Hsp90 is a molecular chaperone that plays a role in the folding and assembly of other proteins. Current ideas suggest that it binds to substrate proteins at a "client" site and this either encourages folding into the proper conformation or prevents aggregation. The binding and release of polypeptides is accompanied by hydrolysis of ATP to ADP + Pi.
The bacterial version of hsp90 is called HtpG. Eukaryotes have several different members of the Hsp90 family including one that resides in the endoplasmic reticulum. The cytosolic protein is called Hsp90 and the ER version is called GRP94. Hsp90 is a highly conserved protein showing significant sequence identity between prokaryotic and eukaryotic proteins. The HSP90 family shares many of the same characteristics of the more highly conserved HSP70s [Heat Shock and Molecular Chaperones, Gene HSPA5 Encodes BiP-a Molecular Chaperone, The Evolution of the HSP70 Gene Family].
Daniel Gewirth and his colleagues have just published the complete structure of GRP94 from dog (Canis familiaris). The article appears in Molecular Cell and their structure is on the cover of the journal (Dollins et al. 2007). This is the endoplasmic reticulum version of Hsp90 and its the only ER version of this protein whose structure is known. Gewirth has been working on the structure since 2001 and he deposited the first structural coordinates of a fragment of this protein back in February 2004. (See the Protein Data Base (PDB) for the structures. Search for "hsp90".)
The complete protein is a dimer of two identical subunits. Each monomer has three distinct domains; an N-terminal domain (N); a middle domain (M); and a C-terminal domain (C). The ATP hydrolysis site sits at the interface between the N and M domains. The C domains interact to form the dimer. The presumed site of binding for misfolded proteins ("client" site") is in the V-shaped pocket formed when the C domains come together.
The mechanism of action of Hsp90 proteins is not known although it presumably involves a conformational change induced by ATP hydrolysis. This paper provides an important clue to that mechanism because the dimer structure differs from that seen with the yeast protein (Hsp82) and the E. coli protein (HtpG) (below).
Each of the structures seems to identify a protein in one of the conformations adopted in vivo. The most likely explanation is that the wings of the protein open and close to capture and release the substrate protein. This conformational change is induced by binding and hydrolysis of ATP.
Now that we have a structure for GRP94 from dog we can compare the structures of proteins from different species to see how closely they resemble each other. Let's look at the N-terminal domain to get an idea of how protein structure is conserved over billions of years. The four structures below are, from left to right, yeast (1zwh), dog (2fyp), human (1us7) and E. coli (2ior).
Aren't they remarkably similar! This is exactly the sort of thing you expect with a highly conserved protein.
By the way, anyone can create these images by going to the PDB site [2ior] and viewing the structures with the MBT SimpleViewer. If you haven't already installed this viewer it will automatically install in your browser and it only takes a few minutes.
Dollins, E.D., Warren, J.J., Immormino, R.M. and Gewirth, D.T. (2007) Structures of GRP94-Nucleotide Complexes Reveal Mechanistic Differences between the hsp90 Chaperones. Molec. Cell 28:41-56.