December 4-7, 2001
Friiberghs Manor, Örsundsbro and Stockholm
University
by Arthur Horwich
Yale School of Medicine, Howard Hughes
Medical Institute
The transfer of information in living systems proceeds from DNA to RNA to protein. The final step of such transfer involves the folding of newly made proteins into their characteristic three-dimensional native conformations. For many years, this was thought to be a spontaneous process, based on the classic studies of Christian Anfinsen and coworkers beginning in the 1950s, who established that the primary amino acid sequence of a protein contains all of the information necessary to direct the polypeptide chain to its native conformation, a state that is typically at the energetic minimum. Yet these studies already recognized that only particular conditions would support such folding, implying that kinetic difficulties could be encountered in reaching the thermodynamic minimum under some conditions, e.g. those of the cell.
In the mid-1980s, a class of specialized proteins known as heat shock proteins, whose abundance is thermally induced, was recognized to bind to other proteins, particularly under conditions of thermal stress. It was suggested by Hugh Pelham that, through the consumption of ATP, these proteins might be acting at the quaternary structure level to dissociate oligomeric aggregates formed by hydrophobic contacts between partially denatured proteins. But experiments observing their association with secretory and mitochondrial proteins, which pass through membranes in unfolded states, soon shifted attention to a role for heat shock proteins in adjusting conformation of single polypeptide chains.
We identified a mutation in a yeast heat shock protein, Hsp60, an abundant homo-oligomeric double-ring complex inside the mitochondrial matrix, that blocked the folding to native form of newly-imported mitochondrial proteins. This protein was found to be essential for cell viability under all conditions, and in the absence of its function, imported proteins aggregated, having apparently misfolded. In collaboration with Ulrich Hartl, physical association of newly imported proteins with the Hsp60 ring assembly was demonstrated, and the bound proteins appeared to occupy non-native forms. The presence of ATP caused dissociation of the proteins, and they were now found to be in their native form. An essential role of Hsp60 in providing ATP-mediated kinetic assistance to protein folding to the native state was proposed. Additional such double-ring assemblies, termed chaperonins by John Ellis to denote this family of molecular chaperones, were identified in other cellular compartments, e.g. GroEL in the bacterial cytoplasm, Rubisco binding protein in chloroplasts, and the CCT chaperonin in the eukaryotic cytosol. These were observed to carry out similar essential actions in assisting the folding of many newly-translated proteins. Such action was reconstituted in vitro by George Lorimer and colleagues, with the productive folding of an otherwise aggregating protein mediated by GroEL, its cooperating single ring cochaperonin GroES, and ATP.
During the past decade, X-ray crystallographic work, carried out in collaboration with the late Paul Sigler, and functional studies have begun to elucidate the mechanism of the GroEL/GroES system. To summarize our current understanding, the central cavity of GroEL provides an environment that assists proper folding through actions associated with specific conformational states. One action is the binding of collapsed, partially structured non-native proteins in the open ring of a GroEL complex via hydrophobic side chains lining the open cavity, which forestalls aggregation and may be associated with partial unfolding. We have observed this interaction to be multivalent, involving contact between the substrate protein and multiple surrounding apical domains. The other major chaperonin action is facilitation of folding, occurring inside a sequestered, enlarged, and now hydrophilic central cavity, produced upon sequential binding of ATP and GroES to the same ring as polypeptide. Such binding produces a 60° elevation and 90° twist to the apical domains of the GroES-bound GroEL ring, removing the hydrophobic polypeptide-binding surface from the central cavity. On the same time scale (t~1 sec), substrate polypeptide is released into the central cavity. In a second, longer, phase (t~12 sec), corresponding to the lifetime of the GroEL-GroES-ATP complex, folding proceeds in the central cavity until ATP hydrolysis, followed by ATP binding to the opposite ring, dissociates the GroES and releases the polypeptide into solution. Because only a fraction of the polypeptide molecules achieve the native state during the lifetime of the folding-active complex, most are released in a non-native state and rebind to the chaperonin for another trial at productive folding. This mechanism, known as cis folding because GroES and polypeptide are bound to the same GroEL ring, is used by substrate proteins small enough to be encapsulated in the central cavity underneath GroES. In contrast, proteins too large to be encapsulated can in some cases be assisted by GroEL/GroES via a trans mechanism, as indicated by our recent study of aconitase, an 82 kDa protein, where binding of GroES to the opposite (trans) ring of GroEL is required for release and productive folding.
There are many exciting questions that remain to be resolved concerning the chaperonin-mediated folding reaction. They particularly concern the fate of polypeptide. How is a misfolding polypeptide captured and "unfolded" by an open GroEL ring? Is it actively unfolded on the binding sites? Or does the chaperonin select the least folded states from an ensemble of non-native states, a so-called thermodynamic partitioning mechanism? To answer this, polypeptide conformation may be able to be monitored during the binding step by carrying out time-resolved fluorescence studies, or by pulsed hydrogen-deuterium exchange coupled with mass spectrometry. What does a polypeptide look like once it is stably bound in the central cavity of a GroEL ring? To answer this, X-ray crystallographic approaches are being taken, at this point with various degrees of order introduced into the system, e.g. attaching a small polypeptide to each GroEL subunit, allowing full occupancy of the apical binding sites. In addition, new NMR techniques, developed by Kurt Wüthrich and his group, are being applied in collaboration with them to observe both the machine and a labeled substrate polypeptide bound in the central cavity. Finally, the question of how the cis chamber favors productive folding remains a mystery. Is it an infinite dilution folding chamber, or is it a box with confining walls, that serves to limit the conformers that can occur on the folding "landscape?" Here, it may be possible to address this with single molecule fluorescence studies, comparing the rate of folding inside with that at high dilution in solvent.
Printer Friendly 
Comments & Questions
Tell a Friend 
