Putting handcuffs on the chaperonin GroEL.

Oligomeric, ring-shaped nano-machines that are fueled by ATP are ubiquitous in all three kingdoms of life and are involved in a wide range of processes that include, for example, protein folding, protein degradation, DNA and RNA remodeling, and protein insertion into membranes (for review, see ref. 1). These assemblies are true machines because they carry out work by undergoing movements that are coordinated in time and space and are driven by energy consumption (i.e., ATP binding and hydrolysis). One fundamental question regarding such machines relates to the coupling between the order of their ATPpromoted conformational changes (Fig. 1) and their biological function. These conformational changes can take place in a concerted fashion according to the MonodWyman-Changeux (MWC) model (2), as proposed and recently demonstrated in the case of the chaperonin GroEL (3, 4). Alternatively, the changes can occur in a domino-like sequential fashion according to the Koshland-Némethy-Filmer model (5) or in a probabilistic manner, as suggested in the case of the proteolytic machine ClpP (6). A second fundamental question regarding such machines concerns the mechanisms by which ATP consumption is converted into work. Both questions are elegantly addressed, in the case of the chaperonin GroEL, in the report by Corsepius and Lorimer (7) published in PNAS. GroEL consists of two back-to-back stacked heptameric rings, with a cavity at each end in which protein folding can take place under confining conditions. GroEL assists protein folding by cycling between protein substrate acceptor and release states upon ATP binding and hydrolysis. ATP binding to GroEL occurs with positive cooperativity within rings and negative cooperativity between rings, both with respect to ATP. A nested allosteric model that accounts for these observations was put forward (3) in which each ring of GroEL switches in a concerted MWC fashion between a T state, with low affinity for ATP and high affinity for protein substrates, and an R state, with high affinity for ATP and low affinity for protein substrates. In the presence of ATP, the GroEL double-ring can, therefore, exist in three states: TT, TR, and RR. Because of the negative interring cooperativity, the TR→RR transition is less favored than the TT→TR transition, thus ensuring that the two rings can operate out of phase with each other. Complexes of protein substrates with GroEL are heterogeneous because bound nonfolded substrates can exist in multiple conformations that interact with different configurations of GroEL subunits (8). It has, therefore, been difficult to quantitate the effects of substrate binding on the allosteric properties of GroEL. Corsepius and Lorimer (7) have neatly circumvented this heterogeneity problem by labeling GroEL with tetramethylrhodamine (TMR) at position 242 in the apical domain near the protein substrate-binding site. This position was chosen so that TMR molecules attached to adjacent subunits can form noncovalently “stacked” dimers in the T state but not in the R state. Addition of ATP, which induces the transition to the R state, therefore leads to dissociation of the TMR dimers. The formation of dimers in the T state tethers pairs of GroEL subunits to each other in a manner that mimics intersubunit tethering by nonfolded proteins. The TMR dimers can, therefore, serve as surrogates for nonfolded protein substrates to quantitate the impact of protein substrate binding on the allosteric properties of GroEL. The TMR labeling can also be used to monitor the T→R allosteric transitions because the TMR dimers formed in the T state are weakly fluorescent compared with TMRmonomers that exist in the R state. It has been shown that stabilization of the T state by protein substrate binding is accompanied by stimulation of ATP hydrolysis by GroEL (9), but it has remained unclear whether the stimulation is because of intersubunit tethering or just due to the binding. Corsepius and Lorimer (7) show that ATP hydrolysis is stimulated by TMR dimer formation but not by labeling with fluorescein5-maleimide, a compound that is structurally similar to TMR that does not form dimers. The authors’ results, therefore, indicate that it is the intersubunit tethering that is responsible for the stimulation of ATPase activity by protein substrates. The mechanism by which this occurs is not known. It has been shown, however, that protein substrates enhance the steady-state ATPase activity of GroEL by increasing the rate of ADP release (10, 11). Hence, it is likely that tethering shifts the equilibrium after hydrolysis in favor of the T state, which has lower affinity for nucleotides, thereby increasing the rate of ADP release and stimulating ATPase activity. The TMR labeling strategy is also used to determine the impact of tethering on the relative stabilities of the T and R states (7). Clever analysis that assumes stochastic labeling is used to extract the distribution of tethered states (i.e., how many rings have 0, 1, 2, or 3 crosslinks) from the average number of TMR molecules per ring that is determined experimentally. This distribution is combined with a modified nested allosteric model that includes the effect of tethering on the stability of the T state. The resulting equation is used to fit the change in fluorescence, of a sample of GroEL with an average number of TMR molecules per ring, as a function of ATP Fig. 1. Scheme showing various mechanisms of allosteric switching for a seven-membered ring. Each subunit can exist in a low(green) or high affinity (cyan) state for the ligand (e.g., ATP). (A) Concerted conformational changes. (B) Sequential conformational changes. (C ) Probabilistic conformational changes.