Characterization of Desmin Disease Mutants and their Association with alphaB-Crystallin in Desminopathy

Abstract

Mutations in intermediate filaments (IFs) and associated proteins have been shown to cause a number of diseases in humans, ranging from blistering skin diseases to premature aging, as well as from cataract to cardiomyopathies. Desminopathy, a disease caused by dysfunctional mutations in type III muscle-specific IF protein desmin, constitutes a distinct sub-group of myofibrillar myopathies (MFM) manifesting as skeletal and / or cardiac myopathy. Mutations in the chaperone aB-crystallin, that supposedly maintains cytoskeletal integrity, have also been identified to cause MFM. Intracytoplasmic aggregates in desminopathy uniformly comprise aberrantly folded desmin that, among other proteins, recruits aB-crystallin. Currently, the molecular basis of sequential events that lead to such aggregates in the myocyte of patients harboring desmin mutations is not well understood. Thus, to unravel the molecular basis of desminopathy, we have investigated the interdependence between filament alterations arising from desmin mutations and their functional consequences in terms of interaction with the small heat shock protein aB-crystallin. We have systematically characterized various mutants spanning the non-a-helical aminoterminal “head”, central a-helical “rod” and non-a-helical carboxy-terminal “tail” domain of desmin. We show by in vitro characterization of the five head mutants, that two mutant variants residing in the conserved nonapeptide motif “SSYRRTFGG” of desmin - DesS13F and DesR16C - interfere with assembly by forming filamentous aggregates. Consistent with in vitro data, both mutants fail to generate a bona fide filament system in cells lacking a type III IF cytoskeleton. In cells expressing vimentin or desmin, both mutants fail to integrate into the endogenous filament network and severely affect its cellular localization. The novel desminopathy-causing mutant DesL377del22fs apparently interacts with wild-type desmin at dimer, tetramer and higher level of filament organization in vitro, but leads to a disruption of the IF cytoskeleton in cells. This mutant is not detectable in the myotubes of a heterozygous carrier even upon proteasome inhibition. Two – DesR454W and DesK449T - out of the six tail mutants form abnormal filaments during in vitro assembly and correspondingly generate aberrant filaments in cells devoid of type III IF protein cytoskeleton. The desmin fragment Des(ESA)delC244, resembling almost “first-half” of a desmin molecule, has deleterious effects on filament assembly in vitro as well as in transfected cells. It exhibits nucleoplasmic aggregates in two of the four investigated cell lines. With respect to characterizing the association of desmin disease mutants with aB-crystallin, data from yeast two-hybrid analyses, electron microscopy (EM), cosedimentation assay and viscometry distinctly suggest that the tail domain of desmin is pivotal in modulating its binding to aB-crystallin. We show that aB-crystallin binds to wild-type desmin filament under optimized buffer condition, but its binding to C-terminal deletion variants is either diminished or abolished. We speculate that this occurs due to differences in hydrophobic surface properties and exposed residues of wild-type desmin and its deletion variants. DesdelRDG is devoid of the conserved tripeptide motif “RDG”, yet it binds to aB-crystallin with similar strength as desmin wild-type. Thus, we propose that the prerequisite for binding of aB-crystallin to desmin is the 3-dimensional desmin protein conformation, which can be altered due to a mutation, and not the linear amino acid sequence involving conserved motifs per se. The two tail mutants – DesI451M and DesR454W - reveal weaker and stronger binding, respectively, to aB-crystallin as compared to wild-type protein. With respect to kinetics of binding, unlike desmin wild-type, DesR454W binds to aB-crystallin at all stages of assembly, and this probably results from its “open” filament structure both alone and in an equimolar mixture with wild-type desmin. Data from R454W and wild-type desmin transfection in 3T3 cells corroborate the in vitro data, showing that DesR454W binds around 50% more cytosolic aB-crystallin than desmin wild-type. Hence, our data suggest that mutations in desmin cause toxic gain-of-function, whereby the desmin mutants show enhanced binding to aB-crystallin. A plausible explanation for aggregate formation in desminopathy could be such modified protein-protein interactions. In summary, our data demonstrate the impact of desmin mutations not only on its structural property, but also on its molecular interaction with aB-crystallin. This adds to our understanding of the molecular basis of desminopathy as we show for the first time that subtle alterations in the nanoarchitecture of desmin filament are sufficient to induce aberrant interaction with an associated protein aB-crystallin. Such a modification might eventually contribute to the pathogenesis of desminopathy.