, plants and yeast that seems to reverse aggregation.

The hsp100 subset is part of the large AAA protein family, ATPases that are involved in a number of different unwinding and unfolding helicases. "We've got a not particularly high resolution structure of hsp104, which I'll show if I have time," she says.

What she will talk about is the fascinating molecular chaperone story as well as more recent research, particularly the results of studies of what a non-native protein actually looks like when it is stuck inside the chaperonin. And she has pictures.

"We have some new results on a very large substrate on a bacteriophage co-protein," she says. "It's at the upper end of the substrate side that will fit into the chaperone box. We have a map now of the structure of the chaperonin container bulging at the seams, with a large protein stuck inside it."