Structural Probing of Hsp26 Activation and Client Binding by Quantitative Cross-Linking Mass Spectrometry

Abstract

Small heat-shock proteins (sHSPs) are important members of the cellular stress response in all species. Their best-described function is the binding of early unfolding states and the resulting prevention of protein aggregation. Many sHSPs exist as a polydisperse composition of oligomers, which undergoes changes in subunit composition, folding status, and relative distribution upon heat activation. To date, only an incomplete picture of the mechanism of sHSP activation exists; in particular, the molecular basis of how sHSPs bind client proteins and mediate client specificity is not fully understood. In this study, we have applied cross-linking mass spectrometry (XL-MS) to obtain detailed structural information on sHSP activation and client binding for yeast Hsp26. Our cross-linking data reveals the middle domain of Hsp26 as a client-independent interface in multiple Hsp26::client complexes and indicates that client specificity is likely mediated via additional binding sites within its αcrystallin domain and C-terminal extension. Our quantitative XL-MS data underpins the middle domain as the main driver of heat-induced activation and client binding but shows that global rearrangements spanning all domains of Hsp26 take place simultaneously. We also investigated a Hsp26::client complex in the presence of Ssa1 (Hsp70) and Ydj1(Hsp40) at the initial stage of refolding and observe that the interaction between refolding chaperones is altered by the presence of a client protein, pointing to a mechanism where the interaction of Ydj1 with the HSP::client complex initiates the assembly of the active refolding machinery.

Kinetic principles of chemical cross-link formation for protein-protein interactions

“…A change in reaction rate for one species will therefore affect the concentrations of all mono-and cross-links competing for that specific lysine, and this propagates to a smaller extent to second neighbors, in a network-like fashion. This, on the other hand, necessitates that distance is not the sole determiner for cross-link quantity; a notion that up to now is widely accepted in field, as it is assumed that a distance change, for example by a conformational change that brings two cross-linked residues in closer contact, will directly result in a change on cross-link abundances, i.e., an increase of the respective cross-link [5][6][7][8][9][10][11][12][13][14][15][16][17] . The simulations of our kinetic model show that if the cross-link reaction was indeed mainly controlled by the physical distance between two linkable residues, one would expect a strong correlation between the final concentrations in our simulation and the kinetic distance.…”

“…However, such an interdependence would imply that the distance between two cross-linked lysine residues cannot be the sole factor that determines relative crosslink yields, as a change in lysine reactivity and mono-link concentration could affect cross-link quantities, directly or indirectly. This is an important point as it is often assumed that a distance change, for example by a conformational change that brings two cross-linked residues in closer contact, will directly result in a change of cross-link abundances, i.e., an increase of the respective cross-link [5][6][7][8][9][10][11][12][13][14][15][16][17] . Our model enables us to directly test this hypothesis, as one would expect a strong correlation between the final quantity (i.e.…”

“…The model can also be used to study the cross-link formation based on the structure of any protein complex, thereby enabling realistic predictions of cross-link yields for a diverse set of systems. This is an important point as it often assumed that a distance change, for example by a conformational change that brings two cross-linked residues in closer proximity, will directly result in a change on cross-link abundances, i.e., an increase of the respective cross-link yield [5][6][7][8][9][10][11][12][13][14][15][16][17] . Our model now puts us into a position to directly test this hypothesis.…”