br Materials and methods br Results br
Materials and methods
Discussion The current model of the pathological mechanism underlying AGel amyloidosis is mostly based on studies performed on the D187N-mutated protein and experiments conducted in vitro and in vivo on transgenic animals expressing this variant [15,16,48]. Results obtained with D187N likely also apply to the later discovered Danish variant (D187Y), although some differences were reported . On the contrary, the mechanism underlying the recently discovered renal disease associated with the N184K and G167R mutations have not been fully elucidated. Interestingly, while the crystallographic structures of isolated G2 domains carrying the N184K or G167R mutation are already available [30,31], until now the D187N/Y structural characterization had not been elucidated. We here demonstrated that crystallization of the D187N G2 domain was possible only in complex with a previously developed nanobody, called Nb11 . Nb11 tightly binds to gelsolin and protects the D187N variant from aberrant furin proteolysis. In the analysis of the crystal structure, we must consider that the binding of Nb11 somehow biases our model. Indeed, we are observing a proteolysis-resistant species that lost, to some extent, the structural determinants of its proteotoxicity. To elucidate the mechanism of protection exerted by Nb11 it is necessary to understand the pathological mechanism of D187N mutation (and vice versa). Clearly, Nb11 acts as a protective chaperone; however, the molecular mechanism behind such function was as yet unclear, mainly because Nb11 binds G2 in a position distant from the furin cleavage site (Fig. 2). Regarding the D187N mutation, a large body of literature is already available, and these studies converged to a general agreement, i.e. that the D187N substitution disrupts calcium binding in G2 and the thermodynamic stability of the mutant is decreased to levels similar to those of the WT protein deprived of calcium [, , ,49]. Ultimately the mutation leads to the exposure of an otherwise buried sequence, which is aberrantly cleaved by furin. However, the correlation between calcium binding impairment and susceptibility to proteolysis has been the object of discussion. One Rostafuroxin is that the mutation somehow induces a conformational change of the native state that leads to the exposure of the furin site. Another possibility is that the loss of coordinated calcium increases the population of (partially) unfolded protein, which is generally prone to proteolysis [, , ,49]. The partial disorder hypothesis is consistent with the crystallographic structures, the thermodynamic data as well as with the MD results, which in the absence of the coordinated cation show a fast (on a timescale of tens of ns) opening of the C-terminal stretch. Calcium-mediated disorder-to-order induction has been shown, e.g. in sortase [50,51], adenylate cyclase toxin , and possibly several others [53,54]. Interestingly, Kazmirsky and coworkers came to a similar conclusion through a detailed NMR analysis of the D187N variant . In the study, they suggested that the C-terminal tail of the pathogenic variant might be less structured compared to the WT. It might be tempting to explain the increased susceptibility to proteolysis solely based on the flexibility of the C-terminus, which indeed interacts with the hinge loop and exerts some steric protection. However, even the crystallographic structures of the D187NG2:Nb11 complex reveal a destabilized tail, and the simulations show a similar dynamic behavior of the stretch irrespective of Nb11 binding. Destabilization of the C-terminus might likely be a necessary condition, but it seems not to be sufficient for the efficient proteolytic cleavage. In conclusion, we believe that the destabilization of the C-terminal tail is the direct consequence of the loss of the Ca2+ ion induced by D187N mutation. Such increased flexibility is the first event that triggers the sequential opening of the gelsolin fold, which results in the exposure of residues of the hydrophobic core. The process is likely reversible and this partially unfolded state of the protein is susceptible to furin proteolysis. In the complex with Nb11, the binding interface, including α0 and α2-loop-β5, becomes more rigidly anchored to the proper WT conformation. The latter Nb11-stabilized region lies at the base of the disorder-prone C terminus, forming an elbow of sorts. We speculate that Nb11 stabilization reverses, at least in part, the entropy gain due to the abolished coordination of Ca2+.