The interactions of molecular chaperones with client proteins: why are they so weak?

Abstract

The major classes of molecular chaperones have highly variable sequences, sizes, and shapes, yet they all bind to unfolded proteins, limit their aggregation, and assist in their folding. Despite the central importance of this process to protein homeostasis, it has not been clear exactly how chaperones guide this process or whether the diverse families of chaperones use similar mechanisms. For the first time, recent advances in NMR spectroscopy have enabled detailed studies of how unfolded, "client" proteins interact with both ATP-dependent and ATP-independent classes of chaperones. Here, we review examples from four distinct chaperones, Spy, Trigger Factor, DnaK, and HscA-HscB, highlighting the similarities and differences between their mechanisms. One striking similarity is that the chaperones all bind weakly to their clients, such that the chaperone-client interactions are readily outcompeted by stronger, intra- and intermolecular contacts in the folded state. Thus, the relatively weak affinity of these interactions seems to provide directionality to the folding process. However, there are also key differences, especially in the details of how the chaperones release clients and how ATP cycling impacts that process. For example, Spy releases clients in a largely folded state, while clients seem to be unfolded upon release from Trigger Factor or DnaK. Together, these studies are beginning to uncover the similarities and differences in how chaperones use weak interactions to guide protein folding.

Keywords: chaperone; nuclear magnetic resonance (NMR); protein aggregation; protein folding; protein–protein interactions (PPIs).

Figure 1

Structurally diverse molecular chaperones direct clients to the native state.A, comparison of four molecular chaperones that differ in size, sequence, and shape. The molecular mass and PDB code for each chaperone are shown. Despite these different architectures, the chaperones all promote client folding and limit aggregation. B and C, core functions of the molecular chaperones are to suppress aggregation and promote client folding by directly binding to the unfolded state(s). While some chaperones (DnaK, green) require ATP to perform these functions (B), others (Spy, blue, and Trigger Factor, orange) work independently of nucleotide turnover (C).

Figure 2

Spy binds clients as they fold.A, Spy binds to unfolded clients, which may then explore diverse conformations. Upon arrival at the native state, burial of hydrophobic residues decreases Spy's affinity for the client and causes client release. B, When Spy's affinity for unfolded clients is increased by point mutations, this leads to unfolding of the native state and decreases the efficiency of folding.

Figure 3

A hierarchy of protein–protein interactions directs trigger factor (TF) to chaperone and release nascent polypeptides. In the cytosol, TF exists in equilibrium between monomeric and dimeric states. Monomeric TF can interact with ribosomes, but its tighter affinity for translating ribosomes allows TF to selectively chaperone nascent polypeptides. As translation continues, TF can remain bound to the polypeptide and eventually dissociate to repeat the cycle.

Figure 4

DnaK directs clients to the native state by binding to the unfolded state and off-pathway misfolded intermediates.A, crystal structure of DnaK SBD bound to model NRLLLTG peptide. B, DnaK employs conformational selection to bind to the unfolded state, but not the native state, driving flux in the pathway toward the native state. C, DnaK can bind to off-pathway intermediates, preventing long-range interactions and increasing the conversion to on-pathway intermediates. DnaK is also able to intervene at multiple other points in the folding landscape, such as at the unfolded state and multiple other nonnative states. Also, there is no “directionality” implied in the action of DnaK on the topological landscape (e.g., DnaK does not know: which direction leads to the native state).

Figure 5

Stepwise protein–protein interactions enable Fe-S cluster biogenesis and installation to promote client release.A, the cycle of Fe-S biogenesis and transfer is depicted. 1, IscU exists in equilibrium between a structured state (S-state) and a disordered state (D-state). 2, the cysteine desulfurase IscS interacts with the D state of IscU. 3, assembly of the Fe-S cluster stabilizes the S state of IscU. 4, holo-IscU is transferred from IscS to HscB. 5, holo-IscU forms a ternary complex with HscB and HscA. 6, Fe-S cluster is transferred to the client protein, HscB is released, and HscA binds to the D state of IscU. 7, nucleotide exchange of HscA causes release of IscU. B, loading of the apo-client with the Fe-S cluster drives client release and folding.