About TSRI
Research & Faculty
News & Publications
Scientific Calendars
Scripps Florida
PhD Program
Campus Services
Work at TSRI
TSRI in the Community
Giving to TSRI
Directory
Library
Contact
Site Map & Search
TSRI Home

Scientific Report 2005


Molecular Biology




Ring Assemblies Mediating ATP-Dependent Protein Folding and Unfolding


A.L. Horwich, W.A. Fenton, E. Chapman, E. Koculi

Large ring assemblies function in many cellular contexts as compartments within a compartment, where actions can be carried out on a substrate bound in the central space inside an oligomeric ring by a high local concentration of surrounding active sites. Both protein folding and unfolding are carried out in an ATP-dependent fashion by such assemblies. We are studying the essential double-ring components, chaperonins, that assist protein folding to the native state. We are focusing on the bacterial chaperonin GroEL and more recently have been examining an opposite number, an “unfoldase,” the bacterial heat-shock protein 100 ring assembly known as ClpA. In the past year, we focused on polypeptide binding and ATP-mediated action by both machines, showing quite different mechanisms.

GroEL-Mediated Folding

We are investigating polypeptide binding by an open ring of GroEL that is mediated through contacts between the exposed hydrophobic surface of nonnative polypeptide and a hydrophobic lining of the open ring. This step is one that potentially mediates unfolding of kinetically trapped states. In collaboration with K. Wüthrich, Department of Molecular Biology, using solution nuclear magnetic resonance and transverse relaxation optimized spectroscopy, we examined the structure of isotope-labeled human dihydrofolate reductase bound to GroEL. The resonances detected indicate that the reductase does not occupy a stable tertiary structure while bound to an open GroEL ring and also suggest that the enzyme is undergoing conformational exchange. This unfolded state was, however, productive; upon addition of ATP and the cochaperonin GroES, a nativelike pattern of resonances was recovered.

The binding of ATP and GroES triggers productive GroEL-GroES–mediated folding in the encapsulated now-hydrophilic cavity of the GroES-bound ring (Fig. 1). By contrast, addition of ADP and GroES does not trigger folding. Surprisingly, however, x-ray and solution electron cryomicroscopy structures of GroEL-GroES-ADP and GroEL–GroES–ADP–aluminum fluoride, which is a folding-active state, are isomorphous. We noted that these structure determinations were all carried out in the absence of substrate polypeptide and that a bound substrate potentially represents a load on the ring to which it is bound, resisting nucleotide/GroES-driven elevation and twist of the apical domain that are associated with ejection of a bound polypeptide off the cavity wall into the GroES-encapsulated cavity where productive folding occurs. Thus, the γ-phosphate of ATP might be critical to exerting a power stroke of apical movement. Consistent with such an idea, we found that addition of aluminum fluoride to a GroEL-GroES-ADP-polypeptide complex triggered productive folding. Further, we found that a substantial amount of free energy was released upon binding of aluminum fluoride to GroEL-GroES-ADP.

To directly monitor apical movement, we used fluorescence resonance energy transfer between a fluorophore placed on the stable equatorial base of a subunit and a fluorophore placed in the apical domain (at a position that moves ~30 Å during the transition of a ring from unbound to GroES bound). Indeed, when no substrate was present, the apical domains opened rapidly (<1 second) upon addition of either ADP-GroES or ATP-GroES. In contrast, in the presence of bound polypeptide, only ATP-GroES could promote such rapid opening; ADP-GroES was either unable to drive opening at all or required a longer time (>40 seconds). Additional studies with fluorophores placed on GroEL and GroES indicated that GroES can associate rapidly with GroEL-polypeptide complexes in ADP, evidently forming a collision complex, but subsequent apical movement is impaired. We are using electron microscopy to examine the putative collision state, because it most likely is a state that is transiently populated in the physiologic nucleotide ATP.

ClpA-Mediated Unfolding

ClpA recognizes terminal peptide tags of proteins that are concordantly unfolded and translocated through its central channel. The polypeptide is generally directly translocated into a double-ring proteasome-like protease, ClpP, where it is degraded. During this past year, we used chemical cross-linkers placed on tag elements to identify channel-facing structures of ClpA that bind the tags and then did mutational analysis of the identified regions. For example, the C-terminal 11-residue ssrA peptide, which is added to proteins stalled at the ribosome to recruit these chains to ClpA, binds to 3 loops in the central channel of ClpA, 2 at the level of the proximal D1 ATPase domain and 1 at the level of the distal D2 ATPase (Fig. 1).

Fig. 1. Protein folding and unfolding by chaperone ring assemblies. In protein folding mediated by the chaperonin GroEL (left), the energy of binding ATP and the cochaperonin GroES are used to produce rigid body movements of a GroEL ring that eject a bound nonnative substrate polypeptide into a GroES-encapsulated central cavity, switched from hydrophobic (shaded) to hydrophilic wall character, where productive folding proceeds. The free energy provided by a set of hydrogen bonds formed between the γ-phosphate of ATP and the nucleotide pocket is critical to producing a power stroke of apical domain movement that can eject the substrate polypeptide into the folding chamber. In contrast, in ClpA-mediated unfolding (right), this chaperone seems to use ATP hydrolysis by its D2 ATPase domain to drive a forceful distalward movement of a loop facing its central channel, exerting mechanical force on a bound protein that is proposed to exert an unfolding action.

Interestingly, a mutation at the point of insertion of the D2 loop into the channel wall allows substrate binding but blocks unfolding/
translocation, suggesting that this loop, connected to the more active D2 ATPase of ClpA, is a translocator that pulls on bound polypeptide in association with ATP hydrolysis, exerting a mechanical force that mediates unfolding. Consistently, x-ray studies of different nucleotide states have shown that 2 other such ring components that act on nucleic acids, the phi12 packaging motor and simian virus 40 T antigen, undergo such movements of channel-facing loops.

Publications

Hinnerwisch, J., Fenton, W.A., Furtak, K., Farr, G.W., Horwich, A.L. Loops in the central channel of ClpA chaperone mediate protein binding, unfolding, and translocation. Cell 121:1029, 2005.

Horst, R., Bertelsen, E.B., Fiaux, J., Wider, G., Horwich, A.L., Wüthrich, K. Direct NMR observation of a substrate protein bound to the chaperonin GroEL. Proc. Natl. Acad. Sci. U. S. A. 102:000, 2005.

Motojima, F., Chaudhry, C., Fenton, W.A., Farr, G.W., Horwich, A.L. Substrate polypeptide presents a load on the apical domains of the chaperonin GroEL. Proc. Natl. Acad. Sci. U. S. A. 101:15005, 2004.

 

Arthur Horwich, M.D.
Visiting Investigator



Faculty